UNIX"

TEXT PROCESSING
 HOWARD u!SAMs &COMPANY
HAYDEN BOOKS

Related Titles
Advanced C Primer++ UNIX@ System V Bible
Stephen Prata, The Waite Group Stephen Prata and Donald Martin,
The Waite Group
Discovering MS-DOS@
Kate O’Day, The Waite Group UNIX@ Communications
Bryan Costales, The Waite Group
Microsoft? C Programming
for the IBM@ C with Excellence:
Robert Lajore, The Waite Group Programming Proverbs
Henry F. Ledgard with John Tauer
MS-DOS@ Bible
Steven Simrin, The Waite Group C Programmer’s Guide
to Serial Communications
MS-DOS@ Developer’s Guide Joe Campbell
John Angermeyer and Kevin Jaeger,
Hayden Books
The Waite Group

Tricks of the MS-DOS@ Masters UNIX System Library

John Angermeyer, Rich Fahringer,
Kevin Jaeger, and Dan Shajer, The Waite Group UNIX@ Shell Programming
Stephen G. Kochan and Patrick H. Wood
Inside XENIX@
Christopher L. Morgan, The Waite Group UNIX@ System Security
Patrick H. Wood and Stephen G. Kochan
UNIX@ Primer Plus
Mitchell Waite. Donald Martin, UNIX@ System Administration
and Stephen Prata, The Waite Group David Fieldler and Bruce H. Hunter

UNIX@ System V Primer, Exploring the UNIX@ System

Revised Edition Stephen G. Kochan and Patrick H. Wood
Mitchell Waite, Donald Martin,
and Stephen Prata, The Waite Group Programming in C
Stephen G. Kochan
Advanced U N I P -
A Programmer’s Guide Topics in C Programming
Stephen Prata. The Waite Group Stephen G. Kochan and Patrick H. Wood

UNIX@ Shell Programming Language

Rod Manis and Marc Meyer

For the retailer nearest you, or io order directly from the publisher,
call 800428-SAMs. In Indiana, Alaska, and Hawaii call 31 7-298-5699.
 TEXT IPROCESSINC

DALE DOUGHERTY AND TIM O'REILLV

and the staff of O'Reilly & Associates, Inc.

CONSULTING EDITORS:

Stephen G. Kochan and Patrick H. Wood

HAYDEN BOOKS
A Division of Howard W Sams G Company
4300 West 62nd Street
Indianapolis, Indiana 46268 USA
Copyright 0 1987 O’Reilly & Associates, Inc.
FIRST EDITION
SECOND PRINTING - 1988
All rights reserved. No part of this book shall be reproduced, stored in a retrieval system, or
transmitted by any means, electronic, mechanical, photocopying, recording, or otherwise, without
written permission from the publisher. No patent liability is assumed with respect to the use of
the information contained herein. While every precaution has been taken in the preparation of
this book, the publisher assumes no responsibility for errors or omissions. Neither is any liability
assumed for damages resulting from the use of the information contained herein.
International Standard Book Number: 0-672-4629 1-5
Library of Congress Catalog Card Number: 87-60537
Acquisitions Editor: Therese Zak
Editor: Susan Pink Bussiere
Cover: Visual Graphic Services, Indianapolis
Design by Jerry Bates
Illustration by Patrick Sarles
Typesetting: O’Reilly & Associates, Inc.
Printed in the United States of America

Trademark Acknowledgements
All terms mentioned in this book that are known to be trademarks or service marks are listed
below. Howard W. Sams & Co. cannot attest to the accuracy of this information. Use of a term
in this book should not be regarded as affecting the validity of any trademark or service mark.
Apple is a registered trademark and Apple Laserwriter is a trademark of Apple Computer, Inc.
devps is a trademark of Pipeline Associates, Inc.
Merge/286 and Merge/386 are trademarks of Locus Computing Corp.
DDL is a trademark of Imagen Corp.
Helvetica and Times Roman are registered trademarks of Allied Corp.
IBM is a registered trademark of International Business Machines Corp.
Interpress is a trademark of Xerox Corp.
LaserJet is a trademark of Hewlett-Packard Corp.
Laserwriter is a trademark of Apple Computer, Inc.
Linotronic is a trademark of Allied Corp.
Macintosh is a trademark licensed to Apple Computer, Inc.
Microsoft is a registered trademark of Microsoft Corp.
MKS Toolkit is a trademark of Mortice Kern Systems, Inc.
Multimate is a trademark of Multimate International Corp.
Nutshell Handbook is a trademark of O’Reilly & Associates, Inc.
PC-Interface is a trademark of Locus Computing Corp.
PostScript is a trademark of Adobe Systems, Incorporated.
PageMaker is a registered trademark of Aldus Corporation.
SoftQuad Publishing Software and SQtroff are trademarks of SoftQuad Inc.
WordStar is a registered trademark of MicroPro International Corp.
UNIX is a registered trademark of AT&T.
VP/ix is a trademark of Interactive Systems Corp. and Phoenix Technologies, Ltd.
C 0 N T E N T S

Preface xi

1 From Typewriters to Word Processors 1

A Workspace . 2
Tools for Editing . 4
Document Formatting . . 6
Printing - 8
Other UNIX Text-Processing Tools . . 10

2 UNIX Fundamentals 12

The UNIX Shell . . 12

Output Redirection . 14
Special Characters . 19
Environment Variables . . 20
Pipes and Filters . . 21
Shell Scripts . 23

3 Learningvi 24

Session 1: Basic Commands . . 25

Opening a File . . 25
Moving the Cursor . 28
Simple Edits . 32
Session 2: Moving around in a Hurry . 41
Movement by Screens . . 42
Movement by Text Blocks . 44
Movement by Searches . 45
Movement by Line Numbers . . 47
Session 3: Beyond the Basics . 48
Command-Line Options . 49
Customizing vi . . 50
 Edits and Movement . 53
More Ways to Insert Text . 54
Using Buffers . 54
Marking Your Place . 57
Other Advanced Edits . . 57

4 nroff and troff 58

What the Formatter Does . 59

Using n r o f f . 63
Using t r o f f . 64
The Markup Language . . 67
Turning Filling On and Off . 69
Controlling Justification . 71
Hyphenation . 73
Page Layout . 75
Page Transitions . . 86
Changing Fonts . . 92
A First Look at Macros . 99

5 The ms Macros 104

Formatting a Text File with m8 . 105

Page Layout . 106
Paragraphs . 106
Changing Font and Point Size . 114
Displays . . 117
Headings . . 120
Cover Sheet Macros . 122
Miscellaneous Features . . 123
Page Headers and Footers . 126
Problems on the First Page . 127
Extensions to ms . 127

6 ThemmMacros 128

Formatting a Text File . . 128

Page Layout . 132
Justification . 137
Word Hyphenation . 137
Displays . . 138
Changing Font and Point Size . 141
More About Displays . . 145
Forcing a Page Break . . 150
Formatting Lists . . 150
 Headings . . 162
Table o f Contents . 168
Footnotes and References . 170
Extensions to nmr . 173

7 Advanced Editing 177

The ex Editor . 178

Using ex Commands in v i . 180
Write Locally, Edit Globally . 180
Pattern Matching . 184
Writing and Quitting Files . 190
Reading In a File . 192
Executing UNIX Commands . 192
Editing Multiple Files . . 195
Word Abbreviation . 198
Saving Commands with map . 198

8 Formatting with t b l 203

Using t b l .204
Specifying Tables . 205
A Simple Table Example . 206
Laying Out a Table . 207
Describing
- Column Formats . 209
Changing the Format within a Table . 219
Putting Text Blocks in a Column . 221
Breaking Up Long Tables . 224
Putting Titles on Tables . 225
A tbl Checklist . 226
Some Complex Tables . . 227

9 Typesetting Equations with eqn 232

A Simple eqn Example . 233

Using eqn . 233
Specifying Equations . . 234
Spaces in Equations . 236
Using Braces for Grouping . 238
Special Character Names . 239
Special Symbols . . 241
Other Positional Notation * 244
Diacritical Marks . 245
Defining Terms . . 247
Quoted Text . 248
 Fine-Tuning the Document . 248
Keywords and Precedence . 250
Problem Checklist . 251

10 Drawing Pictures 253

The pic Preprocessor . . 254

From Describing to Programming Drawings . 281
pic Enhancements . 291
11 A Miscellany of UNIX Commands 293

Managing Your Files . . 293

Viewing the Contents of a File . 298
Searching for Information in a File . . 301
Proofing Documents . 304
Comparing Versions of the Same Document . 312
Manipulating Data . 322
Cleaning Up and Backing Up . . 336
Compressing Files . 338
Communications . 339
Scripts of UNIX Sessions . 341

12 Let the Computer Do the Dirty Work 342

Shell Programming . 343

ex Scripts . 354
Stream Editing (sed) . . 360
A Proofreading Tool You Can Build . 380

13 The a w k Programming Language 387

Invoking a w k . 388
Records and Fields . 389
Testing Fields , 390
Passing Parameters from a Shell Script . 390
Changing the Field Separator . . 391
System Variables . 392
Looping . . 393
a w k Applications . 400
Testing Programs . 410
14 Writing nrof f and t r o f f Macros 412

Comments . 412
Defining Macros . 413
Macro Names . 414
Macro Arguments . 416
Nested Macro Definitions . 418
Conditional Execution . . 418
Interrupted Lines . 423
Number Registers . 424
Defining Strings . . 429
Diversions . 431
Environment Switching . 433
Redefining Control and Escape Characters . 435
Debugging Your Macros . 436
Error Handling . . 439
Macro Style . 441

15 Figures and Special Effects 443

Formatter Escape Sequences . .443

Local Vertical Motions . * 445

Local Horizontal Motions . 447

Absolute Motions . 448
Line Drawing . a 9
Talking Directly to the Printer . 460
Marking a Vertical Position . . 461
Overstriking Words or Characters . 462
Tabs, Leaders, and Fields . 467
Constant Spacing . 471
Pseudo-Fonts . 473
Character Output Translations . 473
Output Line Numbering . 475
Change Bars . 476
Form Letters . 477
Reading in Other Files or Program Output . 479

16 What’s in a Macro Package? 48 1

Just What Is a Macro Package, Revisited . . 481

Building a Consistent Framework . 484
Page Transitions . . 489
Page Transitions in ms . . 491
Some Extensions to the Basic Package . 495
Other Exercises in Page Transition . .500
17 An Extended ms Macro Package 509

Creating a Custom Macro Package . . 510

Structured Technical Documents . 512
Figure and Table Headings . 523
Lists, Lists, and More Lists . 525
Source Code and Other Examples . 528
Notes, Cautions, and Warnings . 530
Table of Contents, Index. and Other End Lists . 532
18 Putting It All Together 542

Saving an External Table of Contents .544

Index Processing . 548
Let make Remember the Details . 562
Where to Go from Here . 567
A Editor Command Summary 569

B Formatter Command Summary 593

C Shell Command Summary 628

D Format of troff Width Tables 635

E Comparing rmn and ms 640

F The format Macros 643

G Selected Readings 646

Index 647
 Preface

Many people think of computers primarily as “number crunchers,” and think of word
processors as generating form letters and boilerplate proposals. That computers can be
used productively by writers, not just research scientists, accountants, and secretaries, is
not so widely recognized. Today, writers not only work with words, they work with
computers and the software programs, printers, and terminals that are part of a computer
system.
The computer has not simply replaced a typewriter; it has become a system for
integrating many other technologies. As these technologies are made available at a rea-
sonable cost, writers may begin to find themselves in new roles as computer program-
mers, systems integrators, data base managers, graphic designers, typesetters, printers,
and archivists.
The writer functioning in these new roles is faced with additional responsibilities.
Obviously, it is one thing to have a tool available and another thing to use it skillfully.
Like a craftsman, the writer must develop a number of specialized skills, gaining con-
trol over the method of production as well as the product. The writer must look for
ways to improve the process by integrating new technologies and designing new tools
in software.
In this book, we want to show how computers can be used effectively in the
preparation of written documents, especially in the process of producing book-length
documents. Surely it is important to learn the tools of the trade, and we will demon-
strate the tools available in the UNIX environment. However, it is also valuable to
examine text processing in terms of problems and solutions: the problems faced by a
writer undertaking a large writing project and the solutions offered by using the
resources and power of a computer system.
In Chapter 1, we begin by outlining the general capabilities of word-processing
systems. We describe in brief the kinds of things that a computer must be able to do
for a writer, regardless of whether that writer is working on a UNIX system or on an
IBM PC with a word-processing package such as WordStar or MuItiMate. Then, hav-
ing defined basic word-processing capabilities, we look at how a text-processing system
includes and extends these capabilities and benefits. Last, we introduce the set of text-

= xi .
xii UNIX Text Processing 0

processing tools in the UNIX environment. These tools, used individually or in combi-
nation, provide the basic framework for a text-processing system, one that can be
custom-tailored to supply additional capabilities.
Chapter 2 gives a brief review of UNIX fundamentals. We assume you are
already somewhat acquainted with UNIX, but we included this information to make
sure that you are familiar with basic concepts that we will be relying on later in the
book.
Chapter 3 introduces the v i editor, a basic tool for entering and editing text.
Although many other editors and word-processing programs are available with UNIX,
v i has the advantage that it works, without modification, on almost every UNIX sys-
tem and with almost every type of terminal. If you learn v i , you can be confident that
your text editing skills will be completely transferable when you sit down at someone
else’s terminal or use someone else’s system.
Chapter 4 introduces the nrof f and t r o f f formatting programs. Because
v i is a text editor, not a word-processing program, it does only rudimentary formatting
of the text you enter. You can enter special formatting codes to specify how you want
the document to look, then format the text using either n r o f f or t r o f f . (The
n r o f f formatter is used for formatting documents to the screen or to typewriter-like
printers; t r o f f uses much the same formatting language, but has additional con-
structs that allow it to produce more elaborate effects on typesetters and laser printers.)
In this chapter, we also describe the different types of output devices for printing
your finished documents. With the wider availability of laser printers, you need to
become familiar with many typesetting terms and concepts to get the most out of
t r o f f ’ s capabilities.
The formatting markup language required by n r o f f and t r o f f is quite com-
plex, because it allows detailed control over the placement of every character on the
page, as well as a large number of programming constructs that you can use to define
custom formatting requests or macros. A number of macro packages have been
developed to make the markup language easier to use. These macro packages define
commonly used formatting requests for different types of documents, set up default
values for page layout, and so on.
Although someone working with the macro packages does not need to know
about the underlying requests in the formatting language used by n r o f f and t r o f f ,
we believe that the reader wants to go beyond the basics. As a result, Chapter 4 intro-
duces additional basic requests that the casual user might not need. However, your
understanding of what is going on should be considerably enhanced.
There are two principal macro packages in use today, m s and mm (named for the
command-line options to nro f f and t r o f f used to invoke them). Both macro
packages were available with most UNIX systems; now, however, m s is chiefly avail-
able on UNIX systems derived from Berkeley 4.x BSD, and mm is chiefly available on
UNIX systems derived from AT&T System V. If you are lucky enough to have both
macro packages on your system, you can choose which one you want to learn. Other-
wise, you should read either Chapter 5, The ms Macros, or Chapter 6, The m m Macros,
depending on which version you have available.
 o Preface xiii

Chapter 7 returns to v i to consider its more advanced features. In addition, it

takes a look at how some of these features can support easy entry of formatting codes
used by n r o f f and t r o f f .
Tables and mathematical equations provide special formatting problems. The
low-level n r o f f and t r o f f commands for typesetting a complex table or equation
are extraordinarily complex. However, no one needs to learn or type these commands,
because two preprocessors, t b l and eqn, take a high-level specification of the table
or equation and do the dirty work for you. They produce a “script” of n r o f f or
t r o f f commands that can be piped to the formatter to lay out the table or equations.
The t b l and eqn preprocessors are described in Chapters 8 and 9, respectively.
More recent versions of UNIX (those that include AT&T’s separate Documenter’s
Workbench software) also support a preprocessor called p i c that makes it easier to
create simple line drawings with t r o f f and include them in your text. We talk about
pi c in Chapter 10.
Chapter 1 1 introduces a range of other UNIX text-processing tools-programs for
sorting, comparing, and in various ways examining the contents of text files. This
chapter includes a discussion of the standard UNIX s p e l l program and the Writer’s
Workbench programs s t y l e and d i c t i o n .
This concludes the first part of the book, which covers the tools that the writer
finds at hand in the UNIX environment. This material is not elementary. In places, it
grows quite complex. However, we believe there is a fundamental difference between
learning how to use an existing tool and developing skills that extend a tool’s capabili-
ties to achieve your own goals.
That is the real beauty of the UNIX environment. Nearly all the tools it provides
are extensible, either because they have built-in constructs for self-extension, like
n r o f f and t r o f f ’s macro capability, or because of the wonderful programming
powers of the UNIX command interpreter, the shell.
The second part of the book begins with Chapter 12, on editing scripts. There are
several editors in UNIX that allow you to write and save what essentially amount to
programs for manipulating text. The e x editor can be used from within v i to make
global changes or complex edits. The next step is to use e x on its own; and after you
do that, it is a small step to the even more powerful global editor sed. After you have
mastered these tools, you can build a library of special-purpose editing scripts that
vastly extend your power over the recalcitrant words you have put down on paper and
now wish to change.
Chapter 13 discusses another program-auk-that extends the concept of a text
editor even further than the programs discussed in Chapter 12. The auk program is
really a database programming language that is appropriate for performing certain kinds
of text-processing tasks. In particular, we use it in this book to process output from
t r o f f for indexing.
The next five chapters turn to the details of writing t r o f f macros, and show
how to customize the formatting language to simplify formatting tasks. We start in
Chapter 14 by looking at the basic requests used to build macros, then go on in Chapter
15 to the requests for achieving various types of special effects. In Chapters 16 and 17,
we’ll take a look at the basic structure of a macro package and focus on how to define
the appearance of large documents such as manuals. We’ll show you how to define
xiv 0 UNIX Text Processing 0

different styles of section headings, page headers, footers, and so on. We’ll also talk
about how to generate an automatic table of contents and index-two tasks that take
you beyond t r o f f into the world of shell programming and various UNIX text-
processing utilities.
To complete these tasks, we need to return to the UNIX shell in Chapter 18 and
examine in more detail the ways that it allows you to incorporate the many tools pro-
vided by UNIX into an integrated text-processing environment.
Numerous appendices summarize information that is spread throughout the text,
or that couldn’t be crammed into it.

***

Before we turn to the subject at hand, a few acknowledgements are in order. Though
only two names appear on the cover of this book, it is in fact the work of many hands.
In particular, Grace Todino wrote the chapters on t b l and e q n in their entirety, and
the chapters on v i and ex are based on the O’Reilly & Associates’ Nutshell Hand-
book, Learning the Vi Editor, written by Linda Lamb. Other members of the O’Reilly
& Associates staff-Linda Mui, Valerie Quercia, and Donna Woonteiler-helped tire-
lessly with copyediting, proofreading, illustrations, typesetting, and indexing.
Donna was new to our staff when she took on responsibility for the job of
copyfitting-that final stage in page layout made especially arduous by the many fig-
ures and examples in this book. She and Linda especially spent many long hours get-
ting this book ready for the printer. Linda had the special job of doing the final con-
sistency check on examples, making sure that copyediting changes or typesetting errors
had not compromized the accuracy of the examples.
Special thanks go to Steve Talbott of Masscomp, who first introduced us to the
power of t r o f f and who wrote the first version of the extended m s macros, f o r -
mat shell script, and indexing mechanism described in the second half of this book.
Steve’s help and patience were invaluable during the long road to mastery of the UNIX
text-processing environment.
We’d also like to thank Teri Zak, the acquisitions editor at Hayden Books, for her
vision of the Hayden UNIX series, and this book’s place in it.
In the course of this book’s development, Hayden was acquired by Howard Sams,
where Teri’s role was taken over by Jim Hill. Thanks also to the excellent production
editors at Sams, Wendy Ford, Lou Keglovitz, and especially Susan Pink Bussiere,
whose copyediting was outstanding.
Through it all, we have had the help of Steve Kochan and Pat Wood of Pipeline
Associates, Enc., consulting editors to the Hayden UNIX Series. We are grateful for
their thoughtful and thorough review of this book for technical accuracy. (We must, of
course, make the usual disclaimer: any errors that remain are our own.)
Steve and Pat also provided the macros to typeset the book. Our working drafts
were printed on an HP LaserJet printer, using d i t r o f f and TextWare International’s
t p l u s postprocessor. Final typeset output was prepared with Pipeline Associates’
devps, which was used to convert d i t r o f f output to PostScript, which was used in
turn to drive a Linotronic LlOO typesetter.
 C H A P T E R

From Typewriters to Word Processors

Before we consider the special tools that the UNIX environment provides for text pro-
cessing, we need to think about the underlying changes in the process of writing that are
inevitable when you begin to use a computer.
The most important features of a computer program for writers are the ability to
remember what is typed and the ability to allow incremental changes-no more retyping
from scratch each time a draft is revised. For a writer first encountering word-
processing software, no other features even begin to compare. The crudest command
structure, the most elementary formatting capabilities, will be forgiven because of the
immense labor savings that take place.
Writing is basically an iterative process. It is a rare writer who dashes out a fin-
ished piece; most of us work in circles, returning again and again to the same piece of
prose, adding or deleting words, phrases, and sentences, changing the order of thoughts,
and elaborating a single sentence into pages of text.
A writer working on paper periodically needs to clear the deck-to type a clean
copy, free of elaboration. As the writer reads the new copy, the process of revision
continues, a word here, a sentence there, until the new draft is as obscured by changes
as the first. As Joyce Carol Oates i s said to have remarked: “No book is ever finished.
It is abandoned.”
Word processing first took hold in the office as a tool to help secretaries prepare
perfect letters, memos, and reports. As dedicated word processors were replaced with
low-cost personal computers, writers were quick to see the value of this new tool. In a
civilization obsessed with the written word, it is no accident that WordStar, a word-
processing program, was one of the first best sellers of the personal computer revolu-
tion.
As you learn to write with a word processor, your working style changes.
Because it is so easy to make revisions, it is much more forgivable to think with your
fingers when you write, rather than to carefully outline your thoughts beforehand and
polish each sentence as you create it.
If you do work from an outline, you can enter it first, then write your first draft by
filling in the outline, section by section. If you are writing a structured document such
2 0 UNlX Text Processing 0

as a technical manual, your outline points become the headings in your document; if
you are writing a free-flowing work, they can be subsumed gradually in the text as you
flesh them out. In either case, it i s easy to write in small segments that can be moved
as you reorganize your ideas.
Watching a writer at work on a word processor is very different from watching a
writer at work on a typewriter. A typewriter tends to enforce a linear flow-you must
write a passage and then go back later to revise it. On a word processor, revisions are
constant-you type a sentence, then go back to change the sentence above. Perhaps
you write a few words, change your mind, and back up to take a different tack; or you
decide the paragraph you just wrote would make more sense if you put it ahead of the
one you wrote before, and move it on the spot.
This is not to say that a written work is created on a word processor in a single
smooth flow; in fact, the writer using a word processor tends to create many more drafts
than a compatriot who still uses a pen or typewriter. Instead of three or four drafts, the
writer may produce ten or twenty. There is still a certain editorial distance that comes
only when you read a printed copy. This is especially true when that printed copy is
nicely formatted and letter perfect.
This brings us to the second major benefit of word-processing programs: they
help the writer with simple formatting of a document. For example, a word processor
may automatically insert carriage returns at the end of each line and adjust the space
between words so that all the lines are the same length. Even more importantly, the
text is automatically readjusted when you make changes. There are probably commands
for centering, underlining, and boldfacing text.
The rough formatting of a document can cover a multitude of sins. A s you read
through your scrawled markup of a preliminary typewritten draft, it is easy to lose track
of the overall flow of the document. Not so when you have a clean copy-the flaws of
organization and content stand out vividly against the crisp new sheets of paper.
However, the added capability to print a clean draft after each revision also puts
an added burden on the writer. Where once you had only to worry about content, you
may now find yourself fussing with consistency of margins, headings, boldface, italics,
and all the other formerly superfluous impedimenta that have now become integral to
your task.
As the writer gets increasingly involved in the formatting of a document, it
becomes essential that the tools help revise the document’s appearance as easily as its
content. Given these changes imposed by the evolution from typewriters to word pro-
cessors, let’s take a look at what a word-processing system needs to offer to the writer.

A Workspace
One of the most important capabilities of a word processor is that it provides a space in
which you can create documents. In one sense, the video display screen on your termi-
nal, which echoes the characters you type, is analogous to a sheet of paper. But the
workspace of a word processor i s not so unambiguous as a sheet of paper wound into a
typewriter, that may be added neatly to the stack of completed work when finished, or
tom out and crumpled as a false start. From the computer’s point of view, your
 0 From Typewriters to Word Processors 0 3

workspace is a block of memory, called a hufSeer, that is allocated when you begin a
word-processing session. This buffer is a temporary holding area for storing your work
and is emptied at the end of each session.
To save your work, you have to write the contents of the buffer to a file. A file is
a permanent storage area on a disk (a hard disk or a floppy disk). After you have saved
your work in a file, you can retrieve it for use in another session.
When you begin a session editing a document that exists on file, a copy of the file
is made and its contents are read into the buffer. You actually work on the copy, mak-
ing changes to it, not the original. The file is not changed until you save your changes
during or at the end of your work session. You can also discard changes made to the
buffered copy, keeping the original file intact, or save multiple versions of a document
in separate files.
Particularly when working with larger documents, the management of disk files
can become a major effort. If, like most writers, you save multiple drafts, it is easy to
lose track of which version of a file is the latest.
An ideal text-processing environment for serious writers should provide tools for
saving and managing multiple drafts on disk, not just on paper. It should allow the
writer to

work on documents of any length;

save multiple versions of a file;
save part of the buffer into a file for later use;
switch easily between multiple files;
. insert the contents of an existing file into the buffer;
summarize the differences between two versions of a document.

Most word-processing programs for personal computers seem to work best for short
documents such as the letters and memos that offices chum out by the millions each
day. Although it is possible to create longer documents, many features that would help
organize a large document such as a book or manual are missing from these programs.
However, long before word processors became popular, programmers were using
another class of programs called text editors. Text editors were designed chiefly for
entering computer programs, not text. Furthermore, they were designed for use by com-
puter professionals, not computer novices. A s a result, a text editor can be more diffi-
cult to learn, lacking many on-screen formatting features available with most word pro-
cessors.
Nonetheless, the text editors used in program development environments can pro-
vide much better facilities for managing large writing projects than their office word-
processing counterparts. Large programs, like large documents, are often contained in
many separate files; furthermore, it is essential to track the differences between versions
of a program.
UNIX is a pre-eminent program development environment and, as such, it is also
a superb document development environment. Although its text editing tools at first
may appear limited in contrast to sophisticated office word processors, they are in fact
considerably more powerful.
4 0 UNIX Text Processing 0

Tools for Editing

For many, the ability to retrieve a document from a file and make multiple revisions
painlessly makes it impossible to write at a typewriter again. However, before you can
get the benefits of word processing, there is a lot to learn.
Editing operations are performed by issuing commands. Each word-processing
system has its own unique set of commands. A t a minimum, there are commands to

move to a particular position in the document;

insert new text;
change or replace text;
delete text;
copy or move text.

To make changes to a document, you must be able to move to that place in the text
where YOU want to make your edits. Most documents are too large to be displayed in
their entirety on a single terminal screen, which generally displays 24 lines of text.
Usually only a portion of a document is displayed. This partial view of your document
i s sometimes referred to as a window.* If you are entering new text and reach the bot-
tom line in the window, the text on the screen automatically scrolls (rolls up) to reveal
an additional line at the bottom. A cursor (an underline or block) marks your current
position in the window.
There are basically two kinds of movement:

scrolling new text into the window

positioning the cursor within the window

When you begin a session, the first line of text is the first line in the window, and the
cursor is positioned on the first character. Scrolling commands change which lines are
displayed in the window by moving forward or backward through the document.
Cursor-positioning commands allow you to move up and down to individual lines, and
along lines to particular characters.
After you position the cursor, you must issue a command to make the desired
edit. The command you choose indicates how much text will be affected: a character, a
word, a line, or a sentence.
Because the same keyboard is used to enter both text and commands, there must
be some way to distinguish between the two. Some word-processing programs assume
that you are entering text unIess you specify otherwise; newly entered text either

*Some editors, such as emacs, can split the terminal screen into multiple windows. In addition, many
high-powered UNIX workstations with large bit-mapped screens have their own windowing software that
allows multiple programs to be run simultaneously in separate windows. For purposes of this book, we
assume you are using the v i editor and an alphanumeric terminal with only a single window.
 0 From Typewriters to Word Processors 0 5

replaces existing text or pushes it over to make room for the new text. Commands are
entered by pressing special keys on the keyboard, or by combining a standard key with
a special key, such as the control key (CTRL).
Other programs assume that you are issuing commands; you must enter a com-
mand before you can type any text at all. There are advantages and disadvantages to
each approach. Starting out in text mode is more intuitive to those coming from a type-
writer, but may be slower for experienced writers, because all commands must be
entered by special key combinations that are often hard to reach and slow down typing.
(We’ll return to this topic when we discuss v i , a UNIX text editor.)
Far more significant than the style of command entry is the range and speed of
commands. For example, though it is heaven for someone used to a typewriter to be
able to delete a word and type in a replacement, it is even better to be able to issue a
command that will replace every occurrence of that word in an entire document. And,
after you start making such global changes, it is essential to have some way to undo
them if you make a mistake.
A word processor that substitutes ease of learning for ease of use by having fewer
commands will ultimately fail the serious writer, because the investment of time spent
learning complex commands can easily be repaid when they simplify complex tasks.
And when you do issue a complex command, it is important that it works as
quickly as possible, so that you aren’t left waiting while the computer grinds away.
The extra seconds add up when you spend hours or days at the keyboard, and, once
having been given a taste of freedom from drudgery, writers want as much freedom as
they can get.
Text editors were developed before word processors (in the rapid evolution of
computers). Many of them were originally designed for printing terminals, rather than
for the CRT-based terminals used by word processors. These programs tend to have
commands that work with text on a line-by-line basis. These commands are often more
obscure than the equivalent office word-processing commands.
However, though the commands used by text editors are sometimes more difficult
to learn, they are usually very effective. (The commands designed for use with slow
paper terminals were often extraordinarily powerful, to make up for the limited capabili-
ties of the input and output device.)
There are two basic kinds of text editors, line editors and screen editors, and both
are available in UNIX. The difference is simple: line editors display one line at a time,
and screen editors can display approximately 24 lines or a full screen.
The line editors in UNIX include ed, sed, and ex. Although these line edi-
tors are obsolete for general-purpose use by writers, there are applications at which they
excel, as we will see in Chapters 7 and 12.
The most common screen editor in UNIX is v i . Learning v i or some other
suitable editor is the first step in mastering the UNIX text-processing environment.
Most of your time will be spent using the editor.
UNIX screen editors such as v i and emacs (another editor available on many
UNIX systems) lack ease-of-learning features common in many word processors-there
are no menus and only primitive on-line help screens, and the commands are often com-
plex and nonintuitive-but they are powerful and fast. What’s more, UNIX line editors
such as e x and sed give additional capabilities not found in word processors-the
6 0 UNIX Text Processing 0

ability to write a script of editing commands that can be applied to multiple files. Such
editing scripts open new ranges of capability to the writer.

Document Formatting
Text editing is wonderful, but the object of the writing process is to produce a printed
document for others to read. And a printed document is more than words on paper; it is
an arrangement of text on a page. For instance, the elements of a business letter are
arranged in a consistent format, which helps the person reading the letter identify those
elements. Reports and more complex documents, such as technical manuals or books,
require even greater attention to formatting. The format of a document conveys how
information is organized, assisting in the presentation of ideas to a reader.
Most word-processing programs have built-in formatting capabilities. Formatting
commands are intermixed with editing commands, so that you can shape your document
on the screen. Such formatting commands are simple extensions of those available to
someone working with a typewriter. For example, an automatic centering command
saves the trouble of manually counting characters to center a title or other text. There
may also be such features as automatic pagination and printing of headers or footers.
Text editors, by contrast, usually have few formatting capabilities. Because they
were designed for entering programs, their formatting capabilities tend to be oriented
toward the formats required by one or more programming languages.
Even programmers write reports, however. Especially at AT&T (where UNIX
was developed), there was a great emphasis on document preparation tools to help the
programmers and scientists of Bell Labs produce research reports, manuals, and other
documents associated with their development work.
Word processing, with its emphasis on easy-to-use programs with simple on-
screen formatting, was in its infancy. Computerized phototypesetting, on the other
hand, was already a developed art. Until quite recently, it was not possible to represent
on a video screen the variable type styles and sizes used in typeset documents. As a
result, phototypesetting has long used a markup system that indicates formatting instruc-
tions with special codes. These formatting instructions to the computerized typesetter
are often direct descendants of the instructions that were formerly given to a human
typesetter-center the next line, indent five spaces, boldface this heading.
The text formatter most commonly used with the UNIX system is called n r o f f .
To use it, you must intersperse formatting instructions (usually one- or two-letter codes
preceded by a period) within your text, then pass the file through the formatter. The
n r o f f program interprets the formatting codes and reformats the document “on the
fly” while passing it on to the printer. The n r o f f formatter prepares documents for
printing on line printers, dot-matrix printers, and letter-quality printers. Another pro-
gram called t r o f f uses an extended version of the same markup language used by
n r o f f , but prepares documents for printing on laser printers and typesetters. We’ll
talk more about printing in a moment.
Although formatting with a markup language may seem to be a far inferior system
to the “what you see is what you get” (wysiwyg) approach of most office word-
processing programs, it actually has many advantages.
 0 From Typewriters to Word Processors 0 7

First, unless you are using a very sophisticated computer, with very sophisticated
software (what has come to be called an electronic publishing system, rather than a
mere word processor), it is not possible to display everything on the screen just as it
will appear on the printed page. For example, the screen may not be able to represent
boldfacing or underlining except with special formatting codes. Wordstar, one of the
grandfathers of word-processing programs for personal computers, represents underlin-
ing by surrounding the word or words to be underlined with the special control charac-
ter AS (the character generated by holding down the control key while typing the letter
S). For example, the following title line would be underlined when the document is
printed:
^Sword Processing with WordStar”S

Is this really superior to the following nrof f construct?

.u l
Text Processing with vi and nroff

It is perhaps unfair to pick on Wordstar, an older word-processing program, but very

few word-processing programs can complete the illusion that what you see on the
screen is what you will get on paper. There is usually some mix of control codes with
on-screen formatting. More to the point, though, is the fact that most word processors
are oriented toward the production of short documents. When you get beyond a letter,
memo, or report, you start to understand that there is more to formatting than meets the
eye.
Although “what you see is what you get” is fine for laying out a single page, it is
much harder to enforce consistency across a large document. The design of a large
document is often determined before writing is begun, just as a set of plans for a house
are drawn up before anyone starts construction. The design is a plan for organizing a
document, arranging various parts so that the same types of material are handled in the
same way.
The parts of a document might be chapters, sections, or subsections. For instance,
a technical manual is often organized into chapters and appendices. Within each
chapter, there might be numbered sections that are further divided into three or four lev-
els of subsections.
Document design seeks to accomplish across ?he entire document what is accom-
plished by the table of contents of a book. It presents the structure of a document and
helps the reader locate information.
Each of the parts must be clearly identified. The design specifies how they will
look, trying to achieve consistency throughout the document. The strategy might
specify that major section headings will be all uppercase, underlined, with three blank
lines above and two below, and secondary headings will be in uppercase and lowercase,
underlined, with two blank lines above and one below.
If you have ever tried to format a large document using a word processor, you
have probably found it difficult to enforce consistency in such formatting details as
these. By contrast, a markup language-especially one like nrof f that allows you to
define repeated command sequences, or macros-makes it easy: the style of a heading
is defined once, and a code used to reference it. For example, a top-level heading might
be specified by the code .H 1 , and a secondary heading by H2. I
8 0 UNlX Text Processing 0

Even more significantly, if you later decide to change the design, you simply
change the definition o f the relevant design elements. If you have used a word proces-
sor to format the document as it was written, it is usually a painful task to go back and
change the format.
Some word-processing programs, such as Microsoft WORD, include features for
defining global document formats, but these features are not as widespread as they are
in markup systems.

Printing
The formatting capabilities of a word-processing system are limited by what can be out-
put on a printer. For example, some printers cannot backspace and therefore cannot
underline. For this discussion, we are considering four different classes of printers: dot
matrix, letter quality, phototypesetter, and laser.
A dot-matrix printer composes characters as a series o f dots. It is usually suitable
for preparing interoffice memos and obtaining fast printouts o f large files.

TL,:
II,L>
- paraqraph was printed ~ l t ha dot-fiatrig p r i n t e r . It m e 5 a print
head cc.ntaining 9 pins, which are adjuster! t o produce the shape ci each
c h a r a c t x . More rophicated dot-aatrix p r i n t e r s h a r e p r i n t heads
contaising up t o 24 pins. The greater t h e nufiber o f pins, the finer
t h e d o t s ?hat a r e printed, and t h e mure psssible i t 1 5 tcr io01 the eye
into t h i n k i n g i t sees a solid character. Got atatrix prioters are ais0
c a p i h j ~r ~ f prifiting o u t graphic disp!ays.

A letter-quality printer is more expensive and slower. Its printing mechanism

operates like a typewriter and achieves a similar result.

This paragraph was printed with a letter-

quality printer. It is essentially a
computer-controlled typewriter and, like a
typewriter, uses a print ball or wheel
containing fully formed characters.

A letter-quality printer produces clearer, easier-to-read copy than a dot-matrix printer.

Letter-quality printers are generally used in offices for formal correspondence as well as
for the final drafts of proposals and reports.
Until very recently, documents that needed a higher quality o f printing than that
available with letter-quality printers were sent out for typesetting. Even if draft copy
was word-processed, the material was often re-entered by the typesetter, although many
typesetting companies can read the files created by popular word-processing programs
and use them as a starting point for typesetting.
 0 From Typewriters to Word Processors 9

This paragraph, like the rest of this book, was phototypeset. In photo-
typesetting, a photographic technique is used to print characters on film or
photographic paper. There is a wide choice of type styles, and the charac-
ters are much more finely formed that those produced by a letter-quality
printer. Characters are produced by an arrangement of tiny dots, much like
a dot-matrix printer-but there are over 1000 dots per inch.

There are several major advantages to typesetting. The high resolution allows for the
design of aesthetically pleasing type. The shape of the characters is much finer. In
addition, where dot-matrix and letter-quality type is usually constant width (narrow
letters like i take up the same amount of space as wide ones like m), typesetters use
variable-width type, in which narrow letters take up less space than wide ones. In addi-
tion, it’s possible to mix styles (for example, bold and italic) and sizes of type on the
same page.
Most typesetting equipment uses a markup language rather than a wysiwyg
approach to specify point sizes, type styles, leading, and so on. Until recently, the tech-
nology didn’t even exist to represent on a screen the variable-width typefaces that
appear in published books and magazines.
AT&T, a company with its own extensive internal publishing operation,
developed its own typesetting markup language and typesetting program-a sister to
n r o f f called t r o f f (typesetter-rofJ). Although trof f extends the capabilities of
n r o f f in significant ways, it is almost totally compatible with it.
Until recently, unless you had access to a typesetter, you didn’t have much use for
t r o f f . The development of low-cost laser printers that can produce near typeset-
quality output at a fraction of the cost has changed all that.

This paragraph was produced on a laser printer. Laser printers produce

high-resolution characters-300 to 500 dots per inch-though they are not
quite as finely formed as phototypeset characters. Laser printers are not
only cheaper to purchase than phototypesetters, they also print on plain
paper, just like Xerox machines, and are therefore much cheaper to operate.
However, as i s always the case with computers, you need the proper
software to take advantage of improved hardware capabilities.

Word-processing software (particularly that developed for the Apple Macintosh, which
has a high-resolution graphics screen capable of representing variable type fonts) is
beginning to tap the capabilities of laser printers. However, most of the
microcomputer-based packages still have many limitations. Nonetheless, a markup
language such as that provided by t rof f still provides the easiest and lowest-cost
access to the world of electronic publishing for many types of documents.
The point made previously, that markup languages are preferable to wysiwyg sys-
tems for large documents, is especially true when you begin to use variable size fonts,
leading, and other advanced formatting features. I t is easy to lose track of the overall
format of your document and difficult to make overall changes after your formatted text
is in place. Only the most expensive electronic publishing systems (most of them based
on advanced UNIX workstations) give you both the capability to see what you will get
on the screen and the ability to define and easily change overall document formats.
10 0 UNIX Text Processing 0

Other UNIX Text-Processing Tools

Document editing and formatting are the most important parts of text processing, but
they are not the whole story. For instance, in writing many types of documents, such as
technical manuals, the writer rarely starts from scratch. Something is already written,
whether it be a first draft written by someone else, a product specification, or an out-
dated version of a manual. It would be useful to get a copy of that material to work
with. If that material was produced with a word processor or has been entered on
another system, UNIX’s communications facilities can transfer the file from the remote
system to your own.
Then you can use a number of custom-made programs to search through and
extract useful information. Word-processing programs often store text in files with dif-
ferent internal formats. UNIX provides a number of useful analysis and translation
tools that can help decipher files with nonstandard formats. Other tools allow you to
“cut and paste” portions of a document into the one you are writing.
As the document is being written, there are programs to check spelling, style, and
diction. The reports produced by those programs can help you see if there is any
detectable pattern in syntax or structure that might make a document more difficult for
the user than it needs to be.
Although many documents are written once and published or filed, there is also a
large class of documents (manuals in particular) that are revised again and again. Docu-
ments such as these require special tools for managing revisions. UNIX program
development tools such as SCCS (Source Code Control System) and dif f can be
used by writers to compare past versions with the current draft and print out reports of
the differences, or generate printed copies with change bars in the margin marking the
differences.
In addition to all of the individual tools it provides, UNIX is a particularly fertile
environment for writers who aren’t afraid of computers, because it is easy to write com-
mand files, or shell scripts, that combine individual programs into more complex tools
to meet your specific needs. For example, automatic index generation is a complex task
that is not handled by any of the standard UNIX text-processing tools. We will show
you ways to perform this and other tasks by applying the tools available in the UNIX
environment and a little ingenuity.
We have two different objectives in this book. The first objective is that you
learn to use many of the tools available on most UNIX systems. The second objective
is that you develop an understanding of how these different tools can work together in a
document preparation system. We’re not just presenting a UNIX user’s manual, but
suggesting applications for which the various programs can be used.
To take full advantage of the UNIX text-processing environment, you must do
more than just learn a few programs. For the writer, the job includes establishing stan-
dards and conventions about how documents will be stored, in what format they should
appear in print, and what kinds of programs are needed to help this process take place
efficiently with the use of a computer. Another way of looking at it is that you have to
make certain choices prior to beginning a project. We want to encourage you to make
your own choices, set your own standards, and realize the many possibilities that are
open to a diligent and creative person.
 0 From Typewriters to Word Processors 0 11

In the past, many of the steps in creating a finished book were out of the hands of
the writer. Proofreaders and copyeditors went over the text for spelling and grammati-
cal errors. It was generally the printer who did the typesetting (a service usually paid
by the publisher). A t the print shop, a typesetter (a person) retyped the text and speci-
fied the font sizes and styles. A graphic artist, performing layout and pasteup, made
many of the decisions about the appearance of the printed page.
Although producing a high-quality book can still involve many people, UNIX
provides the tools that allow a writer to control the process from start to finish. An
analogy is the difference between an assembly worker on a production line who views
only one step in the process and a craftsman who guides the product from beginning to
end. The craftsman has his own system of putting together a product, whereas the
assembly worker has the system imposed upon him.
After you are acquainted with the basic tools available in UNIX and have spent
some time using them, you can design additional tools to perform work that you think
is necessary and helpful. To create these tools, you will write shell scripts that use the
resources of UNIX in special ways. We think there is a certain satisfaction that comes
with accomplishing such tasks by computer. It seems to us to reward careful thought.
What programming means to us is that when we confront a problem that normally
submits only to tedium or brute force, we think of a way to get the computer to solve
the problem. Doing this often means looking at the problem in a more general way and
solving it in a way that can be applied again and again.
One of the most important books on UNIX is The UNIX Programming Environ-
ment by Brian W. Kernighan and Rob Pike. They write that what makes UNIX effec-
tive ?is an approach to programming, a philosophy of using the computer.? A t the
heart of this philosophy ?is the idea that the power of a system comes more from the
relationships among programs than from the programs themselves.?
When we talk about building a document preparation system, it is this philosophy
that we are trying to apply. A s a consequence, this is a system that has great flexibility
and gives the builders a feeling of breaking new ground. The UNIX text-processing
environment is a system that can be tailored to the specific tasks you want to accom-
plish. In many instances, it can let you do just what a word processor does. In many
more instances, it lets you use more of the computer to do things that a word processor
either can?t do or can?t do very well.
 C H A P T E R

UNIX Fundamentals

The UNIX operating system is a collection of programs that controls and organizes the
resources and activities of a computer system. These resources consist of hardware
such as the computer’s memory, various peripherals such as terminals, printers, and disk
drives, and software utilities that perform specific tasks on the computer system. UNIX
is a multiuser, multitasking operating system that allows the computer to perform a
variety of functions for many users. It also provides users with an environment in
which they can access the computer’s resources and utilities. This environment is
characterized by its command interpreter, the shell.
In this chapter, we review a set of basic concepts for users working in the UNIX
environment. As we mentioned in the preface, this book does not replace a general
introduction to UNIX. A complete overview is essential to anyone not familiar with the
file system, input and output redirection, pipes and filters, and many basic utilities. In
addition, there are different versions of UNIX, and not all commands are identical in
each version. In writing this book, we’ve used System V Release 2 on a Convergent
Technologies’ Miniframe.
These disclaimers aside, if it has been a while since you tackled a general intro-
duction, this chapter should help refresh your memory. If you are already familiar with
UNIX, you can skip or skim this chapter.
As we explain these basic concepts, using a tutorial approach, we demonstrate the
broad capabilities of UNIX as an applications environment for text-processing. What
you learn about UNIX in general can be applied to performing specific tasks related to
text-processing.

TheUNIXShell
As an interactive computer system, UNIX provides a command interpreter called a
shell. The shell accepts commands typed at your terminal, invokes a program to per-
form specific tasks on the computer, and handles the output or result of this program,
normally directing it to the terminal’s video display screen.

12
 0 UNIX Fundamentals 0 13

UNIX commands can be simple one-word entries like the date command:
$ date
Tue A p r 8 13:23:41 EST 1 9 8 7

Or their usage can be more complex, requiring that you specify options and arguments,
such as filenames. Although some commands have a peculiar syntax, many UNIX
commands follow this general form:
command option(s) argument(s)
A command identifies a software program or utility. Commands are entered in
lowercase letters. One typical command, Is, lists the files that are available in your
immediate storage area, or directory.
An option modifies the way in which a command works. Usually options are
indicated by a minus sign followed by a single letter. For example, Is -1 modifies
what information is displayed about a file. The set of possible options is particular to
the command and generally only a few of them are regularly used. However, if you
want to modify a command to perform in a special manner, be sure to consult a UNIX
reference guide and examine the available options.
An argument can specify an expression or the name of a file on which the com-
mand is to act. Arguments may also be required when you specify certain options. In
addition, if more than one filename is being specified, special metacharacters (such as
* and ?) can be used to represent the filenames. For instance, Is -1 ch* will
display information about all files that have names beginning with c h .
The UNIX shell is itself a program that is invoked as part of the login process.
When you have properly identified yourself by logging in, the UNIX system prompt
appears on your terminal screen.
The prompt that appears on your screen may be different from the one shown in
the examples in this book. There are two widely used shells: the Bourne shell and the
C shell. Traditionally, the Bourne shell uses a dollar sign ($) as a system prompt, and
the C shell uses a percent sign (%). The two shells differ in the features they provide
and in the syntax of their programming constructs. However, they are fundamentally
very similar. In this book, we use the Bourne shell.
Your prompt may be different from either of these traditional prompts. This is
because the UNIX environment can be customized and the prompt may have been
changed by your system administrator. Whatever the prompt looks like, when it
appears, the system is ready for you to enter a command.
When you type a command from the keyboard, the characters are echoed on the
screen. The shell does not interpret the command until you press the RETURN key.
This means that you can use the erase character (usually the DEL or BACKSPACE key)
to correct typing mistakes. After you have entered a command line, the shell tries to
identify and locate the program specified on the command line. If the command line
that you entered is not valid, then an error message is returned.
When a program is invoked and processing begun, the output it produces is sent
to your screen, unless otherwise directed. To interrupt and cancel a program before it
has completed, you can press the interrupt character (usually CTRL-C or the DEL key).
If the output of a command scrolls by the screen too fast, you can suspend the output by
14 0 UNIX Text Processing 0

pressing the suspend character (usually CTRL-S) and resume it by pressing the resume
character (usually CTRL-0).
Some commands invoke utilities that offer their own environment-with a com-
mand interpreter and a set of special “internal” commands. A text editor is one such
utility, the mail facility another. In both instances, you enter commands while you are
“inside” the program. In these kinds of programs, you must use a command to exit
and return to the system prompt.
The return of the system prompt signals that a command is finished and that you
can enter another command. Familiarity with the power and flexibility of the UNIX
shell is essential to working productively in the UNIX environment.

Output Redirection
Some programs do their work in silence, but most produce some kind of result, or out-
put. There are generally two types of output: the expected result-referred to as staan-
durd output-and error messages-referred to as standard error. Both types of output
are normally sent to the screen and appear to be indistinguishable. However, they can
be manipulated separately-a feature we will later put to good use.
Let’s look at some examples. The echo command is a simple command that
displays a string of text on the screen.
$ echo my name
my name

In this case, the input echo m y name is processed and its output is m y name.
The name of the command-echo-refers to a program that interprets the command-
line arguments as a literal expression that is sent to standard output. Let’s replace
echo with a different command called c a t :
$ cat my name
cat: Cannot open m y
cat: C a n n o t o p e n name

The c a t program takes its arguments to be the names of files. If these files existed,
their contents would be displayed on the screen. Because the arguments were not
filenames in this example, an error message was printed instead.
The output from a command can be sent to a file instead of the screen by using
the output redirection operator (>). In the next example, we redirect the output of the
echo command to a file named r e m i n d e r s .
$ echo C a l l home a t 3:QO > reminders
$

No output is sent to the screen, and the UNIX prompt returns when the program is fin-
ished. Now the c a t command should work because we have created a file.
$ cat reminders
C a l l home at 3:OO

The c a t command displays the contents of the file named r e m i n d e r s on the

screen. If we redirect again to the same filename, we overwrite its previous contents:
 0 UNIX Fundamentals 0 15

$ echo Pick up expense voucher > reminders

$ cat reminders
Pick up expense voucher
We can send another line to the file, but we have to use a different redirect operator to
append (>>) the new line at the end of the file:
$ echo Call home at 3:OO > reminders
$ echo Pick up expense voucher >> reminders
$ cat reminders
Call home at 3 : O O
Pick up expense voucher
The cat command is useful not only for printing a file on the screen, but for con-
catenating existing files (printing them one after the other). For example:
$ cat reminders todolist
Call home at 3 : O O
Pick up expense voucher
Proofread Chapter 2
Discuss output redirection
The combined output can also be redirected:
$ cat reminders todolist > do-now

The contents of both reminders and todolist are combined into do-now.
The original files remain intact.
If one of the files does not exist, an error message is printed, even though stan-
dard output is redirected:
$ rm todolist
$ cat reminders todolist > do now -
cat: todolist: not found
The files we’ve created are stored in our current working directory.

Files and Directories

The UNIX file system consists of files and directories. Because the file system can
contain thousands of files, directories perform the same function as file drawers in a
paper file system. They organize files into more manageable groupings. The file sys-
tem is hierarchical. It can be represented as an inverted tree structure with the root
directoiy at the top. The root directory contains other directories that in turn contain
other directories.*

*In addition to subdirectories, the root directory can contain otherfile systems. A file system is the skeletal
structure of a directory tree, which is built on a magnetic disk before any files or directories are stored on it.
On a system containing more than one disk, or on a disk divided into several partitions, there are multiple
file systems. However, this is generally invisible to the user, because the secondary file systems are
mounted on the root directory, creating the illusion of a single file system.
16 0 UNIX Text Processing 0

On many UNIX systems, users store their files in the /usr file system. (As disk
storage has become cheaper and larger, the placement of user directories is no longer
standard. For example, on our system, /usr contains only UNIX software: user
accounts are in a separate file system called /work.)
Fred’s home directory is /usr/fred. It is the location of Fred’s account on
the system. When he logs in, his home directory is his current working directory. Your
working directory is where you are currently located and changes as you move up and
down the file system.
A pathname specifies the location of a directory or file on the UNIX file system.
An absolute pathname specifies where a file or directory is located off the root file sys-
tem. A relative pathname specifies the location of a file or directory in relation to the
current working directory.
To find out the pathname of our current directory, enter pwd.
$ pwd
/usr/fred
The absolute pathname of the current working directory is /usr/fred. The Is
command lists the contents of the current directory. Let’s list the files and subdirec-
tories in /usr/ fred by entering the 1s command with the -F option. This option
prints a slash ( / ) following the names of subdirectories. In the following example,
oldstuff is a directory, and notes and reminders are files.
$ IS -F
reminders
notes
oldstuf f/
When you specify a filename with the 1s command, it simply prints the name of
the file, if the file exists. When you specify the name of directory, it prints the names
of the files and subdirectories in that directory.
$ 1s reminders
reminders
$ 1s oldstuff
chOl-draft
letter.212
memo
In this example, a relative pathname is used to specify oldstuf f. That is, its loca-
tion is specified in relation to the current directory, /usr/fred. You could also
enter an absolute pathname, as in the following example:
$ 1s /usr/fred/oldstuff
chOl-draft
letter.212
memo
Similarly, you can use an absolute or relative pathname to change directories using the
cd command. To move from /usr/fred to /usr/fred/oldstuff,you can
enter a relative pathname:
 0 UNlX Fundamentals 0 17

$ cd oldstuff
$ pwd
/usr/fred/oldstuff

The directory / u s r / f red/oldstuff becomes the current working directory.

The cd command without an argument returns you to your home directory.
$ cd
When you log in, you are positioned in your home directory, which is thus your current
working directory. The name of your home directory is stored in a shell variable that is
accessible by prefacing the name of the variable (HOME) with a dollar sign ($). Thus:
$ echo $HOME
/usr/ f red

You could also use this variable in pathnames to specify a file or directory in your
home directory.
$ 1s $HOME/oldstuff/memo
/usr/fred/oldstuff/memo

In this tutorial, /usr/fred is our home directory.

The command to create a directory is mkdir. An absolute or relative pathname
can be specified.
$ mkdir /usr/fred/reports
$ mkdir reports/monthly

Setting up directories is a convenient method of organizing your work on the system.

For instance, in writing this book, we set up a directory /work/textp and, under
that, subdirectories for each chapter in the book (/work/textp/chOl,
/work/textp/ch02,etc.). In each of those subdirectories, there are files that
divide the chapter into sections (sectl, sect2,etc.). There is also a subdirectory
set up to hold old versions or drafts of these sections.

Copying and Moving Files

You can copy, move, and rename files within your current working directory or (by
specifying the full pathname) within other directories on the file system. The cp com-
mand makes a copy of a file and the mv command can be used to move a file to a new
directory or simply rename it. If you give the name of a new or existing file as the last
argument to cp or mv,the file named in the first argument is copied, and the copy
given the new name. (If the target file already exists, it will be overwritten by the copy.
If you give the name of a directory as the last argument to c p or mv, the file or files
named first will be copied to that directory, and will keep their original names.)
Look at t h e following sequence of commands:
$ pwd Prinr working directory
/usr/fred
~

18 0 UNIX Text Processing 0

$ IS -F List contents of current directory

meeting
oldstuf f/
notes
reports/
$ IUV notes oldstuff Move notes to oldstuff directory
$ 1s List contents of current directory
meeting
oldstuff
reports/
$ mv meeting meet.306 Rename meeting
$ 1s oldstuff List contents of oldstuff subdirectory
chOl-draft
letter.212
memo
notes
In this example, the m v command was used to rename the file meeting and to move
the file notes from /usr/fred to /usr/fred/oldstuff. You can also
use the mv command to rename a directory itself.

Permissions
Access to UNIX files is governed by ownership and permissions. If you create a file,
you are the owner of the file and can set the permissions for that file to give or deny
access to other users of the system. There are three different levels of permission:

r Read permission allows users to read a file or make a copy of it.

W Write permission allows users to make changes to that file.
X Execute permission signifies a program file and allows other users to
execute this program.

File permissions can be set for three different levels of ownership:

owner The user who created the file is its owner.

group A group to which you are assigned, usually made up of those users
engaged in similar activities and who need to share files among them-
selves.
other All other users on the system, the public.

Thus, you can set read, write, and execute permissions for the three levels of own-
ership. This can be represented as:

rwxrwxrwx
I I \
owner group other
 0 UNIX Fundamentals 0 19

When you enter the command Is -1, information about the status of the file is
displayed on the screen. You can determine what the file permissions are, who the
owner of the file is, and with what group the file is associated.
$ 1s -1 meet.306
-rw-rw-r-- 1 fred techpubs 126 March 6 10:32 meet.306

This file has read and write permissions set for the user f r e d and the group
techpubs. All others can read the file, but they cannot modify it. Because f r e d is
the owner of the file, he can change the permissions, making it available to others or
denying them access to it. The chmod command is used to set permissions. For
instance, if he wanted to make the file writeable by everyone, he would enter:
$ chmod o+w meet.306
$ 1s -1 meet.306
-rw-rw-rw- 1 fred techpubs 126 March 6 10:32 meet.306

This translates to “add write permission (+w)to others (o).” If he wanted to remove
write permission from a file, keeping anyone but himself from accidentally modifying a
finished document, he might enter:
$ chmod go-w meet.306
$ 1s -1 meet.306
-rw-r--r-- 1 fred techpubs 126 March 6 10:32 meet.306

This command removes write permission (-w) from group (9) and other ( 0 ) .
File permissions are important in UNIX, especially when you start using a text
editor to create and modify files. They can be used to protect information you have on
the system.

Special Characters
As part of the shell environment, there are a few special characters (metacharacters) that
make working in UNIX much easier. We won’t review all the special characters, but
enough of them to make sure you see how useful they are.
The asterisk (*) and the question mark (?) are filename generation metacharac-
ters. The asterisk matches any or all characters in a string. By itself, the asterisk
expands to all the names in the specified directory.
$ echo *
meet.306 oldstuff reports

In this example, the echo command displays in a row the names of a11 the files and
directories in the current directory. The asterisk can also be used as a shorthand nota-
tion for specifying one or more files.
$1s meet*
meet. 306
$ 1s /work/textp/ch*
/work/textp/chOl
/work/textp/ch02
 1
20 0 UNlX Text Processing 0

/work/textp/ch03
/work/textp/chapter -make

The question mark matches any single character.

$ 1s /work/textp/chOl/sect?
/work/textp/chOl/sectl
/work/textp/chOl/sect2
/work/textp/chOl/sect3

Besides filename metacharacters, there are other characters that have special meaning
when placed in a command line. The semicolon (;) separates multiple commands on
the same command line. Each command is executed in sequence from left to right, one
before the other.
$cd o1dstuff;pwd;ls
/usr/fred/oldstuff
chOl-draft
letter.212
memo
notes

Another special character is the ampersand (&). The ampersand signifies that a com-
mand should be processed in the background, meaning that the shell does not wait for
the program to finish before returning a system prompt. When a program takes a signi-
ficant amount of processing time, it is best to have it run in the background so that you
can do other work at your terminal in the meantime. We will demonstrate background
processing in Chapter 4 when we look at the nrof f /t rof f text formatter.

Environment Variables
The shell stores useful information about who you are and what you are doing in
environment variables. Entering the s e t command will display a list of the environ-
ment variables that are currently defined in your account.
$ set
PATH .:bin:/usr/bin:/usr/local/bin:/etc
argv 0
cwd /work/textp/ch03
home /usr/fred
shell /bin/sh
status 0
TERM wy50

These variables can be accessed from the command line by prefacing their name with a
dollar sign:
$ echo $TERM
wy50

The TERM variable identifies what type of terminal you are using. It is important that
you correctly define the TERM environment variable, especially because the v i text
 0 UNIX Fundamentals 0 21

editor relies upon it. Shell variables can be reassigned from the command line. Some
variables, such as TERM, need to be exported if they are reassigned, so that they are
available to all shell processes.
$ TERM=tvi925; export TERM Tell UNIX I ' m using a Televideo 925
You can also define your own environment variables for use in commands.
$ friends="alice ed ralph"
$ echo $friends
alice ed ralph

You could use this variable when sending mail.

$ mail $friends
A message t o friends
<CTRL-D>

This command sends the mail message to three people whose names are defined in the
friends environment variable. Pathnames can also be assigned to environment vari-
ables, shortening the amount of typing:
$ pwd
/usr/fr e d
$ book="/work/textp"
$ cd $book
$ pwd
/work/textp

Pipes and Filters

Earlier we demonstrated how you can redirect the output of a command to a file. Nor-
mally, command input is taken from the keyboard and command output is displayed on
the terminal screen. A program can be thought of as processing a stream of input and
producing a stream of output. As we have seen, this stream can be redirected to a file.
In addition, it can originate from or be passed to another command.
A pipe is formed when the output of one command is sent as input to the next
command. For example:
$ 1s I wc

might produce:
10 10 72

The 1s command produces a list of filenames which is provided as input to w c . The

w c command counts the number of lines, words, and characters.
Any program that takes its input from another program, performs some operation
on that input, and writes the result to the standard output is referred to as afilter. Most
UNIX programs are designed to work as filters. This i s one reason why UNIX pro-
grams do not print "friendly" prompts or other extraneous information to the user.
22 0 UNIX Text Processing 0

Because all programs expect-and produce-nly a data stream, that data stream can
easily be processed by multiple programs in sequence.
One of the most common uses of filters is to process output from a command.
Usually, the processing modifies it by rearranging it or reducing the amount of informa-
tion it displays. For example:
$ who List who is on the system, and at which terminal
peter ttyOOl Mar 6 1 7 : 1 2
Walter tty003 Mar 6 13:51
Chris tty004 Mar 6 15:53
Val tty020 Mar 6 15:48
tim tty005 Mar 4 17:23
ruth tty006 Mar 6 17:02
fred ttyOOO Mar 6 10:34
dale tty008 Mar 6 15:26
$ who I sort List the same information in alphabetic order
Chris ttyOO4 Mar 6 15:53
dale ttyO08 Mar 6 15:26
fred ttyOOO Mar 6 10:34
peter ttyOOl Mar 6 17:12
ruth tty006 Mar 6 17:02
t im tty005 Mar 4 17:23
val tty020 Mar 6 15:48
Walter tty003 Mar 6 13:51
$

The s o r t program arranges lines of input in alphabetic or numeric order. It

sorts lines alphabetically by default. Another frequently used filter, especially in text-
processing environments, is g r e p , perhaps UNIX’s most renowned program. The
g r e p program selects lines containing a pattern:
$ who I grep ttyOOl Find out who is on terminal I
peter ttyOOl Mar 6 17:12

One of the beauties of UNIX is that almost any program can be used to filter the output
of any other. The pipe is the master key to building command sequences that go
beyond the capabilities provided by a single program and allow users to create custom
“programs” of their own to meet specific needs.
If a command line gets too long to fit on a single screen line, simply type a
backslash followed by a carriage return, or (if a pipe symbol comes at the appropriate
place) a pipe symbol followed by a carriage return. Instead of executing the command,
the shell will give you a secondary prompt (usually >) so you can continue the line:
$ echo This is a long line shown here as a demonstration I
> wc
1 10 49
This feature works in the Bourne shell only.
 0 UNIX Fundamentals 0 23

Shell Scripts 1

A shell script is a file that contains a sequence of UNIX commands. Part of the flexi-
bility of UNIX is that anything you enter from the terminal can be put in a file and exe-
cuted. To give a simple example, we’ll assume that the last command example (grep)
has been stored in a file called whoison:
$ cat whoison
who I grep t t y O O l
The permissions on this file must be changed to make it executable. After a file
is made executable, its name can be entered as a command.
$ chmod + xwhoison
$ 1s -1 whoison
-rwxrwxr-x 1 fred doc 123 Mar 6 17:34 whois
$ whoison
peter ttyOOl Mar 6 17:12
Shell scripts can do more than simply function as a batch command facility. The basic
constructs of a programming language are available for use in a shell script, allowing
users to perform a variety of complicated tasks with relatively simple programs.
The simple shell script shown above is not very useful because it is too specific.
However, instead of specifying the name of a single terminal line in the file, we can
read the name as an argument on the command line. In a shell script, $ 1 represents
the first argument on the command line.
$ cat whoison
who I grep $1
Now we can find who is logged on to any terminal:
$ whoison t t y O O 4
Chris tty004 Mar 6 15:53
Later in this book, we will look at shell scripts in detail. They are an important part of
the writer’s toolbox, because they provide the “glue” for users of the UNIX system-
the mechanism by which all the other tools can be made to work together.
 C H A P T E R
rn

Learning vi

UNIX has a number of editors that can process the contents of readable files, whether
those files contain data, source code, or text. There are line editors, such as ed and
ex, which display a line of the file on the screen, and there are screen editors, such as
v i and emacs, which display a part of the file on your terminal screen.
The most useful standard text editor on your system is vi. Unlike emacs, it is
available in nearly identical form on almost every UNIX system, thus providing a kind
of text editing linguafranca. The same might be said of ed and ex, but screen edi-
tors are generally much easier to use. With a screen editor you can scroll the page,
move the cursor, delete lines, insert characters, and more, while seeing the results of
your edits as you make them. Screen editors are very popular because they allow you
to make changes as you read a file, much as you would edit a printed copy, only faster.
To many beginners, v i looks unintuitive and cumbersome-instead of letting
you type normally and use special control keys for word-processing functions, it uses all
of the regular keyboard keys for issuing commands. You must be in a special insert
mode before you can type. In addition, there seem to be so many commands.
You can’t learn vi by memorizing every single vi command. Begin by learn-
ing some basic commands. As you do, be aware of the patterns of usage that com-
mands have in common. Be on the lookout for new ways to perform tasks, experiment-
ing with new commands and combinations of commands.
As you become more familiar with vi, you will find that you need fewer key-
strokes to tell v i what to do. You will learn shortcuts that transfer more and more of
the editing work to the computer-where it belongs. Not as much memorization is
required as first appears from a list of vi commands. Like any skill, the more editing
you do, the more you know about it and the more you can accomplish.
This chapter has three sections, and each one corresponds to a set of material
about vi that you should be able to tackle in a single session. After you have finished
each session, put aside the book for a while and do some experimenting. When you
feel comfortable with what you have learned, continue to the next session.

24
 0 Learning vi 0 25

Session 1: Basic Commands

The first session contains the basic knowledge you need to operate the v i editor.
After a general description of vi, you are shown some simple operations. You will
learn how to

open and close a file;

give commands and insert text;
move the cursor;
edit text (change, delete, and copy).

You can use v i to edit any file that contains readable text, whether it is a report, a
series of shell commands, or a program. The v i editor copies the file to be edited into
a buffer (an area temporarily set aside in memory), displays as much of the buffer as
possible on the screen, and lets you add, delete, and move text. When you save your
edits, v i copies the buffer into a permanent file, overwriting the contents of the old
file.

Opening a File
The syntax for the v i command is:
vi filename]

wherefilename is the name of either an existing file or a new file. I f you don’t specify
a filename, v i will open an unnamed buffer, and ask you to name it before you can
save any edits you have made. Press RETURN to execute the command.
A filename must be unique inside its directory. On AT&T (System V) UNIX sys-
tems, it cannot exceed 14 characters. (Berkeley UNIX systems allow longer filenames.)
A filename can include any ASCII character except /, which is reserved as the separa-
tor between files and directories in a pathname. You can even include spaces in a
filename by “escaping” them with a backslash. In practice, though, filenames consist
of any combination o f uppercase and lowercase letters, numbers, and the characters .
(dot) and (underscore). Remember that UNIX is case-sensitive: lowercase filenames
are distinctfrom uppercase filenames, and, by convention, lowercase is preferred.
If you want to open a new file called notes in the current directory, enter:
$ vi notes
The v i command clears the screen and displays a new buffer for you to begin work.
Because notes is a new file, the screen displays a column of rzldes (-) to indicate
that there is no text in the file, not even blank lines.
26 0 UNlX Text Processing 0

If you specify the name of a file that already exists, its contents will be displayed on the
screen. For example:
$ vi l e t t e r

might bring a copy of the existing file 1e t t e r to the screen.

Mr. John Fust

Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02154

Dear Mr. Fust:

In o u r conversation last Thursday, we discussed a

documentation project that would produce a user's manual
on the Alcuin product. Yesterday, I received the product
demo and other materials that you sent me.
-
-
-
-
"letter" 11 lines, 250 characters

The prompt line at the bottom of the screen echoes the name and size of the file.
 0 Learning vi 0 27

Sometimes when you invoke vi, you may get either of the following messages:
[using open mode]

or:
Visual needs addressable cursor or upline capability

In both cases, there is a problem identifying the type of terminal you are using. You
can quit the editing session immediately by typing :q.
Although vi can run on almost any terminal, it must know what kind of terminal
you are using. The terminal type is usually set as part of the UNIX login sequence. If
you are not sure whether your terminal type is defined correctly, ask your system
administrator or an experienced user to help you set up your terminal. If you know
your terminal type (wy50 for instance), you can set your TERM environment variable
with the following command:
TERM=wy50; export TERM

vi Commands
The v i editor has two modes: command mode and insert mode. Unlike many word
processors, vi’s command mode is the initial or default mode. To insert lines of text,
you must give a command to enter insert mode and then type away.
Most commands consist of one or two characters. For example:
i insert
C change

Using letters as commands, you can edit a file quickly. You don’t have to
memorize banks of function keys or stretch your fingers to reach awkward combinations
of keys.
In general, vi commands

are case-sensitive (uppercase and lowercase keystrokes mean different things;

e.g., I is different from i);
are not echoed on the screen;
do not require a RETURN after the command.

There is also a special group of commands that echo on the bottom line of the
screen. Bottom-line commands are indicated by special symbols. The slash ( / ) and the
question mark (?) begin search commands, which are discussed in session 2. A colon
( :) indicates an ex command. You are introduced to one ex command (to quit a file
without saving edits) in this chapter, and the ex line editor is discussed in detail in
Chapter 7.
To tell vi that you want to begin insert mode, press i. Nothing appears on the
screen, but you can now type any text at the cursor. To tell v i to stop inserting text,
press ESC and you will return to command mode.
28 0 UNIX Text Processing 0

For example, suppose that you want to insert the word introduction. If you type
the keystrokes iintroduction, what appears on the screen is
introduction

Because you are starting out in command mode, v i interprets the first keystroke (i)as
the insert command. All keystrokes after that result in characters placed in the file,
until you press ESC. If you need to correct a mistake while in insert mode, backspace
and type over the error.
While you are inserting text, press RETURN to break the lines before the right
margin. An autowrap option provides a carriage return automatically after you exceed
the right margin. To move the right margin in ten spaces, for example, enter :set
wm=lO.
Sometimes you may not know i f you are in insert mode or command mode.
Whenever vi does not respond as you expect, press ESC. When you hear a beep, you
are in command mode.

Saving a File
You can quit working on a file at any time, save the edits, and return to the UNIX
prompt. The vi command to quit and save edits is ZZ. (Note that Z Z is capital-
ized.)
Let’s assume that you create a file called letter to practice vi commands
and that you type in 36 lines o f text. To save the file, first check that you are in com-
mand mode by pressing ESC, and then give the write and save command, ZZ. Your
file is saved as a regular file. The result is:
“letter” [New file] 36 lines, 1331 characters

You return to the UNIX prompt. I f you check the list of files in the directory, by typ-
ing Is at the prompt, the new file is listed:
$ Is
chOl ch02 letter
You now know enough to create a new file. As an exercise, create a file called
letter and insert the text shown in Figure 3-1. When you have finished, type Z Z to
save the file and return to the UNIX prompt.

. Moving the Cursor

Only a small percentage of time in an editing session may be spent adding new text in
insert mode. Much of the time, you will be editing existing text.
In command mode, you can position the cursor anywhere in the file. You start all
basic edits (changing, deleting, and copying text) by placing the cursor at the text that
you want to change. Thus, you want to be able to quickly move the cursor to that
place.
 0 Learning v i 0 29

April 1, 1987

Mr. John Fust

Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02159

Dear Mr. Fust:

In our conversation last Thursday, we discussed a

documentation project that would produce a user's
manual on the Alcuin product. Yesterday, I received
the product demo and other materials that you sent me.

Going through a demo session gave me a much better

understanding of the product. I confess to being
amazed by Alcuin. Some people around here, looking
over my shoulder, were also astounded by the
illustrated manuscript I produced with Alcuin. One
person, a student of calligraphy, was really impressed.

Today, I'll start putting together a written plan

that shows different strategies for documenting
the Alcuin product. After I submit this plan, and
you have had time to review it, let's arrange a
meeting at your company to discuss these strategies.

Thanks again for giving us the opportunity to bid on

this documentation project. I hope we can decide upon
a strategy and get started as soon as possible in order
to have the manual ready in time for the first customer
shipment. I look forward to meeting with you towards
the end of next week.

Sincerely,

Fred Caslon

Fig. 3-1. A sample letter entered with vi

30 0 UNlX Text Processing 0

There are v i commands to move

up, down, left, or right, one character at a time;

forward or backward by blocks of text such as words, sentences, or paragraphs;
forward or backward through a file, one screen at a time.

To move the cursor, make sure you are in command mode by pressing ESC. Give the
command for moving forward or backward in the file from the current cursor position.
When you have gone as far in one direction as possible, you’ll hear a beep and the cur-
sor stops. You cannot move the cursor past the tildes (-) at the end of the file.

Single Movements
The keys h, j, k, and 1,right under your fingertips, will move the cursor:
left one space
down one line
up one line
right one space

You could use the cursor arrow keys (t, &, +, t)or the RETURN and BACK-
SPACE keys, but they are out of the way and are not supported on all terminals.
You can also combine the h, j , k, and 1 keys with numeric arguments and
other v i commands.

Numeric Arguments
You can precede movement commands with numbers. The command 4 1 moves the
cursor (shown as a small box around a letter) four spaces to the right, just like typing
the letter 1 four times (1111).

move right
4 characters

This one concept (being able to multiply commands) gives you more options (and
power) for each command. Keep i t in mind as you are introduced to additional com-
mands.

Movement by Lines
When you saved the file letter, the editor displayed a message telling you how
many lines were in that file. A line in the file is not necessarily the same length as a
 0 Learning vi 0 31

physical line (limited to 80 characters) that appears on the screen. A line is any text
entered between carriage returns. If you type 200 characters before pressing RETURN,
v i regards all 200 characters as a single line (even though those 200 characters look
like several physical lines on the screen).
Two useful commands in line movement are:
0 <zero> move to beginning of line
$ move to end of line

In the following file, the line numbers are shown. To get line numbers on your screen,
enter : s e t nu.
1 W i t h t h e s c r e e n e d i t o r you c a n s c r o l l t h e page,
2 move t h e c u r s o r , d e l e t e l i n e s , a n d i n s e r t c h a r a c t e r s ,
w h i l e s e e i n g t h e r e s u l t s o f e d i t s a s y o u make t h e m .
3 Screen editors a r e v e r y popular.

The number of logical lines (3) does not correspond to the number of physical lines (4)
that you see on the screen. If you enter $, with the cursor positioned on the d in the
word delete, the cursor would move to the period following the word them.
1 With t h e s c r e e n editor you can scroll t h e page,
2 move t h e c u r s o r , d e l e t e l i n e s , a n d i n s e r t c h a r a c t e r s ,
w h i l e s e e i n g t h e r e s u l t s o f e d i t s a s y o u make them,
3 S c r e e n e d i t o r s a r e very p o p u l a r .

If you enter 0 (zero), the cursor would move back to the letter t in the word the, at the
beginning of the line.
1 With t h e s c r e e n editor you can scroll t h e page,
2 move t h e c u r s o r , d e l e t e l i n e s , a n d i n s e r t c h a r a c t e r s ,
w h i l e s e e i n g t h e r e s u l t s o f e d i t s a s y o u make t h e m .
3 S c r e e n e d i t o r s a r e very p o p u l a r .

If you do not use the automatic wraparound option (: s e t wm=lO) in v i , you

must break lines with carriage returns to keep the lines of manageable length.

Movement by Text Blocks

You can also move the cursor by blocks of text (words, sentences, or paragraphs).
The command w moves the cursor forward one word at a time, treating symbols
and punctuation marks as equivalent to words. The following line shows cursor move-
ment caused by ten successive w commands:
move t h e c u r s o r , delete l i n e s , and i n s e r t characters,

You can also move forward one word at a time, ignoring symbols and punctuation
marks, using the command W (note the uppercase W). It causes the cursor to move to
the first character following a blank space. Cursor movement using W looks like this:
move t h e c u r s o r , delete l i n e s , and i n s e r t characters,
32 0 UNlX Text Processing 0

To move backward one word at a time, use the command b. The B command allows
you to move backward one word at a time, ignoring punctuation.
With either the w, W, b, or B commands, you can multiply the movement with
numbers. For example, 2w moves forward two words; 5 B moves back five words,
ignoring punctuation. Practice using the cursor movement commands, combining them
with numeric multipliers.

Simple Edits
When you enter text in your file, i t is rarely perfect. You find errors or want to
improve a phrase. After you enter text, you have to be able to change it.
What are the components of editing? You want to insert text (a forgotten word or
a missing sentence). And you want to delete text (a stray character or an entire para-
graph). You also need to change letters and words (correct misspellings or reflect a
change of mind). You want to move text from one place to another part of your file.
And on occasion, you want to copy text to duplicate it in another part of your file.
There are four basic edit commands: i for insert (which you have already seen),
c for change, d for delete, d then p for move (delete and put), and y for yank
(copy). Each type of edit is described in this section. Table 3-1 gives a few simple
examples.

TABLE 3-1. Basic Editing Commands

I Obiect Change Delete CODV(Yank)

One word cw dw YW
Two words 2cw 2d W 2YW
Three words back 3cb 3db 3Yb
One line cc dd YY o r Y
To end of line c$orC d$orD y$
To beginning of line c0 dO YO
Single character r X Yl

Inserting New Text

You have already used the insert command to enter text into a new file. You also use
the insert command while editing existing text to add characters, words, and sentences.
Suppose you have to insert Today, at the beginning of a sentence. Enter the follow-
ing sequence of commands and text:
 0 Learningvi 0 33

1 1’11 s t a r t p u t t i n g
together a written
p l a n t h a t shows
d i f f e r e n t strategies
3k
moveup3
lines
I‘ll start p u t t i n g
together a written
p l a n t h a t shows

t
~

I’ll start putting

- Today, I‘ll start putting
together a written ‘ T o d a y , <ESC together a written
p l a n t h a t shows insert p l a n t h a t shows
different strategies Today, different strategies
I

In the previous example, v i moves existing text to the right as the new text is inserted.
That is because we are showing v i on an “intelligent” terminal, which can adjust the
screen with each character you type. An insert on a “dumb” terminal (such as an
adm3a) will look different. The terminal itself cannot update the screen for each char-
acter typed (without a tremendous sacrifice of speed), so v i doesn’t rewrite the screen
until after you press ESC. Rather, when you type, the dumb terminal appears to
overwrite the existing text. When you press ESC, the line i s adjusted immediately so
that the missing characters reappear. Thus, on a dumb terminal, the same insert would
appear as follows:

I‘ll start putting

- Today, a r t p u t t i n g
together a written iToday together a written
p l a n t h a t shows insert p l a n t h a t shows
different strategies Today, d i f f e r e n t strategies

Today,-I‘ll start putting

together a written <ESC> together a written
p l a n t h a t shows leave p l a n t h a t shows
different strategies insert mode different strategies
I
34 0 UNlX Text Processing 0

Changing Text
You can replace any text in your file with the change command, c. To identify the
amount of text that you want replaced, combine the change command with a movement
command. For example, c can be used to change text from the cursor
cw to the end of a word
2cb back two words
CS to the end of a line

Then you can replace the identified text with any amount of new text: no characters at
all, one word, or hundreds of lines. The c command leaves you in insert mode until
you press the ESC key.

Words
You can replace a word ( c w ) with a longer word, a shorter word, or any amount of text.
The c w command can be thought of as “delete the word marked and insert new text
until ESC is pressed.”
Suppose that you have the following lines in your file letter and want to
change designing to putting together. You only need to change one word.

I‘ll start I‘ll start

designing a cw -
designin$ a
change a
word

Note that the c w command places a $ at the last character of the word to be changed.

designin$ a putting putting together a

together
<ESC>
enter change

The c w command also works on a portion of a word. For example, to change

putting to puts, position the cursor on the second t, enter c w , then type s and press
ESC. By using numeric prefixes, you can change multiple words or characters immedi-
ately. For example:
3 cw change three words to the right of the cursor
5cl change five letters to the right of the cursor

You don’t need to replace the specified number of words, characters, or lines with a like
amount of text. For example:
 -

0 Learning v i 35

I'll s t a r t I'll start

Eutting together a 2c w designing a
de s i g n i n g
<ESC>

Lines
To replace the entire current line, there is the special change command cc. This com-
mand changes an entire line, replacing that line with the text entered before an ESC.
The cc command replaces the entire line of text, regardless of where the cursor i s
located on the line.
The C command replaces characters from the current cursor position to the end
of the line. It has the same effect as combining c with the special end-of-line indica-
tor, $ (as in cS).

Characters
One other replacement edit is performed with the r command. This command replaces
a single character with another single character. One of its uses is to correct misspel-
lings. You probably don't want to use c w in such an instance, because you would
have to retype the entire word. Use r to replace a single character at the cursor:

Yasterday, I received Yesterday, I received

re
replace a
with e

The r command makes only a single character replacement. You do not have to press
ESC to finish the edit. Following an r command, you are automatically returned to
command mode.

Deleting Text
You can also delete any text in your file with the delete command, d. Like the change
command, the delete command requires an argument (the amount of text to be operated
on). You can delete by word (dw), by line (dd and D), or by other movement com-
mands that you will learn later.
With all deletions, you move to where you want the edit to take place and enter
the delete command (d) followed by the amount of text to be deleted (such as a text
object, w for word).
36 0 UNlX Text Processing 0

Words
Suppose that in the following text you want to delete one instance of the word srurt in
the first line.

Today, I'll s t a r t Today, 1'11-

s t a r t putting together dw s t a r t putting together
a written plan delete word a written plan
t h a t t h shows d i f f e r e n t t h a t t h shows d i f f e r e n t

The dw command deletes from the cursor's position to the end of a word. Thus, d w
can be used to delete a portion of a word.

that+h shows d i f f e r e n t
II I
thatshows d i f f e r e n t
1
I
dw
delete word

As you can see, d w deleted not only the remainder of the word, but also the space
before any subsequent word on the same line. To retain the space between words, use
de, which will delete only to the end of the word.

that+h shows d i f f e r e n t that-shows different

word end

You can also delete backwards (db) or to the end or beginning of a line (dS or do).

Lines
The dd command deletes the entire line that the cursor is on. Using the same text as
in the previous example, with the cursor positioned on the first line as shown, you can
delete the first two lines:
 0 Learning v i 0 37

Today, 1 ’ 1 1 - a written plan

start putting together 2 dd t h a t shows d i f f e r e n t
a written plan delete first
t h a t shows d i f f e r e n t 2 lines

If you are using a dumb terminal or one working at less than 1200 baud, line deletions
look different. The dumb or slow terminal will not redraw the screen until you scroll
past the bottom of the screen. Instead the deletion appears as:

I,“ written plan

t h a t shows d i f f e r e n t

An @ symbol “holds the place” of the deleted line, until the terminal redraws the
entire screen. (You can force v i to redraw the screen immediately by pressing either
CTRL-L or CTRL-R, depending on the terminal you’re using.)
The D command deletes from the cursor position to the end of the line:

Today, I ‘ l l s t a r t Today, I ’ l l s t a r t
putting together a putting together a
written plan delete to written plan that
t h a t shows d i f f e r e n t end of line that-

You cannot use numeric prefixes with the D command.

Characters
Often, while editing a file, you want to delete a single character or two. Just as r
changes one character, x deletes a single character. The x command deletes any char-
acter the cursor is on. In the following line, you can delete the letter 1 by pressing x.

Today, 1’111 s t a r t Today, I’llstart

putting X putting
delete
character
38 0 UNlX Text Processing 0

The X command deletes the character before the cursor. Prefix either of these com-
mands with a number to delete that number of characters. For example, 5 X will delete
the five characters to the left of the cursor.

Moving Text
You can move text by deleting it and then placing that deleted text elsewhere in the file,
like a “cut and paste.” Each time you delete a text block, that deletion is temporarily
saved in a buffer. You can move to another position in the file and use the put com-
mand to place the text in a new position. Although you can move any block of text,
this command sequence is more useful with lines than with words.
The put command, p, places saved or deleted text (in the buffer) after the cursor
position. The uppercase version of the command, P, puts the text hefore the cursor. If
you delete one or more lines, p puts the deleted text on a new line(s) below the cursor.
If you delete a word, p puts the deleted text on the same line after the cursor.
Suppose that in your file 1e t t e r you have the following lines and you want to
move the fourth line of text. Using delete, you can make this edit. First delete the line
in question:

~ ~~

Today, I’ll s t a r t Today, I‘ll s t a r t

putting together a dd putting together a
p l a n for d o c u m e n t i n g delete line p l a n f o r documenting
Lhe A l c u i n p r o d u c t t h a t shows
-
t h a t shows

Then use p to restore the deleted line at the next line below the cursor:

Today, I‘ll s t a r t Today, I‘ll s t a r t

putting together a P putting together a
p l a n for d o c u m e n t i n g 1 restore deleted p l a n for d o c u m e n t i n g
t h a t shows line t h a t shows

You can also use xp (delete character and put after cursor) to transpose two letters.
For example, in the word mvoe, the letters Y O are transposed (reversed). To correct this,
place the cursor on v and press x then p.
After you delete the text, you must restore it before the next change or delete
command. If you make another edit that affects the buffer, your deleted text will be
lost. You can repeat the put command over and over, as long as you don’t make a new
edit. In the advanced v i chapter, you will learn how to retrieve text from named and
numbered buffers.
 ~

0 Learning vi 0 39

Copying Text
Often, you can save editing time (and keystrokes) by copying part of your file to
another place. You can copy any amount of existing text and place that copied text
elsewhere in the file with the two commands y (yank) and p (put). The yank com-
mand is used to get a copy of text into the buffer without altering the original text.
This copy can then be placed elsewhere in the file with the put command.
Yank can be combined with any movement command (for example, yw, y $ . or
4yy). Yank is most frequently used with a line (or more) of text, because to yank and
put a word generally takes longer than simply inserting the word. For example, to yank
five lines of text:

on the Alcuin product. on the Alcuin product.

Yesterday, I received 5YY Yesterday, I received
the product demo yank 5 the product demo
and other materials lines and other materials
that you sent me. that you sent me.
- -
- I

... 5 lines yanked

To place the yanked text, move the cursor to where you want to put the text, and
use the p command to insert it below the current line, or P to insert it above the
current line.

that you sent me. that you sent me.

- P on the Alcuin product.
I
place yanked Yesterday, I received
- text the product demo
- and other materials
I
that you sent me.

5 more lines

The yanked text will appear on the line below the cursor. Deleting uses the same buffer
as yanking. Delete and put can be used in much the same way as yank and put. Each
new deletion or yank replaces the previous contents of the yank buffer. As we’ll see
later, up to nine previous yanks or deletions can be recalled with put commands.
40 0 UNlX Text Processing 0

Using Your Last Command

Each command that you give is stored in a temporary buffer until you give the next
command. If you insert the after a word in your file, the command used to insert the
text, along with the text that you entered, is temporarily saved. Anytime you are mak-
ing the same editing command repeatedly, you can save time by duplicating the com-
mand with . (dot). To duplicate a command, position the cursor anywhere on the
screen, and press . to repeat your last command (such as an insertion or deletion) in
the buffer. You can also use numeric arguments (as in 2 .) to repeat the previous com-
mand more than once.
Suppose that you have the following lines in your. file letter. Place the cur-
sor on the line you want to delete:

Yesterday, I received Yesterday, I received

the product demo. dd the product demo.
-
Yesterday, I received delete line Qther materials
other materials

Yesterday, I received
t h e product demo. the product demo.
other materials repeat last
command (dd)

In some versions of v i , the command CTRL-@ ("e) repeats the last insert (or
append) command. This is in contrast to the command, which repeats the last com-
mand that changed the text, including delete or change commands.
You can also undo your last command if you make an error. To undo a com-
mand, the cursor can be anywhere on the screen. Simply press u to undo the last com-
mand (such as an insertion or deletion).
To continue the previous example:

Yesterday, I received
Lhe product demo. U the product demo.
undo last Qther materials
command

The uppercase version of u (U) undoes all edits on a single line, as long as the cursor
remains on that line. After you move off a line, you can no longer use U.
 ~

0 Learning v i 0 41

Joining Two Lines with J

Sometimes while editing a file, you will end up with a series of short lines that are dif-
ficult to read. When you want to merge two lines, position the cursor anywhere on the
first line and press J to join the two lines.

Yesterday.
- Yesterday,
- I received
I received J t h e product demo.
the product demo. join lines

A numeric argument joins that number of consecutive lines.

Quitting without Saving Edits

When you are first learning vi, especially if you are an intrepid experimenter, there is
one other command that is handy for getting out of any mess that you might create.
You already know how to save your edits with Z Z , but what if you want to wipe out
all the edits you have made in a session and return to the original file?
You can quit vi without saving edits with a special bottom-line command based
on the ex line editor. The ex commands are explained fully in the advanced vi
chapter, but for basic vi editing you should just memorize this command:
:q! <RE TURN>

The q ! command quits the file you are in. All edits made since the last time you
saved the file are lost.
You can get by in v i using only the commands you have learned in this session.
However, to harness the real power of v i (and increase your own productivity) you
will want to continue to the next session.

. Session 2: Moving Around in a Hurry .

You use v i not only to create new files but also to edit existing files. You rarely open
to the first line in the file and move through it line by line. You want to get to a
specific place in a file and start work.
All edits begin with moving the cursor to where the edit begins (or, with ex line
editor commands, identifying the line numbers to be edited). This chapter shows you
how to think about movement in a variety of ways (by screens, text, patterns, or line
numbers). There are many,ways to move in vi, because editing speed depends on get-
ting to your destination with only a few keystrokes.
42 UNlX Text Processing 0

In this session, you will learn how to move around in a file by

screens;
text blocks;
searches for patterns;
lines.

Movement by Screens
When you read a book you think of “places” in the book by page: the page where you
stopped reading or the page number in an index. Some v i files take up only a few
lines, and you can see the whole file at once. But many files have hundreds of lines.
You can think of a v i file as text on a long roll of paper. The screen is a win-
dow of (usually) 24 lines of text on that long roll. In insert mode, as you fill up the
screen with text, you will end up typing on the bottom line of the screen. When you
reach the end and press RETURN, the top line rolls out of sight, and a blank line for
new text appears on the bottom of the screen. This is called scrolling. You can move
through a file by scrolling the screen ahead or back to see any text in the file.

Scrdiing the Screen

There are v i commands to scroll forward and backward through the file by full and
half screens:
^F forward one screen
^B backward one screen
^D forward half screen
^U backward half screen

(The symbol represents the CTRL key.

A “F means to simultaneously press the
CTRL key and the F key.)

In our conversation last Thursday, we

discussed a documentation project that would
produce a user‘s manual on t h e Alcuin product.
Yesterday, I received t h e product demo and
other materials that you sent me.

Going through a demo session gave m e a

much better understanding of t h e product. I
confess t o being amazed by Alcuin. Some
 0 Learning v i 0 43

If you press "F, the screen appears as follows:

better understanding of t h e product. I

confess t o being amazed by Alcuin. Some
people around here, looking over my shoulder,
were a l s o astounded by t h e illustrated
manuscript I produced with Alcuin. One
person, a student of calligraphy, was really
impressed.

I Today, I'll start putting together a written

There are also commands to scroll the screen up one line ("E) and down one line ("Y).
(These commands are not available on small systems, such as the PDP-11 or Xenix for
the PC-XT.)

Movement within a Screen

You can also keep your current screen or view of the file and move around within the
screen using:
H home-top line on screen
M middle line on screen
L last line on screen
nH to n lines below top line
nL to n lines above last line

The H command moves the cursor from anywhere on the screen to the first, or home,
line. The M command moves to the middle line, L to the last. To move to the line
below the first line, use 2 H .

Today, I ' l l s t a r t T o d a y , I'll s t a r t

putting together a Eutting together a
written plan that written plan that
shows t h e d i f f e r e n t second line shows t h e d i f f e r e n t
s t r a t e g i e s for t h e s t r a t e g i e s for t h e

These screen movement commands can also be used for editing. For example, dH
deletes to the top line shown on the screen.
44 0 UNlX Text Processing 0

Movement within Lines

Within the current screen there are also commands to move by line. You have already
learned the line movement commands $ and 0.
RETURN beginning of next line
A
to first character of current line
+ beginning of next line
- beginning of previous line

~~ ~~

Going through a demo Going through a demo

session gave m e a much - session gave me a much
better understanding go to start better understanding
of t h e product. of previous of t h e product.
line

The command moves to the first character of the line, ignoring any spaces or tabs.
A

( 0 , by contrast, moves to the first position of the line, even if that position is blank.)

. Movement by Text Blocks

Another way that you can think of moving through a v i file is by text blocks-words,
sentences, or paragraphs. You have already learned to move forward and backward by
word (w or b).
end of word
end of word (ignore punctuation)
beginning of previous sentence
beginning of next sentence
beginning of previous paragraph
beginning of next paragraph
The vi program locates the end of a sentence by finding a period followed by at
least two spaces, or a period as the last nonblank character on a line. I f you have left
only a single space following a period, the sentence won’t be recognized.
A paragraph is defined as text up to the next blank line, or up to one of the
-
default paragraph macros ( I P , .P , .P P , or .QP) in the mm or m s macro pack-
ages. The macros that are recognized as paragraph separators can be customized with
the :s e t command, as described in Chapter 7.

____~

In our conversation -
In our conversation
last Thursday, we . .. I last Thursday, w e ...
go to S t a n
Going through a demo of previous Going through a demo
paragraph session gave me --.
 Learning v i 45

Most people find it easier to visualize moving ahead, so the forward commands
are generally more useful.
Remember that you can combine numbers with movement. For example, 3 )
moves ahead three sentences. Also remember that you can edit using movement com-
mands: d) deletes to the end of the current sentence, 2 y } copies (yanks) two para-
graphs ahead.

. Movement by Searches
One of the most useful ways to move around quickly in a large file is by searching for
text, or, more properly, for a pattern of characters. The pattern can include a “wild-
card” shorthand that lets you match more than one character. For example, you can
search for a misspelled word or each occurrence of a variable in a program.
The search command i s the slash character (/). When you enter a slash, it
appears on the bottom line of the screen; then type in the pattern (a word or other string
of characters) that you want to find:
/text<RETURN> search forward for text

A space before or after text will be included in the search. As with all bottom-line com-
mands, press RETURN to finish.
The search begins at the cursor and moves forward, wrapping around to the start
of the file if necessary. The cursor will move to the first occurrence of the pattern (or
the message “Pattern not found” will be shown on the status line if there is no match).
If you wanted to search for the pattern shows:

Today, I ’ l l start Today, I‘ll s t a r t

putting together a /shows<CR> p u t t i n g t o g e t h e r a
w r i t t e n Elan that search for w r i t t e n plan that
shows t h e d i f f e r e n t shows shows t h e d i f f e r e n t
-
- -
- -
I
/shows

Today, I ’ l l start Today, I’ll start

putting together a /th<CR> putting together a
written plan that search for written plan that
shows t h e d i f f e r e n t th shows t h e d i f f e r e n t
I I

.-, -
.-,
/th
46 0 UNlX Text Processing 0

The search proceeds forward from the present position in the file. You can give any
combination of characters; a search does not have to be for a complete word.
You can also search backwards using the ? command:
?text<RETURN> search backward for text

The last pattern that you searched for remains available throughout your editing
session. After a search, instead of repeating your original keystrokes, you can use a
command to search again for the last pattern.
n repeat search in same direction
N repeat search in opposite direction
/ <RET URN > repeat search in forward direction
? <RE TURN > repeat search in backward direction
Because the last pattern remains available, you can search for a pattern, do some
work, and then search again for the pattern without retyping by using n, N, /, or ?.
The direction of your search (/=forwards, ?=backwards) is displayed at the bottom left
of the screen.
Continuing the previous example, the pattern th is still available to search for:

Today, I'll s t a r t Today, I'll s t a r t

putting together a n putting together a
written plan t h a t search for written plan that
shows t h e d i f f e r e n t next rh shows t h e d i f f e r e n t
I

Today, I'll s t a r t Today, I'll s t a r t

putting together a ?<CR> putting together a
written plan that search back written plan that
shows t h e d i f f e r e n t for th shows t h e d i f f e r e n t
- -
I -
- ?the

Today, I'll s t a r t Today, I'll s t a r t

putting together a putting together a
written plan that repeat search written plan that
shows t h e d i f f e r e n t in opposite shows t h e d i f f e r e n t
direction
 0 Learning vi 0 47

This section has given only the barest introduction to searching for patterns. Chapter 7
will teach more about pattern matching and its use in making global changes to a file.

Current Line Searches

There is also a miniature version of the search command that operates within the current
line. The command f moves the cursor to the next instance of the character you name.
Semicolons can then be used to repeat the “find.” Note, however, that the f com-
mand will not move the cursor to the next line.
fx find (move cursor to) next occurrence of x in the line, where x can be
any character
I repeat previous find command

Suppose that you are editing on this line:

I
Today, I‘ll s t a r t Today, Iill s t a r t
f‘
find first ’
in line

Use d f ‘ to delete up to and including the named character (in this instance ’). This

=
command is useful in deleting or copying partial lines.
The t command works just like f , except it positions the cursor just before the
character searched for. As with f and b. a numeric prefix will locate the nth
occurrence. For example:

Today, 1‘11 s t a r t Today, 1’11 s f a r t

2ta
place cursor
before 2nd a
in line

Movement by Line Numbers

A file contains sequentially numbered lines, and you can move through a file by speci-
fying line numbers. Line numbers are useful for identifying the beginning and end of
large blocks of text you want to edit. Line numbers are also useful for programmers
because compiler error messages refer to line numbers. Line numbers are also used by
e x commands, as you will learn in Chapter 7.
48 0 UNlX Text Processing 0

If you are going to move by line numbers, you need a way to identify line
numbers. Line numbers can be displayed on the screen using the :set nu option
described in Chapter 7. In v i , you can also display the current line number on the
bottom of the screen.
The command "G displays the following on the bottom of your screen: the
current line number, the total number of lines in the file, and what percentage of the
total the present line number represents. For example, for the file letter, "G might
display:
'.letter" line 10 of 4 0 - - 2 5 % - -
^G is used to display the line number to use in a command, or to orient yourself if you
have been distracted from your editing session.
The G (go to) command uses a line number as a numeric argument, and moves to
the first position on that line. For instance, 4 4 G moves the cursor to the beginning of
line 44. The G command without a line number moves the cursor to the last line of the
file.
Two single quotes ( ' ') return you to the beginning of the line you were origi-
nally on. Two backquotes (' ' ) return you to your original position exactly. If you
* .
have issued a search command ( / or ?), will return the cursor to its position when
you started the search.
The total number of lines shown with "G can be used to give yourself a rough
idea of how many lines to move. If you are on line 10 of a 1000-line file:
"chOl" line 1 0 of 1 0 0 0 --1%--

and know that you want to begin editing near the end of that file, you could give an
approximation of your destination with:
800G

Movement by line number can get you around quickly in a large file.

Session 3: Beyond the Basics

You have already been introduced to the basic v i editing commands, i, c, d, and
y. This session expands on what you already know about editing. You will learn

additional ways to enter v i ;

how to customize v i ;
how to combine all edits with movement commands;
additional ways to enter insert mode;
how to use buffers that store deletions, yanks, and your last command;
. how to mark your place in a file.
 0 Learning vi 0 49

Command-Line Options
There are other options to the v i command that can be helpful. You can open a file
directly to a specific line number or pattern. You can also open a file in read-only
mode. Another option recovers all changes to a file that you were editing when the sys-
tem crashes.

Advancing to a Specific Place

When you begin editing an existing file, you can load the file and then move to the first
occurrence of a pattern or to a specific line number. You can also combine the open
command, v i , with your first movement by search or by line number. For example:
$ vi +n l e t t e r
opens l e t t e r at line number n. The following:
$ vi + letter

opens 1 e t t e r at the last line. And:

$ v i +/pattern l e t t e r

opens 1 e t t e r at the first occurrence of pattern.

To open the file l e t t e r and advance directly to the line containing Afcuin,
enter:
$ v i +/Alcuin l e t t e r

7
Today I'll s t a r t p u t t i n g t o g e t h e r a
written plan that presents the different
strategies for the Blcuin
-

There can be no spaces in the pattern because characters after a space are interpreted as
filenames.
If you have to leave an editing session before you are finished, you can mark your
place by inserting a pattern such as Z Z Z or HERE. Then when you return to the file,
all you have to remember is / Z Z Z or /HERE.
50 0 UNlX Text Processing 0

Read-only Mode
There will be times that you want to look at a file, but you want to protect that file from
inadvertent keystrokes and changes. (You might want to call in a lengthy file to prac-
tice v i movements, or you might want to scroll through a command file or program.)
If you enter a file in read-only mode, you can use all the v i movement commands, but
you cannot change the file with any edits. To look at your file l e t t e r in read-only
mode, you can enter either:
$ v i -R letter
or:
$ view l e t t e r

Recovering a Buffer
Occasionally, there will be a system failure while you are editing a file. Ordinarily, any
edits made after your last write (save) are lost. However, there is an option, -r, which
lets you recover the edited buffer at the time of a system crash. (A system program
called p r e s e r v e saves the buffer as the system is going down.)
When you first log in after the system is running again, you will receive a mail
message stating that your buffer is saved. The first time that you call in the file, use the
-r option to recover the edited buffer. For example, to recover the edited buffer of the
file l e t t e r after a system crash, enter:
$ vi -r l e t t e r
If you first call in the file without using the -r option, your buffered edits are lost.
You can force the system to preserve your buffer even when there i s not a crash
by using the command :pre. You may find this useful if you have made edits to a
file, then discover you can’t save your edits because you don’t have write permission.
(You could also just write a copy of the file out under another name or in a directory
where you do have write permission.)

Customizing vi
A number of options that you can set as part of your editing environment affect how
v i operates. For example, you can set a right margin that will cause v i to wrap lines
automatically, so you don’t need to insert carriage returns.
You can change options from within v i by using the :s e t command. In addi-
tion, v i reads an initialization file in your home directory called . e x r c for further
operating instructions. By placing s e t commands in this file, you can modify the
way v i acts whenever you use it.
You can also set up .e x r c files in local directories to initialize various options
that you want to use in different environments. For example, you might define one set
of options for editing text, but another set for editing source programs. The . e x r c
file in your home directory will be executed first, then the one on your current direc-
tory.
 0 Learning vi 0 51

Finally, if the shell variable E X I N I T is set in your environment (with the

Bourne shell e x p o r t command, or the C shell setenv command), any commands
it contains will be executed by v i on startup. If E X I N I T is set, it will be used
instead of - exrc; v i will not take commands from both.

The set Command

There are two types of options that can be changed with the s e t command: toggle
options, which are either on or off, and options that take a numeric or string value (such
as the location of a margin or the name of a file).
Toggle options may be on or off by default. To turn a toggle option on, the com-
mand is:
:set option

To turn a toggle option off, the command is:

:set nooption

For example, to specify that pattern searches should ignore case, you type:
:set ic

If you want v i to return to being case-sensitive in searches, give the command:

:set noic

Some options have values. For example, the option window sets the number of
lines shown in the screen “window.” You set values for these options with an equals
sign (=). For example:
set window=20

During a v i session, you can check what options are available. The command:
:set all

displays the complete list of options, including options that you have set and defaults
that v i has chosen. The display will look something like this:
\

noautoindent open tabst op=8

autoprint prompt taglength=O
noautowrite noreadonly term=wy5 0
nobeautify redraw noterse
directory=/tmp /remap timeout
noedcompatible report=5 ttytype=wy50
noerrorbells scrolls=ll warn
hardtabs=8 sections=AhBhChDh window=2 0
noignorecase shell=/bin/csh wrapscan
nolisp shiftwidth=8 wrapmargin=lO
nolist noshowmatch nowriteany
magic noslowopen
me sg paragraphs=IPLPPPQP LIpplpipbb
number tags=tags /usr/lib/tags
nooptimize
/
52 UNlX Text Processing 0

You can also ask about the setting for any individual option by name, using the com-
mand:
:set option?

The command :s e t shows options that you have specifically changed, or set, either in
your . e x r c file or during the current session. For example, the display might look
like this:
number window=20 wrapmargin=lO

See Appendix A for a description of what these options mean.

The .e x r c File
The e x r c file that controls the vi environment for you is in your home directory.
Enter into this file the s e t options that you want to have in effect whenever you use
v i or e x .
The .e x r c file can be modified with the v i editor, like any other file. A sam-
ple .e x r c file might look like this:
set wrapmargin=lO window=20

Because the file is actually read by e x before it enters visual mode (vi),commands in
.e x r c should not have a preceding colon.
Alternate Environments
You can define alternate v i environments by saving option settings in an .e x r c file
that is placed in a local directory. If you enter v i from that directory, the local
.e x r c file will be read in. If it does not exist, the one in your home directory will be
read in.
For example, you might want to have one set of options for programming:
set number lisp autoindent sw=4 tags=/usr/lib/tags terse
and another set of options for text editing:
set wrapmargin=15 ignorecase

Local . e x r c files are especially useful when you define abbreviations, which are
described in Chapter 7.

Some Useful Options

As you can see when you type :s e t a l l , there are many options. Most options are
used internally by vi and aren’t usually changed. Others are important in certain
cases, but not in others (for example, n o r e d r a w and window can be useful on a
dialup line at a low baud rate). Appendix A contains a brief description of each option.
We recommend that you take some time to play with option setting-if an option looks
interesting, try setting it (or unsetting it) and watch what happens while you edit. You
may find some surprisingly useful tools.
 0 Learning vi 0 53

There is one option that is almost essential for editing nonprogram text. The
w r a p m a r g i n option specifies the size of the right margin that will be used to
autowrap text as you type. (This saves manually typing carriage returns.) This option
is in effect if its value is set to greater than 0. A typical value is 10 or 15:
set wrapmargin=15

There are also three options that control how v i acts in conducting a search. By
default, it differentiates between uppercase and lowercase (foo does not match Foo),
wraps around to the beginning of the file during a search (this means you can begin
your search anywhere in the file and still find all occurrences), and recognizes wildcard
characters when matching patterns. The default settings that control these options are
n o i g n o r c a s e , w r a p s c a n , and magic, respectively. To change any of these
defaults, set the opposite toggles: i g n o r e c a s e , n o w r a p s c a n , or n o m a g i c .
Another useful option is s h i f t w i d t h . This option was designed to help pro-
grammers properly indent their programs, but it can also be useful to writers. The >>
and << commands can be used to indent (or un-indent) text by s h i f t w i d t h char-
acters. The position of the cursor on the line doesn’t matter-the entire line will be
shifted. The s h i f t w i d t h option is set to 8 by default, but you can use : s e t to
change this value.
Give the >> or << command a numeric prefix to affect more than on line. For
example :
lo>>

will indent the next 10 lines by shiftwidth.

Edits and Movement .

You have learned the edit commands c, d, and y. and how to combine them with
movements and numbers (such as 2 c w or 4 d d ) . Since that point, you have added
many more movement commands to your repertoire. Although the fact that you can
combine edit commands with movement is not a “new” concept to you, Table 3-2
gives you a feel for the many editing options you now have.

TABLE 3-2. Combining vi Commands

From Cursor to Change Delete COPY

Bottom of screen cL dL YL
Next line C+ d+ Y+
Next sentence C) d) Y)
Next paragraph Cl d) Yl
Pattern c /pattern d /pattern y /pattern
End of file cG dG YG
Line number 13 c13G d13G y13G
54 UNlX Text Processing 0

You can also combine numbers with any of the commands in Table 3-2 to multi-
ply them. For example, 2 c ) changes the next two sentences. Although this table may
seem forbidding, experiment with combinations and try to understand the patterns.
When you find how much time and effort you can save, combinations of change and
movement keys will no longer seem obscure, but will readily come to mind.

More Ways to Insert Text

You have inserted text before the cursor with the sequence:
itext <ESC>
There are many insert commands. The difference between them is that they insert text
at different positions relative to the cursor:
a append text after cursor
A append text to end of current line
i insert text before cursor
I insert text at beginning of line
0 open new line below cursor for text
0 open new line above cursor for text
R overstrike existing characters with new characters
All these commands leave you in insert mode. After inserting text, remember to press
ESC to escape back to command mode.
The A (append) and 1 (insert) commands save you from having to move the
cursor to the end or beginning of the line before invoking insert mode. For example, A
saves one keystroke over $a. Although one keystroke might not seem like a
timesaver, as you become a more adept (and impatient) editor, you’ll want to omit any
unnecessary keystrokes.
There are other combinations o f commands that work together naturally. For
example, e a i s useful for appending new text to the end o f a word. (It sometimes
helps to train yourself to recognize such frequent combinations so that invoking them
becomes automatic.)

Using Buffers
While you are editing, you have seen that your last deletion (d or x) or yank ( y ) is
saved in a buffer (a place in stored memory). You can access the contents of that buffer
and put the saved text back in your file with the put command (p or P).
The last nine deletions are stored by v i in numbered buffers. You can access
any o f these numbered buffers to restore any (or all) of the last nine deletions. You can
also place yanks (copied text) in buffers identified by fetters. You can fill up to 26
buffers (a through z) with yanked text and restore that text with a put command any
time in your editing session.
 0 Learning vi 0 55

The v i program also saves your last edit command (insert, change, delete, or
yank) in a buffer. Your last command is available to repeat or undo with a single key-
stroke.

Recovering Deletions
Being able to delete large blocks of text at a single bound i s all well and good, but what
if you mistakenly delete 53 lines that you need? There is a way to recover any of your
past nine deletions, which are saved in numbered buffers. The last deletion is saved in
buffer 1 ; the second-to-last in buffer 2, and so on.
To recover a deletion, type (quotation mark), identify the buffered text by
number, and then give the put command. For example, to recover your second-to-last
deletion from buffer 2, type:
"2p

Sometimes it's hard to remember what's in the last nine buffers. Here's a trick
that can help.
The . command (repeat last command) has a special meaning when used with p
and u. The p command will print the last deletion or change, but 2p will print the
last two. By combining p, . (dot), and u (undo), you can step back through the
numbered buffers.
The l l l pcommand will put the last deletion, now stored in buffer 1 , back into
your text. If you then type u, it will go away. But when you type the . command,
instead of repeating the last command ("lp), it will show the next buffer as if you'd
typed 'I2p. You can thus step back through the buffers. For example, the sequence:
I, 1pu.u.u.u.u.

will show you, in sequence, the contents of the last six numbered buffers.

Yanking to Named Buffers

With unnamed buffers, you have seen that you must put (p or P) the contents of the
buffer before making any other edit, or the buffer is overwritten. You can also use y
with a set of 26 named buffers (a through z), which are specifically for copying and
moving text. If you name a buffer to store the yanked text, you can place the contents
of the named buffer at any time during your editing session.
To yank into a named buffer, precede the yank command with a quotation mark
(") and the character for the name of the buffer you want to load. For example:
"dYY yank current line into buffer d
"a6yy yank next six Iines into buffer a

After loading the named buffers and moving to the new position, use p or P to
put the text back.
" dP put buffer d before cursor
" ap put buffer a after cursor
56 0 UNlX Text Processing 0

-
In our conversation last In o u r conversation last
-
Thursday, we discussed a "a6yy Thursday, we discussed a
documentation project yank 6 lines documentation project
that would produce a to buffer Q that would produce a
user's manual on t h e user's manual on t h e
Alcuin product. Alcuin product.

6 lines yanked

Blcuin product -
I 'lap
Alcuin product.
-
In our conversation last

I put buffer (I
after cursor
Thursday, w e discussed a
documentation project
that would produce a
user's manual on t h e
Alcuin product.

There is no way to put part of a buffer into the text-it is all or nothing.
Named buffers allow you to make other edits before placing the buffer with p.
After you know how to travel between files without leaving v i , you can use named
buffers to selectively transfer text between files.
You can also delete text into named buffers, using much the same procedure. For
example:
"a5dd delete five lines into buffer a
If you specify the buffer name with a capital latter, yanked or deleted text will be
appended to the current contents of the buffer. For example:
"bYY yank current line into buffer h
"B5dd delete five lines and append to buffer h
3) move down three paragraphs
"bP insert the six lines from buffer b above the cursor
When you put text from a named buffer, a copy still remains in that buffer; you can
repeat the put as often as you like until you quit your editing session or replace the text
in the buffer.
For example, suppose you were preparing a document with some repetitive ele-
ments, such as the skeleton for each page of the reference section in a manual. You
could store the skeleton in a named buffer, put it into your file, fill in the blanks, then
put the skeleton in again each time you need it.
 0 Learning v i 0 57

Marking Your Place

During a v i session, you can mark your place in the file with an invisible “book-
mark,” perform edits elsewhere, then return to your marked place. In the command
mode:
I’ mu marks current position with x ( x can be any letter)
‘I ‘x moves cursor to beginning of line marked by x
‘I’X
moves cursor to character marked by x
8,. . returns to previous mark or context after a move

Sincerely,
putting together a mx G
written plan that mark and move
to end of file
Fred C a s l o n

Sincerely, Today, I’ll s t a r t

putting together a
written plan that

Ered Caslon

Place markers are set only during the current v i session; they are not stored in the file.

Other Advanced Edits

You may wonder why we haven’t discussed global changes, moving text between files,
or other advanced ex topics. The reason is that, to use these tools, it helps to learn
more about ex and a set of UNIX pattern-matching tools that we discuss together in
Chapter 7.
 C H A P T E R

nrof f and t r o f f

The v i editor lets you edit text, but it is not much good at formatting. A text file such
as program source code might be formatted with a simple program like pr, which
inserts a header at the top of every page and handles pagination, but otherwise prints the
document exactly as it appears in the file. But for any application requiring the
preparation of neatly formatted text, you will use the n r o f f (“en-roff”) or t r o f f
(“tee-roff”) formatting program.
These programs are used to process an input text file, usually coded or “marked
up” with formatting instructions. When you use a wysiwyg program like most word
processors, you use commands to lay out the text on the screen as it will be laid out on
the page. With a markup language like that used by n ro f f and t r o f f , you enter
commands into the text that tell the formatting program what to do.
Our purpose in this chapter is twofold. We want to introduce the basic formatting
codes that you will find useful. But at the same time, we want to present them in the
context of what the formatter is doing and how it works. If you find this chapter
rough-going-especially if this is your first exposure to n r o f f /t rof f-skip ahead
to either Chapter 5 or Chapter 6 and become familiar with one of the macro packages,
m s or mm; then come back and resume this chapter. We assume that you are reading
this book because you would like more than the basics, that you intend to master the
complexities of nrof f / t r o f f . A s a result, this chapter is somewhat longer and
more complex than it would be if the book were an introductory user’s guide.

Conventions
To distinguish input text and requests shown in examples from formatter output,
we have adopted the convention of showing “page comers” around output from
n r o f f or t r o f f . Output from n r o f f is shown in the same constant-width
typeface as other examples:

Here is an example of nroff output.

58
 0 n r o f f and t r o f f 0 59

Output from t r o f f is shown in the same typeface as the text, but with the size of the
type reduced by one point, unless the example calls for an explicit type size:

I I
I Here is an example of troff output. I
In representing output, compromises sometimes had to be made. For example, when
showing nrof f output, we have processed the example separately with n r o f f, and
read the results back into the source file. However, from there, they have been typeset
in a constant-width font by t r o f f. As a result, there might be slight differences from
true nroff output, particularly in line length or page size. However, the context
should always make clear just what is being demonstrated.

What the Formatter Does

Take a moment to think about the things you do when you format a page on a wysiwyg
device such as a typewriter:

You set aside part of the page as the text area. This requires setting top, bot-
tom, left, and right margins.
You adjust the lines that you type so they are all approximately the same
length and fit into the designated text area.
You break the text into syntactic units such as paragraphs.
You switch to a new page when you reach the bottom of the text area.

Left to themselves, nroff or t r o f f will do only one of these tasks: they will
adjust the length of the lines in the input file so that they come out even in the output
file. To do so, they make two assumptions:

They assume that the line length is 6.5 inches.

They assume that a blank line in the input signals the start of a new paragraph.
The last line of the preceding text is not adjusted, and a blank line is placed in
the output.

The process of filling and adjusting is intuitively obvious-we’ve all done much the
same thing manually when using a typewriter or had it done for us by a wysiwyg word
processor. However, especially when it comes to a typesetting program like t r o f f ,
there are ramifications to the process of line adjustment that are not obvious. Having a
clear idea of what is going on will be very useful later. For this reason, we’ll examine
the process in detail.
60 0 UNlX Text Processing

Line Adjustment
There are three parts to line adjustment: filling, justification, and hyphenation. Filling
is the process of making all lines of text approximately equal in length. When working
on a typewriter, you do this automatically, simply by typing a camage return when the
line is full. Most word-processing programs automatically insert a carriage return at the
end of a line, and we have seen how to set up v i to do so as well.
However, n r o f f and t r o f f ignore carriage returns in the input except in a
special “no fill” mode. They reformat the input text, collecting all input lines into
even-length output lines, stopping only when they reach a blank line or (as we shall see
shortly) a formatting instruction that tells them to stop. Lines that begin with one or
more blank spaces are not filled, but trailing blank spaces are trimmed. Extra blank
spaces between words on the input line are preserved, and the formatter adds an extra
blank space after each period, question mark, or exclamation point.
Justification is a closely related feature that should not be confused with filling.
Filling simply tries to keep lines approximately the same length; justification adjusts the
space between words so that the ends of the lines match exactly.
By default, n r o f f and t r o f f both fill and justify text. Justification implies
filling, but it is possible to have filling without justification. Let’s look at some exam-
ples. First, we’ll look at a paragraph entered in v i . Here’s a paragraph from the letter
you entered in the last chapter, modified so that it offers to prepare not just a user’s
guide for the Alcuin illuminated lettering software, but a reference manual as well. In
the course of making the changes, we’ve left a short line in the middle of the paragraph.
In our conversation last Thursday, we discussed a
documentation project that would produce a user’s guide
and reference manual
for the Alcuin product. Yesterday, I received t h e product
demo and other materials that you sent me.

Now, let’s look at the paragraph after processing by n r o f f:

In our conversation last Thursday, we discussed a

documentation project that would produce a user‘s
guide and reference manual for t h e Alcuin product.
Yesterday, I received t h e product demo and other
materials that you sent me.
The paragraph has been both filled and justified. If the formatter were told to fill, but
not to justify, the paragraph would look like this:

In our conversation last Thursday, we discussed a

documentation project that would produce a user‘s guide
and reference manual for t h e Alcuin product. Yesterday,
I received t h e product demo and other materials that
you sent me.
 0 nroff andtroff 0 61

As you can see, n r o f f justified the text in the first example by adding extra space
between words.
Most typewritten material is filled but not justified. In printer’s terms, it is typed
rugged right. Books, magazines, and other typeset materials, by contrast, are usually
right justified. Occasionally, you will see printed material (such as ad copy) in which
the right end of each line is justified, but the left end is ragged. It is for this reason that
we usually say that text is right or left justifzed, rather than simply justified.
When it is difficult to perform filling or justification or both because a long word
falls at the end of a line, the formatter has another trick to fall back on (one we are all
familiar with)-hyphenation.
The n r o f f and t r o f f programs perform filling, justification, and hyphena-
tion in much the same way as a human typesetter used to set cold lead type. Human
typesetters used to assemble a line of type by placing individual letters in a tray until
each line was filled. There were several options for filling as the typesetter reached the
end of the line:

The next word might fit exactly.

The next word might fit if the typesetter squeezed the words a little closer
together.
The next word could be hyphenated, with part put on the current line and part
on the next line.

If, in addition to being filled, the text was to be justified, there was one additional issue:
after the line was approximately the right length, space needed to be added between
each word so that the line length came out even.
Just like the human typesetter they replace, n r o f f and t r o f f assemble one
line of text at a time, measuring the length of the line and making adjustments to the
spacing to make the line come out even (assuming that the line is to be justified). Input
lines are collected into a temporary storage area, or hufSeer, until enough text has been
collected for a single output line. Then that line is output, and the next line collected.
It is in the process of justification that you see the first significant difference
between the two programs. The n r o f f program was designed for use with
typewriter-like printers; t r o f f was designed for use with phototypesetters.
A typewriter-style printer has characters all of the same size-an i takes up the
same amount of space as an m. (Typical widths are 1/10 or 1/12 inch per character.)
And although some printers (such as daisywheel printers) allow you to change the style
of type by changing the daisywheel or thimble, you can usually have only one typeface
at a time.
A typesetter, by contrast, uses typefaces in which each letter takes up an amount
of space proportional to its outline. The space allotted for an i is quite definitely nar-
rower than the space allotted for an m. The use of variable-width characters makes the
job of filling and justification much more difficult for t r o f f than for n r o f f .
Where n r o f f only needs to count characters, t r o f f has to add up the width of
each character as it assembles the line. (Character widths are defined by a “box”
around the character, rather than by its natural, somewhat irregular shape.)
62 0 UNlX Text Processing 0

The t r o f f program also justifies by adding space between words, but because
the variable-width fonts it uses are much more compact, it fits more on a line and gen-
erally does a much better job of justification.*
There’s another difference as well. Left to itself, n r o f f will insert only full
spaces between words-that is, it might put two spaces between one pair of words, and
three between another, to fill the line. If you call n r o f f with the -e option, it will
attempt to make all interword spaces the same size (using fractional spaces if possible).
But even then, nrof f will only succeed if the output device allows fractional spacing.
The t r o f f program always uses even interword spacing.
Here’s the same paragraph filled and justified by t r o f f:

In our conversation last Thursday, we discussed a documentation project that would

produce a user’s guide and reference manual for the Alcuin product. Yesterday, I
received the product demo and other materials that you sent me.
To make matters still more difficult, typeset characters come in a variety of dif-
ferent designs, orfonts. A font is a set of alphabetic, numeric, and punctuation charac-
ters that share certain design elements. Typically, fonts come in families of several
related typefaces. For example, this book is typeset for the most part in the Times
Roman family of typefaces. There are three separate fonts:
roman
bold
italic
Typesetting allows for the use of multiple fonts on the same page, as you can see from
the mixture of fonts throughout this book. Sometimes the fonts are from the same fam-
ily, as with the Times Roman, Times Bold, and Times Italic just shown. However, you
can see other fonts, such as Helvetica, in the running headers on each page. Bold and
italic fonts are generally used for emphasis; in computer books such as this, a constant-
width typewriter font is used for examples and other “computer voice” statements.
Even within the same font family, the width of the same character varies from
font to font. For example, a bold “m” is slightly wider than a Roman “m.”
To make things still more complicated, the same font comes in different sizes. If
you look at this book, you will notice that the section headings within each chapter are
slightly larger for emphasis. Type sizes are measured in units called points. We’ll talk
more about this later, but to get a rough idea of what type sizes mean, simply look at
the current page. The body type of the book is 10-point Times Roman: the next head-
ing is 12-point Times Bold. The spacing between lines is generally proportional to the
point size, instead of fixed, as it i s with n r o f f .

*The very best typesetting programs have the capability to adjust the space between individual characters as
well. This process i s called kerning. SoftQuad Publishing Software in Toronto sells an enhanced version of
t r o f f called SQroff that does support kerning.
 0 nroff and troff 0 63

The t r o f f program gets information about the widths of the various characters
in each font from tables stored on the system in the directory / u s r / l i b / f o n t .
These tables tell t r o f f how far to move over after it has output each character on the
line.
We’ll talk more about t r o f f later. For the moment, you should be aware that
the job of the formatting program is much more complicated when typesetting than it is
when preparing text for typewriter-style printers.

Using n r o f f
As mentioned previously, left to themselves, n r o f f and t r o f f perform only rudi-
mentary formatting. They will fill and justify the text, using a default line length of 6.5
inches, but they leave no margins, other than the implicit right margin caused by the
line length. To make this clearer, let’s look at the sample letter from the last chapter
(including the edit we made in this chapter) as it appears after formatting with n r o f f .
First, let’s look at how to invoke the formatter. The n r o f f program takes as an
argument the name of i? file to be formatted:
$ nroff l e t t e r

Alternatively, it can take standard input, allowing you to preprocess the text with some
other program before formatting it:
$ t b l report I nroff
There are numerous options to n r o f f . They are described at various points in this
book (as appropriate to the topic) and summarized in Appendix B.
One basic option is -T, which specifies the terminal (printer) type for which out-
put should be prepared. Although n r o f f output is fairly straightforward, some differ-
ences between printers can significantly affect the output. (For example, one printer
may perform underlining by backspacing and printing an underscore under each under-
lined letter, and another may do it by suppressing a newline and printing the under-
scores in a second pass over the line.) The default device is the Teletype Model 37
terminal-a fairly obsolete device. Other devices are listed in Appendix B. If you
don’t recognize any of the printers or terminals, the safest type is probably lp:
$ n r o f f -Tlp file

In examples in this book, we will leave off the -T option, but you may want to experi-
ment, and use whichever type gives the best results with your equipment.
Like most UNIX programs, n r o f f prints its results on standard output. So,
assuming that the text is stored in a file called letter, all you need to do is type:
$ nroff letter

A few moments later, you should see the results on the screen. Because the letter will
scroll by quickly, you should pipe the output of n r o f f to a paging program such as
pgor more:
64 0 UNlX Text Processing 0

$ nroff l e t t e r I pg
or out to a printer using 1 p or l p r :
$ nroff l e t t e r I lp

Usingtroff
The chief advantage of t r o f f over nrof f is that it allows different types of charac-
ter sets, or fonts, and so lets you take full advantage of the higher-quality printing avail-
able with typesetters and laser printers. There are a number of requests, useful only in
t r o f f , for specifying fonts, type sizes, and the vertical spacing between lines. Before
we describe the actual requests though, we need to look at a bit of history.
The t r o f f program was originally designed for a specific typesetter, the Wang
C/A/T. Later, it was modified to work with a wide range of output devices. We’ll dis-
cuss the original version of t r o f f (which is still in use at many sites) first, before
discussing the newer versions. The C/A/T typesetter was designed in such a way that it
could use only four fonts at one time.
(Early phototypesetters worked by projecting light through a film containing the
outline of the various characters. The film was often mounted on a wheel that rotated
to position the desired character in front of the light source as it flashed, thus photo-
graphing the character onto photographic paper or negative film. Lenses enlarged and
reduced the characters to produce various type sizes. The C/A/T typesetter had a wheel
divided into four quadrants, onto which one could mount four different typefaces.)
Typically, the four fonts were the standard (roman), bold, and italic fonts of the
same family, plus a “special” font that contained additional punctuation characters,
Greek characters (for equations), bullets, rules, and other nonstandard characters. Fig-
ure 4-1 shows the characters available in these standard fonts.

The Coming of d i t r o f f
Later, t rof f was modified to support other typesetters and, more importantly (at least
from the perspective of many readers of this book), laser printers. The later version of
t r o f f is often called d i t r o f f (for device-independent t r o f f ) , but many UNIX
systems have changed the name of the original t r o f f to o t r o f f and simply call
d i t r o f f by theoriginalname, t r o f f .
The d i t r o f f program has not been universally available because, when it was
developed, it was “unbundled” from the basic UNIX distribution and made part of a
separate product called Documenter’s Workbench or DWB. UNIX system manufactur-
ers have the option not to include this package, although increasingly, they have been
doing so. Versions of DWB are also available separately from third party vendors.
The newer version of t ro f f allows you to specify any number of different
fonts. (You can mount fonts at up to ten imaginary “positions” with .fp and can
request additional fonts by name).
 0 n r o f f and t r o f f 0 65

Times Roman

abcdefghijklmnopqrstuv wxyz
ABCDEFGHIJKLMNOPQRSTUVWXYZ
1234567890
! $ % & ( ) ' ' * + - . , I :; = ? [ ] I
- - - '14 '12 3/4 fi fl "t't 80
Times Italic

abcdefghijklrnnopqrstuvwxyz
ABCDEFGHIJKLMNOPQRSTUVWXYZ
1234567890
!$%&a()" * + - . , I : ;=
- - - '14 ' / z 3 / 4 f i f E " f 't.8 0

Times Bold

abcdefghijklmnopqrstuvwxyz
ABCDEFGHIJKLMNOPQRSTUVWXYZ
1234567890
!$%&()"*+-.,/:;=?[]I
0 - - - '14 I12 3/4 fi fl "t'C80
Special Mathematical Font

Fig. 4-2. The Four Standard Fonts

There may also be different font sizes available, and there are some additional corn-
rnands for line drawing ( d i t r o f f can draw curves as well as straight lines). For the
most part, though, d i t r o f f is very similar to the original program, except in the
greater flexibility it offers to use different output devices.
One way to find out which version of t r o f f you have on your system (unless
you have a program explicitly called d i t r o f f ) is to list the contents of the directory
/usr/lib/font:
66 0 UNlX Text Processing 0

$18 -F /usr/lib/font
d e v l j/
devps/
ftB
ftI
ftR
ftS

If there are one or more subdirectories whose name begins with the letters dev, your
system is using d i t r o f f . Our system supports both d i t r o f f and o t r o f f , so
we have both a device subdirectory (for d i t r o f f ) and font files (for o t r o f f )
directly in / u s r / l i b / f o n t .
We’ll talk more about font files later. For the moment, all you need to know is
that they contain information about the widths of the characters in various fonts for a
specific output device.
Contrary to what a novice might expect, font files do not contain outlines of the
characters themselves. For a proper typesetter, character outlines reside in the typesetter
itself. All t r o f f sends out to the typesetter are character codes and size and position
information.
However, t r o f f has increasingly come to be used with laser printers, many of
which use downloadable fonts. An electronic image of each character is loaded from
the computer into the printer’s memory, typically at the start of each printing job.
There may be additional “font files” containing character outlines in this case, but
these files are used by the software that controls the printer, and have nothing to do
with t r o f f itself. In other cases, font images are stored in ROM (read-only memory)
in the printer.
If you are using a laser printer, it is important to remember that t r o f f itself has
nothing to do with the actual drawing of characters or images on the printed page. In a
case like this, t r o f f simply formats the page, using tables describing the widths of
the characters used by the printer, and generates instructions about page layout, spacing,
and so on. The actual job of driving the printer is handled by another program, gen-
erally referred to as a printer driver or t r o f f postprocessor.
To use t r o f f with such a postprocessor, you will generally need to pipe the
output of t r o f f to the postprocessor and from there to the print spooler:
. t r o f f file
$ I postprocessor I lp
If you are using the old version of t r o f f , which expects to send its output directly to
the C/mtypesetter, you need to specify the -t option, which tells t r o f f to use
standard output. If you don’t, you will get the message:
Typesetter busy.

(Of course, if by any chance you are connected to a C/A/T typesetter, you don’t need
this option. There are several other options listed in Appendix B that you may find use-
ful.) When you use d i t r o f f , on the other hand, you will need to specify the -T
command-line option that tells it what device you are using. The postprocessor will
then translate the device-independent t r o f f output into instructions for that particular
type of laser printer or typesetter. For example, at our site, we use t r o f f with an
 0 nroff and t r o f f 0 67

Apple Laserwriter and Pipeline Associates’ devps postprocessor, which translates

t r o f f output for the Laserwriter. Our command line looks something like this:
$ ditroff -Tps files I devps I lp

You can print the same file on different devices, simply by changing the -T option and
the postprocessor. For example, you can print drafts on a laser printer, then switch to a
typesetter for final output without making extensive changes to your files. (To actually
direct output to different printers, you will also have to specify a printer name as an
option to the lp command. In our generic example, we simply use lp without a n y
options, assuming that the appropriate printer is connected as the default printer.)
Like all things in life, this is not always as easy as it sounds. Because the fonts
used by different output devices have different widths even when the nominal font
names and sizes are the same, pagination and line breaks may be different when you
switch from one device to another.
The job of interfacing d i t r o f f to a wide variety of output devices i s becoming
easier because of the recent development of industry-wide page description languages
like Adobe Systems’ PostScript, Xerox’s Interpress, and Imagen’s DDL. These page
description languages reside in the printer, not the host computer, and provide a device-
independent way of describing placement of characters and graphics on the page.
Rather than using a separate postprocessor for each output device, you can now
simply use a postprocessor to convert t r o f f output to the desired page description
language. For example, you can use Adobe Systems’ Transcript postprocessor (or an
equivalent postprocessor like devps from Pipeline Associates) to convert t r o f f
output to PostScript, and can then send the PostScript output to any one of a number of
typesetters or laser printers.
From this point, whenever we say t r o f f , we are generally referring to
d i t r o f f . In addition, although we will continue to discuss n r o f f as it differs from
t r o f f , our emphasis is on the more capable program. It is our opinion that the grow-
ing availability of laser printers will make t r o f f the program of choice for almost all
users in the not too distant future.
However, you can submit a document coded for t r o f f to n r o f f with entirely
reasonable results. For the most part, formatting requests that cannot be handled by
n r o f f are simply ignored. And you can submit documents coded for n r o f f to
t r o f f , though you will then be failing to use many of the characteristics that make
t r o f f desirable.

The Markup Language

The n r o f f and t r o f f markup commands (often called requests) typically consist
of one or two lowercase letters and stand on their own line, following a period or apos-
trophe in column one. Most requests are reasonably mnemonic. For example, the
request to leave space is:
- SP
There are also requests that can be embedded anywhere in the text. These requests are
commonly called escape sequences. Escape sequences usually begin with a backslash
68 0 UNlX Text Processing 0

(\) . For example, the escape sequence \ 1 will draw a horizontal line. Especially in
t r o f f, escape sequences are used for line drawing or for printing various special char-
acters that do not appear in the standard ASCII character set. For instance, you enter
\ ( b u to get 0 , a bullet.
There are three classes of formatting instructions:

Instructions that have an immediate one-time effect, such as a request to space

down an inch before outputting the next line of text.
Instructions that have a persistent effect, such as requests to set the line length
or to enable or disable justification.

= Instructions that are useful for writing macros. There is a “programming

language” built into the formatter that allows you to build up complex requests
from sequences of simpler ones. As part of this language there are requests for
storing values into variables called strings and number registers, for testing
conditions and acting on the result, and so on.

For the most part, we will discuss the requests used to define macros, strings, and
number registers later in this book.
At this point, we want to focus on understanding the basic requests that control
the basic actions of the formatter. We will also learn many of the most useful requests
with immediate, one-time effects. Table 4-1 summarizes the requests that you will use
most often.

TABLE 4-1. Basic n r o f f k r o f f Requests

Request Meaning Request Meaning

.ad Enable line adjustment .n a No justification of lines
- br Line break .ne Need lines to end of page
- bP Page break .n f No filling of lines
.ce Center next line .n r Define and set number register
.de Define macro - PO Set page offset
- ds Define string - PS Set point size
.fi Fill output lines .so Switch to source file and return
.ft Set current font - SP Space
.in Set indent .t a Set tab stop positions
.Is Set double or triple spacing .ti Set temporary indent
-11 Specify line length .vs Set vertical line spacing

Looking at nrof f Output

When we discussed the basic operations of the text formatter, we saw that n r o f f and
t r o f f perform rudimentary formatting. They will fill and justify the text, using a
 0 nroff and troff 0 69

default line length of 6.5 inches, but they leave no margins, other than the implicit right
margin caused by the line length.
To make this clearer, let’s look at the sample letter from the last chapter as it
appears after formatting with n r o f f, without any embedded requests, and without
using any macro package. From Figure 4-2, you can see immediately that the formatter
has adjusted all of the lines, so that they are all the same length-ven in the address
block of the letter, where we would have preferred them to be left as they were. Blank
lines in the input produce blank lines in the output, and the partial lines at the ends of
paragraphs are not adjusted.
The most noticeable aspect of the raw formatting is a little difficult to reproduce
here, though we’ve tried. No top or left margin is automatically allocated by nrof f.

Turning Filling On and Off

Even though filling of uneven text lines resulting from editing is probably the most
basic action we want from the formatter, it is not always desirable. For example, in our
letter, we don’t want the address block to be filled. There are two requests we could
use to correct the problem: .b r (break) and .n f (no f i l l ) .
A .b r request following a line outputs the current contents of the line buffer and
starts the next line, even though the buffer is not yet full. To produce a properly for-
matted address block, we could enter the following requests in the file:
Mr. John F u s t
.br
Vice President, Research and Development
.br
Gutenberg Galaxy Software
.br
Waltham, Massachusetts 02159
Each individual input line will be output without filling or justification. We could also
use the .n f request, which tells n r o f f to stop filling altogether. Text following this
request will be printed by the formatter exactly as it appears in the input file. Use this
request when you want text to be laid out as it was typed in.
Because we do want the body of the letter to be filled, we must turn filling back
on with the .f i fill) request:
April 1, 1987
- nf
Mr. John Fust
Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02159
.fi
Dear Mr. Fust:
70 0 UNlX Text Processing 0

April 1, 1987

r - John Fust Vice President, Research and

evelopment Gutenberg Galaxy Software Waltham,
assachusetts 02159

ear Mr. Fust:

n our conversation last Thursday, we discussed a

ocumentation project that would produce a user's
uide and reference manual for the Alcuin product.
esterday, I received the product demo and other
aterials that you sent me. After studying them,
want to clarify a couple of points:

oing through a demo session gave me a much better

nderstanding of the product. I confess to being
mazed by Alcuin. Some people around here,
ooking over my shoulder, were also astounded by
he illustrated manuscript I produced with Alcuin.
ne person, a student of calligraphy, was really
mpre ssed.

omorrow, I'll start putting together a written

lan that presents different strategies for
ocumenting the Alcuin product. After I submit
his plan, and you have had time to review it,
et's arrange a meeting at your company to discuss
hese stratgies.

hanks again for giving us the opportunity to bid

n this documentation project. I hope we can
ecide upon a strategy and get started as soon as
ossible in order to have the manual ready in time
or first customer ship. I look forward to meeting
ith you towards the end of next week.

Sincerely,

Fred Caslon

Fig. 4-2. A Raw nr o ff-formatted File

 0 nroffandtroff 0 71

If you look carefully at the previous example, you will probably notice that we entered
the two formatting requests on blank lines in the letter. If we were to format the letter
now, here is what we?d get:

April 1, 1987
Mr. John Fust
Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02159
Dear Mr. Fust:

As you may notice, we?ve lost the blank lines that used to separate the date from the
address block, and the address block from the salutation. Lines containing formatting
requests do not result in any space being output (unless they are spacing requests), so
you should be sure not to inadvertently replace blank lines when entering formatting
codes.

Controlling Justification
Justification can be controlled separately from filling by the .a d (adjust) request.
(However, filling must be on for justification to work at all.) You can adjust text at
either margin or at both margins.
Unlike the . b r and .nf requests introduced, . a d takes an argument, which
specifies the type of justification you want:

1 adjust left margin only

r adjust right margin only
b adjust both margins
C center filled line between margins

There is another related request, .n a (no adjust). Because the text entered in a
file is usually left justified to begin with, turning justification off entirely with - n a
produces similar results to .ad 1 in most cases.
However, there is an important difference. Normally, if no argument is given to
the .a d request, both margins will be adjusted. That is, .a d is the same as .a d b.
However, following an .na request, .a d reverts to the value last specified. That is,
the sequence:
.ad r
Some text
.ad 1
Some text
-
ad
Some texr
will adjust both margins in the third block of text. However, the sequence:
72 0 UNlX Text Processing 0

.ad r
Some text
.na
Some text
.ad
Some text
will adjust only the right margin in the third block of text.
It’s easy to see where you would use .ad b or .ad 1. Let’s suppose that
you would like a ragged margin for the body of your letter, to make it look more like it
was prepared on a typewriter. Simply follow the .f i request we entered previously
with . a d 1.
Right-only justification may seem a little harder to find a use for. Occasionally,
you’ve probably seen ragged-left copy in advertising, but that’s about it. However, if
you think for a moment, you’ll realize that it is also a good way to get a single line over
to the right margin.
For example, in our sample letter, instead of typing all those leading spaces before
the date (and having it fail to come out flush with the margin anyway), we could enter
the lines:
.ad r
April 1, 1987
.ad b

As it turns out, this construct won’t quite work. If you remember, when filling is
enabled, nro f f and t r o f f collect input in a one-line buffer and only output the
saved text when the line has been filled. There are some non-obvious consequences of
this that will ripple all through your use of nrof f and t r o f f . If you issue a
request that temporarily sets a formatting condition, then reset it before the line is out-
put, your original setting may have no effect. The result will he controlled by the
request that is in effect ut the time the line is output, not ut the time that it i s first col-
lected in the line buffer.
Certain requests cause implicit line breaks (the equivalent of carriage returns on a
typewriter) in the output, but others do not. The .ad request does not cause a break.
Therefore, a construction like:
.ad r
April 1, 1987
.ad b
Mr. John F u s t

will result in the following output:

I April 1, 1987 Mr. John F u s t

and not:

I Mr. John F u s t -
April 1, 1987
 0 nroff and t r o f f 0 73

To make sure that you get the desired result from a temporary setting like this, be sure
to follow the line to be affected with a condition that will cause a break.” For instance,
in the previous example, you would probably follow the date with a blank line or an
.sp request, either of which will normally cause a break. If you don’t, you should put
in an explicit break, as follows:
.ad r
April 1, 1 9 8 7
br-
.ad b
Mr. John F u s t

A final point about justification: the formatter adjusts a line by widening the blank
space between words. If you do not want the space between two words adjusted or split
across output lines, precede the space with a backslash. This is called an unpaddable
space.
There are many obscure applications for unpaddable spaces; we wil1 mention
them as appropriate. Here’s a simple one that may come in handy: n r o f f and
t r o f f normally add two blank spaces after a period, question mark, or exclamation
point. The formatter can’t distinguish between the end of a sentence and an abbrevia-
tion, so if you find the extra spacing unaesthetic, you might follow an abbreviation like
Mr.with an unpaddable space: M r . \ J o h n F u s t .

Hyphenation
As pointed out previously, hyphenation is closely related to filling and justification, in
that it gives n r o f f and t r o f f some additional power to produce filled and justified
lines without large gaps.
The n r o f f and t K O f f programs perform hyphenation according to a general
set of rules. Occasionally, you need to control the hyphenation of particular words.
You can specify either that a word not be hyphenated or that it be hyphenated in a cer-
tain way. You can also turn hyphenation off entirely.

Specifying Hyphenation for Individual Words

There are two ways to specify that a word be hyphenated a specific way: with the
.hw request and with the special hyphenation indicator \%.
The .hw (hyphenate word) request allows you to specify a small list of words
that should be hyphenated a specific way. The space available for the word list is small
(about 128 characters), so you should use this request only for words you use fre-
quently, and that n r o f f and t r o f f hyphenate badly.

*The following requests cause a break:

.bp .br .ce .fi .nf .sp .in .ti
All other requests can be interspersed with text without causing a break. In addition, as discussed later,
even these requests can be introduced with a special “no break” control character ( ’ instead of .) so that
they too will not cause a break.
74 UNlX Text Processing 0

.
To use hw, simply specify the word or words that constitute the exception list,
typing a hyphen at the point or points in the word where you would like it to be
hyphenated:
.hw hy-phen-a-tion

You can specify multiple words with one - hw request, or you can issue multiple .hw
requests as you need them.
However, if it is just a matter of making sure that a particular instance of a word
is hyphenated the way you want, you can use the hyphenation indication character
sequence \ % . As you type the word in your text, simply type the two characters \ % at
each acceptable hyphenation point, or at the front of the word if you don’t want the
word to be hyphenated at all:
\%acknowledge the word acknowledge will not be hyphenated
ac\%know\%ledge the word acknowledge can be hyphenated only
at the specified points
This character sequence is the first instance we have seen of a formatting request that
does not consist of a request name following a period in column one. We will see
many more of these later. This sequence is embedded right in the text but does not
print out.
In general, nrof f and t r o f f do a reasonable job with hyphenation. You will
need to set specific hyphenation points only in rare instances. In genera1, you shouldn’t
even worry about hyphenation points, unless you notice a bad break. Then use either
.hw or \ % to correct it.
The UNIX hyphen command can be used to print out all of the hyphenation
points in a file formatted with n r o f f or trof f -a.
$ n r o f f optionsfiles I hyphen

or:
$ t r o f f options -a files I hyphen
If your system doesn’t have the hyphen command, you can use grep instead:
$ nroff oprionsjiles I grep ’ -$’
(The single quotation marks are important because they keep g r e p from interpreting
the - as the beginning of an option.)

Turning Hyphenation Off and On

If you don’t want any hyphenation, use the .nh ( n o hyphenation) request. Even if you
do this, though, you should be aware that words already containing embedded hyphens,
em dashes (-), or hyphen indication characters ( \ % ) will still be subject to hyphena-
tion.
After you’ve turned hyphenation off, you can turn it back on with the .hy
(hyphenate) request. This request has a few twists. Not only does it allow you to turn
hyphenation on, it also allows you to adjust the hyphenation rules that n r o f f and
t r o f f use. It takes the following numeric arguments:
 0 n r o f f and t r o f f 0 75

0 turn hyphenation off

1 turn hyphenation on
2 do not hyphenate the last line on a page
4 do not hyphenate after the first two characters of a word
8 do not hyphenate before the last two characters of a word

Specifying .h y with no argument is the same as specifying - hy 1. The other

numeric values are additive. For example, .h y 1 2 ( - h y 4 plus .hy 8 ) will keep
n r o f f and t r o f f from breaking short syllables at the beginning or end of words,
and .h y 1 4 will put all three hyphenation restrictions into effect.

PageLayout
Apart from the adjusted address block, the biggest formatting drawback that you prob-
ably noticed when we formatted the sample letter is that there was no left or top margin.
Furthermore, though it is not apparent from our one-page example, there is no bottom
margin either. If there were enough text in the input file to run onto a second page, you
would see that the text ran continuously across the page boundary.
In normal use, these layout problems would be handled automatically by either
the m s or mm macro packages (described later). Here, though, we want to understand
how the formatter itself works.
Let’s continue our investigation of the n r o f f and t r o f f markup language
with some basic page layout commands. These commands allow you to affect the
placement of text on the page. Some of them (those whose descriptions begin with the
word s e t ) specify conditions that will remain in effect until they are explicitly changed
by another instance of the same request. Others have a one-time effect.
As shown in Table 4-2, there are two groups of page layout commands, those that
affect horizontal placement of text on the page and those that affect vertical placement.
A moment’s glance at these requests wiIl tell you that, before anything else, we need to
talk about units.

TABLE 4-2. Layout Commands

.1 1 n Set the line length to n

.PO n Set the left margin (page offset) to n
Horizontal Layout . in n Indent the left margin to n
.t i n Temporarily indent the left margin to n
.c e n Center the following n lines
.pl n Set the page length to n
.sp n Insert n spaces
Vertical Layout .b p n Start a new page
.w h n Specify when (at what vertical position
on the page) to execute a command
76 0 UNlX Text Processing 0

Units of Measure
By default, most n r o f f and t r o f f commands that measure vertical distance (such
as sp) do so in terms of a number of ‘‘lines’’ (also referred to as vertical spaces, or
vs). The n r o f f program has constant, device-dependent line spacing; t r o f f has
variable line spacing, which is generally proportional to the point size. However, both
programs do allow you to use a variety of other units as well. You can specify spacing
in terms of inches and centimeters, as well as the standard printer’s measures picas and
points. (A pica is 1/6 of an inch; a point is about 1/72 of an inch. These units were
originally developed to measure the size of type, and the relationship between these two
units is not as arbitrary as it might seem. A standard 12-point type is 1 pica high.)
Horizontal measures, such as the depth of an indent, can also be specified using
any of these measures, as well as the printer’s measures ems and ens. These are relative
measures, originally based on the size of the letters m and n in the current type size and
typeface. By default, horizontal measures are always taken to be in ems.
There is also a relationship between these units and points and picas. An em is
always equivalent in width to the height of the character specified by the point size. In
other words, an em in a 12-point type is 12 points wide. An en is always half the size
of an em, or half of the current point size. The advantage of using these units is that
they are relative to the size of the type being used. This is unimportant in n r o f f,
but using these units in t r o f f gives increased flexiblility to change the appearance of
the document without recoding.
The n r o f f and t r o f f programs measure not in any of these units, but in
device-dependent basic units. Any measures you specify are converted to basic units
before they are used. Typically, n r o f f measures in horizontal units of 1/240 of an
inch and o t r o f f uses a unit of 1/432 inch. These units too are not as arbitrary as
they may seem. According to Joseph Osanna’s NrofSITroff User’s Manual-the origi-
nal, dense, and authoritative documentation on t r o f f published by AT&T as part of
the UNlX Programmer’s Manual-the n r o f f units were chosen as “the least com-
mon multiple of the horizontal and vertical resolutions of various typewriter-like output
devices.” The units for o t r o f f were based on the C/A/T typesetter (the device for
which t r o f f was originally designed), which could move in horizontal increments of
1/432 of an inch and in vertical increments of exactly one-third that, or 1/144 inch.
Units for d i t r o f f depend on the resolution of the output device. For example, units
for a 300 dot-per-inch (dpi) laser printer will be 1/300 of an inch in either a vertical or a
horizontal direction. See Appendix D for more information on d i t r o f f device
units.
You don’t need to remember the details of all these measures now. You can gen-
erally use the units that are most familiar to you, and we’ll come back to the others
when we need them.
To specify units, you simply need to add the appropriate scale indicator from
Table 4-3 to the numeric value you supply to a formatting request. For example, to
space down 3 inches rather than 3 lines, enter the request:
.sp 3i

The numeric part of any scale indicator can include decimal fractions. Before the speci-
fied value is used, nro f f and t r o f f will round the value to the nearest number of
device units.
 nroff and t r o f f 77

TABLE 4-3. Units of Measure

Indicator Units
C Centimeters
i Inches
m Ems
n Ens
P Points
P Picas
U Device Units
V Vertical spaces (lines)
none Default

In fact, you can use any reasonable numeric expression with any request that
expects a numeric argument. However, when using arithmetic expressions, you have to
be careful about what units you specify. All of the horizontally oriented requests-
.
.11, i n , .ti, .t a, .PO, It, and . .
mc-assume you mean ems unless you
specify otherwise.
Vertically oriented requests like .sp assume v’s unless otherwise specified.
The only exceptions to this rule are - p s and .vs, which assume points by default-
but these are not really motion requests anyway.
As a result, i f you make a request like:
- 1 1 7i/2

what you are really requesting is:

- 1 1 7i/2m

The request:
- 1 1 7i/2i

is not what you want either. In performing arithmetic, as with fractions, the formatter
converts scaled values to device units. In o t r o f f , this means the previous expres-
sion is really evaluated as:
-11 (7*432u)/ (2*432u)

I f you really want half o f 7 inches, you should specify the expression like this:
-11 7i/2u

You could easily divide 7 by 2 yourself and simply specify 3 . 5 . The point of this
example is that when you are doing arithmetic-usually with values stored in variables
called number registers (more on these later)-you will need to pay attention to the
interaction between units. Furthermore, because fractional device units are always
rounded down, you should avoid expressions like 7 i / 2 .5u because this is equivalent
to 7i/2u.
78 0 UNlX Text Processing 0

In addition to absolute values, many n r o f f and t r o f f requests allow you to

specify relative values, by adding a + or a - before the value. For example:
. 1 1 -.5i

will subtract '/2 inch from the current line length, whatever it is.

Setting Margins
In n r o f f and t r o f f, margins are set by the combination of the .PO (page ofSset)
and - 1 1 (line length) requests. The .PO request defines the left margin. The .11
request defines how long each line will be after filling, and so implicitly defines the
right margin:

right
PO 11
margin

The n r o f f program's default line length of 6.5 inches i s fairly standard for an 8[/2-
by-1 1 page-it allows for l-inch margins on either side.
Assuming that we'd like 11/4-inch margins on either side of the page, we would
issue the following requests:
-11 6 i
. P O 1.25i

This will give us 1 1 / 4 inches for both the right and left margins. The - P O request
specifies a left margin, or page offset, of 11/4 inches. When the 6-inch line length is
added to this, it will leave a similar margin on the right side of the page.
Let's take a look at how our sample letter will format now. One paragraph of the
output should give you the idea.

I In
discussed
our
a
conversation
documentation
last
project
Thursday,
that w o u
we
ld
p r o d u c e a user's g u i d e and r e f e r e n c e manua for
the Alcuin product. Yesterday, I received t h e
p r o d u c t demo a n d other m a t e r i a l s t h a t y o u s e n t m e .

As we saw earlier, n r o f f assumes a default page offset of 0. Either you or the macro
package you are using must set the page offset. In t r o f f, though, there is a default
page offset of 26/27 inch, so you can get away without setting this value.
(Keep in mind that all n r o f f output examples are actually simulated with
t r o f f , and are reduced to fit on our own 5-inch wide printed page. As a result, the
widths shown in our example output are not exact, but are suggestive of what the actual
result would be on an S1/2-by-l1 inch page.)
 0 nroffandtroff 0 79

Setting Indents
In addition to the basic page offset, or left margin, you may want to set an indent, either
for a single line or an entire block of text. You may also want to center one or more
lines of text.
To do a single-line indent, as is commonly used to introduce a paragraph, use the
. t i (temporary indent) request. For example, if you followed the blank lines between
paragraphs in the sample letter with the request . t i 5, you’d get a result like this
from n r o f f :

7
...Yesterday,I received the product demo and other
materials that you sent me.

Going through a demo session gave me a

much better understanding of the product. I
confess to being amazed by Alcuin ...
The .i n request, by contrast, sets an indent that remains in effect until it is changed.
For example, if you had entered the line . i n 5 between the paragraphs, (instead of
.t i 5), the result would have looked like this:

. . .Yesterday,
I received the product demo and other
materials that you sent me.

Going through a demo session gave me a

much better understanding of the product.
I confess to being amazed by Alcuin ...
All succeeding paragraphs will continue to be indented, until the indent is reset. The
default indent (the value at the left margin) is 0.
These two indent requests can be combined to give a “hanging indent.”
Remember that you can specify negative values to many requests that take numeric
arguments. Here is the first case where this makes sense. Let’s say we would like to
modify the letter so that it numbers the points and indents the body of the numbered
paragraph:
. . .Yesterday,
I received the product demo and other materials
that you sent me. After studying them, I want to clarify
a couple of points:

.in 4
.ti -4
1. Going through a demo session gave me a much better
understanding of the product. I confess to being amazed by
Alcuin - - -
80 0 UNlX Text Processing 0

The first line will start at the margin, and subsequent lines will be indented:

I
...Yesterday, I received t h e product demo and other
materials that you sent me. After studying them,
I want t o clarify a couple of points:

1. Going through a demo session gave m e

better understanding of t h e product.
t o being amazed by Alcuin...
a much
I confess
1
To line up an indented paragraph like this in nrof f,just count the number of charac-
ters you want to space over, then use that number as the size of the indent. But this
trick is not so simple in t r o f f . Because characters, and even spaces, are not of con-
stant width, it is more difficult to create a hanging indent. Ens are a good unit to use
for indents. Like ems, they are relative to the point size, but they are much closer to the
average character width than an em. As a result, they are relatively intuitive to work
with. An indent of 5n is about where you expect a 5-character indent to be from fami-
liarity with a typewriter.

Centering Output Lines

Centering is another useful layout tool. To center the next line, use the .ce request:
.ce
This line will be centered.

Here’s the result:

This line will b e centered.

Centering takes into account any indents that are in effect. That is, if you have used
. i n to specify an indent of 1 inch, and the line length is 5 inches, text will be centered
within the 4-inch span following the indent.
To center multiple lines, specify a number as an argument to the request:
.ce 3
Documentation for t h e Alcuin Product

A Proposal Prepared by
Fred Caslon

Here’s the result:

 I
0 nroff and t r o f f 0 81

Documentation f o r the Alcuin Product

A Proposal Prepared by
Fred Caslon

Notice that .ce centered all three text lines, ignoring the blank line between.
To center an indeterminately large number of lines, specify a very large number
with the .ce request, then turn it off by entering .ce 0:
.ce 1 0 0 0
Many lines of text here.
.ce 0
In looking at the examples, you probably noticed that centering automatically dis-
ables filling and justification. Each line is centered individually. However, there is also
the case in which you would like to center an entire filled and justified paragraph.
(This paragraph style is often used to set off quoted material in a book or paper.) You
can do this by using both the - i n and - 1 1 requests:
I was particularly interested by one comment that I
read in your company literature:

.in +5n
- 1 1 -5n
The development of Alcuin can be traced back to our
founder’s early interest in medieval manuscripts.
He spent several years in the seminary before
becoming interested in computers. After he became
an expert on typesetting software, he resolved to
put his two interests together.
.in -5n
. 1 1 +5n

Here’s the result:

I was particularly interested by one comment that I

read in your company literature:

The development of Alcuin can be traced back to

our founder’s early interest in medieval
manuscripts. He spent several years in the
seminary before becoming interested in comput-
ers. After he became an expert on typesetting
software, he resolved to put his two interests
together.
82 UNlX Text Processing 0

Remember that a line centered with .ce takes into account any indents in effect at the
time. You can visualize the relationship between page offset, line length, indents, and
centering as follows:

in I ce

Setting Tabs
No discussion of how to align text would be complete without a discussion of tabs. A
tab, as anyone who has used a typewriter well knows, is a horizontal motion to a prede-
fined position on the line.
The problem with using tabs in n r o f f and t r o f f is that what you see on the
screen is very different from what you get on the page. Unlike a typewriter or a
wysiwyg word processor, the editor/formatter combination presents you with two dif-
ferent tab settings. You can set tabs in v i , and you can set them in n r o f f and
t r o f f, but the settings are likely to be different, and the results on the screen defin-
itely unaesthetic.
However, after you get used to the fact that tabs will not line up on the screen in
the same way as they will on the printed page, you can use tabs quite effectively.
By default, tab stops are set every .8 inches in n r o f f and every .5 inches in
t r o f f . To set your own tab stops in n r o f f or t r o f f , use the . t a request. For
example:
.ta li 2 . 5 i 3i
will set three tab stops, at 1 inch, 2'/2 inches, and 3 inches, respectively. Any previous
or default settings are now no longer in effect.
You can also set incremental tab stops. The request:
.ta li +1.5i +.5i
will set tabs at the same positions as the previous example. Values preceded with a
plus sign are added to the value of the last tab stop.
You can also specify the alignment of text at a tab stop. Settings made with a
numeric value alone are left adjusted, just as they are on a typewriter. However, by
adding either the letter R or C to the definition of a tab stop, you can make text right
adjusted or centered on the stop.
For example, the following input lines (where a tab character is shown by the
symbol :1)-
.n f
.ta l i 2 . 5 i 3.5i
I 1 First I I Second I I Third
.fi

will produce the following output:

 0 nroff andtroff 0 83

First Second Third I

But:
.nf
.ta li 2.5iR 3.5iC
I I First I I Second I I Third
.fi

will produce:

First Second Third

Right-adjusted tabs can be useful for aligning numeric data. This is especially
true in t r o f f, where all characters (including blank spaces) have different sizes, and,
as a result, you can’t just line things up by eye. If the numbers you want to align have
an uneven number of decimal positions, you can manually force right adjustment of
numeric data using the special escape sequence \ 0 , which will produce a blank space
exactly the same width as a digit. For example:
. t a liR
I I500.2\0
I -
I 125 3 5
I 150. \ O \ O

will produce:

As on a typewriter, if you have already spaced past a tab position (either by print-
ing characters, or with an indent or other horizontal motion), a tab in the input will push
text over to the next available tab stop. If you have passed the last tab stop, any tabs
present in the input will be ignored.
You must be in no-fill mode for tabs to work correctly. This is not just because
filling will override the effect of the tabs. Using .n f when specifying tabs is an
important rule of thumb; we’ll look at the reasoning behind it in Chapter 15.

Underlining
We haven’t yet described how to underline text, a primary type of emphasis in
n r o f f, which lacks the trof f ability to switch fonts for emphasis.
There are two underlining requests: .u l (underfine) and .c u (continuous
underline). The .u l request underlines only printable characters (the words, but not
the spaces), and .c u underlines the entire text string.
84 0 UNlX Text Processing 0

.
These requests are used just like ce. Without an argument, they underline the
text on the following input line. You can use a numeric argument to specify that more
than one line should be underlined.
Both of these requests produce italics instead of underlines in t r o f f . Although
there is a request, .u f , that allows you to reset the underline font to some other font
than italics,* there is no way to have these requests produce underlining even in
t rof f . (The m s and mm macro packages both include a mucro to do underlining in
t ro f f , but this uses an entirely different mechanism, which is not explained until
Chapter 15.)

Inserting Vertical Space

A s you have seen, a blank line in the input text results in a blank line in the output.
You can leave blank space on the page (for example, between the closing of a letter and
the signature) by inserting a number of blank lines in the input text.
However, particularly when you are entering formatting codes as you write, rather
than going back to code an existing file like our sample letter, it is often more con-
.
venient to specify the spacing with the sp request.
For example, you could type:
Sincerely,
.sp 3
Fred Caslon

In t r o f f , the .sp request is even more important, because t r o f f can space in

much finer increments.
For example, if we were formatting the letter with t rof f , a full space between
paragraphs would look like this:

I 1
In our conversation last Thursday, we discussed a documentation project that would
produce a user’s guide and reference manual for the Alcuin product. Yesterday, I
received the product demo and other materials that you sent me.

Going through a demo session gave me a better understanding of the product. I con-
fess to being amazed by Alcuin. Some people around here, looking over my
shoulder, were also astounded by the illuminated manuscript I produced with Alcuin.
One person, a student of calligraphy, was really impressed.
The output would probably look better if there was a smaller amount of space between
the lines. If we replace the line between the paragraphs with the request - sp - 5 ,
here is what we will get:

*This request is generally used when the document is being typeset in a font family other than Times
Roman. It might be used to set the “underline font” to Helvetica Italic, rather than the standard Italic.
 nroff and troff a5

In our conversation last Thursday, we discussed a documentation project that would

produce a user’s guide and reference manual for the Alcuin product. Yesterday, I
received the product demo and other materials that you sent me.

Going through a demo session gave me a much better understanding of the product.
I confess to being amazed by Alcuin. Some people around here, looking over my
shoulder, were also astounded by the illuminated manuscript I produced with Alcuin.
One person, a student of calligraphy, was really impressed.
Although it may not yet be apparent how this will be useful, you can also space to an
absolute position on the page, by inserting a vertical bar before the distance. The fol-
lowing:
- s p 13i

will space down to a position 3 inches from the top of the page, rather than 3 inches
from the current position.
You can also use negative values with ordinary relative spacing requests. For
example:
.sp -3

will move back up the page three lines. Of course, when you use any of these requests,
you have to know what you are doing. I f you tell n r o f f or t r o f f to put one line
on top of another, that’s exactly what you’ll get. For example:
This is t h e first line.
.sp -2
This is the second line.
.
br
This is t h e third line.

will result in:

I I
I This is t h e second line.
This i s t h e flhrsd line.

Sure enough, the second line is printed above the first, but because we haven’t restored
the original position, the third line is then printed on top of the first.
When you make negative vertical motions, you should always make compensatory
positive motions, so that you end up at the correct position for future output. The previ-
ous example would have avoided disaster if it had been coded:
This is t h e first line.
.sp -2
This is t h e second line.
-
SP
This is t h e third line.
86 0 UNlX Text Processing 0

(Notice that you need to space down one less line than you have spaced up because, in
this case, printing the second line ?uses up? one of the spaces you went back on.)
These kind of vertical motions are generally used for line drawing (e-g., for draw-
ing boxes around tables), in which all of the text is output, and the fonnatter then goes
back up the page to draw in the lines. At this stage, it is unlikely that you will find an
immediate use for this capability. Nonetheless, we are sure that a creative person,
knowing that it is there, will find it just the right tool for a job. (We?ll show a few
creative uses of our own later.)
You probably aren?t surprised that a typesetter can go back up the page. But you
may wonder how a typewriter-like printer can go back up the page like this. The
answer is that it can?t. If you do any reverse line motions (and you do when you use
certain macros in the standard packages, or the t b l and e q n preprocessors), you
must pass the n r o f f output through a special filter program called col to get all of
the motions sorted out beforehand, so that the page will be printed in the desired order:
$ nroff files I col I lp

Double or Triple Spacing

Both nrof f and troff provide a request to produce double- or triple-spaced output
without individually adjusting the space between each line. For example:
.Is 2
Putting this at the top of the file produces double-spaced lines. An argument of 3 speci-
fies triple-spaced lines.

Page Transitions
If we want space at the top of our one-page letter, it is easy enough to insert the com-
mand:
- s p li

before the first line of the text. However, n r o f f and troff do not provide an
easy way of handling page transitions in multipage documents.
By default, n r o f f and t r o f f assume that the page length is 1 1 inches. How-
ever, neither program makes immediate use of this information. There is no default top
and bottom margin, so text output begins on the first line, and goes to the end of the
page.
The .b p (break page) request allows you to force a page break. If you do this,
the remainder of the current page will be filled with blank lines, and output will start
again at the top of the second page. If you care to test this, insert a .bp anywhere in
the text of our sample letter, then process the letter with n r o f f . If you save the
resulting output in a file:
$ nroff letter > 1etter.out
 0 nroff and t r o f f 0 87

you will find that the text following the .bp begins on line 67 ( 1 1 inches at 6 lines per
inch equals 66 lines per page).
To automatically leave space at the top and bottom of each page, you need to use
the .wh (when) request. In nrof f and t r o f f parlance, this request sets a trap-a
position on the page at which a given macro will be executed.
You’ll notice that we said mucro, not request. There’s the rub. To use .wh,
you need to know how to define a macro. It doesn’t work with single requests.
There’s not all that much to defining macros, though. A macro is simply a
sequence of stored requests that can be executed all at once with a single command.
We’ll come back to this later, after we’ve looked at the process of macro definition.
For the moment, let’s assume that we’ve defined two macros, one containing the
commands that will handle the top margin, and another for the bottom margin. The
first macro will be called .TM, and the second .BM. (By convention, macros are
often given names consisting of uppercase letters, to distinguish them from the basic
n r o f f and t r o f f requests. However, this is a convention only, and one that is not
always followed.)
To set traps that will execute these macros, we would use the .wh request as fol-
lows:
.wh 0 TM
.wh -li BM
The first argument to .wh specifies the vertical position on the page at which to exe-
cute the macro. An argument of 0 always stands for the top of the page, and a nega-
tive value is always counted from the bottom of the page, as defined by the page length.
In its simplest form, the .TM macro need only contain the single request to space
down 1 inch, and - BM need only contain the single request to break to a new page. If
.wh allowed you to specify a single request rather than a macro, this would be
equivalent to:
.wh 0 .sp l i
.wh -1i .bp

With an 1 1-inch page length, this would result in an effective 9-inch text area, because
on every page, the formatter’s first act would be to space down 1 inch, and it would
break to a new page when it reached 1 inch from the bottom.
You might wonder why n r o f f and t r o f f have made the business of page
transition more complicated than any of the other essential page layout tasks. There are
two reasons:

The n r o f f and t r o f f programs were designed with the typesetting heri-

tage in mind. Until fairly recently, most typesetters produced continuous out-
put on rolls of photographic paper or film. This output was manually cut and
pasted up onto pages.
Especially in t r o f f , page transition is inherently more complex than the
other tasks we’ve described. For example, books often contain headers and
footers that are set in different type sizes or styles. At every page transition,
the software must automatically save information about the current type style,
aa 0 UNIX Text Processing 0

switch to the style used by the header or footer, and then revert to the original
style when it returns to the main text. Or consider the matter of footnotes-the
position at which the page ends is different when a footnote is on the page.
The page transition trap must make some allowance for this.

In short, what you might like the formatter to do during page transitions can vary. For
this reason, the developers of n r o f f and t r o f f have allowed users to define their
own macros for handling this area.
When you start out with n r o f f or t r o f f , we advise you to use one of the
ready-made macro packages, m s or mm. The standard macro package for UNIX sys-
tems based on System V is mm; the standard on Berkeley UNIX systems is m s .
Berkeley UNIX systems also support a third macro package called m e . In addition,
there are specialized macro packages for formatting viewgraphs, standard UNIX refer-
ence manual pages (man), and UNIX permuted indexes (mptx). Only the m s and
mm packages are described in this book. The macro packages have already taken into
account many of the complexities in page transition (and other advanced formatting
problems), and provide many capabilities that would take considerable time and effort
to design yourself.
Of course, it is quite possible to design your own macro package, and we will go
into all of the details later. (In fact, this book is coded with neither of the standard
macro packages, but with one developed by Steve Kochan and Pat Wood of Pipeline
Associates, the consulting editors of this series, for use specifically with the Hayden
UNIX library.)

Page Length Revisited

Before we take a closer look at macros, let’s take a moment to make a few more points
about page length, page breaks, and the like.
Assuming that some provision has been made for handling page transitions, there
are several wrinkles to the requests we have already introduced, plus several new
requests that you will probably find useful.
First, let’s talk about page length. It’s important to remember that the printing
area is defined by the interaction of the page length and the location of the traps you
define. For example, you could define a text area 7.5 inches high (as we did in prepar-
ing copy for this book) either by

changing the page length to 9.5 inches, and setting I-inch margins at the top
and bottom;
leaving the page length at 1 1 inches, and setting 1.75-inch margins at the top
and bottom.

In general, we prefer to think of .pl as setting the paper length, and use the page
transition traps to set larger or smaller margins.
However, there are cases where you really are working with a different paper size.
A good example of this is printing addresses on envelopes: the physical paper height is
about 4 inches (24 lines on a typewriter-like printer printing 6 lines per inch), and we
 0 n r o f f and t r o f f 0 a9

want to print in a narrow window consisting of four or five lines. A good set of defini-
tions for this case would be:
-pl 4i
.wh 0 TM
.wh -9v BM
with .TM containing the request .s p 9v, and with .BM, as before, containing
.bp.
There is more to say about traps, but it will make more sense later, so we’ll leave
the subject for now.

Page Breaks without Line Breaks

Page breaks-we’ve talked about their use in page transition traps, but they also have a
common use on their own. Often, you will want to break a page before it would nor-
mally end. For example, if the page breaks right after the first line of a paragraph, you
will probably want to force the line onto the next page, rather than leaving an
“orphaned” line. Or you might want to leave blank space at the bottom of a page for
an illustration. To do this, simply enter a .bp at the desired point. A new page will
be started immediately.
However, consider the case in which you need to force a break in the middle of a
paragraph to prevent a “widowed” line at the top of the next page. If you do this:
The medieval masters of calligraphy and illumination
are largely unknown to us. We thankfully have examples
of their work, and even
- bP
marginal notes by the copyists of some manuscripts,
but the men who produced these minute masterpieces
are anonymous.
the .bp request will also cause a line break, and the text will not be filled properly:

The medieval masters of call graphy and illumination

are largely unknown to u s . We thankfully have examples
of their work, and even

New page begins here

marginal notes by the copyists of some manuscripts, but

the men who produced these minute masterpieces are
anonymous.

Fortunately, there is a way around this problem. If you begin a request with an apos-
trophe instead of a period, the request will not cause a break.
90 0 UNlX Text Processing 0

The medieval masters of calligraphy and illumination

are largely unknown t o us. W e thankfully have examples
of their work, and even
‘bP
marginal notes by t h e copyists of some manuscripts,
but t h e men who produced these minute masterpieces
are anonymous.

Now we have the desired result:

The medieval masters of calligraphy and illumination

are largely unknown t o us. We thankfully have examples

New page begins here

of their work, and even marginal notes by t h e copyists

of some manuscripts, but t h e men who produced these
minute masterpieces are anonymous.

(In fact, most page transition macros use this feature to make paragraphs continue
across page boundaries. We’ll take a closer look at this in later chapters.)
Another very useful request is the conditional page break, or .ne (need) request.
If you want to make sure an entire block of text appears on the same page, you can use
this request to force a page break if there isn’t enough space left. If there is sufficient
space, the request is ignored.
For example, the two requests:
.ne 3.2i
. s p 3i

might be used to reserve blank space to paste in an illustration that is 3 inches high.
The .n e request does not cause a break, so you should be sure to precede it with
. b r or another request that causes a break if you don’t want the remnants of the
current line buffer carried to the next page if the .ne is triggered.
It is often better to use .ne instead of .bp, unless you’re absolutely sure that
you will always want a page break at a particular point. If, in the course of editing, an
.ne request moves away from the bottom of the page, it will have no effect. But a
.bp will always start a new page, sometimes leaving a page nearly blank when the text
in a file has been changed significantly.
There are other special spacing requests that can be used for this purpose.
(Depending on the macro package, these may have to be used.) For example, .s v
(save space) requests a block of contiguous space. If the remainder of the page does
not contain the requested amount of space, no space is output. Instead, the amount of
space requested is remembered and is output when an .os (output saved space)
request is encountered.
These are advanced requests, but you may need to know about them because most
macro packages include two other spacing requests in their page transition macros:
.ns (no space) and .rs (restore space). An .n s inhibits the effect of spacing
requests; .r s restores the effectiveness of such requests.
 0 n r o f f and troff 0 91

Both the m s and mm macros include an .ns request in their page transition
macros. As a result, if you issue a request like:
.sp 3i

with 1 inch remaining before the bottom of the page, you will not get 1 inch at the bot-
tom, plus 2 inches at the top of the next page, but only whatever remains at the bottom.
The next page will start right at the top. However, both macro packages also include an
- o s request in their page top macro, so if you truly want 3 inches, use .s v 3 i,and
you will get the expected result.
However, if you use .s v , you will also have another unexpected result: text
following the spacing request will “float” ahead of it to fill up the remainder of the
current page.
We’ll talk more about this later. We introduced it now to prevent confusion when
spacing requests don’t always act the way you expect.

Page Numbering
The nrof f and t rof f programs keep track of page numbers and make the current
page number available to be printed out (usually by a page transition macro). You can
.
artificially set the page number with the pn request:

.pn 5 Set the current page number to 5

- p n +5 Increment the current page number by 5
.pn -5 Decrement the current page number by 5

You can also artificially set the number for the nexf page whenever you issue a .bp
request, simply by adding a numeric argument:

.bp 5 Break the page and set the next page number to 5
.bp + 5 Break the page and increment the next page number by 5
.bp -5 Break the page and decrement the next page number by 5

In addition to inhibiting .sp, the -

ns request inhibits the action of bp, unfess a
page number is specified. This means (at least in the existing macro packages), that the
sequence:
- bP
- bP
will not result in a blank page being output. You will get the same effect as if you had
-
specified only a simple bp. Instead, you should specify:
.bp +1

The starting page number (usually 1) can also be set from the command line, using the
-n option. For example:
92 0 UNlX Text Processing 0

$ n r o f f -ms -n10 file

will start numbering file at page number 10. In addition, there is a command-line
option to print only selected pages of the output. The -0 option takes a list of page
numbers as its argument. The entire file (up to the last page number in the list) is pro-
cessed, but only the specified pages are output. The list can include single pages
separated by commas, or a range of pages separated by a hyphen, or both. A number
followed by a trailing hyphen means to output from that page to the end. For example:
$ n r o f f -ms -01,5,7-9,13- file
will output pages 1, 5, 7 through 9, and from 13 to the end o f the file. There should be
no spaces anywhere in the list.

= Changing Fonts
In old t rof f (otrof f), you were limited to four fonts at a time, because the fonts
had to be physically mounted on the C/A/T typesetter. With ditrof f and a laser
printer or a modem typesetter, you can use a virtually unlimited number of fonts in the
same document.
In o t rof f you needed to specify the basic fonts that are in use with the - fp
(font position) request. Normally, at the front o f a file (or, more likely, in the macro
package), you would use this request to specify which fonts are mounted in each of the
four quadrants (positions) of the typesetter wheel. B y default, the roman font is
mounted in position 1 , the italic font in position 2, the bold font in position 3, and the
special font in position 4. That is, t rof f acts as though you had included the lines:
-fp 1 R
-fp 2 I
.fp 3 B
.fp 4 s

In dit rof f, up to ten fonts are automatically mounted, with the special font in posi-
tion 10. Which fonts are mounted, and in which positions, depends on the output dev-
ice. See Appendix D for details. The font that is mounted in position 1 will be used
for the body type of the text-it is the font that will be used if no other specification is
given. The special font is also used without any intervention on your part when a char-
acter not in the normal character set is requested.
To request one of the other fonts, you can use either the .ft request, or the
inline font-switch escape sequence \ f.
For example:
.ft €3

This line will be set i n b o l d t y p e .

- br
.ft R
This line will again be set in roman t y p e .

will produce:
 0 n r o f f and t r o f f 0 93

I I
I This line will be set in bold type.
This line will again be set in roman type.
You can also change fonts using an inline font escape sequence. For example, the
preceding sentence was coded like this:
...a n i n l i n e f o n t \fIescape sequence\fP.

You may wonder at the \ fP at the end, rather than \ fR. The P command is a spe-
cial code that can be used with either the .ft request or the \ f escape sequence. It
means “return to the previous font, whatever it was.” This is often preferable to an
explicit font request, because it i s more general.
All of this begs the question of fonts different than Times Roman, Bold, and
Italic. There are two issues: first, which fonts are available on the output device, and
second, which fonts does t roff have width tables for. (As described previously,
troff uses these tables to determine how far to space over after it outputs each char-
acter.) For otroff these width tables are in the directory /usr/lib/font, in
files whose names begin with ft. If you list the contents of this directory, you might
see something like this for o t KO ff:
$ 1s /usr/lib/font
ftB ftBC ftC ftCE ftCI
ftCK ftCS ftCW ftFD ftG
ftGI ftGM ftGR ftH ftHB
ftHI ftI ftL ftLI ftPA
ftPB f t P I ftR f t S ftSB
ftSI ftSM ftUD

You can pick out the familiar R, I, B, and S fonts, and may guess that ftH, ftHI,
and ftHB refer to Helvetica, Helvetica Italic, and Helvetica Bold fonts. However,
unless you are familiar with typesetting, the other names might as well be Greek to you.
In any event, these width tables, normally supplied with troff,are for fonts that are
commonly used with the C/A/T typesetter. If you are using a different device, they may
be of no use to you.
The point is that if you are using a different typesetting device, you will need to
get information about the font names for your system from whoever set up the equip-
ment to work with troff. The contents of /usr/lib/font will vary from
installation to installation, depending on what fonts are supported.
For ditroff,there is a separate subdirectory in /usr/lib/font for each
supported output device. For example:
$ 1s /usr/lib/font
devl j devps
$ 1s /usr/lib/font/devps
B.out BI.out CB.out CI.out CW.out CX.out
DESC.out H . o u t HB o u t . HI. out HK. o u t HO. o u t
HX.out I.out L I out . PA. o u t .
PB o u t PI. out
PX-out R.out 0. o u t RS.out S.out s1. o u t
94 0 UNlX Text Processing 0

Here, the font name is followed by the string .out.

Again, the font names themselves are probably Greek to you. However, with
ditrof f, you can actually use any of these names, and see what results they give
you, because all fonts should be available at any time.
For the sake of argument, let's assume that your typesetter or other t r o f f -
compatible equipment supports the Helvetica font family shown in Figure 4-3, with the
names H, H I , and HB. (This is a fairly reasonable assumption, because Helvetica is
probably the most widely available font family after Times.)

Helvetica
abcdefghijklrnnopqrstuvwxyz
ABCDEFGHIJKLMNOPQRSTUVWXYZ
1234567890
!$%&a()"*+-.,/:;=?[]I
0 - - - l/4 ' / z 3/4 fi fl "t'80
Helvetica Italic
abcdefghijklrnnopqrstuvwxyz
ABCDEFGHIJKLMNOPQRS TUVWXYZ
1234567890
!$ % & () " + - . ,/: ;= ? [ ] I
0 -- - '/4 1' 3/4
'2 fi fl 't '80

Helvetica Bold
abcdefghijklmnopqrstuvwxyz
ABCDEFGHIJKLMNOPQRSTUVWXYZ
1234567890
!$%&()"*+-.,/:;=?[]I
0 -- - l/4 112 314 fi fl " t * @ @

Fig. 4-3. Helvetica Fonts

When specifying two-character font names with the \f escape sequence, you
must add the ( prefix as well. For example, you would specify Helvetica Italic by the
inline sequence \ f ( H I , and Helvetica Bold by \ f (HB.
 0 n r o f f and t r o f f 0 95

There is another issue when you are using fonts other than the Times Roman fam-
ily. Assume that you decide to typeset your document in Helvetica rather than Roman.
You reset your initial font position settings to read:
-fp 1 H
.fp 2 H I
. f p 3 HB
.fp 4 s

However, throughout the text, you have requests of the form:

.ft B

or:
\fB

You will need to make a set of global replacements throughout your file. To insulate
yourself in a broader way from overall font change decisions, t rof f allows you to
specify fonts by position, even within .f t and \f requests:

.ft 1 or \fl Use the font mounted in position 1

.ft 2 or \f2 Use the font mounted in position 2
.ft 3 or \f3 Use the font mounted in position 3
.ft 4 or \f4 Use the font mounted in position 4

Because you don’t need to use the .f p request to set font positions with d i t K O f f,
and the range of fonts is much greater, you may have a problem knowing which fonts
are mounted in which positions. A quick way to find out which fonts are mounted is to
run d i t r o f f on a short file, sending the output to the screen. For example:
$ ditroff -Tps junk 1 more
x T ps
x res 7 2 0 1 1
x init
x font 1 R
x font 2 I
x font 3 B
x font 4 BI
x font 5 CW
x font 6 CB
x font 7 H
x font 8 HB
x font 9 HI
x font 10 S
.-.
The font positions should appear at the top of the file. In this example, you see the fol-
lowing fonts: (Times) Roman, (Times) Bold, (Times) Italic, (Times) Bold Italic, Con-
stant Width, Constant Bold, Helvetica, Helvetica Bold, Helvetica Italic, and Special.
Which font is mounted in which position is controlled by the file DESC. o u t in the
device subdirectory of / u s r / l i b / f o n t . See Appendix D for details.
I
96 0 UNlX Text Processing 0

Special Characters
A variety of special characters that are not part of the standard ASCII character set are
supported by n r o f f and t r o f f . These include Greek letters, mathematical sym-
bols, and graphic characters. Some of these characters are part of the font referred to
earlier as the special font. Others are part of the standard typesetter fonts.
Regardless of the font in which they are contained, special characters are included
in a file by means of special four-character escape sequences beginning with \ (.
Appendix B gives a complete list of special characters. However, some of the
most useful are listed in Table 4-4, because even as a beginner you may want to include
them in your text. Although n r o f f makes a valiant effort to produce some of these
characters, they are really best suited for t r o f f .

TABLE 4-4. Special Characters

Name Escape Sequence Output Character

em dash \(em -
bullet \(bu 0

square \(sq 0
baseline rule \(nJ -
underrule \(ul -
114 \( 14 ’14
112 \( 12 ‘12

314 \( 34 v4
degrees \(de 0

dagger \(dg t
double dagger \(dd $
registered mark kg 8
copyright symbol \(co 0
section mark \( sc 0
square root \(sq .I
greater than or equal \(>= 2
less than or equal \( <= I
not equal \(!= #
multiply \(mu X
divide \(di -
plus or minus \(+- k
right arrow \( -> +
left arrow \(<- t
up arrow \( ua T
down arrow \(da J8

We’ll talk more about some of these special characters as we use them. Some are
used internally by eqn for producing mathematical equations. The use of symbols
such as the copyright, registered trademark, and dagger is fairly obvious.
 0 n r o f f and t r o f f 0 97

However, you shouldn’t limit yourself to the obvious. Many of these special
characters can be put to innovative use. For example, the square root symbol can be
used to simulate a check mark, and the square can become an alternate type of bullet.
As we’ll show in Chapter 15, you can create additional, effective character combina-
tions, such as a checkmark in a box, with overstriking.
The point i s to add these symbols to your repertoire, where they can wait until
need and imagination provide a use for them.

Type Size Specification

Typesetting also allows for different overall sizes of characters. Typesetting character
sizes are described by units called points. A point is approximately 1/72 of an inch.
Typical type sizes range from 6 to 72 points. A few different sizes follow:

This line 1s set in 6-point type.

This line is set in 8-point type.

This line is set in 10-point type.

This line i s set in 12-point type.
This line is set in 14-point type.
This line is set in 18-point type.
(The exact size of a typeface does not always match its official size designation.. For
example, 12-point type is not always 1/6 inch high, nor is 72-point type 1 inch high.
The precise size will vary with the typeface.)
As with font changes, there are two ways to make size changes: with a request
and with an inline escape sequence. The .ps request sets the point size. For exam-
ple:

.ps 10 Set the point size to 10 points

A - p s request that does not specify any point size reverts to the previous point size
setting, whatever it was:

.ps 1 0

Some text here

- PS Revert to the point size before we changed it

To switch point size in the middle of the line, use the \ s escape sequence. For exam-
ple, many books reduce the point size when they print the word UNIX in the middle of a
line. The preceding sentence was produced by these input lines:
 98 0 UNlX Text Processing 0

F o r example, many books reduce t h e point size when

they print t h e word \s8UNIX\sO in t h e middle of a line.

As you can probably guess from the example, \ S O does not mean to use a point size
of 0, but to revert to the previous size.
In addition, you can use relative values when specifying point sizes. Knowing
that the body of the book is set in 10-point type, we could have achieved the same
result by entering:
For example, many books reduce t h e point size when
they print t h e word \s-2UNIX\sO in t h e middle of a line.

You can increment or decrement point sizes only using a single digit; that is, you can’t
increment or decrement the size by more than 9 points.
Only certain sizes may be available on the typesetter. (Legal point sizes in
o t r o f f are 6, 7 , 8, 9, 10, 1 1 , 12, 14, 16, 18, 20, 22, 24, 28, and 36. Legal point sizes
in d i t r o f f depend upon the output device, but there will generally be more sizes
available.) If you request a point size between two legal sizes, o t r o f f will round up
to the next legal point size; d i t r o f f will round to the nearest available size.

Vertical Spacing
In addition to its ability to change typefaces and type sizes on the same page, a
typesetter allows you to change the amount o f vertical space between lines. This spac-
ing is sometimes referred to as the baseline spacing because it is the distance between
the base o f characters on successive lines. (The difference between the point size and
the baseline spacing is referred to as leading, from the old days when a human compo-
sitor inserted thin strips of lead between successive lines of type.)
A typewriter or typewriter-style printer usually spaces vertically in 1/6-inch incre-
ments (Le.. 6 lines per inch). A typesetter usually adjusts the space according to the
point size. For example, the type samples shown previously were all set with 20 points
of vertical space. More typically, the vertical space will vary along with the type size,
like this:

This line i s set in 6-point type and 8-point spacing.

is set in 8-point type and IO-point spacing.
This line
This line is set in 10-point type and 12-point spacing.
This line is set in 12-point type and 14-point spacing.
This line is set in 14-point type and 16-point spacing.
1 This line is set in 18-point type and 20-poi
Typically, the body o f a book is set with a single size o f type (usually 9 or 10 point,
with vertical spacing set to 1 1 or 12 points, respectively). Larger sizes are used occa-
sionally for emphasis, for example, in chapter or section headings. When the type size
is changed, the vertical spacing needs to be changed too, or the type will overrun the
previous line, as follows, where 14-point type i s shown with only 10-point spacing.
 0 nroff and t r o f f 0 99

I Here is typ larg r than

f c f
the space a lotte for it.
Vertical spacing is changed with the .vs request. A vertical space request will
typically be paired with a point size request:
-ps 1 0
.vs 1 2
After you set the vertical spacing with .vs,this becomes the basis of the v unit for
t rof f . For example, if you enter .v s 1 2 , the request .s p will space down 12
points; the request:
.sp 0.5v

will space down 6 points, or half the current vertical line spacing. However, if you
change the baseline vertical spacing to 16, the .s p request will space down 16 points.
Spacing specified in any other units will be unaffected. What all this adds up to is the
commonsense observation that a blank line takes up the same amount of space as one
containing text.
When you use double and triple spacing, it applies a multiplication factor to the
baseline spacing. The request - 1 s 2 will double the baseline spacing. You can
specify any multiplication factor you like, though 2 and 3 are the most reasonable
values.
The - 1 s request will only affect the spacing between output lines of text. It
does not change the definition of v or affect vertical spacing requests.

A First Look at Macros =

Although we won’t go into all the details of macro design until we have discussed the
existing macro packages in the next two chapters, we’ll cover some of the basic con-
cepts here. This will help you understand what the macro packages are doing and how
they work.
To define a macro, you use the . d e request, followed by the sequence of
requests that you want to execute when the macro is invoked. The macro definition is
terminated by the request . . (two dots). The name to be assigned to the macro is
given as an argument to the .de request.
You should consider defining a macro whenever you find yourself issuing a
repetitive sequence of requests. If you are not using one of the existing macro packages
(which have already taken care of this kind of thing), paragraphing is a good example of
the kind of formatting that lends itself to macros.
Although it is certainly adequate to separate paragraphs simply by a blank line,
you might instead want to separate them with a blank line and a temporary indent.
What’s more, to prevent “orphaned” lines, you would like to be sure that at least two
lines of each paragraph appear at the bottom of the page. So you might define the fol-
lowing macro:
100 0 UNlX Text Processing 0

.de P
- SP
.ne 2
.ti 5n

This is the simplest kind of macr-a straightforward sequence of stored commands.

However, macros can take arguments, take different actions depending on the presence
or absence of various conditions, and do many other interesting and wonderful things.
We'll taik more about the enormous range of potential in macros in later chapters.
For the moment, let's just consider one or two points that you will need to understand
in order to use the existing macro packages.

Macro Arguments
Most basic t r o f f requests take simple arguments-single characters or letters. Many
macros take more complex arguments, such as character strings. There are a few simple
pointers you need to keep in mind through the discussion of macro packages in the next
two chapters.
First, a space is taken by default as the separator between arguments. If a single
macro argument is a string that contains spaces, you need to quote the entire string to
keep it from being treated as a series of separate arguments.
For example, imagine a macro to print the title of a chapter in this book. The
macro call looks like this:
.CH 4 "Nroff and Troff"

A second point: to skip an argument that you want to ignore, supply a null string ("").
For example:
.CH " " '' P re face"

As you can see, it does no harm to quote a string argument that doesn't contain spaces
( " P r e f a c e " ) , and it is probably a good habit to quote all strings.

Number Registers
When you use a specific value in a macro definition, you are limited to that value when
you use the macro. For example, in the paragraph macro definition shown previously,
the space will always be 1, and the indent always 5n.
However, n r o f f and t r o f f allow you to save numeric values in special vari-
ables known as number registers. If you use the value of a register in a macro defini-
tion, the action of the macro can be changed just by placing a new value in the register.
For example, in m s , the size of the top and bottom margins i s not specified with an
absolute value, but with a number register. As a result, you don't need to change the
macro definition to change these margins; you simply reset the value of the appropriate
number register. Just as importantly, the contents of number registers can be used as
flugs (a kind of message between macros). There are conditional statements in the
markup language of n r o f f and t r o f f , so that a macro can say: "If number register
 0 nroff and troff 0 101

Y has the value x, then do thus-and-so. Otherwise, do this.? For example, in the mm
macros, hyphenation is turned off by default. To turn it on, you set the value of a cer-
tain number register to 1. Various macros test the value of this register, and use it as a
signal to re-enable hyphenation.
To store a value into a number register, use the .n r request. This request takes
two arguments: the name of a number register,* and the value to be placed into it.
For example, in the m s macros, the size of the top and bottom margins is stored
in the registers HM (header margin) and F M (footer margin). To reset these margins
from their default value of 1 inch to 1.75 inches (thus producing a shorter page like the
one used in this book), all you would need to do is to issue the requests:
.nr HM 1 . 7 5 i
.nr F M 1 . 7 5 i

You can also set number registers with single-character names from the command line
by using the -r option. (The mm macros make heavy use of this capability.) For
example:
$ ntoff -nrm -rN1 file
will formatfile using the m macros, with number register N set to the value 1 . We
will talk more about using number registers later, when we describe how to write your
own macros. For the moment, all you need to know i s how to put new values into
existing registers. The next two chapters will describe the particular number registers
that you may find useful with the mm and m s macro packages.

Predefined Strings
The m m and m s macro packages also make use of some predefined text strings. The
n r o f f and t r o f f programs allow you to associate a text string with a one- or two-
character string name. When the formatter encounters a special escape sequence includ-
ing the string name, the complete string is substituted in the output.
To define a string, use the . d s request. This request takes two arguments, the
string name and the string itself. For example:
. d s nt N r o f f and T r o f f

The string should not be quoted. It can optionally begin with a quotation mark, but it
should not end with one, or the concluding quotation mark will appear in the output. If
you want to start a string with one or more blank spaces, though, you should begin the
definition with a quotation mark. Even in this case, there is no concluding quotation
mark. A s always, the string is terminated by a newline.

*Number register names can consist of either one or two characters, just like macro names. However, they
are distinct-that is, a number register and a macro can be given the same name without conflict.
102 0 UNIX Text Processing 0

You can define a multiline string by hiding the newlines with a backslash. For
example:
.ds LS This is a very long string that goes over \
more than one line.

When the string is interpolated, it will be subject to filling (unless no-fill mode is in
effect) and may not be broken into lines at the same points as you’ve specified in the
definition. To interpolate the string in the output, you use one of the following escape
sequences:
\*a
\ * (ab
where a is a one-character string name, and ab is a two-character string name.
To use the nr string we defined earlier, you would type:
\ * (nt
It would be replaced in the output by the words Nroff and Troff.
Strings use the same pool of names as macros. Defining a string with the same
name as an existing macro will make the macro inoperable, so it is not advisable to go
around wildly defining shorthand strings. The v i editor’s abbreviation facility
(described in Chapter 7) i s a more effective way to save yourself work typing.
Strings are useful in macro design in much the same way number registers are-
they allow a macro to be defined in a more general way. For example, consider this
book, which prints the title of the chapter in the header on each odd-numbered page.
The chapter title is not coded into the page top macro. Instead, a predefined string is
interpolated there. The same macro that describes the format of the chapter title on the
first page of the chapter also defines the string that will appear in the header.
In using each of the existing macro packages, you may be asked to define or
interpolate the contents of an existing string. For the most part, though, string defini-
tions are hidden inside macro definitions, so you may not run across them. However,
there are a couple of handy predefined strings you may find yourself using, such as:
\ * (DY
which always contains the current date in the m s macro package. (The equivalent
string in mm is \ * (DT.) For example, if you wanted a form letter to contain the date
that it was formatted and printed rather than the date it was written, you could interpo-
late this string.

Just What Is a Macro Package?

Before leaving the topic of macros, we ought to take a moment to treat a subject we
have skirted up to this point: just what is a macro package?
As the name suggests, a macro package is simply a collection of macro defini-
tions. The fact that there are command-line options for using the existing packages may
seem to give them a special status, but they are text files that you can read and modify
(assuming that your system has the UNIX file permissions set up so you can do so).
 0 nroffandtroff 0 103

There is no magic to the options -ms and -mm. The actual option to n r o f f
and troff is -mx, which tells the program to look in the directory
/usr/lib/tmac for a file with a name of the form tmac .x. As you might expect,
this means that there is a file in that directory called tmac .s or tmac .m (depending
on which package you have on your system). It also means that you can invoke a
macro package of your own from the command line simply by storing the macro defini-
tions in a file with the appropriate pathname. This file will be added to any other files
in the formatting run. This means that if you are using the ms macros you could
achieve the same result by including the line:
.so /usr/lib/tmac/tmac.s

at the start of each source file, and omitting the command-line switch -ms. (The .so
request reads another file into the input stream, and when its contents have been
exhausted, returns to the current file. Multiple .s o requests can be nested, not just to
read in macro definitions, but also to read in additional text files.)
The macros in the standard macro packages are no different (other than in com-
plexity) than the macros you might write yourself. In fact, you can print out and study
the contents of the existing macro packages to learn how they work. We’ll be looking
in detail at the actions of the existing macro packages, but for copyright reasons we
can’t actually show their internal design. We’ll come back to all this later. For now,
all you need to know is that macros aren’t magic-just an assemblage of simple com-
mands working together.
 C H A P T E R
rn 8 rn

Thems Macros

The UNIX shell i s a user interface for the kernel, the actual heart of the operating sys-
tem. You can choose the C shell or Korn shell instead of the Bourne shell, without
worrying about its effects on the low-level operations of the kernel. Likewise, a macro
package is a user interface for accessing the capabilities of the n r o f f / t r o f f for-
matter. Users can select either the m s or m macro packages (as well as other pack-
ages that are available on some systems) to use with n r o f f/tr o f f .
The m s package was the original Bell Labs macro package, and is available on
many UNIX systems, but it is no longer officially supported by AT&T. Our main rea-
son for giving m s equal time is that many Berkeley UNIX systems ship m s instead of
mm. In addition, it is a less complex package, so it is much easier to learn the principles
of macro design by studying m s than by studying mm.
A third general-purpose package, called m e , is also distributed with Berkeley
UNIX systems. It was written by Eric Allman and is comparable to m s and m.
(Mark Horton writes us: I think of m s as the FORTRAN of n r o f f, mm as the PL/I,
and m e as the Pascal.) The m e package i s not described in this book.
In addition, there are specialized packages-mv, for formatting viewgraphs,
m p t x , for formatting the permuted index found in the UNIX Reference Manual, and
man, for formatting the reference pages in that same manual. These packages are sim-
ple and are covered in the standard UNIX documentation.
Regardless of which macro package you choose, the formatter knows only to
replace each call of a macro with its definition. The macro definition contains the set of
requests that the formatter executes. Whether a definition is supplied with the text in
the input file or found in a macro package i s irrelevant to n r o f f l t r o f f. The for-
matter can be said to be oblivious to the idea of a macro package.
You might not expect this rather freely structured arrangement between a macro
package and n r o f f/tr o f f . Macros are application programs of sorts. They organ-
ize the types of functions that you need to be able to do. However, the actual work is
accomplished by n r o f fjtrof f requests.
In other words, the basic formatting capabilities are inherent in n r o f f and
t r o f f ; the user implementation of these capabilities to achieve particular formats is

104
 0 The m s Macros 0 105

accomplished with a macro package. If a macro doesn’t work the way you expect, its
definition may have been modified. It doesn’t mean that n r o f f / t r o f f works dif-
ferently on your system. It is one thing to say “nrof f / t r o f f won’t let me do it,”
and another to say “I don’t have the macro to do it (but I could do it, perhaps).”
A general-purpose macro package like m s provides a way of describing the for-
mat of various kinds of documents. Each document presents its own specific problems,
and macros help to provide a simple and flexible solution. The m s macro package is
designed to help you format letters, proposals, memos, technical papers, and reports.
For simple documents such as letters, m s offers few advantages to the basic for-
mat requests described in Chapter 4. But as you begin to format more complex docu-
ments, you will quickly see the advantage of working with a macro package, which pro-
vides specialized tools for so many of the formatting tasks you will encounter.
A text file that contains m s macros can be processed by either n r o f f or
t r o f f , and the output can be displayed on a terminal screen or printed on a line
printer, a laser printer, or a typesetter.

Formatting a Text File with ms

If you want to format an m s document for a line printer or for a terminal screen, enter
this command line:
$ nroff -ms file(s)
To format for a laser printer or typesetter, enter this command line:
$ troff -ms file(s) I devicepostprocessor
Be sure to redirect the output to a file or pipe it to the printer; if you do not, the output
will be sent to your terminal screen.

Problems in Getting Formatted Output

There are two ways for a program to handle errors. One is to have the program ter-
minate and issue an error message. The other way is to have it keep going in hopes that
the problems won’t affect the rest of the output. The m s macros take this second
approach.
In general, m s does its best to carry on no matter how scrambled the output
looks. Sometimes the problems do get corrected within a page or two; other times the
problem continues, making the remaining pages worthless. Usually, this is because the
formatter had a problem executing the codes as they were entered in the input file.
Most of the time input errors are caused by not including one of the macros that must
be used in pairs.
Because m s allows formatting to continue unless the error is a “fatal” one, error
correction is characteristic of the m s macro definitions. Apart from the main function
of the macro, some of them, such as the paragraph macro, also invoke another macro
called .RT to restore certain default values.
106 0 UNlX Text Processing 0

Thus, if you forget to reset the point size or indentation, you might notice that the
problem continues for a while and then stops.

PageLayout
A s suggested in the last chapter, one of the most important functions of a macro pack-
age is that it provides basic page layout defaults. This feature makes it worthwhile to
use a macro package even if you don’t enter a single macro into your source file.
At the beginning of Chapter 4, we showed how n r o f f alone formatted a sample
letter. If we format the same letter with m s , the text will be adjusted on a page that
has a default top and bottom margin of 1 inch, a default left margin, or page offset, of
about 1 inch, and a default line length of 6 inches.
All of these default values are stored in number registers so that you can easily
change them:

LL Line Length
HM Header (top) Margin
FM Footer (bottom) Margin
PO Page offset (left margin)

For example, if you like larger top and bottom margins, all you need to do is
insert the following requests at the top of your file:
.nr HM 1.5i
.nr F M 1.5i

Registers such as these are used internally by a number of m s macros to reset the
formatter to its default state. They will not take effect until one of those “reset” mac-
ros is encountered. In the case of HM and FM, they will not take effect until the next
page unless they are specified at the very beginning of the file.*

Paragraphs
A s we saw in the last chapter, paragraph transitions are natural candidates for macros
because each paragraph generally will require several requests (spacing, indentation,) for
proper formatting.
There are four paragraph macros in m s :

*These “reset” macros (those that call the internal macro .RT) include .LP, .PP, .IP, -QP,
.
.SH, NH, .RS, . R E , .T S , and .TE. The very first met macro calk a special initialization
macro called .B G that is used only once, on the first page. This macro prints the cover sheet, if any (see
“Cover Sheet Macros” later in this chapter), as well as performing some special first-page initialization.
 0 The m s Macros 0 107

.LP Block paragraph

.PP First line of paragraph indented
- QP Paragraph indented from both margins
.I P Paragraph with hanging indent (list item)

The LP macro produces a justified, block paragraph. This is the type of para-
graph used for most technical documentation. The . P P macro produces a paragraph
with a temporary indent for the first line. This paragraph type is commonly used in
published books and magazines, as well as in typewritten correspondence.
Let’s use the same letter to illustrate the use of these macros. In the original
example (in Chapter 4), we left blank lines between paragraphs, producing an effect
similar to that produced by the .LP macro.
In contrast, . P P produces a standard indented paragraph. Let’s code the letter
using . P P macros. Because this is a letter, let’s also disable justification with an
.na request. And of course, we want to print the address block in no-fill mode, as
shown in Chapter 4. Figure 5-1 shows the coded letter and Figure 5-2 shows the for-
matted output.

Spacing between Paragraphs

With n r o f f , all of the paragraph macros produce a full space between paragraphs.
However, with t rof f , the paragraph macros output a blank space of 0 . 3 ~ Basically,
.
this means that a blank line will output one full space and the paragraph macros will
output about a third of that space.
The amount of spacing between paragraphs is contained in the number register
PD (paragraph distance). If you want to change the amount of space generated by any
of the paragraph macros, simply change the contents of this register.
For example, if you don’t want to leave any space between paragraphs in the
letter, you could put the following line at the start of your file:
. n r PD 0
This flexibility afforded by macro packages is a major advantage. It is often possible to
completely change the appearance of a coded document by resetting only a few number
registers at the start of a file. (As we’ll see, this statement is even more true of of mm
than of ms.)

Quoted Paragraphs
A paragraph that is indented equally from the left and right margins is typically used to
display quoted material. It is produced by .QP. For example:
- QP
In t h e next couple of d a y s , I’ll be putting together a .__
108 0 UNlX Text Processing 0

.ad r
April 1, 1987
.sp 2
.ad
.nf
M r . John Fust
Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02159
.fi
- SP
.na
Dear M r . Fust:
.PP
In our conversation last Thursday, we discussed a documentation
project that would produce a user's manual on the Alcuin
product. Yesterday, I received the product demo and other
materials that you sent me.
.PP
Going through a demo session gave me a much better understanding
of the product. I confess to being amazed by Alcuin.
Some people around here, looking over my shoulder, were also
astounded by the illustrated manuscript I produced with Alcuin.
One person, a student of calligraphy, was really impressed.
* PP

In the next couple of days, I'll be putting together a written

plan that presents different strategies for documenting the
Alcuin product. After I submit this plan, and you have had time
to review it, let's arrange a meeting at your company to discuss
these strategies.
.PP
Thanks again for giving us the opportunity to bid on this
documentation project. I hope we can decide upon a strategy
and get started as soon as possible in order to have the manual
ready in time for the first customer shipment. I look forward to
meeting with you towards the end of next week.
- SP
S incerely,
.sp 3
Fred Caslon

Fig. 5-1. Letter Coded with ms Macros

 0 Thems Macros 109

April 1, 1987

Mr. John Fust

Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02159
Dear Mr. Fust:
In our conversation last Thursday, we discussed
a documentation project that would produce a user's
manual on the Alcuin product. Yesterday, I received
the product demo and other materials that you sent
me -
Going through a demo session gave me a much
better understanding of the product. I confess to
being amazed by Alcuin. Some people around here,
looking over my shoulder, were also astounded by the
illustrated manuscript I produced with Alcuin. One
person, a student of calligraphy, was really
impressed.
In the next couple of days, I'll be putting
together a written plan that presents different
strategies for documenting the Alcuin product. After
I submit this plan, and you have had time to review
it, let's arrange a meeting at your company to dis-
cuss these strategies.
Thanks again for giving us the opportunity to
bid on this documentation project. I hope we can
decide upon a strategy and get started as soon as
possible in order to have the manual ready in time
for the first customer shipment. I look forward to
meeting with you towards the end of next week.
Sincerely,

Fred Caslon

Fig. 5-2. Formatted Output

110 0 UNlX Text Processing 0

The .QP macro produces a paragraph indented on both sides. The pair of macros
.QS and .QE can be used to mark a section longer than one paragraph that is
indented. This is useful in reports and proposals that quote at length from another
source.
- LP
I w a s p a r t i c u l a r l y i n t e r e s t e d i n t h e f o l l o w i n g comment
found i n t h e product s p e c i f i c a t i o n :
- QS
Users f i r s t n e e d a b r i e f i n t r o d u c t i o n t o what
t h e p r o d u c t does. Sometimes t h i s i s m o r e f o r t h e
b e n e f i t o f p e o p l e who h a v e n ' t y e t b o u g h t t h e
p r o d u c t , and a r e j u s t l o o k i n g a t t h e manual.
However, i t a l s o serves t o p u t t h e r e s t o f t h e
manual, a n d t h e p r o d u c t i t s e l f , i n
t h e proper c o n t e x t .
- QE
The result of formatting is:

I was p a r t i c u l a r l y i n t e r e s t e d i n t h e f o l l o w i n g comment
found i n t h e product s p e c i f i c a t i o n :

U s e r s f i r s t n e e d a b r i e f i n t r o d u c t i o n t o what t h e
p r o d u c t does. S o m e t i m e s t h i s i s more for t h e b e n e -
f i t o f p e o p l e who h a v e n ' t yet bought t h e product,
and a r e j u s t l o o k i n g a t t h e manual. However, it
a l s o serves t o p u t t h e r e s t o f t h e m a n u a l , a n d t h e
p r o d u c t i t s e l f , i n t h e proper c o n t e x t .

Use the .QP macro inside a .QS!. QE block to break up paragraphs.

Indented Paragraphs
The I P macro produces an entire paragraph indented from the left margin. This is
especially useful for constructing lists, in which a mark of some kind (e.g., a letter or
number) extends into the left margin. We call these labeled item lists.
The .IP macro takes three arguments. The first argument i s a text label; if the
label contains spaces, it should be enclosed within quotation marks. The second argu-
ment i s optional and specifies the amount of indentation; a default of 5 is used if the
second argument is not specified. A third argument of 0 inhibits spacing before the
indented paragraph.
Item lists are useful in preparing command reference pages that describe various
syntax items, and in glossaries that present a term in one column and its definition in
the other. The following example shows a portion of the input file for a reference page:
 0 The m s Macros 0 111

.IP figure 10
is the name of a cataloged figure. If
a figure has not been cataloged, you need to use
the LOCATE command.
.IP f:p 1 0
is the scale of the
figure in relation to the page.
.IP font 10
is the two-character abbreviation or
full name of one of the available fonts
from the Alcuin library.
The following item list is produced:

I I
figure is the name of a cataloged figure. If a figure
has not been cataloged, you need to use the
LOCATE command.

f:p is the scale of the figure in relation to the

Page -
font is the two-character abbreviation or full name
of one of the available fonts from the Alcuin
library.
An .LP or .PP should be specified after the last item so that the text following the
list is not also indented.
If you want to indent the label as well as the paragraph, you can use the . i n
request around the list. The following example:
.in 10
.IP figure 1 0
is the name of a cataloged figure. If
a figure has not been cataloged, you need to use
the LOCATE command.
.in 0
will produce:

7
figure is the name of a cataloged figure.
figure has not been cataloged, you need to
use the LOCATE command. I
You can specify an absolute or relative indent. To achieve the effect of a nested list,
you can use the .RS (you can think of this as either relative start or right shift) and
.RE (relative end or retreat) macros:
112 0 UNlX Text Processing 0

.IP font 10
is the two-character abbreviation or
full name of one of the available fonts
from the Alcuin library.
.RS
.IP cu
Cursive
.IF RS
Slanted
.RS
.IF LH 5 0
Left handed
.IP RH 5 0
Right handed
.RE
.IF BL
Block
.RE
The labels on the second level are aligned with the indented left margin of paragraphs
on the first level.

font is the two-character abbreviation or full name of

one of the available fonts from the Alcuin
library.
CU Cursive
RS Slanted
LH Left handed
RH Right handed
I BL Block
One thing you need to watch out for in using the .IP macro is not to include space in
the label argument. Because of the way the macro i s coded, the space may be expanded
when the finished line is adjusted. The first line will not be aligned with the rest. For
example:
.IP "font name" 10
is the two-character abbreviation or full name . - .

r
might produce the following:

font name is the two-character abbreviation or full

name of one of the available fonts from the
Alcuin library.
 o The ms Macros 0 113

To avoid this problem, always use an unpaddable space (a backslash followed by a

space) to separate words in the label argument to .IP. This caution applies to many
other formatting situations as well.
Automatically numbered and alphabetized lists are not provided for in m s .
(Chapter 16 shows how to write your own macros for this.) However, by specifying the
number or letter as a label, you can make do with the .IP macro. For example:
User-oriented documentation recognizes three things:
.in +3n
.IP 1) 5n
that a new user needs
to learn the system in stages, getting a sense of the
system as a whole while becoming proficient in performing
particular tasks;
.IP 2) 5n
that there are different levels of users, and not
every user needs to learn all the capabilities
of the system in order to be productive;
.IP 3) 5n
that an experienced user must be able to rely on
the documentation for accurate and thorough reference
information.
.in -3n
This produces:

ented documentation recognizes three things:

that a new user needs to learn the system in

stages, getting a sense of the system as a
whole while becoming proficient in performing
particular tasks;

that there are different levels of users, and

not every user needs to learn all the capabil-
ities of the system in order to be productive;

that an experienced user must be able to rely on

the documentation for accurate and thorough
reference information.
The number is indented three ens and the text is indented five more ens. (Note: If you
are using nroff, you don’t need to specify units on the indents. However, if you are
using t r o f f, the default scaling for both the .IP macro and the .i n requests
shown in the previous example is ems. Remember that you can append a scaling indi-
cator to the numeric arguments of most macros and t r o f f requests.)
 114 UNlX Text Processing 0

Changing Font and Point Size

When you format with n r o f f and print on a line printer, you can put emphasis on
individual words or phrases by underlining or overstriking. When you are use t r o f f
and send your output to a laser printer or typesetter, you can specify variations of type,
font, and point size based on the capabilities of the output devices.

Roman, Italic, and Bold Fonts

Most typefaces have at least three fonts available: roman, bold, and italic. Normal
body copy i s printed in the roman font. You can change temporarily to a bold or italic
font for emphasis. In Chapter 4, you learned how to specify font changes using the
.f t request and inline \ f requests. The m s package provides a set of mnemonic
macros for changing fonts:

.B bold
.I italic
.R roman

Each macro prints a single argument in a particular font. You might code a single sen-
tence as follows:
.B Alcuin
revitalizes an
.I age-old
tradition.
The printed sentence has one word in bold and one in italic.

I Alcuin revitalizes an age-old tradition.

If no argument is specified, the selected font is current until it is explicitly changed:
The a r t of
.B
calligraphy
.R
is, quite simply,
T
.l

beautiful
.R
handwriting;

The example produces:

fo- calligraphy is, quite simply, beautiful handwriting;

 0 The ms Macros 0 115

You've already seen that the first argument is changed to the selected font. If you
supply a second argument, it is printed in the previous font. (You are limited to two
arguments, set off by a space; a phrase must be enclosed within quotation marks to be
taken as a single argument.) A good use for the alternate argument is to supply punc-
tuation, especially because of the restriction that you cannot begin a line with a period.
its opposite is
.B cacography .
This example produces:

I i t s opposite is cacography.

If the second argument is a word or phrase, you must supply the spacing:
The ink pen has been replaced by a
.I light " pen."
This produces:

I 1
I The ink pen has been replaced by a light pen. I
If you are using nro f f, specifying a bold font results in character overstrike; specify-
ing an italic font results in an underline for each character (not a continuous rule).
Overstriking and underlining can cause problems on some printers and terminals.
The chief advantage of these macros over the corresponding t r o f f constructs is
the ease of entry. It is easier to type:
.B calligraphy

than:
\fBcalligraphy\fP

However, you'll notice that using these macros changes the style of your input consider-
ably. As shown in the examples on the preceding pages, these macros require you to
code your input file using short lines that do not resemble the resulting filled output
text.
This style, which clearly divorces the form of the input from the form of the out-
put, is recommended by many nrof f and t rof f users. They recommend that you
use macros like these rather than inline codes, and that you begin each sentence or
clause on a new line. There are advantages in speed of editing. However, there are
others (one of the authors included) who find this style of input unreadable on the
screen, and prefer to use inline codes, and to keep the input file as readable as possible.
(There is no difference in the output file.)
116 UNIX Text Processing 0

Underlining
If you want to underline a single word, regardless of whether you are using nrof f or
t rof f, use the .UL macro:
the
.UL art
of calligraphy.

It will print a continuous rule beneath the word. You cannot specify more than a sin-
gle word with this macro.

Changing Point Size

As discussed in Chapter 4, you can change the point size and vertical spacing with the
.ps and .vs requests. However, if you do this in m s , you will find that the point
size and vertical spacing revert to 10 and 12 points, respectively, after the next para-
graph macro. This is because the paragraph macro, in addition to other tasks, resets the
point size and vertical spacing (along with various other values) to default values stored
in number registers.
The default point size and vertical spacing for a document are kept in the registers
P S and VS, respectively. If you want to change the overall point size or vertical spac-
ing, change the value in these registers. (The default values are 10 and 12, respec-
tively.) For example, to change the body type to 8 points and the spacing to 10 points,
enter the following requests at the top of your document:
.nr P S 8
.nr VS 12

At the top of a document, these settings will take effect immediately. Otherwise, you
must wait for the next paragraph macro for the new values to be recognized. If you
need both immediate and long-lasting effects, you may need a construct like:
.ps 8
.nr P S 8
.vs 12
.nr VS 12

There are also several macros for making local point size changes. The . L G macro
increases the current point size by 2 points; the .SM macro decreases the point size by
2 points. The new point size remains in effect until you change it. The .NL macro
changes the point size back to its default or normal setting. For example:

.LG
Alcuin
.NL
is a graphic arts product f o r
.SM
UNIX
.NL
systems.
 0 ThemsMacros 0 117

The following line is produced:

I I
1 Alcuin is a graphic arts product for UNIX systems. I
The .LG and .S M macros simply increment or decrement the current point size
by 2 points. Because you change the point size relative to the current setting, repeating
a macro adds or subtracts 2 more points. If you are going to change the point size by
more than 2, it makes more sense to use the - p s request. The .NL macro uses the
value of the number register PS to reset the normal point size. Its default value is 10.
In the following example, the .p s request changes the point size to 12. The
.L G and .S M macros increase and decrease the point size relative to 12 points. The
.NL macro is not used until the end because it changes the point size back to 10.
.ps 12
.L G
Alcuin
- SM
is a graphic a r t s p r o d u c t for
.SM
UNIX
.L G
systems.
.NL
It produces the following line:

I Alcuin is a graphic arts product for UNIX systems. I

A change in the point size affects how much vertical space is needed for the larger or
smaller characters. Vertical spacing i s usually 2 points larger than the point size (10 on
12). Use the vertical spacing request to temporarily change the vertical spacing, if
necessary.

Displays
A document often includes material-such as tables, figures, or equations-that are not
a part of the running text, and must be kept together on the page. In ms and mm,such
document elements are referred to generically as displays.
The macros .D S , .DE, . I D , .CD, and .LD are used to handle displays in
ms. The display macros can be relied upon to provide

adequate spacing before and after the display;

horizontal positioning of the display as a left-justified, indented, or centered
block;
proper page breaks, keeping the entire display together.
118 UNlX Text Processing 0

The default action of the .DS macro is to indent the block of text without filling lines:
Some of t h e typefaces that are currently available are:
.DS
Roman
Caslon
Baskerville
Helvet i ca
-DE
This produces:

Some of t h e typefaces that are currently available are:

Roman
Caslon
Baskerville
Helvetica

You can select a different format for a display by specifying a left-justified or

centered display with one of the following arguments:

I Indented (default)
L Left-justified
C Center each line
B Block (center entire display)

The L argument can be used for formatting an address block in a letter:

.DS L
Mr. John F u s t
Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02154
-DE
The display macro prevents these lines from being filled; it “protects” the carriage
returns as they were entered in the file.
A display can be centered in two ways: either each individual line in the display
is centered (c),or the entire display is centered as a block (B) based on the longest line
of the display.
The use of tabs often presents a problem outside of displays. Material that has
been entered with tabs in the input file should be formatted in no-fill mode, the default
setting of the display macros. The following table was designed using tabs to provide
the spacing.
 The m s Macros 0 119

.DS L
Dates Description of Task

June 30 Submit audience analysis

July 2 Meeting t o review audience analysis
July 15 Submit detailed outline
August 1 Submit first draft
August 5 Return of first draft
August 8 Meeting t o review comments
and establish revisions
.DE
This table appears in the output just as it looks in the file. If this material had not been
processed inside a display, the columns would be improperly aligned.

Static and Floating Displays

One of the basic functions of a display is to make sure the displayed material stays
together on one page. If the display is longer than the distance to the bottom of the
page, there is a page break.
If the display is large, causing a page break can leave a large block of white space
at the bottom of the page. To avoid this problem, m s provides a set of macros for
floating displays, as well as macros for the static displays we’ve already discussed. If a
floating display doesn’t fit on the page, the formatter doesn’t force a page break.
Instead, it simply holds the displayed text in reserve while it fills up the remainder of
the page with the text following the display. It prints the display at the top of the next
page, then continues where it left off.
We have already used .DS and . D E to mark the beginning and end of a static
display. To specify a floating display, the closing mark is the same but the beginning is
marked by a different macro:

.ID Same as .DS I(indented) but floating

LD Same as .DS L (left justified) but floating
.CD Same as .DS C (center each line) but floating
.BD Same as .D S B (center display) but floating
In the following example of an input file, numbers are used instead of actual lines
of text to make the placement of the display more obvious:

2
3
4
5
.LD
Long Display
.DE
6
120 0 UNlX Text Processing 0

7
8
9
10

The following two formatted pages might be produced, assuming that there are a suffi-
cient number of lines to cause a page break:

-1- -2 -

Long Display

8
9
10

If there had been room on page 1 to fit the display, it would have been placed there, and
lines 6 and 7 would have followed the display, as they did in the input file.
If a static display had been specified in the previous example, the display would
be placed in the same position on the second page, and lines 6 and 7 would have fol-
lowed it, leaving extra space at the bottom of page 1. A floating display attempts to
make the best use of the available space on a page.
The formatter maintains a queue to hold floating displays that it has not yet out-
put. When the top of a page is encountered, the next display in the queue is output.
The queue is emptied in the order in which it was filled (first in, first out).
The macros called by the display macros to control output of a block of text are
available for other uses. They are known as “keep and release” macros. The pair
.K S / . KE keep a block together and output it on the next available page. The pair
.KF/. KE specify a floating keep; the block saved by the keep can float and lines of
text following the block may appear before it in the text.

Headings
In m s , you can have numbered and unnumbered headings. There are two heading
macros: .NH for numbered headings and .S H for unnumbered section headings.
Let’s first look at how to produce numbered headings. The syntax for the .Nfl
macro is:
.NH [level]
[heading text]
. LP
 The ms Macros 0 121

(The brackets indicate optional arguments.) You can supply a numerical value indicat-
ing the level of the heading. If no value is provided for level, then a top-level heading
is assumed. The heading text begins on the line following the macro and can extend
over several lines. You have to use one of the paragraph macros, either .LP or - PP,
after the last line of the heading. For example:
.N H
Quick Tour of Alcuin
-LP
The result is a heading preceded by a first-level heading number:

r u i c k Tour of Alcuin 1
The next time you use this macro the heading number will be incremented to 2, and
after that, to 3.
You can add levels by specifying a numeric argument. A second-level heading is
indicated by 2:
.NH 2
Introduction t o Calligraphy
.LP
The first second-level heading number is printed:
1.1 Introduction t o Calligraphy

When another heading is specified at the same level, the heading number is automati-
cally incremented. If the next heading i s at the second level:
.NH 2
Digest of Alcuin Commands
.LP

m s produces:

L D D i g e s t o f Alcuin Commands

Each time you go to a new level, . 1 is appended to the number representing the exist-
ing level. That number is incremented for each call at the same level. When you back
out of a level (for instance, when you go from level 5 to 4) the counter for the level (in
this case level 5) is reset to 0.
The macro for unnumbered headings is SH:
.SH
Introduction t o Calligraphy
-LP
Unnumbered headings and numbered headings can be intermixed without affecting the
numbering scheme:
122 0 UNIX Text Processing 0

1. Quick Tour of A l c u i n

Introduction to Calligraphy

1.1 D i g e s t o f A l c u i n Commands

Headings are visible keys to your document’s structure. Their appearance can
contribute significantly to a reader recognizing that organization. If you are using
unnumbered headings, it becomes even more important to make headings stand out. A
simple thing you can do is use uppercase letters for a first-level heading.

Cover Sheet Macros -

In their original incarnation at Bell Laboratories, the m s macros were called on to for-
mat many internal AT&T documents. Accordingly, it is not surprising that there were
quite a few macros that controlled the format of specific internal document types. What
i s surprising is that these macros are still present in copies of the m s macros distributed
outside of AT&T.
You have the option of specifying that your document contains Engineer’s Notes
(.EG), an Internal Memorandum (. I M ) , a Memorandum for Record (.MR), a
Memorandum for File ( MF), a Released Paper ( RP), a Technical Reprint ( * TR), or a
letter (. LT).
Many of these formats are quite useless outside of AT&T, unless you customize
them heavily for other institutions. We prefer simply to ignore them.
In general, what these document type macros control is the appearance of the
document’s cover sheet. The content of that cover sheet is specified using the following
macros:

.T L Title
.AU Author
.AI Author’s Institution
.AB Abstract Start
.AE Abstract End

These macros are general enough that you can still use them even if you aren’t from
Bell Laboratories.
Each macro takes its data from the following line(s) rather than from an argument.
They are typically used together. For example:
- TL
UNIX Text P r o c e s s i n g
- AU
Dale D o u g h e r t y
- AU
Tim O‘Reilly
 0 The ms Macros 0 123

.AI
O’Reilly & Associates, Inc.
.AB
This book provides a comprehensive introduction t o t h e major
UNIX text-processing tools. It includes a discussion of
vi, ex, nroff, and troff, as
well as many other text-processing programs.
.AE
-LP
Exactly how the output will look depends on which document types you have selected.
If you don’t specify any of the formats, you will get something like this:

I-- UNIX Text Processing

Dale Dougherty
Tim 0 ’Reilly
O’Reilly & Associates, Inc.

ABSTRACT
This book provides a comprehensive introduction to
the major UNIX text-processing tools. It includes a
discussion of v i , e x , nroff, and troff, as
well as many other text-processing programs.
You can specify as many title lines as you want following .TL. The macro will be
terminated by any of the other cover sheet macros, or by any paragraph macro. For
multiple authors, .A U and .A I can be repeated up to nine times.
The cover sheet isn’t actually printed until a reset (such as that caused by any of
the paragraph macros) is encountered, so if you want to print only a cover page, you
should conclude it with a paragraph macro even if there i s no following text.
In addition, if you use these macros without one of the overall document type
macros like .RP, the cover sheet will not be printed separately. instead, the text will
immediately follow. insert a .bp if you want a separate cover sheet.

Miscellaneous Features

Putting Information in a Box

Another way of handling special information i s to place it in a box. Individual words
can be boxed for emphasis using the .B X command:
124 0 UNlX Text Processing 0

T o m o v e t o t h e n e x t menu, press t h e
.BX RETURN
key -
This draws a box around the word RETURN.
T o m o v e t o t h e n e x t menu, press t h e
IRETURN I
key.

As you can see, it might be a good idea to reduce the point size of the boxed word.
You can enclose a block of material within a box by using the pair of macros
.B1 and .B2:
.B1
.B
.ce
Note t o Reviewers
.R
- LP
Can y o u g e t a copy o f a m a n u s c r i p t w i t h o u t a n n o t a t i o n s ?
It seems t o m e t h a t you s h o u l d be
a b l e t o mark up a page w i t h comments o r
o t h e r scribbles w h i l e i n A n n o t a t i o n M o d e and
s t i l l o b t a i n a p r i n t e d copy w i t h o u t t h e s e m a r k s .
Any i d e a s ?
- SP
.B 2
This example produces the following boxed section in t ro f f:
Note to Reviewers
Can you get a copy of a manuscript without annotations? It seems to me that you
should be able to mark up a page with comments or other scribbles while in Annota-
tion Mode and still obtain a printed copy without these marks. Any ideas?

You may want to place boxed information inside a pair of keep or display macros. This
will prevent the box macro from breaking if it crosses a page boundary. If you use
these macros with n r o f f, you must also pipe your output through the c o l postpro-
cessor as described in Chapter 4.

Footnotes
Footnotes present special problems-the main is printing the text at the bottom of the
page. The .F S macro indicates the start of the text for the footnote, and .F E indi-
cates the end of the text for the footnote. These macros surround the footnote text that
will appear at the bottom of the page. The .F S macro i s put on the line immediately
following some kind of marker, such as an asterisk, that you supply in the text and in
the footnote.
 0 Thems Macros a 125

... in a n article on desktop publishing.*

.FS
* "Publish or Perish: Start-up grabs early page language
lead," Computerworld, April 21, 1986, p - 1.
-FE

All the footnotes are collected and output at the bottom of each page underneath a short
rule. The footnote text is printed in smaller type, with a slightly shorter line length then
the body text. However, you can change these if you want.
Footnotes in m s use an nrof f /trof f feature called environments (see
Chapter 14), so that parameters like line length or font that are set inside a footnote are
saved independently of the body text. So, for example, if you issued the requests:
.F S
.ft B
-11 -5n
.in +5n
Some text
-
-
-
.FE
the text within the footnote would be printed in boldface, with a 5-en indent, and the
line length would be shortened by 5 ens. The text following the footnote would be
unaffected by those formatting requests. However, the next time a footnote was called,
that special formatting would again be in effect.

I I
*"Publish or Perish: Start-up grabs early page language
lead," Computerworld, April 21, 1986, p. 1.

If a footnote is too long to fit on one page, it will be continued at the bottom of the next
page.

Two-Column Processing
One of the nice features of the m s macros is the ease with which you can create multi-
ple columns and format documents, such as newsletters or data sheets, that are best
suited to a multicolumn format.
To switch to two-column mode, simply insert the .2 C macro. To return to
single-column mode, use . 1 C . Because of the way two-column processing works in
m s , you can switch to two-column mode in the middle of a page, but switching back to
a single column forces a page break. (You'll understand the reason for this when we
return to two-column processing in Chapter 16.)
The default column width for two-column processing i s 7/15th of the line length.
It is stored in the register CW (column width). The gutter between the columns is
126 0 UNlX Text Processing 0

1/15th of the line length, and is stored in the register GW (gutter width). By changing
the values in these registers, you can change the column and gutter width.
For more than two columns, you can use the .MC macro. This macro takes two
arguments, the column width and the gutter width, and creates as many columns as will
fit in the line length. For example, if the line lengths are 7 inches, the request:
.MC 2i .3i

would create three columns 2 inches wide, with a gutter of .3 inches between the
columns.
Again, .1 C can be used to return to single-column mode. In some versions of
m s , the . R C macro can be used to break columns. If you are in the left column, fol-
lowing text will go to the top of the next column. If you are in the right column, .RC
will start a new page.

Page Headers and Footers .

When you format a page with m s , the formatter is instructed to provide several lines at
the top and the bottom of the page for a header and a footer. Beginning with the
second page, a page number appears on a single line in the header and only blank lines
are printed for the footer.
The m s package allows you to define strings that appear in the header or footer.
You can place text in three locations in the header or footer: left justified, centered, and
right justified. For example, we could place the name of the client, the title of the
document, and the date in the page header and we could place the page number in the
footer.

.ds LH GGS
.ds CH Alcuin Project P r o p o s a l
.ds RH \ * ( D Y
.ds CF P a g e %

You may notice that we use the string DY to supply today’s date in the header. In the
footer, we use a special symbol (%) to access the current page number. Here are the
resulting header and footer:

Alcuin Project P r o p o s a l

Page 2 1
April 26, 1987

Normally, you would define the header and footer strings at the start of the document,
so they would take effect throughout. However, note that there is nothing to prevent
you from changing one or more of them from page to page. (Changes to a footer string
 0 Thems Macros 0 127

will take effect on the same page; changes to a header string will take effect at the top
of the next page.)

Problems on the First Page

Because m s was originally designed to work with the cover sheet macros and one of
the standard Bell document types, there are a number of problems that can occur on the
first page of a document that doesn’t use these macros.*
First, headers are not printed on the first page, nor is it apparent how to get them
printed there if you want them. The trick is to invoke the internal .NP (new page)
macro at the top of your text. This will not actually start a new page, but will execute
the various internal goings-on that normally take place at the top of a page.
Second, it is not evident how to space down from the top if you want to start your
text at some distance down the page. For example, if you want to create your own title
page, the sequence:
- s p 3i
.ce
\sl6The Invention of Movable Type\sO

will not work.

The page top macro includes an .n s request, designed to ensure that all leftover
space from the bottom of one page doesn’t carry over to the next, so that all pages start
evenly. To circumvent this on all pages after the first one, precede your spacing request
with an r s (restore spacing) request. On the first page, a .f 1 request must precede
a .rs request.

Extensions toms
In many ways, m s can be used to give you a head start on defining your own macro
package. Many of the features that are missing in m s can be supplied by user-defined
macros. Many of these features are covered in Chapters 14 through 18, where, for
example, we show macros for formatting numbered lists.

*This problem actually can occur on any page, but is most frequently encountered on the first page.
 C H A P T E R

The mm Macros

A macro package provides a way of describing the format of various kinds of docu-
ments. Each document presents its own specific problems, and macros help to provide
a simple and flexible solution. The mm macro package is designed to help you format
letters, proposals, memos, technical papers, and reports. A text file that contains mm
macros can be processed by either n r o f f or t rof f, the two text formatting pro-
grams in UNIX. The output from these programs can be displayed on a terminal screen
or printed on a line printer, a laser printer, or a typesetter.
Some users of the mm macro package learn only a few macros and work produc-
tively. Others choose from a variety of macros to produce a number of different for-
mats. More advanced users modify the macro definitions and extend the capabilities of
the package by defining their own special-purpose macros.
Macros are the words that make up a format description language. Like words,
the result of a macro is often determined by context. That is, you may not always
understand your output by looking up an individual macro, just like you may not under-
stand the meaning of an entire sentence by looking up a particular word. Without exa-
mining the macro definition, you may find it hard to figure out which macro is causing
a particular result. Macros are interrelated; some macros call other macros, like a sub-
routine in a program, to perform a particular function.
After finding out what the macro package allows you to do, you will probably
decide upon a particular format that you like (or one that has evolved according to the
decisions of a group of people). To describe that format, you are likely to use only a
few of the macros, those that do the job. In everyday use, you want to minimize the
number of codes you need to format documents in a consistent manner.

Formatting a Text File

To figure out the role of a macro package such as mm,it may help to consider the dis-
tinction between formatting and format. Formatting i s an operation, a process of sup-
plying and executing instructions. You can achieve a variety of results, some pleasing,

128
 ThemMacros 0 129

some not, by any combination of formatting instructions. A format is a consistent pro-

duct, achieved by a selected set of formatting instructions. A macro package makes it
possible for a format to be recreated again and again with minimal difficulty. It
encourages the user to concentrate more on the requirements of a document and less on
the operations of the text formatter.
Working with a macro package will help reduce the number of formatting instruc-
tions you need to supply. This means that a macro package will take care of many
things automatically. However, you should gradually become familiar with the opera-
tions of the n r o f f/tr o f f formatter and the additional flexibility it offers to define
new formats. If you have a basic understanding of how the formatter works, as
described in Chapter 4, you will find it easier to learn the intricacies of a macro pack-
age.

Invoking n r o f f /trof f with nun

The mm command is a shell script that invokes the n r o f f formatter and reads in the
files that contain the mm macro definitions before processing the text file(s) specified
on the command line.
$ mm option(s)filenarne(s)
If more than one file is specified on the command line, the files are concatenated before
processing. There are a variety of options for invoking preprocessors and postproces-
sors, naming an output device, and setting various number registers to alter default
values for a document. Using the m m command is the same as invoking n r o f f
explicitly with the -mm option.
Unless you specify otherwise, the mm command sets n r o f f ’ s -T option to
the terminal type set in your login environment. By default, output is sent to the termi-
nal screen. If you have problems viewing your output, or if you have a graphics termi-
nal, you may want to specify another device name using the -T option. For a list of
available devices, see Appendix B. The mm command also has a - c option, which
invokes the col filter to remove reverse linefeeds, and options to invoke t b l (-t)
and e q n (-e).
When you format a file to the screen, the output usually streams by too swiftly to
read, just as when you c a t a file to the screen. Pipe the output of the mm command
through either of the paging programs, pg or more, to view one screenful at a time.
This will give you a general indication that the formatting commands achieved the
results you had expected. To print a file formatted with mm, simply pipe the output to
the print spooler (e.g., lp) instead of to a screen paging program.
Many of the actions that a text formatter performs are dependent upon how the
document is going to be printed. If you want your document to be formatted with
t r o f f instead of n r o f f , use the mmt command (another shell script) or invoke
t r o f f directly, using the -mm option. The mmt command prepares output for laser
printers and typesetters. The formatted output should be piped directly to the print
spooler (e.g., lp) or directed to a file and printed separately. You will probably need
to check at your site for the proper invocation of mmt if your site supports more than
one type of laser printer or typesetter.
130 0 UNlX Text Processing 0

If you are using o t r o f f , be sure you don’t let t r o f f send the output to your
terminal because, in all probability, it will cause your terminal to hang, or at least to
scream and holler.
In this chapter, we will generally show the results of the mm command, rather
than mmt-that is, we’ll be showing n r o f f rather than t r o f f . Where the subject
under discussion is better demonstrated by t r o f f , we will show t r o f f output
instead. We assume that by now, you will be able to tell which of the programs has
been used, without our mentioning the actual commands.

Problems in Getting Formatted Output

When you format an mm-coded document, you may only get a portion of your format-
ted document. Or you may get none of it. Usually, this is because the formatter has
had a problem executing the codes as they were entered in the input file. Most of the
time it is caused by omitting one of the macros that must be used in pairs.
When formatting stops like this, one or more error messages might appear on your
screen, helping you to diagnose the problems. These messages refer to the line numbers
in the input file where the problems appear to be, and try to tell you what is missing:

ERROR:(filename) line number

Error message

Sometimes, you won’t get error messages, but your output will break midway. Gen-
erally, you have to go in the file at the point where it broke, or before that point, and
examine the macros or a sequence of macros. You can also run a program on the input
file to examine the code you have entered. This program, available at most sites, is
called checkmm.

Default Formatting
In Chapter 4, we looked at a sample letter formatted by n r o f f . It might be interest-
ing, before putting any macros in the file, to see what happens if we format l e t t e r
as it is, this time using the mm command to read in the mm macro package.
Refer to Figure 6-1 and note that

a page number appears in a header at the top of the page;

the address block still forms two long lines;
lines of input text have been filled, forming block paragraphs;
the right margin is ragged, not justified as with n r o f f ;
the text is not hyphenated
. space has been allocated for a page with top, bottom, left, and right margins.
 0 ThemmMacros 0 131

- 1 -

April 1 , 1987

Mr. John Fust Vice President, Research and

Development Gutenberg Galaxy Software Waltham,
Massachusetts 02159

Dear Mr. Fust:

In our conversation last Thursday, we discussed a

documentation project that would produce a user's
manual on the Alcuin product. Yesterday, I
received the product demo and other materials that
you sent me.

Going through a demo session gave me a much better

understanding of the product. I confess to being
amazed by Alcuin. Some people around here,
looking over my shoulder, were also astounded by
the illustrated manuscript I produced with Alcuin.
One person, a student of calligraphy, was really
impressed.

In the next couple of days, I'll be putting

together a written plan that presents different
strategies for documenting the Alcuin product.
After I submit this plan, and you have had time to
review it, let's arrange a meeting at your company
to discuss these strategies.

Thanks again for giving us the opportunity to bid

on this documentation project. I hope we can
decide upon a strategy and get started as soon as
possible in order to have the manual ready in time
for the first customer shipment. I look forward to
meeting with you towards the end of next week.

Sincerely,

Fred Caslon

Fig. 6-1. A Raw mm-formatted File

132 UNlX Text Processing 0

Page Layout
When you format a page with mm, the formatter is instructed to provide several lines at
the top and the bottom of the page for a header and a footer. By default, a page number
appears on a single line in the header and only blank lines are printed for the footer.
There are basically two different ways to change the default header and footer.
The first way is to specify a command-line parameter with the mm or m m t commands
to set the number register N. This allows you to affect how pages are numbered and
where the page number appears. The second way is to specify in the input file a macro
that places text in the header or footer. Let’s look at both of these techniques.

Setting Page Numbering Style

When you format a document, pages are numbered in sequence up to the end of the
document. This page number is usually printed in the header, set off by dashes.
-1-

Another style of page numbering, used in documents such as technical manuals,

numbers pages specific to a section. The first page of the second section would be
printed as:
2-1

The other type of change affects whether or not the page number is printed in the
header at the top of the first page.
The number register N controls these actions. This register has a default setting
of 0 and can take values from 0 through 5. Table 6-1 shows the effect of these values.

TABLE 6-1. Page Number Styles, Register N

Value Action
0 The page number prints in the header on all pages.
This is the default page numbering style.
1 On page 1 , the page number is printed in place of
the footer.
2 On page 1 , the page number in not printed.
3 All pages are numbered by section, and the page
number appears in the footer. This setting affects
the defaults of several section-related registers and
macros. It causes a page break for a top-level head-
ing (E j=l), and invokes both the .F D and .RP
macros to reset footnote and reference numbering.
 0 ThemmMacros 0 133

TABLE 6-1. -(Cont'd)

Value Action
4 The default header containing the page number is
suppressed, but it has no effect on a header supplied
by a page header macro.
5 All pages are numbered by section, and the page
number appears in the footer. In addition, labeled
displays (.FC, .TB, .EX, and .EC) are also
numbered by section.

The register N can be set from the command line using the -r option. If we set
it to 2, no page number will appear at the top of page 1 when we print the sample letter:
$ ram -rN2 letter I lp

Header and Footer Macros

The mm package has a pair of macros for defining what should appear in a page header
(. PH) and a page footer ( -PF). There is also a set of related macros for specifying
page headers and footers for odd-numbered pages (.OH and .OF) or for even-
numbered pages (. EH and - EF). All of these macros have the same form, allowing
you to place text in three places in the header or footer: left justified, centered, and right
justified. This is specified as a single argument in double quotation marks, consisting of
three parts delimited by single quotation marks.
' left' center' right'
For e x a m p l e , w e c o u l d p l a c e t h e name o f a c l i e n t , t h e
t i t l e o f t h e document, and t h e date i n t h e p a g e h e a d e r ,
a n d w e c o u l d p l a c e t h e p a g e number i n t h e f o o t e r :
.P H"'GGS' Alcuin Project Proposal' \* (DT' "
.PF "''Page % '"'
You may notice that we use the string DT to supply today's date in the header. The
following header appears at the top of the page.

GGS Alcuin Project Proposal I

April 26, 1987

In the footer, we use a special symbol (%)to access the current page number. Only text
to be centered was specified; however, the four delimiters were still required to place
the text correctly. This footer appears at the bottom of the page:
134 0 UNlX Text Processing 0

Page 2

The header and footer macros override the default header and footer.
1
Setting Other Page Control Registers
The mm package uses number registers to supply the values that control line length,
page offset, point size, and page length, as shown in Table 6-2.

TABLE 6-2. Number Registers

Register Contains t r o f f Default n r o f f Default

0 Page offset (left margin) -75 Si
N Page numbering style 0 0
P Page length 66v 66 lines
S Point size ( t r o f f only) 10 NA
W Line length or width 6i 60

These registers must be defined before the mm macro package is read by nrof f
or t r o f f . Thus, they can be set from the command line using the -r option, as we
showed when we gave a new value for register N. Values of registers 0 and W for
n r o f f must be given in character positions (depending on the character size of the
output device for nrof f , S i might translate as either 5 or 6 character positions), but
t r o f f can accept any of the units descibed in Chapter 4. For example:
$ mm -rN2 -rW65 -r10 file
but:
$ rnmt -rN2 -rW6.5i -rOli file
Or the page control registers can be set at the top of your file, using the .s o request to
read in the m m macro package, as follows:
.nr N 2
.nr W 65
.nr 0 10
.so /usr/lib/tmac/tmac.m

If you do it this way, you cannot use the mm command. Use n r o f f or t r o f f

without the -mm option. Specifying -mm would cause the mm macro package to be
read twice; mrn would trap that error and bail out.
 ThemMacros 0 135

Paragraphs
The .P macro marks the beginning of a paragraph.
.P
In our conversation last Thursday, we discussed a
This macro produces a left-justified, block paragraph. A blank line in the input file also
results in a left-justified, block paragraph, as you saw when we formatted an uncoded
file.
However, the paragraph macro controls a number of actions in the formatter,
many of which can be changed by overriding the default values of several number regis-
ters. The .P macro takes a numeric argument that overrides the default paragraph
type, which is a block paragraph. Specifying 1 results in an indented paragraph:
.P 1
Going through a demo session gave me a much better
The first three paragraphs formatted for the screen follow:

In our conversation last Thursday, we discussed a

documentation project that would produce a user's manual
on the Alcuin product. Yesterday, I received the product
demo and other materials that you sent me.

Going through a demo session gave me a much better

understanding of the product. I confess to being amazed
by Alcuin. Some people around here, looking over my
shoulder, were also astounded by the illustrated
manuscript I produced with Alcuin. One person, a student
of calligraphy, was really impressed.

In the next couple of days, I'll be putting together a

written plan that presents different strategies for
documenting the Alcuin product. After I submit this plan,
and you have had time to review it, let's arrange a
meeting at your company to discuss these strategies.

The first line of the second paragraph is indented five spaces. (In t r o f f the default
indent is three ens.) Notice that the paragraph type specification changes only the
second paragraph. The third paragraph, which is preceded in the input file by .P
without an argument, i s a block paragraph.
If you want to create a document in which all the paragraphs are indented, you
can change the number register that specifies the default paragraph type. The value of
P t is 0 by default, producing block paragraphs. For indented paragraphs, set the value
of P t to 1. Now the .P macro will produce indented paragraphs.
.nr Pt 1
If you want to obtain a block paragraph after you have changed the default type,
specify an argument of 0:
136 0 UNlX Text Processing 0

.P 0

When you specify a type argument, it overrides whatever paragraph type is in effect.
There is a third paragraph type that produces an indented paragraph with some
exceptions. If Pt is set to 2, paragraphs are indented except those following section
headings, lists, and displays. It is the paragraph type used in this book.
The following list summarizes the three default paragraph types:

0 Block
1 Indented
2 Indented with exceptions

Vertical Spacing
The paragraph macro also controls the spacing between paragraphs. The amount of
space is specified in the number register P s. This amount differs between nrof f
and troff.
With nrof f, the .P macro has the same effect as a blank line, producing a full
space between paragraphs. However, with t r o f f , the .P macro outputs a blank
space that i s equal to one-half of the current vertical spacing setting. Basically, this
means that a blank line will cause one full space to be output, and the .P macro will
output half that space.
The P macro invokes the .SP macro for vertical spacing. This macro take a
numeric argument requesting that many lines of space.
Sincerely,
.SP 3
F r e d Caslon

Three lines of space will be provided between the salutation and the signature lines.
You do not achieve the same effect if you enter - SP macros on three consecu-
tive lines. The vertical space does not accumulate and one line of space is output, not
three.
Two or more consecutive .SP macros with numeric arguments results in the
spacing specified by the greatest argument. The other arguments are ignored.
.SP 5
.SP
.SP 2
In this example, five lines are output, not eight.
Because the .P macro calls the .SP macro, it means that two or more consecu-
tive paragraph macros will have the same effect as one.
 0 ThemmMacros 0 137

The .SP Macro versus the .sp Request

There are several differences between the .SP
macro and the .s p request. A series
of .s p requests does cause vertical spacing to accumulate. The following three
requests produce eight blank lines:
.sp 5
-SP
.sp 2

The argument specified with the .SP macro cannot be scaled nor can it be a
negative number. The .SP macro automatically works in the scale (v) of the current
vertical spacing. However, both .SP and .s p accept fractions, so that each of the
following codes has the same result:
.sp . 3 v .SP . 3 .sp . 3

Justification .
A document formatted by n r o f f with mm produces, by default, unjustified text (an
uneven or ragged-right margin). When formatted by t r o f f , the same document is
automatically justified (the right margin is even).
If you are using both n r o f f and t r o f f , it is probably a good idea to expli-
citIy set justification on or off rather than depend upon the default chosen by the for-
matter. Use the .S A macro (set adjustment) to set document-wide justification. An
argument of 0 specifies no justification; 1 specifies justification.
If you insert this macro at the top of your file:
.SA 1

both n r o f f and t r o f f will produce right-justified paragraphs like the following:

In our conversation last Thursday, w e discussed

a documentation project that would produce a user's
manual on t h e Alcuin product. Yesterday, I received t h e
product demo and other materials that you sent me.

Word Hyphenation
One way to achieve better line breaks and more evenly filled lines is to instruct the for-
matter to perform word hyphenation.
Hyphenation is turned off in the m m macro package. This means that the for-
matter does not try to hyphenate words to make them fit on a line unless you request it
by setting the number register H y to 1. If you want the formatter to automatically
hyphenate words, insert the following line at the top of your file:
138 UNlX Text Processing 0

.nr Hy 1
Most of the time, the formatter breaks up a word correctly when hyphenating. Some-
times, however, it does not and you have to explicitly tell the formatter either how to
split a word (using the .h y request) or not to hyphenate at all (using the .nh
request).

Displays
When we format a text file, the line breaks caused by carriage returns are ignored by
n r o f f / t r o f f . How text is entered on lines in the input file does not affect how
lines are formed in the output. It doesn’t really matter whether information is typed on
three lines or four; it appears the same after formatting.
You probably noticed that the name and address at the beginning of our sample
file did not come out in block form. The four lines of input ran together and produced
two filled lines of output:
Mr. John Fust Vice President, Research and Development
Gutenberg Galaxy Software Waltham, Massachusetts 02159

The formatter, instead of paying attention to carriage returns, acts on specific macros or
requests that cause a break, such as .P, .SP, or a blank line. The formatter request
b r is probably the simplest way to break a line:
Mr. John Fust
.br
Vice President, Research and Development
The .b r request is most appropriate when you are forcing a break of a single line.
For larger blocks of text, the mrn macro package provides a pair of macros for indicat-
ing that a block of text should be output just as it was entered in the input file. The
.DS (display start) macro is placed at the start of the text, and the . D E (display end)
macro is placed at the end:
.DS
Mr. John Fust
Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02159
-DE
The formatter does not fill these lines, so the address block is output on four lines, just
as it was typed. In addition, the . D E macro provides a line of space following the
display.

Our Coding Efforts, So Far

W e have pretty much exhausted what we can do using the sample letter. Before going
on to larger documents, you may want to compare the coded file in Figure 6-2 with the
n r o f f-formatted output in Figure 6-3. Look them over and make sure you understand
what the different macros are accomplishing.
 0 ThemmMacros 0 139

.nr Pt 1
.SA 1
April 1, 1987
.SP 2
.DS
Mr. John Fust
Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02159
.DE
Dear Mr. Fust:
.P
In our conversation last Thursday, we discussed a
documentation project that would produce a user's manual
on the Alcuin product. Yesterday, I received the product
demo and other materials that you sent me.
.P
Going through a demo session gave me a much better
understanding of the product. I confess to being amazed
by Alcuin. Some people around here, looking over my
shoulder, were also astounded by the illustrated
manuscript I produced with Alcuin. One person, a student
of calligraphy, was really impressed.
.P
In the next couple of days, I'll be putting together a
written plan that presents different strategies f o r
documenting the Alcuin product. After I submit this plan,
and you have had time to review it, let's arrange a
meeting at your company to discuss these strategies.
.P
Thanks again f o r giving us the opportunity to bid on this
documentation project. I hope we can decide upon a
strategy and get started as soon as possible in order to
have the manual ready in time for the first customer
shipment. I look forward to meeting with you towards the
end of next week.
.SP
Sincerely,
.SP 2
Fred Caslon

Fig. 6-2. Coded File

140 UNlX Text Processing 0

- 1 -

April 1, 1987

Mr. John Fust

Vice President, Research and Development
Gutenberg Galaxy Software
Waltham, Massachusetts 02159

Dear Mr. Fust:

In our conversation last Thursday, we

discussed a documentation project that would
produce a user's manual on the Alcuin product.
Yesterday, I received the product demo and other
materials that you sent me.

Going through a demo session gave me a much

better understanding of the product. I confess to
being amazed by Alcuin. Some people around here,
looking over my shoulder, were also astounded by
the illustrated manuscript I produced with Alcuin.
One person, a student of calligraphy, was really
impressed.

In the next couple of days, I'll be putting

together a written plan that presents different
strategies for documenting the Alcuin product.
After I submit this plan, and you have had time to
review it, let's arrange a meeting at your company
to discuss these strategies.

Thanks again for giving us the opportunity to

bid on this documentation project. I hope we can
decide upon a strategy and get started as soon as
possible in order to have the manual ready in time
for the first customer shipment. I look forward to
meeting with you towards the end of next week.

Sincerely,

Fred Caslon

Fig. 6-3. Formatted Output

 0 ThemmMacros 0 141

We have worked through some of the problems presented by a very simple one-
page letter. A s we move on, we will be describing specialized macros that address the
problems of multiple page documents, such as proposals and reports. In many ways,
the macros for more complex documents are the feature performers in a macro package,
the ones that really convince you that a markup language is worth learning.

Changing Font and Point Size

When you format with n r o f f and print on a line printer, you can put emphasis on
individual words or phrases by underlining or overstriking. When you are using
t r o f f and send your output to a laser printer or typesetter, you can specify variations
of type, font, and point size based on the capabilities of the output device.

Roman, Italic, and Bold Fonts

Most typefaces have at least three fonts available: roman, bold, and italic. Normal
body copy is printed in the roman font. You can change temporarily to a bold or italic
font for emphasis. In Chapter 4, you learned how to specify font changes using the
.f t request and inline \f requests. The mm package provides a set of mnemonic
macros for changing fonts:

.B Bold
.I Italic
.R Roman

Each macro prints a single argument in a particular font. You might code a single sen-
tence as follows:
.B Alcuin
revitalizes an
.I age-old
tradition.

The printed sentence has a word in bold and one in italic. (In nroff, bold space is
simulated by overstriking, and italics by underlining.)

AIcuin revitalizes an age-old tradition. 1

142 UNlX Text Processing 0

If no argument i s specified, the selected font is current until it is explicitly changed:

The art of
.E
calligraphy
.R
is, quite simply,

beaut i fu 1
.R
handwriting;

The previous example produces:

=of calligraphy is, quite simply, beautiful handwriting;

You've already seen that the first argument is changed to the selected font. If you
supply a second argument, it is printed in the previous font. Each macro takes up to six
arguments for alternating font changes. (An argument is set off by a space; a phrase
must be enclosed within quotation marks to be taken as a single argument.) A good use
for the alternate argument is to supply punctuation, especially because of the restriction
that you cannot begin an input line with a period.
its opposite is
.B cacography .
This example produces:

I its opposite is cacography. 1

If you specify alternate arguments consisting of words or phrases, you must supply the
spacing:
The ink pen has been replaced by a
.I light " pen."

This produces:

=pepen has been replaced by a light pen. 1

Here's an example using all six arguments:
Alcuin uses three input devices, a
.E "light pen" ", a " "mouse" ", and a " "graphics tablet."

This produces:

Alcuin uses three input devices, a light pen, a mouse, and a graphics tablet.
 0 ThemmMacros 0 143

There are additional macros for selecting other main and alternate fonts. These macros
also take up to six arguments, displayed in alternate fonts:

.BR Alternate bold and roman

.I B Alternate italic and bold
.R I Alternate roman and italic
.BI Alternate bold and italic
- IR Alternate italic and roman
.RB Alternate roman and bold

If you are using n r o f f , specifying a bold font results in character overstrike; specify-
ing an italic font results in an underline for each character (not a continuous rule).
Overstriking and underlining can cause problems on some printers and terminals.

Changing Point Size

When formatting with t ro f f , you can request a larger or smaller point size for the
type. A change in the point size affects how much vertical space is needed for the
larger or smaller characters. Normal body copy is set in 10-point type with the vertical
spacing 2 points larger.
You learned about the .p s (point size) and .v s (vertical spacing) requests in
Chapter 4. These will work in mm; however, mm also has a single macro for changing
both the point size and vertical space:

.S [point size] [vertical spacing]

The values for point size and vertical spacing can be set in relation to the current set-
ting: + increments and - decrements the current value. For example, you could
specify relative point size changes:
.s + 2 +2

or absolute ones:
.s 12 1 4

By default, if you don’t specify vertical spacing, a relation of 2 points greater than the
point size will be maintained. A null value (””) does not change the current setting.
The new point size and vertical spacing remain in effect until you change them.
Simply entering the .S macro without arguments restores the previous settings:
.s
The mm package keeps track of the default, previous, and current values, making it
easy to switch between different settings using one of these three arguments:

D Default
P Previous
C Current
144 0 UNIX Text Processing 0

To restore the default values, enter:

.s D
The point size returns to 10 points and the vertical spacing is automatically reset to 12
points. To increase the vertical space to 16 points while keeping the point size the
same, enter:
.S C 16

In the following example for a letterhead, the company name is specified in 18-point
type and a tag line in 12-point type; then the default settings are restored:
.S 18
Caslon Inc.
.s 12
Communicating Expertise
.S D

The result is:

1I
r

Caslon Inc.
Communicating Expertise
You can also change the font along with the point size, using the 1 macro described
previously. Following is the tag line in 12-point italic.

=Communicating Expertise
A special-purpose macro in mm reduces by 1 point the point size of a specified string.
The .SM macro can be followed by one, two, or three strings. Only one argument is
reduced; which one depends upon how many arguments are given. If you specify one
or two arguments, the first argument will be reduced by 1 point:
using
.SM UNIX ,
you will find

The second argument is concatenated to the first argument, so that the comma immedi-
ately follows the word UNIX:

I- using UNIX, you will find I

If you specify three arguments:
.SM [ UNIX 1
The second argument is reduced by one point, but the first and third arguments are
printed in the current point size, and all three are concatenated:
 0 ThemmMacros 0 145

I
More about Displays
Broadly speaking, a display is any kind of information in the body of a document that
cannot be set as a normal paragraph. Displays can be figures, quotations, examples,
tables, lists, equations, or diagrams.
The display macros position the display on the page. Inside the display, you
might use other macros or preprocessors such as t b l or eqn. You might simply
have a block of text that deserves special treatment.
The display macros can be relied upon to provide

adequate spacing before and after the display;

horizontal positioning of the display as a left justified, indented, or centered
block;
proper page breaks by keeping the entire display together.

The default action of the DS macro is to left justify the text block in no-fill mode. It
provides no indentation from the current margins.
You can specify a different format for a display by specifying up to three argu-
ments with the .DS macro. The syntax is:

.DS ybrmat] lfill mode] [right indent]

The format argument allows you to specify an indented or centered display. The argu-
ment can be set by a numeric value or a letter corresponding to the following options:

O L N o indent (default)
1 1 Indented
2 c Center each line
3 CB Center entire display

For consistency, the indent of displays is initially set to be the same as indented para-
graphs (five spaces in n r o f f and three ens in t r o f f ) , although these values are
maintained independently in two different number registers, P i and Si. (To change
the defaults, simply use the .nr request to put the desired value in the appropriate
register.)
A display can be centered in two ways: either each individual line in the display
is centered (C) or the entire display is centered as a block based on the longest line of
the display (CB).
For instance, the preceding list was formatted using t b l , but its placement was
controlled by the display macro.
146 UNlX Text Processing 0

.DS CB
.TS
table specifications
- TE
.DE
Thefill mode argument is represented by either a number or a letter

0 N No-fill mode (default)

1 F Fill mode

The right indent argument is a numeric value that is subtracted from the right
margin. In nrof f, this value is automatically scaled in ens. In t r o f f, you can
specify a scaled number; otherwise, the default is ems.
The use of fill mode, along with other indented display options, can provide a
paragraph indented on both sides. This is often used in reports and proposals that quote
at length from another source. For example:
.P
I was particularly interested in the following comment
found in the product specification:
.DS I F 5
Users first need a brief introduction to what the product
does. Sometimes this is more for the benefit of people
who haven't yet bought the product, and
are just looking at the manual.
However, it also serves to put the rest of
the manual, and the product itself, in the proper context.
-DE
The result of formatting is:

I was particular y interested in the following comment

found in the the product specification:

I Users first need a brief introduction to

what the product does. Sometimes this is
more for the benefit of people who haven't
yet bought the product, and are just looking
at the manual. However, it also serves to
put the rest of the manual, and the product
itself, in the proper context.

The use of tabs often presents a problem outside of displays. Material that has
been entered with tabs in the input file should be formatted in no-fill mode, the default
setting of the display macros. The following table was designed using tabs to provide
the spacing:
 0 ThemmMacros 0 147

.DF I
Dates Description of Task

June 30 Submit audience analysis

July 2 Meeting t o review audience analysis
July 15 Submit detailed outline
August 1 Submit first draft
August 5 Return of first draft
August 8 Meeting to review comments
.DE

This table appears in the output just as it looks in the file. If this material had not been
processed inside a display in no-fill mode, the columns would be improperly aligned.

Static and Floating Displays

There are two types of displays, static andfloating. The difference between them has to
do with what happens when a display cannot fit in its entirety on the current page.
Both the static and the floating display output the block at the top of the next page if it
doesn't fit on the current page; however, only the floating display allows text that fol-
lows the display to be used to fill up the preceding page. A static display maintains the
order in which a display was placed in the input file.
We have already used .D S and . D E to mark the beginning and end of a static
display. To specify a floating display, the closing mark is the same, but the beginning
is marked by the .DF macro. The options are the same as for the .D S macro.
In the following example of an input file, numbers are used instead of actual lines
of text:
1
L

3
4
5
.DF
Long Displuy
.DE
6
7
8
9
10

The following two formatted pages might be produced, assuming that there are a suffi-
cient number of lines in the display to cause a page break:
148 0 UNlX Text Processing 0

-1- -2-

Long Display

8
9
10

If there had been room on page 1 to fit the display, it would have been placed there, and
lines 6 and 7 would have followed the display, as they did in the input file.
If a static display had been specified, the display would be placed in the same
position on page 2, and lines 6 and 7 would have to follow it, leaving extra space at the
bottom of page 1 . A floating display attempts to make the best use of the available
space on a page.
The formatter maintains a queue to hold floating displays that it has not yet out-
put. When the top of a page is encountered, the next display in the queue is output.
The queue is emptied in the order in which it was filled, (first in, first out). Two
number registers, D e and D f , allow you to control when displays are removed from
the queue and placed in position.
At the end of a section, as indicated by the section macros - H and .HU (which
we will see shortly), or at the end of the input file, any floating displays that remain in
the queue will be placed in the document.

Display Labels
You can provide a title or caption for tables, equations, exhibits, and figures. In addi-
tion, the display can be labeled and numbered in sequence, as well as printed in a table
of contents at the end of the file. The following group of macros are available:

.EC Equation
.EX Exhibit
.FG Figure

All of these macros work the same way and are usually specified within a pair of
.O S / . DE macros, so that the title and the display appear on the same page. Each
macro can be followed by a title. If the title contains spaces, it should be enclosed
within quotation marks. The title of a table usually appears at the top of a table, so it
must be specified before the .TS macro that signals to t b l the presence of a table
(see Chapter 8).
 0 ThemmMacros 149

.TB "List of Required Resources"

- TS
The label is centered:
I I
ITable 1. List of Required Resources I
If the title exceeds the line length, then it will be broken onto several lines. Addi-
tional lines are indented and begin at the first character of the title.

Table 1. List of Required Resources

Provided by Gutenberg Galaxy
I Software
The label for equations, exhibits, and figures usually follows the display. The fol-
lowing:
.FG "Drawing with a Light Pen"
produces a centered line:

I 1
I Figure 1. Drawing with a Light Pen I
The default format of the label can be changed slightly by setting the number
register O f to 1. This replaces the period with a dash.

I 1
I Figure 1 - Drawing with a Light Pen I
Second and third arguments, specified with the label macros, can be used to
modify or override the default numbering of displays. Basically, the second argument
is a literal and the third argument a numeric value that specifies what the literal means.
If the third argument is

0 then the second argument will be treated as a prefix;

1 then the second argument will be treated as a suffix;
2 then the second argument replaces the normal table number.

Thus, a pair of related tables could be specified as l a and l b using the following labels:
.TB "Estimated Hours: June, July, and August" a 1
.TB "Estimated Hours: September and November," lb 2

(These labels show two different uses of the third argument. Usually, you would con-
sistently use one technique or the other for a given set of tables.)
For tbl, the delimiters for tables are - TS/.TE. For e q n , the delimiters for
equations are .EQ/.EN. For p i c , the delimiters for pictures or diagrams are
.PS/. PE. These pairs of delimiters indicate a block to be processed by a specific
150 0 UNlX Text Processing 0

preprocessor. You will find the information about each of the preprocessors in Chapters
8 through 10. A s mentioned, the preprocessor creates the display, the display macros
position it, and the label macros add titles and a number.
Although it may seem a minor point, each of these steps is independent, and
because they are not fully integrated, there is some overlap.
The label macros, being independent of the preprocessors, do not make sure that a
display exists or check whether a table has been created with tbl. You can create a
two-column table using tabs or create a figure using character symbols and still give it a
label. Or you can create a table heading as the first line of your table and let t b l pro-
cess it (tbl won’t provide a number and the table won’t be collected for the table of
contents).
In t bl, you can specify a centered table and not use the .DS/.DE macros.
But, as a consequence, nrof f / trof f won’t make a very good attempt at keeping the
table together on one page, and you may have to manually break the page. It is recom-
mended that you use the display macros throughout a document, regardless of whether
you can get the same effect another way, because if nothing else you will achieve con-
sistency.

Forcing a Page Break

Occasionally, you may want to force a page break, whether to ensure that a block of
related material is kept together or to allow several pages for material that will be
manually pasted in, such as a figure. The .SK (skip) macro forces a page break. The
text following this macro is output at the top of the next page. If supplied with an argu-
ment greater than 0, it causes that number of pages to be skipped before resuming the
output of text. The “blank” pages are printed, and they have the normal header and
footer.
On the next page, you will find a sample page from an
Alcuin manuscript printed with a 16-color plotter.
.SK 1

Formatting Lists
The mm macro package provides a variety of different formats for presenting a list of
items. You can select from four standard list types:

bulleted
dashed
numbered
alphabetized
 0 ThemmMacros 0 151

In addition, you have the flexibility to create lists with nonstandard marks or text labels.
The list macros can also be used to produce paragraphs with a hanging indent.
Each list item consists of a special mark, letter, number, or label in a left-hand
column with a paragraph of text indented in a right-hand column.

Structuring a List
The list macros help to simplify what could be a much larger and tedious formatting
task. Here’s the coding for the bulleted list just shown:
.BL
*LI
bulleted

dashed
.LI
numbered
.LI
alphabetized
.LE
The structure of text in the input file has three parts: a list-initialization macro (. BL),
an item-mark macro (. LI), and a list-end macro ( - LE).
First, you initialize the list, specifying the particular macro for the type of list that
you want. For instance, BL initializes a bulleted list.
You can specify arguments with the list-initialization macro that change the
indentation of the text and turn off the automatic spacing between items in the list. We
will examine these arguments when we look at the list-initialization macros in more
detail later.
Next, you specify each of the items in the list. The item-mark macro, .LI, is
placed before each item. You can enter one or more lines of text following the macro.
.B L
.LI
Item 1
.LI
Item 2
.LI
Item 3

When the list is formatted, the .LI macro provides a line of space before each item.
(This line can be omitted through an argument to the list-initialization macro if you
want to produce a more compact list. We’ll be talking more about this in a moment.)
The .LI macro can also be used to override or prefix the current mark. If a
mark is supplied as the only argument, it replaces the current mark. For example:
152 0 UNlX Text Processing 0

.LI 0
Item 4

If a mark is supplied as the first argument, followed by a second argument of 1 , then

the specified mark is prefixed to the current mark. The following:
.LI - 1
Item 5

would produce:

r* Item 5 1
A text label can also be supplied in place of the mark, but it presents some addi-
tional problems for the proper alignment of the list. We will look at text labels for
variable-item lists.
The .LI macro does not automatically provide spacing after each list item. An
argument of 1 can be specified if a line of space is desired.
The end of the list is marked by the list-end macro .LE. It restores page format-
ting settings that were in effect prior to the invocation of the last list-initialization
macro. The . L E macro does not output any space following the list unless you
specify an argument of 1. (Don’t specify this argument when the list is immediately
followed by a macro that outputs space, such as the paragraph macro.)
Be sure you are familiar with the basic structure of a list. A common problem is
not closing the list with .LE. Most of the time, this error causes the formatter to quit
at this point in the file. A less serious, but nonetheless frequent, oversight is omitting
the first .LI between the list-initialization macro and the first item in the list. The list
is output but the first item will be askew.
Here is a sample list:
- BL
.LI
Item 1

Item 2
.LI
Item 3
.LI 0
Item 4
.LI - 1
Item 5
.LE
 0 ThemmMacros 0 153

The t r o f f output produced by the sample list is:

Item 1
Item 2

Item 3
o Item 4
-* Item 5

Complete list structures can be nested within other lists up to six levels. Different
types of lists can be nested, making it possible to produce indented outline structures.
3ut, like nested if-then structures in a program, make sure you know which level you
are at and remember to close each list.
For instance, we could nest the bulleted list inside a numbered list. The list-
initialization macro .AL generates alphabetized and numbered lists.

.A L
.LI
Don't worry, we'll get t o the list-initialization macro .AL.
You can specify five different variations of
alphabetic and numbered lists.
.BL
.LI
Item 1
.LI
Item 2
.
LI
Item 3
.LE
.
LI
We'll also look at variable-item lists.
.LE

This input produces the following formatted list from t rof f :

154 0 UNlX Text Processing 0

Item 1
Item 2
Item 3

re’ll also look at variLJle-item lists.

You may already realize the ease with which you can make changes to a list. The
items in a list can be easily put in a new order. New items can be added to a numbered
list without readjusting the numbering scheme. A bulleted list can be changed to an
alphabetized list by simply changing the list-initialization macro. And you normally
don’t have to be concerned with a variety of specific formatting requests, such as setting
indentation levels or specifying spacing between items.
On the other hand, because the structure of the list is not as easy to recognize in
the input file as it is in the formatted output, you may find it difficult to interpret com-
plicated lists, in particular ones that have been nested to several levels. The code-
checking program, checkmm, can help; in addition, you may want to format and print
repeatedly to examine and correct problems with lists.

Marked Lists
Long a standby of technical documents, a marked list clearly organizes a group of
related items and sets them apart for easy reading. A list of items marked by a bullet
( 0 ) is perhaps the most common type of list. Another type of marked list uses a dash
(-). A third type of list allows the user to specify a mark, such as a square ( ). The
list-initialization macros for these lists are:

.BL [text indent] [ 11

.DL [text indent] [ 11
.M L [mark] [text indent] [ 11

With the .BL macro, the text is indented the same amount as the first line of an
indented paragraph. A single space is maintained between the bullet and the text. The
bullet is right justified, causing an indent of several spaces from the left margin.
As you can see from this n r o f f-formatted output, the bullet is simulated in
n r o f f by a + overstriking an 0 :
 0 ThemmMacros 0 155

cb Alcuin/UNIX interface definition

I cb Programmer's documentation for Alcuin I

If you specify a text indent, the first character of the text will start at that position. The
position of the bullet is relative to the text, always one space to its left.
If the last argument is 1, the blank line of space separating items is omitted. If
you want to specify only this argument, you must specify either a value or a null value
("">for a text indent.
.BL '"' 1
It produces a much more compact list:

I I
cb G G S Technical Memo 3 2 0 0
6 G G S Product Marketing Spec
cb Alcuin/UNIX interface definition
cb Programmer's documentation for Alcuin

Because the bullets produced by n r o f f are not always appropriate due to the
overstriking, a dashed list provides a suitable alternative. With the ,DL macro, the
dash is placed in the same position as a bullet in a bulleted list. A single space is main-
tained between the dash and the text, which, like the text with a bulleted list, is indented

r
by the amount specified in the number register for indented paragraphs (Pi).
The n r o f f formatter supplies a dash that is a single hyphen, and t r o f f sup-
plies an em dash. Because the em dash is longer, and the dash is right justified, the
alignment with the left margin is noticeably different. It appears left justified in
t r o f f ; in n r o f f,the dash appears indented several spaces because it is smaller.

The third chapter on the principles of computerized

font design should cover the following topics:

- Building a Font Dictionary

- Loading a Font

- Scaling a Font

You can specify a text indent and a second argument of 1 to inhibit spacing between
items.
156 0 UNlX Text Processing 0

With the .ML macro, you have to supply the mark for the list. Some possible
candidates are the square (enter \ (sq to get o),the square root (enter \ (sr to get
d), which resembles a check mark, and the gradient symbol (enter \ ( g r to get v).
The user-specified mark is the first argument.
.ML \(sq
Not all of the characters or symbols that you can use in t r o f f will have the same
effect in nroff.

r
Unlike bulleted and dashed lists, text is not automatically indented after a user-
specified mark. However, a space is added after the mark. The following example of
an indented paragraph and a list, which specifies a square as a mark, has been formatted
using n r o f f . The square appears as a pair of brackets.

[] Remove old initialization files.

[] Run install program.

[] Exit t o main menu and choose selection 3.

The user-supplied mark can be followed by a second argument that specifies a text
indent and a third argument of 1 to omit spacing between items.
The following example was produced using the list-initialization command:
.ML \ ( s q 5 1

I
The specified indent of 5 aligns the text with an indented paragraph:

Check t o see that you have completed the following

steps :

[I Remove old initialization files.

[I Run install program.
[I Exit t o main menu and choose selection 3.

Numbered and Alphabetic Lists

The .A L macro is used to initialize automatically numbered or alphabetized lists. The
syntax for this macro is:

.AL [type] [text indent] [l]

If no arguments are specified, the .A L macro produces a numbered list. For instance,
we can code the following paragraph with the list-initialization macro .AL:
 0 ThemMacros 0 157

User-oriented documentation recognizes three things:

- AL
- LI
that a new user needs to learn the system in stages,
getting a sense of the system as a whole while becoming
proficient in performing particular tasks;
.LI
that there are different levels of users, and not every
user needs to learn all the capabilities of the system
in order to be productive;
.LI
that an experienced user must be able to rely on the
documentation for accurate and thorough reference
information.
.LE
to produce a numbered list:
I

User-oriented documentation recognizes three things:

1. that a new user needs to learn the system in stages,

getting a sense of the system as a whole while
becoming proficient in performing particular tasks;

2. that there are different levels of users, and not

every user needs to learn all the capabilities of
the system in order to be productive;

3. that an experienced user must be able t o rely on the

documentation for accurate and thorough reference
information.
The number is followed by a period, and two spaces are maintained between the period
and the first character of text.
The level of text indent, specified in the number register Li, is 6 in nrof f and
5 in t r o f f . This value is added to the current indent. If a text indent is specified,
that value is added to the current indent, but it does not change the value of Li.
The third argument inhibits spacing between items in the list. Additionally, the
number register LS can be set to a value from 0 to 6 indicating a nesting level. Lists
after this level will not have spacing between items. The default is 6, the maximum
nesting depth. If Ls were set to 2, lists only up to the second level would have a
blank line of space between items.
Other types of lists can be specified with .AL, using the first argument to
specify the list type, as follows:
158 UNIX Text Processing 0

Value Sequence Description

1 1, 2, 3 Numbered
A A, B, C Alphabetic (uppercase)
a a, b, c Alphabetic (lowercase)
I I, 11, I11 Roman numerals (uppercase)
i i, ii, iii Roman numerals (lowercase)

You can produce various list types by simply changing the type argument. You can
create a very useful outline format by nesting different types of lists. The example we
show of such an outline is one that is nested to four levels using I, A, 1 , and a, in
that order. The rather complicated looking input file is shown in Figure 6-4 (indented
for easier viewing of each list, although it could not be formatted this way), and the
n K O f f -formatted output is shown in Figure 6-5.
Another list-initialization macro that produces a numbered list is RL (reference
list). The only difference i s that the reference number is surrounded by brackets ([I).

.RL [text indent] [ 11

The arguments have the same effect as those specified with the .AL macro. To initial-
ize a reference list with no spacing between items, use:

.RL "" 1

It produces the following reference list:

I

[l] The Main Menu

[2] Menus or Commands?
[3] Error Handling
[4] Getting Help
[SI Escaping to UNIX

Variable-Item Lists
With a variable-item list, you do not supply a mark; instead, you specify a text label
with each .L I . One or more lines of text following .LI are used to form a block
paragraph indented from the label. If no label is specified, a paragraph with a hanging
indent is produced. The syntax is:

.VI, text indent [label indent] [ll

Unlike the other list-initialization macros, a text indent is required. By default, the label
is left justified, unless a label indent is given. If you specify both a text indent and a
label indent, the indent for the text will be added to the label indent.
 ~

ThemMacros 0 159

.AL I
- LI
Quick Tour of Alcuin
.AL A
.LI
Introduction to Calligraphy
- LI
Digest of Alcuin Commands
.AL 1
.LI
Three Methods of Command Entry
.AL a
- LI
Mouse
- LI
Keyboard
.LI
Light Pen
LE
.LI
Starting a Page
- LI
Drawing Characters
.AL a
.LI
Choosing a Font
.LI
Switching Fonts
.LE
.LI
Creating Figures
.LI
Printing
.LE
.LI
Sample Illuminated Manuscripts
.LE
.LI
Using Graphic Characters
.AL A
- LI
Modifying Font Style
.LI
Drawing Your Own Font
.LE
* LI

Library of Hand-Lettered Fonts

.LE

Fig. 6-4. Input for a Complex List

160 a UNlX Text Processing 0

- 1 -

I. Quick Tour of Alcuin

A. Introduction to Calligraphy

B. Digest of Alcuin Commands

1. Three Methods of Command Entry

a. Mouse

b. Keyboard

c. Light Pen

2. Starting a Page

3. Drawing Characters

a. Choosing a Font

b. Switching Fonts

4. Creating Figures

5. Printing

C. Sample Illuminated Manuscripts

11. Using Graphic Characters

A. Modifying Font Style

€3. Drawing Your Own Font

111. Library of Hand-Lettered Fonts

Fig. 6-5. Output of a Complex List

 0 ThemmMacros 0 161

Variable-item lists are useful in preparing command reference pages, which

describe various syntax items, and glossaries, which present a term in one column and
its definition in the other. The text label should be a single word or phrase. The fol-
lowing example shows a portion of the input file for a reference page:
.VL 15 5
.LI figure
is the name of a cataloged figure. If
a figure has not been cataloged, you need to use
the LOCATE command.
.LI f:p
is the scale of the
figure in relation to the page.
.LI font
is the two-character abbreviation or
full name of one of the available fonts
from the Alcuin library.
- LE
The following variable-item list is produced:

figure is the name of a cataloged figure. If a

figure has not been cataloged, you need to
use the LOCATE command.

f :p is the scale of the figure in relation to

the page.

~ font is the two-character abbreviation or full

name of one of the available fonts from the
Alcuin library.
If you don't provide a text label with .LI or give a null argument you will
("'I),

get a paragraph with a hanging indent. If you want to print an item without a label,
specify a backslash followed by a space ( \ ) or \ 0 after .LI. Similarly, if you want
to specify a label that contains a space, you should also precede the space with a
backslash and enclose the label within quotation marks:
.LI "point\ size"
or simply substitute a \ 0 for a space:
. L I point\Osize

The first line of text is left justified (or indented by the amount specified in label
indent) and the remaining lines will be indented by the amount specified by text indent.
This produces a paragraph with a hanging indent:
162 0 UNlX Text Processing 0

.VL 15
-LI
There are currently 16 font dictionaries in t h e Alcuin
library. Any application may have u p t o 12 dictionaries
resident in memory at t h e same time.
.LE
When formatted, this item has a hanging indent of 15:

There are currently 16 font dictionaries in t h e Alcuin

library. Any application may have up t o
12 dictionaries resident in memory at t h e
same time .

Headings
Earlier we used the list macros to produce an indented outline. That outline, indented
to four levels, is a visual representation of the structure of a document. Headings per-
form a related function, showing how the document is organized into sections and sub-
sections. In technical documentation and book-length manuscripts, having a structure
that i s easily recognized by the reader is very important.

Numbered and Unnumbered Headings

Using mm, you can have up to seven levels of numbered and unnumbered headings,
with variable styles. There are two heading macros: .H for numbered headings and
.HU for unnumbered headings. A different style for each level of heading can be speci-
fied by setting various number registers and defining strings.
Let's first look at how to produce numbered headings. The syntax for the .H
macro is:

.H level [heading text] [heading sufJix3

The simplest use of the .H macro is to specify the level as a number between 1 and 7
followed by the text that is printed as a heading. If the heading text contains spaces,
you should enclose it within quotation marks. A heading that is longer than a single
line will be wrapped on to the next line. A multiline heading will be kept together in
case of a page break.
If you specify a heading suffx, this text or mark will appear in the heading but
will not be collected for a table of contents.
A top-level heading is indicated by an argument of 1:
.H 1 "Quick Tour of Alcuin"
 ThemMacros 0 163

The result is a heading preceded by a heading-level number. The first-level heading has
the number 1.
1. Quick Tour of Alcuin

A second-level heading is indicated by an argument of 2:

.H 2 "Introduction t o Calligraphy"

The first second-level heading number is printed:

1.1 Introduction t o Calligraphy

When another heading is specified at the same level, the heading-level number is
automatically incremented. If the next heading is at the second level:
.H 2 "Digest of Alcuin Commands"

it produces:
1.2 Digest of Alcuin Commands
Each time you go to a new (higher-numbered) level, .1 is appended to the number
representing the existing level. That number i s incremented for each call at the same
level. When you back out of a level (for instance, from level 5 to 4), the counter for the
level (in this case level 5), is reset to 0.
An unnumbered heading is really a zero-level heading:
.H 0 "Introduction t o Calligraphy"

A separate macro, .HU, has been developed for unnumbered headings, although
its effect is the same.
.HU "Introduction t o Calligraphy"

Even though an unnumbered heading does not display a number, it increments the

r
counter for second-level headings. Thus, in the following example, the heading "Intro-
duction to Calligraphy" is unnumbered, but it has the same effect on the numbering
scheme as if it had been a second-level heading (1.1).

Quick Tour of Alcuin

Introduction to Calligraphy

1.2 Digest of Alcuin Commands

If you are going to intermix numbered and unnumbered headings, you can change
the number register Hu to the lowest-level heading that i s in the document. By chang-
ing H u from 2 to a higher number:
.nr Hu 5
.H 1 "Quick Tour of Alcuin"
.HU "Introduction t o Calligraphy"
.H 2 "Digest of Alcuin Commands"
164 0 UNlX Text Processing 0

rT
the numbering sequence is preserved for the numbered heading following an unnum-
bered heading:

Quick Tour of Alcuin

Introduction t o Calligraphy
Digest o f Alcuin Commands
Headings are meant to be visible keys to your document’s structure. If you are using
unnumbered headings, it becomes even more important to make headings stand out. A
simple thing you can do is use uppercase letters for a first-level heading.
Here is a list of some of the other things you can do to affect the appearance of
headings, although some of the items depend upon whether you are formatting with
n r o f f or t r o f f :

change to roman, italic, or bold font

change the point size of the heading
adjust spacing after the heading
center or left justify the heading
change the numbering scheme
select a different heading mark

The basic issue in designing a heading style is to help the reader distinguish between
different levels of headings. For instance, in an outline form, different levels of indent
show whether a topic is a section or subsection. Using numbered headings is an effec-
tive way to accomplish this. If you use unnumbered headings, you probably want to
vary the heading style for each level, although, for practical purposes, you should limit
yourself to two or three levels.
First, let’s look at what happens if we use the default heading style.
The first two levels of headings are set up to produce italicized text in t r o f f
and underlined text in n r o f f . After the heading, there is a blank line before the first
paragraph of text. In addition, a top-level heading has two blank lines before the head-
ing; all the other levels have a single line of space.

I---- 1.2 Introduction to Calligraphy

7
Alcuin revitalizes an age-old tradition. Calligraphy, quite simply, is the art of
beautiful handwriting.

Levels three through seven all have the same appearance. The text is italicized or
underlined and no line break occurs. Two blank lines are maintained before and after
the text of the heading. For example:
 0 ThemmMacros 0 165

I 1
1.2.1.3 Light Pen The copyist’s pen and ink has been replaced by a light pen.
To change the normal appearance of headings in a document, you specify new
values for the two strings:

HF Heading font
HP Heading point size

You can specify individual settings for each level, up to seven values.
The font for each level of heading can be set by the string H F . The following
codes are used to select a font:

1 Roman
2 Italic
3 Bold

By default, the arguments for all seven levels are set to 2, resulting in italicized head-
ings in t ro f f and underlining in nrof f. Here the .HF string specifies bold for
the top three levels followed by two italic levels:
.ds HF 3 3 3 2 2

If you do not specify a level, it defaults to 1. Thus, in the previous example, level 6
and 7 headings would be printed in a roman font.
The point size is set by the string HP. Normally, headings are printed in the
same size as the body copy, except for bold headings. A bold heading is reduced by 1
point when it is a standalone heading, as are the top-level headings. The HP string can
take up to seven arguments, setting the point size for each level.
.ds HP 1 4 1 4 12

If an argument is not given, or a null value or 0 is given, the default setting of 10 points
is used for that level. Point size can also be given relative to the current point size:
.ds HP + 4 + 4 + 2

A group of number registers control other default formats of headings:

Ej Eject page
Hb Break follows heading
Hc Center headings
H i Align text after heading
Hs Vertical spacing after heading

For each of these number registers, you specify the number of the level at which some
action is to be turned on or off.
166 0 UNlX Text Processing 0

The E j register is set to the highest-level heading, usually 1 , that should start on
a new page. Its default setting is 0. This ensures that the major sections of a document
will begin on their own page.
.nr Ej 1
The Hb register determines if a line break occurs after the heading. The Hs register
determines if a blank line is output after the heading. Both are set to 2 by default. Set-
tings of 2 mean that, for levels 1 and 2, the section heading is printed, followed by a
line break and a blank line separating the heading from the first paragraph of text. For
lower-level headings (an argument greater than 2), the first paragraph follows irnmedi-
ately on the same line.
The H c register is set to the highest-level heading that you want centered. Nor-
mally, this is not used with numbered headings and its default value is 0. However,
unnumbered heads are often centered. A setting of 2 will center first- and second-level
headings:
.nr Hc 2

With unnumbered headings, you also have to keep in mind that the value of H c must be
greater than or equal to Hb and Hu. The heading must be on a line by itself; therefore
a break must be set in Hb for that level. The Hu register sets the level of an unnum-
bered heading to 2, requiring that Hc be at least 2 to have an effect on unnumbered
headings.
There really is no way, using these registers, to get the first and second levels left
justified and have the rest of the headings centered.
The number register H i determines the paragraph type for a heading that causes a
line break (Hb). It can be set to one of three values:

0 Left justified
1 Paragraph type determined by P t
2 Indented to align with first character in heading

If you want to improve the visibility of numbered headings, set Hi to 2:

.nr Hi 2
It produces the following results:

4.1 Generating Output

An Alcuin manuscript is a computer representation

that has t o be converted for output on various kinds
of devices, including plotters and laser printers.
 0 ThemmMacros 0 167

Changing the Heading Mark

Remember how the list-initialization macro .A L allowed you to change the mark used
for a list, producing an alphabetic list instead of a numbered list? These same options
are available for headings using the .HM macro.
The .HM macro takes up to seven arguments specifying the mark for each level.
The following codes can be specified:

1 Arabic
001 Arabic with leading zeros
A Uppercase alphabetic
a Lowercase alphabetic
I Uppercase roman
i Lowercase roman

If no mark is specified, the default numbering system (arabic) is used. Uppercase

alphabetic marks can be used in putting together a series of appendices. You can
specify A for the top level:
.HM A
and retain the default section numbering for the rest of the headings. This could pro-
duce sections in the following series:

A , A.1, A.2, A.2.1, etc.

Marks can be mixed for an outline style similar to the one we produced using the list
macros:
.HM I A 1 a i

Roman numerals can be used to indicate sections or parts. If you specify:

.HM I i
the headings for the first two levels are marked by roman numerals. A third-level head-
ing is shown to demonstrate that the heading mark reverted to arabic by default:

I. Quick Tour of Alcuin

1.i Introduction t o Calligraphy

I.ii Digest of Alcuin Commands

I.ii.1 Three Methods of Command Entry

168 0 UNlX Text Processing 0

When you use marks consisting of roman numerals or alphabetic characters, you might
not want the mark of the current level to be concatenated to the mark of the previous
level. Concatenation can be suppressed by setting the number register H t to 1 :
.HM I i
.nr Ht 1
Now, each heading in the list has only the mark representing that level:

I. Quick Tour of Alcuin

i. Introduction to Calligraphy

ii. Digest of Alcuin Commands

1. Three Methods of Command E n t r y

Table of Contents
Getting a table of contents easily and automatically is almost reason enough to justify
all the energy, yours and the computer’s, that goes into text processing. You realize
that this is something that the computer was really meant to do.
When the table of contents page comes out of the printer, a writer attains a state
of happiness known only to a statistician who can give the computer a simple instruc-
tion to tabulate vast amounts of data and, in an instant, get a single piece of paper list-
ing the results.
The reason that producing a table of contents seems so easy is that most of the
work is performed in coding the document. That means entering codes to mark each
level of heading and all the figures, tables, exhibits, and equations. Processing a table
of contents is simply a matter of telling the formatter to collect the information that’s
already in the file.
There are only two simple codes to put in a file, one at the beginning and one at
the end, to generate a table of contents automatically.
At the beginning of the file, you have to set the number register C 1 to the level
of headings that you want collected for a table of contents. For example, setting C 1 to
2 saves first- and second-level headings.
Place the . T C macro at the end of the file. This macro actually does the pro-
cessing and formatting of the table of contents. The table of contents page is output at
the end of a document.
A sample table of contents page follows. The header “CONTENTS” is printed
at the top of the page. At the bottom of the page, lowercase roman numerals are used
as page numbers.
 0 ThemMacros 0 169

r- CONTENTS

1 . Quick Tour of Alcuin .......................................................................... 1

1 . 1 Introduction to Calligraphy .......................................................... 3
1.2 Digest of Alcuin Commands ........................................................ 8
1.3 Sample Illuminated Manuscripts .................................................. 21

2. Using Graphic Characters ................................................................... 31

2.1 Scaling a Font ............................................................................... 33
2.2 Modifying Font Style ................................................................... 37
2.3 Drawing Your Own Font .............................................................. 41

I 3. Library of Hand-Lettered Fonts .......................................................... 51

- 1 -

One blank line is output before each first-level heading. AI1 first-level headings are left
justified. Lower-level headings are indented so that they line up with the start of text
for the previous level.
If you have included various displays in your document, and used the macros
.FG, .TB, and .EX to specify captions and headings for the displays, this informa-
tion is collected and output when the .TC macro is invoked. A separate page is
printed for each accumulated list of figures, tables, and exhibits. For example:

LIST OF TABLES

TABLE 1. List of Required Resources ..................... 7

TABLE 2. List of Available Resources .................... 16

If you want the lists of displays to be printed immediately following the table of con-
tents (no page breaks), you can set the number register C p to 1.
If you want to suppress the printing of individual lists, you can set the following
number registers to 0:

Lf If 0, no figures
Lt If 0, no tables
Lx If 0, no exhibits
170 0 UNlX Text Processing 0

In addition, there is a number register for equations that is set to 0 by default. If you
want equations marked by .E C to be listed, specify:
.nr Le 1

There are a set of strings, using the same names as the number registers, that define the
titles used for the top of the lists:

Lf LIST OF FIGURES
Lt LIST OF TABLES
Lx LIST OF EXHIBITS
Le LIST OF EQUATIONS

You can redefine a string using the . d s (define string) request. For instance, we can
redefine the title for figures as follows:
.ds Lf LIST OF ALCUIN DRAWINGS

. Footnotes and References

Footnotes and references present special problems, as anyone who has ever typed a term
paper knows. Fortunately, mm has two pairs of specialized macros. Both of them fol-
low a marker in the text and cause lines of delimited text to be saved and output either
at the bottom of the page, as a footnote, or at end of the document, as a reference.

Footnotes
A footnote is marked in the body of a document by the string \*F. It follows immedi-
ately after the text (no spaces).
in an article on desktop publishing.\*F

The string F supplies the number for the footnote. It is printed (using t r o f f) as a
superscript in the text and its value is incremented with each use.
The .F S macro indicates the start, and .F E the end, of the text for the footnote.
These macros surround the footnote text that will appear at the bottom of the page. The
.FS macro is put on the line immediately following the marker.
.FS
"Publish or Perish: Start-up grabs early page language lead,"
\fIComputerworld\fR, April 21, 1986, p - 1.
-FE

You can use labels instead of numbers to mark footnotes. The label must be specified
as a mark in the text and as an argument with .FS.
...in accord with t h e internal specs.[APS]
.FS [ A P S ]
"Alcuin Product Specification," March 1986
.F E
 0 ThemmMacros 0 171

You can use both numbered and labeled footnotes in the same document. All the foot-
notes are collected and output at the bottom of each page underneath a short line rule.
If you are using t r o f f, the footnote text will be set in a type size 2 points less than
the body copy.
If you want to change the standard format of footnotes, you can specify the F D .
macro. It controls hyphenation, text adjustment, indentation, and justification of the
label.
Normally, the text of a footnote i s indented from the left margin and the mark or
label i s left justified in relation to the start of the text. It is possible that a long footnote
could run over to the next page. Hyphenation is turned off so that a word will not be
broken at a page break. These specifications can be changed by giving a value between
.
0 and 11 as the first argument with F D , as shown in Table 6-3.

TABLE 6-3. .FD Argument Values

Text Label
Argument Hyphenation Adjust Indent Justification
0 no Yes Yes left
1 Yes Yes Yes left
2 no no Yes left
3 Yes no Yes left
4 no Yes no left
5 Yes Yes no left
6 no no no left
7 Yes no no left
8 no Yes Yes right
9 Yes Yes Yes right
10 no no Yes right
11 Yes no Yes right

.
The second argument for F D , if 1, resets the footnote numbering counter to 1.
This can be invoked at the end of a section or paragraph to initiate a new numbering
sequence. If specified by itself, the first argument must be null:
.FD I'" 1

References
A reference differs from a footnote in that all references are collected and printed on a
single page at the end of the document. In addition, you can label a reference so that
you can refer to it later.
172 UNlX Text Processing 0

A reference is marked where it occurs in the text with \ * ( R f . The formatter

converts the string into a value printed in brackets, such as [l]. The mark is followed
by a pair of macros surrounding the reference text. The .RS macro indicates the start,
and .RF the end, of the text for the reference.
You will find information on this page description language
in their reference manual, which has been published
as a book.\*(Rf
.RS
Adobe Systems, Inc. PostScript Reference Manual.
Reading, Massachusetts: Addison-Wesley; 1985.
.RF

You can also give as a string label argument to .RS the name of a string that will be
assigned the current reference number. This string can be referenced later in the docu-
ment. For instance, if we had specified a string label in the previous example:
.RS A s
We could refer back to the first reference in another place:
The output itself is a readable file which you can interpret
with the aid o f t h e PostScript manual.\*(As

At the end of the document, a reference page is printed. The title printed on the
reference page is defined in the string Rp. You can replace “REFERENCES” with
another title simply by redefining this string with - ds.

REFERENCES

1. Adobe Systems, Inc.; PostScript Reference Manual.

Reading, Massachusetts: Addison-Wesley; 1985.

In a large document, you might want to print a list of references at the end of a chapter
or a long section. You can invoke the .RP macro anywhere in a document.
- RP
.H 1 “Detailed Outline of User Guide”
It will print the list of references on a separate page and reset the reference counter to 0.
A reset argument and a paging argument can be supplied to change these actions. The
reset argument i s the first value specified with the .RP macro. It i s normally 0, reset-
ting the reference counter to 1 so that each section is numbered independently. If refer-
ence numbering should be maintained in sequence for the entire document, specify a
value of 1 .
 0 ThemmMacros 0 173

The paging argument is the second value specified. It controls whether or not a
page break occurs before and after the list. It is normally set to 0, putting the list on a
new page. Specifying a value of 3 suppresses the page break before and after the list;
the result is that the list of references is printed following the end of the section and the
next section begins immediately after the list. A value of 1 will suppress only the page
break that occurs after the list and a value of 2 will suppress only the page break that
occurs before the list.
If you want an effect opposite that of the default settings, specify:
.RE’ 1 3

The first argument of 1 saves the current reference number for use in the next section or
chapter. The second argument of 3 inhibits page breaks before and after the list of
references.

Extensions to mm
So far, we have covered most but not all of the features of the mm macro package.
We have not covered the Technical Memorandum macros, a set of specialized
macros for formatting technical memos and reports. L i e the ones in the m s macro
package, these macros were designed for internal use at AT&T’s Bell Laboratories,
reflecting a company-wide set of standards. Anyone outside of Bell Labs will want to
make some modifications to the macros before using them. The Technical Memoran-
dum macros are a good example of employing a limited set of user macros to produce a
standard format. Seeing how they work will be especially important to those who are
responsible for implementing documentation standards for a group of people, some of
whom understand the basics of formatting and some of whom do not.
Writing or rewriting macros i s only one part of the process of customizing mm.
The mm macros were designed as a comprehensive formatting system. As we’ve seen,
there are even macros to replace common primitive requests, like .sp. The develop-
ers of mm recommend, in fact, that you not use n r o f f or t r o f f requests unless
absolutely necessary, lest you interfere with the action of the macros.
Furthermore, as you will see if you print out the mm macros, the internal code of
mm is extraordinarily dense, and uses extremely un-mnemonic register names. This
makes it very difficult for all but the most experienced user to modify the basic struc-
ture of the package. You can always add your own macros, as long as they don’t con-
flict with existing macro and number register names, but you can’t easily go‘in and
change the basic macros that make up the mm package.
At the same time, the developers of mm have made it possible for the user to
make selective modifications-those which mm has allowed mechanisms for in
advance. There are two such mechanisms:

mm’s use of number registers to control all aspects of document formatting

mm’s invocation of undefined (and therefore user-definable) macros at various
places in the mrn code
174 UNlX Text Processing 0

The mm package is very heavily parameterized. Almost every feature of the formatting
system-from the fonts in which different levels of heading are printed to the size of
indents and the amount of space above and below displays-is controlled by values in
number registers. By learning and modifying these number registers, you can make sig-
nificant changes to the overall appearance of your documents.
In addition, there are a number of values stored in strings. These strings are used
like number registers to supply default values to various macros.
The registers you are most likely to want to change follow. Registers marked
with a dagger can only be changed on the comand line with the -r option (e.g.,
-rN4).

c1 Level of headings saved for table of contents. See .TC macro.

Default is 2.
CP If set to 1, lists of figures and tables appear on same page as table of
contents. Otherwise, they start on a new page. Default is 1 .
Ds Sets the pre- and post-space used for static displays.
Fs Vertical spacing between footnotes.
Hb Level of heading for which break occurs before output of body text.
Default is 2 lines.
Hc Level of heading for which centering occurs. Default is 0.
Hi Indent type after heading. Default is 1 (paragraph indent). Legal
values are: O=left justified (default); 1 =indented; 2=indented except
after .H, .LC, .DE.
Hs Level of heading for which space after heading occurs. Default is 2,
i.e., space will occur after first- and second-level headings.
HY Sets hyphenation. If set to 1 , enables hyphenation. Default is 0.
L? Sets length of page. Default is 66v.
L i Default indent of lists. Default is 5.
Ls List spacing between items by level. Default is 6, which is spacing
between all levels of list.
Nt Page numbering style. O=all pages get header; l=header printed as
footer on page 1; 2=no header on page 1 ; 3=section page as footer;
4=no header unless .PH defined; 5=section page and section figure
as footer. Default is 0.
NP Numbering style for paragraphs. O=unnumbered; l=numbered.
0 Offset of page. For nroff, this value is an unscaled number
representing character positions. (Default is 9 characters; about -75.)
For t r o f f , this value is scaled (Si).
 1
0 ThemmMacros 0 175

Of Figure caption style. O=period separator; l=hyphen separator.

Default is 0.
Pi Amount of indent for paragraph. Default is 5 for nrof f, 3n for
troff.
Ps Amount of spacing between paragraphs. Default is 3v.
Pt Paragraph type. Default is 0.
S? Default point size for trof f. Default is 10. Vertical spacing i s
\nS+2.

Si Standard indent for displays. Default is 5 for n r o f f, 3 for t r o f f .

w Width of page (line and title length). Default i s 6 in t r o f f , 60
characters in n r o f f.

There are also some values that you would expect to be kept in number registers
that are actually kept in strings:

HF Fonts used for each level of heading (l=roman, 2=italic, 3=bold)

HP Point size used for each level of heading

For example, placing the foIlowing register settings at the start of your document:
.nr Hc 1
.nr Hs 3
.nr Hb 4
.nr Hi 2
.ds HF 3 3 3 3 2 2 2
.ds HP 1 6 1 4 1 2 1 0 1 0 10 1 0

will have the following effects:

Top-level headings (generated by .H1) will be centered.

The first three levels of heading will be followed by a blank line.
The fourth-level heading will be followed by a break.
Fifth- through seventh-level headings will be run-in with the text.
All headings will have the following text indented under the first word of the
heading, so that the section number hangs in the margin.
The first five levels of heading will be in bold type; the sixth and seventh will
be italic.
A first-level heading will be printed in 16-point type; a second-level heading in
14-point type; a third-level heading in 12-point type; and all subsequent levels
in 10-point type.
176 0 UNlX Text Processing 0

There isn?t space in this book for a comprehensive discussion of this topic. However, a
complete list of user-settable mm number registers is given in Appendix B. Study this
list, along with the discussion of the relevant macros, and you will begin to get a picture
of just how many facets of mm you can modify by changing the values in number
registers and strings.
The second feature-the provision of so-called ?user exit macros? at various
points-is almost as ingenious. The following macros are available for user definition:
.HX .HY .HZ .PX .TX .TY

The .HX, .HY, and .HZ macros are associated with headings. The .HX macro is
executed at the start of each heading macro, . H Y in the middle (to allow you to
respecify any settings, such as temporary indents, that were lost because of nun?s own
processing), and .HZ at the end.
By default, these macros are undefined. And, when t r o f f encounters an unde-
fined macro name, it simply ignores it. These macros thus lie hidden in the code until
you define them. By defining these macros, you can supplement the processing of
headings without actually modifying the mm code. Before you define these macros, be
sure to study the nun documentation for details of how to use them.
Similarly, .P X is executed at the top of each page, just after .PH. Accordingly,
it allows you to perform additional top-of-page processing. (In addition, you can rede-
fine the - TP macro, which prints the standard header, because this macro is relatively
self-contained.)
There is a slightly different mechanism for generalized bottom-of-page processing.
The .B S / . B E macro pair can be used to enclose text that will be printed at the bot-
tom of each page, after any footnotes but before the footer. To remove this text after
you have defined it, simply specify an empty block.
The .VM (verticd margins) macro allows you to specify additional space at the
top of the page, bottom of the page, or both. For example:
.VM 3 3
will add three lines each to the top and bottom margins. The arguments to this macro
should be unscaled. The first argument applies to the top margin, the second to the bot-
tom.
The .T X and .T Y macros allow you to control the appearance of the table of
contents pages. The .TX macro i s executed at the top of the first page of the table of
contents, above the title; .TY is executed in place of the standard title (?CON-
TENTS?).
In Chapter 14, you will learn about writing macro definitions, which should give
you the information you need to write these supplementary ?user exit macros.?
I
A P E
C H T R
.

Advanced Editing

Sometimes, in order to advance, you have to go backward. In this chapter, we are

going to demonstrate how you can improve your text-editing skills by understanding
how line editors work. This doesn’t mean you’ll have to abandon full-screen editing.
The v i editor was constructed on top of a line editor named ex, which was an
improved version of another line editor named ed. So in one sense we’ll be looking at
the ancestors of v i . We’ll look at many of the ways line editors attack certain prob-
lems and how that applies to those of us who use full-screen editors.
Line editors came into existence for use on “paper terminals,” which were basi-
cally printers. This was before the time of video display terminals. A programmer, or
some other person of great patience, worked somewhat interactively on a printer. Typi-
cally, you saw a line of your file by printing it out on paper; you entered commands
that would affect just that line; then you printed out the edited line again. Line editors
were designed for this kind of process, editing one line at a time.
People rarely edit files on paper terminals any more, but there are diehards who
still prefer line editors. For one thing, it imposes less of a burden on the computer.
Line editors display the current line; they don’t update the entire screen.
On some occasions, a line editor is simpler and faster than a full-screen editor.
Sometimes, a system’s response can be so slow that it is less frustrating to work if you
switch to a line editor. Or you may have occasion to work remotely over a dial-up line
operating at a baud rate that is too slow to work productively with a full-screen editor.
In these situations, a line editor can be a way to improve your efficiency. It can reduce
the amount of time you are waiting for the computer to respond to your commands.
The truth is, however, that after you switch from a screen editor to a line editor,
you are likely to feel deprived. But you shouldn’t skip this chapter just because you
won’t be using a line editor. The purpose of learning e x is to extend what you can do
in v i .
178 UNlX Text Processing 0

The e x Editor
The e x editor is a line editor with its own complete set of editing commands.
Although it is simpler to make most edits with v i , the line orientation of e x is an
advantage when you are making large-scale changes to more than one part of a file.
With e x , you can move easily between files and transfer text from one file to another
in a variety of ways. You can search and replace text on a line-by-line basis, or glo-
bally. You can also save a series of editing commands as a macro and access them with
a single keystroke.
Seeing how e x works when it is invoked directly will help take some of the
“mystery” out of line editors and make it more apparent to you how many e x com-
mands work.
Let’s open a file and try a few e x commands. After you invoke e x on a file,
you will see a message about the total number of lines in the file, and a colon command
prompt. For example:
$ ex intro
“intro” 20 lines, 731 characters

You won’t see any lines in the file, unless you give an e x command that causes one or
more lines to be printed.
All e x commands consist of a line address, which can simply be a line number,
and a command. You complete the command with a carriage return. A line number by
itself is equivalent to a print command for that line. So, for example, if you type the
numeral 1 at the prompt, you will see the first line of the file:
:1
Sometimes, t o advance,

To print more than one line, you can specify a range of lines. Two line numbers are
specified, separated by commas, with no spaces in between them:
:1,3
Sometimes, t o advance,
y o u have t o g o backward.
Alcuin i s a computer graphics tool

The current line is the last line affected by a command. For instance, before we issued
the command 1 , 3, line 1 was the current line; after that command, line 3 became the
current line. It can be represented by a special symbol, a dot (.).
: .,+3
that lets you design and create hand-lettered, illuminated
manuscripts, such as were created in t h e Middle Ages.

The previous command results in three more lines being printed, starting with the
current line. A + or - specifies a positive or negative offset from the current line.
 0 Advanced Editing 0 179

The e x editor has a command mode and an insert mode. To put text in a file,
you can enter the append or a command to place text on the line following the
current line. The i n s e r t or i command places text on the line above the current
line. Type in your text and when you are finished, enter a dot ( - ) on a line by itself:
:a
Monks, skilled i n calligraphy,
labored t o make copies of ancient
documents and preserve in a
library the works of many Greek and
Roman authors.

Entering the dot takes you out of insert mode and puts you back in command mode.
A line editor does not have a cursor, and you cannot move along a line of text to
a particular word. Apart from not seeing more of your file, the lack of a cursor (and
therefore cursor motion keys) is probably the most difficult thing to get used to. After
using a line editor, you long to get back to using the cw command in v i .
If you want to change a word, you have to move to the line that contains the
word, tell the editor which word on the line you want to change, and then provide its
replacement. You have to think this way to use the s u b s t i t u t e or s command.
It allows you to substitute one word for another.
We can change the last word on the first line from tool to environment:
:1
Alcuin is a computer graphics tool
:s/tool/environment/
Alcuin is a computer graphics environment

The word you want to change and its replacement are separated by slashes (/). As a
result of the substitute command, the line you changed is printed.
With a line editor, the commands that you enter affect the current line. Thus, we
made sure that the first line was our current line. We could also make the same change
by specifying the line number with the command:
:ls/environment/tool/
Alcuin is a computer graphics tool

If you specify an address, such as a range of line numbers, then the command will
affect the lines that you specify:
:1,20s/Alcuin/ALCUIN/
ALCUIN is named after an English scholar

The last line on which a substitution was made is printed.

Remember, when using a line editor, you have to tell the editor which line (or
lines) to work on as well as which command to execute.
180 0 UNIX Text Processing 0

Another reason that knowing ex is useful is that sometimes when you are work-
ing in vi, you might unexpectedly find yourself using “open mode.” For instance, if
you press Q while in vi, you will be dropped into the ex editor. You can switch to
vi by entering the command vi at the colon prompt:
:vi

After you are in vi, you can execute any ex command by first typing a :
(colon). The colon appears on the bottom of the screen and what you type will be
echoed there. Enter an ex command and press RETURN to execute it.

= Using ex Commands in v i
Many ex commands that perform normal editing operations have equivalent vi com-
mands that do the job in a simpler manner. Obviously, you will use d w or dd to
delete a single word or line rather than using the delete command in ex. How-
ever, when you want to make changes that affect numerous lines, you will find that the
ex commands are very useful. They allow you to modify large blocks of text with a
single command.
Some of these commands and their abbreviations follow. You can use the full
command name or the abbreviation. whichever is easier to remember.

delete d Delete lines

move m Move lines
COPY co Copy lines
substitute s Substitute one string for another

The substitute command best exemplifies the ex editor’s ability to make editing easier.
It gives you the ability to change any string of text every place it occurs in the file. To
perform edits on a global replacement basis requires a good deal of confidence in, as
well as full knowledge of, the use of pattern matching or “regular expressions.”
Although somewhat arcane, learning to do global replacements can be one of the most
rewarding experiences of working in the UNIX text-processing environment.
Other e x commands give you additional editing capabilities. For all practical
purposes, they can be seen as an integrated part of vi. Examples of these capabilities
are the commands for editing multiple files and executing UNIX commands. We will
look at these after we look at pattern-matching and global replacements.

Write Locally, Edit Globally

Sometimes, halfway through a document or at the end of a draft, you recognize incon-
sistencies in the way that you refer to certain things. Or, in a manual, some product
that you called by name is suddenly renamed (marketing!). Often enough, you have to
go back and change what you’ve already written in several places.
 0 Advanced Editing 181

The way to make these changes is with the search and replace commands in ex.
You can automatically replace a word (or string of characters) wherever it occurs in the
file. You have already seen one example of this use of the substitute command, when
we replaced Alcuin with ALCUIN:
:1,20s/Alcuin/ALCUIN/

There are really two steps in using a search and replace command. The first step is to
define the area in which a search will take place. The search can be specified locally to
cover a block of text or globally to cover the entire file. The second step is to specify,
using the substitute command, the text that will be removed and the text that will
replace it.
At first, the syntax for specifying a search and replace command may strike you
as difficult to learn, especially when we introduce pattern matching. Try to keep in
mind that this is a very powerful tool, one that can save you a lot of drudgery. Besides,
you will congratulate yourself when you succeed, and everyone else will think you are
very clever.

Searching Text Blocks

To define a search area, you need to be more familiar with how line addressing works
in ex. A line address simply indicates which line or range of lines an e x command
will operate on. If you don't specify a line address, the command only affects the
current line. You already know that you can indicate any individual line by specifying
its number. What we want to look at now are the various ways of indicating a block of
text in a file.
You can use absolute or relative line numbers to define a range of lines. Identify
the line number of the start of a block of text and the line number of the end of the
block. In v i , you can use " G to find the current line number.
There are also special symbols for addressing particular places in the file:

- Current line
$ Last line
9.
0 All lines (same as 1 , $)

The following are examples that define the block of text that the substitute command
will act upon:

:. , $ s Search from the current line to the end of the file

: 2 0 , .s Search from line 20 through the current line
: . , .+ 2 0 s Search from the current line through the next 20 lines
:1 0 0 , $ s Search from line 1 0 0 through the end of the file
:% S Search all lines in the file

Within the search area, as defined in these examples, the substitute command will look
for one string of text and replace it with another string.
182 0 UNlX Text Processing 0

You can also use pattern matching to specify a place in the text. A pattern -is del-
imited by a slash both before and after it.

lpatternIl,lpattern2 / s Search from the first line containing pattern1 through the
first line containing pattern2
:.,lpattern / s Search from the current line through the line containing
pattern

It is important to note that the action takes place on the entire line containing the pat-
tern, not simply the text up to the pattern.

Search and Replace

You’ve already seen the substitute command used to replace one string with another
one. A slash is used as a delimiter separating the old string and the new. By prefixing
the s command with an address, you can extend its range beyond a single line:
:1,20s/Alcuin/ALCUIN/

Combined with a line address, this command searches all the lines within the block of
text. But it only replaces the first occurrence of the pattern on each line. For instance,
if we specified a substitute command replacing roman with Roman in the following
line:
after t h e roman hand. In teaching t h e roman script

only the first, not the second, occurrence of the word would be changed.
To specify each occurrence on the line, you have to add a g at the end of the
command:
:s/roman/Roman/g

This command changes every occurrence of roman to Roman on the current line.
Using search and replace is much faster than finding each instance of a string and
replacing it individually. It has many applications, especially if you are a poor speller.
So far, we have replaced one word with another word. Usually, it’s not that easy.
A word may have a prefix or suffix that throws things off. In a while, we will look at
pattern matching. This will really expand what you are able to do. But first, we want
to look at how to specify that a search and replace take place globally in a file.

Confirming Substitutions
It is understandable if you are over-careful when using a search and replace command.
It does happen that what you get is not what you expected. You can undo any search
and replacement command by entering u. But you don’t always catch undesired
changes until it is too late to undo them. Another way to protect your edited file is to
save the file with :w before performing a replacement. Then, at least you can quit the
file without saving your edits and go back to where you were before the change was
made. You can also use :e ! to read in the previous version of the buffer.
 0 Advanced Editing 183

It may be best to be cautious and know exactly what is going to be changed in

your file. If you’d like to see what the search turns up and confirm each replacement
before it is made, add a c at the end of the substitute command:
:1,30s/his/the/gc

It will display the entire line where the string has been located and the string itself will
be marked by a series of carets (AAA).
copyists at his school
.-.An

If you want to make the replacement, you must enter y and press RETURN.
If you don’t want to make a change, simply press RETURN.
this can b e used for invitations, signs, and menus.
h h h

The combination of the v i commands // (repeat last search) and . (repeat last com-
mand) is also an extraordinarily useful (and quick) way to page through a file and make
repetitive changes that require a judgment call rather than an absolute global replace-
ment.

Global Search and Replace

When we looked at line addressing symbols, the percent symbol, %, was introduced. If
you specify it with the substitute command, the search and replace command will affect
all lines in the file:
:%s/Alcuin/ALCUIN/g

This command searches all lines and replaces each occurrence on a line.
There is another way to do this, which is slightly more complex but has other
benefits. The pattern is specified as part of the address, preceded by a g indicating that
the search is global:
:g/Alcuin/s//ALCUIN/g

It selects all lines containing the pattern Alcuin and replaces every occurrence of that
pattern with ALCUIN. Because the search pattern is the same as the word you want to
change, you don’t have to repeat it in the substitute command.
The extra benefit that this gives is the ability to search for a pattern and then
make a different substitution. We call this context-sensitive replacement.
The gist of this command is globally search for a pattern:
:g /pattern/

Replace it:
:g/pattern/ s / /

or replace another string on that Iine:

184 0 UNlX Text Processing 0

:g/pattern/ s /string/

with a new string:

:g/pattern/ s/string/new/

and do this for every occurrence on the line:

:g/pattern/ s /string/new/g

For example, we use the macro .BX to draw a box around the name of a special key.
To show an ESCAPE key in a manual, we enter:
.BX E s c

Suppose we had to change Esc to ESC, but we didn’t want to change any references to
Escape in the text. W e could use the following command to make the change:
:g/BX/s/Esc/ESC/

This command might be phrased: “Globally search for each instance of B X and on
those lines substitute the Esc with ESC”. We didn’t specify g at the end of the
command because we would not expect more than one occurrence per line.
Actually, after you get used to this syntax, and admit that it is a little awkward,
you may begin to like it.

Pattern Matching
If you are familiar with grep, then you know something about regular expressions. In
making global replacements, you can search not just for fixed strings of characters, but
also for patterns of words, referred to as regular expressions.
When you specify a literal string of characters, the search might turn up other
occurrences that you didn’t want to match. The problem with searching for words in a
file is that a word can be used in many different ways. Regular expressions help you
conduct a search for words in context.
Regular expressions are made up by combining normal characters with a number
of special characters. The special characters and their use follow.*

Matches any single character except newline.

* Matches any number (including 0) of the single character (including a
character specified by a regular expression) that immediately precedes
it. For example, because . (dot) means any character, - * means
match any number of any character.

*\( and\), and \{n,rn\] are not supported in all versions of v i . \<, \>, \u,\u,U, a n d k are supported
only in v i / e x , and not in other programs using regular expressions.
 0 AdvancedEditing 0 185

Matches any one of the characters enclosed between the brackets.

For example, [AB]matches either A or B. A range of consecutive
characters can be specified by separating the first and last characters
in the range with a hyphen. For example, [A-21 will match any
uppercase letter from A to Z and [ 0-91 will match any digit from 0
to 9.
Matches a range of occurrences of the single character (including a
character specified by a regular expression) that immediately precedes
it. The n and m are integers between 0 and 256 that specify how
many occurrences to match. \(n\} will match exactly n occurrences,
\{ n,\] will match at least n occurrences, and \( n,m\} will match any
number of occurrences between n and rn. For example, A \ { 2 , 3 \ }
will match either AA (as in AARDVARK or AAA but will not match
the single letter A).
Requires that the following regular expression be found at the begin-
ning of the line.
Requires that the preceding regular expression be found at the end of
the line.
Treats the following special character as an ordinary character. For
example, \ . stands for a period and \* for an asterisk.
Saves the pattern enclosed between \( and \) in a special holding
space. Up to nine patterns can be saved in this way on a single line.
They can be “replayed” in substitutions by the escape sequences \ 1
to \ 9 .
Matches the nth pattern previously saved by \( and \), where n is a
number from 0 to 9 and previously saved patterns are counted from
the left on the line.
Matches characters at the beginning (\<) or at the end ( \ > ) of a
word. The expression \<ac would only match words that begin
with ac, such as action but not react.
Prints the entire search pattern when used in a replacement string.
Converts the first character of the replacement string to uppercase.
Converts the replacement string to uppercase as in :/Unix/ \U& /.
Converts the first character of the replacement string to lowercase, as
in : s / Act/\l&/.
Converts the replacement string to lowercase.

Unless you are already familiar with UNIX’s wildcard characters, this list of special
characters probably looks complex. A few examples should make things clearer. In the
examples that follow, a square (0)is used to mark a blank space.
186 0 UNlX Text Processing 0

Let’s follow how you might use some special characters in a replacement. Sup-
pose you have a long file and you want to substitute the word balls for the word ball
throughout that file. You first save the edited buffer with :w, then try the global
replacement:
:g/ball/s//balls/g

When you continue editing, you notice occurrences of words such as ballsoon, glo-
ballsy, and ballss. Returning to the last saved buffer with :e ! , you now try specifying
a space after bull to limit the search:
:g/ba110/s//ballsO/g
But this command misses the occurrences ball., ball,,ball:, and so on.
:g/\<ball\>/s//balls/g

By surrounding the search pattern with \ < and \>, we specify that the pattern should
only match entire words, with or without a subsequent punctuation mark. Thus, it does
not match the word balls if it already exists.
Because the \ < and \ > are only available in ex (and thus v i ) , you may have
occasions to use a longer form:
:g/ball\ ( [a, .; : !7 1 \ ) /s//balls\l/g
This searches for and replaces ball followed by either a space (indicated by n) or any
one of the punctuation characters ,.;:! ?. Additionally, the character that is matched
is saved using \ ( and \ ) and restored on the right-hand side with \ 1. The syntax
may seem complicated, but this command sequence can save you a lot of work in a
similar replacement situation.

Search for General Classes of Words

The special character & is used in the replacement portion of a substitution command
to represent the pattern that was matched. It can be useful in searching for and chang-
ing similar but different words and phrases.
For instance, a manufacturer decides to make a minor change to the names of
their computer models, necessitating a change in a marketing brochure. The HX5000
model has been renamed the Series HX.5000,along with the HX6000 and HX8500
models. Here’s a way to do this using the & character:
:g/HX[568][05]00/s//Series & / g
This changes HX8500 to Series HX8500. The & character is useful when you want to
replay the entire search pattern and add to it. If you want to capture only part of the
search pattern, you must use \ ( and \ ) and replay the saved pattern with
\ l .. . \n.)
For instance, the same computer manufacturer decides to drop the HX from the
model numbers and place Series after that number. We could make the change using
the following command:
 0 AdvancedEditing 187

: g / \ (Series\) HX\ ([568]) [05]O O \ ) / s / / \ 2 \l/g

This command replaces Series HX8500 with 8500 Series.
Suppose you have subroutine names beginning with the prefixes mgi, mgr,and
mga.

mgibox routine
mgrbox routine
mgabox routine

If you want to save the prefixes, but want to change the name box to square, either of
the following replacement commands will do the trick:
:g/mg([iar])box/s//mg\lsquare/

The global replacement keeps track of whether an i, a,or r is saved, so that only
box is changed to square. This has the same effect as the previous command:
:g/mg[iar]box/s/box/square/g
The result is:

mgisquare routine
mgrsquare routine
mgasquare routine

Block Move by Patterns

You can edit blocks of text delimited by patterns. For example, assume you have a 150
page reference manual. All references pages are organized in the same way: a para-
graph with the heading SYNTAX, followed by DESCRIPTION, followed by PARAME-
TERS. A sample of one reference page follows:
.Rh 0 "Get status of named file" "STAT"
.Rh "SYNTAX"
.nf
integer*4 stat, retval
integer*4 status (11)
character*123 filename
...
retval = stat (filename, status)
.fi
.Rh "DESCRIPTION"
Writes the fields of a system data structure into the
status array. These fields contain (among other
things) information about the file's location, access
privileges, owner, and time of last modification.
.Rh "PARAMETERS"
.IP "filename" 15n
188 0 UNIX Text Processing 0

A character string variable or constant containing

the U N I X pathname f o r the file whose status you want
t o retrieve. You can give the...

Suppose that you decide to move DESCRIPTION above the SYNTAX paragraph. With
pattern matching, you can move blocks of text on all 150 pages with one command!
:g/SYNTAX/,/DESCRIPTION/-l,mo/PARAMETERS/-1
This command moves the block of text between the line containing the word SYNTAX
and the line just before the word DESCRIPTION ( / D E S C R I P T I O N / - l ) to the line
just before PARAMETERS. In a case like this, one command literally saves hours of
work.
This applies equally well to other ex commands. For example, if you wanted to
delete all DESCRIPTION paragraphs in the reference chapter, you could enter:
:g/SYNTAX/,/DESCRIPTION/-l,d

This very powerful kind of change is implicit in the e x editor’s line addressing syntax,
but is not readily apparent. For this reason, whenever you are faced with a complex,
repetitive editing task, take the time to analyze the problem and find out if you can
apply pattern-matching tools to do the job.

More Examples
Because the best way to learn pattern matching i s by example, the following section
gives a list of examples with brief explanations. Study the syntax carefully, so that you
understand the principles at work. You should then be able to adapt them to your situa-
tion.

1. Delete all blank lines:

:g/*S/d

What you are matching is the beginning of the line followed by the end of
the line, with nothing in between.
2. Put t rof f italic codes around the word RETURN
:g/RETURN/s//\\fIRETURN\\fR/g
Notice that two backslashes ( \ \) are needed in the replacement, because
the backslash in the t r o f f italic code will be interpreted as a special
character. (\ f I alone would be interpreted as f I; it takes \ \ f I to get
\ f I.>
3. Modify a list of pathnames in a file:
:g/\/usr\/tim/s//\/usr\/linda/g
A slash (used as a delimiter in the global replacement sequence) must be
escaped with a backslash when it is part of the pattern or replacement; use
\ / to get /. Another way to achieve this same effect is to use a different
 0 AdvancedEditing 189

character as the pattern delimiter. For example, you could make the previ-
ous replacement as follows:
:g:/usr/tim:s::/usr/linda:g

4. Change all periods to semicolons in lines 1 to 10:

:l,lOq/\-/s//;/q
A period i s a special character and must be escaped with a backslash.
5. Reverse the order of all hyphen-separated items in a list:
:g/\~.*\~O-O\~.*\~/s//\ZO-~\l/
The effect of this command on several items is:
more-display filesbecomes display files-more
1p-print files becomes print files-lp

6. Standardize various uses of a word or heading:

:g/"Example [Os:]/s//Examples:O/g

Note that the brackets enclose three characters: a space (represented in the
example by n), a colon, and the letter s. Therefore, this command searches
for Examplen Examples, or Example: at the beginning of a line and
replaces it with Examples:. (If you don't include the space, Examples
would be replaced with Exampless:.)
As another similar example, change all occurrences of the word help (or
Help) to HELP:
:g/[Hhlelp/s//HELP/g

7. Replace one or more spaces with a single space:

:g/m*/s//o/g

Make sure you understand how the asterisk works as a special character.
An asterisk following any character (or any regular expression that matches
a single character, such as . or [a-z] ) matches zero or more instances
of that character. Therefore, you must specify two spaces followed by an
asterisk to match one or more spaces (one plus zero or more).
8. Replace one or more spaces following a colon with two spaces:
:g/:m*/s//:m/g

9. Replace one or more spaces following a period or a colon with two spaces:
:q/\ ( [:.I \ ) m * / s / / \ i m / g
Either of the two characters within brackets can be matched. This character
is saved, using parentheses, and restored on the right-hand side as 1. Note
that a special character such as a period does not need to be escaped within
brackets.
190 UNIX Text Processing 0

10. Delete all leading blanks on a line:

: g / v n * \ ( - * \ ) /s//\i/g

Search for one or more blanks at the beginning of a line; save the rest of the
line and replace it without any leading blanks.
11. Delete all trailing blanks:
:g/m*s/~///
12. Remove manual numbering from section headings (e.g., 1.1 Introduction) in
a document:
:g/[l-9]\. [1-9]*\(.*\)/s//\l/g
A hyphen-separated pair of letters or digits enclosed in square brackets (e.g,
[ 1 - 9 ] ) specifies a range of characters.

13. Change manually numbered section heads (e.g., 1.1, 1.2) to a t r o f f

macro (e.g., .A h for an A-level heading):
:g/^[l-9]\. [l-g]/s//\.Ah/

14. Show macros in the output by protecting them from interpretation. Putting
\ & in front of a macro prevents t rof f from expanding them. This com-
mand was used frequently throughout this book to print an example that
contained macros. Three backslashes are needed in the replacement pattern:
two to print a backslash and one to have the first ampersand interpreted
literally.
:g/*\./s//\\\&&/

Writing and Quitting Files

You have learned the v i command Z Z to quit and write (save) your file. But you
will usually want to exit a file using e x commands, because these commands give you
greater control.

:W Writes (saves) the buffer to the file but does not exit. You can use :w
throughout your editing session to protect your edits against system
failure or a major editing error.
:q Quits the file (and returns to the UNIX prompt).
:w q Both writes and quits the file.

The v i editor protects existing files and your edits in the buffer. For example, if
you want to write your buffer to an existing file, v i will give you a warning, because
this would delete the original file. Likewise, if you have invoked v i on a file, made
 0 Advanced Editing 191

edits, and want to quit without saving the edits, v i will give you an error message
such as:
N o write since last change.

These warnings can prevent costly mistakes, but sometimes you want to proceed with
the command anyway. An exclamation mark (!) after your command ovemdes this
warning:
:w! filename
:q!

The :q ! command is an essential editing command that allows you to quit without
affecting the original file, regardless of any changes you made in the session. The con-
tents of the buffer are discarded.

Renaming the Buffer

You can also use :w to save the entire buffer (the copy of the file you are editing)
under a new filename.
Suppose that you have a file l e t t e r that contains 600 lines. You call in a
copy and make extensive edits. You want to quit and save both the old version of
l e t t e r and your new edits for comparison. To rename your buffer l e t t e r -new,
give the command:
:wq letter-new

Saving Part of a File

In an editing session, you will sometimes want to save just part of your file as a
separate, new file. For example, you might have entered formatting codes and text that
you want to use as a header for several files.
You can combine e x line addressing with the write command, w, to save part of
a file. For example, if you are in the file l e t t e r and want to save part of l e t t e r
as the file newf i l e , you could enter:
:230, S w newfile

which saves from line 230 to the end of the file, or:
.
: , 6 0 0 ~newf ile

which saves from the current line to line 600 in newf i l e .

Appending to a Saved File

You can use the UNIX redirect and append operator (>>) with w to append the con-
tents of the buffer to an existing file. For example:
192 0 UNIX Text Processing 0

:1,1Ow newfile
:340, $w>>newfile

The existing file, newf i l e , will contain lines 1 through 10, and from line 340 to the
end of the buffer.

Reading In a File
Sometimes you want to copy text or data already entered on the system into the file you
are editing. In vi, you can read in the contents of another file with the ex command:
:read filename

or :
:r filename
This reads in the contents offilename on the line after the cursor position in the file.
Let's suppose that you are editing the file l e t t e r , and want to read in data
from a file in another directory called /work/ a 1c u i n / c h 0 1 . Position the cursor
just above the line where you want the new data inserted, and enter:
:r /work/alcuin/chOl

The entire contents of /work/alcuin/chOl are read into l e t t e r , beginning

below your cursor position.

Executing UNIX Commands

You can also display or read in the results of any UNIX command while you are editing
in v i . An exclamation mark (!) tells ex to create a shell and regard what follows as
a UNIX command.
: !command

So, if you are editing and want to check the time or date without exiting v i , you can
enter:
: !date

The time and date will appear on your screen; press RETURN to continue editing at the
same place in your file. If you want to give several UNIX commands in a row, without
returning to v i in between, you can create a shell with the ex command:
:sh

When you want to exit the shell and return to vi, press "D.
You can combine :read with a call to UNIX, to read the results of a UNIX
command into your file. As a very simple example:
 0 Advanced Editing 0 193

:r !date

This will read in the system's date information into the text of your file.
Suppose that you are editing a file, and want to read in four phone numbers from
a file called phone, but in alphabetical order. The phone file is in the following
order:
Willing, Sue 333-4444
Walsh, Linda 555-6666
Quercia, Valerie 777-8888
Dougherty, Nancy 999-0000

The command:
:r !sort phone

reads in the contents of phone after they have been passed through the s o r t filter:
Dougherty, Nancy 999-0000
Quercia, Valerie 777-8888
Walsh, Linda 555-6666
Willing, Sue 333-4444

Suppose that you are editing a file and want to insert text from another file in the direc-
tory, but you can't remember the new file's name.
You could perform this task the long way: exit your file, give the 1s command,
note the correct filename, reenter your file, and search for your place.
Or, you could do the task in fewer steps. The command : ! Is will display a list
of files in the directory. Note the correct filename. Press RETURN to continue editing.
file1
file2
letter
newfile

The command:
: r newfile

will read in the new file:

"newfile" 35 lines, 949 characters

Filtering Text through a Command

You can also send a block of text as standard input to a UNIX command. The output
from this command replaces the block of text in the buffer. Filtering text through a
command can be done either from ex or v i . The main difference between the two
methods is that the block of text is indicated with line addresses in e x and with text
objects in v i .
The first example demonstrates how to do this with ex. Assume that instead of
being contained in a separate file called phone, the list of names in the preceding
example was already contained in the current file, on lines 96 to 99.
194 UNIX Text Processing 0

You simply type the addresses of the lines you want affected, followed by an
exclamation mark and the UN1X command line to be executed. For example, the com-
mand:
:9 6 , 9 9 ! sort

will pass lines 96 to 99 through the s o r t filter, and replace those lines with the output
of s o r t .
In v i , this sequence is invoked by typing an exclamation mark followed by any
v i objects that indicate a block of text, and then the UNIX command line to be exe-
cuted. For example:
! ) command

will pass the next sentence through command.

There are some unusual features about how v i acts when you use this feature.
First, the exclamation mark that you type is not echoed right away. When you type the
symbol for the text object to be affected, the exclamation mark appears at the bottom of
the screen, but the symbol you type to reference the object does not.
Second, only objects that refer to more than one line of text (G, { 1 , ( ) , [ I )
can be used. A number may precede either the exclamation mark or the object to repeat
the effect. Objects such as w do not not work unless enough of them are specified so
as to exceed a single line. A slash (/) followed by a pattern and a RETURN can also be
specified, taking the text up to the pattern as input to the command.
Third, there is a special object that is used only with this command syntax. The
current line can be specified by entering a second exclamation mark:
! !command

Either the entire sequence or the text object can be preceded by a number to repeat the
effect. For instance, to change the same lines as in the previous example, you could
position the cursor on line 96, and enter:
4 ! ! sort

or:
! 4 !sort

As another example, assume you have a portion of text in a file that you want to
change from lowercase to uppercase letters. You could process that portion with the
t r command. In these examples, the second sentence is the block of text that will be
filtered to the command. An exclamation mark appears on the last line to prompt you
for the UNIX command:
 0 Advanced Editing 0 195

of the product. of the product.

-
I confess t o being !I -
I confess to being
amazed by Alcuin. ! appears on amazed by Alcuin.
Some people around last line Some people around
!-

Enter the UNIX command and press RETURN. The input is replaced by the output.

of the product. of the product.

-
I confess t o being tr’[a-z] ‘ -
I CONFESS TO BEING
amazed by Alcuin. ‘[A-Z] ‘ AMAZED BY ALCUIN.
Some people around input replaced Some people around
by output

To repeat the previous command, the syntax is:

! block !

It is sometimes useful to send sections of a coded document to nrof f to be replaced

by formatted output. However, remember that the “original” input is replaced by the
output.
If there i s a mistake, such as an error message being sent instead of the expected
output, you can undo the command and restore the lines.

. Editing Multiple Files

The e x commands enable you to edit multiple files. The advantage to editing multiple
files is speed. When you are sharing the system with other users, it takes time to exit
and reenter v i for each file you want to edit. Staying in the same editing session and
traveling between files is not only faster in access time: you save abbreviations and
command sequences you have defined and keep named buffers so that you can copy
text from one file to another.

Invoking v i on Multiple Files

When you first invoke v i , you can name more than one file to edit files sequentially,
and then use e x commands to travel between the files. The following:
196 UNlX Text Processing 0

$ vi file1 file2
invokes file1 first. After you have finished editing the first file, the ex command :w
writes (saves)filel, and :n calls in the next file (file2).
Suppose that you know you want to edit two files, l e t t e r and note. Open
the two files by typing:
$ vi letter note
The message:
Two f i l e s t o edit

appears on the screen. The first named file, l e t t e r , appears. Perform your edits to
l e t t e r , and then save it with the ex command :w. Call in the next file, note,
with the ex command : n and press RETURN. Perform any edits and use : w q to
quit the editing session.
There is no practical limit to the number of files you can invoke v i on at one
time. You can use any of the shell's pattern-matching characters, or even more com-
plex constructions. Suppose you were writing a program, and wanted to change the
name of a function call, for example, g e t c u r s o r . The command:
$ vi 'grep -1 getcursor *'
would invoke v i on all of the files in the current directory containing the string
g e t c u r s o r . The command:
$ grep -1
prints the names of all files containing a string; using a command enclosed in
backquotes (") as an argument to another command causes the shell to use the output
of the command in backquotes as the argument list for the first command.
The v i editor will print a message similar to:
5 f i l e s t o edit

before displaying the first file.

If you try to quit without editing all of the files, v i will issue a warning mes-
sage:
4 more f i l e s t o e d i t

You must type :q ! if you want to exit without editing all of the files.

Calling In New Files

You don't have to call in multiple files at the beginning of your editing session. Any
time in v i , you can switch to another file with the ex command :e. If you want to
edit another file within v i , first save your current file (:w), then give the command:
:e filename
Suppose that you are editing the file l e t t e r , and want to edit the file note and
then return to l e t t e r .
 0 AdvancedEditing 0 197

Save l e t t e r with w and press RETURN. The file l e t t e r is saved and

remains on the screen. You can now switch to another file, because your edits are
saved. Call in the file l e t t e r with :e and press RETURN.
The v i editor “remembers” two filenames at a time as the current and alternate
filenames. These can be referred to by the symbols % (current filename) and # (alter-
nate filename). The # symbol is particularly useful with :e, because it allows you to
switch easily back and forth between files. In the example just given, you could return
to the first file, l e t t e r , by typing the command :e#.
If you have not first saved the current file, v i will not allow you to switch files
with :e or :n unless you tell it imperatively to do so by adding an exclamation mark
after the command. For example, if after making some edits to note, you wanted to
discard the edits and return to l e t t e r , you could type :e ! #.
The command:
e!
is also useful. It discards your edits and returns to the last saved version of the current
file. The % symbol, by contrast, is useful mainly when writing out the contents of the
buffer to a new file. For example, a few pages earlier we showed how to save a second
version of the file 1 e t t e r with the command:
:w l e t t e r - n e w

This could also have been typed:

:w %.new

Edits between Files

Named buffers provide one convenient way to move text from one file to another.
Named buffers are not cleared when a new file is loaded into the v i buffer with the
:e command. Thus, by yanking text in one file (into multiple named buffers if neces-
s a r y ) , reading in a new file with :e, and putting the named buffer into the new file,
material can be transferred selectively between files.
The following example illustrates transfemng text from one file to another.

-
In our conversation I n our conversation
last Thursday, we “f 6yy last Thursday, w e
discussed a yank 6 lines discussed a
documentation project to bufferf documentation project
that would produce a that would produce a
user‘s manual o n t h e . .- user‘s manual on t h e . -.
6 lines vanked

Save the file with the :w command. Enter the file note with :e, and move the
cursor to where the copied text will be placed.
198 0 UNlX Text Processing 0

Dear Mr. Caslon, Dear Mr. Caslon,

Thank you f o r . . . " fP -
In our conversation
put yanked text last Thursday, we dis-
below cursor cussed a documentation
project that would
produce a user's
manual on t h e . . .
Thank you for - . .

Word Abbreviation
Often, you will type the same long phrases over and over in a file. You can define
abbreviations that v i will automatically expand into the full text whenever you type
the abbreviation in insert mode. To define an abbreviation, use the e x command:
:a b abbr phrase
Where abbr is an abbreviation for the specified phrase. The sequence of characters that
make up the abbreviation will be expanded in insert mode only if you type it as a full
word; abbr will not be expanded within a word.
Suppose that in the file letter you want to enter text that contains a fre-
quently recurring phrase, such as a difficult product or company name. The command:
:ab IMRC International Materials Research Center

abbreviates International Materials Research Center to the initials IMRC.

Now when you type IMRC in insert mode:
i t h e IMRC
IMRC expands to the full text:
t h e International Materials Research Center
When you are choosing abbreviations, select combinations of characters that don't ordi-
narily occur while you are typing text.

Saving Commands with map

While you are editing, you may use a particular command sequence frequently, or you
may occasionally use a very complex command sequence. To save keystrokes, or the
time that it takes to remember the sequence, you can assign the sequence to an unused
key.
The map command acts a lot like ab except that you define a macro for com-
mand mode instead of insert mode.
 0 Advanced Editing 0 199

:map x sequence Define character x as a sequence of editing commands

:unmap x Disable the sequence defined for x
:map List the characters that are currently mapped

Before you can start creating your own maps, you need to know the keys not used in
command mode that are available for user-defined commands:
^A g K ^K
^O q ^T V
V ^W ^X ^Z
* \ - (underscore)
Depending on your terminal, you may also be able to associate map sequences with spe-
cial function keys. With maps, you can create simple or complex command sequences.
As a simple example, you could define a command to reverse the order of words. In
v i , with the cursor as shown:
you can t h e scroll page

the sequence to put the after scroll would be d w e l p : delete word, d w ; move to the
end of next word, e; move one space to the right, 1 ; put the deleted word there, p.
Saving this sequence:
:map v dwelp

enables you to reverse the order of two words anytime in the editing session with the
single keystroke v.
Note that when defining a map, you cannot simply type certain keys, such as
RETURN, ESC, TAB, BACKSPACE, and DELETE, as part of the map command. If
you want to include one of these keys as part of the command sequence, preface that
key with a ^v. The keystroke ^V appears in the map as the character. Characters
A

following the ^ V also do not appear as you expect. For example, a carriage return
appears as ^M,escape as * [, tab as ^ I, and so on.
You can undo the effect of any map sequence with the u command. Fortunately,
the undo restores the file as it was before you executed the map sequence, treating the
series of commands as though it were a single v i command.
Unless you use unmap to remove a mapped key, its special meaning is in effect
for as long as your current session, even if you move between files. It can therefore be
a convenient way of making the same edits in a number of files.
All the v i and e x commands can be used in map sequences, with the exception
that the p or p u t command cannot be used to replace entire lines yanked in the same
mapping. If you try to yank and then put back a deleted line within a map, you will get
the error message:
Cannot put inside global macro.

If you want to move lines from one place to another within a mapping, you can usually
get around this restriction using the e x editor’s c o p y or co command.
200 0 UNlX Text Processing

Complex Mapping Example

Assume that you have a glossary with entries like this:
map - an ex command that allows you to associate
a complex command sequence with a single key.
You would like to convert this glossary list to nrof f format, so that it looks like this:
.IP "map" 10n
An ex command...

The best way to do this is to perform the edit on one of the entries and write down the
sequence of commands. You want to:

1. Insert the macro for an indented paragraph at the beginning of the line.
2. Press ESC to terminate insert mode.
3. Move to the end of the word and add the size of the indent.
4. Press RETURN to insert a new line.
5. Press ESC to terminate insert mode.
6. Remove the hyphen and capitalize the next word.

That's quite an editing chore if you have to repeat it more than a few times! With
:map, you can save the entire sequence so that it can be re-executed with a single key-
stroke:
:map z I.IP ""[ea" 10nAM"[3x-

The sequence " [ appears when you type "v followed by ESC. The sequence "M is
shown when you type "v RETURN.
Now, simply typing z will perform the entire series of edits. On a slow terminal,
you can actually see the edits happening individually. On a fast terminal, it will seem
to happen by magic.
Don't be discouraged if your first attempt at key mapping fails. A small error in
defining the map can give you very different results than you expect. Simply type u to
undo the edit, and try again.
Remember, the best way to define a complex map is to do the edit once manually,
writing down each keystroke that you must type.

Mapping Keys for Insert Mode

Normally, maps apply only to command mode-after all, in insert mode, keys stand for
themselves, and shouldn't be mapped as commands.
However, by adding an exclamation mark ( ! ) to the m a p command, you can
force it to override the ordinary meaning of a key and produce the map in insert mode.
You may find this feature appropriate for tying character strings to special keys that you
wouldn't otherwise use. It is especially useful with programmable function keys, as
we'll see in a minute. Many terminals have programmable function keys. You can
 0 Advanced Editing 0 201

usually set up these keys to print whatever character or characters you want using a spe-
cial setup mode on the terminal. But this will limit you to a particular terminal, and
may limit the actions of programs that want to set up those function keys themselves.
The e x editor allows you to map function keys by number, using the syntax:
:map #I commands

for function key number 1 , and so on. (It can do this because the editor has access to
the entry for that terminal found in either the t e r m c a p or t e r m i n f o database and
knows the escape sequence normally output by the function key. )
As with other keys, maps apply by default to command mode, but by using the
m a p ! commands as well, you can define two separate values for a function key-ne
to use in command mode, the other in insert mode. For example, if you are a t r o f f
user, you might want to put font-switch codes on function keys. For example:
:map # 1 i\f ( C w ^ [
:map! # 1 \fI

If you are in command mode, the first function key will enter insert mode, type in the
three characters \ f I, and return to command mode. If you are already in insert mode,
the key will simply type the three-character t r o f f code.
Note: If function keys have been redefined in the terminal's setup mode, the #n
syntax might not work because the function keys no longer put out the expected control
or escape sequence as described in the terminal database entry. You will need to exam-
ine the t e r m c a p entry (or t e r m i n f o source) for your terminal and check the
definitions for the function keys. The terminal capabilties kl, k2 through k 9 , k0
describe the first ten function keys. The capabilities 1 1 ,12 through 1 9 , 1 0
describe the remaining function keys. Using your terminal's setup mode, you can
change the control or escape sequence output by the function key to correspond with the
t e r m c a p or t e r m i n f o entry. (If the sequence contains "M, which is a carriage
return, press "M, not the RETURN key.) For instance, to have function key 1 available
for mapping, the terminal database entry for your terminal must have a definition of
kl, such as k l = ^ A @ . In turn, the definition ^A@ must be what is output when you
press that key. To test what the function key puts out, press the key at the UNIX
prompt, followed by a RETURN if necessary. The shell should dispiay the sequence
output by the function key after trying unsuccessfully to execute it as a command.

(3 Functions
Named buffers provide yet another way to create macros+omplex command
sequences that you can repeat with only a few keystrokes.
If you type a command line in your text (either a v i sequence or an e x com-
mand preceded by a colon), then yank or delete it into a named buffer, you can execute
the contents of that buffer with the @ command. It works in the same way as a m a p
sequence, except that you enter the command line in the file instead of at the colon
prompt; this is helpful if the command sequence is long and might need editing to work
properly. Let's look at a simple but not very useful example of an @ function. In your
file, enter this key sequence:
202 0 UNlX Text Processing 0

cw\fIgadfly\fR*VESC

This will appear on your screen as:

cw\fIgadfly\fR^[

Then delete your command line into buffer g by typing "gdd. Now, whenever you
place the cursor at the beginning of a word and type @g, that word in your text will be
changed to gadfly. Because @ is interpreted as a v i command, . will repeat the entire
sequence, even if it i s an e x command. The command @ @ repeats the last @, and u
or U can be used to undo the effect of @. The @ function is useful because you can
create very specific commands. It is especially useful when you are making specific
editing commands between files, because you can store the commands in named buffers
and access them in any file you edit.
 C H A P T E R
9 9 9 9 S 9

Formatting with tbl

Some information is best presented in tabular format, that is, displayed in rows and
columns. You can structure data in columns using tabs, but that can be difficult, espe-
cially if the table consists of long lines of text. The t b l preprocessor was designed to
make it easier to prepare complicated tables, such as the following.

' Production of Audio Equipment

(units: loo0 sets)
Product 1984 1985
1 General radio 8,895 8,770
Clock radio 5,467 6,500
Radiokassette 29,734 27,523
Tape deck 11,788 14,300
Car radio 9,450 10,398
Car stereo 15.670 17.456

With t b l , you can center, left justify, and right justify columns of data or align
numeric data within a column. You can put headings that span one or more columns or
rows, and draw horizontal and vertical lines to box individual entries or the whole table.
An entry may contain equations or consist of several lines of text, as is usually the case
with descriptive tables. A table can have as many as 35 columns and essentially an
unlimited number of rows.
When you use t b l , you should have an idea or, better still, a written design of
the table. Then, using a few t b l specifications, you can define how a formatted table
should look, The data is entered row by row; each column is separated by ordinary
tabs.
For example, the t b l description for the previous table looks like this:
204 0 UNlX Text Processing 0

.TS
center,box;
c s s
c s s
c c c
1 r r.
Production of Audio Equipment
(units: 1000 sets)
-
Product 1984 1985
-
General radio 8,895 8,770
Clock radio 5,467 6,500
Radio/cassette 29,734 27,523
Tape deck 11,788 14,300
C a r radio 9,450 10,398
Car stereo 15,670 17,456
.TE

When t b l processes the specifications, it calculates all the values needed to produce
the table and passes these values to nrof f or t r o f f, which formats or outputs the
final table.
In this chapter, we will show you how to use t b l to specify the general appear-
ance o f a table. We begin with some very simple examples, then gradually work up to
more complicated ones to show all o f t b l ’ s capabilities.

Usingtbl
The t b l description can be written in a file or as part of a larger file that contains
other tables and text. You can format a table in a file using the t b l command as in
the following:
$ tblfile I troff
$ tblfile I nroff
The t b l command writes its results to standard output. Because you will probably
not be interested in the generated formatting requests, you would normally pipe the out-
put to n r o f f or t r o f f and then to a printer.
The t b l command also accepts a list of filenames as input and processes them
one by one in the order in which they are named on the command line. I f you don’t
give any filenames, t b l reads from standard input. The standard input may also be
read in the middle of a list of files by typing a minus sign at the desired place.
If you’re using a line printer that doesn’t have fractional or reverse line motions,
use the -T option of nrof f and give the type of output device you’re using. This is
important when you’re using n r o f f together with t b l to create boxed tables. For
example, if you’re using a regular line printer, the option should read -Tlp. You
 0 Formatting with t b l 0 205

must also pipe the n r o f f output to a program called c o l , which filters the reverse
linefeeds. The command line for a table with boxes would then read:
$ tblfile I n r o f f -Tlp I c o l

t b l with eqn
When you have equations within your table and you use the eqn preprocessor to for-
mat them, invoke t b l before eqn. The tbl command usually executes faster
because eqn normally produces a larger amount o f output. To use e q n with t b l ,
use the following command line:
$ tblfile I eqn I t r o f f
There is a possible complication that can occur with any o f the preprocessors
( t b l , e q n , or pic). If you read in subsidiary files with the .so request, those
files will never be passed through the preprocessor, since the .s o request has not been
encountered yet by the preprocessor. Some UNIX systems support a program called
soelim, which works just like c a t , except that it reads in files called by .so
requests. If any subsidiary files contain data that must be processed, start your com-
mand line with soelim:
$ soelim file I t b l I eqn . . . I n r o f f

Specifying Tables
A table is always indicated by a .TS (table start) at the beginning o f the table descrip-
tion and a .TE (table end) at the end. The general format of each table looks like this:
- TS
global options line;
format section.
data
- TE
These delimiters serve two functions. First, they signal to tbl the beginning and end
of the table description. The t b l program processes the table, and enables formatting
requests into the text o f the table. The TS and .TE lines remain after processing by
t b l . This allows them to be used as macro calls by n r o f f and t r o f f . Both ms
and mm define these macros; however, an enterprising user can redefine them, and sur-
round a table with consistent formatting effects. If the macros are undefined, t b l will
not suffer in any way because the use o f .TS/ .T E as delimiters is separate from their
secondary use as macros.
As you can see from the general format, t b l sees a table in terms o f three dis-
tinct parts:
206 0 UNlX Text Processing 0

1. The overall layout of the table described in the global options line. For
example, this line describes whether the table is to be centered on the page
or made as wide as the rest of the document. The global options line is
optional.
2. The layout of each column in the table described in the format section. For
example, in this section, you specify whether a column is to be left or right
justified. The format section is required and may contain one or more for-
mat lines.
3. The actual text or numbers, datu, to be entered in the table.

A Simple Table Example

Let?s start with a simple table like the following to show the different parts of the tbl
description:
1 User console
2 Monochromatic graphics terminal
3 Color graphics terminal
4 Line printer
5 Digitizer
6 Laser printer
7 Unallocated

You can lay out this table using the following tbl requests:
.TS Table Start macro
tab ( e ) ; Options line
c 1. Format line
l@User console
2@Monochromatic graphics terminal
3@Color graphics terminal
4@Line printer
5@Digitizer Table entries
6@Laser printer
7@Unallocated
-TE Table End macro

Now let?s see what these lines mean:

1. The .TS at the beginning says that a table follows.

2. The options line applies to the layout of the table as a whole. The option
tab ( e ) means that you will be using the @ character as a tab character
when you input data to the table. Normally, tbl expects the columns in
 0 Formatting with tbl 0 207

the table to be separated by actual tabs. But it is much easier to figure out
whether you have the right number of columns if you use a visible character
that is not part of the data. This i s useful in debugging a table error when
the formatted data doesn’t appear in the proper columns. The options line
always ends with a semicolon (;).
3. The format section applies to the lines of data in the table. Each format line
contains a key letter for each column of the table. The layout of the key
letters resembles the layout of actual data in the table.
Each format line corresponds to a single line in the table. However, you
can have fewer format lines than lines in the table. In this case, the last line
of the description applies to all remaining lines of data. In our example, we
have only one format line, so all lines in the table will follow this format.
For example:
c 1.

means that there are two columns in each line. The first column will be
centered (c),and the second left justified (1). The format section ends with
a period at the end of the last format line.
4. The data itself. Each line of data corresponds to one line in the table. If
you have very long input lines, they can be broken into smaller line seg-
ments. A backslash o) at the end of a line segment means that it continues
to the next line and i s part of a longer input line. Each of the columns in
our table is separated by an @ sign, which we are using in place of a tab
character, as we have specified in the options line.
5. A TE signals the end of the table description.

Laying Out a Table

The global options line is an optional line that controls the overall appearance of the
table. Normally, a table is positioned on the left-hand side of the page. Because the
table is probably part of a larger document, you may want to center the table and
enclose it in a box to make it stand out. Let’s modify the options line in our example to
produce this new layout:

.TS
center, box, tab ( @ ); New options line
c 1.
l@User console
2@Monochromatic graphics terminal
3@Color graphics terminal
etc.
208 0 UNlX Text Processing 0

When formatted, the table looks like this:

1 User console
2 Monochromatic graphics terminal
3 Color graphics terminal
4 Line printer
5 Digitizer
6 Laser printer
7 Unallocated
8 Pen plotter
9 Raster plotter
IO. 1 1 . 12 Unall ocated

Now you know how to use three of t k option names: center, box, and
t a b ( ) . If you use one or more option names, they must be separated by spaces, tabs,
or commas. The options line, if present, must immediately follow the .TS line. There
are other options that you can use:

expand Make the table as wide as the current line length

allbox Enclose each item in the table in a box
doublebox Box the whole table with a double line
linesize ( n ) Set lines (for box, allbox, and doublebox) in n-
point type
delim ( x y ) Set x and y as eqn delimiters. See Chapter 9 for infor-
mation on the equation preprocessor eqn.

The difference between a table that i s centered or left justified and one that is
expanded is the amount of space between columns. If you specify center or the
default, the width between columns will be three ens. If you specify expand, t b l
will expand the width of the overall columns until the table i s as wide as the current
margins.
If the overall width of the table calculated by t b l i s greater than the width of the
text, n r o f f /trof f will ignore any positioning option you specify. The table will be
printed as is necessary to fit everything, even if the table runs to the edge of the paper.
The linesize option changes the width of the lines used in enclosing tables to
a given point size. Normally, the lines are 10 point. You can specify an absolute line
size, such as linesize ( 2 4 ) , to print thicker box lines, or a relative size, such as
linesize ( + 1 4 ) , to produce the same effect.
Let’s try one more example by enclosing all the data entries in boxes. The
options line for the table now reads:
center, allbox, t a b ( @ );
 0 Formatting with tbl 0 209

The new table would look like this:

1 User console
2 Monochromatic graphics terminal
3 Color graphics terminal
4 Line Drinter
5 Digitizer
6 Laser printer
7 Unallocated
8 Pen plotter
9 Raster plotter
10, 11, 12 Unallocated

The t b l program isn’t very good at keeping boxed tables on one page. If you
have a long table, t b l may break it up at an awkward point (for example, placing the
last line of a table on another page). To keep a boxed table together on one page,
enclose it in a .D S / . D E macro pair (in either m s or mm). Alternatively, you can
give t b l the latitude to split a table and print each section with its own table heading
using the .TS H macro, as you will see later.

Describing Column Formats

Each column in the table is described by a key letter in the format section. Key letters
are separated from each other by spaces or tabs for readability. The basic set of key
letters includes:

Lor 1 Left justify the data within a column.

Ror r Right justify the data within a column.
Cor c Center the data within a column.
Sor s Extend data in the previous column to this column (horizontal
span).
Nor n Align numbers by their decimal points. If there are no decimal
points, align them by the units digit.
Aor a Indent characters in the column from the standard left align-
ment by one em.
h
Extend entry from previous row down through this row (verti-
cal span). Text will be centered between the specified rows.
Tor t Also vertical span, but text will appear at the top of the column
instead of midway within the specified area.
21 0 0 UNlX Text Processing 0

If all columns of the table follow the same format, you need only one format line
for the entire table. However, not all tables contain the same number of columns
throughout. For example, you might have a table where the upper half consists of three
columns, and the lower half contains only two.
The rule in writing format lines is to specify key letters for the largest number of
columns in the table and carry that number for all format lines. That way, if you
specify three columns, and you’re using only two, you can use two consecutive tab
characters (with nothing in between) to denote an empty field for the unused column.
The longest format fine defines the number of columns in the table.
Suppose you defined four columns in the first format line, and then defined only
three columns in the succeeding lines. The t b l program will still format your table,
but it assumes that the undefined column is left justified.
In the following sections, we will show some typical applications of these and
other key letters to format table headings and columns of data.

Tables with Headers

You can think of a table header as an extra row of data that may or may not have the
same format as the actual data. If the format of the header is different, you must add
another line at the beginning of your format section to describe the header.
For example, we’ll change the first column in the previous table to have the
header Perf and the second to have the header Device, so that we get the following
table.

I 1 User console
Device

2 Monochromatic graphics terminal

3 Color graphics terminal
4 Line printer
5 Digitizer
6 Laser printer
7 Unallocated
8 Pen plotter
9 Raster plotter
10, 1 1 , 12 Unallocated

The relevant lines that produced this table follow:

.T S
c e n t e r , box, t a b ( @ );
c c
c 1.
Port@Device
- SP
l @ U s e r console
 0 Formatting with t b l 0 21 1

2@Monochromatic graphics terminal

etc.

The first line of the format description (c c) says that there are two columns of data,
each one centered within each column. (Note that there is no period at the end of this
line.) Because this is the first line of the format description, it applies to the first line
of our data, which happens to be the table heading. This means that the words Port and
Device will be centered in each column. The second (and last) format line i s the same
as in the previous example and applies to the rest of the table. Note the period at the
end of this line.
We used - s p to produce a blank line after the table header. The t b l com-
mand assumes that any non-numeric string preceded by a dot is a t r o f f or n r o f i
request and passes it unchanged to the formatter. Thus, you can vary spacing between
rows, or use other n r o f f /t r o f f commands within a table.

Tables with Spanned Headers

Our previous table now contains a header for each column. We now want to have an
overall title or header that spans the width of the table. As before, you can think of the
spanned header as an extra data line with its own format description.
We want the header to be only one column, centered across the whole table like
the following.
Output Device Configuration
Port Device
1 User console
2 Monochromatic graphics terminal
3 Color graphics terminal
4 Line printer
5 Digitizer
6 Laser printer
7 Una1loc ated
8 Pen plotter
9 Raster plotter
10, 1 1 , 12 Unallocated

Because we should keep the number of columns the same throughout the table,
we use the spun format option ( s ) to tell t b l that the entry in a preceding column
continues on to the other columns. The relevant portion of our table description con-
tains the following lines:
.TS
center, box, t a b (e);
c s
c c
c 1.
212 UNlX Text Processing 0

Output Device Configuration

.sp .5v
Port@Device
.sp .5v
l@User console
etc.
W e now have three format lines: the first describes the main header, the second
describes each column header, and the third applies to the rest of the data in the table.

Numeric and Alphabetic Columns

You can align numeric data by the decimal point or the units digit using the key letter
n in the format line. When you use n,numbers in a column will be aligned as fol-
lows:

23.6
155
98.08.6
5.26
12798
0.2365
980.

You should never enter non-numeric data in a column that is designated as n.

On the other hand, you can enter numbers in columns that are aligned using any of the
other key letters. The numbers will just be treated as if they were ordinary alphabetic
characters. Thus, a column of numbers might also be centered, left justified, or right
justified.
You should also avoid putting equations in numeric columns because t b l
attempts to split numeric format items into two parts. To prevent this from happening,
use the delim (q) global option. For example, if the e q n delimiters are $$, a
d e 1 im ( $ $ ) option causes a numeric column such as:
79.909 $+- .157$

to be divided after 79.909 and not after -157.

Columns designated as a are always slightly indented relative to left-justified
columns. If necessary, t b l increases the column width to force this. Data in an a
format is positioned so that the widest entry is centered within the column.
A note about n and a: when you have several command lines, do not use both
n and a to format different rows in the same column. For example, the format lines:
r n r
r a r

are not allowed. This i s because n and a share the same number register location in
n r o f f / t ro f f ’ s memory.
 a Formatting with tbl 0 213

The special nonprinting character string \ & may be used to override the normal
alignment of numeric or alphabetic data. For example, if you use \ & before a digit,
then the digit will line up with the decimal point and \ & will not appear in the output.
The effect of \ & is as follows.

Input Form output

9.65 9.65
12.4.8 12.4.8
15.\&1.32 15.7 - 3 2
2\&0.9.19 20.9.19
processor p r o c e s s o r
half half
half\& half

Vertically Spanned Columns

Let's see how the vertical span key (") is used in a table like the following.

kcaU
Fuel Substance
gram mol. wt.

Hydrogen 68.4
Methane 211
Gases
Butane 680
Ethane 368

Benzene 782
Liquids Ethyl alcohol 328
Methyl alcohol 171

The t b 1 description for this table is:

.TS
t a b ( @ );
c c c
A -
C

1 1 n.
Fuel@Substance@kcal/
@@gram mol. wt.
- SP
Gases@[email protected]
\^@Methanee211
\"@Butane@680
\"@Ethane@368
f SP
21 4 0 UNlX Text Processing 0

Liquids@Benzene@82
\^@Ethyl alcohol@328
\^@Methyl alcohol@171
.TE

There are three lines in the format section: the first two describe the column headings,
and the last describes the format of the data.
We can imagine the first line of the header as consisting of the words Fuel Sub-
stance kcall and the second line as Fuel Substance gram mol. wt. The words Fuel Sub-
stance don't actually appear twice, but are centered relative to the two lines that form
the third column header. We use the caret key (A) in the second format line to tell t b l
that these two column names vertically span their respective columns. Note the first
two data lines that correspond to the first two format lines.
. We could have also used the same approach to describe the rest of the data, but
this would mean writing seven more format lines, one for each of the lines of data. The
table really has three columns with the same format throughout, so you can use just one
format line to describe all of them. Then you can enter the characters \ " in place of a
column entry to tell t b l that the entry in the previous row for that column vertically
spans this row also.
You can use the " key letter in the format section and at the same time enter \ "
in the data section as we did previously. You don't lose anything by doing this and
t b 1 doesn't complain.
Another way of describing a vertically spanned column is by using the key letter
t (or T) in the format line. Any corresponding vertically spanned item will begin at
the top of its range. Thus, if we specify t instead of in the format line, the words
Fuel and Substance will be in line with kcall.

Drawing Lines in Tables

Horizontal rules are specified by underscores and by equal signs entered between the
appropriate lines of data. An underscore on a line by itself entered between two rows
of data produces a single rule running the whole width of the table. An equal sign on a
line by itself produces a double rule.
If you want a horizontal rule to be only as wide as the contents of the column,
enter an underscore or equal sign in that column as part of the data. The underscore or
equal sign must be separated from the other columns by tabs or the tab character we've
specified in the options line. To print these characters explicitly, they should be pre-
ceded by a \ & or followed by a space before the usual tab or newline character.
You can also use these two characters in place of a key letter in the format line.
If an adjacent column contains a horizontal or vertical line, the horizontal line is
extended to meet nearby lines. If you enter any data in this column, the data will be
ignored and you will get a warning message. The following table has a fairly compli-
cated heading:
 0 Formatting with t b l 0 215

I 1984 (Jan.-July) I
198411983
Items Units

I ‘R”, I
(%I
3,889,543 145.7
Color 2,766,004 110.7
1,123,539 12.5

The t b l description for this table looks like this:

.TS
center,box,tab(@) ;
c s s
c c -
A A
I C
* A
I C
1 r n.
1984 (Jan.-July)
ItemseUnits
@@1984/1983
@ @ (%I
-
TV@3,889,[email protected]
Color@2,766,004@110 -7
B/W@1,123,[email protected]
.TE

As you can see from the preceding description, vertical lines are drawn by specifying
bars within the format lines. A single vertical bar between two key letters draws a sin-
gle vertical line between those two columns in the table. You can enter the bar after the
first key letter or before the second key letter. A vertical bar to the left of the first key
letter or to the right of the last one produces a vertical line at the edge of the table.
Two vertical bars (I I) draw a double rule.
These characters are really more useful for drawing lines inside the table rather
than for manually enclosing a table in a box because there are global options that
automatically do this. To draw vertical and horizontal lines in our table “Fuels,” we
modify the relevant format and data lines as follows:

c I IC IC

A I IA IC

1 I
In. I1
Fuel@Substance@kcal/
@@gram mol. wt.
-

Gases@[email protected]
erc.
216 0 UNlX Text Processing 0

-
Liquids@Benzene@782
etc.

kcaU
Fuel Substance
gram mol. wt.
Hydrogen 68.4
Methane 21 1
Gases
Butane 680
Ethane 368
Benzene 782
Liquids Ethyl alcohol 328
Methyl alcohol 171

Changing Fonts and Sizes

The t b l program assumes that the table is always set in roman type. However, you
can always change the typeface of all entries in a column to italic or boldface. You can
add one of the following letters after the column key letter:
fb fB b B Boldface
fi fI i I Italic
f cw fcw cw CW Constant width

If you want to change the font of only some of the entries, you should use explicit
nrof f /t ro f f requests rather than specifying the font in the format line. For exam-
ple, let's change the headers in the previous table to boldface and the words Gases and
Liquids to italic. The format lines would look like this:

c I IcB IcB
" I I" IcB
1 I I1 In.

Gases will be written as \fIGases\fR and Liquids as \fILiquids\fR. The

effect would be as follows:
 0 Formatting with tbl 0 21 7

kcal/
Substance
ram mol. wt.
Hydrogen
Methane
Butane

782
328

The type size in which headings and data are printed is normally 10 points. You
can also change the size of the type by using the key letter p and an absolute or rela-
tive point size. To specify a change in size relative to the existing point size, use a +
or - before the value. For example, a column specification of c p 1 2 or c p + 2 will
both result in a centered column using 12-point type.

Changing the Column Width

When you’re not using the expand option, the normal spacing between any two
columns is three ens. You can change the spacing by specifying a numeric value
between the key letters representing those columns. The number specifies the separa-
tion in ens. When you’re using the e x p a n d option and you specify a column space,
the number is multiplied by a constant such that the table is as wide as the current line
length.
If you don’t want any spaces between the columns, simply write 0, as in:
rO 1
which yields:

Hydrogen68.4
Methane2 I 1
Butane680

These spacings are only nominal spacings. The data may be so irregular in length
that no two columns will actually appear to be separated by the specified distance.
However, varying the amount of separation between two columns still leaves t b l free
to make each column as wide or as narrow as is necessary.
You can specify a minimum width for any column by entering the letter w (or
W) after the key letter, followed by the desired width in parentheses. You can use any
unit of measurement recognized by n r o f f /trof f when specifying a width dimen-
sion. You can also enter a value without a unit of measurement, in which case t b l
assumes the value is in ens. Thus. the format:
218 0 UNlX Text Processing 0

rw (15)

specifies a column that is 15 ens wide with the text right justified within the column,
and:
-
lw (2 2 5 i )

specifies a left-justified column that is 2.25 inches wide.

You can also force t b l to make the width of particular columns equal by using
the letter e (or E) after the key letter for those columns. This allows a group of regu-
larly spaced columns.
To show that t b l can be used for any text that needs to be laid out in columns
(as opposed to tables), we can print the following text:

Signature
August 3 1, J. White K. Kimura
1987

using this t b l description:

.TS
expand, tab(@);
c c c
cew(l.3i) ce ce.
Signature@@
\-@ \-@ \-
August 31,@J. White@K. Kimura
1987@@
- TE

In the last format line, we specified that all three columns be 1.3 inches wide. Because
all columns will be of equal width, we need to specify the width only once.

Other Key Letters

We already showed you some of the more widely used key letters. Additional features
that can be used with the basic set of key letters are:

Vor v Used with a number to indicate the vertical line spacing used
within a table entry. Used only with text blocks (discussed in a
later section).
uor u Move the corresponding entry up by one-half line to produce
staggered columns. This doesn’t work with the allbox glo-
bal option.
 0 Formatting with t b l 0 219

zor z Ignore the data entry in calculating column width. This is use-
ful in allowing headings to run across adjacent columns where
spanned headings might be inappropriate.

Key letters for a column can be written in any order. They do not need to be
separated, except when you specify both a point size (p) and a column separation
number. Thus, a numeric column entered in bold 18-point type with a minimum
column width of 1.5 inches and separated from the next column by 12 ens can be wnt-
ten as:
npl8w (1.Si)B 1 2

Two or more format lines can also be written on one line by separating them with com-
mas. For example, the format lines:
c c c
1 1 n.
can be written as:
c c c , 1 1 n.

Changing the Format within a Table

All our examples so far have shown tables that consist of somewhat complicated head-
ings followed by identical rows of data. Thus, we can keep the number of format lines
comparatively small. This may not be the case when a table is divided into sections,
each of which has its own heading. Let’s look at the following table (from AT&T’s
Documenter’s Workbench Text Formatter’s Reference):

Horizontal Local Motions

. Effect in
Function
troff nroff
b”’ Move distance N
\(space) Unpaddable space-size space
\O Digit-size space
\I 1/6 em space ignored
\A 1/12 em space ignored

It has both a main header and column headers. The body of the table is divided
into two parts. The upper part contains two columns, and the lower part contains three.
To format each part correctly, we must enter a command line for each row of data so
that tbl can keep track of which rows of the table have which format. This process is
tedious and prone to error. Fortunately, t b l has a way around this.
To change the format of columns within a table, t b l has the table continue
request .T & . We can change the format of a table at any time by entering .T& fol-
220 0 UNlX Text Processing 0

lowed by the new format line(s) and the additional data. The general format for the
t b l description is as follows:

- TS
option line;
format section.
data
.T&
new format section.
data
.T&
another new format section.
data
.TE
There are two things we cannot change after a .T & request: L e global options line
and the number of columns specified. Our original options line holds for the entire
table.
Let's see how we can use the .T& request to produce the previous table:
- TS
center, box, linesize (6),t a b ( @ );
cB s s.
Horizontal Local Motions
-
- T&
CI I CI s
CI I CI s
CI I CI I CI
c I 1 s.
Function@Effect in
\eA@-
\e"@troff@nroff
-
\eh'N'@Move distance N
\e(space)@Unpaddable space-size space
\eO@Digit-size space
-
- T&
c I l I 1 .
\el@1/6 em space@ignored
\e*@1/12 em space@ignored
.TE
We take the largest number of columns in the table, which is three. W e have two - T &
requests to break up the table into three parts with their own format sections. The first
part applies to the main header only. The second describes the column headers and the
 0 Formatting with tbl 0 221

three-column segment of the table. Finally, the lower part applies to the last part of the
table.
Although you can have hundreds of lines in a table, t b l uses only the first 200
lines to set up the table. Any format changes you make after the 200th column will not
be processed by tbl. In this case, you should break up the table into smaller table
segments.
Should you specify . T S H but forget to follow it with .TH, some strange
things will happen. One recent instance of this caused the table to be output in a nearly
endless succession of pages. (In t rof f terms, a diversion created to capture the table
heading filled up with the table instead; this caused the first page break that triggered
the output of the diversion at the top of the next page; each time the diversion was out-
put, it caused a new page break and the diversion was output again.)

Putting Text Blocks in a Column

Some tables consist of column entries that cannot be conveniently typed as a simple
string between tabs. Descriptive tables, for example, require ordinary flowing text justi-
fied between the margins of the specific column in which it appears in the table. These
sections of flowing text are called text blocks.
Each block of text is preceded by a T { and followed by a T } . The T { marker
must be at the end of a line, and the T } must be at the start of a line:
...T{
Block of
text
T} ...
When a text block is included in a row that contains other columns of data or text, the
T{ that marks the beginning of the text block must appear at the end of the line in the
text. Even a single blank space following the T{ will cause the table to fail. Like-
wise, the T} symbol must always begin the line:
... Data@T {
Block of
text
T } @data...

This makes it easy for you to revise text when necessary and also allows you to insert
any special nrof f /trof f commands before or after the text block.
Let?s lay out the following table:
222 0 UNlX Text Processing 0

Sor :Pattern-Matching Characters in v i

Special Characters Usage
Matches any single character except newline.

* Matches any number (including zero) of the single char-

acter (including a character specified by a regular expres-
sion) that immediately precedes it.

[...I Matches any one of the characters enclosed between the

brackets. A range of consecutive characters can be
specified by separating the first and last characters in the
range with a hyphen.

Requires that the preceding regular expression be found

at the end of the line.

Matches a range of occurrences of the single character

(including a character specified by a regular expression)
that immediately precedes it. n and m are integers
between 0 and 256 that specify how many occurrences to
match.

The t b l description of this table is:

- TS
box,tab(@);
cb s
CII CI
cw(1.25i) I lw(3.25i) -
Some Pattern-Matching Characters in \fIvi\fR
-
Special Characters@Usage
-
\fI.\fR@Matches any single character\
except \fInewline\fR.

*@T {
Matches any number (including zero) of t h e
single character (including
a character specified by a regular expression)
that immediately precedes it.
TI
 0 Formatting with t b l 0 223

[...]@TI
Matches any \fIone\fR of the characters enclosed
between the brackets.
A range of consecutive characters can be
specified by separating the
first and last characters in the range with a hyphen.
T}

$@T 1
Requires that the preceding regular
expression be found at the end of the line.
TI

\ { \ f In,m\fR\ }@TI
Matches a range of occurrences of the
single character (including a
character specified by a regular expression)
that immediately precedes
it. n and m are integers between
0 and 256 that specify how many occurrences to match.
TI
.TE

What might confuse you about this source text is that each block of text occupies two
or more lines. Just think of everything that comes between a T { and a T } as a single
entry that occupies a single column in that row. I t is separated from its neighbors by
tabs. If you keep track of the tabs, you will be able to sort out quite easily the sequence
of columns.
In the previous description, we specified a minimum width for each column. If a
width is not given, t b l uses the default:
L * c/(N+l)
where L is the current line length, C is the number of table columns spanned by the
text, and N is the total number of columns in the table. I t is sometimes better to define
a column width because tbl might make the table too narrow by default.
You can also use the nrof f/ t r o f f commands .na and .ad to left justify
text blocks if the output doesn’t come out fully justified. The t b l description would
be:
...T{
- na
Block of
text
.ad
TI
224 0 UNIX Text Processing 0

The nrof f and t rof f forrnatters can accept only about twenty or thirty small text
blocks in a table without exceeding certain internal limits. If the limits are exceeded,
you will get error messages like “too many strindmacro names” or “too many number
registers.”
In this case, you should divide the table into two or more independent tables, each
with its own . T S and . T E requests. The final formatted sections can be “joined”
and made to appear as one table by inserting minus . s p requests (such as
.s p -12p) between the sections. This will cause the formatter to draw them
together.
You can also change the vertical line spacing within a text block using a key
letter followed by v (or V) and a number. The number may be a signed digit and is
taken as an increase or decrease from the current vertical spacing.

1 Breaking Up Long Tables

If you have a very long table that will fill many pages, it might be helpful to break up
the table into several smaller ones, with the main heading reproduced at the top of each
page. Then the reader doesn’t have to keep returning to the first page to see what the
columns indicate. The t b l program also automatically breaks a boxed table if it runs
over one page.
You can use the .T S H and TH macros to reproduce the original heading at
the top of each page of the table:

.TS H
options;
format section.
main header
.TH
data
.TE

The .TH (table header) macro is a feature of the m s macro package (not tbl). This
macro can take the letter N as an argument; this causes the table header to be printed
only if it i s the first table header on a page. This is useful when you have to build a
long table from smaller .T S H/ .TE segments. For example:
 n Formatting with tbl 0 225

.TS H
global options;
format section.
main header
.TH
data
.TE
.TS H
global options;
format section.
main header
.TH N
data
- TE
This causes the table header to appear at the top of the first table segment. The header
will not appear on top of the second segment when both segments appear on the same
page. If the table continues to another page, the heading will still appear at the top of
the new page. This feature is useful when breaking a long complex tabie into segments.

Putting Titles on Tables

The mm macro .T B can be used to automatically number and title a table. All tables
with .T B are numbered consecutively. The title is centered above the table if it can fit
on one line. If the title is longer than one line, all succeeding lines of the title are
indented to line up with the first character of the title. The . T B macro is normally
used inside a .DS / .DE pair.
The . T B macro is not part of tbl. Thus, it can be used to generate titles or
headers for tables that are created using only tabs and none of the t b l commands.
The general format of the .T B macro is:
- TB [title] [n] wug]
where n is used to ovemde the normal numbering. Theflag option can take one of the
following values:

0 n is used as a prefix to the normal table number

1 n is used as a suffix to the normal table number
2 n replaces the normal table number

If you put the .T B macro before the .T S macro, the title is placed above the table.
You can also put the title below the table by using the TB macro after .TE.
For example, we can modify one of our tables by adding a title and labeling it as
Table 5. We add the following lines before the .T S :
226 UNlX Text Processing 0

.D S
.TB "Horizontal Local Motions" "5" n21v
- SP
And we add a .DE after the .TE. The table now looks like this.

Table 5. Horizontal Local Motions

- Effect in
Function
troff nroff
h'N' Move distance N
\(space) Unpaddable space-size space
\O Digit-size space
\I 1/6em space ignored
\A 1/12 ern space ignored

Another useful mm macro is the TC macro. The .TC macro is placed at the
end of the file. When the file is formatted, .TC collects the titles of tables that were
generated using .T B for the table of contents. Thus, if we had used .T B to put
headers in our examples, the table of contents might look like this:

LIST OF TABLES

TABLE 1. Production of Audio Equipment . . . . . . . . 2

TABLE 2. Output Device Configuration . . . . . . . . . .14

TABLE 3. Heating Value of Fuels . . . . . . . . . . . . . . 17

A tbl Checklist
Most table formatting errors come from specifying too few columns in the format sec-
tion, forgetting a tab character between column entries in a table, or omitting one or
more of the characters that t b l expects in a table description. After you've finished
laying out a table, check that you have the following:

a . T S with a . T E
. a .THwitha .TS H
a semicolon at the end of the options line (if there is one)
 0 Formatting with t b l 0 227

= a period at the end of the last format line (including format sections with a
.T&)
in the format section, an item for each column and a format line for each line
of the table
a tab symbol for each column in each line of the table, except for the first
column when horizontally spanning, and within text blocks
. for text blocks, a T { with every T}
. no extra blanks after:
any .TS, .TE, .TS H, .TH, or .T&
the end of the options and format lines
any TI or TI
. no periods at the beginning of any “data” text lines (add a
period, if necessary)
\& before the

a space after each table entry of - and = unless you want the lines to extend
across the column

Some Complex Tables

Surely, the best way to learn more about t b l is to study tables of greater complexity
than the ones we’ve look at so far. The tbl article by M.E. Lesk in the UNIX
Programmer’s Manual provides many fine examples of difficult tables. Look at the
formatted tables and try to “break” the code that produced them. In this section, you’ll
find two complicated tables followed by the tbl input for you to decipher.
The weight table shown in Figure 8-1 i s taken from a manual that describes the
safe operation of mobile cranes. This table was coded by an associate, Daniel Gilly,
over several hours. The code is listed in Figure 8-2. Look at how the vertical line indi-
cator ( I ) is used between entries to draw a line at the end of each column. Note also
the use of the alphabetic (a) format specification to produce indented text.
The financial table shown in Figure 8-3 is adapted from a prospectus prepared by
t rof f users at a large New York law firm. The code for this table is listed in Figure
8-4. Note the use of a leader character Oa) in the first entry, coupled with a fixed width
specification for the first column, to produce leaders that fill out the column. Also,
notice how the table headings are printed in a smaller point size than the rest of the
table, using the format specification (p8).
228 0 UNlX Text Processing

WEIGH7 OF MATE: ALS (Based On Volume)

Approx. Approx.
Weight, Weight,
Material Material
Lbs. Per Lbs. Per
Cubic Foot Cubic Foot
METALS TIMBER, AIR-DRY
Aluminum 165 Cedar 22
Brass 535 Fir, Douglas, seasoned 34
Bronze 500 Fir, Douglas, unseasoned 40
Copper 560 Fir, Douglas, wet 50
Iron 480 Fir, Douglas, glue
Lead 710 laminated 34
Steel 490 Hemlock 30
Tin 460 Pine 30
MASONRY Poplar 30
Ashlar masonry 140-160 Spruce 28
Brick masonry, soft 110 LIQUIDS
Brick masonry, com- Alcohol, pure 49
mon (about 3 tons Gasoline 42
per thousand) 125 Oil 58
Brick masonry, pressed 140 Water 62
Clay tile masonry, EARTH
average 60 Earth, wet 100
Rubble masonry 130-155 Earth, dry (about 2050
Concrete, cinder, lbs. per cu. yd.) 75
haydite 100-1 10 Sand and gravel, wet 120
Concrete, slag 130 Sand and gravel, dry 105
Concrete, stone 144 River sand (about 3240
Concrete, stone, Ibs. per cu. yd.) 120
reinforced (4050 lbs. VARIOUS BUILDING
per cu. yd.) 150 MATERIALS
[CE AND SNOW Cement, Portland, loose 94
Ice 56 Cement, Portland, set 183
Snow, dry, fresh fallen 8 Lime, gypsum, loose 53-64
Snow, dry, packed 12-25 Mortar, cement-lime,
Snow, wet 27-40 set 103
MISCELLANEOUS Crushed rock (about
Asphalt 80 2565 lbs. per
TZU 75 cu. yd.) 90-110
Glass 160
Paper 60

Fig. 8-1. A Complex Table

 0 Formatting with t b l 0 229

.ps 8
.vs 10
.TS
center,box,tab ( @ );
c b s s s
c IcI c I c
AlclAlc
AICIAIC
AlclAlc-
WEIGHTS O F MATERIALS (Based On Volume)

Material@Approx.@Material@Approx.
@Weight,@@Weight,
@Lbs. Per@@Lbs. Per
@Cubic Foot@@Cubic Foot
-
.sp - 5
- T&
lblcllblc.
METALS@@TIMBER, AIR-DRY@
.T&
alclalc.
Aluminum@165@Cedar@\O22
Brass@535@Fir, Douglas, seasoned@\034
Bronze@500@Fir, Douglas, unseasoned@\040
Copper@560@Fir, Douglas, wet@\050
Iron@480@Fir, Douglas, glue@
Lead@710@\0\01aminated@\O34
Steel@490@Hemlock@\O30
Tin@460@Pine@\030
.T&
lblclalc.
MASONRY@@Poplar@\030
- T&
alclalc.
Ashlar masonry@140-160@Spruce@\O28
.T&
a I c I lb I c.
Brick masonry, soft@llO@LIQUIDS@
- T&
alclalc.
Brick masonry, corn-@@Alcohol, pure@\049
\O\Omon (about 3 tons@@Gasoline@\042
\O\Oper thousand)@125@Oil@\058
Brick masonry, pressed@140@Water@\062
- T&
a I c I lb I a.
Clay tile masonry,@@EARTH@
.T&
alclalc.
\O\Oaverage@\06O@Earth, wet@100
Rubble masonry@130-155@Earth, dry (about 2050@

Fig. 8-2. Input for Figure 8-1

230 UNIX Text Processing 0

Concrete, cinder,@@\O\Olbs. per cu. yd.)@\075

\O\Ohaydite@100-110@Sand and gravel, wet@120
Concrete, slag@130@Sand and gravel, dry@105
Concrete, stone@144@River sand (about 3240@
Concrete, stone,@@\O\Olbs. per cu. yd.)@120
.T&
a I c I lbl c.
\O\Oreinforced (4050 lbs.@@VARIOUS BUILDING@
\O\Oper cu. yd.)@15O@\O\OMATERIALS@
.T&
lb I c la1 c.
ICE AND SNOW@@Cement, Portland, loose@\094
.T&
alclalc.
Ice@\056@Cement, Portland, set@183
Snow, dry, fresh fallen@\0\08@Limer gypsum, loose@53-64
Snow, dry, packed@12-25@Mortarr cement-lime,@
Snow, wet@27-40@\0\0set@103
.T&
lb I c I a I c.
MISCELLANEOUS@@Crushed rock (about@
.T&
alclalc.
Asphalt@\080@\0\02565 lbs. per@
Tar@\075@\0\0cu. yd.)@90-110
Glass@160@@
Paper@\060@@
-sp .5
- TE
Fig. 8-2. -(Cont’d)

Year Ending December 31

1986 1985 1984 1983
(Dollars in millions)
Premiums .............................................. $ 10,922.7 $ 10,330.7 $ 9,252.4 $ 9,071.8
Investment income ............................... 3,671.7 3,146.0 2,749.7 2,3 08.9
Federal income taxes ............................ 24.4 91.6 71.9 20.8
Operating income ................................. 359.8 346.1 342.6 309.6
Realized gains (losses) ......................... 15.4 27.0 (30.2) (15.2)
Net income ........................................... 375.2 373.1 312.4 295.8
Cash provided by operations ............... 4,123.2 3 3 60.8 3,514.9 3,067.4
Assets ................................................... 41,645.8 34,434.7 32,876.6 27,987.6

Fig. 8-3. Financial Table

 0 Formatting with tbl 0 231

.TS
expand, tab ( @ );
lw(13P) cbp8 s s s
lw(13P) c s s s
lw(13P) cbp8 cbp8 cbp8 cbp8
lw(13P) cbp8 s s s
lw(13P) n n n n.
@Year Ending December 31
.sp .2v
@-
@1986@1985@1984@1983
@ (Dollars in millions)
- s p .5v
Premiums\a@$\010,922.7@$\010,330~7@$\0\09,252.4@$\0\09,071~8
Investment income\a@3,671.7@3,146.0@2,749.7@2,308.9
Federal income taxes\[email protected]@[email protected]@20.8
Operating income\[email protected]@[email protected]@309-6
Realized gains (losses)\[email protected]@27.0@(30.2)@(15.2)
Net income\[email protected]@[email protected]@295.8
Cash provided by operations\a@4,123.2@3,560.8@3,514-9@3,067-4
Assets\a@41,645.8@34,434.7@32,876-6@27,987.6
.TE
Fig. 8-4. Input for Figure 8-3
 C H A P T E R
rn

Typesetting Equations with eqn

Typesetting mathematical equations has always been a problem for users who have a
limited knowledge of mathematics or typesetting. This is because mathematical expres-
sions are often a mixture of standard text and special characters in different point sizes.
For example, the equation:
m m

requires three special characters (E,00, and +) and roman and italic characters in two
different sizes. Expressions also may require horizontal and vertical printing motions
(as in subscripts and superscripts).
You could code this example using t rof f requests, but the syntax for describ-
ing the printing motions, sizes, and fonts are difficult to learn and difficult to type in
correctly. UNIX has formatting tools specifically designed for documents containing
mathematical symbols-the programs e q n and neqn. The e q n program is a
preprocessor for trof f; n e q n is a preprocessor for nrof f .
With e q n you can typeset both inline equations and equations that are set off
from the body of the text like the example shown. It takes an English-like description
of a mathematical equation and generates a t rof f script. You don’t need to under-
stand what you are typing.
The e q n preprocessor was designed to be easy to learn and even easier to use.
This implies that normal mathematical conventions such as operator precedence and
parentheses cannot be used. Nor does e q n assume that parentheses are always bal-
anced, or that an expression is better written in another form. There are only a few
rules, keywords, special symbols, and operators to remember. If something works in
one situation, it should work everywhere.
This section shows you how to typeset mathematical equations using a set of spe-
cial words that belong to the e q n vocabulary. With e q n , you can format the follow-
ing quite easily:

232
 0 Typesetting Equations with eqn 233

the Greek alphabet

special symbols, such as summations (C), products (n)integrals d),and square
roots ( C )
positional notation, such as subscripts and superscripts, fractions, matrices, and
vertical piles
diacritical marks
sizes and fonts
horizontal and vertical spacing

You can even define a string that appears repeatedly throughout the document so that
you do not need to type it in each time it appears.

A Simple eqn Example

To best illustrate how e q n works and how easy it is to learn the syntax, let’s take a
simple example:
a2
-
h
If you were to read this mathematical expression aloud to another person, you might say
“a sub 2 over b.” This is exactly how you would describe the expression to e q n .
The word s u b denotes a subscript; the word over denotes a fraction. You will see
the other words that e q n treats as special (i.e., that belong to the e q n vocabulary) as
we move along in this section.
When you use e q n , it assumes that you have a two-dimensional picture of how
the equation should appear in the document. The key in writing the eqn decription is
to familiarize yourself with the special words used by e q n in printing mathematical
characters. Then, describe the equation as if you were reading it aloud to another per-
son.
The e q n preprocessor takes care of the standard things that you would expect to
happen automatically, such as printing superscripts and subscripts in an appropriately
smaller size, and adjusting the length and size of fraction bars. Following mathematical
convention, variables are made italic, parentheses, operators, and digits are made roman,
and normal spacing is automatically adjusted to make the expression look better.

Usingeqn
The e q n preprocessor is used not only for typesetting equations, but also for typeset-
ting nontechnical documents. For example, many documents contain subscripted or
superscripted words. Using e q n can be easier than formatting the subscript or super-
script using t r o f f commands.
234 0 UNlX Text Processing 0

To format a document with eqn, you would enter:

$ eqn /usr/pub/eqnchar files I t r o f f [options]
You can then pipe the output to the desired printer. The file /usr/pub/eqnchar
contains definitions of additional special characters that can be used by eqn. It is not
essential that you use it, but you may get better results with certain equations if you do.
If you use eqn with the tbl preprocessor to print tables containing mathemati-
cal expressions, invoke t b l before eqn to mimimize the data passed through the
pipe:
$ tbl /usr/pub/eqncharfile I eqn I t r o f f
If you are using nrof f instead of trof f, you can get a reasonable approximation
of eqn output by using neqn. However, printers used with nrof f may be unable
to print many of the special characters used in equations.

Specifying Equations
Mathematical documents contain both displayed equations and standard text mixed with
mathematical expressions. The eqn preprocessor allows you to typeset both forms.

Displayed Equations
For equations that appear outside the body of the text, mark the beginning of each equa-
tion with an . E Q and the end with an .EN. Note that these delimiters may or may
not also be defined as macros. They are recognized by eqn as flags to begin and end
processing.
If they are not defined as macros by the package you are using, you can define
them yourself, or can simply supplement them with trof f requests (such as .ce to
center the equation) as desired.
If you are using the m s macro package, . E Q and . E N are defined as macros,
and the equation is centered by default. Thus, if you type:
- EQ
C=Ax+By
.EN
the output will be:
C =Ax +By
In m s , you can also left justify the equation using - E Q L or indent it using
.EQ I. You can further specify an arbitrary equation number or label that will be
printed at the right margin. For example, the lines:
.EQ I (13a)
C=Ax+By
-EN
 0 Typesetting Equations with eqn 0 235

produce the following:

C=Ax+By
The mathematical symbols +, -, =, and ( ) are typed in just as they appear in the
equation.
If you’re using the mm macro package, put the .EQ/ . E N pair inside a
.D S / .DE pair so that the format looks like this:
.DS
.EQ
equation
.EN
-DE

This automatically centers the displayed equation. You can also use a break producing
request (such as .b r or .s p ) immediately following the .D S macro but before the
.EQ macro to display the equation at the left margin of the text.

Inline Expressions
If you are using m s or nun, .EQ and .EN imply a displayed equation and so cannot
be used for short inline expressions. But eqn provides a shorthand notation for
displaying this type of expression. You can define any two characters as delimiters to
mark the beginning and end of an inline equation, and then type the expression right in
the middle of the text. To do this, define the equation delimiters within an .EQ and an
.E N at the beginning of your file.
For example, to set both delimiters to #, add the following lines:
.EQ
delim # #
-EN
If you’re using mm,do not use the .D S / .DE pair to enclose a .EQ/ . E N pair
that only defines the delimiters for inline equations. If you do, extra blank lines will
appear in the output.
Do not use braces ( { }), a circumflex (A), a tilde (-), or double quotation marks
(”) as delimiters because these have a special meaning to eqn. Choose characters that
you are unlikely to use within any equation in the document. After you have defined
your delimiter, you can begin using it within a line of text as in the following example:
The possible prices of an ice cream cone in cents are
# y sub 1 = 7 5 # , # y sub 2 = 8 5 # , and # y sub 3 = 9 5 # .

This produces the line:

Assume that the possible prices of an ice cream cone in cents are y 1=76, y2=85,
and y 3=95.
The e q n program leaves enough room before and after a line containing inline expres-
sions with fractions or large characters so that they don’t interfere with the surrounding
lines.
236 UNlX Text Processing 0

To turn off the delimiters, use:

.EQ
delim o f f
.EN
Throughout this section, we will use the delimiters ## in our
I e q n examples. However,
we will typically show the results as a displayed equation.

Spaces in Equations
You may have noticed in the previous example that the word sub is surrounded by
blanks, and the subscript is separated from the = sign with a blank. Spaces and new
lines are used to tell e q n that certain words belong to the e q n vocabulary and
deserve special treatment. The spaces and new lines that you type in the input equation
do not appear in the printed output.
For example, all of the following equations:
#C=Ax+By#
# C = Ax iB y #
#C= A x +
BY #
produce the same output:
C =Ax +By
Note that the spaces and newlines were ignored by e q n .
You should use spaces as freely as possible to break up more complex equations
and make your input more readable and easier to edit. Remember that any spaces or
newlines you enter within an equation are not printed out. This i s often a point of con-
fusion for new users. If your equation doesn't turn out the way it should, chances are
you missed typing in a space somewhere. A useful rule of thumb is: when in doubt,
use a space.

Printing Spaces in the Output

You may want to fine-tune the printed appearance of an equation by adding spaces
between groups of terms. If you want to print spaces in the output, use a tilde (-) for
each space. A circumflex ( A ) gives a space half the width of a tilde. For example:
#C-=-Ax-+-By#

yields:
c =Ax +By
and:
#C"=^Ax^+^By#

yields:
C =Ax + B y
 0 Typesetting Equations with eqn 0 237

You can also use tabs to separate parts of an equation, but the tab stops must set by the
t r o f f .t a request. For example:
.ta li 1.5i 2i 2.5i
- EQ
x sub 1
+x sub 2
+ s sub 1
=10
.EN
- EQ
- 2 ~sub 1
+ s sub 1
=42
.EN

yields:
X! +x2 +SI =10
-& 1 +s I =42
(Note that each equation must have its own pair of .EQ/ .EN delimiters.) Another
way of aligning equations uses the e q n words m a r k and lineup, as you will see
later.

Subscripts and Superscripts: A Common Use

Perhaps the most common application of eqn is in generating subscripts and super-
scripts within a line of text or a table. As you have seen in previous examples, sub-
scripts are denoted by the word sub. Superscripts are designated by sup. For exam-
ple:
# y sub 1 = x s u p 2"+^1X

yields:
y ,=x2+ 1
There are two simple rules to remember in writing subscripts and superscripts:

1. Put at least one space or space delimiter (such as or -) before and after
the words sup and sub.
2. Leave at least one space or space delimiter after the subscript or superscript.

Let's see the effect on the output when you omit necessary spaces. For example:
# y sub 1 =x sup2^+*1#
yields:
y1zxsup2+ 1
238 0 UNlX Text Processing 0

and:
B y s u b 1 =x s u p 2+"1#

yields:
2+ 1
Y l=x
If you don't leave a space after s u b or s u p (as in the first example), eqn will not
recognize them as special words, and so will not produce a subscript or superscript.
Also, if you don't leave a space after the subscript or superscript, eqn thinks that the
character(s) following it are still part of the subscript or superscript. This is a very
common mistake made by new users.
You can also write subscripted subscripts and superscripted superscripts. If a
superscript and subscript both appear for the same item, s u b should come before
sup. Therefore:
# a sub k s u p 2 #

yields:
a:
Reversing the order of the words:
#a s u p 2 s u b k#

yields:

Some equations also require you to type chemical symbols like:

2He4

Because s u p technically means a superscript on something, you must use a place-

holder (a pair of double quotation marks) before the word sup and write this expres-
sion as:
#"" sup 2 He sub 4 #

Using Braces for Grouping

Normally, you would use a blank or a space delimiter to signal the end of a subscript or
superscript. But if your subscript or superscript consists of two or more characters or
words separated by blanks, or if you are writing nested subscripts or superscripts, this
will not work. In this case, use braces to mark the beginning and end of your subscript
or superscript.
I
0 Typesetting Equations with eqn 0 239

For example, the line:

#r s u b {i=5;t=10Ayears]#

yields:
'i=5;f=lOyears

In contrast, this line without the braces:

# r sub i=5;t=10Ayears#

yields:
ri=5;,=10

In the first example, we used braces to force eqn to treat the string:
i=5;t=10 y e a r s

as a subscript. Use braces to make your intent perfectly clear whenever you are unsure
of how eqn will treat the equation. You can also use braces within braces, as in the
line:
#e sup { i sup {k+l}]#

which yields:
ik+l
e
Make sure that a left brace always has a corresponding right brace.
If you have to print braces in your document, enclose them in double quotation
marks like " { " and " ) ".

Special Character Names

In many mathematical equations, you use the Greek alphabet to define variables. To
print Greek letters, spell them out in the case that you want. For example, delta
produces 6, and DELTA gives A. Thus, you only need to spell out the character n, as
in:
#pi r sup 2X

to print:
nr
Note that special names don't exist for all uppercase Greek letters, such as
ALPHA or ETA, because they are identical to the equivalent English letters. See Table
9-1 for a list of Greek letters.
240 0 UNlX Text Processing 0

TABLE 9-1. Names for Greek Letters

gamma Y

A common mistake is to forget to put a space around the Greek name. For exam-
ple, typing:
#f (theta)#

yields:
f (theta )

and not:

which i s what we want. Because there are no spaces surrounding the word t h e t a ,
e q n doesn't recognize it as a special word.
You can also use t rof f four-character names for characters, as in the descrip-
tion:
#c = a \(pl b#

which yields:
C=a+b
 0 Typesetting Equations with eqn 241

Special Symbols .
The eqn program recognizes the sequences in Table 9-2 as belonging to the e q n
vocabulary, and translates them to the appropriate symbols.

TABLE 9-2. eqn Special Symbols

Sequence Symbol Sequence Symbol

>= 2 approx -
I

<= <
- nothing
-- - cdot
!= + times X
+- k de 1 V
-> + grad V
W<- t ... ...
<< << 1 - - - 1

Fn
, . e - 9

>> >> sum

in€ 00 int
partial a prod
ha1 € ‘/2 union V
prime , inter n
The following examples illustrate the use of these character sequences.
#C s u b 0 prime
yields:
co
and:
#O <= a <= 1#

yields:
OIaI 1
and:
# d e l y / d e l x#

yields:
vylVX
and:
# p a r t i a l x / p a r t i a l t#

yields:
ax/at
242 0 UNlX Text Processing 0

Digits, parentheses, brackets, punctuation marks, and the following mathematical words
are converted into roman font instead of the italic font used for other text:
sin cos tan sinh cosh t a n h arc
max min l i m log In exP
Re Im and if €or det

Summations, Integrals, Products, and Limits

Summations, integrals, products, and limits often require an upper and lower part
around the symbol. The word from indicates the character sequence to be entered at
the lower part; the word t o indicates the upper part. These parts are both optional, but
if they are used, they should appear in that order. For example, you would type:
#Expected-Value-=-sum from {i=l) t o inf pi s u b 1 X sub i#

to print the following expression:

m

Expected Value = z.xiXi

i=l

Notice that we used braces around the from part although this was not neccessary
because there were no imbedded blanks in the string i = l . But if the from and t o
parts contain any blanks to separate special words, you must use braces around them.
A from does not necessarily need an accompanying t o , as you will see in the
following example:
#lim from {m -> inf} sum from i = O t o m c sup i #

which yields:
m

Square Root Signs

To draw a square root sign, use the word s q r t . For example:
Isqrt { b sup 2 - 4ac)#

yields:
qb 2-4ac
Square roots of tall quantities appear too dark and heavy. Big square root quantities are
better written to the power ‘ / 2 , as in:
 Typesetting Equations with eqn 0 243

Creating a cube root or a higher root sign requires a little imagination. You can think
of a cube root sign, for example, as consisting of two parts: a superscript 3 (with noth-
ing before it) and a square root sign. However, you can’t type:
#sup 3 sqrt X X

because a s u p is a superscript on something. You must use a pair of double quotation

marks as a placeholder for sup. For example:
#““ sup 3 sqrt x#

yields:

Enclosing Braces and Brackets

You can generate big brackets [] , braces { } , parentheses
, and bars I around()
quantities by using the words left and right, followed by the desired character.
For example:
#P-=-R-left [ l A - A { l + i sup n } over i right I #

yields:

P=R 1
r
1 - l+in
i

The resulting brackets (and any character you specify) are made big enough to enclose
the quantity. (Braces are typically bigger than brackets and parentheses.) Note the
spaces surrounding the words left and right and the character to be expanded.
Two other characters that you can use are the floor and ceiling characters
shown in the following example:
#left floor a over b right floor !=
left ceiling x over y right ceiling#

which yields:

A left does not need a corresponding right. If the right part is omitted, use
braces to enclose the quantity that you want the left bracket to cover. This is useful
when you are making piles, as you will see in the next section.
You can also omit the left part, although technically you can’t have a right
without an accompanying left. To get around this, you must type:
#left ‘I“ expression right ) #

The left ‘11‘ in this equation means a “left nothing.”

244 0 UNlX Text Processing 0

Other Positional Notation

In addition to subscripts and superscripts, eqn lets you format expressions containing
fractions, matrices, and vertical piles.

Fractions
Making a fraction, such as one-third, is as simple as typing “ 1 over 3.” For more com-
plex fractions, eqn automatically positions the fraction bar and adjusts its length.
Thus, the line:
#Income over Capital-=-Income over Sales-times-Sales
over Capitals

yields:
Income - Income Sales
X
Capital Sales Capital
When you have both a s u p and an o v e r in the same equation, eqn does sup
before over. However, you can always use braces to tell eqn what part goes with
what. For example, you would type:
#e sup {k over t]#

to yield:
-k
e‘
You would not type:
#e s u p k over t #

The latter form produces:

which i s not what we want.

Arrays and Matrices

To make an array or a matrix, use the word matrix , and the words l c o l , c c o l ,
and r c o l to denote the position of the columns. For example:
.EQ
matrix {
lcol 11 above 0 )
rcol {half above -1)
1
- EN
yields:
 0 Typesetting Equations with eqn 0 245

1 '12
0 -1
This produces a matrix with the first column left justified and the second column right
justified. Each item is separated from the item below it by the word above. You can
also center the columns using ccol. You can adjust each column separately and use
as many columns as you like. However, each column must have the same number of
items in it as the other columns.
A matrix should be used when the items in the columns don't all have the same
height (for example, when you have fractions mixed with whole numbers). This forces
the items to line up because matrix looks at the entire structure before deciding what
spacing to use.

Vertical Piles
To make vertical piles or columns of items, use the word pile before the equation
description and the keyword above to separate the items. You can also enclose the
piles in big braces or big brackets. For example:
- EQ
P-=-left [
p i l e ( nu sub 1 above nu sub 2 above cdot
above cdot above cdot above nu sub N }
right ]
.EN
yields:

The items are centered one above the other and separated by the word above. Braces
enclose the entire pile list. The items in the pile can themselves contain piles.
You can left justify (lpile), right justify (rpile), or center (cpile) the ele-
ments of the pile. (A cpile i s the same as a regular pile.) However, the vertical
spacing you get using these three forms will be somewhat larger than the normal pile.
For example:
- EQ
f sub X (x)^=^left {
r p i l e { 0 above 2 x above 0 }
- - l p i l e { x < 0 above 0 <= x <= 1 above x > 1 )
- EN
yields:
246 0 UNlX Text Processing 0

I 0 x<o
f x ( x ) = 2x O I X I I
0 x>l

Note that in this example, we have a left brace without a corresponding right brace.

Diacritical Marks
With e q n , writing diacritical marks on top of letters is straightforward. The words
known by e q n follow, with examples of how they appear on top of the letter x:
-
bar X
under -
X
dot X
..
dotdot X
hat 2
tilde
I

X
vec jt
dyad ft

The following examples show how these keywords are used:

#cr e h a t pes#

yields:
crGpes
and:
#Citr o d o t d o t en#

yields:
Citroen
and:
#a vec f b vec#

yields:
d+2
and:
#X b a r sub s t #

yields:
-
x.7,
 0 Typesetting Equations with eqn 0 247

The e q n program positions the diacritical marks at the appropriate height. It also
makes bar and under the right length to cover the entire item. Other marks are
centered above the character(s).
Typing words with diacritical marks may seem confusing at first because you
have to leave spaces around the letter and its corresponding mark. Just remember that
e q n doesn't print the spaces you type in.

DefiningTerms
In some documents, you type a string of characters often, either within the text or
within several equations. If you notice a string that is frequently used, you can name it
using a d e f i n e statement within an .EQ and . E N . Then you can use the name
within an expression instead of typing the whole string.
Suppose you notice that the string 2 sup i appears repeatedly in equations.
You can avoid retyping by naming it 2 i, for example, as in the following commands:
* EQ
define 2 i '2 sup i'
.EN
You should enclose the string between single quotation marks or between any two char-
acters that don't appear inside the definition. After you've defined a term, you can use
it as a convenient shorthand in other equations, just as if it were one of e q n ' s special
keywords.
A note about using definitions: although a definition can use a previous defini-
tion, do not define something in terms of itself. Thus:
.EQ
d e f i n e 2i '2 s u p i'
d e f i n e 1 / 2 i '1 over 2i'
-EN

is acceptable, but:
- EQ
d e f i n e X 'X bar'
- EN
is not because X is defined in terms of itself. If you want to do this, protect the X in
the definition with double quotation marks, as in:
- EQ
define X ' "X" bar '
- EN
You can also redefine e q n keywords. For example, you can make / mean over by
typing:
248 0 UNlX Text Processing 0

- EQ
define / 'over'
.EN

Quoted Text
You have seen the use of double quotation marks as placeholders (in the sup, sqrt,
and d e f i n e examples) when e q n needs something grammatically but you don't
want anything in the output. Quotation marks are also used to get braces and other
e q n keywords printed in the output. For example:
#"{ size b e t a } " #
prints the words:
{ size beta}
instead of looking up the two words s i z e and b e t a in the e q n vocabulary and
converting them. (The word s i z e is used to change the size of the characters from
the 10-point default.)
Any string entirely within quotation marks is not subject to font changes and
spacing adjustments normally done by t r o f f or n r o f f on the equation. This pro-
vides for individual spacing and adjusting, if needed. Thus, the line:
#italic "COS (x)" + cos (x) #

yields :
cos(x)+cos@ )
To print a literal quotation mark, you must escape it with a backslash character in the
form Y'.

Fine-Tuning the Document

Typesetting a technical document is not only a matter of getting the e q n vocabulary
right so you can print the appropriate mathematical expressions. Although e q n tries to
make some actions automatic and puts items in the proper places, some fine-tuning is
occasionally needed. With eqn, you can line up equations, define font sizes and
types, and vary horizontal and vertical spacing.

Lining Up Equations
Earlier we showed you how to line up pieces of an equation using tabs. Another
method of doing this is to use the commands mark and l i n e u p . This is useful
when you have to line up a series of equations at some horizontal position, often at an
equal sign.
 0 Typesetting Equations with e q n 0 249

For example, you would type in:

-EQ
mu-mark =-lambda t
.EN
.EQ
lineup =-int from 0 to t lambda d z
.EN

to line up the two equations:

p=h2
t

= J?dz
0

The word mark can appear only once at any place in an equation. Successive equa-
tions should also contain lineup only once. Thus, when you have a series of equa-
tions that require you to line up items in more than one position, like the following:
+XI +x2 = 34
+ 4a2 = 28
+4x 2 = 56
it might be better to line up the pieces of the equation on the left-hand side using tabs,
and those on the right-hand side using m a r k and lineup.
If at all possible, you should type in the longest expression first to serve as the
marking point. If you type in shorter expressions first, m a r k will not have enough
room to line up successive longer expressions.

Changing Fonts and Sizes

In e q n , equations are automatically set in 10-point type, with standard mathematical
conventions to write some characters as roman or italic. To change sizes and fonts, use
the following keywords:

size Change to any of the following legal sizes:

12, 14, 16, 18, 20, 22, 24, 28, 36
You can also change the size by a relative amount, such as
size +2 to make a character 2 points bigger, or s i z e -2
to make it 2 points smaller,
bold Change to bold.
fat Widen the current font by overstriking.
250 0 UNlX Text Processing 0

italic Change to italic.

roman Change to roman.

Like sup and sub, these keywords only apply to the character(s) immediately
following them, and revert to the original size and font at the next space. To affect
more complex or longer strings (such as a whole equation), use braces. Consider the
following examples:

#bold qP# qp
#roman alpha-beta# aP
#fat half# '/2

#size + 3 x =y# x =y
#size 8 {A + B } # A +B

If the entire paper is to be typeset in a nonstandard size or format, you can avoid rede-
fining each and every character sequence by setting a global size (gsize) or font
(gfont) that will affect the whole document. You can set this up at the top of your
file (or wherever the font and size changes begin) within an .EQ and - EN.
For example, to change the fonts to roman and the size to 12, you could enter:
- EQ
gfont R
gsize 12
-EN

The rest of the equations in the document (up to another gfont or gsize) will be
set in 12-point roman type. You can use any other t r o f f font names in place of R.

Horizontal and Vertical Motions

You have already learned how to obtain small extra horizontal spaces in the output
using - and A . To move terms at some arbitrary length backward or forward, use the
commands back n and fwd n, where n denotes how far you want to move, in 1/1oOs
of an em. (An em is about the width of the letter m).
You can also move items up or down using up n or down n, where n is the
same unit of measure as described. These local horizontal and vertical motions affect
only the character(s) next to the keyword. To move larger strings or whole expressions,
enclose them in braces.

Keywords and Precedence

Braces are used to group items or change the precedence of operations if you are unsure
of how eqn will treat multiple keywords in a single expression. If you don't use
braces, eqn performs the operations in the following order:
 0 Typesetting Equations with eqn 251

d y a d vec u n d e r b a r t i l d e h a t d o t d o t d o t
f w d b a c k down u p
f a t roman i t a l i c b o l d s i z e
s u b sup s q r t over
from to

All operations group to the right, except for the following, which group to the left:
over sqrt l e f t r i g h t

Problem Checklist
The e q n program usually displays self-explanatory messages when it encounters a
syntax error or any other formatting error. To check a document before printing, type:
$ eqn jiies > /dev/null

This discards the output but prints the error message. Some of the error messages you
might encounter are:
eqn: s y n t a x error b e t w e e n l i n e s 1 4 a n d 42, f i l e book

A syntax error (such as leaving out a brace, having one too many braces, having a s u p
with nothing before it, or using a wrong delimiter) has occurred between lines 14 and
42, approximately, in the file b o o k . These line numbers are not accurate, so you have
to look at nearby lines as well. If the following message is displayed:
word o v e r f l o w

you have exceeded the limits of t r o f f ? s internal buffer. If you print the equation as
a displayed equation, this message will usually go away. If the message is l i n e
o v e r f l o w , the only solution is to break up the equation across multiple lines, mark-
ing each with a separate .EQ and .EN. The e q n program does not warn about
equations that are too long for one line. If the following message is displayed:
eqn: f a t a l error: U n e x p e c t e d e n d o f i n p u t a t 2 sub a

you forgot to put a closing quotation mark after the string 2 s u b a when you named
it in the d e f i n e statement.
It is also easy to leave out an ending delimiter in an equation. In this case, e q n
thinks that successive character sequences (which may run to hundreds of lines) are still
part of the inline expression. You may then get an overflow error or a garbled docu-
ment. The c h e c k e q program checks for misplaced or missing inline delimiters and
similar problems.
For example, when run on a draft of this chapter, c h e c k e q produced the fol-
lowing report:
$ checkeq sectl
sectl :
New d e l i m s ##, l i n e 6
2 l i n e t t , l i n e s 618-619
252 0 UNlX Text Processing 0

2 line # # , lines 619-620

2 line # # , lines 620-621

EQ in # # , line 689
EN in # # , line 691
13 line # # , lines 709-721

2 line # # , lines 1300-1301

2 line # # , lines 1301-1302
Unfinished # #

This report (which ran to 66 lines) was telling us that somewhere before line 618 there
was an unclosed inline equation using the delimeter #. Sure enough, the following error
was found:
B#f( theta )
Because there was only one delimiter, eqn gets “out of phase” and all subsequent
delimiters are misplaced. After we fixed this one error, checkeq printed the follow-
ing “null” report:
$ checkeq sect1
sect 1 :

Because a simple problem like the one shown here can cause every subsequent equation
in the file to be garbled, and can waste an entire formatting run, it makes sense to run
checkeq before you format any files containing equations.
 C H A P T E R

10

Drawing Pictures

If you are one of those who can’t draw a straight line, let alone a decent picture or
graph, you probably replace pictures with verbal descriptions. Perhaps you know what
it is like to describe a drawing to a person who knows how to draw. The pic prepro-
cessor requires you to follow the process of using “words” to describe something pic-
torial.
The pic preprocessor has a dual purpose. The first is to provide a “natural
language” method of describing simple pictures and graphs in your documents. The
second i s to offer a “programming language” for generating pictures and graphs with
minimal user input. Learning pic is an iterative process: describe what you want and
then look at what you get. We have included many examples that show both the
description and the resulting picture or graph. Take the time to create variations of
these descriptions, making modifications and improvements.
The pic preprocessor was designed to produce output on a typesetter, which
makes pic expensive and difficult to learn. Fortunately, some graphics terminals and
most laser printers can be set up to display or print pic drawings. Access to one or
the other is essential if you are going to get enough practice to know how pic
responds.
As a preprocessor, pic is a program that processes a specific portion of an input
file before the whole document goes to the t r o f f formatter. (The nrof f formatter
cannot produce pic output for terminals or line printers.) The preprocessors translate
your description into low-level formatter requests for t r o f f.
Just like with t b l and e q n , a pair of macros in the input file mark the begin-
ning and end of input to be processed by pic. The delimiters for pic are:
- PS
pic description
.PE

When you format a document that contains pic descriptions, you must invoke the
pi c preprocessor as follows:
$ picfile I troff I device

253
254 0 UNlX Text Processing 0

The p i c Preprocessor
Imagine that you have to describe over the telephone the following picture:

You might say: “There’s an ellipse at the top. Arrows are connected to two boxes and
a circle below it.” Now, think about describing this picture to someone who is attempt-
ing to draw it. No matter how careful you are, you realize that it is difficult to translate
a drawing into words.
“First, draw an ellipse. Move down and draw a circle below it. Then draw
one box to the left and draw another box of the same size to the right.
Then draw an arrow from the bottom of the ellipse to the top of the left-
hand box. Then draw a line from the bottom of the ellipse to the top of the
right-hand box. The last thing to do is draw a line between the circle and
the ellipse and put arrowheads on both ends. ”

Here’s what the actual pic description looks like:

.PS
down
ellipse
move down 1 . 2 5
circle radius - 3 5
move left l i from left of last circle; box
move right li from right of last circle; box
arrow from lower left of last ellipse t o top of 1st box
arrow from lower right of last ellipse t o top of 2nd box
arrow <-> from bottom of last ellipse t o t o p of last circle
-PE
Even though you may know nothing about pic, you should be able to make some
sense out of this description. It names several objects: an ellipse, two boxes, a circle,
 0 Drawing Pictures 0 255

and three arrows. It specifies motion in inches as well as changes in direction. It also
arranges some objects in relation to others, locating the boxes to the left and right of the
circle and drawing arrows between the ellipse and the circle.
Having seen a full description of a pic drawing in this example, you should be
able to get something of the flavor of pic. The simpler the drawing, the less explain-
ing you have to do. We won’t go into any more detail about this pic description right
now. We’ll look at it later in this chapter after we’ve covered the basics of the pic
language.

Naming Objects
The pic program is easy to use if you are describing only a single box or a circle. To
draw a circle, you name that object within the .P S / .PE macros:
.PS
circle
-PE

When this description is processed by pic it produces:

0There are seven graphics primitives: arc, arrow, box, circle,

ellipse, line, and spline. We will show these primitives in examples that
present additional aspects of pic.
In using a computer language, you have to be precise, using as few words as pos-
sible to get the picture you want. This means that you allow the program to make as
many of the decisions about the drawing as is practical. After you understand pic’s
normal behavior, you will know what pic will do on its own.
For instance, we didn’t specify the size of the circle in the last example. By
default, pic draws a circle with a diameter of ‘12 inch (or a radius of .25 inch). You
can get a circle of a different size, but you have to specify the size.
.P S
circle r a d i u s .5
- PE
The pic program understands any number to be in inches. You specify the size of a
circle by giving its radius, which can be abbreviated as rad, or its diameter,
which can be abbreviated as diam. The previous input produces a circle twice the
size of the standard circle:
256 0 UNlX Text Processing 0

Similarly, if you specify box, you will get a box with a height of .5 inch and a
width of .75 inch. You can get a larger or smaller box by changing its dimensions:
.PS
box height li width . 5
.PE
The output for this example is a box twice as high as it is wide:

You can also use the abbreviations h t and wid for these attributes. The order
in which you specify the dimensions does not matter, and you can change one attribute
without changing the other. That is how we can draw a square:
.PS
box ht - 7 5
.PE
The default width is already .75 inch, so this pic description produces:

With the attribute same, you can reuse the dimensions specified for a previous
object of the same type. For instance, after you had described the square box, box
same would duplicate a square of the same size.
 0 Drawing Pictures 0 257

Labeling Objects
To provide a label for any object, specify the text within double quotation marks after
the name of the object. The label is placed at the center of the object.
.P S
box ht - 7 5 " S q u a r e One"
-PE

This pi c description produces:

0Square One

Even if a label does not contain blank spaces, you must enclose it within double quota-
tion marks. Each individually quoted item will be output on a new line.
box wid . 5 "Second" "Square"

This description produces:

Second
Square

Because trof f, not pic, actualfy handles the text, pic doesn't really try to fit a
label inside an object. You must determine the amount of text that will fit. The pic
program ignores lines beginning with a period, permitting you to use t r o f f requests
to change the point size, font, or typeface. It is best to avoid spacing requests, and be
sure to reset any change in point size.
When you specify a single text label with a line, pic centers it on the line.
For instance, inline t r o f f requests can be used to print a label in 14-point italic (i.e.,
4 points larger than the current point size).
- PS
line " \ f I \ s 14pic\ s 10 \ f R"
.PE

It produces:

*
Because the standard placement of labels is not always useful, you can specify the attri-
butes above or below. In the following example, the point size is specified using
the following .ps request:
258 0 UNlX Text Processing 0

.ps +2
l i n e "\fIPIC\fR" above
.ps -2
It produces:

PIC

If you supply two quoted arguments with l i n e , the first will be printed above the line
and the second printed below.
You can also select a line or box that is d o t t e d or dashed, as you can
see in the next example:
b o x d o t t e d "\f (CWbox d o t t e d \ f P " above

Note the inline request to invoke the constant-width font for the label. The above
keyword places the label above the center line of the box. This description produces:

................
ox dotted

................

The box, composed of dots, contains a label printed in constant-width font. It is obvi-
ous here that p i c made no attempt to fit the label "inside" the box. The above
attribute does not place text above the box, but rather above the center of the box. The
description:
l i n e dashed "sign here" below

produces a dashed line:

-----
sign here

If the attributes of texture are followed by a value, pic will try to keep that amount of
spacing between the dashes or dots. The description dashed .1 will result in
dashes spaced .1 inch apart.

pic's Drawing Motion

After you have named an object and determined its size, you have to think about where
p i c is going to draw it. (Indentation and other matters concerning the placement of
the drawing on the page are supplied by either the .PS/.PE or .DS/.DE macros.
The p i c program places a single object at the left margin. If you name three objects
in the same description, where will pic draw them?
 0 Drawing Pictures 0 259

.PS
circle "A"
line ,,1" 1.2..
box "E"
.P E
The following output is produced:

Objects are placed one after another from left to right. The pic program assumes that
objects should be connected, as in the following example:
.PS
box ht 1.25
box ht 1
box ht - 7 5
box ht - 5
.
PE
This description produces a row of boxes of decreasing size:

If you don't want objects to be connected, you can move before specifying the next
object. In the next example, move places a box to the right of a c i r c l e :
.P S
circle "A" ; move; box "B"
.PE
As shown in this example, pic commands can be entered on the same line, separated
by semicolons, instead of on separate lines. This description produces:
260 0 UNlX Text Processing 0

Changing Direction
As you have seen, pic places objects in a continuous motion from left to right. You
can also get pic to change direction using the attributes left, right, up, or
down. We'll see examples of their use shortly.
The distance of a move is the same length as a line (.5 inch). If you want to
change the distance of a move or the length of a line, then the change must be
accompanied by an attribute of direction. Although it seems natural to write:
line 2; move 1; arrow 1 Wrong
pic does not accept this command unless you specify directions for all three cases.
When pic objects to your choice of words, it will display the offending line, using a
caret (") to mark the error.
pic: syntax error near line 1, file test
context i s
line 2 "; move 1

Only the first error on the line is marked. (It is acceptable to write line; move,
using the standard length and distance.) The next example shows how to draw a line of
a specified length and how to move a specified distance. The pic program assumes
that any value i s in inches; thus you can say 2i or simply 2 to indicate 2 inches.
line u p 2; move down 1; arrow right 1

Note that the attribute of direction precedes the distance. The preceding description
produces:

You cannot specify down 1 or right 1 without also specifying either a line or
move. These attributes change the direction of the motion used to draw objects. They
do not cause movement. The attributes of direction affect the position of the objects
that follow it, as shown in the next example.
 0 Drawing Pictures 0 261

.P S
, line; box "B"
down; circle "A"-
.P E
These objects are drawn from top to bottom:

If you describe a change of motion, it affects the points where objects are con-
nected. Look what happens if we specify the attribute down after the circle:
- PS
-
circle "A" , down; line; box "Br'
.PE

Now the line begins at a different position:

The pic program keeps track of the start and end points for each object, and
their relationship to the direction in which objects are being drawn. The next object is
drawn from the exit point of the previous object. Entry and exit points may seem obvi-
ous for a line, but not so obvious with circles. When the motion i s from left to right, a
circle's entry point is at 9 o'clock and its exit point is at 3 o'clock. When we specified
down after the circle in the first example, the exit point of the circle did not
change; only the direction in which the line was drawn from that point changed. Entry
and exit points are reversed when the motion is from right to left, as specified by the
left attribute.
262 0 UNlX Text Processing 0

left; arrow; circle "A"; arrow; box "B"

This description produces:

You can draw a diagonal line by applying two changes in direction. Look at how
we describe a right triangle:
.P S
line down li
line right li
line up li left li
.PE

This description produces:

The diagonal line is drawn by combining two attributes of direction, up and left.
You can describe a continuous line using then. In the next example we use arrow
to demonstrate that we are describing a single object.
- PS
arrow down li t h e n right li then up li left li
.PE

When using then, you have to define the motion on a single line or escape the end of
the line with a backslash (\). It produces:

If the description ended with:

 Drawing Pictures 0 263

then u p l i then left li

we would have a 1-inch square instead of a right triangle.

An a r c is a portion of a circle. Naming four arcs consecutively will draw a cir-
cle. An arc is drawn counterclockwise from the current position (from 6 o'clock to 3
o'clock, for instance). The next example uses arcs to produce a box with rounded
comers:
line right 1; arc; line up ; arc
line left 1; arc; line down; arc

This description starts with the bottom line of the curved box. The motion is counter-
clockwise.

The attribute c w draws an arc in a clockwise direction:

arc "A"; arc "B*' cw

This description produces:

Note that text is placed at what pic considers to be the center of the arc, which is the
center of the corresponding circle.
A spline is a cross between an a r c and a line. It is used to draw
smoothed curves. In this example, a spline traces a path between two circles.
circle rad - 2 5
spline right 1 then down . 5 left 1 then right 1
circle same

This description produces:

264 0 UNlX Text Processing 0

A spline is used in the same way as a line. When drawn continuously using
then, a spline that changes direction draws a curve. (Similarly, a line would
produce an angle.) We’ll see more examples of spline later.

Placing Objects
It isn’t always useful to place objects in a continuous motion. Look at the following
example, which seems like it ought to work but doesn’t:
.P S
down; arrow; box
right; arrow; e l l i p s e ; arrow
.P E
This pic description produces:

Note the short arrow, drawn from the box to the circle. What happened? The end point
of the box was not on the right, but on the bottom, because the motion in effect where
the box is drawn is down. Changing direction (right) affects only the direction in
which the arrow is drawn: it does not change where the arrow begins. Thus, the arrow
is drawn along the bottom line of the box.
Sometimes, it is best to place an object in relation to previously placed objects.
The pic program provides a natural way to locate objects that have been drawn. For
example, the attribute first locates the first occurrence of an object, and the attribute
from specifies that the object serves as a starting point for the next object drawn.
.P S
c i r c l e ; move; c i r c l e ; arrow u p from f i r s t c i r c l e
- PE
It produces:
 0 Drawing Pictures 0 265

You can reference each type of object using an ordinal number. Referring to the order
in which an object is drawn, you can say first box (1st box is also acceptable)
or 2nd circle. You can also work back from the last object, specifying the
l a s t b o x o r 2nd l a s t box.
The center of each object is used as the reference point. In the last example, the
arrow was drawn from the center of the circle. The attribute chop can be used to
chop off the part of the line that would extend to the center of each circle. In the next
example, a chopped line is drawn between the first and third circles:
.P S
circle " 1 " ; move down from last circle
circle "2"; move right from last circle; circle " 3 "
line from 1st circle t o last circle chop
.PE

This description produces:

The amount that is chopped is by default equal to the radius of the circle. You can
specify how much of the line is chopped, for use with other objects or text, by supply-
ing either one or two values after the attribute. If a single value is given, then both
ends of the line are chopped by that amount. If two values are given, the start of the
line is chopped by the first amount and the end of the line chopped by the second
amount.
It is important to remember that movement from a referenced object is meas-
ured from its center, unless otherwise specified. Look at these four circles:

The second circle is produced by the description:

move right from last circle; circle "2"

Because the distance (.5 inch by default) is measured from the center of the circle, there
is only -2.5 inch between the two circles. The third circle is produced by the descrip-
tion:
266 0 UNlX Text Processing 0

move right from right of last circle; circle ?3?

Now the distance is measured from the right of the second circle. There is twice as
much space between the second and third circle as between the first and second. The
fourth circle is produced by the description:
move right from bottom of last circle; circle ? 4 ?

The starting point of the fourth circle (its left ?side?) is .5 inch right from the bottom
of the previous circle.
Using bottom, top, right, and left, you can locate specific points on
any object. In the next example, we solve the problem of turning a comer by specify-
ing the place from which the arrow will be drawn:
.P S
down; arrow; box
right; arrow from right of last box; ellipse; arrow ; box
up; arrow from t o p of last box
.P E
In our earlier example, the arrow was drawn from the bottom of the box; now we
change the starting point of the arrow to the right of the previous box. This descrip-
tion produces:

With boxes and ellipses, you can refer to an upper or lower position:
.PS
box; arrow from upper right of last box;
arrow down from lower left of last box
.P E
This description produces:
 0 Drawing Pictures 0 267

With objects like lines and arcs, it i s more useful to refer to the start and e n d of
the object. For example, here’s another way to draw a triangle:
- PS
line down li
line right
line from start of 1st line to end of 2nd line
.PE
The last line could also be written:
line to start of 1st line
The pic description produces:

You now know enough of the basic features of pic to benefit from a second look at
the pic description shown at the beginning of this chapter. The only thing we haven’t
covered is how to get a double-headed arrow. Because an arrow can also be speci-
fied as line -> or line <-, you can get a double-headed arrow with line
<->.
.P S
1 down
2 ellipse
3 move down 1.25
4 circle radius .35
5 move left li from left of last circle; box
6 move right li from right of last circle; box
7 arrow from lower left of last ellipse to top of \
1st box
8 arrow from lower right of last ellipse to top of \
2nd box
9 line <-> from bottom of last ellipse to top of last \
circle
.PE
The lines in this description are numbered for easy reference in the following exercise.
As i s true with almost anything you describe, a pic description could be written
in several different ways. In fact, you will learn a lot about pic by making even
minor changes and checking the results. See if you can answer these questions:
268 0 UNlX Text Processing 0

Why is down specified before the ellipse? If you removed down,

would the circle be centered underneath the ellipse?
down changes direction of movement. Does pic allow you to say move
1 . 2 5 as well as move down 1 - 2 5 ?
b Where is the exit point of the circle when it is drawn with a downward
motion in effect? If lines 5 and 6 were replaced by:
move left li; box
move right 2i; box

where would the boxes be drawn?

There is 1 inch between the circle and each box. How much space would there
be i f lines 5 and 6 were replaced by:
move left from last circle; box
move right from last circle; box

Hint: The distance of a move is .5 inch, and this would be measured from the
center of the circle, which has a radius of -35 inch.
Line 8 draws an m o w from the lower right of the ellipse to the top of the
right-hand box. If it were simplified to:
arrow from last ellipse t o 2nd box

where would the beginning and ending of the arrow be?

This drawing can present an interesting problem if the circle is omitted. How
would you draw the two boxes if the circle was not there as a reference point?

Fortunately, there is a simple way to deal with the problem presented in the last ques-
tion. Lacking a reference object, you can create an invisible one using the i n v i s
attribute. This lets you specify a circle that is not drawn but still holds a place that you
can reference.
- PS
down
ellipse
move down 1.25
circle radius - 3 5 invis
move left li from left of last circle; box
move right li from right of last circle; box
arrow from lower left of last ellipse to t o p of 1st b o x
arrow from lower right of last ellipse to t o p of 2nd box
.PE
 Drawing Pictures 0 269

This p i c description produces:

One thing that seems hard to get used to is that your current position always changes
after an object is drawn, based on the motion in effect. This means you have to keep in
mind the location of the starting point for the next object that you are going to draw.
You can use braces to enclose an object (or a series of objects or motions) so that
the starting point is unchanged. In the last drawing, if the invis attribute didn’t
solve the problem so easily, we could have used braces to maintain a central point
below the ellipse from which you can move to draw the boxes. Here’s a different
example that illustrates how braces can be used to control your position:
.PS
{arrow d o w n )
[arrow u p )
[arrow left}
arrow right
.
PE
Each object, except the last, is enclosed in braces; all objects share the same starting
point. This description produces:
270 0 UNlX Text Processing 0

Placing Text
Text can be placed in a drawing just like an object. You have to take care in placing
text, as in the next example, where we specify a move so that the compass points are
not drawn on top of the arrowheads:
*PS
{arrow down; move; " S " 1
{arrow up; move; "N" 1
{arrow left; move; "W" 1
{arrow right; move; "E" 1
.PE
Notice that the attributes of direction cause the object to be drawn in that direction and
establish a new motion for successive objects. This description produces:

As mentioned, pic does not really handle text, allowing t r o f f to do the work. In
some ways, this i s unfortunate. The thing to remember is that pic does not know
where the text begins or ends. (You can use the attributes 1 just or r just to have
the text left justified-the first character is positioned at that p o i n t d r right justified-
the last character is at that point. These attributes can also be used with text labels.)
The pic program does not keep track of the start and the end of a text object. It
only knows a single point which is the point where t r o f f centers the text. In other
words, a text item does not cause a change in position. Two consecutive quoted items
of text (not used as labels to another object) will overwrite one another. Objects are
drawn without regard to where the text item is, as shown in the next example:
"Start"; line; arrow; line; "Finish"

This description produces:

This example can be improved by right justifying the first text item ("Start"
rjust) and left justifying the last text item ("Finish" 1 just). As you'll
 0 Drawing Pictures 0 271

notice, though, the picture starts at the margin, and the label is forced out where it
doesn't belong.

Start - Finish

The location of the point that pic knows about is unchanged. Most of the time, you
will have to use the move command before and after inserting text.
Because pic works better with objects than text, the i n v i s attribute can be
used to disguise the object behind the text, and give you a way to place text where you
can point to it.
.PS
down
e l l i p s e i n v i s "DECISION?"
move down 1.25
c i r c l e rad - 3 5 i n v i s "Maybe"
move l e f t l i from l e f t o f l a s t c i r c l e ; box i n v i s "Yes"
move r i g h t l i from r i g h t o f l a s t c i r c l e ; box i n v i s "NO"
arrow f r o m lower l e f t o f l a s t e l l i p s e t o t o p o f 1 s t box
arrow from lower r i g h t o f l a s t e l l i p s e t o t o p o f 2nd box
l i n e <-> from bottom o f l a s t e l l i p s e t o t o p o f l a s t c i r c l e
.
PE

This description produces:

DECISION?

Yes Maybe No

You may have recognized that the description for this drawing is basically the same one
that produced the drawing at the beginning of this chapter. The invis attribute
makes text labels, not objects, the subject of this picture. This should lead you to the
idea that p i c descriptions can be reused. Try to think of the form of a drawing
separately from its content. Most drawings contain forms that can be reworked in the
service of new material.
272 UNlX Text Processing 0

Place and Position Notation

Can you locate the starting points of the arrows on this ellipse?

To write the description for this example is a test of patience and thoroughness, if noth-
ing else. We start at the upper left of the ellipse and move clockwise around the
ellipse.
.P S
ellipse
arrow up left from upper left of last ellipse
arrow up from top of last ellipse
arrow up right from upper right of last ellipse
arrow right from right of last ellipse
arrow right down from lower right of last ellipse
arrow down from bottom of last ellipse
arrow left down from lower left of last ellipse
arrow left from left of last ellipse
.PE
Although you can say upper left or lower l e f t , you cannot say t o p
l e f t or b o t t o m right.
Sometimes pic’s English-like input can get to be cumbersome. Fortunately,
pic supports several different kinds of place and position notations that shorten
descriptions.
You can reduce the phrase:
f r o m bottom o f last ellipse
to either of the following:
from .b of last ellipse
from last e1lipse.b
You can use this notation for the primary points of any object. You can also refer to
the compass points of an object, which provides a way to specify comers. Table 10-1
lists the placename notations.
 0 Drawing Pictures 0 273

TABLE 10-1. pic Placename Notation

Value Position
t TOP
b Bottom
1 Left
r Right
n North
e East
W West
s South
nw Northwest
SW Southwest
ne Northeast
se Southeast

Instead of writing:
from lower left of last ellipse
you might write:
from last ellipse-sw
Another simple way to shorten a description is to give an object its own name. The
name must begin with an uppercase letter. If we assign the placename E l p to the
ellipse:
Elp: ellipse
then we have either of the following ways to refer to specific points:
arrow up left from upper left of Elp
arrow up left from Elp.nw
Here’s the condensed version of the description for the previous example:
.PS
Elp: ellipse
arrow up left from Elp.nw
arrow up from E1p.n
arrow up right from Elp-ne
arrow right from E1p.e
arrow right down from Elp.se
arrow down from E1p.s
arrow left down from Elp.sw
arrow left from E1p.w
.PE
At least it helps to keep you from confusing the placement of the arrow with the draw-
ing motion.
274 0 UNlX Text Processing 0

If you want to specify a point that is not at one of the compass points, you can
indicate a point somewhere in between two places. You can use the following kind of
construction:
fraction o f t h e way b e t w e e n first .position a n d second .position
or use the following notation:
fraction < first .position, second .position >
The following example shows both forms:
box
a r r o w down l e f t from 1 / 2 o f t h e way b e t w e e n l a s t b o x . s w \
and l a s t b o x - w
a r r o w down r i g h t f r o m 1/2 < l a s t box.se, l a s t b 0 x . e >

Although you may not want to intermix different forms for the sake of someone reading
the description, pic does allow it. The preceding description produces:

The a t attribute can be used to position objects in a drawing.

b o x "A'" , b o x w i t h .se a t l a s t b o x - n w "B"
b o x w i t h . s w a t l a s t b o x - n e "C"

This description produces:

The next example illustrates again the problem of placing text. This time we
want to position callouts above and below the text.
 0 Drawing Pictures 0 275

UNIX System
Programs

PATH=.:/bin:/usr/bin:/usr/local/bin:/usr/~~e~/~in
1

Site-Specific
Directory

We position the text inside a long box. Because the callout lines will point to the box
that surrounds the text rather than to the text itself, we try to specify approximately
where to draw the lines.
.P S
# " # " introduces a comment
#
Ibescribe box; escape end of line to include
# text on separate line
#
Path: box ht - 2 5 wid 4 \
"\f(CWPATH=.:/bin:/usr/bin:/usr/local/bin:/usr/fred/bin\fR"
#
# Describe line down from box and put top of ellipse
# at end of last line; label will be printed
# in 9-point italic.
#
line down from 1/3 <Path.sw, Path.s>
ellipse "\fI\s9Currentw "Directory\sO\fP" with .t at \
end of last line
#
# Describe two lines, one up from box
# and a second down to the point right of it.
#
line up from 1/2 <Path.nw, Path.n>
276 UNlX Text Processing 0

line to 2/3 <Path.nw, Path.n>

ellipse "\fI\s9UNIX System" "Programs\sO\fP" with .b at \
start of last line
#
# Describe the third callout below the box.
#
line down from Path.s
ellipse "\fI\s9Site-Specific" "Programs\sO\fP" with .t at \
end of last line
.PE
Admittedly, positioning callouts is guesswork; it took several iterations to align the cal-
louts with the appropriate text.

Defining Object Blocks

You can describe any sequence of motions or objects as a block. A block is defined
between a pair of brackets ( [ ] ). You can also give the block a placename, beginning
with an uppercase letter. Some of the objects that we have created in this chapter, such
as a square, triangle, or compass, could be defined as blocks and named so that we can
refer to each one as a single object.
Rtriangle: [
linewid = 1
line down then right then up left
1
.ps 18
.ft I
I. 1 I. at Rtriangle.w
"2" at Rtriangle.s
" 3 " at Rtriangle
.ft R
-ps 10
This description produces:

We are able to refer to the compass points of the block, although these points may not
always be where you expect. The number 3 is printed at the center of Rtriangle
according to pic. But in fact its position is the side opposite the right angle. The
"center" of this block is at the center of a box that shares the bottom and left sides of
the right triangle.
 1
Drawing Pictures 0 277

You can also refer to positions for a single block using brackets. The reference
[ 3 .w is a position at the west end of the block.
In this example, instead of specifying individual line lengths, we redefined the
variable linewid. This is the variable that pic accesses to determine how long a
line should be. Shortly, we'll look at all the variables preset by pic. Generally, what
you describe within a block remains local to the block. Thus, linewid would not
affect other lines outside the block. Otherwise, resetting a variable has an effect not
only on other objects in that drawing but also on other drawings in that file.
The best use of blocks in a drawing is to define significant portions so that you
can position them accurately. Blocks usually relate to the content of a drawing. In the
next example, we describe a two-dimensional box to represent a modem.
MOD: [
BOXA: box w i d 1 ht .25 " \(bu \(bu \(bu \(bu \(bu 1.

line from B O X A - n w u p 1 right .5 x

then right 1 then down 1 left - 5 t o BOXA.ne
line from BOXA.se up 1 right .5 then up .25
1
The block, named MOD, consists of a box followed by a series of lines. The box is
given a name, BOXA. The special character sequence \ (bu represents a bullet (inter-
preted by t r o f f , not pic). This description produces:

.....
The next block, named WALL, describes a drawing of a telephone wall socket. It con-
tains two objects, a box named BOXB and a circle inside the box named C I R .
WALL: [
BOXB: box wid - 2 5 ht .5
CIR: circle at center o f BOXB radius - 0 5
] with . s at M O D - n e + (.5,1)
To position this block in relation to MOD, we describe a position 1 inch up and .5 inch
to the left of the top right-hand corner of MOD. Then we draw a spline from the
modem to the wall socket. This introduces us to the fact that no matter how we specify
an object, pic locates that object in a Cartesian coordinate system. We'll look at this
in more detail in a later section. For now, it is sufficent to note how we change posi-
tion by adding or subtracting from the position on the x-axis and y-axis.
MOD. ne+ ( .5, 1 ) adds .5 to the x-axis (moving to the right) and 1 to the y-axis
(moving up) from the coordinates of MOD. ne.
278 0 UNlX Text Processing n

spline from M0D.n up .25 right .5 then right 1 t o center \

of W A L L - C I R
Notice that we can refer to objects inside a block. If we had not named the circle, we
could still refer to it as WALL. circle.
The last thing to do i s to position the text:
move right 1 f r o m WALL.e; " Telephone Line"
move down . 5 from M0D.s "Modem"

This entire description produces the following drawing:

Telephone Line

Modem

Resetting Standard Dimensions

The p i c program has a number of built-in variables that define the values used to
draw standard p i c objects.
Refer to Table 10-2. You can redefine these variables anywhere in a p i c
description. A variable set inside one p i c description will remain in effect for other
descriptions within the same file. One exception i s a variable defined within a block;
that definition is local to the block.
 0 Drawing Pictures 0 279

TABLE 10-2. pic System Variables

Variable Default Value Meaning

arcrad -25 Radius of arc
arrowwid .05 Width or thickness of arrowhead
arrowht .1 Height or length of arrowhead
boxwid .75 Width of box
boxht .5 Height of box
circlerad .25 Radius of circle
dashwid .05 Width of dash
ellipseht .5 Height of ellipse
linewid .5 Length of horizontal line
lineht .5 Length of vertical line
movewid .5 Distance of horizontal motion
moveht .5 Distance of vertical motion
scale 1 Scale dimensions
textwid 0 Width of area used for drawing
textht 0 Height of area used for drawing

For instance, we can specify an oversize arrow by changing the following vari-
ables:
arrowwid = 1
arrowht = 1
linewid = 2
arrow

It produces the following pic drawing:

Controlling the Dimensions of a Drawing

The textwid and textht variables control the width and height respectively, of
the area used by pic on a page. (It doesn’t refer to the amount of space occupied by
an item of text .) These values can also be set as arguments to the .P S macro.
.P S width height
280 0 UNlX Text Processing 0

When you specify the width or height or both, pic scales the drawing to that size
regardless of the absolute dimensions that are specified within the drawing. The only
thing it doesn’t scale adequately is text. It can be easier to describe a drawing with
simple units and have them scaled more precisely to fit on the page than to work with
exact sizes.
A good example of scaling i s turning the rounded box described previously in this
chapter into a representation of a terminal screen.
.PS 2 4
line right 1; arc; line up ; arc
line left 1; arc; line down; arc
.PE

Although the pic description is made up of 1-inch lines, the actual screen produced
by pic will be 4 inches wide and 2 inches high.
Normally, you want trof f to output the regular lines of text on lines that fol-
low the pic drawing. If the . P F (F forflyback) macro is used in place of .PE,
trof f will return to the position it held before the pic drawing was output. This
feature is useful if we want to put formatted text within our large screen.
.PS 2 4
line right 1; arc; line up ; arc
line left 1; arc; line down; arc
.P E
.ft cw
.sp 2
Alcuin Development System 5/31/87
- SP
Please login:
.sp 6

This description produces:

I Alcuin Development System

 0 Drawing Pictures 0 281

You have to remember to provide the space after the text to push the current position
past the end of the screen. Otherwise subsequent text will also appear within the box.

Debugging pic Descriptions

You can invoke the pic preprocessor on its own to have it check through your file
and report any syntax errors. This can save a lot of time, especially if your file contains
other text that will be sent to trof f , assuming that you wouldn?t want the file pro-
cessed unless the pic descriptions succeeded. If you have the file circles, for
example, that contains a pic description, you can invoke pic as:
$ pic circles
If processing is successful, pic output will stream past on your terminal screen. If
pic finds an error in your description, it will print the error message.
If you have several pic descriptions in a file, or you have regular text surround-
ing a pic description, you can send the output to /dev/null, and only the error
messages will be displayed on your screen.
You may want to invoke pic on its own simply to look at the output p i c pro-
duces. For a discussion of the output that pic sends to t rof f , read about line
drawing in Chapter 14.

From Describing to Programming Drawings

As we look at more advanced examples of pic, you may begin to question the amount
of description that is required to produce a drawing. You may be amazed that drawings
that look so simple require so many words. After you realize that you are approaching
the limits of what can be described using an English-like syntax, you may want to look
at pic from another perspective. You can view pi c as a programming language for
generating graphics.
Looking at this other side of pic, you will find that the descriptions are perhaps
more difficult to read but much easier to write. The purpose of a ?programmed? pic
description is not to imitate a verbal description, but to minimize user input, to provide
structures that can be used to produce several kinds of drawings, and to make it easier
to change a drawing.
The focus of the rest of this chapter will be to introduce many of these special
features of pic, including variables, expressions, and macros. But there are more pos-
sibilities than we can attempt to describe. The pic program follows the general UNIX
philosophy that any program should be able to accept input from any program and
direct its output to another program, trof f . Thus, pic descriptions can be built by
other UNIX utilities. For instance, you might develop an awk program specifically
designed for creating flow charts.
282 0 UNlX Text Processing 0

Locating Objects Using Cartesian Coordinates

For more exact positioning of objects and text, pic uses a standard Cartesian coordi-
nate system. The first object drawn, including a move, starts at position 0,O. The x
and y position of a circle, an ellipse, or a box is at the center of the object. For lines,
this position refers to the beginning. For arcs, this position is at the center point of the
related circle. You can position objects using the at attribute:
circle "0,O" at 0,O
circle "1.1" at 1,l
circle "1.0" at 1,0
circle "2.1" at 2,l
This description produces:

The center of the circle is placed at the specified coordinates. You could also write:
c i r c l e w i t h .t a t 1 , l

and it would place the top of the circle at that position. A reference to last c i r -
c l e would still locate the center of the circle, but a line drawn from 1 , 1 would
begin at the top of the circle.
Note that the position of 0,O will not always be the same place on the page. The
first object drawn is the point of reference; movement to the left of that point will cause
0 , O to be moved over towards the center of the page.
box h t 0 . 3 wid 0 . 3 "0,O"
move t o 1,0
box "1.0" same
move t o - 1 , 0
box " - 1 , O " same

This description produces:

It may be helpful to sketch a drawing on graph paper and then translate it into a pic
description. Standard graph paper is divided into quarter-inch squares. When you use
 Drawing Pictures 0 283

graph paper, you might find it useful to set the scale variable to 4. All dimensions
and positions will be divided by the value of scale, which is 1 by default.
It is much easier to describe a drawing in full units rather than fractions. Look at
the following description:
scale=4
line from 0 , O to 0,3 then t o 6,3 then t o 6 , O then t o 0 , O
line from 6,O to 8,l then t o 8,4 then t o 2,4 then t o 0,3
line from 6,3 to 8,4

The distance between 0 and 1 is normally 1 inch. Because we are scaling this drawing
by 4, the actual distance is '/4 inch. It seems easier to describe a point as 2,3 rather than
5,.75. This description produces a two-dimensional box:

Although pic scales the location of text, it is your responsibility to reduce the
size of the text to fit a scaled object. You can also use scale to change the basic
unit of measurement from inches to any other unit. For instance, setting scale to 6
will cause all dimensions and coordinates to be interpreted in picas (6 picas to the inch).
Splines and arcs are much easier to draw using coordinates. In the following
example, we use a spline to draw a smooth curve between several points on a crude
graph.

This graph is produced by the following description:

scale=4
line from 0.0 to 0 , 4
line from 0 , O to 9,0
spline from 0 , O t o 3,3 then to 5,.25 then t o 8,1.5

You can also specify relative coordinates as an expression within parentheses. It

has the effect of adding or subtracting from the absolute coordinates of a particular
place.
284 0 UNlX Text Processing 0

circle rad .5
circle same at last circle+(.25,0)

The same attribute allows us to duplicate the size of a previous object. The expres-
sion circle same means “the same size as the last circle.” This description pro-
duces:

Similarly, you can achieve finer control over positioning by moving from a compass
point:
box with . s w at last box.ne+(.05,-.05)

Expressions and User-Defined Variables

An expression can be used to supply the dimensions or the position of an object. Any
of the following operators can be used in an expression: +, -, *, /, and %
(modulo).” Expressions can be used to manipulate the built-in variables as follows:
circle rad circlerad/2

This will draw a circle with a radius that is half the size of the default radius. An
expression can also refer to the value of placenames. The coordinates of any object can
be specified as .x and .y. Here’s a list of some of the possibilities:
BoxA x The x-coordinate of the center of BoxA
last b0x.y The y-coordinate of the center of the last box
BoxA. s .y The y-coordinate of the southern compass point of BoxA
BoxA.wid The width of BoxA
last circle. rad The radius of the last circle
The next description defines a box and then divides the specified height and width of
that box to produce a second box half that size.
Boxa: box ht 2 wid 3; arrow
box ht Boxa.ht/2 wid Boxa.wid/2
The p i c program also has a number of functions that can be evaluated in an expres-
sion, as shown in Table 10-3:
 0 Drawing Pictures 0

TABLE 10-3. pic Functions

Function Description
s i n (a) Sine of a
cos ( a ) Cosine of a
a t a n 2 (a,h) Arctangent of alb
log ( a ) Natural logarithm of a
sqrt ( a ) Square root of a
i n t (a) Integer a
max ( a & ) Maximum value of a,b
m i n (a&) Minimum value of a,b
rand (a) Random number generator

In giving the size or length of an object, you can name and define your own vari-
ables. A variable is any lowercase name that is not reserved as part of the p i c
language. A variable can be defined as a constant or an expression.
a=ellipsewid*3
b=ellipseht/2
ellipse wid a ht b

This description produces:

Defining Macros
With macros, you can predefine a series of objects or motions that will be included in
the description each time you refer to the macro by name.
define name%
definition
YO

A percent sign (%) is used here as the delimiter but any character not in the definition
can be used. The format of the d e f i n e statement is shown on three lines for reada-
bility only; a simple macro could be put on a single line. The definition can extend
across as many lines as necessary.
When you refer to name in your description, The p i c program will replace it
with the definition.
Macros can also take arguments. These arguments are specified in the definition
as $1 thru $9. They will be replaced by the arguments supplied when the macro is
invoked.
286 0 UNlX Text Processing 0

name ( a r g l , arg2, arg3)

A macro does not exist as a place or position as far as pic is concerned. The pic
program simply replaces the macro name with the lines defined in the macro. You can-
not refer to the macro as you would refer to a block. However, you can set positions
from within a macro.
in the following example, the "tail" hanging down from the box and the list of
items drawn to the right of it were produced by a macro.

Alcuin Product Operation

Controller
Marketing
Engineering
Documentation
Quality Assurance
Customer Support

In the p i c description that produced this drawing, the box is drawn explicitly and a
short line is started down from the bottom of the box. Then a macro named dept is
invoked to produce each item on the list.
d e f i n e dept %
l i n e down - 2 5
{ l i n e right -15; move r i g h t - 2 ; "$1" ljust }
%

in this macro, after a line down is described, the rest of the description is put within
braces to reserve the starting position for the next macro call. A short line is drawn,
followed by a move to place the text in the correct position. Quotation marks are
placed around the argument because the argument will contain a text label.
This macro is invoked for the first item as:
d e p t (Control1 e r )

Controller is supplied as the first argument, which the macro inserts as a text
object. Notice that the argument in the definition is quoted ('l$1
'l) so that the actual
text when specified does not have to be quoted.
 o Drawing Pictures 0 287

The previous drawing was modeled after an example shown in Estimating Illus-
tration Costs: A Guide published by the Society for Technical Communication. The
guide considered this drawing to be of medium difficulty and estimated that it would
require an hour of an illustrator's time. It took ten to fifteen minutes to design and exe-
cute this description for pic, including correcting some syntax errors and formatting
for the laser printer. Here's the complete description of the drawing:
- PS
box ht -75 wid 1.75 "Alcuin Product Operation"
line down .25 from bottom of last box
define dept %
line down - 2 5
{ line right -15; move right - 2 ; "$1" ljust }
%
dept (Controller)
dept(Marketing)
dept (Engineering)
dept (Documentat ion )
dept(Qua1ity Assurance)
dept(Customer Support)
.PE
The second example of macro use is probably harder to read than it i s to write.
Let's look at it in portions. The purpose of the drawing is to represent a network of
computers. We decided to use three types of objects to represent each type of com-
puter: a square, a triangle, and small circle. These objects will appear in a hierarchy
and lines will be drawn to connect an object on one level with an object on the level
below it. Before starting to describe it in pic terms, we prepared a rough sketch of
the drawing on graph paper. This made us realize that we could easily determine the
coordinate points of objects; thus, all the macros are set up to accept coordinate posi-
tions.
Comments, beginning with # , describe the user-supplied arguments. Following
are the definitions for three macros: backbone (a box), local (a triangle), and
endpt (a small circle).
scale = 4
top = 10
define backbone %
# $ 1 = x coordinate ; $2 = label
ycoord = top-2
BB$1: box wid 1 ht 1 with .sw at $l,ycoord
"$2" at ($1,ycoord)+ (2,l) 1just
%
define local %
# $1 x coordinate; $2 = label
=
ycoord = top-5
LOS1: move to $l,ycoord
line down 1 left 1 then right 2 then up 1 left 1
288 0 UNlX Text Processing 0

"$2" at ($l,ycoord)-(0, -7)

%
define endpt %
# $1 = x c o o r d i n a t e
ycoord = t o p - 8
c i r c l e r a d -125 w i t h .n a t $ l , y c o o r d
EP$1: l a s t c i r c 1 e . n
z
Because each type of object maintained the same height (or position on the y-axis), a
variable ycoord was set up to supply that position from the t o p of the drawing.
(The top of the drawing i s defined by another variable.)
Each of these macros requires that you supply an x-axis coordinate as the first
argument. This argument is also used to assign a unique placename that is used later
when we draw lines between objects.
The b a c k b o n e and l o c a l macros also take a second argument for a label.
Handling text inside a macro definition is especially convenient if you are going to
change the font and point size.
The next task is to connect the backbone systems to the local systems and the
local systems to endpoints. Although we know which types of objects are connected,
not all objects are connected in the same way. We decided that the macros require two
arguments to supply the x-coordinate for each of the two objects.
d e f i n e BtoL %
# $1 = x c o o r d o f b a c k b o n e ; $ 2 = x coord o f
# local
l i n e f r o m B B $ l - ( O , . 5 ) t o LO$2
%
define LtoE %
# $1 = x c o o r d o f l o c a l ; $ 2 = x coord o f e n d p t
l i n e from L0$1-(0,1) t o EP$2
%

The BtoL and LtoE macros draw lines between the placenames set up by the
b a c k b o n e , l o c a l , and endpt macros.
Here are the actual macro calls:
b a c k b o n e ( 1 0 , IBM/370)
b a c k b o n e (18, DEC VAX)
l o c a l (8,68K-1)
local (13,68K-2)
l o c a l (17,68K-3)
endpt ( 7 )
endpt (9)
endpt (12)
endpt (13)
endpt ( 1 4 )
endpt (16)
endpt (18)
 0 Drawing Pictures 0 289

BtoL (10,8)
BtoL (10,13)
BtoL (18,17)
LtoE (8,7); LtoE (8,9)
LtoE (13,l.Z') ; LtoE (13,13); LtoE (13,14)
LtoE (17,16); LtoE (17,18)
line from LO13 to LO17
"\s8Personal Computers\sO" at 13,l
"\sl2\fBA Network of Computers\sO\fR" ljust at l0,top
Notice that arguments supplied to a macro are separated by commas and that an argu-
ment may contain a space. Here's what the description produces:

A Network of Computers
DEC VAX

Personal Computers

Twelve objects are specified and eleven lines are drawn between the objects. One line
is explicitly drawn connecting the second triangle to the third triangle. It didn't make
sense to define a macro for this single instance. But if you were setting this up for oth-
ers to use, such a macro would be necessary.
Shortly, we will be looking at several relatively new features that make pic
even more powerful for generating pictures. In particular, these features allow us to
improve our effort to generate a diagram of a computer network.

pic's Copy Facility

The pic program provides an interesting copy facility that has two different uses: it
allows you to read a pic description from a remote file, and it allows you to read
lines of data and pass them as individual arguments to a macro.
If you are going to use pic regularly, you should think about maintaining a
macro library. You might define frequently used objects, such as triangles, and place
290 UNlX Text Processing 0

them in their own file. You can include the file in your description with the following
line:
copy " u s r / l i b / r n a c r o s / p i c / t r i a n g l e s "

Putting the filename in double quotation marks is required. Any .PS/.P E macros that
are found in the remote file are ignored.
You might also define a set of related macros for a particular type of drawing,
such as an organizational chart or a flow diagram. After you have taken the time to
create and test a description, you should consider structuring it so that the forms can be
easily used in other drawings.
This copy facility replaces an older construct that allowed you to redirect input
from another file through the .P S macro.
. P S < triangles

A second use of the copy facility is to read data through a macro. We'll show
how the e n d p t macro from our last example can be designed to use this facility. In a
file where we had already defined a macro named e n d p t , we could invoke this macro
with the following command:
copy t h r u endpt
7
9
12
13
14
16
18
The p i c program reads each line of data up to the .P E and replaces each argument
in the macro definition with the corresponding field from each line. In this example,
the macro is executed seven times, once for each line of data.
We could put the data in a separate file, named e n d p t .d, for example. Then
you enter this version of the copy command:
copy "endpt - d " thru endpt

The double quotation marks are required. Now the e n d p t macro will be executed for
each line in the file e n d p t .d. (The filename suffix .d is optional and signifies that
the file contains data for a macro call.)
You can specify a string that p i c will recognize in the remote file as a signal to
stop reading input. Used with copy t h r u , u n t i l is followed by the string. In
the following example, the word STOP is used as the string:
copy "endpt-d" thru endpt until STOP

You can also use u n t i l when you are taking input from the same file:
copy thru local until STOP
8 68K-1
13 68K-2
 0 Drawing Pictures 0 291

17 68K-3
STOP

In both cases, p i c will read lines of data until it comes across the string STOP.
Another way to use copy t h r u is to supply the macro definition. This is a
compact, single-step method:
-
copy "endpt d" thru %
# $1 = x coordinate
ycoord = top-8
circle rad - 1 2 5 with .n at $l,ycooxd
EP$1: last circ1e.n
%

Although the percent sign is used as the delimiter, any character not found in the defini-
tion could be used. The copy t h r u statement with the macro definition can be put
on a single line, which is helpful for short definitions.
copy t h r u % box at $ 1 , $ 2 %
1 1
1 2
1 3
1 4

Because you can get a description down to this level, basically consisting of functions,
you could have a standard description file associated with independent data files. You
could write a program to build the data files from user input or from some other source.

Executing UNIX Commands

You can execute any UNIX command from p i c , using the following syntax:
sh % command %
Again, the percent sign represents any valid delimiter character. The p i c program
submits this command to the shell for execution and then returns to interpret the next
line of the description. You could issue a command to obtain data from another file:
sh % awk -F: {print$l} /etc/passwd %

picEnhancements
Most of the enhancements found in new versions of p i c are aimed at developing
p i c as a graphics programming language. Additional capabilities include f o r loops
and i f conditional statements. A f o r loop allows one or more p i c commands to
be executed as long as a condition is met.
292 0 UNlX Text Processing 0

f o r i=l t o 3 by .05
do %
box ht i;move
9i

Each time through the loop the value of the variable i is incremented by .05, produc-
ing five boxes of increasing height. The by clause specifies the amount that the van-
able is incremented each time through the loop. If the by clause is omitted, then the
variable is incremented by 1. The % is used as the delimiter marking the commands to
be executed on each pass.
The i f statement evaluates an expression to determine if it is true or false. If
true, t h e n specified pic commands are executed. If false, the t h e n clause is not
acted upon; instead, an e l s e clause, if specified, is read and commands specified
inside it are executed.
if x > y then % x = y % e l s e % x = x + 1%
This conditional statement evaluates the expression x > y. If true, x is set to y; if
false, the value of x is incremented by 1 . The % is a delimiter marking the beginning
and end of the commands specified for both t h e n and e l s e clauses. The expres-
sion inside an i f statement can use any of the relational operators that are shown in
Table 10-4.

TABLE 10-4. pic Relational Operators

Operator Meaning
__ Equal to
I = Not equal to
> Greater than
>= Greater than or equal to
< Less than
<= Less than or equal to
&& And
I Or
! Not

In addition to enhancements that add more graphics programming features to

pic, progress has been made in allowing input to be taken from bit-mapped graphic
terminals and translated into pic output. A separate program called cip, available
on some systems, allows users to create drawings using a mouse (a la MacDraw for the
Macintosh). The cip program generates a pic description of a drawing tnat can be
included in any file to be processed by t r o f f .
 C H A P T E R

11

A Miscellany of UNIX Commands

In this chapter, we present a miscellany of UNIX programs with text-processing appli-

cations. In addition, we introduce several UNIX utilities for communications and for
reading and writing to tapes and floppy disks. These utilities are not specifically
designed for text processing, but we have found them necessary for working in the
UNIX environment. Although you can find more detailed information on these utilities
in books aimed at a general audience, we’ve included brief discussions of them to
encourage you to learn them.
UNIX has many standard programs, as a run-down of the table of contents for the
UNIX Reference Manual will demonstrate. The challenge of UNIX is knowing which
programs are appropriate for a specific situation. No one learns all the commands, but
becoming familiar with a great number of them can be helpful. It is rather like those of
us who collect far more books on our shelves than are “needed,” knowing the reward
of finding the right book for the right occasion.
At times, you will be surprised when you discover a program with rather unusual
or specialized capabilities; at other times, you may be frustrated by a demanding pro-
gram or confused by inconsistencies from one program to the next. These qualities
seem to originate from the open design of UNIX, and serve to distinguish this text-
processing environment from the closed systems of most word processors.
In some ways, what we are trying to do in this chapter i s to address problems that
arise in typical documentation projects and show how one or more UNIX programs can
be applied as solutions to these problems. The emphasis is on the interactive use of
these programs, although many of them can be used effectively in shell scripts or as
parts of other programs. (In the next chapter, we go into more detail about shell
scripts.) The commands are presented in sections, grouped by function.

Managing Your Files

One of the realities of using a computer is that you begin to think of a document in
terms of files, rather than chapters or sections. You edit and print files; create and copy

293
294 0 UNIX Text Processing 0

files; delete files accidentally and lose your edits; and look through files to find the
information that is contained in them. Increasingly, files contain the goods that you
trade. You exchange not only printed copies of documents, but using floppy disks,
tapes, or modems, you take files off one system and put them on another system.
Learning to organize and maintain files is essential to working on a computer.

Using the File System to Your Advantage

One obvious feature of UNIX that makes it easy to handle large numbers of files is the
hierarchical file system. With carefully named files and directories, the pathname,
which specifies a file's unique place in the file system hierarchy, can tell a lot about not
only how to get at the file, but its contents as well.
For example, on our system, we keep all source files for various books in progress
on a file system called /work;work for a given client is kept in a directory named for
the client, with a subdirectory for each separate manual. Within each manual's sub-
directory, individual chapters are named consistently, chOl, ch02, and so on. As a
result, it is easy both to locate a file (Chapter 1 of the FORTRAN manual for ABC
Corp. can predictably be found in /work/abc/fortran/chOl) and to guess its
contents.
If you are using the C shell, you can create an alias that provides a shorthand
way of entering a command. In the following example, the alias allows you to think in
terms of manuals instead of directories:
% alias fortran "cd /work/abc/fortran; pwd"
% pwd
/work/fred
% fortran
/work/abc/fortran

You can place an alias definition in your cshrc file so that it becomes part of
your environment.
In the Bourne shell, you achieve a similar result by using an environment variable
called CDPATH to define a search path for the cd command. For example:
$ CDPATW=/work/abc:/work/textp:/usr
$ cd fortran
/work/abc/fortran
$ cd jane
/usr/ jane
$ cd ch03
/work/textp/ch03

When you issue a cd command, the shell searches for a subdirectory with that name
under any of the directories in the path, changes to it, and prints the full directory name.
The search directories in CDPATH are specified between colons. Directories
listed in CDPATH are searched in order from left to right.
 A Miscellany of UNIX Commands 0 295

Shell Filename Metacharacters

Even with files organized into directories, you can still accumulate a lot of files.
Developing some consistent naming conventions that take advantage of shell metachar-
acters (wildcards) can save you a lot of trouble. Most users are familiar with metachar-
acters but many don’t make full use of them.
In UNIX, you can match any part of a filename with a wildcard. Remember that
* matches zero or more characters. This gives you more power to select a specific
group of files out of a directory. In the following example, assume that you want to
delete the files lock, filelocks, and lock-release,but ignore the files
filelist, lecture, and stocks. c.
$ 1s
filelist
f ilelocks
lecture
lock
lock-release
stocks. c
$ rm *lock*

Because * can match zero characters, * l o c k * will match lock as well as

filelocks.
The shell interprets the pattern-matching character ? to match any single character,
and the construct [m-n] to match a range of consecutive characters.
If you name your files consistently, you can use these characters to select groups
of files. For example, in a directory containing a BASIC manual, you might have the
following list of files:
$ 1s
aPPa
aPPb
changes
chOl
ch0l.old
ch02
ch03
ch03.example.s
ch03. out
ch04
ch04.example.s
ch05
letter.613

As usual in any directory, there are a number of auxiliary files. Some of these
files apply to the work on this project, but they are not actually part of the book. If
you’ve carefully chosen the names of related files, you can use metacharacters to select
only the files in a particular group. For example:
296 0 UNIX Text Processing 0

$ 1s chO?
chOl
ch02
ch03
ch04
ch05
You could select a range of files, using brackets:
$ IS ch0[3-5]
ch03
ch04
ch05
If you had entered chO *, miscellaneous files such as chO 1. old would have been
included. (Note that whenever you use numbers in filenames, as shown here, to con-
sistently name a group of related files, you should begin the numbering sequence with
01, 02 . . . rather than 1 , 2. . . . This will cause 1 s to list the files in proper alphabet-
ical order. Otherwise, Is will list chl, then c h l l , ch12 . . . ch2, ch20 . . . and so on.)
Metacharacters have broader applications than for simply listing files. Look at
this example of running spe 11on an entire book:
$ spell chO? app? > spell.out

(We?ll be looking at the s p e l l command later in the section ?Proofing Docu-

ments.?) This command is run on the seven files that match one of the two patterns
specified on the command line.
Metacharacters are also useful in moving and copying files from one directory to
another:
$ cp basic/chO? /work/backup

Locating Files
Although a hierarchical file system with consistent naming conventions helps a lot, it is
still easy to lose track of files, or just to have difficulty specifying the ones you want to
manipulate. The number of files contained on even a small hard disk can be enormous,
and complex directory hierarchies can be difficult to work with.
It is possible to lose a file on the file system when you have forgotten in which
directory you put it. To look through an entire file system or a large directory hierar-
chy, you need a utility called f i n d . The f i n d utility looks at the external charac-
teristics of a file-who created it, when it was last accessed, its name, and so on.
The f i n d utility probably wins top honors for having the most cumbersome
command-line syntax in UNIX. It?s not that f i n d i s a difficult command; its syntax
is simply difficult to recall. You might expect that all you have to enter is f i n d and
the name of the file that you want to look for. This is not the way it works, however,
which is a nuisance to new users. The f i n d command requires repeated trips to the
UNIX Reference Manual before you grasp its atypical format.
 o A Miscellany of UNlX Commands 0 297

To use find, specify the pathnames of the directories that you want to search;
then place one or more conditions upon the search. The name of a particular file that
you want to search for is considered one of these conditions. It is expressed as:

-name filename

To obtain a listing of the pathnames of files that are found, you have to specify the
- p r i n t condition as well (-name must precede - p r i n t ) .
If you wanted to find any file named notes on the / w o r k file system, here’s
the command to enter:
$ find /work -name notes -print
/work/alcuin/notes
/work/textp/ch02/notes

The output is the pathname (starting with the specified file system or directory) of each
file that is found. More than one pathname can be supplied. A slash (/) represents the
root directory and thus is used if you want to search the entire file system. Note that
the search can take some time, and that if you do not have read permissions to a direc-
tory you will get a message saying that it cannot be opened.
In the next example, we add another condition, - u s e r , and limit the search to
files named m e m o that are owned by the user fred. This is helpful when you are
searching for a file that has a fairly common name and might exist in several users’
accounts. Filename metacharacters can be used but they must be protected from the
shell using backslashes or single quotation marks. (If you don’t do this, the metachar-
acters will be interpreted by the shell as referring to files in the current directory, and
will not be passed to the f i n d command.)
$ find /work /usr -name ’memo*’ --user fred -print
/usr/fred/alcuin/memo
/work/alcuin/memo.523
/work/caslon/memo.214

Two directory hierarchies are searched, /work and / u s r . If you did not specify the
-name condition, this command would locate all the files owned by fred in these
two file systems.
Many f i n d conditions have uses for other tasks besides locating files. For
instance, it can be useful to descend a directory hierarchy, using f i n d to print the
complete pathname of each file, as in the following example:
$ find /work/alcuin -print
/work/alcuin
/work/alcuin/chOl
/work/alcuin/chOl.old
/work/alcuin/commands/open
/work/alcuin/commands/stop
...
This usage provides a kind of super Is that will list all files under a given directory,
not just those at the current directory level. As you’ll see, this becomes very useful
when it comes time to back up your files.
298 0 UNIX Text Processing 0

The longer you work with a UNIX system, the more you will come to appreciate
find. Don’t be put off by its awkward syntax and many options. The time you spend
studying this command will be well repaid.

File Characteristics
Most of us are concerned only with the contents of a file. However, to look at files
from UNIX’s point of view, files are labeled containers that are retrieved from storage
and soon put back in the same place. It might be said that the operating system reads
(and writes) the label but doesn’t really care to look inside the container. The label
describes a set of physical or external characteristics for each file. This information is
displayed when the 1 s command produces a long listing.
$ 1s -1 /work/textp/chOl
total 2 0
-rw-rw-r-- 1 fred doc 9 4 9 6 Jun 10 15:18 chOl
To the operating system, the file ( c h O l ) contains a certain number of bytes (9496),
each representing a character. The date and time ( J u n 1 0 1 5 : 1 8 ) refer to the last
time the file was modified. The file has an owner ( f r e d ) , who is usually the person
who created the file. The owner belongs to a group of users (doc) who can be given
different permissions from all other users. The operating system keeps track of the file
permissions (-rw-rw-r--) for the owner, group, and other users-determining who
can read, write, or execute the file.
All of these characteristics can be modified either by use of the file or by com-
mands such c h m o d (change permissions) and c h o w n (change owner). You may
need to become a super-user to change these characteristics.
There are some options for 1s that allow you to make use of this information.
For instance, if you had recently made some changes to a set of files, but couldn’t
remember which ones, you could use the -t option to sort a list of files with the most
recently modified files first. The -r option reverses that order, so that 1s -rt pro-
duces a list with the oldest files first.
In addition, f i n d has a number of options that make use of external file charac-
teristics. As we’ve seen, you can look for files that belong to a particular user. You
can also look for files that are larger than a particular size, or have been modified more
recently than a certain date.
Don’t get stuck thinking that the only handle you can pick a file up with is the
file’s name.

Viewing the Contents of a File

You are probably familiar with a number of UNIX commands that let you view the con-
tents of a file. The c a t command streams a file to the screen at a rate that is usually
too swift. The pg and m o r e commands display a file one page at a time. They are
frequently used asfilters, for instance, to supply paging for n r o f f output.
 A Miscellany of UNIX Commands 0 299

$ nroff -mm chOl I pg

You can also use these commands to examine unformatted files, proofing formatting
codes as well as text. Although these are frequently used commands, not everyone is
aware that they have interactive subcommands, too. You can search for a pattern; exe-
cute a UNIX command; move to another file specified on the command line; or go to
the end of the file.
You can list these subcommands by entering h when the program pauses at the
bottom of a page. Here's the help screen pg provides:
______________--________________________--_--_____---__
h help
q or Q quit
< b l a n k > o r \n n e x t page
1 next l i n e
d o r *D d i s p l a y h a l f a p a g e more
- o r "L r e d i s p l a y current page
s k i p t h e next page forward
next f i l e
previous f i l e
l a s t page
w or z s e t window s i z e and d i s p l a y n e x t page
s savefile save current f i l e i n s a v e f i l e
/pattern/ s e a r c h f o r w a r d for p a t t e r n
? p a t t e r n ? or
*pattern" s e a r c h b a c k w a r d for p a t t e r n
! command e x e c u t e command
M o s t commands c a n be p r e c e d e d b y a number, a s i n :
+ l \ n ( n e x t p a g e ) ; -l\n ( p r e v i o u s p a g e ) ; l \ n (page 1 ) .
S e e t h e manual p a g e f o r more d e t a i l .
____________________---_-___-___----__--______----_--__
One advantage of p g is that you can move backward as well as forward when going
through a file. A special feature of more is the ability to invoke v i at the current
point in the file. When you quit v i , more resumes paging through the rest of the
file.
Another command used for examining a file is pr. Its most common use is to
perform minor page formatting for a file on the way to a line printer. It breaks the
input file into pages (66 lines to a page) and supplies a header that contains the date, the
name of the file, and the current page number. Top, bottom, and side margins are also
added.
The pr command also has many options that can be used to perform some odd-
ball tasks. For example, the -n option adds line numbers:
$ pr -n test

The following is displayed:

300 0 UNlX Text Processing 0

Jul 4 14:27 1987 test Page 1

1 apples
2 oranges
3 walnuts
4 chestnuts

You can adjust the page length using the -1 option. If you are printing to a terminal,
the -p option specifies a pause at the beginning of each page. You can also display an
input file in -n columns.
The -m option simultaneously merges two or more files and prints each of them,
one per column:
$ pr -m -t t e s t *
In this example, we display four files side-by-side:
apples apples apples oranges
oranges oranges oranges walnuts
walnuts walnuts grapes chestnuts
chestnuts

The t e s t * file specification i s expanded to four filenames: t e s t , t e s t 1,

t e s t 2 , and t e s t 3 . The -t option suppresses the heading and does not print
linefeeds to fill a page, which is especially useful when you are sending the output of
pr to a file or the terminal.
We found a use for pr when working on this book. We wanted to include
n r o f f -formatted examples in the text. We had difficulty because n r o f f inserts
tabs, instead of spaces, to optimize horizontal positioning on printers. To remove the
tabs, we used pr with the -e option to expand the tabs to their equivalent in blank
spaces. The following shell script implements this process so that it can be invoked as
a single command:
$ nroff -mm -roo examples/$l I pr -e -t
The p r command works as a filter for n r o f f . The -r option is used with n r o f f
to set register 0 (page offset or left margin) to zero.
Sometimes it can be useful to examine just the beginning or the end of a file.
Two commands, head and tail, print the first or last ten lines of a file. The
h e a d command can be used to look at the initial settings of number registers and
strings that are often set at the top of a file.
$ head ch02
.nr W 65
.nr P 3
.nr L 60
. s o /usr/lib/tmac/tmac .m
.nr Pt 2
.ds U X \ s - ~ U N I X \ S O
.ds RP 3321
I
0 A Miscellany of UNlX Commands 0 301

.H1 "Product Overview"

.ds HM 1 1 A

This output could be redirected to a file as a way of starting a new chapter. The t a i l
command has the same syntax; it can save time when you want to check the end of a
large file.

. Searching for Information in a File

The many benefits provided by g r e p to the user who doesn't remember what his or
her files contain are well known. Even users of non-UNIX systems who make fun of
its obscure name wish they had a utility with its power to search through a set of files
for an arbitrary text pattern, known as a regular expression. We have already discussed
regular expressions and their use in search and replace commands in v i (see Chapter
7). In this section, we show some of the ways to perform pattern-matching searches
using g r e p and its siblings, e g r e p and f grep.
The main function of grep i s to look for strings matching a regular expression
and print only those lines that are found. Use g r e p when you want to look at how a
particular word is used in one or more files.
$ grep "run[-]time" ch04
This procedure avoids run-time errors for not-assigned
and a run-time error message is produced.
run-time error message is produced.
program aborts and a run-time error message is produced.
DIMENSION statement in BASIC is executable at run time.
This means that arrays can be redimensioned at run time.
accessible or not open, the program aborts and a run-time
This example lists the lines in the file ch04 that contain either r u n - t i m e or r u n
time.
Another common use is to look for a specific macro in a file. In a file coded with
mm macros, the following command will list top-level and second-level headings:
$ grep ""\ .H[12] chO [123
I'

chOl: .H1 "Introduction"

chOl: .H1 "Windows, Screens, and Images"
chOl: .H2 "The Standard Screen-stdscr"
chOl : .H2 "Adding Characters"
...
ch02: .HI "Introduction"
ch02: .H1 "What Is Terminal Independence?"
ch02 : - H2 "Termcap"
ch02 : .H2 "Terminfo"
In effect, it produces a quick outline of the contents of these files. When more than one
file is specified, the name of the file appears with each line. Note that we use brackets
302 UNlX Text Processing 0

as metacharacters both in the regular expression and when specifying the filename.
Because metacharacters (and spaces) have meaning to the shell, they will be interpreted
as such unless the regular expression is placed within quotation marks.
There are several options commonly used with grep. The -i option specifies
that the search ignore the distinction between uppercase and lowercase. The -c option
tells g r e p to return only a count of the number of lines matched. The -1 option
returns only the name of the file when g r e p finds a match. This can be used to
prepare a list of files for another command.
The shell construct commandl 'command2' causes the output of command2 to be
used as an argument to commandl. For example, assume that you wanted to edit any
file that has a reference to a function call named g e t c u r s o r . The command:
$ vi 'grep -1 getcursor *'
would invoke v i on all of the files in the current directory containing the string
g e t c u r s o r . Because the g r e p command is enclosed in single backquotes (' '), its
output becomes the list of files to be edited.
The grep command can work on the results of a find command. You can use
f i n d to supply a list of filenames and grep to search for a pattern in those files. For
example, consider the following command, which uses f i n d to look for all files in the
specified directory hierarchy and passes the resulting names to grep to scan for a par-
ticular pattern:
$ find /work/docbook -exec grep " [aA]lcuin" { } \;
Alcuin product. Yesterday, I received the product demo
Alcuin. Some people around here, looking over my shoulder,
with Alcuin. One person, a student of calligraphy,
presents different strategies for documenting the Alcuin
The development of Alcuin can be traced to our founder's
the installation file "alcuin.install"> and the font
configuration file "alcuin.ftables."
The -exec condition allows you to specify a command that is executed upon each file
that is found ( { } indicates the pathname of the file). The command must end with an
escaped semicolon.
Although this is a good way to introduce the very useful -exec option to
f i n d , it is actually not the best way to solve the problem. You'll notice that even
though g r e p is working on more than one file, the filenames are not printed because
the data is actually passed to grep from a pipe. The reason is that grep is being
invoked many times (once for each file that is found), and is not really working on
many files at once. If you wanted to produce a list of the selected files, you could use
the -1 option with grep. But more to the point, this is a very inefficient way to do
the job.
In this case, it would be preferable to write:
$ grep " [aA]lcuin" .find /work/docbook -print'
Because grep is invoked only once, this command will run much faster.
 A Miscellany of UNlX Commands 0 303

There is a potential danger in this approach. If the list of files is long, you may
exceed the total allowable length of a command line. The best approach uses a com-
mand we haven't shown yet-xargs. This command provides an extended version of
the same function the shell provides with backquotes. It converts its input into a form
that can be used as an argument list by another command. The command to which the
argument list is passed is specified as the first argument to x a r g s . So, you would
write:
$ f i n d /work/docbook -print I x a r g s grep " [ a A ] l c u i n "
Or you could generalize this useful tool and save it as the following shell script, which
could be called m f g r e p (multifile grep). This script takes the pathname for f i n d
as the first argument and the pattern for g r e p as the second. The list of files found is
passed to g r e p by x a r g s :
f i n d $1 1 x a r g s grep "$2"

The f g r e p (fast grep)" command performs the same function as g r e p , except it

searches for a fixed string rather than a regular expression. Because it doesn't interpret
metacharacters, it often does a search faster than g r e p . For interactive use, you may
not find enough difference to keep this command in your active repertoire. However, it
may be of more benefit inside shell scripts.
The e g r e p command is yet another version of g r e p , one that extends the syn-
tax of regular expressions. A + following a regular expression matches one or more
occurrences of the regular expression; a ? matches zero or one occurrences. In addi-
tion, regular expressions can be nested within parentheses.
$ egrep "Lab (oratorie) ? s " name. l i s t
AT&T Bell L a b o r a t o r i e s
AT&T B e l l L a b s

Parentheses surround a second regular expression and ? modifies this expression. The
nesting helps to eliminate unwanted matches; for instance, the word Labors or oratories
would not be matched.
Another special feature of e g r e p is the vertical bar (I), which serves as an or
operator between two expressions. Lines matching either expression are printed, as in
the next example:
$ egrep "stdscr I curscr" ch03
i n t o t h e stdscr, a c h a r a c t e r a r r a y .
When s t d s c r i s r e f r e s h e d , t h e
stdscr i s refreshed.
curscr.
i n i t s c r ( ) c r e a t e s t w o windows: stdscr
and curscr.

*Despite what the documentation says, egrep is usually the fastest of the three g r e p programs.
304 UNIX Text Processing 0

Remember to put the expression inside quotation marks to protect the vertical bar from
being interpreted by the shell as a pipe symbol. Look at the next example:
$ egrep "Alcuin (User I Programmer) ( ' s) ? Guide" docguide
Alcuin Programmer's Guide is a thorough
refer t o the Alcuin User Guide.
Alcuin User's Guide introduces new users to

You can see the flexibility that egrep's syntax can give you, matching either User or
Programmer and matching them if they had an ' s or not.
Both e g r e p and f g r e p can read search patterns from a file using the -f
option.

Proofing Documents
There are no computer tools that completely replace the close examination of final
printed copy by the human eye. However, UNIX does include a number of proofing
aids, ranging from a simple spelling checker to programs for checking style and diction,
and even sexist usage.
We'll look at some of these programs in this section. Not all of the programs
we'll discuss are available on all UNIX systems. Keep in mind, though, that g r e p is
also a very powerful proofing aid, which you can use to check for consistent usage of
words and phrases.

Looking for Spelling Errors

The spe 1 1 command reads one or more files and prints a list of words that are possi-
bly misspelled. You can redirect the output to a file, then use g r e p to locate each of
the words, and v i or ex to make the edits. In the next chapter, though, we introduce
a shell script named p r o o f for running s p e l l interactively and correcting spelling
errors in place in a file. You will probably prefer to use s p e l l in that manner rather
than invoking it manually.
Even if you do build that script, you can use s p e l l on its own if you are
unsure about which of two possible spellings is right. Type the name of the command,
followed by a RETURN, then type the alternative spellings you are considering. Press
*D (on a line by itself) to end the list. The s p e l l command will echo back the
word(s) in the list that it considers to be in error.
$ spell
misspelling
mispelling
"D
mispelling

You can invoke s p e l l in this way from within v i , by typing the ex colon prompt,
an exclamation point, and the name of the s p e l l command.
 0 A Miscellany of UNIX Commands 0 305

When you run spell on a file, the list of words it produces usually includes a
number of legitimate words or terms that the program does not recognize. You must
cull out the proper nouns and other words spell doesn't know about to arrive at a
list of true misspellings. For instance, look at the results on this sample sentence:
$ cat sample
Alcuin uses Transcript to convert d i t r o f f i n t o
PostScript output for t h e Laserwriter printerr.
$ spell sample
Alcuin
ditroff
printerr
Laserwriter
PostScript
Transcript

Only one word in this list is actually misspelled.

On many UNIX systems, you can supply a local dictionary file so that spell
recognizes special words and terms specific to your site or application. After you have
run spell and looked through the word list, you can create a file containing the
words that were not actual misspellings. The spell command will check this list
after it has gone through its own dictionary.
If you added the special terms in a file named d i c t , you could specify that file
on the command line using the + option:
$ spell +dict sample
printerr

The output is reduced to the single misspelling.

The spell command will also miss words specified as arguments to nroff
or t roff macros, and, like any spelling checker, will make some errors based on
incorrect derivation of spellings from the root words contained in its dictionary. If you
understand how spell works, you may be less surprised by some of these errors.
The directory /usr/lib/spell contains the main program invoked by the
spe11 command along with auxiliary programs and data files.
$ 1s -1 /usr/lib/spell
t o t a l 604
-rwxr-xr-x 1 bin bin 20176 Mar 9 1985 h a s h c h e c k
-rwxr-xr-x 1 bin bin 14352 Mar 9 1985 h a s hmake
-J-w-r--r-- 1 bin bin 53872 Mar 9 1985 h l i s t a
-r"-r--r-- 1 bin bin 53840 Mar 9 1985 h l i s t b
-r--r--r-- 1 bin bin 6328 Mar 9 1985 h s t o p
-rw-rw-rw- 1 root root 102892 J u l 12 16:lO s p e l l h i s t
-rwxr-xr-x 1 bin bin 23498 Mar 9 1985 s p e l l i n
-rwxr-xr-x 1 bin bin 27064 Mar 9 1985 s p e l l p r o g
The spell command pipes its input through deroff -w and sort -u to
remove formatting codes and prepare a sorted word list, one word per line. (The
deroff and sort commands are discussed later in this chapter.) Two separate
306 0 UNlX Text Processing n

spelling lists are maintained, one for American usage and one for British usage (invoked
with the -b option to spell). These lists, hlista and hlistb, cannot be read
or updated directly. They are compressed files, compiled from a list of words
represented as nine-digit hash codes. (Hash-coding is a special technique for quick
search of information.)
The main program invoked by spell is spellprog. It loads the list of hash
codes from either hlista or hlistb into a table, and looks for the hash code
corresponding to each word on the sorted word list. This eliminates all words (or hash
codes) actually found in the spelling list. For the remaining words, spellprog tries
to see if it can derive a recognizable word by performing various operations on the
word stem, based on suffix and prefix rules. A few of these manipulations follow:

-y+iness
+ness
-y+i+less
+less
-y+ies
-t+ce
-t+cy

The new words created as a result of these manipulations will be checked once more
against the spell table. However, before the stem-derivative rules are applied, the
remaining words are checked against a table of hash codes built from the file hstop.
The stop list contains typical misspellings that stem-derivative operations might
allow to pass. For instance, the misspelled word thier would be converted into thy
using the suffix rule -y+ier. The hstop file accounts for as many cases of this type
of error as possible.
The final output consists of words not found in the spell list, even after the pro-
gram tried to search for their stems, and words that were found in the stop list.
You can get a better sense of these rules in action by using the -v or -x option.
The -v option eliminates the last lookup in the table, and produces a list of
words that are not actually in the spelling list along with possible derivatives. It allows
you to see which words were found as a result of stem-derivative operations, and prints
the rule used.
$ spell -v sample
Alcuin
ditroff
Laserwriter
PostScript
printerr
Transcript
+out output
+s uses

The -x option makes spell begin at the stem-derivative stage, and prints the vari-
ous attempts it makes to find the word stem of each word.
 0 A Miscellany of UNlX Commands 0 307

$ s p e l l --x sample
___
=into
=Laserwriter
=Laserwrite
=Laserwrit
=laserwriter
=laserwrite
=laserwrit
=output
=put
...
Laserwriter
...
The stem is preceded by an equals sign. At the end of the output are the words whose
stem does not appear in the spell list.
One other file you should know about is spellhist. Each time you run
spell, the output is appended through a command called t e e into spellhist, in
effect creating a list of all the misspelled or unrecognized words for your site. The
spellhist file is something of a “garbage” file that keeps on growing. You will
want to reduce it or remove it periodically. To extract useful information from this
spellhist, you might use the sort and uniq -c commands shown later in this
chapter to compile a list of misspelled words or special terms that occur most fre-
quently. It is possible to add these words back into the basic spelling dictionary, but
this is too complex a process to describe here.

Checking Hyphenation
The hyphen command is used on nroff-formatted files to print a list of words that
have been hyphenated at the end of a line. You can check that nroff has correctly
hyphenated words.
$ hyphen ch03. out
ch03. out :
applica-tion
p r o -gr am
charac-ter

If you disagree with the hyphenation of a word, you can go back into your source file
and use either the .hw request to specify hyphenation points or the .nh request to
inhibit hyphenation of the word.
If you don’t have the hyphen command on your system, you can print the lines
ending in hyphens using grep:
$ grep ’-$’ ch03.out

This will not display the second half of the hyphenated word on the following line, but
it should give you enough of an idea. Alternatively, you could use awk or sed,
308 0 UNIX Text Processing 0

described in the next chapter, to create a version of this command that would print both
lines.

Counting Words
In the past, writers were paid by the word. The w c command will count words for
you:
$ wc chOl
180 1529 9496 chOl

The three numbers printed represent the number of lines, words, and characters, respec-
tively. (The presence of formatting commands in the input file will make this measure-
ment somewhat inaccurate.)

Writer’s Workbench
No book on UNIX text processing can avoid some discussion of Writer’s Workbench
(WWB), a collection of programs for the analysis of writing style.
Unfortunately, unlike most of the programs described in this book, the Writer’s
Workbench is not available on all UNIX systems. It was originally developed for inter-
nal use at Bell Labs, and was available in early releases of UNIX to the academic com-
munity. But it was made into a separate product when UNJX was commercially
released.
The three original programs, s t y l e , d i c t i o n , and e x p l a i n , are available
in Berkeley UNIX systems and in Xenix, but not in System V.
AT&T has released a greatly improved and expanded version, including additional
programs for proofreading, that is controlled from a master program called w w b . How-
ever, this version is only available as a separately priced package for 3B2 and 3B5 com-
puters. The unfortunate result is that one of UNIX’s most unusual contributions to text
processing is not officially part of UNIX and has never been ported to many UNIX sys-
tems.
In this section, we’ll describe the original s t y l e and d i c t i o n programs,
with a brief discussion of w w b .
The s t y l e program analyzes a document’s style and computes readability
indexes based on several algorithms widely accepted in the academic community. For
example, when run on a draft of this section, s t y l e gave the following report:
readability grades:
(Kincaid) 11.1 (auto) 11.6 (Coleman-Liau) 1 1 . 0
(Flesch) 11.5 (52.7)
sentence info:
no. sent 53 no. wds 1110
av sent leng 20.9 av word leng 4.79
no. questions 0 no. imperatives 0
no. nonfunc w d s 624 56.2% av leng 6.25
short sent (<16) 3 4 % (18) long sent ( > 3 1 ) 17% (9)
longest sent 46 wds at sent 4;
 A Miscellany of UNlX Commands 0 309

shortest sent 5 wds at sent 47

sentence types :
simple 32% (17) complex 47% (25)
compound 4 % ( 2 ) compound-complex 17% (9)
word usage:
verb types as % of total verbs
tobe 29% (33) aux 28% (32) inf 15% ( 1 7 )
passives as % of non-inf verbs 9% (9)
types as % of total
prep 12.0% (133) conj 3.6% ( 4 0 ) adv 5.0% (56)
noun 26.8% (298) adj 15.5% (172) pron 7.3% (81)
nominalizations 3 % (30)
sentence beginnings:
subject opener: noun (22) pron (5) pos (1) adj ( 2 )
art ( 4 ) tot 64%
prep 17% ( 9 ) adv 9% (5)
verb 0 % ( 0 ) s u b-conj 6% (3) conj 0% ( 0 )
expletives 4% (2)

Even if you aren’t an English teacher and don’t know the Kincaid algorithm from the
Flesch, this report can be very useful.
First, regardless of the differences between the algorithms, they all give you a
general idea of the required reading level for what you have written. It is up to you to
adjust your style according to the audience level you want to reach. This may not be a
trivial task; however, it may be a vital one if you are writing a book for a specific audi-
ence. For example, if you were writing an instruction manual for heavy equipment to
be used by people reading at the sixth-grade level, a s t y 1e report like the one shown
would be a dire warning that the manual would not be successful.
In general, to lower the reading level of a document, use shorter sentences and
simpler constructions. (Incidentally, most writing in newspapers and general circulation
magazines is at the sixth-grade level. But you shouldn’t get the impression that text
written for a lower reading level is better. Writing can be clear and effective at any
level of complexity. At the same time, each of us must recognize, and adjust for, the
skills of our intended reader.)
The analysis of reading level is only a small part of what s t y l e offers. The
detailed analysis of sentence length and type, word usage, and sentence beginnings can
give you considerable insight into your writing. If you take the time to read the report
carefully at the same time as you reread your text, you will begin to see patterns and
can make intelligent decisions about editorial changes.
As an exercise, run s t y l e on a short passage you have written, read the report
carefully, then rewrite your work based on the report. See what difference this makes
to the s t y l e report. You will eventually get a feel for what the program provides.
In some cases, d i c t i o n , the other major program in the Writer’s Workbench,
can also help you find areas to change.
The d i c t i o n program relies on a library of frequently misused words and
phrases. It relentlessly searches out these words and flags them as inappropriate by
enclosing them in brackets. For example, when run on a previous draft of this section,
d i c t i o n made the following recommendations:
31 0 0 UNlX Text Processing 0

wwb
style performs stylistic analysis of a document and
computes readability indexes based on a[ number of ]
algorithms widely accepted in the academic community.

this may not be a trivial task however it may be a

[ vital ] one if you are writing a book with a specific
target audience.

for example if you were writing an instruction manual

for heavy equipment to be used by people reading at the
sixth grade level a style report like the one shown above
would be a dire warning that the manual would not be
[ very ]successful.

[ in some cases ] diction the other major program in the

writer s workbench can help you to find possible areas to
change.

in the latest official release of wwb there are a

[ number of ] additional programs including .

morestyle which looks for abstract words as well as

listing the frequency with which each word is used
and the word diversity the[ number of ]different words
divided by the total[ number of 3 words .

morestyle also gives a count of the[ number of ]negative

constructions contained in your writing.

spellwwb which lists possible spelling errors in a

slightly more usable format than the standard spell
program and spelladd which allows you to build a local
dictionary word of spelling exceptions words that spell
regards as errors but[ which ]you know to be correct -

you can run these programs individually or using one of

several [ overall ]control programs.

running wwb will run[ all of ]these programs.

number of sentences 37 number of hits 10

The d i c t i o n program lists “problem” sentences from your source file, with words
or phrases it has taken exception to enclosed in brackets. You can redirect this output
 0 A Miscellany of UNfX Commands 0 311

to a file, or page through it on the screen. Punctuation and macros are first stripped by
the d e r o f f program, which explains the odd appearance of the text.
We find that we ignore diction's advice much of the time-the exception list
is applied across the board, without regard for context. For example, you'll notice that
it flagged the phrase number of several times, though that was exactly what we meant in
all but one case. However, the twenty percent of its recommendations that we agree
with are worth the effort of running the program.
If you don't understand why d i c t i o n complains about a phrase, you can use
e x p l a i n to ask for help. For example:
$ explain
phrase?
which
use "that" when clause is restrictive" for '*which"
use "when" for "at which time"
phrase?
number of
use "many" for "a large number of"
use "several, many, some" for "a number of"
use "usually" for "except in a small number of cases"
use "some" for "in a number of cases"
use "enough" for "sufficient number of"
use "often" for "in a considerable number of cases''
phrase?
perform
use "do" for "perform"
use "measure" for "perform a measurement"
phrase?
AD
The official release of WWB for 3B computers contains improved versions of
s t y l e and d i c t i o n , as well as many additional programs. These programs include

abst, which evaluates the abstractness of your writing.

acro, which looks for acronyms (any word printed in all capital letters) so
you can check that they have been properly defined.
dictadd, which allows you to add to the dictionaries used by d i c t i o n ,
s p e l l , and s e x i s t .
double, which looks for double words.
f indbe, which looks for syntax that may be difficult to understand.
morestyle, which looks for abstract words and lists the frequency with
which each word is used and the word diversity (the number of different words
divided by the total number of words). The m o r e s t y l e program also gives
a count of the number of negative constructions contained in your writing.
I
31 2 UNIX Text Processing 0

I o r g , which prints the first and last sentence of each paragraph, so you can
analyze paragraph transitions and the flow of ideas within your writing.
I punct, which checks punctuation (e.g., the placement of commas and periods
with quotation marks).
I sexist, which checks your writing against a dictionary of sexist words and
phrases.
I spellwwb, which lists possible spelling errors in a slightly more usable for-
mat than the standard spell program, and spelladd, which allows you to
build a local dictionary of spelling exceptions (words that spell regards as
errors, but that you know to be correct).
I split rules, which finds split infinitives.
I syl, which prints the average number of syllables in the words you use.

You can run these programs individually or use one of several control programs. The
wwb program will run just about everything. The proof r program will run those
programs that help you proofread (such as spell, double, punct, and dic-
tion). The prose program will run those that analyze style (such as style and
sexist).
There is also an interactive version of proof r called proofvi, which stores
its output in a temporary file and then allows you to edit your original, stepping through
each flagged problem.

Comparing Versions of the Same Document

UNIX provides a number of useful programs for keeping track of different versions of
documents contained in two or more files:

the dif f family of programs, which print out lines that are different between
two or more files
the SCCS system, which lets you keep a compact history of differences
between files, so that you can go back and reconstruct any previous version
the make program, which keeps track of a predefined list of dependencies
between files

Checking Differences
The dif f command displays different versions of lines that are found when compar-
ing two files. It prints a message that uses ed-like notation (a for append, c for
change, and d for delete) to describe how a set of lines has changed. This is followed
by the lines themselves. The < character precedes lines from the first file and > pre-
cedes lines from the second file.
 A Miscellany of UNIX Commands 0 313

Let’s create an example to explain the output produced by d i f f. Look at the

contents of three sample files:

TEST1 TEST2 TEST3

apples apples oranges

oranges oranges walnuts
walnuts grapes chestnuts

When you run d i f f on these files, the following output is produced:

$ diff t e s t l t e s t 2
3c3
< walnuts
___
> grapes

The d i f f command displays the only line that differs between the two files. To
understand the report, remember that d i f f is prescriptive, describing what changes
need to made to the first file to make it the same as the second file. This report speci-
fies that only the third line is affected, exchanging walnuts for grapes. This is more
apparent if you use the -e option, which produces an editing script that can be submit-
ted to ed, the UNIX line editor. (You must redirect standard output to capture this
script in a file.)
$ diff -e t e s t l t e s t 2
3c
grapes

This script, if run on t e s t 1 , will bring t e s t 1 into agreement with t e s t 2. (Later

in this section, we’ll look at how to get e d to execute this script.) If you compare the
first and third files, you find more differences:
$ diff t e s t l t e s t 3
Id0
< apples
3a3
> chestnuts

To make t e s t 1 the same as t e s t 3, you’d have to delete the first line (apples) and
append the third line from t e s t 3 after the third line in t e s t l . Again, this can be
seen more clearly in the editing script produced by the -e option. Notice that the
script specifies editing lines in reverse order; otherwise, changing the first line would
alter all succeeding line numbers.
314 0 UNlX Text Processing 0

$ d i f f -e t e s t l test3
3a
chestnuts

Id

You can use the di f f 3 command to look at differences between three files.
For each set of differences, it displays a row of equals signs (====) followed by 1 , 2, or
3, indicating which file is different; if no number is specified, then all three files differ.
Then, using ed-like notation, the differences are described for each file.
$ d i f f 3 t e s t l test2 test3
____3
1 : IC
2:lc
apples
3:Oa
__--
___- 3
1:3c
2:3c
grapes
3:2,3c
walnuts
chestnuts

With the output of di f f 3 , it is easy to keep track of which file is which; however, the
prescription given is a little harder to decipher. To bring these files into agreement, you
would have to add apples at the beginning of the third file; change line 3 of the second
file to line 3 of the first file; and change lines 2 and 3 of the third file, effectively drop-
ping the last line.
The di f f 3 command also has a -e option for creating an editing script for
ed. It doesn’t quite work the way you might think. Basically, it creates a script for
building the first file from the second and third files.
$ d i f f 3 -e t e s t l test2 test3
3c
walnuts
chestnuts

IC

If you reverse the second and third files, a different script is produced:
$ d i f f 3 -e t e s t l t e s t 3 test2
3c
grapes
 0 A Miscellany of UNlX Commands 0 315

As you might guess, this is basically the same output as doing

a d i f f on the first and third files. (The only difference in the output is the result of a
rather errant inconsistency between d i f f and d i f f 3. The latter produces an e d
script that ends with the commands that save the edited version of the file; d i f f
requires that you supply them.)
Another useful program is s d i f f (side-by-side d i f f ). Its most straightfor-
ward use is to display two files in two columns on the screen. In a gutter between the
two columns, the program displays a < if the line is unique to the first file, a > if the
line is unique to the second file, and a I if the line is different in both files. Because
the default line length of this program (130 characters) is too wide for most terminals, it
is best to use the -w option to specify a smaller width. Here are the results of running
s d i f f on two different pairs of files:
$
. s d i f f -w60 t e s t l test2
apples apples
oranges oranges
walnuts I grapes
$ sdiff - w 6 0 t e s t l t e s t 3
apples <
oranges oranges
walnuts walnuts
> chestnuts
The -s option to the s d i f f command only shows the differences between the two
files. Identical lines are suppressed. One of the most powerful uses of s d i f f is
interactive, building an output file by choosing between different versions of two files.
You have to specify the - 0 option and the name of an output file to be created. The
s d i f f command then displays a % prompt after each set of differences. You can
compare the different versions and select the one that will be sent to the output file.
Some of the possible responses are 1 to choose the left column, r to choose the right
column, and q to exit the program.
$ sdiff -w60 -0 t e s t t e s t l test3
apples <
% 1
oranges oranges
walnuts walnuts
> chestnuts
% r
$ cat t e s t
apples
oranges
walnuts
chestnuts
316 0 UNlX Text Processing 0

Having looked at these commands in simplified examples, let's now consider some
practical applications for comparing documents.
When working on a document, it is not an uncommon practice to make a copy of
a file and edit the copy rather than the original. This might be done, for example, if
someone other than the writer is inputting edits from a written copy. The d i f f com-
mand can be used to compare the two versions of a document. A writer could use it to
proof an edited copy against the original.
$ diff brochure brochure.edits
49c43,44
< environment for program development and communications,

> environment for multiprocessing, program development

> and communications, programmers
56~51
< offering even more power and productivity f o r commericial

> offering even more power and productivity for commercial

76~69
< Languages such as FORTRAN, COBOL, Pascal, and C can be

> Additional languages such as FORTRAN, COBOL, Pascal, and

Using d i f f in this manner is a simple way for a writer to examine changes without
reading the entire document. By capturing d i f f output in a file, you can keep a
record of changes made to any document.
As another example, suppose a company has a number of text files that comprise
its help facility. These files are shipped with the product and maintained online by the
customer. When there is a documentation update, these files also need to be updated.
One way to accomplish this is to replace each text file in its entirety, but that involves
distributing a lot of material that remains unchanged. Another way is to use d i f f and
simply send a record of changes between the old and the new. The -e option creates
an editing script for e d that can be used to recreate the second file from the first.
$ d i f f -e help.txt help.new > help.chgs
$ cat help.chgs
153,199d
65c
$INCLUDE {filename} program.name

56a
.Rh 0 " " "$CHAIN Statement"
-Rh "Syntax"
.in 5n
-nf
$CHAIN {filename} program.name
.fi
.in 0
 0 A Miscellany of UNlX Commands 0 317

.Rh "Description"
Use the $CHAIN statement to direct the compiler to read
source code from program-name and compile it along
....
The company could ship the file help. chgs with instructions on how to input this
editing script to ed. You'd want to create a shell script to automate this process, but
that is really an extension of knowing how it might be done from the command line.
The following command pipes the editing script to ed:
$ (cat help.chgs; echo ' w ' ) I ed - help-txt
To save the changes, a w command is submitted through echo. (In fact, if you have
any concern about sparing the original file, you could change the w to 1 , Sp. which
will cause the edited contents to be printed to standard output, but not saved in the file.
Redirect standard output to a new file to keep both copies.)
As a further example, let's take the instance where two people have made copies
of a file and made changes to their own copies, and now you want to compare them
both against the original. In this example, chOl is the original; chOl .t o m contains
edits made by Tom; and chO 1 .a n n contains changes made by Ann.
$ d i f f 3 chOl ch0l.ann ch0l.tom
____3
1:56a
2: 56a
3: 5 7 , 1 0 3 ~
.mc I
.Rh 0 n n "$CHAIN Statement"
.XX "BASIC statements, $CHAIN"
.XX "$CHAIN statement"
.Rh "Syntax"
.UN
.in 5n
.nf
$CHAIN {file) program-name
.fi
.in 0
.Rh "Description"
Use the $CHAIN statement to direct the compiler to read
source code from program-name and compile it along
....
_--_
____ 3
1 :65c
2:65c
$INCLUDE file
3:112c
$INCLUDE file program-name
_-__
____2
1: 136c
318 UNIX Text Processing 0

2: 136c
Nesting of $INSERT statements is not permitted.
3:183c
Nesting of $INSERT statements is permitted.
_ _ _ -_
1 :143,144~
program-name is converted to a valid UNIX filename.
- LP
2: 143,152~
program-name is converted to a valid UNIX filename using
the following conversion rules:
.TS
center, tab ( @ );
c 1 c.
/@is converted to@?
?@is converted to@??
Null@is converted to@?O
An initial .@is converted to@?.
.TE
3:190,191~
program-name is converted to a valid UNIX filename using
a set of conversion rules.
You often find that one version has some things right and another version has other
things right. What if you wanted to compile a single version of this document that
reflects the changes made to each copy? You want to select which version is correct for
each set of differences. One effective way to do this would be to use sdif f.
We'll use the -s option to suppress the printing of identical lines. To make the
example fit on the printed page, we specify a 45-character line length. (You would gen-
erally use an 80-character line length for the screen.) Because the total line length is
limited to 45 characters, s d i f f will be able to display only the first 15 or so charac-
ters of the line for each file; the rest of the line will be truncated.
$ s d i f f -w45 -a -0 ch0l.new ch0l.ann ch0l.tom
56a57,103
> .Rh 0 '"I "$CHAIN Statement"
> .XX "BASIC statements, $CHAIN"
> .XX "$CHAIN statement"
> .Rh "Syntax"
> .UN
> .in 5n
> .nf
> $CHAIN {\fIfile\fP} \fI
> .fi
> .in 0
> .Rh "Description"
> Use the $CHAIN statement to de
 1
0 A Miscellany of UNlX Commands 0 31 9

> code from \fIprogram.name\fP

--.....
% r
65~112
$ INCLUDE {\fIfile\fP) I $INCLUDE {\fIfile\fP)
% r
% 143,152~190,191
\fIprograrn.name\fP is I \fIprograrn.name\fP is
following rules. I following r u l e s .
- TS <
center, t a b ( @ ); <
c 1 c. <
/@is converted to@? <
?@is converted to@?? <
Null@is converted t o @ ? O <
A n initial .@is converted<
. TE <
% 1
The file c h 0 1 .new contains the portions of each file that were selected along with all
the lines that both files have in common.
Another program worth mentioning is b d i f f (big file d i f f ) . It is used on
files too large for d i f f . This program breaks up a large file into smaller segments
and then passes each one through d i f f . It maintains line numbering as though
di f f were operating on one large file.

SCCS
We’ve shown an example using di f f to produce a file that described the changes
made to a text file for a help facility. It allowed the distribution of a smaller file
describing changes instead of a wholly new version of the file. This indicates a poten-
tial application for di f f, which is fully realized in the Source Code Control System
or SCCS. SCCS i s a facility for keeping track of the changes to files that take place at
different stages of a software development or documentation project.
Suppose you have a first draft of a manual. (This is referred to as a delta when it
is saved in a special SCCS format.) The second draft, of course, is based on changes to
the first draft.
When you make the delta for the second draft, SCCS, instead of keeping a
separate copy for each draft, uses d i f f to record the changes to the first draft that
resulted in the second draft. Only the changes, and the instructions for having an editor
make them, need to be maintained. SCCS allows you to regenerate earlier drafts, which
saves disk space.
SCCS is quite complex-too complex to describe here-but we seriously suggest
that you investigate it if you are working on a large, frequently-revised or multiple-
author writing project.
320 0 UNIX Text Processing 0

Using make
The make program is a UNIX facility for describing dependencies among a group of
related files, usually ones that are part of the same project. This facility has enjoyed
widespread use in software development projects. Programmers use make to describe
how to “make” a program-what source files need to be compiled, what libraries must
be included, and which object files need to be linked. By keeping track of these rela-
tionships in a single place, individual members of a software development team can
make changes to a single module, run make, and be assured that the program reflects
the latest changes made by others on the team.
We group make with the other commands for keeping track of differences
between files only by a leap of the imagination. However, although it does not compare
two versions of the same source file, it can be used to compare versions such as a
source file and the formatted output.
Part of what makes UNIX a productive environment for text processing is discov-
ering other uses for standard programs. The make utility has many possible applica-
tions for a documentation project. One such use is to maintain up-to-date copies of for-
matted files that make up a single manual and provide users with a way of obtaining a
printed copy of the entire manual without having to know which preprocessors or
n r o f f / t r o f f options need to be invoked.
The basic operation that make performs is to compare two sets of files, for
example, formatted files and unformatted files, and determine if any members of one
set, the unformatted files, are more recent than their counterpart in the other set, the for-
matted files. This is accomplished by simply comparing the date or time stamp of pairs
of files. If the unformatted source file has been modified since the formatted file was
made, make executes the specified command to “remake” the formatted file.
To use make, you have to write a description file, usually named makefile
(or Makefile), that resides in the working directory for the project. The
make f i 1 e specifies a hierarchy of dependencies among individual files, called com-
ponents. At the top of this hierarchy is a target. For our purposes, you can think of the
target as a printed copy of a book; the components are formatted files generated by pro-
cessing an unformatted file with n r o f f.
Here’s the makef i le that reflects these dependencies.
manual: ch0l.fmt ch02.fmt ch03.fmt
l p chO [ 1-31 fmt .
ch0l.fmt: chOl
nroff -mm chOl > ch0l.fmt
ch02.fmt: ch02
t b l ch02 I nroff -mm > ch0l.fmt
ch03.fmt: ch03a ch03b ch03c
nroff -mm ch03? > ch03.fmt

This hierarchy can be represented in a diagram:

 0 A Miscellany of UNlX Commands 0 32 1

zh01. f m t zh02. f m t zh03 f m t-

t t t
n r o f f -mm tbl I n r o f f -mm n r o f f -mm

The target is manual and it is made up of three formatted files whose names appear
after the colon. Each of these components has its own dependency line. For instance,
ch 0 1 . fmt is dependent upon a coded file named c h 0 1. Underneath the dependency
line is the command that generates chOl .fmt. Each command line must begin with
a tab.
When you enter the command make, the end result is that the three formatted
files are spooled to the printer. However, a sequence of operations is performed before
this final action. The dependency line for each component is evaluated, determining if
the coded file has been modified since the last time the formatted file was made. The
formatting command will be executed only if the coded file is more recent. After all
the components are made, the lp command is executed.
As an example of this process, we’ll assume that all the formatted files are up-to-
date. Then by editing the source file ch03a, we change the modification time. When
you execute the make command, any output files dependent on ch03a are reformat-
ted.
$ make
nroff -mm ch03? > c h 0 3 . f m t
lp chO [l-31 .f m t
Only ch03. fmt needs to be remade. As soon as that formatting command finishes,
the command underneath the target manual is executed, spooling the files to the
printer.
Although this example has actually made only limited use of make’s facilities,
we hope it suggests more ways to use make in a documention project. You can keep
your makef iles just this simple, or you can go on to learn additional notation, such
as internal macros and suffixes, in an effort to generalize the description file for
increased usefulness. We’ll return to make in Chapter 18.
322 0 UNIX Text Processing 0

Manipulating Data

Removing Formatting Codes

The de r o f f command removes n r o f f/tr o f f requests, macros, inline backslash
sequences, and eqn and t b l specifications.
$ cat temp
. C H 11 “A Miscellany of UNIX Commands”
In this chapter, we present a miscellany of \s-2UNIX\sO
programs with text-processing applications.
.P
In addition, we introduce several \s-2UNIX\sO utilities
$ deroff temp
Miscellany UNIX Programs
In this chapter, we present a miscellany of UNIX programs
with text-processing applications.
In addition, we introduce several UNIX utilities
Special rules are applied to text specified as arguments to a macro so that they are not
passed through derof f. A word in a macro call must contain at least three letters.
Thus, A and of are omitted.
The d e r o f f -w command is used by spell to remove t r o f f requests
and place each word on a separate line. You can use d e r o f f in a similar manner to
prepare a word list.
$ deroff -w temp
Miscellany
UNIX
Programs
In
this
chapter
we
present
miscellany
of
UNIX
p r og rams
with
text
processing
applications
In
addition
Again, not all “words” are recognized as words. The d e r o f f command requires
that a word consist of at least two characters, which may be letters, numerals,
 0 A Miscellany of UNlX Commands 0 323

ampersands, or apostrophes. (As mentioned above, it applies slightly different rules to

text specified as an argument to a macro.)
We had hoped d e r o f f might be useful for our clients who wanted online
copies of a document but used a word processor. Because d e r o f f drops words, it
was not practical for stripping out t r o f f-specific constructs. Perhaps the best way to
do this is to use n r o f f to process the file, and then use a combination of terminal
filters to strip out tabs, backspaces (overstrikes), and reverse linefeeds.

The s o r t and uniq Commands

The s o r t command puts lines of a file in alphabetic or numeric order. The u n i q
command eliminates duplicate lines in a file.
The s o r t command works on each line of a text file. Normally, it is used to
order the contents of files containing data such as names, addresses, and phone
numbers. In the following example, we use g r e p to search for index entries, coded
with the macro .X X or .XN, and sort the output in alphabetic order.
ch04 I s o r t -df
$ grep " . X [ X N ] "
.XX "ABORT statement"
.XX "ASSIGNMENT statement"
.XX "BASIC statements, ABORT"
.XX "BASIC statements, ASSIGNMENT"
.XX "BASIC statements, BEGIN CASE''

The -f option folds uppercase and lowercase words together (that is, it ignores case
when performing the sort). The -d option sorts in dictionary order, ignoring any spe-
cial characters.
The u n i q command works only on sorted files, comparing each adjacent line.
The s o r t command has a -u option for removing all but one indentical set of lines.
Usually this is sufficient, but u n i q does have several options, which gives you addi-
tional flexibility. For example, here's the sorted output of four files:
$ sort test*
apples
apples
apples
chestnuts
chestnuts
grapes
oranges
oranges
oranges
oranges
walnuts
walnuts
walnuts

The -d option prints one line for each duplicate line, but does not print lines that are
unique.
324 0 UNlX Text Processing 0

5 sort t e s t * I u n i q -d
apples
chestnuts
oranges
walnuts

In this example, grapes has been filtered out. The -u option prints only unique lines.
If we used the -u option, only grapes would appear.
You wouldn’t expect sort to be useful on a document containing long lines of
text. However, if you bothered to start sentences on a new line when creating the input
file (as we recommended in Chapter 3), scanning a sorted file can produce some
interesting things. The following command sorts the contents of ch03 and pipes the
output through pg:
5 sort -u ch03 I pg
Looking at the results gives you a slightly turned about view of your document. For
instance, you might notice inconsistencies among arguments to formatter requests:
f SP
.sp .2i
- s p .3v
.sp .5

Or you could check the frequency with which sentences begin in the same manner:
It i s dangerous t o use mvcur()
It is designed so that each piece of code
It is possible that some programs

In the next example, we use derof f to create a word list. Then we sort it and
use uniq to remove duplicates. The -c option with u n i q provides a count of the
occurrences of identical lines. (It overrides -u and -d.)
$ deroff -w ch03 I s o r t -fd I u n i q - c
1 abort
1 aborted
3 about
4 above
1 absolute
1 absorb
1 accepting
1 accomplishes
1 active
2 actual
5 actually
2 Add
7 add
...
68 you
3 Your
 A Miscellany of UNIX Commands 0 325

13 your
2 zero

In the next example, we repeat the previous command, this time adding another
sort at the end to order the words by frequency. The -r option is used to reverse the
comparison, putting the greatest number first.
$ deroff -w ch03 I sort -fd I uniq -c I s o r t -rfd
666 the
234 to
219 is
158 window
156 of
1 4 8 and
1 1 4 in
111 s c r e e n
105 t h a t
83 c h a r a c t e r
76 a r e
...
1 aborted
1 abort
You will find other examples of s o r t in the next section, where we look at sorting
particular fields. Be sure to read the UNIX command pages for s o r t and uniq and
experiment using different options.

The join Command

The j oin command compares lines contained in separate files and joins lines that
have the same key. (When you use sort or join,each line is separated intofields
by blanks or tabs. Normally, the first field is the key field, on which the sort or
join is performed. However, there are options that allow you to change the key
field.) The file must be sorted in ascending ASCII sequence before being processed by
join.
$ cat 85
j an 19
feb 05
mar 14
aPr 15
may 15
j un 18
jul 19
aug 20
s eP 19
nov 18
dec 18
$ cat 8 6
326 0 UNlX Text Processing

jan 09
feb 15
mar 04
aPr 06
may 14
jun 13
jul 13
aug 10
seP 14
nov 13
de c 12
$ sort 85 > 8 5 .,temp; sort 86 > 86 t-p
First we sort both of these files, creating temporary files. Then we perform the j o i n ,
followed by a sort with the -M option, to reorder them by month.
$ join 85.temp 86.temp I sort -M > joiner
$ cat joiner
jan 1 9 09
f e b 05 15
mar 1 4 04
apr 1 5 06
may 15 14
jun 18 13
jul 1 9 13
aug 20 10
sep 1 9 14
nov 1 8 13
dec 1 8 12
$

After the data is joined in this manner, it can be sorted by field. Fields are separated by
blank spaces or tabs. The sort can be performed on specific fields, using + to indicate
the first sort field and - to indicate the last sort field. The first field is +O. To sort on
the second field, use + 1 .
$ sort +1 joiner
feb 05 1 5
mar 14 04
apr 15 0 6
may 15 14
dec 1 8 12
jun 18 13
nov 18 13
jan 19 09
jul 19 13
sep 1 914
aug 20 1 0
 0 A Miscellany of UNIX Commands 0 327

The corn Command

The corn command reads the contents of two sorted files and produces for output a
three-column listing of lines that are found

only in the first file;

. only in the second file;
in both the first and second files.

For example, let’s suppose that we had generated a list of UNIX commands found in
Berkeley 4.2 and another list of commands found in AT&T System V.2. We can use
corn to produce a compact listing of commands found exclusively in one version and
commands common to both. For obvious reasons, this example uses only the beginning
o f the list.
$ cat bsd4.2
adb
addbib
apply
apropos
ar
as
at
awk

$ cat attV.2
adb
admin
ar
as
as.3
at
awk

Note that both files have already been sorted.

$ comm bsd4.2 attV.2
adb
addbib
admin
apply
apropos
ar
as
asa
at
awk
328 0 UNlX Text Processing 0

Commands found only on systems running Berkeley 4.2 are in the left-hand column,
and those found only on AT&T System V.2 are in the center column. Commands
found in both versions are listed in the right-hand column.
You can also suppress the display of one or more columns. For instance, if you
wanted to display only the commands that were found on both systems, you’d enter:
$ comrn -12 bsd4.2 a t t V . 2
Only the third column would be shown.
B y specifying - instead of a filename, you can also use standard input. In the
next example, we produce a listing of filenames from two directories on the system, sort
them, and compare them against the commands named in the bsd4.2 file. This
allows us to compare commands found on our system with those on the list of Berkeley
commands.
$ ( cd /bin ; 1s ; cd /usr/bin ; 1s ) I sort I comm - bsd4.2
a c ct corn
adb
addbib
admin
apnum
apply
apropos
ar
as
asa
at
awk

Parentheses are used to group a series of commands, combining their output into a sin-
gle stream; we want a list of command names without pathnames from several direc-
tories. Because a new shell is created to execute these commands, notice that we do not
change our current working directory when the commands in parentheses have finished
executing.

The cut and paste Commands

The c u t and p a s t e commands modify a table or any other data in fields or
columns. You can extract specific columns of data using c u t , and join them horizon-
tally using p a s t e .
For our examples, we’ll make use of a portion of a table of ASCII characters that
specifies their decimal and hexadecimal values. (This example is probably unneces-
sarily complex; you can use cut and paste for much simpler jobs than this!) Here’s
what the table looks like to begin with:
$ cat appc
.T S
c e n t e r , box;
cb cb cb
 A Miscellany of UNIX Commands 0

n n 1.
Decimal Hexadecimal ASCII
-
-

000 00 NUL
001 01 so
002 02 STX
003 03 ETX
004 04 EOT
005 05 ENQ
006 06 AC K
007 07 BEL
008 08 BS
009 09 HT
.TE
Each column i s separated by a tab. A tab is the default field delimiter for c u t ; the
-d option can be used to change it. The -c option allows you to specify character
positions or ranges. The command c u t - c 6 - 8 0 would print characters beginning at
position 6 through 80, truncating the first five characters. The -f option is used to
specify one or more fields that are passed to standard output. (Given the name of the
command, one might reasonably think you’d specify the fields or column position you
wanted cut out, but.. . )
In the next example we extract the third field, which contains the ASCII names:
$ cut -f3 -s appc
ASCII
NUL
so
STX
ETX
EOT
ENQ
ACK
BEL
BS
HT

We use the -s option to remove all lines that do not have any delimiters, thus drop-
ping the t b l constructs from the output. Normally, c u t passes lines without delim-
iters straight through, and that i s what we really want for our next feat. W e are going to
reorder the table so that it can be referenced by the ASCII name rather than by decimal
number. All of this can be done from the command line, with only a brief entry into
the editor at the end.
We’ll look at this in stages. First, we extract the third column and send it along
to p a s t e :
330 0 UNlX Text Processing 0

$ cut -f3 appc I paste - appc

- TS .T S
center, box; center, box;
cb cb cb cb cb cb
n n l.n n 1.
ASCII Decimal Hexadecimal ASCII

NUL 000 00 NUL

so 001 01 so
S TX 002 02 STX
E TX 003 03 ETX
EOT 004 04 EOT
ENQ 005 05 ENQ
ACK 006 06 AC K
BEL 007 07 BEL
BS 008 08 BS
HT 009 09 HT
.TE .TE

The paste command reads one or more files or standard input (the - option) and
replaces the newline with a tab in all but the last file. This gives us four columns.
(Yes, it doubled the t b l specifications, but we have an editor.) Now, all we have to
do is extract the first three columns from the output. Only c u t -f 1 , 2 , 3 has been
added to the previous command, and the output is redirected to a file.
$ cut -f3 appc I paste - appc I cut - f 1 , 2 , 3 > ascii.table
$ cat ascii.table
- TS .T S
center, box; center, box;
cb cb cb cb cb cb
n n 1.n n 1.
ASCII Decimal Hexadecimal
-
- -
-

NUL 000 00
so 001 01
S TX 002 02
ETX 003 03
EOT 004 04
ENQ 005 05
ACK 006 06
BEL 007 07
BS 008 08
HT 009 09
.TE .TE

This gives us three columns in the correct order. We can go into v i to rearrange the
t b l constructs and execute a sort command on just the data portion of the table to
bring it all together.
 0 A Miscellany of UNlX Commands 0 331

$ cat ascii.table
.TS
center, box;
cb cb c b
n n 1.
ASCII Decimal Hexadecimal
-

ACK 006 06
BEL 007 07
BS 008 08
ENQ 005 05
EOT 004 04
ETX 003 03
HT 009 09
NUL 000 00
so 001 01
S TX 002 02
-TE

The p a s t e command can be used in several interesting ways. Normally, in order to

merge two files, p a s t e replaces the newline in the first file with a tab. The -d
option allows you to specify a substitute for the tab. This can be any single character or
a list of characters. Special characters can be represented as follows: newline (\n), tab
( \ t), backslash ( \ \), and empty string (\ 0). Each character in the list i s assigned in
sequence to replace a newline, and the list is recycled as many times as necessary. We
can use p a s t e to present our three-column table in six columns:
$ paste -s -d"\t\n" appci
.TS center, box;
c b c b cb n n 1.
Decimal Hexadecimal ASCII =
000 00 NUL 001 01 SO
002 02 STX 003 03 ETX
004 04 EOT 005 05 ENQ
006 06 ACK 007 07 BEL
008 08 BS 00 9 09 HT
.TE
The - s option is used when only a single file is specified. It tells p a s t e to merge
subsequent lines in the same file rather than to merge one line at a time from several
files. In this example, the first line's newline is replaced by a tab while the second line
retains the newline. To get nine columns out of three-column input, you'd specify
-d" \ t \ t \ n
'I.

A little work needs to be done to the t b l specifications. You could also execute
the p a s t e command from within v i so that it only affects the data portion.
You would probably want to go to this much trouble for a large table (or many
small tables) rather than the small examples shown here. A more practical example that
uses p a s t e alone would be to construct a multi-column table from a single long list
332 0 UNIX Text Processing 0

of words. Simply split the list into equal-sized chunks, then paste them together side by
side.

The tr Command
The t r command is a character translation filter, reading standard input and either
deleting specific characters or substituting one character for another.
The most common use of t r is to change each character in one string to the
corresponding character in a second string. (A string of consecutive ASCII characters
can be represented as a hyphen-separated range.)
For example, the command:
$ tr "A-2" "a-z" < fire

will convert all uppercase characters in fife to the equivalent lowercase characters. The
result is printed on standard output.
As described in Chapter 7, this translation (and the reverse) can be useful from
within v i for changing the case of a string. You can also delete specific characters.
The -d option deletes from the input each occurrence of one or more characters speci-
fied in a string (special characters should be placed within quotation marks to protect
them from the shell). For instance, the following command passes to standard output
the contents of file with all punctuation deleted:
$ cat file I tr -d 'I, . !?; :'I

The -s (squeeze) option of t r removes multiple consecutive occurrences of the

same character. For example, the command:
$ t r -a " '' < file

will print on standard output a copy of file in which multiple spaces in sequence have
been replaced with a single space.
We've also found t r useful when converting documents created on other sys-
tems for use under UNIX. For example, one of our writers created some files using an
IBM PC word processor. When we uploaded the files to our system, and tried to edit
them with v i , we got the message:
N o t an ascii f i l e

and a blank screen. The v i editor could not read the file. However, using a program-
ming utility that lists the actual binary values that make up the contents of a file (od, or
octal dump), we were able to determine that the word processor used nulls (octal 000)
instead of newlines (octal 012) to terminate each line.
The t r command allows you to specify characters as octal values by preceding
the value with a backslash, so the command:
$ tr ' \ O O O ' '\012'

was what we needed to convert the file into a form that could be edited with v i .
 0 A Miscellany of UNIX Commands 0 333

Splitting Large Files

Splitting a single large file into smaller files can be done out of necessity-when you
come across a program that can't handle a large f i l e - o r as a matter of preference-
when you find it easier to work with smaller files. UNIX offers two different programs
for breaking up files, s p l i t and c s p l i t .
The s p l i t command divides a file into chunks, consisting of the same number
of lines. This is IO00 lines, unless specified differently. In the following example of
s p l i t , we break up a 1700-line file into 500-line chunks. The w c command supplies
a summary of the number of lines, words, and characters in a text file.
$ wc ch03
1708 8962 59815 ch03
$ split -500 ch03
$ w c ch03*
500 2462 16918 ch03aa
500 2501 16731 ch03ab
500 2976 19350 ch03ac
208 1023 6816 ch03ad
1708 8962 59815 ch03
The s p l i t command created four files. It appended aa, ab, ac, etc. to the end of
the original filename to create a unique filename for each file. You can also specify, as
a third argument, a different filename to be used instead of the original filename.
Look at the end of one of these files:
$ tail ch03ac
.Bh "Miscellaneous Functions"
.in 5n
.TS
t a b ( @ );
1 1 1.

Unfortunately, the file breaks in the middle of a table. The split command pays no
attention to content, making it inadequate for breaking a file into manageable, but com-
plete, sections.
The c s p l i t command offers an alternative, allowing you to break a file in con-
text. There are two ways to use it. The first is to supply one or more line numbers.
You could enter the following command:
$ csplit ch03 100 145 200
Four files would be created (0-99, 100-144, 145-199, 200-end). The naming convention
for files created by c s p l i t is different than s p l i t . Files are named xx00,
xxOl, xxO2, and so on. If you want to specify a prefix that is different than xx,
you can do so with the -f option.
Because we do not know in advance which line numbers to specify, we can use
grep to get this information. The -n option to g r e p causes line numbers to be
returned. In this example, we specify a pattern to match the section header macros, Ah
and Bh:
334 UNlX Text Processing 0

$ grep -n 'I. [ABIh" ch03

It produces the following listing:

5: .Ah '' Int roduct ion''
30: .Ah "Using the Curses Library"
175: .Ah "The Curses Functions"
398 :.Bh "Adding Characters to the Screen Image"
638: .Bh "Standout Mode"
702: .Bh "Getting Characters from the Terminal"
777: .Bh "Input Modes"
958 : .Bh "Erasing and Clearing"
1133: .Bh "Creating and Removing Multiple Windows"
1255: .Bh "Window-Specific Functions"
1301: .Bh "Mani pu 1ating Mu 1 t ip 1e W indows"
1654: .Bh "Terminal Manipulation"
From this listing, we select the appropriate places at which to split the file and supply
these numbers to split. The -f option is used to supply a filename prefix.
$ c s p l i t -f ch03. ch03 175 1133
6803 Number of bytes in each segment
32544
20468
$ Is ch03.*
ch03.00
ch03.01
ch03.02
The csplit command prints a character count for each of the three files it created.
(This count can be suppressed using the - s option.)
The second way to use csplit is to supply a list of patterns. For instance, if
you had prepared an outline that you wanted to break into files correponding to sections
I, 11, and 111, you could specify:
$ c s p l i t -s -f s e c t . o u t l i n e /I-/ /II./ /III./
$ Is s e c t . *
sect.0 1
sect.0 2
sect.03
You can also repeat a pattern. In one project we were working on, one large file
contained a number of commands in reference page format. We decided it would be
easier if we put each command in its own file. The beginning of a reference header was
marked by the macro .Rh 0 . First, we used g r e p to determine the number of
times this macro occurred.
$ grep -c *' .Rh 0" ch04
43
We reduce this number by 1 and surround it with braces:
 0 A Miscellany of UNIX Commands 0 335

$ csplit -s -f ch04. ch04 "/.Rh O / " 442)

The pattern is enclosed within double quotation marks because it contains a space. (If
you use the C shell, you must protect the braces from being interpreted by placing them
in double quotation marks as well.) This command creates 43 files:
$ 1s ch04*
ch04
-
ch04 00
-
ch04 01
ch04.02
ch04.03
...
ch04.39
ch04.40
ch04.41
ch04.42
-
ch04 43

The only task remaining is to rename the files, using the name of the command listed as
the first argument to the . R h macro. (We'd have to write an awk or s h e l l script
to do this automatically.)
After you have divided a large file into a number of smaller files, you might
organize them in a subdirectory. Let's look at a small example of this:
$ mkdir ch04.files
$ mv ch04.?? ch04.files
Again, the usefulness of filename metacharacters is apparent, giving us the ability to
move 43 files without typing 43 filenames.

Encryption
The cloak-and-dagger set and the security conscious will find uses for the encryption
facilities of UNIX. (These facilities are not available on UNIX systems sold outside the
United States.) The crypt command reads a file from standard input, asks you to
supply a key for encoding the file, and passes to standard output an encrypted version of
the file. You should redirect standard output to a new file because the encrypted file is
not readable text.
$ cat message I crypt > encrypted.msg
Enter key:alabaster
Just as when you enter a password, the key does not appear on the screen as you enter
it. If you prefer, you can enter the key as an argument to crypt. To decode an
encrypted file, you simply cat the file to crypt and supply the key.
The UNIX editors ed, ex, and v i , can be invoked with the -x option to read
or edit an encrypted file. (Some versions of these programs recognize this option but
do not support the encryption feature.) Of course, you have to supply the correct key.
336 0 UNLX Text Processing 0

Cleaning Up and Backing Up

In this section, we show some procedures for backing up active files to some other
medium such as tape or floppy disk. At many sites, backups are the responsibility of
one person, who performs these tasks on a regular basis to ensure that users can recover
much of their data in case there is a serious system crash. At other sites, individual
users might be responsible for doing their own backups, especially if there are only a
few users on the system. Whoever does it must ensure that backups of important files
are made periodically.
A second reason for learning a backup procedure is to enable you to store files on
an off-line medium. For users of PCs, this is the standard method of operation (and
therefore much simpler to do), but all UNIX systems have hard disks as the primary
storage medium. No matter how large a disk drive is, sooner or later, users will fill it to
capacity. Frequently, there are useless files that can be deleted. Other inactive files,
such as an early draft of a document, might be removed from the system after you have
made a copy on floppy disk or tape. After a project is finished, you probably want to
make several copies of all important files. At a later time, should you need files that
have been stored off-line, you can easily restore them to the system.
We are going to describe how to use the cpio command for backing up one or
more working directories. There are other UNIX commands that might be used as well
( t a r and dd, for instance). At your site, you may even have simpler shell scripts that
prevent you from having to deal with cpio directly. Ask an expert user at your site
about backup procedures and go through it once or twice. Apart from learning about
cpio, you will need:

1. The UNIX filename of the device (ldevlxuxu) to which you are directing
the output of the cpio command.
2. Familiarity with operating the device, such as being able to load a tape in
the tape drive and knowing how to format a floppy disk prior to use.

You can use cpio in two basic ways, either to back up or to restore files. You use
cpio with the -0 option and > to redirect output to the device for backup, or with
the -i option and < to redirect input from the device to restore files.
Unlike many of the commands we’ve looked at, cpio depends exclusively on
reading a list of filenames from standard input. This list identifies the files that will be
backed up. For practical purposes, this involves doing an 1s command on the direc-
tory you want backed up and piping the results to cpio.
You need to know the UNIX filename for the backup device. This name is site
specific, so you need to check with a knowledgeable user. At our site, we have a
floppy disk drive named /dev/rfp021. A tape drive might be named
/dev/mt0.
After you have loaded the tape in the tape drive or placed the floppy disk in the
disk drive, you can perform the backup using your own version of this command:
 0 A Miscellany of UNlX Commands 0 337

$ 1s /work/docbook/chl3 I cpio -ov > /dev/rfpO21

sect 3
dict
shellstuff
..-
384 b l o c k s

The -v (verbose) option prints a list of filenames on the screen.

The -i option to c p i o reads or restores files from a tape or floppy disk device.
Sometimes, before you actually restore files, you want to list the contents of the tape or
disk. The -t option prints a table of contents but does not actually read these files
onto the system.
$ cpio -it < /dev/rfp021
384 b l o c k s
sect3
dict
shellstuff
.--
Using the -v option along with the -t option produces a long (verbose) listing of
files, as if you had entered Is -1.
You don't have to extract all the files from disk or tape. You can specify certain
files, using filename metacharacters to specify a pattern.
$ cpio -iv "sect?" < /dev/rfp021
No m a t c h .

Remember to refer to the full pathname if the files were saved using a complete path-
name, and to put pathnames that include metacharacters within double quotation marks.
$ cpio -i "/work/docbook/chl3/sect?" < /dev/rfp021
384 b l o c k s
sect3
sect2
sect 1

Before restoring a file, c p i o checks to see that it won't overwrite an existing file of
the same name that has been modified more recently than the file being read.
You can also use the f i n d command with the - c p i o condition to do a back
up. The advantage of using f i n d i s that it descends all the way down a directory
hierarchy.
$ find /work/docbook/chl3 -cpio /dev/rfp021

To restore a directory hierarchy, use the -d option to cpio. Administrators fre-

quently use f i n d to generate a list of files that have been modified within a certain
time period. The conditions -mt i m e (modification time) and - a t i m e (access time)
can be followed by a number indicating a number of days. This number can be pre-
ceded by a plus sign, indicating more than that number of days, or a minus sign, indi-
cating less than that many days. If there is no sign, the condition indicates exactly that
number of days.
338 0 UNlX Text Processing 0

This example uses f i n d to produce a list of files that have been modified within
the last seven days. These active files are good candidates for backups.
$ find /work/docbook -mtime -7 -print
/work/docbook
/work/docbook/oshell
/work/docbook/chOl
...
Don’t forget you have to specify - p r i n t to see the results of a f i n d command.
You could work up your own version of this command to look for your own files
that have not been accessed in the last 21 days. Add the option - a t i m e with an
argument of + 2 1 to list the files and directories that have not been accessed in over 21
days. Add the -user option to look only for your own files, the -cpio option to
backup these files, and the -ok option to execute an r m command to delete them
from the system after they’ve been backed up.
$ find /work -atime +21 -user -cpio /dev/rfp021 -ok rm { I \;
The -ok option is the same as the -exec option; however, instead of executing the
command specified within parentheses on all files selected by find, it prompts you
first to approve the command for each file.

1 Compressing Files
You can conserve the amount of disk space that text files take up by storing some of
your files in a compressed form. The p a c k command can be used to compress a file.
It generally reduces a text file by 25 to 40 percent.
$ 1s -1 ch04/sectl
-rw-rw-rw- 1 €red doc 29350 Jun 10 1 5 : 2 2 ch04/sectl
$ pack ch04/sectl
pack: ch04/sectl: 39.9% Compression
The original file is replaced by a packed file with a .z appended to the original
filename.
$ 1s -1 ch04/sectl.z
-rw-rw-rw- 1 €red doc 17648 Jun 10 1 5 : 2 9 ch04/sectl.z
The p a c k command reduced the size of this file from 29K to 17K bytes. If used
system-wide, it could save a significant amount of disk space, although the amount of
compression will vary from file to file. Obviously, there is less benefit in packing small
files.
To expand a packed file, use the u n p a c k command. You can specify the name
of the file with or without the .z suffix.
$ unpack ch04/sectl
unpack: ch04/sectl: unpacked
 A Miscellany of UNIX Commands 0 339

Another way to temporarily unpack a file is to use a special version of cat for packed
files, called pcat. Use this command to view a packed file (pipe it through more or
pg) or send it as input to another command, as in the following example:
$ pcat chOl/sectl I n r o f f -mrn

Communications .
More and more, we find that our projects require us to work on several different com-
puter systems, some of them UNIX systems, some not. Given this situtation, the abil-
ity to work remotely on other systems and to transfer files has been essential. For-
tunately, a number of useful communications programs are part of the standard UNIX
shipment.
Two basic types of connections between computer systems are a dial-up line,
using a modem to communicate across phone lines, and a direct line, when two com-
puter systems are in close proximity and can be connected by a single cable. The
u u c p and c u commands establish communication links using both types of connec-
tions.
The cu command (Berkeley’s version is called tip) is a UNIX program for
conducting a login session on a remote computer system. UUCP (UNIX-to-UNIX
copy) is a series of related programs for transferring files between UNIX systems. Its
main program is called uucp.
We cannot provide full descriptions of these facilities here. A good way to learn
is to ask an expert user to help you transfer files or begin a remote login session. Keep
notes on the procedure and when following it, if things don’t work as expected, get
more help.
The UUCP programs are quite straightforward and easy to use after you are
accustomed to the conventions. Each system on the UUCP network has a file that
describes the other systems linked to it and what types of links are available. This file
is created by the system administrator of each system. You can find out the names of
these remote systems by entering the uuname command. If your system is properly
configured and you have a login on a remote system, such as b o s t o n , you can begin
a remote session by entering:
$ cu boston

After you are connected to the remote system, you should get a login message. To quit
-
a remote session, log out and then enter . (tilde dot) to return to your own machine.
There are a number of commands you can enter while under the control of c u ,
permitting, for instance, the execation of commands on the local system while you are
still logged in to the remote system. Check the reference page in your UNIX documen-
tation.
You can also dial direct to a non-UNIX system by specifying a telephone number
on the command line (providing, of course, that the files accessed by these communica-
tions programs have been properly configured by the system administrator).
340 UNlX Text Processing 0

You can send mail to users on these remote systems and transfer files. Gen-
erally, file transfers take place between public directories on both systems, usually
/usr/spool /uucppublic. File transfers between other directories will contend
with file and directory access permissions as well as uucp permissions set by the sys-
-
tem administrator. The character serves as a shorthand for the public directory.
For instance, when working on site for a client, we often create files that we want
to send to our own system. If we are logged in on their system, we can send the file
outline to our system named o r a by entering:
$ uucp -m outline ora!-/fred/
The UUCP facility is batch oriented, accepting requests and acting upon them in the
order in which they are received. Although it may execute your request immediately, if
it is busy or encounters difficulty making the connection, UUCP will carry out the
request at a later time.
The -m option is used so that we are sent mail when the copy i s actually com-
pleted. The system name is followed by an exclamation mark (if you use the C shell,
escape ! with a backslash). Then you specify a tilde (-) followed by the user’s name.
Putting a slash after the user name (fred) ensures that the user name will be inter-
preted as a directory (or a directory will be created if one does not exist).
Occasionally, you will need to transfer a large number of files or, perhaps, an
entire directory hierarchy. There are some simple tricks you can use to combine multi-
ple files into a single file, making it easier to transmit to another system. They are
especially helpful when you transfer between public directories.
You must first create a list of the files to be included. (You can do this either
manually or with a command like Is or find.) Then use cpio to create what we
can call a file archive on standard output rather than on a backup device. Redirect stan-
dard output to a file, then use UUCP to send the archive. Use the same backup pro-
gram on the target system to restore the archive. For example, if you had a book made
up of files chOl, ch02, etc., you could “package” that book for transfer to another
system using cpio:
b o s t o n $ cd /usr/proj/book
boston$ find . -name ‘chO?‘ -print I cpio -oc > book.archive
or using a manually generated list of filenames:
boston$ 1s chO? > filelist
b o s t o n $ cpio -oc < filelist > book-archive

Then, after transferring book. archive (instead of numerous individual files) to the
remote system with UUCP, a user can restore the archive:
calif$ mkdir /usr/proj/book
calif$ mv /usr/spool/uucppublic/book.aschive /usr/proj/book
calif$ cd /usr/proj/book
calif$ cpio -icd < book.archive
(The -c option of cpio writes header information in ASCII for portability; -d tells
cpio to create directories if needed when doing the restore.)
 1

A Miscellany of UNIX Commands 0 34 1

(On Berkeley UNIX systems, you can do something similar with tar. See your
UNIX manual for details.)

Scripts of UNIX Sessions

Throughout this chapter, we have provided examples of UNIX commands. These
examples were made using a command called script (which is not a standard Sys-
tem V command). The script command allows you to make a file copy of a UNIX
session. Without this facility, we?d have to simulate the examples by hand.
After you invoke script, your input and output is copied to a file. By default,
the name of this file is typescript, but you can supply a different name on the
command line.
$ script
S c r i p t s t a r t e d on Thu J u l 10 12:49:57 1987
$ echo hello
hello
$

To quit, you enter CTRL-D.

$ cat typescript
S c r i p t s t a r t e d on Thu J u l 10 12:49:57 1987
$ echo h e l l o
hello
$
s c r i p t d o n e o n Thu J u l 10 12:50:11 1987
After we make a script, we simply read the file into our text using vi.
Keeping a script of a procedure is also a good start for building a shell script that
performs a routine task automatically.
 C H A P T E R

12

Let the Computer Do the Dirty Work

Computers are very good at doing the same thing repeatedly, or doing a series of very
similar things one after another. These are just the kinds of things that people hate to
do, so it makes sense to learn how to let the computer do the dirty work.
As we discussed in Chapter 7, you can save e x commands in a script, and exe-
cute the script from within v i with the :so command. It is also possible to apply
such a script to a file from the outside-without opening the file with v i . As you can
imagine, when you apply the same series of edits to many different files, you can work
very quickly using a script.
In addition, there is a special UNIX editor, called sed (stream editor), that only
works with scripts. Although sed can be used to edit files (and we will show many
useful applications in this chapter), it has a unique place in the UNIX editing pantheon
not as a file editor, but as a filter that performs editing operations on the fly, while data
is passed from one program to another through a pipe.
The sed editor uses an editing syntax that is similar to that used by ex, so it
should not be difficult to learn the basics.
The awk program, which is discussed in the next chapter, is yet another text-
processing program. It is similar to sed, in that it works from the outside and can be
used as a filter, but there the resemblance ends. It is really not an editor at all, but a
database manipulation program that can be turned into an editor. Its syntax goes
beyond the global substitution/regular expression syntax we’ve already seen, and so
awk may be the last thing that many writers learn. Nonetheless, it has some important
capabilities that you may want to be familiar with.
Finally, to make best use of these tools, you need to know a bit about shell pro-
gramming. In fact, because the shell provides a framework that you can use to put all
these other tools together, we need to discuss it first.
If you are a programmer, and have already worked with the shell, this discussion
may be too elementary; however, we are assuming that many of our readers are writers
with only minimal exposure to programming. They, like us when we started working
with UNIX, need encouragement to branch out into these untried waters that have so lit-
tle apparent connection to the task at hand.

342
 0 Let the Computer Do the Dirty Work 0 343

This chapter is different from those in the first part of the book in that it not only
teaches the basics of some new programs, but also puts them to work building some
useful text-processing tools. At times, material is organized according to what i s
needed to build the tools, rather than as a comprehensive attempt to teach the program
itself. As a result, the material presented on sed, for example, is less complete than
our earlier treatment of v i . We cover the most important points, but in many ways
this chapter is suggestive. If you come away with a sense of possibility, it has done its
job.

Shell Programming
A shell script, or shell program, can be no more than a sequence of stored commands,
entered in a file just as you would type them yourself to the shell.
There are two shells in common use in the UNIX system, the Bourne shell (sh),
championed by AT&T, and the C shell (csh), developed at the University of California
at Berkeley. Although the C shell has many features that make it preferable for interac-
tive use, the Bourne shell is much faster, so it is the tool of choice for writing shell
scripts. (Even if you use the C shell, scripts written using Bourne shell syntax will be
executed in the Bourne shell.)
We discuss the Bourne shell exclusively in this chapter, although we make refer-
ence to differences from the C shell on occasion. This should pose no problem to C
shell users, however, because the basic method of issuing commands is identical. The
differences lie in more advanced programming constructs, which we will not introduce
in detail here.

Stored Commands
The .profile (or .login if you use the C shell) file in your home directory is a
good example of a shell program consisting only of stored commands. A simple
. p r o f i l e might look like this:
stty erase "'
H
' echoe kill '-X' intr ' - C '
PATH=/bin:/usr/bin:/usr/local/bin:.;export PATH
umask 2
date
mail

This file does some automatic housekeeping to set up your account environment every
time you log in. Even if you aren't familiar with the commands it contains, you can get
the basic idea. The commands are executed one line at a time; it is a tremendous time-
saving to be able to type one command instead of five.
You can probably think of many other repetitive sequences of commands that
you'd rather not type one at a time. For example, let's suppose you were accustomed to
working on an MS-DOS system, and wanted to create a dir command that would
print out the current directory and the names and sizes of all of your files, rather than
just the names. You could save the following two commands in a file called dir:
344 0 UNlX Text Processing 0

pwd
Is -1
To execute the commands saved in a file, you can simply give its name as an argument
to the sh command. For example:
$ sh dir
/work/docbook/chl3
total 21
-rw-rw-r-- 3 fred doc 263 Apr 12 09:17 abbrevs
-rw-rw-r-- 1 fred doc 10 May 1 14:Ol dir
-rw-rw-r-- 1 fred doc 6430 Apr 12 15:OO sectl
-rw-rw-r-- 1 fred doc 14509 Apr 15 16:29 sect2
-rw-rw-r-- 1 fred doc 1024 Apr 28 10:35 stuff
-rw-rw-r-- 1 fred doc 1758 Apr 28 1O:OO tmp
Or you can make a file executable by changing its file permissions with the chmod
command:
$ 1s -1 dir
-rw-rw-r-- 1 fred doc 10 May 1 14:Ol dir
$ chmod + x d i r
$ 1s -1 dir
-rwxrwxr-x 1 fred doc 10 May 1 14:Ol dir
After a file has executable permission, all you need to do to execute the commands it
contains is to type the file’s name:
$ dir
/work/docbook/chl3
total 21
-rw-rw-r-- 3 fred doc 263 Apr 12 09:17 abbrevs
-rwxrwxr-x 1 fred doc 10 May 1 14:Ol dir
-rw-rw-r-- 1 fred doc 6430 Apr 12 15:OO sectl
-rw-rw-r-- 1 fred doc 14509 Apr 15 16:29 sect2
-rw-rw-r-- 1 fred doc 1024 Apr 28 10:35 stuff
-rw-rw-r-- 1 fred doc 1758 Apr 28 1O:OO tmp
The next step is to make the shell script accessible from whatever directory you happen
to be working in. The Bourne shell maintains a variable called PATH, which i s set up
during the login process, and contains a list of directories in which the shell should look
for executable commands. This list is usually referred to as your search path.
To use the value of a variable, simply precede its name with a dollar sign ($).
This makes it easy to check the value of a variable like PATH-simply use the echo
command:
$ echo $PATH
/bin:/usr/bin:/usr/local/bin:.
The Bourne shell expects the list of directory names contained in the PATH variable to
be separated by colons. If your search path is defined as shown, the following direc-
tories will be searched, in order, whenever you type the name of a command:
 0 Let the Computer Do the Dirty Work 0 345

/bin
/usr/bin
/usr/local/bin
. (shorthand for the current directory)
The allocation of system commands to the three bin directories is historical and
somewhat arbitrary, although /usr/local/bin tends to contain commands that are
local to a specific implementation of UNIX. It is sometimes called /usr/lbin or
some other name.
To ensure that any shell scripts you create are automatically found whenever you
type their names, you can do one of two things:

1. You can add shell scripts to one of the directories already in your search
path. However, in most cases, these directories are only writable by the
super-user, so this option is not available to all users.
2. You can create a special “tools” directory of your own, and add the name
of that directory to your search path. This directory might be a subdirectory
of your own home directory, or could be a more globally available directory
used by a group of people.

For example, you could put the following line in your .profile:
PATH=/usr/fred/tools:.:/bin:/usr/bin:/usr/local/bin:

The /usr/fred/tools directory would be searched before any of the standard

search directories. (This means that you can define an alternate command with the
same name as an existing command. The version found first in the search path is exe-
cuted, and the search is stopped at that point. You should not put local directories
before the standard directories if you are concerned at all with system security, because
doing so creates a loophole that can be exploited by an intruder.)
If you are using the C shell, the search path i s stored in a variable called path,
and has a different format; see your UNIX documentation for details. In addition, you
must use the rehash command whenever you add a command to one of the search
directories.

Passing Arguments to Shell Scripts

The previous example is very simple; the commands it used took no arguments. In con-
trast, consider a case in which you want to save a single complex command line in a
file. For example, if you use tbl and eqn with nrof f, your typical command line
might look like this:
$ tbl file I eqn I nroff -ms I col I lp

How much easier it would be to save that whole line in a single file called format,
and simply type:
346 UNlX Text Processing 0

$ format file
The question then becomes: how do you tell your format script where in the com-
mand line to insert the file argument?
Because all of the programs in the script are designed to read standard input as
well as take a filename argument, we could avoid the problem by writing the script
thus:
tbl I eqn I nroff -ms I col I lp

and using it like this:

$ cat file I format

or like this:
.$ format < file
But this still begs the question of how to pass an argument to a shell script.
Up to nine arguments can be represented by positional notation. The first argu-
ment is represented in the shell script by the symbol $ 1 , the second by $ 2 , and so on.
So, for example, we could write our script:
t b l $1 I eqn I nroff -ms 1 col I lp
When specified as an argument to the format command:
$ format chOl

the filename would be substituted in the script for the symbol $ 1 .

But what if you want to specify several files at once? The symbol $ * means
“use all arguments,” so the script:
t b l $* I eqn I nroff -ms I col I l p
will allow us to write:
.$ format file1 file2. ..
Now consider the slightly more complex case in which you’d like to support
either the m s or the m m macros. You could write the script like this:
t b l $2 I eqn I nroff $ 1 1 col I lp
The first argument will now follow the invocation of nrof f, and the second will
represent the filename:
$ format -ms file

However, at this point we have lost the ability to specify “all arguments,” because the
first argument is used differently than all the rest. There are several ways to handle this
situation, but we need to learn a few things first.
 0 Let the Computer Do the Dirty Work 347

Conditional Execution
Commands in a shell script can be executed conditionally using either the
i f . . . t h e n . . . e l s e or case command built into the shell. However, any condi-
tional commands require the ability to test a value and make a choice based on the
result. As its name might suggest, the t e s t command does the trick.
There are different kinds of things you can test, using various options to the com-
mand. The general form of the command is:
$ test condition

Condition is constructed from one or more options; some of the most useful are listed in
Table 12-1.

TABLE 12-1. Useful t e s t Options

Option Meaning
-d file True iffile exists and is a directory
-f file True if file exists and is a regular file
-n sl True if the length of string sl is nonzero
-r file True if file exists and is readable
-s file True if file exists and has a size greater than zero
-w file True iffile exists and is writable
-x file True iffile exists and is executable
-z sz True if the length of string sl is zero
strl = str2 True if strings strl and str2 are identical
strl != str2 True if strings strl and str2 are not identical
strl True if string strl is not the null string
nl -eq n2 True if the integers nZ and n2 are algebraically equal
(any of the comparisons -ne, -gt, -ge, -It,
and - l e may be used in place of -eq)

The t e s t command has a special form just for use in shell scripts. Instead of
using the word test, you can simply enclose condition in square brackets. The expres-
sion must be separated from the enclosing brackets by spaces.
So, for example, to return to our format script, we could write:
if [ = I1 --*I ]
then
t b l $2 I eqn I n r o f f -mm I c o l I lp
else
t b l $2 I eqn I n r o f f -ms I col I l p
fi

We've simply used the t e s t command to compare the value of two strings-the first
argument, and the string "-mm"-and executed the appropriate command line as a
result. If the strings are equal, the first command line is executed; if they are not equal,
the second line is executed instead. (Notice that there are spaces surrounding the equals
sign in the test.)
348 0 UNIX Text Processing 0

The syntax of i f . . . then . . . else clauses can get confusing. One trick is to
think of each keyword ( i f , t h e n , and e l s e ) as a separate command that can take
other commands as its argument. The else clause is optional. (That is, you can say,
“if the condition is met, do this,” and give no alternatives. If the condition is not met,
the script will simply go on to the next line, or exit if there is no next line.) The entire
sequence is terminated with the f i keyword.
After you realize that each part of the sequence is really just a separate command,
like other UNIX commands, the abbreviated form, which uses semicolons rather than
newlines to separate the commands, will also make sense:

i f condition; t h e n command; fi

An i f . .. t h e n . . . else clause allows you to make a choice between at most

two options. There is also an e l i f statement that allows you to create a sequence of
i f clauses to deal with more conditions. For example, suppose your system supports a
third macro package-ne you’ve written yourself, and called mS because it’s a super-
set of m s . (More on this in Chapter 17!) You could write the script like this:
i f [ ..$1“ = w-rnrnwv
1
t h e n t b l $ 2 I e q n I n r o f f -mm I col I lp
e l i f [ # - $ 1 r * = ~ ~ - m s ”]
then t b l $2 I eqn I n r o f f -ms I col I lp
e l i f [ -,$1,* ~ ~ - m S ”1
t h e n t b l $2 I e q n I n r o f f -mS I col I lp
fi

This syntax can get awkward for more than a few conditions. Fortunately, the shell
provides a more compact way to handle multiple conditions: the case statement.
The syntax of this statement looks complex (even in the slightly simplified form given
here):
c a s e value i n
pattern 1 command; ;
..
pattern) command; ;
esac

In fact, the statement is quite easy to use, and is most easily shown by example. We
could rewrite the previous script as follows:
c a s e $1 in
-mm) t b l $ 2 I eqn I n r o f f -mm I col 1 lp;;
-ms) t b l $2 I eqn I n r o f f -ms I col I lp;;
-mS) t b l $2 I eqn I n r o f f -mS I col I lp;;
esac

This form i s considerably more compact, especially as the number of conditions grows.
(Be sure to note the ;; at the end of each line. This is an important part of the syn-
tax.)
 0 Let the Computer Do the Dirty Work 0 349

Here's how the c a s e statement works. Each value in turn is compared (using
standard shell metacharacters like * and ?, if present) against the paftern before the
close parenthesis at the start of each line. If the pattern matches, the line is executed.
If not, the script tries again with the next line in the c a s e statement. After the value
has been compared against each case, the process starts over with the next value (if
more than one has been specified).

Discarding Used Arguments

All of the conditions we've tested for so far are mutually exclusive. What if you want
to include more than one potentially true condition in your script? The trick to dealing
with this situation requires two more shell commands: w h i l e and s h i f t .
Consider the following example. You realize that it is inefficient to pass your
files through e q n every time you use f o r m a t . In addition, you sometimes use
p i c . You want to add options to your f o r m a t shell script to handle these cases as
well.
You could decree that the macro package will always be the first argument to
your script, the name of the preprocessor the second, and the file to be formatted the
third. To delay execution of the command until all of the options have been assembled,
you can use the c a s e statement to set shell variables, which are evaluated later to
make up the actual command line. Here's a script that makes these assumptions:
case $1 in
-mm) macros="-mm"-- I .

-ms) macros="-ms"; ;
-mS) macros="-mS" ,.
--
esac
case $2 in
-E) pre="l eqn"
. .I
-P) pre="l pic
esac
t b l $3 $pre 1 nroff $macros I col I lp

But what if you don't want either preprocessor, or want both e q n and p i c ? The
whole system breaks down. We need a more general approach.
There are several ways to deal with this. For example, there is a program called
g e t o p t that can be used for interpreting command-line options. However, we will
use another techniqueaiscarding an argument after it is used, and shifting the remain-
ing arguments. This is the function of the s h i f t command.
This command finds its most elementary use when a command needs to take more
than nine arguments. There is no $ 1 0 , so a script to echo ten arguments might be
written:
echo The first nine arguments: $ 1 $ 2 $3 $ 4 $ 5 $ 6 $7 $8 $9
shift
echo The tenth argument: $9
350 0 UNlX Text Processing 0

After the shift command, the old $ 1 has disappeared, as far as the shell is con-
cerned, and the remaining arguments are all shifted one position to the left. (The old
$ 2 is the current $ 1 , and so on.) Take a moment to experiment with this if you want.
Shifting works well with conditional statements, because it allows you to test for
a condition, discard the first argument, and go on to test the next argument, without
requiring the arguments to be in a specific order. However, we still can't quite get the
job done, because we have to establish a loop, and repeat the case statement until all
of the arguments are used up.

Repetitive Execution
As we suggested at the start of this chapter, the real secret of programming is to get the
computer to do all the repetitive, boring tasks. The basic mechanism for doing this is
the loop-an instruction or series of instructions that cause a program to do the same
thing over and over again as long as some condition is true.
The while command is used like this:

while condition
do
commands
done

In the script we're trying to write, we want to repeatedly test for command-line argu-
ments as long as there are arguments, build up a command line using shell variables,
and then go ahead and issue the command. Here's how:
w h i l e [ $ # -gt 0 I
do
c a s e $1 i n
-E) eqn="l eqn";;
-P) p i c = " l -.
p i c n ,,
-*) o p t i o n s = " $ o p t i o n s $1"; ;
*) f i l e s = " $ f i l e s $1";;
esac
shift
done
t b l $ f i l e s $eqn $ p i c I nroff $options I col I lp
The special shell variable $ # always contains the number of arguments given to a
command. What this script is saying in English is: As long as there is at least one
argument

test the first argument against the following list of possibilities; if there
is a match, set the variable as instructed;
throw away the argument now that you've used it, and shift the remain-
ing arguments over one place;
 Let the Computer Do the Dirty Work 351

. decrement the shell variable $#, which contains the number of argu-
ments:
. go back to the first line following the do statement, and start over.

The loop will continue as long as the condition specified in the while statement is
met-that is, until all the arguments have been used up and shifted out of existence.
As you've no doubt noticed, to make this work, we had to account for all of the
arguments. We couldn't leave any to be interpreted in the command line because we
had to use them all up to satisfy the while statement. That meant we needed to think
about what other kinds of arguments there might be and include them in the case
statement. We came up with two possibilities: additional n r o f f options and files.
In addition, because of the pattern-matching flexibility in the case statement,
we don't need to call out each of the macro packages separately, but can just treat them
as part of a more general case. Any argument beginning with a minus sign is simply
assumed to be an n r o f f option.
You'll notice that we used a somewhat different syntax for assigning these last
two potential groups of arguments to variables:

variable= 'I $variable additional-value"

Or, as shown in the script:

options="$options $1"
f i l e s= 'I $ f i l e s $1 ''

This syntax is used to add a value to a variable. We know that we can expect at least
one option to n r o f f , so we simply add any other options to the same variable. Simi-
larly, there may be more than one filename argument. The * ) case can be executed
any number of times, each time adding one more filename to the variable.
If you want to become more familiar with how this works, you can simulate it on
the command line:
. f iles=sectl
$
$ f i l e s = " $ f i l e s sect2"
$ echo $ f i l e s
sect1 sect2

As you've seen, in the script we used the standard shell metacharacter *, which
means "any number of any characters," right in the pattern-matching part of the case
statement. You can use any of the shell metacharacters that you can type on the com-
mand line equally well in a shell script. However, be sure you realize that when you do
this, you're making assumptions-that any option not explicitly tested for in the case
statement is an n r o f f option, and that any argument not beginning with a minus sign
is a filename.
This last assumption may not be a safe one-for example, one of the filenames
may be mistyped, or you may not be in the directory you expect, and the file will not be
found. We may therefore want to do a little defensive programming, using another of
the capabilities provided by the test command:
352 0 UNlX Text Processing 0

*) if [ -f $1 ]
then
files="$files $1"
else echo "format: $1: file not found"; exit
fi;;

The [ -f 3 test checks to see whether the argument is the name of an existing file. If it
is not, the script prints an informative message and exits. (The e x i t command is
used to break out of a script. After this error occurs, we don't want to continue with the
loop, or go on to execute any commands.)
This example is also instructive in that it shows how each element in the case
statement's condition list does not need to be on a single line. A line can contain a
complex sequence of commands, separated by semicolons or newlines or both, and is
not terminated till the concluding ;; is encountered.

Setting Default Values

We've considered the case where multiple values are stored in the same variable. What
about the other extreme, where no value is stored?
If an option, such as -E for eqn, is not specified on the command line, the
variable will not be defined. That is, the variable will have no value, and the variable
substitution S e q n on the final line of the script will have no effect-it is as if it isn't
there at all.
On the other hand, it is possible to e x p o r t a variable, so that it will be recog-
nized not just in the shell that created it, but in any subshell. This means that the com-
mands:
$ eqn="l e q n " ; export eqn
$ format -ms myfile

will have the same effect as:

$ format -ms -E myfile
Although there are occasions where you might want to do this sort of thing, you don't
want it to happen unexpectedly. For this reason, it is considered good programming
practice to initialize your variables-that is, to set them to a predefined value (or in
many cases, a null value) to minimize random effects due to interaction with other pro-
grams.
To set a shell variable to a null value, simply equate it to a pair of quotation
marks with nothing in between. For example, it would be a good idea to start off the
f o r m a t script with the line:
eqn="";pic="";optionS=""

In addition to setting arguments to null values, we can also set them to default
values-that is, we can give them values that will be used unless the user explicitly
requests otherwise. Let's suppose that we want the script to invoke t r o f f by default,
but also provide an option to select n r o f f . We could rewrite the entire script like
this:
 0 Let the Computer Do the Dirty Work 0 353

e qn=II I* ;p i c = n II ; r o f f = " d i t r o f f -Tps";post=" I devps"

1p= '' l p - d l a s e I:'I
w h i l e [ $ # -gt 0 ]
do
case $1 i n
-E) eqn="l eqn";;
-P) p i c = " l p i c I*,, ..
-N) roff="nroff"; post=" I c o l " ; l p = " l p - d l i n e " ; ;
- * ) o p t i o n s = " $ o p t i o n s $1"; ;
* ) i f [ -f $1 1; t h e n
f i l e s = " $ f i l e s $1"
e l s e echo "format: $1: f i l e not found"; e x i t
fi;;
esac
shift
done
e v a l " t b l $ f i l e s $eqn $ p i c 1 $ r o f f $ o p t i o n s $ p o s t I $ 1 ~ "

The t r o f f output needs to be passed through a postprocessor before it can be sent to

a printer. (We use devps, but there are almost as many different postprocessors as
there are possible output devices.) The n r o f f output, for some printers, needs to be
passed through c o l , which is a special filter used to remove reverse linefeeds. Like-
wise, the I p command will need a "destination" option. We're assuming that the
system has a printer called l a s e r for t r o f f output, and one called l i n e for line-
printer output from n r o f f. The default case (tr o f f ) for both the postprocessor and
destination printer is set in the variables at the start of the file. The -N option resets
them to alternate values if n r o f f is being used. The e v a l command is necessary in
order for the pipes to be evaluated correctly inside a variable substitution.

What We've Accomplished

You might wonder if this script really saved you any time. After all, it took a while to
write, and it seems almost as complex to use as just typing the appropriate command
line. After all,, was it worth all that work, just so that we can type:
$ format -ma -E -P -N m y f i l e
instead of:
$ tblmyfile I eqn I pic I n r o f f -ms I l p
There are two answers to that question. First, many of the programs used to format a
file may take options of their own-ptions that are always the same, but always need
to be specified-and, especially if you're using t r o f f , a postprocessor may also be
involved. So your actual command line might work out to be something like this:
$ t b l myfile I eqn I p i c -T720 -D I d i t r o f f -ms -Tps I
> devps I lp
354 UNlX Text Processing 0

That’s considerably more to type! You could just save your most frequently used com-
binations of commands into individual shell scripts. But if you build a general tool,
you’ll find that it gives you a base to build from, and opens up additional possibilities
as you go on. For example, later in this book we’ll show how to incorporate some
fairly complex indexing scripts into f ormat-something that would be very difficult
to do from the command line. That is the far more important second reason for taking
the time to build a solid shell script when the occasion warrants.
A s this chapter goes on, we’ll show you many other useful tools you can build for
yourself using shell scripts. Many of them will use the features of the shell we intro-
duced in this section, although a few will rely on additional features we’ve yet to learn.

e x Scripts
We’ve discussed e x already in Chapter 7. As we pointed out, any command, or
sequence of commands, that you can type at ex’s colon prompt can also be saved in a
file and executed with e x ’ s :s o command.
This section discusses a further extension of this concept-how to execute e x
scripts from outside a file and on multiple files. There are certain e x commands that
you might save in scripts for use from within v i that will be of no use from the
outside-maps, abbreviations, and so on. For the most part, you’ll be using substitute
commands in external scripts.
A very useful application of editing scripts for a writer is to ensure consistency of
terminology-or even of spelling-across a document set. For the sake of example,
let’s assume that you’ve run s p e l l , and it has printed out the following list of
misspellings:
$ s p e l l sect1 sect2
chmod
ditrof f
m y f ile
thier
writeable

As is often the case, s p e l l has flagged a few technical terms and special cases it
doesn’t recognize, but it has also identified two genuine spelling errors.
Because we checked two files at once, we don’t know which files the errors
occurred in, or where in the files they are. Although there are ways to find this out, and
the job wouldn’t be too hard for only two errors in two files, you can easily imagine
how the job could grow time consuming for a poor speller or typist proofing many files
at once.
We can write an e x script containing the following commands:
g/thier/s//their/g
g/writeable/s//writable/g
wq

Then we can edit the files as follows:

 0 Let the Computer Do the Dirty Work 0 355

$ ex - sectl < exscript

$ ex - sect2 < e x s c r i p t

(The minus sign following the invocation of e x tells it to accept its commands from
standard input.)
If the script were longer than the one in our simple example, we would already
have saved a fair amount of time. However, given our earlier remarks about letting the
computer do the dirty work, you might wonder if there isn’t some way to avoid repeat-
ing the process for each file to be edited. Sure enough, we can write a shell script that
includes the invocation of e x , but generalizes it, so that it can be used on any number
of files.

Looping in a Shell Script

One piece of shell programming we haven’t discussed yet is the for loop. This com-
mand sequence allows you to apply a sequence of commands for each argument given
to the script. (And, even though we aren’t introducing it until this late in the game, it is
probably the single most useful piece of shell programming for beginners. You will
want to remember it even if you don’t write any other shell programs.)
Here’s the syntax of a for loop:

€or variable i n list

do
commands
done

For example:
for file in $ *
do
ex - $file < exscript
done

(The command doesn’t need to be indented; we indented for clarity.) Now (assuming
this shell script is saved in a file called correct), we can simply type:
$ correct s e c t l sect2

The for loop in correct will assign each argument (each f i l e in $*) to the
variable f i l e and execute the e x script on the contents of that variable.
It may be easier to grasp how the for loop works with an example whose out-
put is more visible. Let’s look at a script to rename files:
for file in $ *
do
m v $file Sfi1e.x
done

Assuming this script is in an executable file called move, here’s what we can do:
356 0 UNlX Text Processing 0

$ 1s
chO 1 ch02 ch03 move
$ move ch??
$ 1s
chOl .x ch02 .x ch03 .x move

With a little creativity, you could rewrite the script to rename the files more specifi-
cally:
for nn in $ *
do
mv ch$nn sectSnn
done
With the script written this way, you'd specify numbers instead of filenames on the
command line:
$ 1s
chOl ch02 ch03 move
$ move 01 02 03
$ 1s
sect01 sect02 sect03 move

The for loop need not take $ * (all arguments) as the list of values to be substituted.
You can specify an explicit list as well, or substitute the output of a command. For
example:
for variable in a b c d

will assign variable to a, b, c, and d i n turn. And:

for variable in 'grep -1 "Alcuin"'
will assign variable in turn to the name of each file in which grep finds the
string Alcuin.
If no list is specified:
for variable
the variable will be assigned to each command-line argument in turn, much as it was in
our initial example. This is actually not equivalent to for variable i n $ * but
to for variable i n $ @, which has a slightly different meaning. The symbols
$ * expand to $ 1 , $ 2 , $3, etc., but $ @ expands to " $ l V 1 V ',$ Z " , " $ 3 " , etc.
Quotation marks prevent further interpretation of special characters.
Let's return to our main point, and our original script:
for file in $ *
do
ex - $file < exscript
done
 0 Let the Computer Do the Dirty Work 0 357

It may seem a little inelegant to have to use two scripts-the shell script and the ex
script. And in fact, the shell does provide a way to include an editing script directly
into a shell script.

Here Documents
The operator << means to take the following lines, up to a specified string, as input to
a command. (This is often called a here document.) Using this syntax, we could
include our editing commands in correct like this:
for file in $ *
do
ex - $file << end-of-script
g/thier/s//their/g
g/writeable/s//writable/g
wq
end-of-script
done

The string end-of-script is entirely arbitrary-it just needs to be a string that

won’t otherwise appear in the input and can be used by the shell to recognize when the
here document is finished. By convention, many users specify the end of a here docu-
ment with the string EOF, or E-0-F, to indicate end offile.
There are advantages and disadvantages to each approach shown. If you want to
make a one-time series of edits and don’t mind rewriting the script each time, the here
document provides an effective way to do the job.
However, writing the editing commands in a separate file from the shell script is
more general. For example, you could establish the convention that you will always put
editing commands in a file called exscript. Then, you only need to write the
correct script once. You can store it away in your personal “tools” directory
(which you’ve added to your search path), and use it whenever you like.

ex Scripts Built by diff

A further example of the use of ex scripts is built into a program we’ve already looked
at-di f f . The -e option to dif f produces an editing script usable with either e d
or ex, instead of the usual output. This script consists of a sequence of a (add), c
(change), and d (delete) commands necessary to recreatefilel fromfile2 (the first and
second files specified on the d i f f command line).
Obviously, there is no need to completely recreate the first file from the second,
because you could do that easily with cp. However, by editing the script produced by
d i f f , you can come up with some desired combination of the two versions.
It might take you a moment to think of a case in which you might have use for
this feature. Consider this one: two people have unknowingly made edits to different
copies of a file, and you need the two versions merged. (This can happen especially
easily in a networked environment, in which people copy files between machines. Poor
coordination can easily result in this kind of problem.)
358 0 UNlX Text Processing 0

To make this situation concrete, let’s take a look at two versions of the same para-
graph, which we want to combine:
Version 1 :
The Book of Kells, now one of the treasures of the Trinity
College Library in Dublin, was found in the ancient
monastery at Ceannanus Mor, now called Kells. It is a
beautifully illustrated manuscript of the Latin Gospels,
and also contains notes on local history.
It was written in the eighth century.
The manuscript is generally regarded as the finest example
of Celtic illumination.

Version 2:
The Book of Kells was found in the ancient
monastery at Ceannanus Mor, now called Kells. It is a
beautifully illustrated manuscript of the Latin Gospels,
and also contains notes on local history.
It is believed to have been written in the eighth century.
The manuscript is generally regarded as the finest example
of Celtic illumination.
As you can see, there is one additional phrase in each of the two files. We would like
to merge them into one file that incorporates both edits.
Typing:
$ diff -e version1 version2 > exscript
will yield the following output in the file exscript:
6c
It is believed to have been written in the eighth century.

1,2c
The Book of Kells was found in the ancient

You’ll notice that the script appears in reverse order, with the changes later in the file
appearing first. This is essential whenever you’re making changes based on line
numbers; otherwise, changes made earlier in the file may change the numbering, render-
ing the later parts of the script ineffective.
You’ll also notice that, as mentioned, this script will simply recreate version 1,
which is not what we want. We want the change to line 5, but not the change to lines 1
and 2. We want to edit the script so that it looks like this:
6c
It is believed to have been written in the eighth century.

W
 0 Let the Computer Do the Dirty Work 0 359

(Notice that we had to add the w command to write the results of the edit back into the
file.) Now we can type:
$ ex - versionl < exscript

to get the resulting merged file:

The Book of Kells, now one of the treasures of the Trinity
College Library in Dublin, was found in the ancient
monastery at Ceannanus Mor, now called Kells. It is a
beautifully illustrated manuscript of the Latin Gospels,
and also contains notes on local history.
It is believed to have been written in the eighth century.
The manuscript is generally regarded as the finest example
of Celtic illumination.
Using d i f f like this can get confusing, especially when there are many changes. It is
very easy to get the direction of changes confused, or to make the wrong edits. Just
remember to do the following:

Specify the file that is closest in content to your eventual target as the first file
on the d i f f command line. This will minimize the size of the editing script
that is produced.
After you have corrected the editing script so that it makes only the changes
that you want, apply it to that same file (the first file).

Nonetheless, because there is so much room for error, it is better not to have your script
write the changes back directly into one of your source files. Instead of adding a w
command at the end of the script, add the command 1, $p to write the results to stan-
dard output. This is almost always preferable when you are using a complex editing
script.
If we use this command in the editing script, the command line to actually make
the edits would look like this:
$ ex - versionl < exscript > version3

The d i f f manual page also points out another application of this feature of the pro-
gram. Often, as a writer, you find yourself making extensive changes, and then wishing
you could go back and recover some part of an earlier version. Obviously, frequent
backups will help. However, if backup storage space is at a premium, it is possible
(though a little awkward) to save only some older version of a file, and then keep incre-
mental di f f -e scripts to mark the differences between each successive version.
To apply multiple scripts to a single file, you can simply pipe them to e x rather
than redirecting input:
cat script1 script2 script3 1 ex - oldfile
But wait! How do you get your w (or 1 , $p) command into the pipeline? You could
edit the last script to include one of these commands. But, there’s another trick that we
ought to look at because it illustrates another useful feature of the shell that many peo-
ple are unaware of.
360 0 UNlX Text Processing 0

If you enclose a semicolon-separated list of commands in parentheses, the stan-

dard output of all of the commands are combined, and can be redirected together. The
immediate application is that, if you type:
cat scriptl script2 script3; echo ‘l,$p‘ I ex - oldfile
the results of the c a t command will be sent, as usual, to standard output, and only the
results of echo will be piped to ex. However, if you type:
(cat scriptl script2 script3; echo ‘l,$p‘) I ex - oldfile

the output of the entire sequence will make it into the pipeline, which is what we want.

Stream Editing
We haven’t seen the sed program yet. Not only is it a line editor rather than a screen
editor, but it takes the process one step further: it is a “noninteractive” line editor. It
can only be used with editing scripts. It was developed in 1978 as an extension to ed
for three specific cases (according to the original documentation):

to edit files too large for comfortable interactive editing

to edit any size file when the sequence of editing commands is too complicated
to be comfortably typed in interactive mode
to perform multiple “global” editing functions efficiently in one pass through
the input

All of these are still good reasons for using sed. But these cases can be solved by the
scripting ability of ex that we have already looked at. Why learn yet another editor?
One answer lies in the third point. Because it was specifically designed to work
with scripts, sed is considerably faster than ex when used with a comparable script.
The other answer lies in sed’s unique capability to be used as an editing
filter-a program that makes edits on the fly as data is being passed through a pipe on
its way to other programs.
The sed program uses a syntax that is very similar to that used by e x , so it is
not very difficult to learn. However, there are some critical differences, which make it
inadvisable for an experienced ed or ex user to just blindly jump in.
We’re going to take a close look at sed, not as a general-purpose editor, but as a
tool to accomplish specific tasks. As a result, we won’t cover every command, but only
those that differ significantly from their ex equivalents or offer specific benefits that
we want to utilize.
First, a brief note on usage. The sed command has two forms:
sed -e command editfiles
sed - f scriptjile editfiles
The first form, using -e, allows you to specify an editing command right on the com-
mand line. Multiple -e options can be specified on the same line.
 0 Let the Computer Do the Dirty Work 0 36 1

The second form, using - f , takes the name of a script containing editing com-
mands. W e prefer this form for using sed.
In addition, you can specify an entire multiline editing script as an argument to
sed, like this:

sed ’
Editing script begins here

Editing script ends here’ editjiles

This last form is especially useful in shell scripts, as we shall see shortly. However, it
can also be used interactively. The Bourne shell will prompt for continuation lines after
it sees the first single quotation mark.
You can also combine several commands on the same line, separating them with
semicolons:
s e d -e ’commandl; command2; .. .’ editfiles
One last point: when using s e d -e, you should enclose the expression in quo-
tation marks. Although this is not absolutely essential, it can save you from serious
trouble later.
Consider the following example:
$ sed -e s/thier/their own/g myfile
The expression s / t h i e r / t h e i r own/g will work correctly in a s e d script used
with the -f option. But from the command line it will result in the message “Com-
mand garbled,” because the shell interprets the space as a separator between arguments,
and will parse the command expression as s / t h i e r / t h e i r and treat the remainder
of the line as two filenames, own/g and m y f i l e . Lacking a closing / for the s
command, s e d will complain and quit.

Differences between e x and sed

The first difference between s e d and interactive line editors like e d and ex is the
way lines are addressed. In e x , the default is to affect only a specifically addressed
line; therefore, commands like g exist to address multiple lines. The s e d program,
on the other hand, works by default on all lines, so it needs commands that allow it to
bypass selected lines. The s e d program is implicitly global. In ex, the default is to
edit the current line, and you must explicitly request global edits, or address particular
lines that you want to have edited. In sed, the default is to edit every line, and line
addresses are used to restrict the operation of the edit.
For example, consider the difference between ex and s e d in how they interpret
a command of the form:
lpatternlsloldstringlnewstringl
362 0 UNlX Text Processing 0

In e x , this means to locate the first line matching pattern and, on that line, perform the
specified substitution. In sed, the same command matches every line containing pat-
tern, and makes the specified edits. In other words, this command in sed works the
same as ex’s global flag:
g/patternls/oldstring/newstring/
In both sed and e x , a command of the form:
lpatternll, lpattern2/command
means to make the specified edits on all lines between patternl and pattern2.
Although you can use absolute line number addresses in sed scripts, you have
to remember that sed has the capability to edit multiple files at once in a stream. And
in such cases, line numbers are consecutive throughout the entire stream, rather than
restarted with each new file.
Besides its addressing peculiarities, you also need to get used to the fact that
sed automatically writes to standard output. You don’t need to issue any special com-
mands to make it print the results of its edits; in fact, you need to use a command-line
option to make it stop.
To make this point clear, let’s consider the following admittedly artificial exam-
ple. Your file contains the following three lines:
The files were writeable by thier owner, not by all.
The files were writeable by thier owner, not by all.
The files were writeable by thier owner, not by all.
You use the following editing script (in a file called edscript):
/thier/s//their/
/writeable/s//writable/
1, SP
Here are the very different results with e x and sed:
$ ex - junk < edscript
The files were writeable by their owner, not by all.
The files were writable by thier owner, not by all.
The files were writeable by thier owner, not by all.

$ sed -f edscript junk

The files were writable by their owner, not by all.
The files were writable by their owner, not by all.
The files were writable by their owner, not by all.
The files were writable by their owner, not by all.
The files were writable by their owner, not by all.
The files were writable by their owner, not by all.
The e x command, lacking the g prefix to make the edits global, applies the first line
in the script to the first line in the file, and then goes to the second line, to which it
applies the second line in the script. No edits are performed on the third line. The con-
 0 Let the Computer Do the Dirty Work 0 363

tents of the buffer are printed to standard output by the final line in the script. This is
analogous to what would happen if you issued the same commands manually in ex.
The sed command, in contrast, applies each line in the script to every line in the
file, and then sends the results to standard output. A second copy of the input is printed
to standard output by the final line in the script.
Although the same script almost works for ex and sed, the sed script can be
written more simply as:
s/thier/their/
s/writeable/writable/

Because edits are applied by default to every line, we can skip the initial pattern address
and simply give the s command. And we want to omit the print command, which
gave us the annoying second copy of the input.
There are also some special added commands that support sed’s noninteractive
operation. We will get to these commands in due course. However, in some ways, the
special commands are easier to learn than the familiar ones. The cautionary example
shown was intended to underline the fact that there is a potential for confusion when
commands that look identical produce very different results.

Some Shell Scripts Using sed

The sed command you are most likely to start with is s (or substitute) because you
can put it to work without knowing anything about sed’s advanced control structures.
Even if you learn no other sed commands, you should read this section, because this
command is easy to learn and will greatly extend your editing power.
Within the constraints just outlined, the s command works similarly to its ex
equivalent. Let’s look at several shell scripts that use sed.
First, because speed is definitely a factor when you’re making large edits to a lot
of files, we might want to rewrite the c o r r e c t script shown previously with ex as
follows:
f o r file in $ *
do
sed -f sedscr $ f i l e > $file.tmp
m v $file.tmp $file
done

This script will always look for a local editing script called sedscr, and wiIl apply
its edits to each file in the argument list given to c o r r e c t . Because sed sends the
result of its work to standard output, we capture that output in a temporary file, then
move it back to the original file.
As it turns out, there is a real danger in this approach! If there is an error in the
sed script, sed will abort without producing any output. As a result, the temporary
file will be empty and, when copied back onto the original file, will effectively delete
the original.
To avoid this problem, we need to include a test in the c o r r e c t shell script:
364 UNlX Text Processing D

for file in $ *
do
sed -f sedscr $file > $file.tmp
if [ - s $file.tmp ]
then
mv $file.tmp $file
else
echo "Sed produced an empty file."
fi
done

The [ - s ] test checks to see whether or not a file is empty-a very useful thing indeed
when you are using editing scripts.
You might want to create another simple shell script that uses sed to correct
simple errors. We'll call this one c h a n g e :
sed -e "s/$1/$2/g" $3 > $3.tmp
if [ - s $3.tmp 1
then
mv $3.tmp $ 3
else
echo "Possible error using regular expression syntax."
This script will simply change the first argument to the second in the file specified by
the third argument:
$ change mispeling misspelling rnyfile
(Because we control the actual editing script, the most likely errors could come from
faulty regular expression syntax in one of the first two arguments; thus, we changed the
wording of the error message.)

Integrating sed into format

Let's consider a brief application that shows sed in its role as a true stream editor,
making edits in a pipeline--edits that are never written back into a file.
To set the stage for this script, we need to turn back briefly to typesetting. On a
typewriter-like device (including a CRT), an em dash is typically typed as a pair of
hyphens (--). In typesetting, it is printed as a single, long dash (-). The t roff pro-
gram provides a special character name for the em dash, but it is inconvenient to type
\ (em in your file whenever you want an em dash.
Suppose we create a sed script like this:
s/--/\\ (em/g
and incorporate it directly into our format script? We would never need to worry
about em dashes-sed would automatically insert them for us. (Note that we need to
double the backslash in the string \ ( e m because the backslash has meaning to sed as
well at to t r o f f , and will be stripped off by s e d . )
 0 Let the Computer Do the Dirty Work 0 365

The format script might now look like this:

eqn="" ;pic="" ; m a c r o s z * * ~ s "
C; -
O ~ = " *,
'roff="ditroff -Tljn
sed="\ sed -e 's/--/\\(em/g'"
while [ $ X -gt 0 I
do
case $1 in
-E) eqn="l eqn";;
-P) pic="l pic n ,, .-
-N) roff="nroff";col=" 1 col";sed="" --
. I

- * ) options="$options $1"; ;
* ) if [ -f $1 1; t h e n
files="$files $1"
else echo "format: $1: file not found"; exit
fi;;
esac
shift
done
eval "cat $files Ssedltbl $eqn $picl$roff $options S c o l J l p "

(Notice that we've set up the -N option for nrof f so that it sets the sed variable
to null, because we only want to make this change if we are using t r o f f .)

Excluding Lines from Editing

Before we go any further, let's take a moment to be sure the script is complete.
What about the case in which someone is using hyphens to draw a horizontal
line? We want to exclude from the edit any lines containing three or more hyphens
together. To do this, we use the ! (don't!) command:
/---/!s/--/\ (em/g

It may take a moment to understand this syntax. It says, simply, "If you find a line
containing three hyphens together, don't make the edit." The sed program will treat
all other lines as fair game. (It's important to realize that the ! command applies to
the pattern match, not to the s command itself. Although, in this case, the effect might
seem to be the same whether you read the command as "Don't match a line containing
--- " or "Match a line containing ---, and don't substitute it," there are other cases
in which it will be very confusing if you don't read the line the same way that sed
does.)
W e might also take the opportunity to improve the aesthetics even further, by put-
ting in a very small space between the ends of the dash and the preceding and following
words, using the trof f construct \ ", which produces a 1/12-em space:
/---/!s/--/\\A\\(em\\A/g

A s it turnsout, changing hyphens to em dashes is not the only "prettying up" edit we
might want to make when typesetting. For example, some laser printers do not have a
true typeset quotation mark (" and " as opposed to " and "). If you are using an output
device with this limitation, you could use sed to change each double quotation mark
366 UNlX Text Processing 0

character to a pair of single open or close quotation marks (depending on context),

which, when typeset, will produce the appearance of a proper double quotation mark.
This is a considerably more difficult edit to make because there are many separate
cases that we need to account for using regular expression syntax. Our script might
need to look like this:
s/^"/' ./
s/"$/ ' '/
s/"? /"? /g
S/"?$/ '?/g-
s/ " / "/g
s/" / " /g
s/ I I"/ I I"/s
s/" I I/"[ I /g
s / " ) / ' ') / g
s/"l/ ''I /g
s / ( " / (. ./g
s/\ [ " / \ 1 ' * /g
s/";/* ' ; / g
s/": / * :/g
s/,"/* " / g
s/",/"./g
s / \ . " / .\ \ \ & * ' / g
s / " \ ./ * .\ \ 1& / g
s / ~ - \ \ ~ \ (em/*'\\
\ (em/g
s/\\ (em\\^"/\\ (em" /g
s/"\\ (em/" \ \ (em/g
s / \ \(em"/\\ (em' '/g
(This list could be shortened by judicious application of \ ( [ . . . ] \ ) regular
expression syntax, but it is shown in its long form for effect. Note that the symbol
I I represents a tab.)

Branching to Selective Parts of a Script

In technical books like this, it i s usually desirable to show examples in a constant-width
font that clearly shows each character as it actually appears. A pair of single quotation
marks in a constant-width font will not appear at all similar to a proper typeset double
quotation mark in a variable-width font. In short, it is not always desirable to make the
substitutions shown previously.
However, we can assume that examples will be set off by some sort of macro pair
(in this book, we used .E S and .EE, for example start and example end), and we can
use those as the basis for exclusion. There are two ways to do this:
 0 Let the Computer Do the Dirty Work 0 367

Use the ! command, as we did before.

Use the b (brunch) command to skip portions of the editing script.

Let's look at how we'd use the ! command first.

We could apply the ! command to each individual line:
/ " \ .ES/ / " \ . E E / ! s / " " / ' ' /
/ " \ .E S / , / " \ .E E / ! s / " $ / ' ' /
/"\.ES/,/"\.EE/!s/"? /"? /g

But there has to be a better way, and there is. The sed program supports the flow
control symbols { and } for grouping commands. So we simply need to write:
/"\.ES/,/"\.EE/!{
S/""/"/

s / " S / '/
J

s/"? / " ? /g

s/\\ ( e m \ \ " " / \ \ ( e m " / g

s/"\\ (em/' \ \ (em/g
s/\\ ( e m " / \ \ ( e m " / g
1
All commands enclosed in braces will be subject to the initial pattern address.
There is another way we can do the same thing. The sed program's b (brunch)
command allows you to transfer control to another line in the script that is marked with
an optional label. Using this feature, we could write the previous script like this:
/"\.ES/, /^\.EE/bend
s/""/"/
S/"$/" /
S/"? /"? /g

s / \ \( e m \ \ " * ' / \ \ (em"/g

s/"\\ ( e m / ' ' \ \ ( e m / g
s/\\ ( e m " / \ \ ( e m " / g
:end

A label consists of a colon, followed by up to eight characters. If the label is missing,

the b command branches to the end of the script. (Because we don't have anything
368 0 UNlX Text Processing 0

past this point at the moment, we don't actually need the label in this case. That is the
form we will use from now on.)
The b command is designed for flow control within the script. It allows you to
create subscripts that will only be applied to lines matching certain patterns and will not
be applied elsewhere. However, as in this case, it also gives you a powerful way to
exempt part of the text from the action of a single-level script.
The advantage of b over ! for our application is that we can more easily specify
multiple conditions to avoid. The ! symbol can apply to a single command, or can
apply to a set of commands enclosed in braces that immediately follows. The b com-
mand, on the other hand, gives you almost unlimited control over movement around the
script.
For example, if we are using multiple macro packages, there may be other macro
pairs besides . E S and . E E that enclose text that we don't want to apply the s e d
script to. So, for example, we can write:
/ A .ES/, / " . E E / b
/".E'S/, / " PE/b.
.
/ " GI/, / " G2/b.
In addition, the quotation mark is used as part of trof f ' s own comment syntax ( \ "
begins a comment), so we don't want to change quotation marks on lines beginning
with either a - or a * :
/ " .' 1 / b
It may be a little difficult to grasp how these branches work unless you keep in mind
how sed does its work:

1. It reads each line in the file into its buffer one line at a time.
2. It then applies all commands in the script to that one line, then goes to the
next line.

When a branch dependent on a pattern match is encountered, it means that if a line that
matches the pattern i s read into the buffer, the branch command will cause the relevant
portion of the script to be skippedfor that line. If a label is used, the script will con-
tinue at the label; if no label is used, the script is effectively finished for that line. The
next line is read into the buffer, and the script starts over.
The previous example shows how to exempt a small, clearly delineated portion of
a file from the action of a sed script. To achieve the opposite effect-that is, to make
a sed script affect only a small part of a file and ignore the rest-we can simply
anchor the desired edits to the enclosing pattern.
For example, if there were some edits we wanted to make only within the con-
fines of our - E S and .EE macros, and not elsewhere, we could do it like this:
/ " \ .ES/, / " \ .EE/ {
Editing commands here
1
 0 Let the Computer Do the Dirty Work 0 369

If the script is sufficiently complex that you’d rather have a more global method of
exclusion, you can reverse the sense of a branch by combining it with ! :
/A\.ES/,/*\.EE/!b

When the first line in the script is applied to each line in the input, it says: “Does the
line match the pattern? No? Branch to the end of the script. (That is, start over on the
next line of the input.) Yes? Go on to the next line in the script, and make the edits.”

Back to format
The edits we’ve shown using sed are very useful, so we want to be sure to properly
integrate them with format. Because we are now making a large series of edits
rather than just one, we need to use sed with a script file rather than a single-line
script using -e. As a result, we’ll change the variable assignment in format to:
sed=“l sed -f /usr/local/cleanup.sed”

where cleanup. sed is the name of the script containing the editing commands, and
/us r / l o c a 1 could be any generally accessible directory. We’ll add additional for-
matting cleanup commands to this file later.

Inserting Lines of Text

The sed program, like ex and v i , has commands for inserting new lines of text.
The i (insert) command adds text before the current line; a (append) adds text after
the current line. In ex, after you enter insert mode, you can type as long as you like,
breaking lines with carriage returns.* Insert mode is terminated by typing a period at
the start of a h e , followed immediately by a carriage return. In sed, you must
instead type a backslash at the end of each inserted line. Insert mode is terminated by
the first newline that is not “escaped” with a backslash in this way. For example, the
sed script:
la\
The backslash is a ubiquitous escape character used by\
many UNIX programs. Perhaps its most confusing appearance\
is at t h e end of a line, when it is used t o “hide a\
newline.” It appears t o stand alone, when in fact it is\
followed by a nonprinting character-a newline.

*The terms “carriage return” and “newline” are used somewhat loosely here. They are actually distinct
characters in the ASCII character set-quivalent to W (carriage return) and AJ (linefeed). The confusion
arises because UNIX changes the carriage return (“M) generated by the carriage return key to a linefeed (V)
on input. (That is, when you type a carriage return when editing a file, what is actually stored is a linefeed.)
On output, the linefeed is mapped to both characters-that is, a AJ in a file actually is output to the terminal
as a carriage retumfiinefeed pair (“MV).
370 UNlX Text Processing 0

will append the five lines shown in the example following line 1 in the file to which the
sed script is applied. The insert ends on the fifth line, when sed encounters a new-
line that is not preceded by a backslash.

A sed Script for Extracting Information from a File

The -n option to sed suppresses normal output and causes sed to print only the
output you explicitly ask for using the p command.
There are two forms of the p command:

rn As an absolute print command. For example:

/pattern/p
will always print the line(s) matched by pattern.
rn In combination with a substitute command, in which case the line will only be
printed if a substitution is actually made. For example:
/pattern/s /oldstring/news t r ing/gp
will not be printed if a line containing pattern is found but oldstring was not
replaced with newstring.

This becomes much clearer if you realize that a line of the form:
s / oldstring / newstring / p

is unrestricted-it matches every line in the file-but you only want to print the result
o f successful substitutions.
Using sed -n with the p command gives you a g r e p - l i k e facility with the
ability to select not just single lines but larger blocks of text.
For example, you could create a simple online quick-reference document, in
which topics are delineated by an initial heading and a distinct terminating string, as in
the following abbreviated example:
$ cat alcuin-online

O u t p u t Devices

A l c u i n requires t h e u s e o f a graphics device w i t h a t l e a s t

300 d p i r e s o l u t i o n , a n d t h e a b i l i t y t o s t o r e a t l e a s t
o n e - h a l f p a g e o f g r a p h i c s a t t h a t r e s o l u t i o n ...
%%%%

Type S t y l e s
 0 Let the Computer Do the Dirty Work 0 371

There are a number of ornamental type styles available on

many typesetters. For example, many have an Old English
font. But no typesetter currently on the market has t h e
capability of Alcuin to create unique characters in t h e
style of medieval illuminated manuscripts.
%%%%

A shell program like the following is all you need to display entries from this "full text
database":
pattern=$*
sed -n "/$pattern/, /%%%%/p" alcuin-online

(The entire argument list supplied to the command ( $ * ) is assigned to the variable
pattern, so that the user can type a smng including spaces without having to type
quotation marks.)
We'll give an example that is perhaps a bit more realistic. Consider that when
you are developing macros for use with an existing package, you may often need to
consult macros in the package you are either using or worried about affecting. Of
course, you can simply read in the entire file with the editor. However, to make things
easier, you can use a simple shell script that uses sed to print out the definition of the
desired macro. We use a version of this script on our own system, where we call it
getmac:
ma c= '* $2 ''
case $ 1 in
-ms ) file=" / u s r / lib/macros / tmac - s " ;;
-mm) file="/usr/lib/macros/mmt";;
-man) f ile="/usr/lib/macros/an"; ;
esac
sed -n -e "/''\.de *$mac/, / " \ . \ . $/p" $file
done

There are a couple of things about this script that bear mention. First, the name of a
macro does not need to be separated from the .de request by a space. The m s pack-
age uses a space, but mm and man do not. This is the reason the search pattern
includes a space followed by an asterisk (this pattern matches zero or more spaces).
Second, we use the -n option of sed to keep it from printing out the entire file.
It will now print out only the lines that match: the lines from the start of the specified
macro definition ( .de * Smac) to the . - that ends the definition.
(If you are new to regular expressions, it may be a little difficult to separate the
regular expression syntax from t rof f and shell special characters, but do make the
effort, because this is a good application of sed and you should add it to your reper-
toire.)
372 U N l X Text Processing 0

The script prints the result on standard output, but it can easily be redirected into
a file, where it can become the basis for your own redefinition. We'll find good use for
this script in later chapters.
Yet another example of how we can use sed to extract (and manipulate) infor-
mation from a file is provided by the following script, which we use to check the struc-
ture of documents we are writing.
The script assumes that t r o f f macros (in this case, the macros used to format
this book) are used to delineate sections, and prints out the headings. To make the
structure more apparent, the script removes the section macros themselves, and prints
the headings in an indented outline format.
There are three things that sed must accomplish:

1. Find lines that begin with the macro for chapter (. CH) or section headings
(.H1 or .H2).
2. Make substitutions on those lines, replacing macros with text.
3. Print only those lines.

The sed command, do. out l i n e , operates on all files specified on the command
line ( $ *). It prints the result to standard output (without making any changes within
the files themselves).
s e d -n '/''\.[CH][H121/ {
s/ "/ /g
s/"\.CH / \
CHAPTER /
s/"\-Hl/ A. /
s / ~ .H2/
\ B. /
P
}' $*

The sed command is invoked with the -n option, which suppresses the automatic
printing of lines. Then we specify a pattern that selects the lines we want to operate on,
followed by an opening brace ( { ). This signifies that the group of commands up to the
closing brace ( )) are applied only to lines matching the pattern. This construct isn't as
unfamiliar as it may look. The global regular expression of e x could work here if we
only wanted to make one substitution (g/*\. [ C H I [HlZI / s / " / / g ) . The sed
command performs several operations:

1. It removes double quotation marks.

2. It replaces the macro for chapter headings with a newline (to create a blank
line) followed by the word CHAPTER.
3. It replaces the section heading with an appropriate letter and tabbed indent.
4. It prints the line.
 0 Let the Computer Do the Dirty Work 0 373

The result of do. outline is as follows:

$ do.outline chl3/sectl
CHAPTER 13 Let t h e Computer D o the Dirty Work
A. Shell Programming
B. Stored Commands
E. Passing Arguments t o Shell Scripts
B. Conditional Execution
B. Discarding Used Arguments
B. Repetitive Execution
E. Setting Default Values
B. What We've Accomplished

Because the command can be run on a series of files or "chapters," an outline for an
entire book can be produced in a matter of seconds. We could easily adapt this script
for m s or mm section heading macros, or to include a C-level heading.

The Quit Command

The q command causes sed to stop reading new input lines (and to stop sending
them to the output). So, for example, if you only want some initial portion of your file
to be edited, you can select a pattern that uniquely matches the last line you want
affected, and include the following command as the last line of your script:
/pattern / g
After the line matching pattern is reached, the script will be terminated.*
This command is not really useful for protecting portions of a file. But, when
used with a complex sed script, it is useful for improving the performance of the
script. Even though sed i s quite fast, in an application like getmac there is some
inefficiency in continuing to scan through a large file after sed has found what it is
looking for.
So, for example, we could rewrite getmac as follows:
ma c= '' $2 *'
case $1 in
-ms) file="/usr/lib/macros/tmac. s " ; ;
-mm) file="/usr/lib/macros/mmt";;
-man) f ile="/usr/lib/macros/an"; ;
esac
shift
sed -n "
/"\.de *$mac/,/^\.\./I

*You need to be very careful not to use q in any program that writes its edits back to the original file (like
our correct shell script shown previously). After q is executed, no further output is produced. It
should not be used in any case where you want to edit the front of the file and pass the remainder through
unchanged. Using q in this case i s a very dangerous beginner's mistake.
374 UNIX Text Processing 0

P
/ A / . / . /q
}" $file
done

The grouping of commands keeps the line:

/ A / * / . /q

from being executed until sed reaches the end of the macro we're looking for. (This
line by itself would terminate the script at the conclusion of the first macro definition.)
The sed program quits on the spot, and doesn't continue through the rest of the file
looking for other possible matches.
Because the macro definition files are not that long, and the script itself not that
complex, the actual time saved from this version of the script is negligible. However,
with a very large file, or a complex, multiline script that needs to be applied to only a
smali part of the file, this script could be a significant timesaver.
For example, the following simple shell program uses sed to print out the top
ten lines of a file (much like the standard UNIX head program):
for f i l e
do
sed 1Oq $ f i l e
done

This example shows a dramatic performance gain over the same script written as fol-
lows:
for f i l e
do
s e d -n 1 , l O p $ f i l e
done

Matching Patterns across Two Lines

One of the great weaknesses of line-oriented editors i s their helplessness in the face of
global changes in which the pattern to be affected crosses more than one line.
Let me give you an example from a recent manual one of our writers was working
on. He was using the m s .BX macro (incorrectly, it turns out) to box the first letter
in a menu item, thus graphically highlighting the sequence of menu selections a user
would select to reach a given command. For example:
m a i n menu
mortfolio commands
Ovaluate portfolios
mhock factors
He had created a menu reference divided into numerous files, with hundreds of
commands coded like this:
 0 Let the Computer Do the Dirty Work 0 375

.in 5n
.BX "\s-2M\sOn \C
ain menu
.in +5n
. " \ s-2P \ s0 \ c
BX 1'

o r t f o l i o commands
.in i5n
- BX " \ s-2E \ S 0 '' \ c
valuate p o r t f o l i o s
.in i 5 n
- BX \ s-2 s \ s 0 \ c
1'

hock f a c t o r s
.in 0

Suddenly, the writer realized that the M in Main Menu should not be boxed because the
user did not need to press this key. He needed a way to remove the box around the M
if-and only if-the next line contained the string ain menu.
(A troff aside: The \c escape sequence brings text from the following line
onto the current line. You would use this, for example, when you don't want the argu-
ment to a macro to be separated from the first word on the next line by the space that
would normally be introduced by the process of filling. The fact that the .BX macro
already makes provision for this case, and allows you to supply continued text in a
second optional argument, is somewhat irrelevant to this example. The files had been
coded as shown here, the mistake had been made, and there were hundreds, perhaps
thousands, of instances to correct.)
The N command allows you to deal with this kind of problem using sed. This
command temporarily "joins" the current line with the next for purposes of a pattern
match. The position of the newline in the combined line can be indicated by the escape
sequence \ n . In this case, then, we could solve the problem with the following two-
line sed script:
/ .EX "\s-~M\sO"/N
s/.BX "\s-2M\sOn\c\nain Menu/Main Menu/

We search for a particular pattern and, after we find it, "add on" the next line using N.
The next substitution will now apply to the combined line.
Useful as this solution was, the number of cases in which you know exactly
where in the input a newline will fall are limited. Fortunately, sed goes even further,
providing commands that allow you to manipulate multiline patterns in which the new-
line may occur at any point. Let's take a look at these commands.

The Hold Space and the Pattern Space

The next set of commands-hold (h or H), get (g or G), and exchange (x>-can be
difficult to understand, especially if you have read the obscure documentation provided
with most UNIX systems. It may help to provide an analogy that reviews some of the
points we've already made about how sed works.
376 UNlX Text Processing 0

The operations of sed can be explained, somewhat fancifully, in terms of an

extremely deliberate scrivener or amanuensis toiling to make a copy of a manuscript.
His work is bound by several spacial restrictions: the original manuscript is displayed in
one room; the set of instructions for copying the manuscript are stored in a middle
room; and the quill, ink, and folio are set up in yet another room. The original
manuscript as well as the set o f instructions are written in stone and cannot be moved
about. The dutiful scrivener, being sounder of body than mind, is able to make a copy
by going from room to room, working on only one line at a time. Entering the room
where the original manuscript is, he removes from his robes a scrap of paper to take
down the first line of the manuscript. Then he moves to the room containing the list of
editing instructions. He reads each instruction to see i f it applies to the single line he
has scribbled down.
Each instruction, written in special notation, consists of two parts: a pattern and a
procedure. The scrivener reads the first instruction and checks the pattern against his
line. If there is no match, he doesn’t have to worry about the procedure, so he goes to
the next instruction. If he finds a match, then the scrivener follows the action or actions
specified in the procedure.
He makes the edit on his piece of paper before trying to match the pattern in the
next instruction. Remember, the scrivener has to read through a series of instructions,
and he reads all of them, not just the first instruction that matches the pattern. Because
he makes his edits as he goes, he is always trying to match the latest version against the
next pattern; he doesn’t remember the original line.
When he gets to the bottom of the list of instructions, and has made any edits that
were necessary on his piece of paper, he goes into the next room to copy out the line.
(He doesn’t need to be told to print out the line.) After that is done, he returns to the
first room and takes down the next line on a new scrap of paper. When he goes to the
second room, once again he reads every instruction from first to last before leaving.
This is what he normally does, that is, unless he is told otherwise. For instance,
before he starts, he can be told not to write out every line (the -n option). In this case,
he must wait for an instruction that tells him to print (p). I f he does not get that
instruction, he throws away his piece of paper and starts over. By the way, regardless
of whether or not he is told to write out the line, he always gets to the last instruction
on the list.
Let’s look at other kinds of instructions the scrivener has to interpret. First of all,
an instruction can have zero, one, or two patterns specified:

I f no pattern is specified, then the same procedure is followed for each line.
If there is only one pattern, he will follow the procedure for any line matching
the pattern.
I f a pattern is followed by a !, then the procedure is followed for all lines that
do not match the pattern.
If two patterns are specified, the actions described in the procedure are per-
formed on the first matching line and all succeeding lines until a line matches
the second pattern.
 0 Let the Computer Do the Dirty Work 0 377

The scrivener can work only one line at a time, so you might wonder how he handles a
range of lines. Each time he goes through the instructions, he only tries to match the
first of two patterns. Now, after he has found a line that matches the first pattern, each
time through with a new line he tries to match the second pattern. He interprets the
second pattern as pattern !, so that the procedure is followed only if there is no match.
When the second pattern is matched, he starts looking again for the first pattern.
Each procedure contains one or more commands or actions. Remember, if a pat-
tern is specified with a procedure, the pattern must be matched before the procedure is
executed. We have already shown many of the usual commands that are similar to
other editing commands. However, there are several highly unusual commands.
For instance, the N command tells the scrivener to go, right now, and get another
line, adding it to the same piece of paper. The scrivener can be instructed to “hold”
onto a single piece of scrap paper. The h command tells him to make a copy of the
line on another piece of paper and put it in his pocket. The x command tells him to
exchange the extra piece of paper in his pocket with the one in his hand. The g com-
mand tells him to throw out the paper in his hand and replace it with the one in his
pocket. The G command tells him to append the line he is holding to the paper in front
of him. If he encounters a d command, he throws out the scrap of paper and begins
again at the top of the list of instructions. A D command has effect when he has been
instructed to append two lines on his piece of paper. The D command tells him to
delete the first of those lines.
If you want the analogy converted back to computers, the first and last rooms in
this medieval manor are standard input and standard output. Thus, the original file is
never changed. The line on the scrivener’s piece of scrap paper is in the pattern space;
the line on the piece of paper that he holds in his pocket is in the hold space. The hold
space allows you to retain a duplicate of a line while you change the original in the pat-
tern space. Let’s look at a practical application, a sed program that searches for a
particular phrase that might be split across two lines.
As powerful as regular expressions are, there is a limitation: a phrase split across
two lines will not be matched. As we’ve shown, even though you can specify a new-
line, you have to know between which two words the newline might be found. Using
s e d , we can write instructions for general-purpose pattern matching across two lines.
N
h
s / *\n/ /
/pattern-matching syntax/{
9
P
d
l
g
D

This sed script will recognize the phrase pattern-matching syntax even when it’s in
the input file on two lines. Let’s see how the pattern space and hold space allow this to
be done.
378 0 UNlX Text Processing 0

At the start, there is one line in the pattern space. The first action (N) is to get
another line and append it to the first. This gives us two lines to examine, but there is
an embedded newline that we have to remove (otherwise we’d have to know where the
newline would fall in the pattern). Before that, we copy (h) the contents of the pattern
space into the hold space so that we can have a copy that retains the newline. Then we
replace the embedded newline (\n), and any blank spaces that might precede it, with a
single blank. (The sed command does not remove a newline when it terminates the
line in the pattern space.) Now we try to match the phrase against the contents of the
pattern space. If there is a match, the duplicate copy that still contains the newline is
retrieved from the hold space (9) and printed (p). The d command sends control back
to the top of the list of instructions so that another line is read into the pattern space,
because no further editing is attempted “on the corpse of a deleted line” (to use the
phrasing of the original sed documentation). If, on the other hand, there is no match,
then the contents of the hold buffer are replaced (9) with the contents of the pattern
space. Now we have our original two lines in the pattern space, separated by a newline.
W e want to discard the first of these lines, and retain the second in order to pair it up
with the next line. The D command deletes the pattern space up to the newline and
sends us back to the top to append the next line.
This script demonstrates the limits of flow contro1 in sed. After the first line of
input is read, the action N is responsible for all input. And, using d and D to avoid
ever reaching the bottom of the instruction list, sed does not print the line automati-
cally or clear the pattern space (regardless of the -n option). To return to our analogy,
after the scrivener enters the second room, an instruction is always telling him which
room to go to next and whether to get another line or to write it out, for as long as there
are lines to be read from the manuscript.
As we have emphasized, you can always refine a script, perfecting the way it
behaves or adding features. There are three problems with the way this script works.
First and most important, it is not general enough because it has been set up to search
for a specific string. Building a shell script around this sed program will take care of
that. Second, the program does not “go with the flow” of sed. We can rewrite it,
using the b (branch) command, to make use of sed’s default action when it reaches
the bottom of its instruction list. Last, this program always prints matching lines in
pairs, even when the search string is found in its entirety on a single line of input. We
need to match the pattern before each new line of input is paired with the previous line.
Here’s a generalized version of this sed script, called phrase, which allows
you to specify the search string as a quoted first argument. Additional command-line
arguments represent filenames.
search=$l
shift
for f i l e
do
sed ‘
/ ’ “ $ s e a r c h “ ‘ /b
N
h
-
s / *\n//
 0 Let the Computer Do the Dirty Work 0 379

/ ' "$search"' /b
4
s / *\n/ /
/ ' "$search"' / {
g
b
1
4
D' $file
done

A shell variable defines the search string as the first argument on the command line.
Now the sed program tries to match the search string at three different points. If the
search string is found in a new line read from standard input, that line is printed. We
use the b command to drop to the bottom of the list; sed prints the line and clears
the pattern space. If the single line does not contain the pattern, the next input line is
appended to the pattern space. Now it is possible that this line, by itself, matches the
search string. We test this (after copying the pattern space to the hold space) by remov-
ing the previous line up to the embedded newline. If we find a match, control drops to
the bottom of the list and the line is printed. If no match is made, then we get a copy
of the duplicate that was put in the hold space. Now, just as in the earlier version, we
remove the embedded newline and test for the pattern. If the match is made, we want
to print the pair of lines. So we get another copy of the duplicate because it has the
newline, and control passes to the bottom of the script. If no match is found, we also
retrieve the duplicate and remove the first portion of it. The delete action causes con-
trol to be passed back to the top, where the N command causes the next line to be
appended to the previous line.
Here's the result when the program is run on this section:
$ phrase "the procedure i s followed" sect3
I f a p a t t e r n i s f o l l o w e d by a \ f ( C W ! \ f P , then t h e procedure
i s f o l l o w e d f o r a l l l i n e s t h a t d o \ f I n o t \ f P match t h e
so t h a t t h e procedure i s f o l l o w e d o n l y i f t h e r e i s

In Conclusion
The examples given here only begin to touch on the power of sed's advanced com-
mands. For example, a variant of the hold command (H) appends matched lines to the
hold space, rather than overwriting the initial contents of the hold space. Likewise, the
G variant of the get command appends the contents of the hold space to the current line,
instead of replacing it. The X command swaps the contents of the pattern space with
the contents of the hold space. As you can imagine, these commands give you a great
deal of power to make complex edits.
However, it's important to remember that you don't need to understand every-
thing about sed to use it. As we've shown, it is a versatile editor, fast enough to
recommend to beginners for making simple global edits to a large set of files, yet com-
plex enough to tackle tasks that you'd never think to accomplish with an editor.
380 0 UNIX Text Processing 0

Although the syntax is convoluted even for experienced computer users, sed
does have flow control mechanisms that, given some thought and experimentation,
allow you to devise editing programs. It is easy to imagine (though more difficult to
execute) a sed script that contains editing “subroutines,” branched to by label, that
perform different actions on parts of a file and quit when some condition has been met.
Few of us will go that far, but it is important to understand the scope of the tool.
You never know when, faced with some thorny task that would take endless repetitive
hours to accomplish, you’ll find yourself saying: “Wait! I bet I could do that with
sed.”*

A Proofreading Tool You Can Build

Now let’s look at a more complex script that makes minimal use of sed but extensive
use of shell programming. It is the first example of a full-fledged tool built with the
shell that offers significantly greater functionality than any of the individual tools that
make it up.
W e call this script proof. It uses s p e l l to check for misspelled words in a
file, shows the offending lines in context, and then uses sed to make the corrections.
Because many documents contain technical terms, proper names, and so on that will be
flagged as errors, the script also creates and maintains a local dictionary file of excep-
tions that should not be flagged as spelling errors.
This script was originally published with the name s p e l l p r o o f e r in Rebecca
Thomas’s column in the June 1985 issue of UNIX World, to which it was submitted by
Mike Elola. The script as originally published contained several errors, for which we
submitted corrections. The following script, which incorporates those corrections, was
published in the January 1986 issue, and is reprinted with permission of UNIX World.
(Actually, we’ve added a few further refinements since then, so the script is not exactly
as published.)
Because the contents of the script will become clearer after you see it in action,
let’s work backward this time, and show you the results of the script before we look at
what it contains. The following example shows a sample run on an early draft of
Chapter 2. In this example, <CR> indicates that the user has typed a carriage return in
response to a prompt.
$ proof sect1
D o y o u want t o u s e a l o c a l d i c t i o n a r y ?so, e n t e r
If
t h e name o r press RETURN f o r t h e d e f a u l t d i c t i o n a r y : <CR>

Using l o c a l dictionary f i l e dict

working ...

*The preceding sections have not covered all sed commands. See Appendix A for a complete list of
sed commands.
 0 Let the Computer Do the Dirty Work 0 38 1

The word Calisthentics appears to be misspelled.

Do you want to see it in context (y or n)?
n

Press RETURN for no change or replace "Calisthentics" with:

Calisthenics

.H1 "UNIX Calisthenics"

Save corrections in "sectl" file (y or n)?
Y

The word metachacters appears to be misspelled.

Do you want to see it in context (y or n)?
n

Press RETURN for no change or replace "metachacters" with:

metacharacters

generation metacharacters. The asterisk matches any or all

Save corrections in "sectl" file (y or n)?
Y

The word textp appears to be misspelled.

Do you want to see it in context (y or n)?
Y
a directory "/work/textp" and under that directories for
each of the chapters in the book, "/work/textp/chOl",
$ cp notes /work/textp/chOl
name in the directory /work/textp/chOl.
$ Is /work/textp/ch*
$ Is /work/textp/chOl/sect?
cud /work/textp/ch03
$ book= " / work /textp I'
/work/textp

Press RETURN for no change or replace 'textp' with: <CR>

Y o u left the following words unchanged

textp

Do you wish to have any of the above words entered

into a local dictionary file (y/n)?
Y
Append to dict (y/n)?
Y
382 0 UNlX Text Processing 0

D o you wish t o b e selective (y/n)?

Y
Include textp (y/n) ?
Y

Done.
$

Now let's look at the script. Because it is more complex than anything we have looked
at so far, we have printed line numbers in the margin. These numbers are not part of
the script but are used as a reference in the commentary that follows. You will find that
the indentation of nested loops and so forth will make the program much easier to read.
1 echo "Do you want to use a local dictionary? If s o , enter"
2 echo "the name o r press RETURN for the default dictionary: "
3 read localfile
4 if [ - 2 "Slocalfile" 1; then
5 localfile=dict
6 echo Using local dictionary file Slocalfile
7 fi
8 echo "working - .."
9 touch Slocalfile
10 file1ist= " $ * " ; excused=""
11 if [ - 2 "Sfilelist" 1; then
12 echo 'Usage: proof file. . . '
13 exit 1
14 fi
15 for word in 'spell $filelist'
16 do
17 found=' grep "^$word$" $localfile
18 if [ - 2 "$found" ] ; then
19 echo
20 echo "The word $word appears to be misspelled."
21 echo " D o you want to see it in context (y or n)? "
22 read context
23 if [ "$context" = y 1; then
24 grep $word Sfilelist
25 fi
26 echo
27 echo "Press RETURN for no change or replace \"$word\" with:"
28 read correction
29 if [ -n "$correction" 1; then
30 hitlist="'grep -1 Sword $filelist' "
31 for file in $hitlist
32 do
33 echo
 0 Let the Computer Do the Dirty Work 0 383

34 sed -n -e "s/$word/$correction/gp" <$file

35 echo "Save corrections in \"$file\" file (y or n)? "
36 read response
37 if [ "$response" = y 1 ; then
38 sed -e "s/$word/$correction/g" <$file>/usr/tmp/$file
39 if test -s /usr/tmp/$file; then
40 mv /usr/tmp/$file $file
41 fi
42 fi
43 done
44 else
45 excused="$excused Sword"
46 fi
47 fi
48 done
49 echo;echo;echo
50 if [ -n "$excused" 1; then
51 echo "YOU left the following words unchanged"
52 echo $excused I tr " \ " "\012" I pr -5 -t
53 echo
54 echo "Do you wish to have any of the above words entered"
55 echo "into a local dictionary file (y/n)? ''
56 read response
57 if [ "$response" = "y" I ; then
58 if [ -n "Slocalfile" 1; then
59 echo "Append to Slocalfile (y/n)? ''
60 read response
61 if [ "$response" ! = y 1; then
62 echo "Enter new/alternate local dictionary file: ''
63 read localfile
64 fi
65 fi
66 echo
67 echo "DO you wish to be selective (y/n)? "
68 read select
69 for word in $excused
70 do
71 if [ "$select" = y 1; then
72 echo "Include $word (y/n)? "
73 read response
74 if test "$response" = y; then
75 echo $word >>$localfile
76 fi
77 else
78 echo $word >>$localfile
79 fi
384 0 UNIX Text Processing 0

80 done
81 f i
82 f i
83 echo
84 e c h o “Done.“

1-8 The UNIX programming philosophy is to create small programs as

general-purpose tools that can be joined in pipelines. Because of this,
programs generally don’t do prompting, or other “user-friendly” things
that will limit the program to interactive operation. However, there are
times, even in UNIX (!), when this is appropriate.
The shell has commands to handle prompting and reading the resulting
responses into the file, as demonstrated here. The echo command
prints the prompt, and read assigns whatever is typed in response (up
to a carriage return) to a variable. This variable can then be used in the
script.
The lines shown here prompt for the name of the local dictionary file,
and, if none is supplied, use a default dictionary in the current directory
called d i c t . In the sample run, we simply typed a carriage return, so
the variable localfile is set to d i c t .
9 If this is the first time the script has been run, there is probably no local
dictionary file, and one must be created. The touch command is a
good way to do this because if a file already exists, it will merely
update the access time that is associated with the file (as listed by Is
-1). If the file does not exist, however, the touch command will
create one.
Although this line is included in the script as a sanity check, so that the
script will work correctly the first time, it is preferable to create the
local dictionary manually, at least for large files. The spe 1 1 program
tends to flag as errors many words that you want to use in your docu-
ment. The p r o o f script handles the job of adding these words to a
local dictionary, but doing this interactively can be quite time-
consuming. It is much quicker to create a base dictionary for a docu-
ment by redirecting the output of spell to the dictionary, then editing
the dictionary to remove authentic spelling errors and leave only the
exception list. The errors can then be corrected with p r o o f without
the tedium of endlessly repeating n for words that are really not errors.
If you use this script, you should run spell rather than p r o o f on
the first draft of a document, and create the dictionary at that time. Sub-
sequent runs of p r o o f for later drafts will be short and to the point.
10-14 In these lines, the script sets up some variables, in much the same way
as we’ve seen before. The lines:
 0 Let the Computer Do the Dirty Work 0 385

f i l e li st="$ * "
i f [ - 2 " $ f i l e l i s t " 1; t h e n
e c h o "Usage: proof f i l e . .. 'I

exit 1
f i

have much the same effect as the test of the number of arguments
greater than zero that we used in earlier scripts. If f i l e l i s t is a
null string, no arguments have been specified, and so it is time to
display an error message and end the program, using the shell's e x i t
command.
15 This line shows a feature of the shell we've seen before, but it is still
worthy of note because it may take a while to remember. The output of
a command enclosed in backquotes ("> can be substituted for the argu-
ment list of another command. That is what is happening here; the out-
put of the s p e l l command is used as the pattern list of a f o r loop.
17-18 You'll notice that s p e l l still flags all of the words it finds as errors.
But the f o r loop then uses g r e p to compare each word in the list
generated by spell with the contents of the dictionary. Only those
words nor found in the dictionary are submitted for correction.
The pattern given to g r e p is "anchored" by the special pattern-
matching characters ,,and $ (beginning and end of line, respectively),
so that only whole words in the dictionary are matched. Without these
anchors, the presence of the word d i t r o f f in the list would prevent
the discovery of misspellings like t r o f .
20-25 Sometimes it is difficult to tell beforehand whether an apparent misspel-
ling is really an error, or if it is correct in context. For example, in our
sample run, the word t e x t p appeared to be an error, but was in fact
part of a pathname, and so correct. Accordingly, p r o o f (again using
grep) gives you the opportunity to look at each line containing the
error before you decide to change it or not.
As an aside, you'll notice a limitation of the script. If, as is the case in
our example, there are multiple occurrences of a string, they must all be
changed or left alone as a set. There is no provision for making indivi-
dual edits.
26-48 After a word is offered as an error, you have the option to correct it or
leave it alone. The script needs to keep track of which words fall into
each category, because words that are not corrected may need to be
added to the dictionary.
If you do want to make a correction, you type it in. The variable
c o r r e c t i o n will now be nonzero and can be used as the basis of a
test ( t e s t -n). If you've typed in a correction, p r o o f first checks
the files on the command line to see which ones (there can be more than
386 0 UNIX Text Processing 0

one) can be corrected. ( g r e p -1 just gives the names of files in

which the string is found into the variable hitlist, and the script
stores the names.) The edit is then applied to each one of these files.
35 Just to be on the safe side, the script prints the correction first, rather
than making any edits. (The -n option causes s e d not to print the
entire file on standard output, but only to print lines that are explicitly
requested for printing with a p command. Used like this, sed per-
forms much the same function as g r e p , only we are making an edit at
the same time.
37-42 If the user approves the correction, s e d is used once again, this time to
actually make the edit. You should recognize this part of the script.
Remember, it is essential in this application to enclose the expression
used by s e d in quotation marks.
50-84 If you?ve understood the previous part of the shell script, you should be
able to decipher this part, which adds words to the local dictionary. The
t r command converts the spaces separating each word in the
excused list into camage returns. They can then be printed in five
tab-separated columns by p r . Study this section of the program until
you do, because it is an excellent example of how UNIX programs that
appear to have a single, cut-and-dry function (or no clear function at all
to the uninitiated) can be used in unexpected but effective ways.
 C H A P T E R
rn m

13

The a w k Programming Language

A program is a solution to a problem, formulated in the syntax of a particular language.

It is a small step from writing complex editing scripts with sed to writing programs
with awk, but it is a step that many writers may fear to take. “Script” is less loaded a
term than “program” for many people, but an editing script is still a program.
Each programming language has its own “style” that lends itself to performing
certain tasks better than other languages. Anyone can scan a reference page and quickly
learn a language’s syntax, but a close examination of programs written in that language
is usually required before you understand how to apply this knowledge. In this sense, a
programming language is simply another tool; you need to know not only how to use it
but also when and why it is used.
We recommend that you learn more than one programming language. We have
already looked at a number of different programs or scripts written for and executed by
the shell, e x , and sed. A s you learn the awk programming language, you will notice
similarities and differences. Not insignificantly, an awk script looks different from a
shell script. The awk language shares many o f the same basic constructs as the shell’s
programming language, yet awk requires a slightly different syntax. The awk
program’s basic operations are not much different from sed’s: reading standard input
one line at a time, executing instructions that consist of two parts, pattern and pr-n-
crdul-e, and writing to standard output.
More importantly, awk has capabilities that make it the tool of choice for certain
tasks. A programming language i s itself a program that was written to solve certain
kinds o f problems for which adequate tools did not exist. The awk program was
designed for text-processing applications, particularly those in which information is
structured in records and fields. The major capabilities o f awk that we will demon-
strate in upcoming pages are as follows:

definable record and field structure

conditional and looping constructs

387 .
388 0 UNIX Text Processing 0

assignment, arithmetic, relational, and logical operators

numeric and associative arrays
formatted print statements
built-in functions

A quick comparison of a single feature will show you how one programming language
can differ from another. You will find it much easier to perform arithmetic operations
in awk than in the shell. To increment the value of x by 1 using the shell, you'd use
the following line:
x='expr $x + 1'

The e x p r command is a UNIX program that is executed as a separate process return-

ing the value of its arguments. In awk, you only have to write:
++X

This is the same as x = x + 1. (This form could also be used in awk.)

Invoking auk
The awk program itself i s a program that, like sed,runs a specified program on lines
of input. You can enter awk from the command line, or from inside a shell script.
$ awk 'program' files

Input is read a line at a time from one or more files. The program, enclosed in single
quotation marks to protect it from the shell, consists of pattern and procedure sections.
If the pattern is missing, the procedure is performed on all input lines:
$ awk ' {print}' sample Prints all lines in sample file

The procedure is placed within braces. If the procedure is missing, lines matching the
pattern are printed:
$ awk '/programmer's guide/' sample Prints lines matching pattern
in sample file

The awk program allows you to specify zero, one, or two pattern addresses, just
like sed. Regular expressions are placed inside a pair of slashes (/). In awk, pat-
terns can also be made up of expressions. An expression (or a primary expression so as
not to confuse it with a regular expression) can be a string or numeric constant (for
example, red or I), a variable (whose value is a string or numeric), or a function (which
we'll look at later).
You can associate a pattern with a specific procedure as follows:
/patternl/ {
procedure 1
I
/pattern2/ {
 0 The auk Programming Language 0 389

procedure 2
I
{ procedure 3 1

Like sed, only the lines matching the particular pattern are the object of a procedure,
and a line can match more than one pattern. In this example, the third procedure is per-
formed on all input lines. Usually, multiline a w k scripts are placed in a separate file
and invoked using the -f option:
$ a w k -f awkscript sample

Records and Fields

Perhaps the most important feature of awk is that it divides each line of input into
fields. In the simplest case, each field contains a single word, delimited by a blank
space. The a w k program allows you to reference these fields by their position in the
input line, either in patterns or procedures. The symbol $ 0 represents the entire input
line. $1, $2, . . . refer, by their position in the input line, to individual fields.
We'll demonstrate some of these capabilities by building an a w k program to
search through a list of acronyms in a file. Each acronym is listed along with its mean-
ing. If we print the first field of each line, we'll get the name of the acronym:
$ awk '{print $l}' sample
BAS IC
CICS
COBOL
DBMS
GIGO
GIRL

W e can construct a useful program that would allow you to specify an acronym and get
its description. We could use awk just like g r e p :
$ a w k '/BASIC/'
sample
BASIC Beginner's All-Purpose Symbolic Instruction Code

However, there are three things we'd like to do to improve this program and make
better use of awk's capabilities:

1. Limit the pattern-matching search.

2. Make the program more general and not dependent on the particular acro-
nym that is the subject of the search.
3. Pnnt only the description.
390 UNlX Text Processing 0

Testing Fields
The pattern as specified will match the word BASIC anywhere on the line. That is, it
might match BASIC used in a description. To see if the first field ( $ 1 ) matches the pat-
tern, we write:
$1 == “BASIC”

The symbol == is a relational operator meaning “equal to” and is used to compare the
first field of each line with the string BASIC. You could also construct this test using a
given regular expression that looks for the acronym at the beginning of the line.
$1 - /“BASIC/

The pattern-matching operator - evaluates as true if an expression ($1) matches a reg-

ular expression. Its opposite, ! -, evaluates true if the expression does not match the
regular expression.
Although these two examples look very similar, they achieve very different
results. The relational operator == evaluates true if the first field is BASIC but false if
the first field is BASK, (note the comma). The pattern-matching operator - locates
both occurrences.
Pattern-matching operations must be performed on a regular expression (a string
surrounded by slashes). Variables cannot be used inside a regular expression with the
exception of shell variables, as shown in the next section. Constants cannot be
evaluated using the pattern-matching operator.

Passing Parameters from a Shell Script

Our program is too specific and requires too much typing. W e can put the awk script
in a file and invoke it with the -f option. Or we can put the command inside a shell
script, named for the function it performs. This shell script should be able to read the
first argument from the command line (the name of the acronym) and pass it as a
parameter to awk. We’ll call the shell script awkronym and set it up to read a file
named acronyms. Here’s the simplest way to pass an argument into an awk pro-
cedure:
$ cat awkronym
a w k ‘ $ 1 == search‘ search=$l acronyms

Parameters passed to an awk program are specified after the program. The search
variable is set up to pass the first argument on the command line to the auk program.
Even this gets confusing, because $ 1 inside the awk program represents the first field
of each input line, while $ 1 in the shell represents the first argument supplied on the
command line. Here’s how this version of the program works:
$ awkronym CICS
C I C S Customer Information Control System

By replacing the search string BASIC with a variable (which could be set to the string
CICS or BASIC), we have a program that is fairly generalized.
 The awk Programming Language 0 391

Notice that we had to test the parameter as a string ( $ 1 == s e a r c h ) . This is

because we can’t pass the parameter inside a regular expression. Thus, the expressions
“ $ 1 - / s e a r c h / ” or “$1 - search” will produce syntax errors.
As an aside, let’s look at another way to import a shell variable into an a w k pro-
gram that even works inside a regular expression. However, it looks complicated:
search=$l
awk ’ $ 1 -
/‘“$search”’ / ’ acronyms

This program works the same as the prior version (with the exception that the argument
is evaluated inside a regular expression.) Note that the first line of the script makes the
variable assignment before a w k is invoked. In the a w k program, the shell variable is
enclosed within single, then double, quotation marks. These quotes cause the shell to
insert the value of $ s e a r c h inside the regular expression before it is interpreted by
awk. Therefore, a w k never sees the shell variable and evaluates it as a constant string.
You will come upon situations when you wish it were possible to place awk vari-
ables within regular expressions. As mentioned in the previous section, pattern match-
ing allows us to search for a variety of occurences. For instance, a field might also
include incidental punctuation marks and would not match a fixed string unless the
string included the specific punctuation mark. Perhaps there is some undocumented
way of getting an a w k variable interpreted inside a regular expression, or maybe there
is a convoluted work-around waiting to be figured out.

Changing the Field Separator

The a w k program is oriented toward data arranged in fields and records. A record is
normally a single line of input, consisting of one or more fields. The field separator is a
blank space or tab and the record separator is a newline. For example, here’s one
record with five fields:
CICS Customer Information Control System

Field three or $ 3 is the string Information. In our program, we like to be able to print
the description as a field. It is obvious that we can’t just say p r i n t $ 2 and get the
entire description. But that is what we’d like to be able to do.
This will require that we change the input file using another character (other than
a blank) to delimit fields. A tab is frequently used as a field separator. We’ll have to
insert a tab between the first and second fields:
$ cat acronyms
awk Aho, Weinstein & Kernighan
BASIC Beginner‘s All-Purpose Symbolic Instruction Code
CICS Customer Information Control System
COBOL Common Business Orientated Language
DBMS Data Base Management System
GIGO Garbage In, Garbage O u t
GIRL Generalized Information Retrieval Language
392 0 UNlX Text Processing 0

You can change the field separator from the command line using the -F option:
$ awk -F"I 1'' '$1 = search { p r i n t $ 2 ) ' search=$l acronyms

Note that I I is entered by typing a double quotation mark, pressing the TAB
key, and typing a double quotation mark. This makes the tab character (represented in
the example as I I ) the exclusive field separator; spaces no longer serve to
separate fields. Now that we've implemented all three enhancements, let's see how the
program works:
$ awkronym G I G O
G a r b a g e In, Garbage O u t

System Variables
The awk program defines a number of special variables that can be referenced or reset
inside a program. See Table 13-1.

TABLE 13-1. awk System Variables

System Variable Meaning

FILENAME Current filename
FS Field separator (a blank)
NF Number of fields in the current record
NR Number of the current record
OFS Output field separator (a blank)
ORS Output record separator (a newline)
RS Record separator (a newline)

The system variable F S defines the field separator used by awk. You can set
F S inside the program as well as from the command line.
Typically, if you redefine the field or record separator, it is done as part of a
BEGIN procedure. The BEGIN procedure allows you to specify an action that is per-
formed before the first input line is read.
BEGIN { F S = "I I" I
You can also specify actions that are performed after all input is read by defining an
END procedure.
The awk command sets the variable NF to the number of fields on the current
line. Try running the following awk command on any text file:
$ awk ' { p r i n t $NF}' t e s t
If there are five fields in the current record, NF will be set to five; SNF refers to the
fifth and last field. Shortly, we'll look at a program, double,that makes good use of
this variable.
 The awk Programming Language 0 393

Looping
The awkronym program can print field two because we restructured the input file and
redefined the field separator. Sometimes, this isn’t practical, and you need another
method to read or print a number of fields for each record. If the field separator is a
blank or tab, the two records would have six and five fields, respectively.
BASIC Beginner’s All-Purpose Symbolic Instruction Code
CICS Customer Information Control S y s t e m

It is not unusual for records to have a variable number of fields. To print all but the
first field, our program would require a loop that would be repeated as many times as
there are fields remaining. In many awk programs, a loop is a commonly used pro-
cedure.
The while statement can be employed to build a loop. For instance, if we want
to perform a procedure three times, we keep track of how many times we go through
the loop by incrementing a variable at the bottom of the loop, then we check at the top
of the loop to see if that variable is greater than 3. Let’s take an example in which we
print the input line three times.
I i = l
while ( i <= 3) {
print
++i
1
1
Braces are required inside the loop to describe a procedure consisting of more than a
single action. Three operators are used in this program: = assigns the value 1 to the
variable i; <= compares the value of i to the constant 3; and ++ increments the
variable by 1. The first time the while statement is encountered, i is equal to 1.
Because the expression i <= 3 is true, the procedure is performed. The last action
of the procedure is to increment the variable i. The while expression is true after
the end of the second loop has incremented i to 3. However, the end of the third loop
increments i to 4 and the expression evaluates as false.
A for loop serves the same purpose as a while loop, but its syntax is more
compact and easier to remember and use. Here’s how the previous while statement
is restructured as a for loop:
for (i = 1; i <= 3; i++)
print

The for statement consists of three expressions within parentheses. The first expres-
sion, i = 1, sets the initial value for the counter variable. The second expression
states a condition that is tested at the top of the loop. (The while statement tested
the condition at the bottom of the loop.) The third expression increments the counter.
Now, to loop through remaining fields on the line, we have to determine how
many times we need to execute the loop. The system variable N F contains the number
of fields on the current input record. If we compare our counter (i) against N F each
time through the loop, we’ll be able to tell when all fields have been read:
394 UNlX Text Processing 0

for (i = 1; i <= NF; i++)

We will print out each field (Si), one to a line. Just to show how awk works, we'll
print the record and field number before each field.
awk ' { for (i = 1; i <= NF; i++)
print NR":"i, $i } ' $ *

Notice that the print statement concatenates NR, a colon, and i. The comma produces
an output field separator, which is a blank by default.
This program produces the following results on a sample file:
1:l awk
1:2 Aho,
1:3 Weinstein
1:4 &
1:5 Kernighan
2:l BASIC
2 :2 Beginner' s
2:3 All-Purpose
2:4 Symbolic
2 :5 Instruction
2:6 Code

Symbolic is the fourth field of the second record. You might note that the sample
file is acronyms, the one in which we inserted a tab character between the first and
second fields. Because we did not change the default field separator, awk interpreted
the tab or blank as a field separator. This allows you to write programs in which the
special value of the tab is ignored.

Conditional Statements
Now let's change our example so that when given an argument, the program returns the
record and field number where that argument appears.
Essentially, we want to test each field to see if it matches the argument; if it does,
we want to print the record and field number. We need to introduce another flow con-
trol construct, the i f statement. The i f statement evaluates an expression-if true, it
performs the procedure; if false, it does not.
In the next example, we use the i f statement to test whether the current field is
equal to the argument. If it is, the current record and field number are printed.
awk ' { for (i = 1; i <= NF; i++) {
if (Si == search) {
print NR":"i
1
1
} ' search=$l acronyms

This new procedure prints 2:1 or 3.4 and isn't very useful by itself, but it demonstrates
that you can retrieve and test any single field from any record.
 The awk Programming Language 0 395

The next program, double, checks if the first word on a line is a duplicate of
the last word on the previous line. W e use double in proofing documents and it
catches a surprisingly common typing mistake.
awk '
N F > O (
if ($1 == lastword) {
p r i n t NR " : double '' $1
1
lastword = $NF
1' $1

When the first line of input is read, if the number of fields is greater than 0, then the
expression in the i f statement is evaluated. Because the variable lastword has
not been set, it evaluates to false. The final action assigns the value of SNF to the
variabIe lastword. ($NF refers to the last field; the value of N F is the number of
the last field.) When the next input line is read, the first word is compared against the
value of lastword. If they are the same, a message is printed.
double s e c t 1
15: double t h e
32: double a
This version of double is based on the program presented by Kernighan and Pike in
The UNIX Programming Environment. (Writer's Workbench now includes this pro-
gram.) Kernighan and Pike's program also checks for duplicate words, side-by-side, in
the same line. You might try implementing this enhancement, using a for loop and
checking the current field against the previous field. Another feature of Kernighan and
Pike's double i s that you can run the program on more than one file. To allow for
additional files, you can change the shell variable from $ 1 to $ * but the record or
line number printed by NR will correspond to consecutive input lines. Can you write a
procedure to reset NR to 0 before reading input from a new file?

Arrays
The double program shows us how we can retain data by assigning it to a variable.
In awk, unlike several other programming languages, variables do not have to be ini-
tialized before they are referenced in a program. In the previous program, we evaluated
lastword at the top, although it was not actually assigned a value until the bottom of
the program. The awk program initialized the variable, setting it to the null string or
0, depending upon whether the variable is referenced as a string or numeric value.
An array is a variable that allows you to store a list of items or elements. An
array is analogous to a restaurant menu. Each item on this menu is numbered:

#1 tuna noodle casserole

#2 roast beef and gravy
#3 pork and beans
396 0 UNlX Text Processing 0

One way of ordering roast beef is to say simply “Number 2.” Using ordinary vari-
ables, you would have had to define a variable two and assign it the value roasr beef
and gravy. An array is a way of referencing a group of related values. This might be
written :
menu [choice]

where menu is the name of the array and choice is the subscript used to reference
items in the array. Thus, menu [ 11 is equal to tuna noodle casserole. In awk, you
don’t have to declare the size of the array; you only have to load the array (before
referencing it). If we put our three menu choices on separate lines in a file, we could
load the array with the following statement:
menu[NRl = $0

The variable NR, or record number, i s used as the subscript for the array. Each input
line i s read into the next element in the array. We can print an individual element by
referring to the value of the subscript (not the variable that set this value).
print menu[3]

This statement prints the third element in the array, which is pork and beans. If we
want to refer to all the elements of this array, we can use a special version of the for
loop. It has the following syntax:

for (element in array)

This statement can be used to descend the array to print all of the elements:
for (choice in menu)
print menu [choice]

Each time through the loop, the variable choice is set to the next element in the
array. The menu array is an example of an array that uses a numeric subscript as an
index to the elements.
Now, let’s use arrays to increase the functionality of awkronym. Our new ver-
sion will read acronyms from a file and load them into an array; then we’ll read a
second file and search for the acronyms. Basically, we’re reading one input file and
defining keywords that we want to search for in other files. A similar program that
reads a list of terms in a glossary might show where the words appear in a chapter.
Let’s see how it works first:
$ awkronym sect1
exposure t o BASIC programming.
in COBOL and t a k e advantage of a DBMS environment.
in COBOL and t a k e advantage of a DBMS environment.
Of t h e high-level languages, BASIC is probably

Let’s look at the program carefully.

 The awk Programming Language 0 397

awk ' {
if ( FILENAME == "acronyms" ) {
acro desc[NR] = $ 1
~

next
1
for ( name in acro-desc )
for (i = 1; i <= NF; i++)
if (Si == acro-desc[name]) {
p r i n t $0
l
}' acronyms $ *
The current filename is stored in the system variable FILENAME. The procedure
within the first conditional statement is only performed while input is taken from
acronyms. The n e x t statement ends this procedure by reading the next line of input
from the file. Thus, the program does not advance beyond this procedure until input is
taken from a different file.
The purpose of the first procedure is to assign each acronym ( $ 1 ) to an element
of the array acro desc; the record number (NR)indexes the array.
In the secondhalf of the program, we start comparing each element in the array to
each field of every record. This requires two for loops, one to cycle through the array
for each input line, and one to read each field on that line for as many times as there are
elements in the array. An i f statement compares the current field to the current ele-
ment of the array; if they are equal, then the line is printed.
The line is printed each time an acronym is found. In our test example, because
there were two acronyms on a single line, the one line is duplicated. To change this,
we could add n e x t after the p r i n t statement.
What if we changed awkronym so that it not only scanned the file for the acro-
nym, but printed the acronym with the description as well? If a line refers to BASIC,
we'd like to add the description (Beginner's All-Purpose Symbolic Insti-uction Code).
We can design such a program for use as a filter- that prints all lines, regardless of
whether or not a change has been made. To change the previous version, we simply
move the p r i n t statement outside the conditional statement. However, there are other
changes we must make as well. Here's the first part of the new version.
awk ' {
if ( FILENAME == "acronyms" ) {
split ($O,fields,"l I *' )
acro-desc [ fields [ 11 ]=fields [ 2 ]
next
I
The records in acronyms use a tab as a field separator. Rather than change the field
separator, we use the s p l i t function (we'll look at the syntax of this function later
on) to give us an array named fields that has two elements, the name of the acronym
and its description. This numeric array is then used in creating an associative array
named acro-desc. An associative array lets us use a string as a subscript to the ele-
ments of an array. That is, given the name of the acronym, we can locate the element
398 0 UNlX Text Processing 0

corresponding to the description. Thus the expression a c r o-desc [ G I G O ] will

access Garbage I n , G a r b a g e O u t .
Now let’s look at the second half of the program:
for ( name in acro desc ) ~

for (i = 1; i <= NF; i++)

if (Si == name) {
Si = Si ’1 ( “ a c r o desc [name]‘I) ~ ’*
1
print $0

Just like the previous version, we loop through the elements of the array and the
fields for each record. At the heart of this section is the conditional statement that tests
if the current field (Si) is equal to the subscript of the array (name). If the value of
the field and the subscript are equal, we concatenate the field and the array element. In
addition, we place the description in parentheses.
It should be clear why we make the comparison between S i and name, and not
a c r o-desc [name] ; the latter refers to an element, while the former refers to the
subscript, the name of the acronym.
If the current field ( S i ) equals BASIC and the index of the array (name) is the
string BASIC, then the value of the field is set to:
BASIC (Beginner’s All-Purpose Symbolic Instruction Code)

For this program to be practical, the description should be inserted for the first
occurrence of an acronym, not each time. (After we’ve inserted the description of the
acronym, we don’t need the description any more.) We could redefine that element in
the array after we’ve used it:
acro-desc[name] = name

In this instance, we simply make the element equal to the subscript. Thus,
a c r o desc [BASIC] is equal to Beginner’s All-Purpose Symbolic Instruction Code
at the beginning of the procedure, and equal to BASIC if a match has been made. There
are two places where we test the element against the subscript with the expression
“ ( a c r o -d e s c [ n a m e ] ! = name).” The first place is after the f o r loop has
read in a new element from a c r o-desc; a conditional statement ensures that we
don’t scan the next input record for an acronym that has already been found. The
second place is when we test $ i to see if it matches name; this test ensures that we
don’t make another match for the same acronym on that line.
if (Si == name && acro-desc[name] ! = name)

This conditional statement evaluates a compound expression. The & & (and) boolean
operator states a condition that both expressions have to be true for the compound
expression to be true.
Another problem that we can anticipate i s that we might produce lines that exceed
80 characters. After all, the descriptions are quite long. W e can find out how many
characters are in a string, using a built-in awk function, l e n g t h . For instance, to
evaluate the length of the current input record, we specify:
 The awk Programming Language 0 399

length ($0)

The value of a function can be assigned to a variable or put inside an expression and
evaluated.
i f (length(S0) > 70) {
i f (i > 2)
$ i = ''\ne* S i
if (i+l < NF )
$(i+l) = "\n" $ ( i + l )
}

The length of the current input record is evaluated after the description has been con-
catenated. If it is greater than 70 characters, then two conditions test where to put the
newline. The first procedure concatenates a newline and the current field; thus we only
want to perform this action when we are not near the beginning of a line (field greater
than 2). The second procedure concatenates the newline and the next field (i+l) so
that we check that we are not near the end of the line. The newline precedes the field
in each of these operations. Putting it at the end of the field would result in a new line
that begins with a space output with the next field.
Another way to handle the line break, perhaps more efficiently, is to use the
length function to return a value for each field. By accumulating that value, we
could specify a line break when a new field causes the total to exceed a certain number.
We'll look at arithmetic operations in a later section.
Here's the full version of awkronyms:
awk ' (
if ( FILENAME == "acronyms" ) {
split ($O,fields,"l I ")
acro-desc[fields[l]J=fields[2]
next
1
for ( name in acro-desc )
if (acro-desc[name] ! = name)
for (i = 1; i <= NF; i + + )
if (Si == name & & acro-desc[name] ! = name) {
S i = Si " ("acro_desc[name]")"
acro-desc[name] = name
if (length(S0) > 7 0 ) {
if (i > 2 )
$ i = "/n" Si
i f (i+l < NF)
$(i+l) = "\n" S (i+l)
}
1
print $0
}' acronyms $ *

And here's one proof that it works:

400 0 UNIX Text Processing 0

$ cat sectl
Most users of microcomputers have had some
exposure to BASIC programming.
Many data-processing applications are written
in COBOL and take advantage of a DBMS environment.
C, the language of the UNIX environment,
is used by systems programmers.
Of the high-level languages, BASIC is probably
the easiest to learn, and C is the most difficult.
Nonetheless, you will find the fundamental programming
constructs common to most languages.

$ awkronym sectl
Most users of microcomputers have had some
exposure to
BASIC (Beginner‘s All-Purpose Symbolic Instruction Code)
programming. Many data-processing applications are
written in COBOL (Common Business Orientated Language)
and take advantage of a
DBMS (Data Base Management System) environment.
C , the language of the UNIX environment,
is used by systems programmers.
Of the high-level languages, BASIC is probably
the easiest to learn, and C is the most difficult.
Nonetheless, you will find the fundamental programming
constructs common to most languages.
Notice that the second reference to BASIC has not been changed. There are other
features we might add to this program. For instance, we could use auk’s pattern-
matching capabilities so that we don’t make the change on lines containing macros, or
on lines within pairs of certain macros, such as .D S / .DE.
Another version of this program could trademark certain terms or phrases in a
document. For instance, you’d want to locate the first occurrence of UNIX and place
\ ( r g after it.

a w k Applications
A shell program is an excellent way to gather data interactively and write it into a file
in a format that can be read by auk. We’re going to be looking at a series of pro-
grams for maintaining a project log. A shell script collects the name of a project and
the number of hours worked on the project. An awk program totals the hours for each
project and prints a report.
The file d a y is the shell script for collecting information and appending it to a
file named d a i l y in the user’s home directory.
 0 The auk Programming Language 0 40 1

$ cat /usr/local/bin/day
case $ # in
0 ) echo "Project: \c";read proj;echo "Hours: \c";read hrs;;
1) proj=$l; echo "Hours: \c"; read hrs;;
2) proj=$l;hrs=$2;;
esac
set 'who am i'; name=$l; month=$3; day=$4;
echo $name" \ t " $month $day'' \ t " $ hr s" \ t " $proj>>$HOME / dai1y
The c a s e statement checks how many arguments are entered on the command line. If
an argument is missing, the user is prompted to enter a value. Prompting is done
through a pair of statements: echo and r e a d . The e c h o command displays the
prompt on the user's terminal; \ c suppresses the carriage return at the end of the
prompt. The r e a d command waits for user input, terminated by a carriage return, and
assigns it to a variable. Thus, the variables p r o j and hrs are defined by the end of
the c a s e statement.
The s e t command can be used to divide the output of a command into separate
arguments ( $ 1 , $ 2 , $ 3 . . . ). By executing the command who am i from within
s e t , we supply the user's name and the day's date automatically. The e c h o com-
mand is used to write the information to the file. There are four fields, separated by
tabs. (In the Bourne shell, the escape sequence \tproduces a tab; you must use quota-
tion marks to keep the backslash from being stripped off by the shell.)
Here's what d a i l y contains for one user at the end of a week:
$ cat /usr/fred/daily
fred Aug 4 7 Course Development
fred Aug 5 4 Training class
fred Aug 5 4 Programmer's Guide
fred Aug 6 2 Administrative
fred Aug 6 6 Text-processing book
fred Aug 7 4 Course Development
fred Aug 7 4 Text-processing book
fred Aug 8 4 Training class
fred Aug 8 3 Programmer's Guide
There are nine records in this file. Obviously, our input program does not enforce con-
sistency in naming projects by the user.
Given this input, we'd like an awk program that reports the total number of
hours for the week and gives us a breakdown of hours by project. At first pass, we
need only be concerned with reading fields three and four. We can total the number of
hours by accumulating the value of the third field.
total += $3
The += operator performs two functions: it adds $ 3 to the current value of t o t a l
and then assigns this value to t o t a l . It is the same as the statement:
402 0 UNlX Text Processing 0

total = total + $3

We can use an associative array to accumulate hours ( $ 3 ) by project ( $ 4 ) .

hours[$4] += $3
Each time a record is read, the value of the third field is added to the accumulated value
of p r o j e c t [ $ 4 ] .
We don’t want to print anything until all input records have been read. An END
procedure prints the accumulated results. Here’s the first version of t o t :
awk ’
BEGIN ( FS=”I I” 1
t
total += $ 3
hours[$4] += $3
1
END {
for (project in hours)
print project, hours[project]
print
print ”Total Hours:”, total
} ‘ $HOME/daily

Let’s test the program:

$ tot
Course Development 11
Administrative 2
Programmer‘s Guide 7
Training class 8
Text-processing book 10

Total Hours: 38

The program performs the arithmetic tasks well, but the report lacks an orderly format.
It would help to change the output field separator (OFS) to a tab. But the variable
lengths of the project names prevent the project hours from being aligned in a single
column. The awk program offers an alternative print statement, p r i n t f , which is
borrowed from the C programming language.

Formatted Print Statements

The p r i n t f statement has two parts: the first i s a quoted expression that describes
the format specifications; the second is a sequence of arguments such as variable names.
The two main format specifications are % s for strings and %d for decimals. (There
are additional specifications for octal, hexadecimal, and noninteger numbers.) Unlike
the regular print statement, p r i n t f does not automatically supply a newline. This
can be specified as \ n. A tab i s specified as \ t .
 0 The auk Programming Language 0 403

A simple p r i n t f statement containing string and decimal specifications is:

printf "%s\t%d\n", project, hours [project]

First project is output, then a tab (\t), the number of hours, and a newline ( b ) .
For each format specification, you must supply a corresponding argument.
Unfortunately, such a simple statement does not solve our formatting problem.
Here are sample lines that it produces:
Course Development 11
Administrative 2
Programer's Guide 7
We need to specify a minimum field width so that the tab begins at the same position.
The print f statement allows you to place this specification between the % and the
conversion specification. You would use % - 2 0 s to specify a minimum field width of
20 characters in which the value is left justified. Without the minus sign, the value
would be right justified, which is what we want for a decimal value.
END {
f o r (project in hours)
printf "%-2Os\t%2d\n", project, hours [projectI
printf "\n\tTotal Hours :\t%2d\n", total
1
Notice that literals, such as the string Total Hours, are placed in the first part, with
the format specification.
Just as we use the END procedure to print the report, we can include a B E G I N
procedure to print a header for the report:
BEGIN { FS="l I"
print f " % 2 0 s % s \ n \ n " ,"PROJECT ", " HOURS ''
1
This shows an alternative way to handle strings. The following formatted report i s
displayed:

PROJECT HOURS

Course Development 11
Administrative 2
Programmer's Guide 7
Training class 8
Text-processing book 10

Total Hours: 38
404 0 UNlX Text Processing 0

Defensive Techniques
After you have accomplished the basic task of a program-and the code at this point is
fairly easy to understand-it is often a good idea to surround this core with "defen-
sive" procedures designed to trap inconsistent input records and prevent the program
from failing. For instance, in the t o t program, we might want to check that the
number of hours is greater than 0 and that the project description i s not null for each
input record. W e can use a conditional expression, using the logical operator & & .
$ 3 > 0 & & $4 ! = "" (
procedure
1
Both conditions must be true for the procedure to be executed. The logical operator
& & signifies that if both conditions are true, the expression is true.
Another aspect of incorporating defensive techniques is error handling. In other
words, what do we want to have happen after the program detects an error? The previ-
ous condition is set up so that if the procedure is not executed, the next line of input is
read. In this example the program keeps going, but in other cases you might want the
program to print an error message and halt if such an error i s encountered.
However, a distinction between "professional" and "amateur" programmers
might be useful. We are definitely in the latter camp, and we do not always feel com-
pelled to write 100% user-proof programs. For one thing, defensive programming is
quite time consuming and frequently tedious. Second, an amateur is at liberty to write
programs that perform the way he or she expects them to; a professional has to write for
an audience and must account for their expectations. Consider the possible uses and
users of any program you write.

awk and n r o f f / t r o f f
It is fairly easy to have an awk program generate the necessary codes for form reports.
For instance, we enhanced the t o t program to produce a t r o f f-formatted report:
a w k ' BEGIN { F S = " I I"
print ' I . ce"
print ".I3 ''
print "PROJECT A C T I V I T Y REPORT"
print ".R"
.
print " sp 2"
1
NR == 1
begday = $2
1
$3 > 0 & & $ 4 ! = " " I
hours[$4] += $3
total += $ 3
endday = $2
logname = $1
1
 0 The auk Programming Language 0 405

END {
printf "Writer: %s\n", logname
print l'.sp"
p r i n t f "Period: % s t o %s\n",begday, endday
p r i n t lr.spn
p r i n t f ''% 2 0 s % s \ n \ n " , PROJECT ", " HOURS"
p r i n t " sp" .
.
p r i n t " n f ''
p r i n t " na" .
f o r ( p r o j e c t i n hours)
p r i n t f "%-20s\t%2d\n1', p r o j e c t , h o u r s [ p r o j e c t ]
p r i n t ' I . sp"
p r i n t f "Total H o u r s :\t %2d\n", t o t a l
p r i n t " sp" -
} $HOME/daily

We incorporated one additional procedure in this version to determine the weekly

period. The start date of the week is taken from the first record (NR == 1). The last
record provides the final day of the week.
As you can see, auk doesn't mind if you mix p r i n t and p r i n t f state-
ments. The regular p r i n t command i s more convenient for specifying literals, such
as formatting codes, because the newline is automatically provided. Because this pro-
gram writes to standard output, you could pipe the output directly to n r o f f / t r o f f .
You can use awk to generate input to t b l and other t rof f preprocessors
such as pic.

Multiline Records
In this section, we are going to take a look at a set of programs for order tracking. We
developed these programs to help operate a small, mail-order publishing business.
These programs could be easily adapted to track documents in a technical publications
department.
Once again, we used a shell program, t a k e . orders, for data entry. The pro-
gram has two purposes: The first i s to enter the customer's name and mailing address
for later use in building a mailing list. The second is to display seven titles and prompt
the user to enter the title number, the number of copies, and the price per copy. The
data collected for the mailing list and the customer order are written to separate files.
Two sample customer order records follow:
C h a r l o t t e Smith
P.0 N61331 87 Y 0 4 5 Date: 03/14/87
#1 3 7.50
#2 3 7.50
X3 1 7.50
f 4 1 7.50
#7 1 7.50
406 0 UNlX Text Processing 0

Martin S. Rossi
P.0 NONE Date: 03/14/87
#1 2 7.50
82 5 6.75

These are multiline records, that is, a newline is used as the field separator. A blank
line separates individual records. For most programs, this will require that we redefine
the default field separator and record separator. The field separator becomes a newline,
and the record separator is null.
BEGIN { F S = "\n"; RS = 8- 11 }

Let's write a simple program that multiplies the number of copies by the price. We
want to ignore the first two lines of each record, which supply the customer's name, a
purchase order number, and the date of the order. We only want to read the lines that
specify a title. There are a few ways to do this. With awk's pattern-matching capabil-
ities, we could select lines beginning with a hash (#) and treat them as individual
records, with fields separated by spaces.
awk ' / " # / {
amount = $2 * $ 3
p r i n t f " % s %6.2f\n1', $0, amount
next
1
{print}' $*

The main procedure only affects lines that match the pattern. It multiplies the second
field by the third field, assigning the value to the variable amount. The p r i n t f
conversion %f prints a floating-point number; 2 specifies a minimum field width of 6
and a precision of 2. Precision is the number of digits to the right of the decimal point;
the default for % f is 6. We print the current record along with the value of the vari-
able amount. If a line is printed within this procedure, the next line is read from
standard input. Lines not matching the pattern are simply passed through. Let's look at
how addem works:
$ addem orders
C h a r l o t t e Smith
P.0 N 6 1 3 3 1 87 Y 0 4 5 Date: 03/14/87
#l 3 7.50 22.50
#2 3 7.50 22.50
#3 1 7.50 7.50
#4 1 7.50 7.50
#7 1 7.50 7.50

M a r t i n S. Rossi
P.0 NONE Date: 03/14/87
#1 2 7.50 15.00
#2 5 6.75 33.75
 The awk Programming Language 407

Now, let's design a program that reads multiline records and accumulates order infor-
mation for a report. This report should display the total number of copies and the total
amount for each title. We also want totals reflecting all copies ordered and the sum of
all orders.
W e know that we will not be using the information in the first two fields of each
record. However, each record has a variable number of fields, depending upon how
many titles have been ordered. First, we check that the input record has at least three
fields. Then a for loop reads all of the fields beginning with the third field:
NF >= 3 {
for ( i = 3; i <= NF; ++i)

In database terms, each field has a value and each value can be further broken up into
subvalues. That is, if the value of a field in a multiline record is a single line, subvalues
are the words on that line. You have already seen the s p l i t function used to break
up an input record; now we'll see it used to subdivide a field. The s p l i t function
loads any string into an array, using a specified character as the subvalue separator.
s p 1 it ( string, array, separator)

The default subvalue separator is a blank. The s p l i t function returns the number of
elements loaded into the array. The string can be a literal (in quotation marks) or a
variable. For instance, let's digress a minute and look at an isolated use of s p l i t .
Here's a person's name and title with each part separated by a comma:
title="George Travers, Research/Development, Alcuin Inc."

We can use s p l i t to divide this string and print it on three lines.

need = split (title, name, " , " )
print ".ne lr, need
for ( part in name)
print name [part]

This procedure prints each part on a separate line. The number of elements in the array
(3) is saved in the variable need. This variable is passed as an argument to an .ne
request, which tells t r o f f to make sure there are at least three lines available at the
bottom of the page before outputting the first line.
The awk program has twelve built-in functions, as shown in Table 13-2. Four of
these are specialized arithmetic functions for cosine, sine, logarithm, and square root.
The rest of these functions manipulate strings, (You have already seen how the
length function works.) See Appendix A for the syntax of these functions.
Going back to our report generator, we need to split each field into subvalues.
The variable S i will supply the value of the current field, subdivided as elements in the
array o r d e r .
sv = split($i, order)
if ( s v == 3 ) {
procedure
1
else print "Incomplete Record"
408 0 UNlX Text Processing 0

TABLE 13-2. awk Built-in Functions

Function Description
cos Cosine
exP Exponent
g e t 1i n e Read input line
index Return position of substring in string
int Integer
length Length of string
109 Logarithm
sin Sine
split Subdivide string into array
sprintf Format string like p r i n t f
sqrt Square root
substr Substring extraction

The number of elements returned by the function is saved in the sv variable. This
allows us to test that there are three subvalues. If there are not, the e l s e statement is
executed, printing the error message to the screen.
Next, we assign the individual elements of the array to a specific variable. This is
mainly to make it easier to remember what each element represents.
t i t l e = order[l]
c o p i e s = o r d e r [21
p r i c e = order[3]

Then a group of arithmetic operations are performed on these values.

amount = c o p i e s * p r i c e
t o t a l -v o l += c o p i e s
t o t a l -a m t += amount
vel [ t i t l e ] += c o p i e s
amt [ t i t l e ] += amount
These values are accumulated until the last input record is read. The E N D procedure
prints the report.
Here's the complete program:
auk ' BEGIN ( F S = "\n"; RS = "" )
NF >= 3
f o r ( i = 3; i <= NF; + + i ) {
sv = s p l i t ( $ i , o r d e r )
i f (sv == 3 ) I
t i t l e = order[l]
copies = order[2]
p r i c e = order[3]
amount = c o p i e s * p r i c e
t o t a l -v o l += c o p i e s
 0 The auk Programming Language 0 409

t o t a l -amt += amount
v o l [ t i t l e ] += c o p i e s
amt [ t i t l e ] += amount
1
else p r i n t "Incomplete Record"
l
1
END {
p r i n t f " % 5 s \ t % 10 s \ t % 6 s \ n \ n ' I , " T I T L E " , \
"COPIES SOLD", "TOTAL"
for ( t i t l e i n v o l )
p r i n t f "%5s\t%lOd\t$%7.2f\n", t i t l e , vol [ t i t l e ] , \
amt [ t i t l e ]
**printfwn n ei%S\nn,

p r i n t f "\t%s%4d\t$%7.2f\n","Total
- _____________ n

" , t o t a l -v o l , t o t a l -a m t } '
$*

In awk, arrays are one dimensional; a two-dimensional array stores two elements
indexed by the same subscript. You can get a pseudo two-dimensional array in awk
by defining two arrays that have the same subscript. W e only need one for loop to
read both arrays.
The addemup file, an order report generator, produces the following output:
$ addemup orders.today
TITLE COPIES SOLD TOTAL

#1 5 $ 37.50
#2 8 $ 56.25
#3 1 $ 7.50
#4 1 $ 7.50
#7 1 $ 7.50
----_________
T o ta1 16 $ 116.25

After you solve a programming problem, you will find that you can re-use that approach
in other programs. For instance, the method used in the awkronym program to load
acronyms into an array could be applied in the current example to read the book titles
from a file and print them in the report. Similarly, you can use variations of the same
program to print different reports. The construction of the next program is similar to
the previous program. Yet the content of the report is quite different.
awk ' B E G I N { F S = "\n*'; RS = " "
pr i n t f '' % - 15 s \ t % 10 s \ t % 6 s \ n \ n I*, "CUSTOMER " , COP IE S SOLD '' , \
" TOTAL"

1
NF >= 3 {
c u s t o m e r = $1
t o t a l-v o l = 0
t o t a l -amt = 0
41 0 UNlX Text Processing 0

for (i = 3; i <= NF; ++i){

split (Si, order)
title = order[l]
copies = order[2]
price = order [ 3 1
amount = copies * price
total-vol += copies
total-amt += amount
}
printf "\t%s%4d\t$%7.2f\n1',"Tota1 ",total -vo1,total -amt)'
1' $*
In this program, named summary, we print totals for each customer order. Notice
that the variables t o t a l vol and t o t a l -amt are reset to 0 whenever a new
record is read. In the previous program, these values accumulated from one record to
the next.
The summary program, reading a multiline record, produces a report that lists
each record on a single line:
$ summary orders
CUS TOMER COPIES SOLD TOTAL

J. Andrews 7 $ 52.50
John Peterson 4 $ 30.00
Charlotte Miller 11 $ 82.50
Dan Aspromonte 105 $ 787.50
Valerie S. Rossi 4 $ 30.00
Timothy P. Justice 4 $ 30.00
Emma Fleming 25 $ 187.50
Antonio Pollan 5 $ 37.50
Hugh Blair 15 $ 112.50

Testing Programs
Part of writing a program is designing one or more test cases. Usually this means creat-
ing a sample input file. It is a good idea to test a program at various stages of develop-
ment. Each time you add a function, test it. For instance, if you implement a condi-
tional procedure, test that the procedure is executed when the expression is true; test
what happens when it is faIse. Program testing involves making sure that the syntax is
correct and that the problem has been solved.
When awk encounters syntax errors it will tell you that it is "bailing out." Usu-
ally it will print the line number associated with the error. Syntax errors can be caused
by a variety of mistakes, such as forgetting to quote strings or to close a procedure with
a brace. Sometimes, it can be as minor as an extra blank space. The awk program's
 The a w k Programming Language 0 41 1

error messages are seldom helpful, and a persistent effort is often required to uncover
the fault.
You might even see a UNIX system error message, such as the dreadful declara-
tion:
Segmentation fault-core dumped.

Not to worry. Although your program has failed badly, you have not caused an earth-
quake or a meltdown. An image of “core” memory at the time o f the error is saved or
dumped in a file named core. Advanced programmers can use a debugging program
to examine this image and determine where in memory the fault occurred. We just
delete core and re-examine our code.
Again, check each construct as you add it to the program. I f you wait until you
have a large program, and it fails, you will often have difficulty finding the error. Not
only that, but you are likely to make unnecessary changes, fixing what’s not broken in
an attempt to find out what is.
Checking that you have solved the problem you set out to tackle is a much larger
issue. After you begin testing your program on larger samples, you will undoubtedly
uncover “exceptions,” otherwise known as bugs. In testing the awkronym program,
we discovered an exception where an acronym appeared as the last word in the sen-
tence. I t was not “found” because o f the period ending the sentence. That is, awk
found that BASIC and BASIC. were not equal. This would not be a problem if we could
test the search string as a regular expression but we have to test the array variable as a
literal string.
Programming is chiefly pragmatic in its aims. You must judge whether or not
specific problems merit writing a program or if certain exceptions are important enough
to adapt the general program to account for them. Sometimes, in large public programs
as well as small private ones, bugs just become part of the program’s known behavior,
which the user i s left to cope with as best as he or she can. The bug found in awkro-
nym is a common enough problem, so it is necessary to implement a fix.
The fix for the awkronym bug does not involve awk at all. We run a a sed
script before the awkronym program to separate punctuation marks from any word. It
converts a punctuation mark to a field containing garbage characters. Another script
processes the awkronym output and strips out these garbage characters. The example
below shows how both scripts are used as bookends for the awkronym program.
sed ’ s / \ ( . .* \ ) \ ( [ ., !; 1 \ ) /\I @ @ @ \ 2 / g $‘ * I
awk ’ {
program lines
) ‘acronyms - I
sed ’ s / @ @ @ \ ( [ -,! ;I \ ) / \ l / g ’
 C H A P T E R

14

Writing n r o f f and t r o f f Macros

The n r o f f and t rof f formatters include a powerful macro definition and substitu-
tion capability. As we suggested when macros were first discussed in Chapter 4, they
are a good way to combine frequently used sequences of formatting requests into a sin-
gle instruction. But after working with the m s and mm macro packages, you must
know that macros are more than that.
Macros are an essential part of nrof f and t r o f f-you cannot escape them if
you want to make serious use of the formatter. Precisely because macros are so essen-
tial, many users never learn to fully use them. The most obviously useful macros are
already included in the existing macro packages, whose complex internal control struc-
tures make them difficult to understand and modify.
The purpose of this chapter is to introduce the fundamental nrof f and t rof f
requests that are used for creating macros. You’ll learn the basics in this chapter. Then,
in later chapters we can examine how to write macros for specific purposes, without
having to make continual asides to introduce a new request.
Chapter 15 describes additional requests for creating special effects (such as pic-
tures) with your macros, and Chapters 16 through 18 discuss how to go beyond writing
individual macros and how to develop or extend an entire macro package.

Comments
Before we start, we’ll introduce the syntax for inserting comments into your macro
definitions. Macros can get quite confusing, so we advise you to put in comments that
explain what you are doing. This will help immensely when you go back weeks or
months later to modify a macro you have written.

412 .
 0 Writing nrof f and trof f Macros 0 41 3

A line beginning with the sequence:

- \"
will not be interpreted or printed by the formatter. Any part of a line following the
sequence \ will be treated the same way. For example:
.\** O'Reilly C Associates, Inc. custom macro set
. \ " Last modified 4/25/87
.de IZ \ " Initialization macro

Note that there is an important difference between:

.\" A full line comment

and:
\ " A partial line comment
If you simply start the sequence \ at the margin, the formatter will insert a blank line
I'

into the output, because this sequence by itself does not suppress newline generation.
(Note that comments can be used at any time, not just in macros. You can write
notes to yourself in your input file and they will never appear in the output. But if you
accidentally type the sequence \ " in your file, the remainder of the line on which it
appears will disappear from the output.)

Defining Macros
As we've already discussed, use the .de request to define a macro:
.de AB \ " Define macro A B
Requests andlor text of macro here
..
There are also requests to remove or add to existing macros. The .r m request
removes a macro:
.rm PQ \ " Remove macro P Q
You may sometimes want to define a macro for local use, and remove it when you are
done. In general, though, this is an advanced request you will not use often.
The .am request appends to the end of an existing macro. It works just like
.de but does not overwrite the existing contents:
.am DS \ " Append t o the existing definition of DS
.ft cw
..
At first, you may think that this request has only limited usefulness. However, as you
work more with macros, you will find unexpected uses for it. We'll mention a few of
these in later chapters.
414 0 UNlX Text Processing 0

Macro Names
A macro name can be one or two characters, and can consist of any character(s), not
just alphanumeric characters. For example:
.de " ( \ " Macro used internally whose name, we hope,
\ v ' never has t o be remembered
You can even use control characters in macro names. Names can be uppercase or
lowercase, or any combination of the two, and uppercase and lowercase are distinct.
For example, the four names . g m , .GM, .gM, and . G m can all be used without
conflict.
If you are starting from scratch, you can use whatever macro or number register
names you like except for the names of existing formatter requests. However, if you
are adding macros to an existing package, you have to work around the existing names,
because creating a new macro with the same name as an old one will discard the previ-
ously read-in definition.
This is not as easy as it sounds, because macro packages include internal macro,
string, and number register definitions that are not visible to the casual user. You may
be surprised when your new macro makes some other part of the package go haywire.
(In an attempt to forestall this problem, most macro developers give odd, unmnemonic
names to internally called macros. However, collisions still can and do occur.)

Finding the Names of Existing Macros

Before you start adding macros to an existing package, it's a good idea to print the
names of all existing macros.
There are two ways to do this. The . p m request will print (in blocks of 128
characters) the names and sizes of all macros defined in a given run of the formatter.
So, for example, creating a file containing the single request:
- Pm
and formatting it like this:
$ n r o f f -ms pmfile
will print on the screen a list of all the macros defined in the m s macro package. (The
output could also be redirected to a file or printer.)
However, macro names are drawn from the same pool as string names (see the
next example), so it might be better to search for macro or string definitions using
g r e p et al, like this:
$ grep ' "\ .d [ e s i a ] ' macrofiles I cut -fl, 2 -d' ' I sort I uniq
( g r e p will select all lines beginning with either .de, .ds, .di, or .da; c u t
will select only the first two space-separated fields on each of those lines; sort and
u n i q together will produce a sorted list consisting of only one copy of each line. Note
that for -mm,which does not use a space before the macro name, you would need to
specify c u t - f l only. You will also need to substitute for macrofiles the actual
filenames containing the macros of interest.)
 0 Writing nroff and troff Macros 0 41 5

You should do the same for number registers:

$ sed -n -e 's/.*.nr *\ ( . . \ ) .*/\l/p' macrofiles I sort I uniq
here, because we can't rely on number registers being set at the start of a line, as we
can with macro definitions. The one-line sed script included here saves the first two
. -
nonspace characters (. ) following the string n r , and substitutes them for the rest
of the line (i.e., it throws away the rest of the line).
You could also just grep for an individual macro, string, or number register
name before you use it! Or you could take the easy way, and check Appendix B, where
we've listed all the names in each of the packages.
In addition to looking for conflicting names, you may also need to look for con-
flicting usage, or to understand in detail the operation of a macro you are intending to
call within a new macro you are writing.
To do this, you can simply read in the entire macro definition file with the editor
and search for what you want. However, to make things easier, we use the getmac
shell script described in Chapter 12 to print out the definition of the desired macro. The
script prints the result on standard output, which can easily be redirected into a file,
where it can become the basis for your own redefinition.

Renaming a Macro
If you do find a conflict, you can rename macros that have already been defined. The
.r n macro renames an existing macro:
.rn "( H1 \ " Rename A ( t o H1; e a s i e r t o remember

The old name will no longer work. You must use the new name to invoke the macro.
A good trick that you can sometimes pull off with .r n is to temporarily redefine
a macro (without ever modifying its contents). For example, the m s macros include a
macro to draw a box around a paragraph; however, these macros do not leave any space
above or below the box. We can add some like this:
.rn E1 b l \" Rename B 1 t o b l
.de E1 \" Now r e d e f i n e B1
. s p .5 \" A d d some s p a c e b e f o r e t h e b o x i s d r a w n
. bl \" E x e c u t e t h e old d e f i n i t i o n
..
. r n B2 b 2 \" Rename B2 t o b2
.de E2 \" Now r e d e f i n e B2
- b2 \" Execute t h e old d e f i n i t i o n
.sp -5 \" Add some s p a c e a f t e r t h e b o x i s drawn
..
This only works for adding extra control lines before or after the current contents of the
macro. Remember it, though, because this trick may come in handy if you don't want
to (or can't) directly modify a macro in one of the existing packages, but do want a
slightly different effect.
41 6 UNlX Text Processing 0

Macro Arguments
The simplest kind of macro is a sequence of stored commands, starting with a .de
request and ending with the two dots (..) at the beginning of a line.
However, as you've seen when you've used mm and m s , macros can take argu-
ments, and can act differently depending on various conditions. It's also possible for a
macro to save information and pass it to other macros to affect their operation. An
understanding of how to do these things is essential if you plan any serious macro
design.
A macro can take up to nine arguments and can use them in any way. Arguments
are described positionally by the character sequences \ \ $ 1 through \ \ $9".
For example, we could define a very simple .B macro to boldface a single argu-
ment:
.de B \" Macro t o boldface first argument
\fB\\Sl\fP

Or, we could write a simple paragraph macro that, instead of having a fixed indent,
might take a numeric argument to specify the depth of the indent:
.de P I \" Simple paragraph macro
- SP
.ne 2 \" Prevent widows
.ti \\$l \" Indent t o t h e depth specified by first
.. \" argument

As you can see in the first example, you can print an argument in your text. Or, shown
in the second example, you can use it inside the macro as an argument to one or more
of the requests that make up the macro.
Notice that there is nothing intrinsic about a macro that causes a break. The .B
macro, for instance, can be placed in the input file as in the following example:
There are a number of ways t o
.B embolden
text.

As long as filling is in effect, it will produce exactly the same output as:
There are a number of ways t o \fBembolden\fP text.

Macro arguments are separated by spaces. If you want to include an explicit space in
an argument, you should enclose the entire string in quotation marks, like this:

* Actually,
the sequences are \ $ l through \ $ 9 , with only a single backslash. But for reasons to be
described shortly, you always need at least two backslashes.
 Writing n r o f f and t r o f f Macros 0 41 7

T h e r e a r e a number o f ways t o
.B “make t e x t s t a n d o u t . ”

If you didn’t enclose the phrase make text stand out in quotation marks, a single word,
make, would have been interpreted as the first argument, the next word, text, as the
second argument, and so on. This wouldn’t cause a program error-there is no require-
ment that arguments to a macro be used by that macro-but the unused arguments
would simply disappear from the output. As shown here, the entire phrase is treated as
a single argument.
To actually print a quotation mark inside a macro argument, double it. For exam-
ple:
.B “ T h e Q u o t e ( ” ” ) Character’’

will produce:

I 1
I The Quote (”) Character I
You’ve probably recognized that the syntax for specifying arguments by position is very
similar to that used with shell scripts. You might wonder, though, about backslashes,
which are used in the shell to prevent interpretation of a special character. In fact, they
serve the same function in t r o f f .
The n r o f f and t r o f f formatters always read a macro at least twice: once
when they read the definition (and store it away for later use), and once when they
encounter it in the text. At the time the macro is defined, there are no arguments, so it
is essential to prevent the formatter from doing any argument substitution.
When the macro definition is read, the formatter operates in what is referred to (in
the Ni-off/Troff User’s Manual) as copy mode. That is, none of the requests are exe-
cuted; they are simply copied (in this case, presumably into memory) without interpreta-
tion. The exception is that various escape sequences that may have a different value at
macro definition time than at macro execution time (most notably \n, for interpolating
number registers, \ *, for interpolating strings, and \ $, for interpolating arguments)
are executed, unless you suppress interpretation with a preceding backslash. (Other
escape sequences are also interpreted, but because they have fixed values, this makes no
difference to the action of the macro.)
A backslash prevents interpretation of the character that follows it by sacrificing
itself. The backslash tells the formatter: “Take me but let the next guy go.” Each time
the character sequence is read, the backslash is stripped off-that is, \ \ is actually
stored as \. (You can think of \ as saying “I really mean. . . . So in the shell, for
”

example, if you want to use an asterisk literally, rather than as a filename expansion
metacharacter, you write \*-that is, “I really mean *.” In a similar way, \ \ says
“I really mean backslash.”)
When macro definitions are nested inside one another, you will need to add addi-
tional backslashes to get what you want. The true argument interpolation escape
sequence is \ Sn, rather than \ \ $n; the extra backslash is needed because the first one
is stripped when the macro is interpreted in copy mode. The same rule applies when
you want to interpolate the value of a number register or a string in a macro definition.
Think through the number of times the definition will be read before it is executed, and
41 8 0 UNlX Text Processing 0

specify the appropriate number of backslashes, so that you get the actual value used at
the point where you need it. A failure to understand this will cause more frustration
than almost any other error when you are writing macros.
In the example of the .B macro, the sequences \ f B and \ f P did not need to
be escaped, because t r o f f could just as easily interpret them at the time the macro is
defined. However, the macro would also work if they were specified with double
backslashes-it is just that the interpretation of the codes would take place when the
macro was used.

Nested Macro Definitions

We said previously that a macro definition begins with a . d e request and ends with
two dots (. .). This is a simplification. The .de request takes an alternate terminator
as an optional second argument. This feature allows you to create nested macro defini-
tions.
.de M1 \“ S t a r t first macro
.de M2 !! \“ S t a r t second macro
*!! \“ End second macro
_. \” End f i r s t m a c r o

You can also nest macros by delaying interpretation of the .. on the second macro:
.de M 1 \ ” S t a r t f i r s t macro
.de M2 \ ” S t a r t second macro
\ \ -- \ “ End second macro
... \” End f i r s t m a c r o

For example, a group of related macros for producing a certain type of document
might be nested inside a “master” macro. A user would have to invoke the master
macro, indicating document type, to make the other macros available for use. Nested
macros could be used to provide alternate versions of the same set of macros within a
single macro package.

Conditional Execution
One of the most powerful features of n r o f f and t r o f f ’ s macro programming
language is its facility for conditional execution. There are three conditional execution
requests: .i f , .i e (if else), and .el (else). The .i f request is used for a single
condition. (“If the condition is met, do this; otherwise, simply go to the next line.”)
The .i e and .e 1 requests are used as a pair, testing a condition and then performing
either one action or the other. (“If the condition is met, do this; otherwise, do that.”)
 0 Writing nrof f and trof f Macros 0 419

Predefined Conditions
There are a number of different conditions that can be tested with .i f and .i e . The
simplest looks to see if a predefined condition is true or false. There are four prede-
fined conditions, as listed in Table 14- 1 .

TABLE 14-1. Built-in Conditions

I Condition True if
0 Current page number is odd
e Current page number is even
n The file is being formatted by n r o f f
t The file is being formatted by t r o f f

For example, in a page bottom macro, to print the page number in the outside
comer, you might write:
.if o .tl " ' % ' \ " If odd, put page number in right corner
.if e .tl ' % " ' \ " If even, put page number in left corner
(The .t 1 request prints three-part titles, at the left, center, and right of the page. And,
within this request, the % character always prints the current page number. We'll
explain these two items in detail later, when we look at how to write a complete page
transition macro. For right now, we just want to understand how the conditions them-
selves work.)
Because the two conditions, odd and even, are mutually exclusive, you could also
write:
.ie o .tl ' " % ' \ " If odd, put page number in right corner
.el .tl ' % " ' \ " Otherwise, put it in left corner

Notice that you do not specify a condition to be tested in the .e l request.

Arithmetic and Logical Expressions

A closely related condition simply tests for a nonzero number or a true arithmetic
expression. This is generally used with number registers, but it could also be used to
test the value of numeric arguments to a macro. For example, we could write a para-
graph macro that was either indented or flush left, depending on the value of its argu-
ment:
.de P
- SP
.ne 2
.if \\$l .ti \\$I \ " If there is an arg, use it for indent
420 0 UNlX Text Processing 0

That is, if there is a nonzero numeric argument, do a temporary indent to the distance
specified by the value of the argument.
Rather than using the simple presence of a numeric argument to satisfy the condi-
tion, you could also use an arithmetic expression to test for a value. Used in this way,
the argument can simply be aflug telling the macro what to do.
.de P
- SP
.ne 2
.if \\$1=1 .ti 5n \ " If first arg = 1, indent 5 ens
..
The operators shown in Table 14-2 can be used in constructing an expression.

TABLE 14-2. Expression Operators

Operator Description
+I -1 1, * Standard arithmetic operators
9-
0 Modulo
>r < Greater than, less than
>=, <= Greater than or equal, less than or equal
- _
- __
_
I Equal
& AND
OR

Expressions are evaluated from left to right, except where indicated otherwise by the
presence of parentheses. There is no precedence of operators.
Frequently, you will see numeric conditions involving number registers. Here are
a few simple examples:
-if \\nb
.if \\nb>l
.if \\nb<\\nc
.if \\nb+\\nc>l

(Be sure to note the double backslash before each number register invocation: we are
assuming that these requests are made within a macro definition. If they were made
outside a macro, you would use only a single backslash.) The first of these conditions
is commonly used in the existing macro packages. It takes a little getting used to-it is
not always obvious to new users what is being tested in an expression like:
.if \\nb

A condition of this form simply tests that the specified expression (the number register
b in this case) has a value greater than 0. A more complex expression that does the
same thing might be:
.if \\nb-1
 0 Writing nroff and t r o f f Macros 0 42 1

Comparing Strings
Another frequent test that you can use as the basis of a condition is whether or not two
strings are equal-for example, whether an argument contains a particular string. The
syntax is simply:
.if "string1 v1string2 *I

(Note that there are a total of three quotation marks-either single or double will do-
and no equals sign. A frequent error among beginners is to use an equals sign to com-
pare string arguments, which will not work.)
For example, suppose you are writing a macro to center the output if the second
argument is the letter C. You could write:
.if " \ \ $ 2 " C " .ce \ " I f 2nd a r g is C, center the next line

You can also test for a null argument in this way:

.i f \ \ $1 do something
Use of this condition or its inverse, the test for a non-null argument (described in the
next section), allows the user to skip over an argument by supplying a null string ("").

Executing Multiple Requests as a Result of a Condition

All of the examples we've shown so far consist of a single request executed on the
basis of a condition. But often you'll want to execute more than one command when a
condition i s met. To do so, you enclose the sequence to be executed in backslashes and
braces, as in this example:
.if o \ { \
-PO + . 2 5 i
.tl " ' % * \ )

The initial sequence is terminated with an additional backslash to "hide the newline."
You could also type:
.if o \( -PO +.25i
.tl '''%*\I
However, the syntax shown in the first example is almost always used, because it is
easier to read. There is one caveat! You can't put any other characters, even a com-
ment, following the slash. For example, if you type:
.if o \ ( \ \ " I f odd ...
you won't be escaping the newline, you'll be escaping the spaces that precede the com-
ment. If you want to include a comment on a condition like this, use the alternate syn-
tax, and follow the brace with a dot, just like you would if the comment started on a
line of its own:
.if o \ { . \ " If odd...
422 0 UNlX Text Processing 0

The concluding \ } can appear on the same line as most requests. However, we
have found a problem when it immediately follows a string definition or a t m request. .
For some reason:
.ds string \ }
appends a *Q character to the end of the string, at least in our version of t r o f f . The
.
concluding \ } should be put o n the next line, after an initial to suppress newline gen-
eration in the output:
- \ }

Another convention followed in multiple-line execution is to separate the initial

request control character ( . or ' ) from the body of the request with a tab. This greatly
enhances readability, and can be used to show nesting of conditions:
.if o \ ( \
PO +.25i
t l "'\\n%'\)

Conditions can be nested within each other using this syntax. However, you might
wonder if a nested condition could instead be expressed using one of the logical opera-
tors & or : in an expression. Suppose, as described previously, you want to put page
numbers on the outside corners of each page, except on the first page, where you want
it in the center. You might put the following requests in the page bottom macro:
.ie \\n%>l \ I \ \"If pageno > 1
if o .tl ' d " O '

if e .tl ' % " ' \ }

.el .tl " % "

You might think to achieve the same result with the following requests:
.if \\n%>l&o .tl " l o s' \"If pageno > 1 and odd
.if \\n%>l&e .tl ' % " ' \"If pageno > 1 and even
.if \\n%=l .tl ' r % r r \"If pageno = 1
Unfortunately, however, this example will not work. The & and : operators can only
be used to construct arithmetic expressions. For example, in the case of
.if \\nX&\\nY do something
something will be done only if both register X and register Y are non-zero. (Notice that
there are n o spaces surrounding the ti operator.)
You can construct a n else if clause by following an .el with another . i f , and
then the request to be executed if the condition is met.
.ie condition do something
. e l .if condition do something else if
 Writing n r o f f and troff Macros 0 423

Inverse Conditions
The meaning of any of the condition types described can be reversed by preceding them
with an exclamation point ( !). For example:
.if !e \'* If t h e page number is not even
.if !\\nc=l \m* If number register c is not equal t o 1
.if ! "\\$1"" \ " If t h e first argument is non-null
It may not be immediately obvious what this adds to your repertoire. However, we will
encounter many cases in which it is easier to detect when a condition is not met than
when it is. In particular, negative conditions can be more comprehensive than
equivalent positive conditions. For example, the condition:
.if !\\nc=l
tests not only for the cases in which number register c has been set to some number
larger than 0, or explicitly to 0, but the case in which it has never been set at all.
The test for a non-null argument is also useful. For example, in the sequence:
.if ! " \ \ $ 3 " " \(\ \" If there is a third argument
.ce \" center it
\\$3\ }

you only want the , c e request to be executed if there is an argument to be centered.

Otherwise, the request will cause unexpected results, perhaps centering the line of text
following the macro. Saying "If the third argument is non-null, then it exists" may be
the inverse of the way you think, and will take some getting used to.
If you are reading through the definitions for the m s or mm macros, you may
also encounter a construct like this:
.if \\n(.$-2
The .$ is a special predefined number register (more on this topic in a moment) that
contains the number of arguments that have been given to a macro. If there are two or
fewer arguments, the value of the conditional expression shown will be 0. However, it
will evaluate true if there are more than two arguments. It is used in mm's .SM macro
because a different action is taken on the second argument if there are three arguments
instead of two.
.if \\n(.$-3 \\Sl\s-2\\$2\~+2\\$3
.if \\n(.S-2 \~-2\\$l\s+2\\$2

. Interrupted Lines .
Occasionally, when writing a complex m a c r w s p e c i a l l y one with multiple
conditions-you may find yourself writing a request that is too long to fit on a single
80-character line.
You could simply let the line wrap on your screen-UNIX recognizes lines much
longer than the 80 columns usually available on a terminal screen. However, you need
424 UNlX Text Processing 0

not do this. Simply putting a backslash at the end of a line will “hide the newline”
and cause the next line to be interpreted as a continuation of the first.

Number Registers
To set a number register, you use the .n r request. Like macros, number registers can
have either one- or two-character names consisting of any character(s), not just
alphanumeric characters. For example:
.nr A ( 1
sets a number register called ( to 1. Number register names are stored separately
from macro names, so there is no conflict in having a number register with the same
name as a macro. Thus, you can create mnemonic number register names, which helps
to make macros that use those number registers more readable.
(If you are writing your own macro package, you can name registers from scratch.
If you are adding to an existing package, check the number registers used by that pack-
age.)
To use the value stored in a number register, use the escape sequence \nu for a
one-character number register name, and \n ( x x for a two-character name. (In the
standard n r o f f and t r o f f documentation, this is referred to as “interpolating” the
value of the number register.) The point made previously, about using backslashes to
delay the interpretation of an argument, applies equally to number registers. In macros,
you will usually see the invocation of number registers preceded by a double backslash,
because you don’t want to interpolate the value until the macro is executed.
The values stored in number registers can be literal numeric values (with or
without scaling indicators), values from other number registers (whose value can be
interpolated at a later time), or expressions. You can also increment or decrement the
value placed in a number register by preceding the value with a plus or minus sign. For
example:
.nr P N 1 \ ” Set number register P N t o 1
.nr P N +1 \ ” Add 1 t o t h e contents of number register P N
When you add scaling indicators to the value supplied to a number register, be aware
that values are converted to basic units before they are stored, and that when you incre-
ment the value of a number register, it is incremented in basic units. So, in the previous
example, in which no units were specified, the value of PN after incrementing is 2, but
in the following case:
.nr LL 6.5i
.nr LL +1

the value initially stored into LL is converted into units (Le., for a 300 dpi output dev-
ice, it contains the value 1950); after incrementing, it contains the value 1951 (again,
assuming a 3 0 0 dpi device). If you want to increment LL by 1 inch, append the proper
scaling indicator. Likewise, when interpolating the value of a number register, specify
that the value is in units. For example, the construct:
 0 Writing nroff and troff Macros 0 425

.nr I N li
. i n \\n(IN

will produce unexpected results. What you are really writing is:
. i n 300m

(assuming a 300 dpi device) because the default scaling for an indent request is ems.
The proper usage is:
.in \\n(INu

Number Registers as Global Variables

Number registers can be used in different ways. First, and probably most important,
they can generalize a macro package. For example, in’ m s , the default line length is
stored in a number register called LL.
Periodically, macros in the package may muck with the line length, and then reset
it to its default state. Requests to reset the line length to its default value thus have the
form:
- 1 1 \n(LLu \ “ S i n g l e b a c k s l a s h w i t h i n t h e body o f t e x t
or
- 1 1 \\n(LLu \ ” Double backslash w i t h i n a macro d e f i n i t i o n
Because the line length is not “hard coded” in the document, users can change the line
length throughout simply by changing the default value stored in the number register.
You might wonder why this is necessary. After all, you can simply set an initial
line length, and then increment it or decrement it as necessary. And many macros take
this approach. But there are other cases where the line length is a factor in another cal-
culation.
For example, the output text can be centered horizontally on the physical page
regardless of the line length if the page offset is set not absolutely, but in terms of the
line length:
.PO (8.5i-\n ( L L u ) /2u
In general, it is good programming practice to place values that are used at many dif-
ferent places in a program into globally accessible variables. To change the action of
the program, it is only necessary to change the value of the variable. It is the same in
nro f f and t r o f f . When we look at the overall design of a macro package in
Chapter 16, we’ll return to this subject inmore detail.

Number Registers as Flags

In the chapters on the existing macro packages, you’ve also seen number registers used
as flags-signals to a macro to act in a certain way. For example, in mm, paragraphs
are flush left by default, but if the user sets the Pt number register to 1, all paragraphs
will be indented.
426 0 UNlX Text Processing 0

Within the paragraph macro, there is a line that tests the P t register, and acts
accordingly:
. i f \\n(Pt=l .ti +\\n(Pin
This line actually uses number registers in both ways. If the number register P t is set
to 1, the macro indents by the value stored in another register, P i .
One-character number register names can also be set from the command line, with
nrof € or t r o f f ’ s -r option. This gives you the ability to construct macros that
will act differently depending on command-line options. We’ll show some examples of
this in Chapter 16, when we discuss how to print a document on either an g1/z-by-ll
inch or a 6-by-9 inch page, simply by specifying a single command-line switch.

Predefined Number Register Names

In addition to number registers set by the various macro packages, or set by macros you
write, there are quite a few number registers whose usage is predefined by the formatter.
You’ve already seen one of these--%, which always contains the current page number.
Table 14-3 (and Table 14-4) list some of the most important preset registers, and
Appendix B includes a complete listing. Not all of these registers will be meaningful at
this point, but we’ll tell you more about them as we go on.

TABLE 14-3. Predefined Number Registers

Register Contents
-% Current page number
dl Width (maximum) of the last completed diversion
dn Height (vertical size) of the last completed diversion
dw Current day of the week ( 1 to 7)
dY Current day of the month ( 1 to 3 1 )
hP Current horizontal place on the inpur line
In Output line number
mo Current month ( 1 to 12)
nl Vertical position of the last printed text baseline
Yr Last two digits of the current year

The registers in Table 14-3 can be reset. For example, if you want to arbitrarily
reset the page number to 1 , you can type:
.nr % 1

The formatter will keep incrementing the register on each new page, but will count from
the new baseline. (You might want to do this, for example, if you are following the
convention used in many technical manuals, which number pages on a chapter-by-
chapter basis, with a number made up of both the chapter number and the page number.
In this case, the page number i s reset to 1 at the start of each new chapter.)
 0 Writing nrof f and t r o f f Macros 0 427

Note that % is a true number register name, and don’t let the special use of the %
character in the .t l request confuse you. In .t 1, % alone will interpolate the
current page number; however, in any other place, you must specify the full number
register interpolation \ n%.
The set of registers in Table 14-4 cannot be modified. In reading their names, be
sure to note that they are two-character names beginning with . (dot). If you are read-
ing through one of the existing macro packages, it is easy either to confuse them with
macros or requests, because they begin with a period, or to miss the period and read
them as one-character names.

TABLE 14-4. Read-only Number Registers

Register Contents
.$ Number of arguments available in the current macro
.c Number of lines read from the current input file
.d Current vertical place in current diversion; equal to
nl if no diversion
.f Current font position ( 1 to 4 in o t ro f f )
.H Available horizontal resolution in machine units
.i Current indent
-J Current adjustment mode (0 = .ad 1 or .na;
1 = . a d b ; 3 = . a d c ; 5 = . a d r)
.L Line spacing set with - 1 s
.1 Current line length
.n Length of text on previous line
.o Current page offset
-P Current page length
.s Current point size
.t Distance to the next trap (usually the page bottom)
.u Equal to 1 in fill mode and 0 in no-fill mode
.v Available vertical resolution in machine units
.v Current vertical line spacing
.w Width of previous character
.z Name of current diversion

The registers in Table 14-4 are particularly useful when you want to temporarily
change some value (for example, the font) and then restore it, without having to know
what was there before.
For example, if you print an italicized footer on each page, you might include the
following requests in your page bottom macro:
. n r F T //n(.f
.ft I

.ft \ \ n ( F T
428 0 UNlX Text Processing 0

This is safer than simply using the .f t request without an argument to restore the pre-
vious font, which can create havoc if a user makes a font change within a definition of
the footer string.
Be aware that registers with scaled values (e.g., - 1 for the line lengths or .v
for the current vertical spacing) contain those values as basic machine units (as do all
number registers containing scaled values). As described previously, this means you
should append a u whenever you want to use the contents of one of these registers as
an argument to a request.

Autoincrementing Registers
We've described how to increment the value stored in a register by prefixing the value
you supply to the .n r request with a plus sign (+), and how to decrement it by speci-
fying a minus sign (-).
You can also autoincrement or autodecrement a register whenever you interpolate
its value. To make this work, you must supply two values to an initial .nr request:
the starting value and the increment value. For example:
.nr TE 1 1
.nr ST 10 3

Then, when you interpolate the contents of the register, instead of using the standard
\nx or \n (xu, specify a plus or a minus after the \n and before the register name.
The value that is interpolated will be the original contents of the number register plus
(or minus) the increment (or decrement) value. At the same time, the value in the regis-
ter will be updated by the increment value. For example, assuming the initial defini-
tions in the previous example:
\n+ (TE \ " Increment TE by 1, and interpolate the new value
\n-(ST\" Decrement ST by 3, and interpolate the new value
Number register interpolations of the normal sort can still be used and will, as always,
simply give you the value currently stored in the register.

Altering the Output Format

As we've seen, sometimes number registers are simply used to supply values to
requests, or to pass information between macros. But there are many cases in which the
value of a number register is actually interpolated into the formatter output and printed.
The page number register % is a good example. Although it might be used as the basis
to test conditions in macros, it is usually printed as well.
The .a f (alter format) request allows you to specify the format in which to
express the value of a number register. This request takes two arguments, the name of
the register to be affected and the format:
.a f register format
The format codes are given in Table 14-5.
 0 Writing nrof f and troff Macros 0 429

TABLE 14-5. Format Codes

Format Description Numbering Sequence

1 Arabic 0,1,2,3,4,5 ,...
i Lowercase roman 0, i, ii, iii, iv, v, . . .
I Uppercase roman 0, I, 11, 111, IV,v,. . .
a Lowercase alphabetic 0, a, b, c, . . . z, aa, ab,. . . zz, aaa, . . .
A Uppercase alphabetic 0, A, B, C, . . .Z , A A , A B , . . . ZZ, AAA, . . .

In addition to the numbering sequences in Table 14-5, an arabic format having

additional digits (e.g., 001) will result in a numbering sequence with at least that many
digits (e.g., 001, 002, 003,. . .).
For example, to change to lowercase roman page numbering in the front matter of
a book, you could write:
.af % i
(Note that, depending on exactly how a macro package implements page numbering,
this may or may not work exactly as shown. Some macro packages interpolate % into
another register and print the contents of that register. For example, m s stores the page
number in the register P N and the request would be .a f PN i.)
Alphabetic formats are generally used in macros for automatically numbered (or
lettered) lists. We'll take a close look at some of these macros in Chapter 17.

Removing Registers
With the very large number of possible register names (nearly 10,OOO names are possi-
ble, given all one- and two-character combinations of the printing character set), it is
unlikely that you will run out of number register names.
However, if your macros create a very large number of registers, the formatter can
run out of internal storage space. For this reason, it may occasionally be necessary (or
at least wise) to remove temporary registers that you no longer need, using the .rr
request. For example:
. r r TE \ " Remove r e g i s t e r TE

Defining Strings
In addition to macros and number registers, n r o f f and t rof f allow you to define
character strings that will be stored and can be re-invoked at will. This is not intended
as a general-purpose abbreviation function, although in certain cases it can be used that
way. Rather, it is designed to allow you to store global string variables for use
throughout a package, in much the same way that number registers provide numeric
variables.
 I
430 0 UNlX Text Processing 0

For example, in both m s and mm,you can define headers, footers, or both that
will be printed on every page. To do this, the header or footer macro contains a refer-
ence to a predefined string. All the user has to do is give the string a value. The user
doesn't have to modify the macro itself.
A s we've already seen, to define a string, use the .ds (define string) request.
For example:
.ds RH Tools for Building Macros \ " Define right header

String names, like macro and number register names, can have either one or two charac-
ters. However, unlike number registers, string names are drawn from the same pool as
macro and request names, so you have to be careful not to conflict with existing names.
To interpolate the value of a string, use the escape sequence \*x for a one-
character name, or \ * (xu for a two-character name. For example, our page top macro
might include the lines:
.if o .tl ' \ \ * (RH"%' \ " Print header string then page #
.if e .tl '%"\\*(RH' \ " Print page # then header string
Another good example of how to use this request (as well as how to use predefined
number registers) is given by the technique used in m s and mm to build a date string.
The t r o f f program reads in the date from the system clock into the predefined
number registers m o (month), dy (day), and y r (year). To set a complete date string
that users can easily reference, we might write the following requests in our macro
package:
.if \n(mo=l .ds MO January
-if \n(mo=2 .ds MO February
.if \n(mo=3 .ds MO March
.if \n(mo=4 .ds MO April
.if \n(mo=5 .ds MO May
.if \n (mo=6 .ds MO June
.if \n(mo=7 .ds MO July
-if \n(mo=8 .ds MO August
.if \n(mo=9 .ds M O September
-if \n(mo=lO .ds MO October
.if \n(mo=ll .ds M O November
-if \n(mo=l2 .ds MO December
-ds DY \*(MO \n(dy, 19\n(yr
(Note that these requests do not need to be executed from within a macro. The register
values can be interpolated when the macro package is first read in. For this reason, the
string and number register interpolations shown here are not escaped with an additional
backslash.)
Another request, as (append [to] string), also allows you to add to the contents
of an existing string. The last line of the previous sequence could also have been writ-
ten:
 0 Writing nroff and troff Macros 0 43 1

. a s MO \n(dy, 1 9 \ n ( y r

to append the day and year to whatever value had been stored into MO. Here, this is a
little contrived-it is better to maintain the month and the date as a whole in separate
strings. However, the technique of appending to a string is used appropriately in the
definition of a macro to produce numbered section headers, as we?ll see in Chapter 17.

Diversions
So far, we have discussed macros that you define in advance as a sequence of stored
requests. There is also another class of macros that are created by a process called
diversion.
A diversion consists of temporary storage of text into a macro, which can be
saved and output at a later time. In reading the chapters on m s or mm,you might have
wondered how t r o f f manages to move footnotes embedded anywhere in the text to
the bottom of the page, or how it ?floats? a figure, table, or block of text to the top of
a succeeding page, after filling the current page with text that comes later in the input
file.
The answer is simple: the formatter stores the text (or other output) in a macro
created by diversion. (Such a macro is often called simply a diversion.) The size of the
diversion is stored into number registers that you (your macros, that is) can test to see if
the diversion will fit on the current page, and how much space you need to allocate for
it. The macro package can then make decisions about how and where to place the con-
tents of the diversion.
To create a diversion, use the .d i (divert) request. This request takes as an
argument the name of a macro. All subsequent text, requests, etc. will be processed
normally, but instead of being output, they will be stored into the named macro. A
.d i request without an argument ends the diversion.
The output that has been stored in the diversion can now be output wherever you
like, simply by invoking the macro named in the initial . d i request. For many pur-
poses, this invocation will be performed automatically by a page transition macro. We
will look at this in more detail in succeeding chapters, but just to get the idea, let?s look
at a simple definition for a pair of keep macros.
(In general, diversions are handled by pairs of macros-ne to start the diversion,
the other to end it. However, there are other cases in which we will see that this is not
necessary.)
Both m s and mm use diversions in their dispIay macros. In m s , the display
macros handle text positioning, and call lower-level macros called keep macros to make
sure the text in the display stays on the same page.
The purpose of the keep macros, in case you are not familiar with this concept
from earlier chapters, is to make sure that a block of text is not split across two pages.
A typical example of a block that should not be split is a figure-whether it is reserved
space for a figure, or an actual picture created with pic or some other graphics tool.
A simple macro to start a keep might look like this:

t
432 0 UNlX Text Processing 0

.de KS \ " Keep S t a r t

.br
.di KK
..
A simple macro to end a keep might look like this:
.de KE \I' Keep End
.br
.di
.ne \\n(dnu
.nr fI \ \ n ( . u
.nf
.KK
.if \\n(fI .fi
..
In both macros, the .b r requests are extremely important; they flush any partial lines
that have not yet been output. In the .KS macro, the break makes sure that the keep
begins with the text following the macro; in .KE, it makes sure that the last partial line
is included in the diversion.
It is also important to output the diversion in no-fill mode. If you don't, the text
contained in the diversion will be filled and adjusted a second time, with unpredictable
results. (Consider, for example, when the diversion includes an already formatted table.
The table would be scrambled by a second pass.)
You can't just switch back to fill mode after you output the diversion, though.
What if the body of the text was meant to be in no-fill mode? To get around this prob-
lem, you should save the value of trof f ' s read-only register .u, and test the saved
value to see whether or not filling should be restored.
There are a few times when you might not want to follow this rule. For example,
what should you do if there i s a chance that the diversion will be output on a page
where the line length is different? You still want to avoid processing the text twice.
You can put the text into the diversion in no-fill mode, and can embed any formatting
requests into the diversion by preceding them with a backslash (e.g., \ i n 5n). Any
requests treated in this way will be acted on when the diversion is output.
As always, it is important to specify the correct units. In the previous example,
the value in d n is stored using basic device units (as is the case with all scaled values
stored in a number register), so you must add a u on the end of the interpolation. For
example, on a 300 dpi device, after a diversion 2 inches high, dn will contain the
value 600. The request:
. n e \\n(dn

will always result in a page break because (in this example) what you are really writing
is:
.ne 600

What you want to write is:

 0 Writing nroff and troff Macros 0 433

.ne \\n(dnu

Any text and requests that are issued between the initial . K S and the terminating
.KE will be stored in the macro called .KK. The height of the last-completed diver-
sion is always stored in the number register dn. We can simply say that we need
( - ne) at least that much space. If the size of the diversion is greater than the distance
to the bottom of the page, we simply start a new page. Otherwise, we output the text
and continue as if the diversion had never happened.
The case of a floating keep, in which text that follows the keep in the source file
floats ahead of it in the output, and fills up the current page, is more difficult to handle
than the simple example just shown. However, this example should give you an idea of
how to use diversions.
There is also a .d a (divert append) request that adds output to an existing diver-
sion. (A second .di given the same macro name as the first will overwrite the
diversion’s previous contents, but .da will add the new material to the end.)
The .da request has numerous applications. For example, consider footnotes.
To calculate where to place the first footnote, you need to calculate the size of all the
footnotes you want to put on the page. That’s easy-just append them to the same
diversion.
However, there are other far less obvious applications for appended diversions.
For example, you can divert and append section headings or index entries to macros that
will be processed at the end of the file to produce a table of contents or an index.

Environment Switching =
The n r o f f and t K O f f formatters allow you to issue many requests that globally
affect the format of a document. The formatter is generally quite thorough in providing
ways to change and restore the value of various parameters. This makes it relatively
easy to change values such as the line length or fill/no-fill mode in order to treat certain
blocks of text differently and then restore the original values.
Nonetheless, if you want to make major changes to a number of values, it can be
awkward to save and restore them all individually. For this reason, n r o f f and
t r o f f provide a mechanism called environment switching. By default, text processing
takes place in what is considered to be environment 0. The .ev request allows you to
switch to either of two additional environments, referred to as environment 1 and
environment 2.
For example, to change to environment 2, you would enter
.ev 2

To restore a previous environment, you simply issue an .ev request without an argu-
ment. Environments are stored in a “push down stack.” So if you are using multiple
environment switches, a sequence of .ev requests without arguments won’t toggle you
between two environments, but will actually backtrack the specified number of environ-
ment switches. That is:
 I
434 UNlX Text Processing 0

.ev 1
do something
.ev 2
do something
.ev \ " G o back to ev 1
- ev \ " G o back to ev 0

If you use .ev with an argument, you will not pop the stack. For example, the
requests:
.ev 2
.ev 0
will leave both environments on the stack. You might get away with this on one occa-
sion, but if you do this in a macro that is used with any frequency, your stack will keep
getting deeper until it overflows and the formatter fails with the message "Cannot do
ev.3,

Within each environment, settings made with the following requests are remem-
bered separately:
.c2 .cc .ce .cu .fi .ft .hc .hy .in .it .IC - 1 1 .Is .It
.mc .nf .nh .nm .nn .ps .sp . s s .ta .tc .ti .ul .vs
Number registers, macros, and strings are common to all environments. However, any
partially collected lines are part of a given environment. If you switch environments
without causing a break, these partial lines will be held till the environment that con-
tains them is restored.
What this means is best shown by example:
\" Set parameters for environment 0
- 1 1 4.5i
.ad b
.ev 1 \ " Switch to environment 1
-11 -10n \ " Set parameters for environment 1
.in +10n
.ad 1
.ev \ " Restore previous environment (ev 0 )
This text will be formatted using the parameters for
environment 0 . Notice that part of the last input
line appears to be lost when we switch environments.
It reappears when the environment is restored.
.ev 1
-SP \ " The break caused by this request is in ev 1
Now we've switched to environment 1. Notice how the text
is now formatted using the parameters for environment 1.
Also notice that this time, we're going to issue an .sp
request after this sentence to cause a break and make sure
the last partial line is output before we leave this
environment.
- SP
 0 Writing nrof f and trof f Macros 0 435

.ev \ " Back once more to environment 0

This sentence will be preceded by the remainder of input
left over from the last time we were in this environment.
Here's the resulting output (from nroff):

This text will be formatted using the

parameters for environment 0. You'll notice
that part of the last input line appears to
be lost when we switch environments. It

Now we've switched to environment 1. Notice

how the text is now formatted using the
parameters for environment 1. A l s o notice
that this time, we're going to issue an .sp
request after this sentence to cause a break
and make sure the last partial line is output
before we leave this environment.

reappears when the environment is restored.

This sentence will be preceded by the
remainder of the input left over from the
last time we were in this environment.
Environments are very powerful and versatile. The example given previously
could have been handled more appropriately with a macro. However, as you will find,
there are tasks that are best handled by an environment switch.
Printing footnotes i s a primary example. Footnotes are usually collected in a
diversion, which must be output at the bottom of the page without causing a break or
other interference with the text.
Unfortunately, you must use environment switching with caution if you are work-
ing within one of the existing macro packages, because they may use different environ-
ments internally, and changing parameters in an environment may affect the operation
of the package. For example, it was necessary to process the preceding example
independently with nroff, and then read the resulting output into the source file,
because the use of environments by the macro package that produced this book was
incompatible with what we were trying to show.

Redefining Control and Escape Characters

There are special requests to reset the control characters that begin requests ( . and ' )
and the escape character:

k
436 0 UNlX Text Processing 0

.eo \"Turn escape character off except for comments

.ec ! ! " Set escape character t o !
.ec \ \ " Set escape character back t o \
.cc # \ " Change control character from to # .
.c2 \ " Change no-break control character from ' t o A

As far as we can tell by experiment, turning the escape character off entirely with .eo
does not affect the comment sequence \ however, if you change the escape character
l l ;

with .ec, comments must be introduced by the new escape character.

We have not found a significant use for these requests in our own work, or in
macros we've studied, although there are no doubt cases where they are precisely what
is needed.
One application that immediately suggests itself is the representation of control
and escape characters in the examples shown in this book. However, in practice there
are many problems.
For example, if you use these requests in a pair of macros to frame examples, the
closing macro must be invoked with the appropriate control character, creating incon-
sistencies for the user. Even more seriously, if control character translations are in
effect during a page transition (something that is difficult to control) or other macro
invoked by a trap, they will render that macro inoperable, unless it has been designed
with the same control and escape characters.
Our preferred solution to this problem is to use the .t r request, which is dis-
cussed in the next chapter.

Debugging Your Macros

When using a markup language as complex as that provided by n r o f f and t ro f f,
it is easy to make mistakes, particularly when you are designing complex macros.
To limit the number of mistakes you make, you can take lessons from program-
mers in more general-purpose languages:

. Start by writing and testing small pieces of a complex macro. Then, after you
know the pieces work, put them together. It is much easier to find a problem
in a simple macro than in one that is already very complex.
. Be aware of interactions between the macro you are writing and other macros
in the package. Initialize variables (number registers and strings) that might
also be used by other macros.
Include extensive comments, so you can reconstruct what you were trying to
do when you go back to the macro later. (Errors often arise unexpectedly after
the macro has been in use for a while, and you have a chance to exercise it
fully. Be sure you can follow what you originally wrote.)
Test each macro thoroughly before you put it into general use.
 Writing nrof f and trof f Macros 0 437

However, even with the best of intentions, you are likely to make mistakes. This short
section is intended to give you a few pointers on how to track them down.
The term debugging is familiar even to nonprogrammers. In general, it refers to
the process of finding errors in programs. I would like to suggest an alternate definition
that may give you better insight into how to go about this process: Debugging is the
process of finding out what your macro really does, instead of what you thought it
should do.”
When you write a program or a macro, you have an idea in your mind of what
you want to accomplish. When it doesn’t do what you expect, you consider it an error.
But as we all know, computers are very literal. They generally do just what they
are told. (The exception being when there is an error in some underlying layer of
software that causes problems on a higher layer.) Therefore, the essence of debugging
is to compare, on a step-by-step basis, exactly what the program or macro is actually
doing with what you expect it to do.
There are several tools that you can use in debugging macros. First, and most
obviously, you can look carefully at the output. Try to reconstruct the sequence of
instructions and text that have been executed to produce the (presumably) undesirable
result. Often, this will be all you need to do-think like a text formatter, and go
through the requests that have been executed, in the order that they are executed.
You will often find that problems are due to an incorrect understanding of the
action of one of the requests or escape sequences, so it may be advisable to consult the
bible of macro programming, Joseph Osanna’s extraordinarily dense but thorough
NrofflTroff User’s Guide.
Secondly, you can use n r o f f or t ro f f interactively. If you simply type:
$ nroff
or:
$ t r o f f -a

the program will take standard input from the keyboard and send its results to standard
output (the screen). The t r o f f -a command creates an ASCII approximation of
what the t r o f f output would be; if you are using d i t r o f f , you can also save the
normal output in a file and look directly at the output. However, this output is in an
obscure format and takes considerable time to learn.
With t r o f f -a, special characters (such as underlines) are represented by their
special character names. For example, underlining will show up as a sequence of
\ ( u l s . Because proportional type is considerably more compact than the characters
that appear on a terminal screen, lines will appear too long, and will wrap around on the
screen. However, what you see does represent how t r o f f will break the lines.

*I am indebted to Andrew Singer of Think Technologies for this definition.Andrew used similar words in
describing to me the debugging philosophy of his company’s innovative Pascal compiler for the Macintosh,
Lightspeed Pascal.
438 0 UNlX Text Processing 0

Now, by typing in your macros (or reading them in from existing files with the
.so request), you can reproduce the environment of the formatter, and watch the
results as you type in text. As each line is completed in the input buffer, the formatted
result will be output. You can force output of a partially completed line with the . f 1
(flush) request, which was designed for this purpose.
This method has definite limits, but has just as definite a place in pinning down
what the commands you type are doing.
Another debugging tool that you may find useful is the - i g (ignore) request. It
tells the formatter to ignore subsequent input, up to a specified terminator (. . by
default). The .i g request acts like .de, only the input is discarded. (The only
exception to this is that autoincremented registers whose values are interpolated within
the ignored block will still be incremented or decremented.)
This request is useful when you are trying to pin down exactly where in an input
file (or a long macro definition) a fatal error (one that causes the formatter to stop pro-
cessing) occurs. By successively isolating parts of the file with .i g , you can locate
the problem more closely.
This request is also useful for “commenting out” extensive blocks of macro
definition or input text that you don’t want in your output. It is much easier to bracket
a large block of input in this way than it is to insert comment characters at the begin-
ning of each line.
Because you may want to “ignore” more than one macro definition, you may
want to get in the habit of specifying a special delimiter for the .i g request, so that
the “ignore” is not accidentally terminated by the end of the first macro definition.
This will also make it much easier to find the end of the ignored section. For example,
if you insert the line:
-ig ++
anywhere in your input, the formatter will ignore the input until it sees the request:
- +t
The final tool provided for debugging is the .t m (terminal message) request, which
prints a message on standard error. This is particularly useful for tracking down errors
involving number registers. For example, if you have set a condition based on the
value of a number register, and the condition never seems to be satisfied, you might
want to insert .t m messages to print out the value of the number register at certain
points in your file. For example:
. t m B e f o r e c a l l i n g B1, t h e v a l u e o f BC is \ n ( B C
.B 1
. t m A f t e r c a l l i n g B1, t h e v a l u e o f BC i s \ n ( B C

(Note that there are no double backslashes before the number register interpolations,
because these requests are not made while you’re inside a macro definition. From
inside a macro, be sure to double the backslashes, or you will get the value of the
number register at the time the macro was defined.)
A read-only number register that is useful in this regard is .c, which contains
the number of lines read from the current input file. This allows you to create messages
 0 Writing nrof f and trof f Macros 0 439

that will help you (or the user of your macros) find out where in the input file an error
(or other event) occurs:
.tm O n i n p u t line \\n(.c, the v a l u e of BC w a s \ \ n ( B C

(Here, there are double backslashes, because this example is intended to be inside a
macro definition.) Sometimes it is helpful to follow just how far the formatter has got-
ten in a file. The most difficult errors to track are those that cause the formatter to quit
without producing a block of output. A series of messages of the form:
.tm At t o p of page \\n%, I‘ve processed \\n(.c input lines

inserted into the page top macro will help you determine how far the formatter has got-
ten, and can thus help locate an error. If the formatter is processing standard input
rather than an individual file, the .c register will be empty.
Another register that you may find useful in printing error messages is .F , which
contains the name of the current file. (Yes, the filename is a string, even though it’s
stored in a number register.)
The .R register i s also useful. It contains the number of free number registers.
You can print its value to see if you are running out of number registers or coming
close to the limit. (tbl and eqn use many dynamic number registers, and it is possi-
ble to run out if you use a lot in your macros as well.)
Although we use the tools described here to debug our macros, we know that they
don’t always help you deal with the complexity of a macro package. The relationships
among different macros are not always apparent. For instance, you can usually tell
from looking at your output what macro is causing a problem; however, when you look
at the macro definition, you might find that this macro is calling several other macros or
testing registers that have been set elsewhere in the macro package. It soon leads to the
wish for a debugging tool that traced the interpretation and execution of macro defini-
tions.
At least one version of t r o f f does support a trace facility. Users of SoftQuad’s
S Q t r o f f can enable a trace mode to show the invocation of each request, diversion,
trap, and macro call. For instance, suppose that a macro tests the value of a number
register to determine whether a request should be executed. In trace mode, you can see
at what point the . i f request was invoked, whether it was evaluated as true or false,
and determine the actual value of the number register at that point. SoftQuad has also
taken another step to make debugging easier by improving t rof f ’ s obscure error mes-
sages. In general, SoftQuad has enhanced standard t rof f in other ways that aid the
process of macro writing and debugging, such as allowing longer names (up to 14 char-
acters) for macros, requests, strings, registers, and fonts.

Error Handling
There are many different ways that users can get into trouble while coding documents,
and your macros can help them identify and recover from problems. The three most
common classes we have encountered are:

t
440 0 UNlX Text Processing 0

A user fails to properly understand the action of the formatter itself. For exam-
ple, he or she begins a text line with a single quote or period, or defines a spe-
cial character (such as %) as an eqn delimiter. This problem becomes more
pronounced as users try out more advanced capabilities without really under-
standing them.
A user fails to properly understand the macro package. For example, he or she
gives the wrong argument to a macro or specifies the wrong units.
A user temporarily resets some condition, either directly or by failing to close a
set of paired macros. This causes undesirable effects to propagate through the
document.

The mm macros attempt to solve the first problem by creating so comprehensive a

macro package that users never need use many low-level formatter requests. However,
in doing so, its developers have created an environment that is in many ways more
complex than the raw formatter environment itself. And in our opinion, no macro pack-
age is comprehensive enough to meet all user needs. Over time, users come up with
formatting problems that they need to know how to solve on their own. There is no
solution to this problem except better user education.
To some extent, you can compensate for the second problem by testing for argu-
ments and printing error messages if a macro is misused. For example, if a macro
requires an argument, consider printing a message if the user fails to supply it:
.if "\\$l'"' .tm Line \\n(.c: .Se requires section \
number as first argument

Of course, by the time the user sees the error message, he or she has already formatted
the document, and it is too late to do anything else but repair the damage and reprint.
However, messages can sometimes make it easier for users to find errors and can give
them warning to look more closely at their printout.
The .ab request takes things one step further-it lets you terminate processing
if the formatter encounters a condition you don't like. For example, you could write a
macro that aborts if it is called without a necessary argument:
.if !\\n(.$ .ab You forgot the argument!

The .ab request prints its argument as an error message, just like - t m . It just takes
the further, definite step of quitting on the spot.
Probably more suitable, though, is a separate tool for checking macro syntax.
Such a tool exists for mm in the mmcheck program. A program like this checks the
syntax of macros and requests used in a document and reports possible errors.
This kind of approach is especially suitable for the third kind of error-the failure
to close a set of paired macros.
 0 Writing nroff and troff Macros 0 44 1

Macro Style 9

As you develop more of your own macros, you might begin thinking about overall
macro style. Developing macros that behave in a consistent, reliable way becomes all
the more important as the number of new macros you have increases along with the
number of people using them. Recognizing different styles of macro writing helps to
suggest alternatives and improvements in the way a macro works.
If you have read the chapters on m s and mm in detail, or if you are already fami-
liar with both of these packages, you have probably noticed that they embody somewhat
different text-processing philosophies.
For example, m s generally attempts to recover and continue when it encounters a
serious error, but mm aborts the formatting run. And although m s allows a certain
amount of user customization (generally by providing a few number registers and
strings that the user is expected to modify), it has nowhere near the complexity of mm
in this regard. An mm user is expected to set up various number registers that affect
the operation of many different macros.
In writing your own macros (especially ones that will be integrated with one of
the existing packages), you should take some time to think about style, and how you
want users to interact with your macros. This is most easily shown by comparing
several different paragraph macros:
.de P \ " A very simple paragraph macro
-br
.ne 2v
.ti 2P

.de LP \ " A n m s style flush left paragraph

.RT
.ne 1.1
.sp \\n(PDu
.ti \\n(.iu
_.
.de PP \ " A n ms style indented paragraph
.RT
.ne 1.1
.sp \\n(PDu
.ti +\\n(PIu
..
.deP \ " An mm style variable paragraph
- br \ " Note that this is much
.sp (\\n ( P s * . 5)u \ " simplified from t r u e mm code
.ne 1 . 5 ~
442 0 UNlX Text Processing 0

.if\\n( . $ > O & ( 0 \ \ $ 1 ) .ti+\\n(Pin

.if\\n(.$=O .if\\n(Pt=l .ti+\\n(Pin
..
The first example shows a very simple paragraph macro using a fixed indent value.
The second and third examples are adapted from m s . They show the use of an
embedded reset macro (discussed in Chapter 16) and the specification of values such as
indents and interparagraph spacing by means of number registers so that users can
change them without rewriting the macro. The different types of paragraphs (flush left
or indented) are handled by defining two different macros.
The fourth example is adapted from mm. It shows how a macro can be controlled
in a number of different ways. First of all, the size of the paragraph indent can be con-
trolled by the user, as in m s . Second, though, users can specify whether they want an
indent for a particular paragraph by specifying an argument to the macro. Finally, they
can specify whether all paragraphs are to be indented or flush left by setting the Pt
(paragraph type) register.
Although you may not want to go as far as mm in giving different ways to affect
the action of a macro, it is good to realize that all of these options are available and to
draw on them as necessary.
However, it does make sense to be consistent in the mechanisms you use. For
example, suppose you create macros to put captions on figures, tables, and examples. If
you allow the user to control the amount of space before the caption with an optional
argument, you ought to do so in all three analogous macros.
As much as possible, a user should be able to infer the action of a macro from its
name, and should be able to guess at its arguments by analogy to other, similar macros
in the same package. If you are capricious in your design, other users will have a much
greater burden to shoulder when it comes time for them to learn your macros. Even if
you are the only user of macros you develop, consistency will help you keep out of
trouble as you gradually extend your package.
The issue of macro style really comes up as you begin to develop your own cus-
tom macro package, as you will see when we examine the elements of a macro package
in Chapters 16 and 17.
 1

C H A P T E R
8

IS

Figures and Special Effects

This chapter discusses a variety of formatter requests that you can use to draw figures
and achieve special effects like overstriking and vertically stacked text. It also dissects
some of the most complex macros we’ve seen so far, so it should advance your
knowledge of how to write macros as well as your knowledge of its explicit subject
matter.

Formatter Escape Sequences

Preprocessors like t b l and pic draw boxes, lines, and simple figures using an
underlying library of formatter escape sequences that you can also use directly. The
eqn preprocessor also uses many of these escape sequences, as well as others that are
more appropriate for creating special effects with text characters.
The escape sequences are listed in Table 15-1. A s you can see, there are quite a
few! Fortunately, many of these need not be learned by the average user. The various
preprocessors often allow a user to achieve the same effect more easily. Although
t b l or eqn might seem difficult to learn, they are far simpler than the formatter com-
mands they replace. For example, an eqn construct like 10% s u p 5 % is easier to
learn and type than an equivalent trof f construct like:
10\s-3\v’-3p’5\v’3p’\s0

When it comes to drawing lines and figures, things get even more complex.
For this reason, many of the escape sequences we are about to discuss are not
often used by the average person trying to achieve special effects. However, they are
extremely useful to a developer of macros.
In this chapter, we’ll cover the sequences for local vertical and horizontal motions
and line drawing, because these requests are most commonly used in macros. In addi-
tion, we will show several large macros that do line drawing in order to demonstrate
both the use of escape sequences and techniques for writing complex macros.

443 .
444 UNlX Text Processing 0

TABLE 15-1. Formatter Escape Sequences

Escape Description
\v' distance' Move distance vertically down the -page.
- Precede distance
with a minus sign to move back up the page.
\ h 'distance ' Move distance horizontally to the right. Precede distance
with a minus sign to move back to the left.
\U Move em up (I/z line in nro f f).
\d Move I12 em down (l/2 line in n r o f f ) .
\r Move 1 em up ( 1 line in n r o f f ) .
\C Join next line to current output line, even across a break.
\P Cause a break, and adjust current partial output line.
\ x distance' Add extra line space for oversize characters.
\(space) Move right one space (distance determined by .s s ) .
\O Move right the width of a digit in the current font and size.
\I Move right 1/6em (ignored in n r o f f).
\ A Move right 1/12 em (ignored in n r o f f).
\ w ' string ' Interpolate width of string.
\kx Mark current horizontal place in register x.
\O'Xy' Overstrike characters x and y .
\ zc Output character c without spacing over it.
\b' string' Pile up characters vertically (used to construct large brackets,
hence its name).
\l'Nc' Draw a horizontal line consisting of repeated character c for
distance N . If c isn't specified, use _.
\L'NC' D;aw a vertical line consisting of repeated character c for dis-
tance N . If c isn't specified, use I.
\D' 1 x,y' Draw a line from the current position to coordinates x,y
( d i t r o f f only).
\D' c d' Draw a circle of diameter d with left edge at current position
( d i t r o f f only).
\D'e d l d2' Draw an ellipse with horizontal diameter dl and vertical
diameter d2, with the left edge at the current position
( d i t r o f f only).
Draw an arc counterclockwise from current position, with
center at XI ,yI and endpoint at XI +x2 ,yl +y2 ( d i t r o f f
only).
\D' - X I y l x2 y 2 , Draw a spline from current position through the specified
coordinates ( d i t r o f f only).
\H' n' Set character height to n points, without changing the width
( d i t r o f f only).
\S'n' Slant output n degrees to the right. Negative values slant to
the left. A value of zero turns off slanting ( d i t r o f f only).
 0 Figures and Special Effects 0 445

Many of the escape sequences in Table 15-1 take arguments that must be delim-
ited from any following text. The delimiter character is most often ‘ or AG (CTRL-G),
but it can be any character. The first character following the escape sequence will be
taken as the delimiter, and the argument list will be terminated when that same charac-
ter is encountered a second time.

Local Vertical Motions

There are a number of escape sequences for local vertical motions. They are so called
because they take place within a line, without causing a break or otherwise interrupting
the filling and justification process.
However, this is not to say that the motions they cause are limited. For example,
you can use \v, the vertical motion escape sequence, to move anywhere on the page,
just as you can with the .s p request. However, the remainder of the line that has
been collected in the formatter’s internal buffers will be output in the new location just
as if the motion had never taken place.
To make this point clearer, let’s look at three examples of input text that use dif-
ferent types of vertical motion.
What happens with .sp:
Input lines:
Especially i n troff, it is sometimes uncanny the way that
vertical motions can occur
- s p 12p
independently from the output of t h e text.

Output lines:

Especially in trof f, it i s sometimes uncanny the way that vertical motions

can occur

independently from the output of the text.

What happens with sp:

Input lines:
Especially in t r o f f , it is sometimes uncanny the way that
vertical motions can occur
‘ s p 12p
independently from the output of the text.
446 0 UNlX Text Processing 0

Output lines.

Especially in trof f, it is sometimes uncanny the way that vertical motions

can occur independently from the output of the text.

What happens with \v' 12p' :

Input lines:
Especially in troff, it is sometimes uncanny t h e way that
vertical motions can occur \v812p'
independently from t h e output of t h e text.

Output lines:

I------
Especially in t rof f, it is sometimes uncanny the way that vertical motions
can occur
independently from the output of the text.

As you can see, .s p causes a break as well as a downward movement on the page.
The partially collected line is output before the movement takes place. With ' sp, the
line currently being collected is completely filled and output before the spacing takes
place. With \v,the motion is completely independent of the process of filling and jus-
tification.
It is also independent of traps, as we discovered once when trying to put a point-
ing finger ( ) at the bottom of a page to indicate that the subject was continued on the
overleaf. We used a macro invoked by the page bottom trap to print the finger. At
first, we made the mistake of using .s p -1 to move back up the page to place the
finger. Unfortunately, this put t r o f f into an endless loop around the trap position.
The \v escape sequence, on the other hand, did the trick nicely. Since it does not
change the current baseline spacing, it will not trigger a trap.
Long-winded examples aside, that is why \v is considered a local motion. In
general, \v escape sequences are used in pairs to go away from, and then back to, the
current vertical position.
A superscript is a good example of vertical motion using \v. For example, you
could create a simple superscript macro like this:
.de S U
\\sl\s-2\v'-3p'\\s2\v"\so\\$3

This macro

. prints its first argument;

reduces the point size;
makes a 3-point reverse vertical motion;
 ~

0 Figures and Special Effects 0 447

prints the second argument;

makes a 3-point vertical motion to return to the original baseline;
restores the original size;
prints an optional third argument immediately following. (This allows punc-
tuation to be introduced immediately following the superscript, rather than on
the next line. If no third argument is supplied, this argument interpolation will
be ignored.)

This macro could also be implemented using the \u (up) and \d (down) escape
sequences, which use a fixed I/z-em distance. If you did t h i s - o r if you specified the
distance for the \v escape sequence in a relative unit like ems, instead of a fixed unit
like points-it would be essential to have both of the vertical motions either inside or
outside the font size change. For example, assuming that the current font size was 10
points:
.de S U
\\S~\U\S-~\\S~\~\SQ\\$~
..
would produce an unbalanced effect, because the upward motion would be 5 points (l/2
em at 10 points), while the downward motion would be only 4 points (I/z em at 8
points). This caution holds true whenever you mix font and size changes with local
motions.

Local Horizontal Motions

Much of what has been said about local vertical motions is true for local horizontal
motions. They take place independently of the process of filling and justification, and
so, if improperly used, can result in horrors like:

G h a t happenyarhemake a mistake with 'h! I

which was produced by the line:
Look what happens when \h'-3m'you make a mistake with \h!

Horizontal motions are not as likely to take place in pairs as vertical motions. For
example, there are cases where you want to close up the space between two special
characters, or open up additional space on a line. For example, >>, produced by
>\h' -1p' >, looks better than >>.
In addition to \h, there are a number of escape sequences that affect horizontal
motion in specific ways.
For example, " \ " (it's quoted so you can see the blank space following the
backslash) will space over to the right by exactly one space. That sounds trivial, but it
isn't. When it justifies a line, t r o f f feels free to expand the spaces between words.
(The default space size is normally 12/36 of an em, but can be reset with the .s s
448 0 UNlX Text Processing 0

request using units of 3 6 t h ~of an em). The "\ " escape sequence makes sure that you
get exactly one space. This is generally thought of as the unpaddable space character
and is used when you want to keep two words together. However, it can also be used
simply as a horizontal motion sequence.
Another useful sequence is \O. It provides exactly the width of a digit in the
current font and size. (Unlike alphabetic characters, all digits are always the same
width on the standard fonts, to allow them to line up properly in numeric displays.)
The \ O sequence is most useful when you want to line up spaces and digits manually.
The two escape sequences \ I and \ ", which give, respectively, a 1/6 em and
1/12 em space, are useful when you want to create just a little bit of fixed space
between two characters. (The normal space size created by an actual space character is
1/3 em, so these two characters give you, respectively, one-half and one-quarter of the
normal interword spacing.) You may remember that we used \ " in Chapter 12 to
create a little bit of space before and after the em dashes we were introducing into our
files with sed.

Absolute Motions
As you've probably gathered from the preceding discussion, you can specify the dis-
tance for a horizontal or vertical motion using any of the units discussed in Chapter 4.
The values can be given explicitly, or by interpolating the value of a number register.
In addition, as discussed in Chapter 4, you can use a vertical bar ( I ) to indicate absolute
motion relative to the top of the page or the left margin.
This is not as simple as it first appears. For vertical motions, you pretty much get
what you expect. For example, .sp I 2 i, \v' I 2 i ' will move you to a position 2
inches from the top of the page. Depending on where you are on the page before you
issue the command, the generated motion will be either positive or negative.
For horizontal motions, things are a little more ambiguous. The absolute position
indicator doesn't move you to an absolute position based on the output line, but on the
input line. For example:
This is a test of absolute horizontal motion\h'lli' -

produces :

I This is a test of absolute horizontal motion

But:
T h i s is a test of
absolute horizontal motion\h' Ili' -

produces :

I This i s a test of absolute horizontal motion 1

 0 Figures and Special Effects 0 449

What is really supplied as an argument to \h when you use the absolute position indi-
cator is the distance from the current position on the input line to the specified position.
Even though it looks the same, the argument will have a different value, depending on
the length of the input line. And again, as with vertical motions, the actual movement
may be positive {to the right) or negative (to the left), depending on the relationship
between the current position and the absolute position specified.
It may appear odd to have these motions relative to the input line. However, as
we will see (especially in line drawing), there is a method to the madness.

Line Drawing
Now we come to the fun part. Moving around on the page is of little use unless you
plan to write something at the point you've moved to. Superscripts, subscripts, and
overprinting provide some application of local motion, but local motions are most use-
ful with the escape sequences for drawing lines and curves.
Applications range from underlining words in t r o f f , to boxing single words (if
you are writing computer manuals, this is very useful for showing the names of keys to
be pressed), to drawing boxes around whole blocks of text, just like t b l does.
The \ 1 sequence draws a horizontal line; \ L draws a vertical line. Both escape
sequences take two arguments, the second of which is optional. Both arguments should
be enclosed together in a single pair of delimiters.
The first argument is the distance to draw the line. A positive value means to
draw a horizontal line to the right, or a vertical line downward (depending on whether
\1 or \ L is used). A negative value means to draw a line back to the left, or back up
the page.
When you draw a line back to the left, either by explicitly specifying a negative
value, or by specifying an absolute value (such as I 0) that results in a negative move-
ment, t r o f f first moves back to the specified position, then draws the line from left
to right. It is as if the Iine is drawn from the specified distance to the current position.
For example:
\1' 3i' draws a line 3 inches to the right
\lr-3i' draws a line from a position 3 inches to the left
\L' 3i' draws a line 3 inches down
\L'-3i' draws a line 3 inches up
\L' I3i' draws a line to a position 3 inches from the top of the page
The optional second argument is the character with which to draw the line. By default,
a horizontal line is drawn with the baseline rule-a horizontal line that is aligned with
the bottom of the other characters on a line. However, if you want to underline text, be
sure to use the underscore, which is printed in the space allotted for characters that des-
cend below the line:
These-words-are-separatedby_baseline-rules.
These-words-are-separated-by-underscores.

The underscore is usually generated by the underscore character that appears above the
hyphen on most keyboards. However, to be on the safe side, you should refer to it by
450 0 UNlX Text Processing 0

its special character name in t r o f f-\ (ul. (The baseline rule can be specified with
the sequence \ ( r u . )
Vertical lines are drawn by default with a character called the box rule (which can
be generated by the \ ( b r escape sequence or the vertical bar character on most key-
boards). The box rule is a zero-width character-that is, when t r o f f draws the box
rule, it does not space over as it does with other characters. This allows t r o f f to
form exact comers with horizontal lines drawn with undermles. However, as you will
see, it may therefore require you to manually specify additional space to keep it from
crowding previous or succeeding characters.
Except in the case where you draw a line to the left, as described previously, the
current position at which text output will continue is changed to the endpoint of the
line. In drawing a box, you will naturally find yourself returning to the starting point.
However, if you are drawing a single line, you may need to use \v or \h to adjust
the position either before or after the line is drawn.
Let’s look at a couple of examples. A simple macro to underline a word in
t r o f f might look like this:
.de UL
\\$l\l’ I o \ (ul’\\$2

This example prints its argument, backs up a distance equal to the length of the argu-
ment on the input line, then draws a line from that point to the current position. The
optional second argument allows you to specify punctuation without separating it with
the space that is required if it were entered on the next input line. (This reverse motion
is implicit in the negative value generated by the absolute position request I 0-that is,
the distance from the end of the word to the beginning of the line. Lines drawn with
\1 and a negative distance generate a negative horizontal motion for the specified dis-
tance. The line is then drawn in a positive direction back to the current position.)
That is:
. U L Hello ,
produces:

-Hello
,

and:
.UL Hello
I

produces:

Hello
- 9

(In n r o f f , you can underline simply by using an italic font switch, or the .ul
request, because italics are represented in n r o f f by underlines.)
 0 Figures and Special Effects 0 451

A macro to enclose a word (like the name of a key) in a box might look like this:
.de BX
\ (br\ I \ \ $ 1 \ I \ (br\l' I O \ (rn'\l' I O \ (u1'\"\\$2
__
For example, the input text:
Press t h e
.BX RETURN
key -
will produce the line:
Press the- 1 key.
This macro prints a single box rule ( \ (br), spaces over 1/6 em ( \ I), prints the argu-
ment, spaces over another 1/6 em space, and prints a concluding box rule. Then it
draws two horizontal lines back to 0 (the beginning of the input line-that is, the width
of the argument plus the two requested 1/6-em spaces).
The first horizontal line is drawn not with \ (ul but with another special charac-
ter, the root en (\ (rn). This character is used when drawing equations to produce the
top bar in a square root symbol, but it is just as useful when you want to draw a line
over the top of some text without moving back up the page. The second horizontal line
is drawn, as usual, with \ ( u l .
Both lines can be drawn back to zero without compensating horizontal motions
because, as we have already noted, horizontal lines drawn backwards actually generate a
reverse horizontal motion followed by a line drawn back to the current position.
The macro concludes with an additional 1/12-em space (\") and an optional
second argument, designed to allow you to specify punctuation following the box.

A macroto box multiple lines of text (like this paragraph) is more complex. It
requires the use of a diversion to capture the text to be boxed. The diversion can then
be measured, and the Iines drawn to fit. And when you are using diversions, you need
I two macros. one to start the diversion. and one to finish it. as in the following macros:

.de BS \ " Box Start

-br \ " Space down one line; cause break
.di bX \ " Start diverting input t o macro bX
..
.de BE \" Box End
.br \ I' Ensure partial line is in bX
.nr bI In \ '' Set "box indent"--space between
\" box and text
- di \" End diversion
.nr bW \\n (dlu \ 'l Set "box width" to diversion width
.nr b H \\n (dnu \" Set "box height" to diversion height
.ne \\n (bHu+\\n ( vu \ " Make sure bH plus one line is
\" left on page
452 0 UNlX Text Processing 0

. n r fI \ \ n ( . u \ " S e t fI t o 1 i f t e x t i s f i l l e d
. nf \ " S p e c i f y n o - f i l l before p r i n t i n g b X
.ti 0
. i n +\\n(bIu Add "box i n d e n t " t o any o t h e r i n d e n t
\"
. bX Output t h e t e x t stored i n macro b X
\"
. i n -\\n(bIu S u b t r a c t b I t o restore prev i n d e n t
\"
. n r b W +2*\\n ( b A d d 2 x "box i n d e n t " t o " b o x w i d t h "
\"
.sp -1 Compensate f o r b a s e l i n e s p a c i n g
\"
\l'\\n (bWu\ (ul'\L'-\\n (bHu' \l' 1 O \ ( u l ' \h' I O ' \L' \\n (bHu'
\ " Draw box
.if \\n(fI . f i \ " R e s t o r e f i l l i f prev t e x t was f i l l e d
- SP \ " S p a c e down 1 l i n e a f t e r b o x i s d r a w n
..
There are a number of interesting things about these macros. First, they provide a
good illustration of the use of diversions. Note that the macro causes a break (with
either .b r or .sp) before the diversion is started and before it is terminated. Note
also how the predefined read-only registers dn and dl are used to measure the height
and width of the diversion and therefore set the dimensions of the box. (The contents
of these registers are not used directly when the lines are drawn because the registers
are read-only, and the width needs to be adjusted to account for a small amount of spac-
ing between the box rule and the text contained in the box.)
Second, because these macros are complex, they use quite a few number registers.
We want to use register names that are mnemonic, but not use up names that might be
useful for user-level macros. We get around this problem by using names that combine
lowercase and uppercase letters. This is entirely a matter of convention, but one that we
find preferable to mm's use of completely obscure internal register names like ;p.
Third, there is the actual line drawing-the point of this presentation. Let's look
at this aspect of these macros in detail.
As we've discussed, bH and bW have been set to the height and width, respec-
tively, of the diversion. Because the box rule is a zero-width character, however, the
macro needs to allow a small amount of space between the sides of the box and the text
it encloses. It does this by specifying a 1-en indent (which is added to any existing
indent, in case the box occurs in a block of text that is already indented). When the
diversion is output, it will thus be indented 1 en.
After the diversion is output, the indent is reset to its previous value. However,
twice the value of the indent is added to the box width. The box will thus be drawn 2
ens wider than the text it encloses. The text will start in 1 en; the right side of the box
will be drawn 1 en beyond the right margin.
The actual line to draw the box:
\l'\\n(BWu\(ul'\L'-\\n(BHu'\l' IO\(ul'\h' IO'\L'\\n(BHu*

draws a horizontal line using \ (ul from the left margin to the distance specified by
bW (box width), which, as we have seen, now includes a small extra margin. It then
draws a line back up the page to the height specified by bH, and back across the page
to the left margin again.
 0 Figures and Special Effects 0 453

At this point, even though we have drawn the bottom, right, and top sides of the
box, we are still at the top right comer of the box. The macro needs to move horizon-
tally back to the left margin, because horizontal lines to the left are actually drawn from
the left, and leave the current position the same as it was before the line was drawn. In
this case we actually want to move to the left as well. Therefore, we must do so expli-
citly, by following the \ 1 ' I 0 \ (ul' request with a \h' I 0 ' . Finally, the box is
closed by drawing a vertical line back down the left side.
The current position is now at the start of the last line of the contents of the box,
so the macro issues an .s p request to move down one line. Alternatively, you could
write this macro in such a way that it leaves no additional space above and below the
box, but lets the user leave space by issuing some kind of spacing or paragraph request.
By default, the box is drawn just long enough to surround the text it contains.
(The number register d l , which is used to set the box width, contains the width of the
text in the diversion.) For short lines in no-fill mode, the box will aIso be shorter:

Here are some short lines of text in no-fill mode.

Let's see how they come out.

This raises the idea that it might be nice to center a box that is shorter. A more com-
plete set of box macros will do this, as well as let the user change the default box indent
(the distance between the text and the edge of the box):
.de BS \" Box S t a r t
- SP
.d i bX
.n r bC 0 \" Clear centering f l a g
.n r b I 0 \" C l e a r box indent
.if "\\Sl"C" . n r bC 1 \" S e t f l a g i f u s e r wants c e n t e r e d
.if !"\\$2"" . n r b I \\$2n\" S e t box indent if specified
.-
.de BE \ " B o x End
.b r
.if !\\n(bI .nr bI I n \ " S e t b I i f n o t a l r e a d y set
.d i
.n r bW \ \ n ( d l u
.n r bH \ \ n ( d n u
.ne \\n (bHu+\\n (.Vu
.n r f I \\n(.u
.n f
.ti 0
.n r iN \ \ n ( . i u \" Save current indent
- i f \\n ( b C . i n + (\\n (. lu-\\n (bWu) /2u
\ " If c e n t e r i n g , adjust indent
.i n +\\n ( b I u
.bX
.i n -\\n (bIu
.n r b W +2*\\n (bIu
454 0 UNlX Text Processing

.sp -1
\l'\\n(bWu\ (ul'\L'-\\n(bHu'\l' IO\(ul'\h' IO'\L'\\n(bHu'
.if \\n(fI .fi
.in \\n(iNu \ " Restore original indent
- SP
..
Using the full macro, and specifying .BS C 5n, the box now looks like this:

Here are some short lines of text in no-fill mode.

Let's see how thev come out with .BS C 5n.

These macros also provide insight into how to use number registers. For example, B1
takes C as a possible argument to indicate that the box should be centered. Because
the B2 macro controls the output, there must be some way to communicate the user
request for centering between B 1 and B2. The B1 macro sets number register BC to
1 as a signal, or flag, to B2 to do the centering. (Note that BC i s first zeroed, to make
sure that centering is not propagated into the current environment from a previous invo-
cation of the box macros.)
Likewise, B Q is set as a flag to indicate whether justification i s enabled. The box
is drawn in no-fill mode, but the macro must reset filling if it was previously enabled.
The read-only number register .u is nonzero if filling is in effect, so the lines:
.nr BQ \ \ n ( . u

.if \\n(BQ .fi

will execute the .f i request only if justification was previously in effect.
Changing Line Weight
You may occasionally want to change the weight of a line you are drawing. The way to
do this is simple: change the point size with either the .ps request or the \ s escape
sequence before drawing the line. For example:
\1'3i'

will produce:

and:
\s20\1' 3i' \ S O

will produce:
 Figures and Special Effects 0 455

(This trick only works with \ 1 and \ L. It will not change the weight of lines drawn
with any of the \D escape sequences.) You might also want to consider the text size
when you are drawing boxes around text. For example, if you are using a macro like
.BX (shown previously) to draw boxes around the names of keys, you might want to
set the text 2 points smaller, either by specifying the font-switch codes as part of the
argument:
.BX "\s-2RETURN\sO"

or by modifying the macro so that they are built right in:

.de BX
\ ( b r \ I \ ~ - 2 \ \ $ l \ s O \I \ (br\l' I O \ (rn' \1' I O \ (ul' \ ^ \ \ $ 2
__
If either of these things were done, our earlier example would look like this, which is
even better:

Press the1 - key.

Drawing Curves
The previous line drawing escape sequences work in n r o f f and o t r o f f as well as
d i t r o f f . There are also additional drawing sequences that only work in d i t r o f f .
These escape sequences allow you to draw circles, arcs, ellipses, splines (curved lines
between a series of coordinates), and straight lines.
Table 15-2 summarizes these sequences. The syntax of the escape sequences is
familiar-an initial escape code is followed by a series of arguments enclosed in single
quotation marks or some other user-supplied delimiter. In this case, though, all of the
escape sequences begin with the same code-\D-with the type of item to be drawn
(circle, arc, ellipse, spline, or straight line) given by the first argument.

TABLE 15-2. ditroff Escape Sequences for Drawing

Escape Description
b' 1 x,y' Draw a line from the current position to coordinates x,y.
b ' c d' Draw a circle of diameter d with left edge at the current
position.
b fe d l d2' Draw an ellipse with horizontal diameter dl and vertical
diameter d2, with the left edge at the current position.
b ' a xl y l x2 y2' Draw an arc counterclockwise from the current position,
with center at xl ,yl and endpoint at x l +x2,yl +y2.
-
\D ' XIy l x2 y 2 . . . ' Draw a spline from the current position through the speci-
fied coordinates.
456 UNlX Text Processing 0

Learning the geometry used by these escape sequences is best accomplished by

example. Although we have shown the arguments to the line, arc, and spline sequences
as if they were x, y coordinates, they are in fact t rof f ' s usual vertical and horizontal
distances. Read x as horizontal distance, and y as vertical distance. You can get very
confused if you treat them as a true coordinate system.
Let's start simple, with individual fixed-size figures. The following input will
produce the output shown in Figure 15-1:
.sp li
.in .5i
The c i r c l e s t a r t s here\D'c li'and ends h e r e .
.sp li
The l i n e s t a r t s here\D'l li -1i'and ends here.
.sp li
The e l l i p s e s t a r t s here\D'e 2 i li'and ends here.
.sp li
The a r c s t a r t s h e r e \ D * a . 5 i 0 0 .5i'and ends here.
.sp li
The spline starts here
\D'- .5i -.5i . 5 i . 5 i .5i . 5 i . 5 i - . 5 i r a n d e n d s h e r e .
.sp .5i
.in 0

As you can see, arcs and splines are the most difficult figures to construct.
Instinct cries out for the ability to draw an arc between two endpoints with the current
position as the center of the arc. Instead, for consistency with the other figures, draw-
ing begins at the current position, and the first set of values specify the center of the
arc. This takes a little getting used to.
With splines, the problem is that distances are additive, and relative to the previ-
ous position, rather than to the initial position. Our familiarity with x, y coordinate sys-
tems leads us to think that the spline should be produced by a request like this:
\D'- .5i - . 5 i li 0 l . 5 i . 5 i 2 i 0'

(in which the x value increases relative to the origin rather than to the previous point)
instead of by the request shown previously.
You may also have noticed something rather odd. Text continues right after the
endpoint of the figure, yet the .s p li requests seem to give us 1 inch of space from
the original baseline, regardless of the endpoint of the figure. This is most obvious with
the line, which clearly moves back up the page. Yet the next figure is also spaced
down 1 inch. This fact becomes even more obvious if we do this:
.sp l i
The l i n e s t a r t s here\D'l l i -.5i'and ends h e r e .
What h a p p e n s t o t e x t t h a t w r a p s a n d c o n t i n u e s i n f i l l mode?
 0 Figures and Special Effects 0 457

nd ends here.

The line starts her

The ellipse starts her d ends here.

The arc starts her

7
L d ends here.

Fig. 15-1. Some Simple Figures

t
458 n UNlX Text Processing 0

Here’s the result:

d ends here. What happens to text that wraps and continues in fill

/
The line starts h e r d
mode?

The current baseline has not been changed. This is a major contrast to lines
drawn with \ L or \ 1. As you play with lines, you’ll also find that lines drawn to the
left with \D really do move the current position to the left, and you don’t need to add
a compensating horizontal motion if you are drawing a complex figure.
You’ll have to experiment to get accustomed to drawing figures. One other prob-
lem is to get figures to start where you want. For example, to get the endpoints of arcs
with various orientations in the right place, you may need to combine arc drawing
requests with vertical and horizontal motions.
You could use these requests to create a box with curved corners similar to the
one done with pic in Chapter 10. The box is drawn starting with the lower left
comer (so it can be drawn after the text it encloses i s output) and will look like this:

The box was drawn using the following drawing commands. These commands are
shown on separate lines for ease of reading. To make them draw continuously, we need
to add the \c escape sequence to the end of each line. This escape sequence joins
succeeding lines as if the line feed were not there. Warning: using fill mode will not
achieve the same result, because the formatter will introduce spaces between each draw-
ing command as if it were a separate word.
 0 Figures and Special Effects 0 459

\v'-.25i' \c Go back up the page 1/4 inch

\D'a .25i 0 0 .25i' \c Draw bottom left arc 1/4 inch down and to the right
\D'l 3 i O'\c Draw horizontal line 3 inches to the right
\D'a 0 -.25i .25i O'\c Drawbottomrightarc 1/4inchupandtotheright
\D'l 0 -2i'\c Draw vertical line 2 inches back up the page
\D'a -.25i 0 0 -.25i'\c Drawtoprightarc1/4inchupandtotheleft
\D'l - 3 i O'\c Draw horizontal line 3 inches to the left
\D'a 0 . 2 5 i -.25i O'\c Drawtopleftarc 1/4inchdownandtotheleft
\D'l 0 2i'\c Draw vertical line 2 inches down the page
\v' . 2 5 i ' Restore original baseline position
To build a complete macro to enclose examples in a simulated computer screen,
we can adapt the .Bl and .B2 macros shown previously:
.de S S \ " Start S c r e e n w i t h
\" Curved Corners
.sp .5v
.ie !"\\$inn . n r BW \\$1\" G e t w i d t h f r o m f i r s t arg
.el . n r BW 4 i \ " or s e t d e f a u l t i f n o t s p e c i f i e d
-ie ! "\\$2"" . n r BH \\$2\" G e t h e i g h t f r o m s e c o n d a r g
.el . n r BH 2 . 5 i \ " or s e t d e f a u l t i f n o t s p e c i f i e d
- br
. d i BB
.-
.de SE \ " S c r e e n E n d
. br
. n r B I In
. i f \\n(.$>O . n r B I \ \ $ l n
-di
.ne \\n (BHu+\\n ( . V u
.nr BQ \ \ n ( . j
.nf
.ti 0
.in +\\n ( B I u
-in + (\\n ( . lu-\\n (BWu) /2u
.sp .5
.BB
.sp + (\\n (BHu-\\n ( d n u )
. i n -\\n(BIu
. n r BH - . 5 i
. n r BW + 2 * \ \ n ( B I u
. n r BW - . 5 i
\v'-.25i'\c
\D'a . 2 5 i 0 0 .25i'\c
\ D ' 1 \\n(BWu O'\c
\D'a 0 -.25i . 2 5 i O'\c
\ D ' 1 0 -\\n(BHu'\c
460 0 UNlX Text Processing 0

\D'a -.25i 0 0 -.25i'\c

\ D ' 1 -\\n(BWu O'\c
\D'a 0 .25i -.25i O ' \ c
\D'l 0 \\n(BHu'\c
\v' .25i'
.sp -1.5
.if \ \ n ( B Q .fi
.br
.sp .5v
._
Because a screen has a fixed aspect ratio, we don't want the box to be proportional to
the text it encloses. Hence, we give the user of the macro the ability to set the box
width and height. If no arguments are specified, we provide default values.
Because the box size is fixed, there are some additional steps necessary in the
closing macro. First, we must decrement the specified box width and height by the dis-
tance used in drawing the curves, so that the user gets the expected size. Second,
because the box is drawn from the lower left comer back up the page, we must make
sure that the lower left comer is correctly positioned before we start drawing.
To do this, we again need to use a diversion. We measure the height of the diver-
sion, then add enough additional space (. s p + ( \ \n ( BHu- \ \ n (dnu)) to bring the
starting point for drawing low enough so that the box i s not drawn back over the text
that precedes the invocation of S S . (If you don't understand why this was done,
delete this line from the macro, and watch the results.)
We've also centered the screen by default, and added a half-line of vertical spac-
ing above and below the box. (As an exercise, modify the .BX macro to produce a
key-cap with curved comers.)

Talking Directly to the Printer

Depending on the output device and postprocessor you are using, you may be able to
send specialized control commands directly to your printer. For example, you may be
able to embed raster graphics images (such as a file created on an Apple Macintosh
with MacPaint) directly in your output. Or if you are using a PostScript-driven printer,
you can integrate figures done with MacDraw, or issue PostScript commands to print
grey screens over your text.
These capabilities are provided by the two requests \ ! and . c f , copy filename
[to standard output] ( d i t r o f f only).
The \ ! request is the transparent output indicator. Any text following this
escape sequence on a line is placed directly into the output stream, without any process-
ing by t r o f f . This makes it possible to insert control lines that will be interpreted
by a postprocessor or an output device. (As mentioned in the last chapter, transparent
output is also useful for embedding control lines in a diversion, to be executed when the
text in the diversion is output.)
 Figures and Special Effects 0 46 1

Likewise, the contents of the file specified as an argument to .c f are placed

directly on standard output, without processing by d i t r o f f .
Unfortunately, there is a catch! PostScript is a page-description language that
resides in the printer. Before you can talk directly to the printer, you must get through
the postprocessor that translates d i t r o f f output into PostScript. I f the postprocessor
mucks with the data you send out, all bets are off.
As of this writing, Transcript, Adobe Systems? own t rof f -PostScript con-
verter, does not allow you to use \ !. However, with Pipeline Associates? devps,
any lines beginning with ! are ignored by the postprocessor, and go directly to the
printer. This allows you to use transparent output by entering the sequence \ ! ! fol-
lowed by the appropriate PostScript commands. Or, if you are sending a PostScript file
created on the Mac, use an editor to insert an exclamation point at the beginning of each
line.
In any event, this is not a job for the novice, since you must learn PostScript as
well as t r o f f. Experiment with your printer and postprocessor, or ask around to see
if other users have solutions you can adapt to your situation.

Marking a Vertical Position

There are many cases, both in macros and in the body of your text, where you may
want to mark a spot and then return to it to lay down additional characters or draw
lines.
The .mk request marks the current vertical position on the page; .r t returns to
that position. This is useful for two-column processing. To give a simple example:
Two columns are useful when you have a linear list
of information that you want t o p u t side-by-side, but don?t
want t o bother rearranging with the cut-and-paste programs.
-sp - 5
- 1 1 2.5i
.nf
-mk
Item 1
Item 2
Item 3
- 1 1 5i
.in 2.75i
.rt
Item 4
Item 5
.in 0
- SP
This example produces the following output:
462 0 UNlX Text Processing 0

Two columns are useful when you have a linear list of

information that you want t o put side-by-side, but
don’t want t o bother rearranging with t h e cut-and-paste
programs.

Item 1 Item 4
Item 2 Item 5
Item 3

Notice that it is entirely your responsibility to make sure that the second column doesn’t
overprint the first. In this example, we did this by manually adjusting the indent and
the line length. In addition, because the second column is shorter than the first, a con-
cluding .s p is necessary to return to the original position on the page. If this had not
been done, subsequent text would overprint the last line of the first column.
Proper two-column processing for an entire document requires a much more com-
plex setup, which must in part be handled by the page bottom macro. We’ll look at that
in detail in Chapter 16, but this example should be enough to give you the idea.
The .mk request can take as an argument the name of a number register in
which to store the vertical position. This allows you to mark multiple positions on a
page, and return to them by name. The .r t request always returns to the last position
marked, but you can go to a position marked in a register using the .s p request:
.mk Q
.sp I\nQu
or (more to the point of the current discussion) with \v:
\v’ I \nQu’

In addition, .r t can take as an argument a distance from the top of the page.
That is:
. r t 3i

will return to a point 3 inches from the top of the page. The .mk request need not be
used in this case.

Overstriking Words or Characters

There are a number of escape sequences that allow you to overstrike words or charac-
ters to create special effects. These include

boldfacing an entire font by overstriking;

marking and returning to a specific horizontal position;
calculating the width of a word and backing up over it;
centering two characters on top of each other;
. stacking characters vertically.
 Figures and Special Effects 0 463

Boldfacing a Font by Overstriking

The .bd request specifies that a font should be artificially boldfaced by overstriking.
The request has two forms, one for ordinary fonts and one for the special font.
A request of the form:
. b d font offset

will overstrike all characters printed infont by overprinting them, with the second strike
offset from the first by onset-1 basic units. The following:
.bd S font offset
will overstrike characters printed in the special font, whilefont is in effect. And:
.bd font
.bd S font

will restore the normal treatment of the font.

This request is particularly useful when you are boldfacing headings and want to
account for special characters or italics in arguments supplied by the user. (This
assumes that you don't have an explicit bold italic font.) Especially at sizes larger than
10 points, the stroke weights of bold and italic fonts can be quite different.
For example, assume that you had a macro that produced a boldface heading for a
table:
.de Th \ " T a b l e Heading
.ft €3
.ce
T a b l e \\$l: \\$2
.ft P

If the user supplied italics or special characters in the arguments to the macro, the con-
trast between the different character weights might not be as pleasing as it could be.
For example:
.Th "3-1" "Special Uses f o r \(sr i n \ f I t r o f f \ f P "

would produce:

I Table 3-1: Special Uses for 4 in trofs

If the macro had .bd requests added like this:

.de Th \ " T a b l e Heading
.ft B
.bd I 3
.bd S B 3
- ce
T a b l e \\$l: \\$2
.ft R
 1
464 u UNlX Text Processing 0

.bd I
.bd S
..
the output would look like this:

Table 3-1: Special Uses for d in troff

Another example is provided by the constant-width (CW) font used in this book.
Because the font is optimized for the Laserwriter, where the ink bleeds slightly into the
paper, the font is very light on the typesetter. Throughout this book, we have embol-
dened this font slightly, with the requests:
.bd CW 4
.bd S CW 4
This sentence shows how the c o n s t a n t w i d t h font looks without these requests.

Marking and Returning to a Horizontal Position

Just as you can mark a vertical position, you can also mark and move to a specific hor-
izontal position. This is chiefly useful for overstriking characters.
Just as you use a value stored into a register with the .mk request to indicate a
fixed vertical location on the page, you mark a horizontal location with \ k. Then, you
can use the absolute position indicator I to specify the distance for \ h.
To borrow an example from Kernighan's TroffTutorial:

\kxword\h' I \nxu+2utword
will artificially embolden word by backing up almost to its beginning, and then over-
printing it. (At the start of word, \ k stores the current horizontal position in register
x. The \h' I \nxu+2u' sequence returns to that absolute position, plus 2 units-a
very small offset. When word is printed a second time, an overstriking effect is
created.)
This sequence might be useful if you were using a font that had no bold
equivalent, and in circumstances where the .bd request could not be used because the
special effect was not desired for all instances of that font on the same line. And, to be
really useful, the sequence should probably be saved into a macro.

The Width Function

The \w escape sequence returns the length of a string in units. For example:

\w'Hi there'

will tell you the length of the string H i there.

 Figures and Special Effects 0 465

This sequence returned by \w can be used as an argument with \h or with any

horizontally oriented request (such as .i n ) . This has many uses, which we'll intro-
duce as we need them.
To give you an idea of how to use \w,though, we can rewrite the example used
with \ k as follows, to produce the same effect:
.de BD \ " A r t i f i c i a l l y embolden word
\\$l\h'-\~' \\$1'-2~' \\$l

This macro prints its first argument, then backs up the width of that argument, less two
units. Then it prints the argument a second time- at a two-unit offset from the first.
Hint: to avoid awkward constructions involving nested \w sequences, first read the
width into a number register. For example, the previous macro could be rewritten like
this:
.de BD \ " Artificially embolden word
.nr W I (\w' \\$I' -2u)
\\Sl\h'-\\n (WIu' \\$I

In this case, the difference isn't so great; however, at other times the sequence can
become too confusing to read easily.

Overstriking Single Characters

Although \ k provides a good method for overstriking an entire word, there are also
more specialized functions for overstriking a single character.
The \ o sequence takes up to nine characters and prints one on top of the other.
This is most useful for producing accents, and so forth. For example, \ o r e A ' pro-
duces C.
You can also produce other interesting character combinations, although you may
need to tinker with the output to get it to look just right. For example, we once tried to
simulate a checkmark in a box with the sequence: \ o r \ ( sq\ ( sr' . (Note that the
special character escape sequences are treated as single characters for the purpose of
overstriking.) This example produced the following output:

The square root symbol is too low in the box, so we tried to introduce some local
motions to improve the effect, like this:
\ o r \ (sq\v'-4p'\ (sr\v'4pr'

Unfortunately, this didn't work. Although you can nest most escape sequences inside
each other (as long as you use the correct number and order of delimiting quotation
marks), local motions do not work with \o. However, there was a solution.
The \ z sequence also allows overstriking, but in a different way. The \ o
sequence knows the width of each character, and centers them on top of each other.
The \ z sequence simply outputs the following character, but does not space over it.
That means the current position after printing the character is the same as it was before
466 UNlX Text Processing 0

the character was printed. A subsequent character will have its left edge at the same
point as the character immediately following the escape sequence. Because \ z does
allow you to mix vertical motions with overstriking, it solved our problem.
Because all these escape sequences can be a bit much to type, we defined the
checkmark in a box as a string:
.ds CK \ z \ ( s ~ \ \ v ' - ~ P ' \( s K \ \ v ' ~ ~ '

After we did that, simply typing \ * (CK will produce Id.

Stacking up Characters
The \b sequence also does a kind of oversmking-it stacks the characters in the fol-
lowing string. It was designed for use with eqn. There are special bracket-building
characters that are meant to stack up on top of each other. See Table 15-3.

TABLE 15-3. Bracket-Building Characters

Character Name Description

r \ ( 1t Left top of big curly bracket
I \ (lb Left bottom
1 \ (rt Right top
J \ (rb Right bottom
{ \ ( 1k Left center of big curly bracket
t \ ( rk Right center of big curly bracket
I \ (bv Bold vertical
1 \ ( 1f Left floor (left bottom of big square bracket)
1 \ ( rf Right floor (right bottom)
r \ (IC Left ceiling (left top)
1 \ (rc Right ceiling (right top)

A typical invocation looks like this:

\b'\ (It\(lk\ (lb'
which produces:

When you're creating a tall construct like this, you need to allow space so that it
doesn't overprint preceding lines. You can create space above or below the line with
.s p requests. However, this will cause breaks. Although ' s p might do the trick, it
is sometimes hard to predict just where the break will fall.
The trof f program has a special construct designed to solve just this problem
of a tall construct in the middle of filled text. The \x request allows you to associate
extra interline spacing with a word. A positive value specifies space above the line; a
negative value specifies space below the line. So, when illustrating the previous
 Figures and Special Effects 0 467

bracket-building function, we could have shown the results inline, like this
than in an example broken out by blank lines. Typing the sequence:

\br\(lt\(lk\(lb'\x'9p'\x'-gP'

gives us the result we want.

The \ x sequence is also useful when you want to allow extra space for an over-
sized letter at the start of a paragraph. (You've probably seen this technique used in
some books on the first paragraph of a new chapter. It was commonly used in
illuminated manuscripts.)
An application of \b that you might find useful is to create vertically stacked
labels. For example, consider the following macro, which will put such a label in the
outside margin of a book:
.de SL
.mk \ " Mark current vertical position
.ft B \ " Change to bold font
.cs B 24 \ " We'll explain this later
- P O -.25i \ " Shorten the page offset by 1 / 4 inch
.It +.Si \ " Extend the title length used by .tl
\ " This request will be explained later
.if e .tl '\b:\\$l:"' \ " Use .tl to put stacked label
.if o .tl "'\b:\\$l:' \" in the margins
.It -.Si \ " Restore original title length
- P O +.25i \ " Restore original page offset
.cs €3 \ " We'll explain this later
.ft \ " Restore original font
.rt \ " Return to original vertical position C
.. I
e
So, for example: v
.SL "Clever Trick!" e
r
will produce the effect shown in the margin.
T
r
i
Tabs, Leaders, and Fields C
k
We discussed tabs in Chapter 4. However, there are a couple of additional points that I

need to be covered. When you enter a tab on a typewriter, the typing position shifts
over to a predefined position, or tab stop. In n r o f f and troff, what is actually
generated is the distance from the current position on the input line to the next tab stop.
What this means is best illustrated by an example that will not work. Suppose
you want to create a table of contents in which one entry (the page number) is all the
way over to the right margin, and the other (the heading) is indented from the left, like
this:
468 UNlX Text Processing 0

7 I
Getting Started
Turning On the Powerl-2
Inserting Diskettes 1-3
You might be tempted to code the example as follows (where a tab is shown by the
symbol I I >:
.ta 6.5iR
Getting Started1 11-1
.in .5i
Turning O n t h e Power1 11-2
Inserting Diskettes1 11-3

This will not work. Indents cannot be combined with tabs. A tab character generates
the distance from the current position on the input line to the tab stop. Therefore, the
page number will be indented an additional half-inch-extending out into the right
margin-instead of staying where you put it.
The way to achieve this effect (in no-fill mode) is to use either spaces or tabs to
manually indent the first text string.
When you use right or center-adjusted tabs, the text to be aligned on the tab is the
entire string (including spaces) from one tab to the next, or from the tab to the end of
the line. Text is aligned on a right-adjusted tab stop by subtracting the length of the
text from the distance to the next tab stop; text is aligned on a center-adjusted tab stop
by subtracting half the length of the text from the distance.

Using Leaders
A leader works like a tab; however, it produces a character string instead of horizontal
motion. A single character is repeated until a specific horizontal position is reached.
There is actually a leader character, just as there is a tab character. But there is no key
for it on most keyboards, so it is not obvious how to generate it. The magic character is
AA (CTRL-A), and you can insert it into a file with v i by typing V A A (CTRL-V,
CTRL-A).
If you insert a AA into your file where you would normally insert a tab (inciden-
tally, the tab itself is equivalent to V, and will show up as such if you display a line
with ex's :1 command), you will generate a string of dots. For example:
.nf
.ta li 2.5i 3.5i
I IFirst"ASecond^AThird
.fi

will produce:

I I
I I

First ....................................... Second ...................Third

 Figures and Special Effects 0 469

You can change the leader character from a period to any other single character with the
ICrequest. For example, you could create a fill-in-the-blanks form like this:
.nf
.ta li 3iR
.IC -
Signatuse:l I
Date: I I ^A
- fi
This example would produce the following output in t r o f f:

Signature:
Date:
As you can see from the example, tabs and leaders can be combined effectively to line
up both ends of the underlines.
A second way to create leaders is to redefine the output of the tab character with
.t c . This request works just like IC, only it redefines what will be output in
response to a tab character. For example, if you issue the request:
.tc .
a tab character (AI) generates a string of repeated dots, just like a leader ('!A). However,
you will then lose the ability to intermix tabs and leaders on the same line, as in the
previous example.
Issuing a .t c request without an argument will restore the default value, which
generates motion only. (Incidentally, the same is true of .lc-that is, .ICwithout
an argument will cause leaders to generate motion only, just like tabs. To reset the
leader character to its default value, you need to request .I C .).

Using Fields
In addition to tabs and leaders, n r o f f and t ro f f support fields, which are blocks
of text centered between the current position on the input line and the next, or between
two tab stops.
The .f c request allows you to specify a delimiter that indicates the boundaries
of the field, and a second character (called the pad character) that divides the contents
of the field into subfields. A blank space is the default pad character. The .f c
request without any arguments turns off the field mechanism. This request is a little
difficult to explain, but easy to illustrate. The requests:
- nf
.ta li 3i
.fc #
I I #Hi t h e r e #
I !#Hi how are you#
- fc
- fi
470 UNlX Text Processing 0

will produce the following output:

Hi there
Hi how are YOU

Within the field, the pad character (a space by default) is expanded so that the text
evenly fills the field. The first line contains only a single space, so the two words are
adjusted at either end of the field. The second line contains three spaces, so the words
are evenly spaced across the field.
By specifying a pad character other than a space, you can achieve fine control
over spacing within the field. For example, if we modify the input like this:
.fc # "
I I#Hi"how are"you#
.fc

we'll get this result:

I Hi how are YOU I-

What's this good for? To return to our fill-in-the-blanks example, the construction:
- nf
.ta .5i 2i 2.5i 4i
.fc # "
.IC -
I I"AI I "A
-sp - 5
I I#"Signature"#l I #*DateA#
.fc
.IC I

.fi
would produce the following output:

I I

Signature Date

You should also know that - f c, like many other advanced formatter requests, is used
by the t b l preprocessor to create complex tables. It is wise to avoid using it inside a
table.

Using Tabs and Leaders in Macros

Within a macro definition, tabs and leader characters are not interpreted. They will take
effect when the macro i s used, not when it is defined. Within a macro definition, you
can also specify tabs and leaders with the escape sequences \t and \a. These
 Figures and Special Effects 0 471

sequences are also not interpreted until the macro is used, and can be substituted for the
actual tab or leader characters whenever interpretation is to be delayed.

Constant Spacing
One font that you may frequently encounter, especially in the d i t r o f f environment,
is called CW (constant widrh). It is the font used in this book for examples. It has
become something of a convention in computer books to print all “computer voice”
examples-input from the keyboard, the contents of a file, or output on the screen-in a
constant-width font. (This convention is based on the fact that in many computer
languages, precise indentation is syntactically or at least semantically significant, and
the Variable-width typesetting fonts cannot preserve the alignment of the original text.)
When you use a constant-width font, you are essentially asking t r o f f to act like
n r o f f-to work in a realm where all characters, and all spaces, are of identical width.
To use the constant-width font in d i t r o f f , request it like any other font, using
either the request .f t CW or the escape sequence \ f (CW. In o t r o f f , depending
on the output device, you could use constant width by using a preprocessor called cw,
which got around the four font t r o f f limit by handling the constant-width font in a
separate pass. See the description of cw in your UNIX Reference Manual if you are
interested in the details. (There are other ways to do this as well, depending on the out-
put device and the postprocessor you are using to drive it. For example, we used
o t r o f f with TextWare International’s t p l u s postprocessor and the HP LaserJet.
To get around the font limit, we set a convention in the postprocessor that 1 1 -point type
was actually constant width, and then used the .c s and .ss requests to give
t r o f f the correct spacing.)
There is also a request that allows you to simulate the effect of a constant-width
font even when you are using a variable-width font. The .c s request tells t r o f f :
“Use the spacing I give you, even if it doesn’t match what you’ve got in your width
tables.” The request takes up to three arguments. The first two arguments are the most
commonly used. They are the font to be so treated and the width to be used, in 36ths of
an em. By default, the em is relative to the current type size. By using the optional
third argument, you can use the em width of a different type size. So, for example:
.cs E 21
Space t h e bold f o n t a t 21/36 o f an e m .
.cs E 21 12
Space the b o l d f o n t a t 2 1 / 3 6 o f a 1 2 - p o i n t em.

Let’s see what we get with these requests:

Space the bold font at 21/36 of an em.

S p a c e t h e b o l d f o n t at 2 1 / 3 6 o f a 12-
P O i n t em.
472 UNlX Text Processing 0

To return to normal spacing for the font, use .cs without a width argument. For
example:
.cs B

will return control of spacing for the bold font to t r o f f ' s width tables.
Although the results are not always aesthetically pleasing, it may be necessary to
use this request if you have a real need to represent constant-width text. It is also use-
ful for special effects. For example, you may have noticed that in the headings of each
chapter of this book, the word Chapter is broadly and evenly spaced, and the boxes
underneath align with the letters. This was done with the c s request.
The .c s request i s also useful when you are creating vertically stacked labels, as
shown earlier in this chapter. Normally, characters are positioned with their left edge at
the current position on the output line. When constant spacing with .cs is in effect,
the left comer of the character box is placed at that position, and the character itself is
centered in the box. You can see the difference between this graphically in the follow-
ing example:
.sp .7i
.ft B
.in li
. mk
\ b' Variable '
.in 3i
. rt
.cs B 24
\b' Constant'
. br
.cs B
. ft
.in 0
.sp .7i

which produces:

V C
a 0
r n
i S
a t
b a
I n
e t

The ss request is a closely related request that sets the space size. The default
size of an interword space in t r o f f is 12/36 of an em; for true constant-width effects,
you should set it to the same size as the font spacing you have set with - cs.
 1

0 Figures and Special Effects 0 473

Pseudo-Fonts
Using the .bd request to create a bold italic is not the only way to simulate a nonstan-
dard font, at least in dit rof f . In dit rof f, there are two new escape sequences,
\ S and \H. The \ S sequence slants characters by a specified number of degrees.
(Positive values slant characters to the right; negative values slant characters back to the
left.) For example:
\ s r 15’
will slant characters 15 degrees to the right. This can be used to create a pseudo-italic
font. The \ S sequence without an argument turns off slanting.
The \ H sequence sets the character height to a specified point size without
changing the width. For example, if type is currently being set at 10 point, the con-
struct:
\ H ‘ 12‘

will create characters that are 12 points high, but only 10 points wide (assuming you are
at the default 10-point size). A height value of 0 turns off the function.
These escape sequences will only work on certain output devices. You’ll have to
experiment to find whether or not they’ll work in the setup you’re using.

Character Output Translations

“Garbage in, garbage out” is a truism of computer science. You get out of a computer
what you put in. However, there are cases in n r o f f and t rof f in which what you
put in i s not the same as what you get out.
The first of these cases is only true for troff. It involves a special class of
characters called ligatures. As we’ve previously discussed, typeset characters have dif-
ferent widths. Even so, when two narrow characters are printed together, such as a pair
o f f s , or anfand an i, there is excess space between the characters.
To get around this problem, there are special characters cdled ligatures, which are
really single characters designed so that they appear the same as a pair of narrow char-
acters. (These are truly single characters, defined as such in t ro f f ’ s character set.)
The ligature characters and the equivalent individual characters are:
Input Ligature Equivalent Characters

\ (fi fi fi
\ (fl fl fl
\ (ff ff ff
\ (Fi ffi ffi
\ (F1 ffl ffl
The t rof f formatter automatically converts any of these groups of characters to the
equivalent ligature, although all ligatures are not supported by every output device.
474 0 UNlX Text Processing n

(For example,fi a n d J are the only ones in the standard PostScript fonts.) You can turn
this conversion off with the request:
.lg 0
and restore it with:
- 1g
Normally, you won’t need to do this, but there are special cases in which it may hang
you up, and you’ll need to know what to do. We’ll get to one of them in a moment.
The .t r (translate) request provides a more general facility for controlling out-
put character conversions. It takes one or more pairs of characters as an argument.
After such a translation list has been defined, t r o f f will always substitute the second
character in each pair for the first, whenever it appears in the input.
Let’s look at some examples. First, consider the case encountered throughout this
book, in which we illustrate the syntax of various requests without actually executing
them. For example, we want to show a period at the start of a line or the backslash that
starts an escape sequence, without actually having them executed.
W e could simply insulate the special characters from execution. For example, we
can put the zero-width character \ & in front of a period that begins a request, and we
can double all backslashes ( \ \ will appear as \ in the output) or use the \e escape
sequence, to print \.
However, this grows tedious and hard to read in the input file. Another approach
is to do a character translation:
.tr #.%\\ \ ” Translate # t o ., % to \

(As usual, we have to double the backslash.) Now, whenever # appears in the input,
. appears in the output, and whenever % appears in the input, \ appears in the output.
So, in our examples, we can actually type:
#sp l i %” Space down one i n c h

But what appears on the page of this book is:

- s p li \” Space down one inch

The translations are built into the example start and end macros. (The end macro resets
the characters to their normal values.)
If you translate characters with .t r, be sure to restore their original values
correctly when you are done. To reset the previous translation to the normal character
values, the request is:
.tr ##%% \ ” T r a n s l a t e # t o #, % to %

In addition, the translation must be in effect at the time the line is output. If you
translate characters without first causing a break, any partially filled line will be affected
by the translation.
I t is also possible (and recommended in some of the t r o f f documentation) to
use - t r to substitute some other character (usually -) for a space. This creates an
equivalent to the unpaddable space.
 0 Figures and Special Effects 0 475

.tr -
This will allow you to type single characters for unpaddable spaces; your input text will
be more readable and will line up properly on the screen.
Yet another application of - t r, and one that you will find useful in designing
macros for chapter headings and so on, is to translate lowercase input into uppercase,
and then back again:
.de UC \ " Translate input t o uppercase
.tr aAbBcCdDeEfFgGhHiIjJkKlLmMnNoOpPqQrRsStTuUvVwWxXyYz2
\\$I
- br
.tr aabbccddeeffgghhiijjkkllmmnnooppqqrrssttuuvvwwxxyyzz
..
(The break is important. These character translations must be in effect at the time the
line is output, not when it is read into the buffer.)
It is in this last case that you may have trouble with ligatures. If the .UC macro
were defined as shown in the previous example, the line:
.UC t rof f

might produce the following output:

I 1
I TROff I
To have the macro work correctly, we would need to turn ligatures off ( - lg 0) for the
duration of the translation.

Output Line Numbering

Do you remember the treatment of the p r o o f shell script in Chapter 12? It was such
a long example that it required line numbers that could be referred to later in the text.
The n r o f f and trof f programs provide requests that allow you to automatically
number output lines as was done in that example.
The nm (number) request turns numbering on or off. The request:
.nm [+IN
will turn numbering on, with the next line numbered N. For example, the next para-
graph is numbered with .nm 1.

1 A 3-digit arabic number followed by a space is placed at the start of each line.
2 (Blank lines and lines containing formatter requests do not count.) The line length is
3 not changed, so this results in a protruding right column, as in this paragraph. You may
4 need to decrease the line length (by \w* 0 0 0 ' u) if you are numbering filled text
5 rather than an example in no-fill mode. (Be sure to notice the space following the three
6 zeroes.) We'll do that from now on, so only the current paragraph will protrude.
476 UNlX Text Processing 0

There are several optional arguments as well: a step value, the separation
2 between the number and the beginning of the line, and an indent that will be added
to the line. By default, the step value is 1, the separation is 1 , and the indent is 0.
4 For example, if you specified:
.nm 1 2

6 every second line would be numbered, as was done at the start of this paragraph.
The .n n (not numbered) request allows you to temporarily suspend number-
ing for a specified number of lines, as was done for this paragraph using the request
.nn 4 . The specified number of lines is not counted. This could be useful if you
were interspersing numbered lines of code with a textual discussion.
To turn numbering off entirely, use .nm without any arguments. We?ll do
8 that now.
The last line number used by nm is saved in the register In, and it is possible
to restart numbering relative to that number by preceding the initial line number you
give to - nm with a + or a -. For example, to restart numbering at exactly the point
it was turned off, you can use this request:
.nm + O

Let?s do that now. As you can see, numbering resumes just where it left off, with
10 the same step value and indent, as if no intervening lines had been present. After
this line, we?ll turn numbering off entirely.

When using .nm in fill mode, you have to watch for breaks. Because .nm itself
does not cause a break, it make take effect on the output line above where you expect it.
You may need to force an explicit break before .n m to make sure numbering starts on
the next line.

Change Bars
The .m c (margin character) request allows you to print ?change bars? or other marks I
in the margin, as was done with this paragraph. This is especially useful if you are I
revising a document, and want to indicate to reviewers which sections have changed.
You can specify any single character as the margin character-so don?t restrict
yourself to change bars when thinking up uses for this request. For example, you could
use an arrow, or the left-hand character (\ ( l h ) to draw attention to a particular point in
the text, like this. (These characters are oddly named. The right-hand character -im
( \ ( r h ) is a left-hand that points to the right (m);the left-hand character ( \ (lh)is a
right hand that points to the left (m). These characters are mapped onto arrows on
some output devices.)
You can control the distance the mark character appears from the margin with an
optional second argument. If no argument is given, the previous value is used; if there
is no previous value, the default distance is 0 - 2 i in n r o f f and lm in t r o f f.
 0 Figures and Special Effects 0 477

Incidentally, on many UNIX systems, there is a version of d i f f , called

d i f fmk, that will compare two versions of a file, and produce a third file containing
. m c requests to mark the differences. Additions and changes are marked with a bar in
the margin, as shown previously. Deletions are marked with an asterisk.
In our business, we find this very useful for producing interim drafts of technical
manuals. We archive the first draft of the manual, as it was turned in to our client.
Then, after review changes have been incorporated, we use d i f f m k to produce an
annotated version for second draft review:
-
$ d i f f m k d r a f t 1 d r a f t 2 marked d r a f t
$ ditroff ... -
marked d r a f t

This could also be done by manually inserting .mc requests as the edits were made.
But, as stated in Chapter 12, why not let the computer do the dirty work?

Form Letters
No formatter would be complete without the ability to create form letters that merge
existing text with externally supplied data. The n r o f f and t r o f f programs are no
exception in providing requests to handle this type of problem.
The .r d (read) request allows you to read from standard input. This request
prints a prompt on standard error (the user's terminal) and reads input up to a pair of
newlines. For example, you could have a form letter constructed like this:
.n f
.rd E n t e r-t h e-p e r s o n ' s -name
. r d E n t e r-t h e-company
. r d E n t e r-t h e -s t r e e t
. r d E n t e r-t h e -c i t y , -s t a t e , -and-z i p
- SP
.fi
Dear
. r d E n t e r-t h e-s a l u t a t i o n
- SP

Unfortunately, .rd terminates the prompt at the first space, and does not recognize
quotation marks to delimit an entire string as the prompt. As a result, for a wordy
prompt, you must tie the string together using an unobtrusive character like an under-
score, as was done here.
Here's what would happen when this letter is formatted:
$ nroff l e t t e r I lp

E n t e r-t h e-p e r s o n ' s -name: Tim O'Reilly

E n t e r-t h e-company: O ' R e i l l y 6 Associates, I n c .

478 0 UNlX Text Processing 0

E n t e r-t h e-s t r e e t : 981 Chestnut Street

E n t e r-t h e-c i t y , -s tate, -a n d-z ip: Newton, MA 02164

E n t e r-t h e-s a l u t a t i o n : Tim:

Note that a colon is appended to the prompt, and that the RETURN key must be pressed
twice after each response. If no prompt is specified, - r d will ring the terminal bell
when it expects input.
In addition, the input need not come from the keyboard. It can come from a pipe
or from a file. There are two other requests that come in handy to create a true form
letter generation capability.
The .n x (next) request causes the formatter to switch to the specified file to con-
tinue processing. Unlike the .s o request discussed in Chapter 4, it doesn’t return to
the current file. The .e x request tells the formatter to quit.
You can put these requests together with .r d . First, create a list of variable data
(like names and addresses) either in a file or as the output of a database program. Then
pipe this file to the formatter while it is processing a letter constructed like this:

.n f
- rd
- rd
- rd
- SP
.f i
Dear
.r d
Body of letter here

Sincerely,

Jane D o e
- bP
.nx l e t t e r

The .nx request at the end of the form letter causes the file to reinvoke itself when
formatting i s complete. Assuming that the standard input contains a sequence of name,
street, city (et al), and salutation lines, one line for each .rd request, and address
block, in the data file, that are each separated by pairs of newlines, you can generate an
endless sequence of letters.
However, be warned that formatting will continue in an endless loop, even when
the standard input has run out of data, unless you terminate processing. This is where
. e x comes in. By putting it at the end of the list of names coming from standard
input, you tell the formatter to quit when all the data has been used.
 Figures and Special Effects 0 479

The command line to produce this marvel (assuming a form letter in a file called
letter and a list of names followed by an .ex request in a file called names)
would be:
$ cat names I nroff letter I lp
or:
$ nroff < names I lp
It is possible to imagine a more extensive data entry facility, in which a variety of blank
forms are constructed using t r o f f, and filled in with the help of a data entry front
end.* To generalize the facility, you could associate the various fields on the form with
number register or string names, and then interpolate the number or string registers to
actually fill in the form.
This approach would allow you to reuse repeated data items without having to
query for them again. Even more to the point, it would allow you to construct the data
entry facility with a program other than t r o f f (which would allow features such as
data entry validation and editing, as well as increased speed). The data entry front end
would simply need to create as output a data file containing string and number register
definitions.

Reading in Other Files or Program Output

In addition to .nx, don’t forget the .s o (source) request, which allows you to read
in the contents of another file, and then return to the current file.
We’ve mentioned this request briefly in the context of reading in macro
definitions. However, you can also use it to read in additional text. In our business,
we’ve found it very useful in certain types of manuals to break the document into many
separate files read in by - so. For example, we often need to write alphabetically-
ordered reference sections in programming manuals. Unfortunately, the developers
often haven’t finalized their procedure names. If the section consists of a list of . s o
requests:
.so BEGIN -M O D U L E
.so BUFFER
.so CONFIGURE

the job of reorganization is trivial-all you need to do is change the filenames and real-
phabetize the list.

*For this idea, I am indebted to a posting on Usenet, the UNIX bulletin board network, by Mark Wallen of
the Institute for Cognitive Science at UC San Diego (Usenet Message-ID: <[email protected]>, dated June
13, 1986).
480 UNlX Text Processing 0

The only caution, which was mentioned previously in Chapter 8, is that you can’t
include data that must be handled by a preprocessor, such as tables and equations. A
quick look at the command line:
$ tbl ,file I nroff

will show you that the preprocessor is done with the file before the formatter ever has a
chance to read in the files called for by the .s o request. Some systems have a com-
mand called soelim that reads in the files called for by .so. If you use soelim to
start the file into the pipeline, there is no problem.
One useful tip: if you are using soelim, but for some reason you don’t want
soelim to read in a file because you would rather it were read in by t rof f, use
s o rather than .s o to read in the file. The soelim command will ignore the s o
request.
Another interesting request is .s y . This request executes a specified system
command. If the command has output, it is not interpolated into the t r o f f output
stream, nor is it saved. However, you can redirect it into a file, and read that file into
t r o f f with .c f (or with .s o , if you want it processed by t rof f instead of sent
directly to the output stream).
 C H A P E
. T R

16

What’s in a Macro Package?

In Chapters 4, 14, and 15, you’ve seen almost all of the individual formatting requests
that nrof f and trof f provide, and many examples of groups of requests working
together in macros. However, writing individual macros is still a far cry from putting
together a complete package.
In Chapters 5 and 6, you’ve seen the features built into the m s and mm macro
packages, so you can imagine the amount and complexity of macro definitions. Perhaps
you have even looked at a macro package and determined that it was impossible to
decipher. Nonetheless, it is possible even as a beginner to write your own macro pack-
age or to make extensions to one of the existing packages.
In this chapter, we’ll look at the structure of a macro package-the essentials that
allow you to handle basic page formatting. Then, in the next chapter, we’ll look at a
macro package with extensions for formatting large technical manuals or books. Even
if you have no plans to write a new macro package, this chapter will help you under-
stand and work with existing packages.

Just What Is a Macro Package, Revisited

When considering what a macro package is, you might think only of the visible features
provided by macros in existing macro packages. But a macro package is more than a
collection of user macros that implement various features. Failing to understand this
fact might cause someone to import an mm macro into an ms-based macro package,
and never understand why this macro fails to work.
Individual macros are dependent upon other elements of the macro package,
which sometimes makes it hard to isolate a particular macro, even for purposes of
understanding what it does. These interdependencies also make it difficult to under-
stand what a macro package is doing. That is why we want to look at the underlying
structure of a macro package, and not just the obvious features it provides. We want to
look first at what a macro package must do before we look at what it can do.

481 .
482 0 UNlX Text Processing 0

A macro package is a structure for producing paged documents. The nroff

and t r o f f formatters do the actual collecting and formatting of lines of text, as
steadily as a bricklayer placing bricks in a row. But they do not define the structure
that is so obvious by the end result. Fundamentally, it is the macro package that defines
the placement of lines on a page. At a minimum, a macro package must set traps and
create macros to handle page transitions. It usually also defines the layout of the physi-
cal page.
A macro package may also provide a way to arrange the parts of a documents and
affect their appearance. Remember the distinction we made earlier between formatting
and formats. A format reflects the type of document being produced, just as a floor
plan reflects the functions of rooms in a building. For instance, a technical manual
might consist of chapters and sections that require headings. Other elements might be
bulleted lists and numbered lists, a table of contents, and an index. These elements help
readers to identify and to locate important parts of the document. But these features-
so obviously important to users-are really not the essential elements in a macro pack-
age.
Page formatting is the foundation of a macro package, and this foundation must
be solid before you can build a variety of custom document formats.

New or Extended?
The first question to ask when you contemplate writing a whole new package is whether
you need to do it all yourself or can simply build on an existing package.
There are benefits to either approach. The existing macro packages are quite
complex (especially mm). It can be easier to start over, writing only the macros you
need, than to learn the intricate internals of m s or mm. A custom macro package can
be quite small, including only macros for page transition (which can be minimal, as we
shall see) and whatever other macros you want. This is the best approach if you have
something specific in mind.
As with all programming projects, though, you may find your package growing
larger than intended, as your needs and understanding grow and you start to add
features. A macro package begun haphazardly can also end that way, without any con-
sistent structure.
If you do find yourself wanting to create an entire macro package, rather than just
a few macros, you should think about modular programming techniques. Modular pro-
gramming suggests that you break the tasks to be performed by a program into the
smallest possible functional units, then build up larger tasks with each of these smaller
units. This not only helps with debugging and testing of new macros, but also makes it
much easier to write macros, because you end up with a library of low-level general-
purpose macros that perform important functions. You don’t have to reinvent the wheel
for each new macro.
There are numerous advantages to building on the existing packages, especially if
you want to have a general-purpose package:
 ~

What’s in a Macro Package? 483

They already contain a wide range of useful macros that you not only can use
directly, but can call on within new macros.
They are tested and proven. Unless you are very experienced at text process-
ing, it is difficult to foresee all of the kinds of problems that can arise. When
you write your own package, you may be surprised by the kinds of errors that
are filtered out by the design of m s or mm.
If you are familiar with m s or mm, adding a few extended macros to your
repertoire is easier than learning an entire new package.
It can be easier than you expect to modify or add to them.

In our own work, we have chosen to extend the m s macro package rather than to build
an entirely new package. In this chapter, though, we’re going to take a hybrid
approach. We’ll build a minimal ms-like package that illustrates the essentials of a
macro package and allows users who don’t have access to the full m s package to make
use of some of the extensions described in this and later chapters.
In this “mini-ms” package, we have sometimes pared down complex aiacros so
it is easier to understand what they are doing. We try to uncover the basic mechanism
of a macro (what it must do). As a caveat to this approach, we realize that simplifying
a macro package can reduce its functionality. However, we see it as part of the learning
process, to recognize that a macro in a certain situation fails to work and understand the
additional code needed to make it work.

Implementing a Macro Package

As discussed in Chapter 4, the actual option to n r o f f and t r o f f to invoke a macro
package is -mx, which tells the program to look in the directory / u s r / l i b / t m a c
for a file with a name of the form tmac.x. This means you can invoke your own
macro package from the command line simply by storing the macro definitions in a file
with the appropriate pathname. This file will be added to any other files in the format-
ting run.
If you don’t have write privileges for / u s r / l i b / t m a c , you can’t create the
tmac.x file (although your system administrator might be willing to do it for you).
But you can still create a macro package. You will simply have to read it into the for-
matter some other way. You can either

include it at the start of each file with the .s o request:

.so /usr/fred/newmacros

or list it on the command line as the first file to be formatted:

$ nroff /usr/fred/newmacros myfile

Nor do the macros need to be stored in a single file. Especially if you are using a pack-
age as you develop it, you may want to build it as a series of small files that are called
484 0 UNlX Text Processing 0

in by a single master file. You may also want to have different versions of some mac-
ros for nrof f and trof f. So, for example, the mh (Hayden) macros used to for-
mat this book are contained in many different files, which are all read in by .s o
requests in /usr/lib/tmac/tmac. h:
.so /work/macros/hayden/startup
.so /work/macros/hayden/hidden
.so /work/macros/hayden/ch.heads.par
. s o /work/macros/hayden/display
. s o /work/macros/hayden/ex.figs
. s o /work/macros/hayden/vimacs
. s o /work/macros/hayden/lists
. s o /work/macros/hayden/stuff
. s o /work/macros/hayden/index
. s o /work/macros/hayden/cols

Or, like mm,you might have two large files, one for n r o f f and one for t rof f. In
/usr/lib/tmac/tmac .m, you find:
.if n .so /usr/lib/macros/mmn
.if t .so /usr/lib/macros/mmt

In extending an existing macro package, you are not restricted to creating a few local
macro definitions that must be read into each file. You can make a complete copy of
one of the existing packages, which you can then edit and add to. Or even better, you
can read the existing package into your own package with - so, and then make addi-
tions, deletions, and changes. For example, you might create a superset of ms as fol-
lows:
- \ " /usr/lib/tmac/tmac.S - superset of m s - invoke as -mS
. s o /usr/lib/tmac/tmac.s \ " Read in existing package
. s o /usr/macros/S.headings
. s o /usr/macros/S.examples
. s o /usr/macros/S.toc

Building a Consistent Framework

One of the chief factors that distinguishes a macro package from a random collection of
macros is that the package builds a consistent framework in which the user can work.
This consistent framework includes:

. Setting traps to define the top and bottom of each page. This is the one essen-
tial element of a macro package, because it is the one thing nroff and
troff do not do.
I
0 What's in a Macro Package? 0 485

Setting default values for other aspects of page layout, such as the page offset
(left margin) and line length. (The default page offset in n r o f f is 0, which
is not likely to be a useful value, and t r o f f ' s default line length of 6.5
inches is really too long for a typeset line.)
Setting default values for typographical elements in t r o f f such as which
fonts are mounted, the point size and vertical spacing of body copy and foot-
notes, adjustment type, and hyphenation.
Giving the user a method to globally modify the default values set in the macro
package, or temporarily modify them and then return to the defaults.

In a very simple macro package, we might set up default values for t r o f f like this:
.PO li \" Set page offset t o o n e inch
-11 6i \" Set line length t o six inches
.ad 1 \" Adjust left margin only
.hy 14 \" Hyphenate, using all hyphenation rules
.wh 0 NP \" Set new page trap at t h e t o p of t h e page
\" (see below for details)
.wh -1i FO \ " Set footer trap

(We are assuming here that t r o f f's default values for point size and vertical spacing
are acceptable. In o t r o f f, we also need to mount the default fonts with .f p , as
described in Chapter 4; in d i t r o f f, a default set of fonts is already mounted.)
Simply setting up explicit default values like this will do the trick, but for a more
effective and flexible macro package, you should take the further step of storing default
values into number registers. This has numerous advantages, as we'll see in a moment.

Using Number Registers to Increase Flexibility

Writing t r o f f macros i s essentially a kind of programming. If you pay heed to the
principles learned by programmers, you will find that your macros are more effective, if
at first somewhat more complex to write and read.
One important lesson from programming is not to use explicit (so called "hard-
coded") values. For example, if you supply the indent in a paragraph macro with an
explicit value, such as:
.in 5n

you make it difficult for users to change this value at a later time. But if you write:
.in \\n(INu

the user can change the indent of all paragraphs simply by changing the value stored in
number register I N . Of course, for this to work, you must give a default value to the
I N register.
In programming, the process of setting variables to a predefined starting value is
called initialization. To give you an idea of the kinds of variables you might want to
initialize, Table 16-1 lists the values stored into number registers by the m s macros.
486 0 UNlX Text Processing 0

TABLE 16-1. Number Registers Used in m s

Description Name Value

trof f nroff
Top (header) margin HM li li
Bottom (footer) margin FM li li
Point size PS 1OP 1P
Vertical spacing vs 12P 1P
Page offset PO 26/27i 0
Line length LL 6i 6i
Title length LT 6i 6i
Footnote line length FL \\n (LLu*ll/l2 \\n (LLu*11/12
Paragraph indent PI 5n 5n
Quoted paragraph indent QI 5n 5n
Interparagraph spacing PD 0 . 3 ~ lv

The mm package uses many more number registers-in particular, it uses number
registers as flags to globally control the operation of macros. For example, in addition
to registers similar to those shown for m s in Table 16-1, there are registers for para-
graph type, numbering style in headings, hyphenation, spacing between footnotes, as
well as counters for automatic numbering of figures, examples, equations, tables, and
section headings. (See Appendix B for a complete listing.) However, the registers used
in m s should give you a sufficient idea of the kinds of values that can and should be
stored in registers.

An Initialization Sequence
In the m s macro package, a major part of the initialization sequence is performed by
the .IZ macro.* This macro is executed at the start of a formatting run; then it is
removed. Let's take a look at a much simplified version of the initialization sequence
for an ms-like package:
.de IZ \" Initialization macro
\" Initialize Number Registers
.nr HM li \" Heading Margin
.nr FM li \" Footing Margin
.nr PS 10 \ '' Point Size
.nr VS 12 \" Vertical Spacing
.nr P O li \" Page Offset
.nr L L 6i \" Line Length

*There's no real reason why this sequence needs to be put in a macro at all, other than the consistency of
putting two backslashes before number registers when they are read in.
 0 What's in a Macro Package? 487

.nr L T 6i \ " Length of Titles for .tl

.nr F L \\n (LLu*11/12 \ " Footnote Length
.nr P I 5n \ " Paragraph Indent
.nr Q 5nI \ " Quoted Paragraph Indent
.nr PD 0 . 3 ~ \ " Interparagraph Spacing
\ " Set P a g e Dimensions through requests
-PS \\n P S
.vs \\n vs
- P O \\n POU
- 1 1 \\n LLU
.It \\n LTu
.hy 1 4 \ " Specify hyphenation rules
\ " Set Page Transition Traps
.wh 0 NP
.wh -\\n(FMu F O
.wh -\\n(FMu/2u BT
..
.IZ \ " Execute IZ
.rm IZ \ " Remove I Z
As you can see, the initialization sequence stores default values into registers, then actu-
ally puts them into effect with individual formatting requests.
A number of the points shown in this initialization sequence will be a bit obscure,
particularly those relating to trap positions for top and bottom margins. We'll return to
the topic of page transitions shortly.

A Reset Macro
After you have initialized number registers, the next question is how to make use of the
default values in coding. Some registers, like a paragraph indent, will be used in a
paragraph macro. But where, for example, might you use the L L register?
First of all, as suggested, putting default values into number registers allows users
to change values without modifying the macro package itself. For instance, a user can
globally change the interparagraph spacing just by putting a new value into the P D
register.
However, the package itself can use these registers to periodically reset the default
state of various formatting characteristics.
The m s package defines a macro called . R T (reset), which is invoked from
within every paragraph macro. The .R T macro

turns off centering-. ce 0 ;

turns off underlining-. ul 0;
restores the original line length-. 1 1 \ \ n ( LLu;
488 0 UNlX Text Processing 0

restores the original point size and vertical spacing-. p s \ \n (PS and .vs
\\n (VS;
restores the indent that was in effect before any .I P , .R S , or . R E macros
were called (too complex to show here);
changes back to the font in position 1-. f t 1;
turns off emboldening for font 1--. bd 1;
sets tabstopsevery5n--.ta 5n 10n 15n 20n ... ;
turns on fill mode-. f i.

This is part of the m s error recovery scheme. Rather than aborting when it encounters
an error, m s frequently invokes the .RT macro to restore reasonable values for many
common parameters.
If you have used m s for a while, and then switch to another package, you may
find all kinds of errors cropping up, because you've come to rely on this mechanism to
keep unwanted changes from propagating throughout a document. For example, sup-
pose you create a macro that decrements the line length:
-11 -5n

but you forget to increment it again. You may never notice the fact, because m s will
restore the line length at the next paragraph macro. Other packages are far less forgiv-
ing.
Unless you plan to explicitly test for and terminate on error conditions, it is wise
to implement a reset facility like that used by m s .
A simple ms-like reset macro follows:
.de RT \ " R e s e t
.ce 0 \" Turn o f f centering, i f i n effect
.ul 0 \" Turn o f f u n d e r l i n i n g , i f i n effect
-11 \\n(LLu \" Restore default l i n e length
-ps \\n(PS \" Restore default point size
.vs \ \ n ( v S \" Restore default v e r t i c a l spacing
.ft 1 \" Return to f o n t i n p o s i t i o n 1
. t a 5 n 1 0 n 1 5 n 2 0 n 2 5 n 3 0 n 3 5 n 4 0 n 4 5 n 5 0 n 5 5 n 6 0 n 65n 7 0 n
.fi \ " R e s t o r e f i l l mode
.-
The m s version of.RT also ends any diversion invoked outside of the standard m s
macros that create diversions. Thus, a reset may occur within a keep ( .KS, .KE),
footnotes (. F S , .FE), boxed material (. B1, - BZ), and tables ( - T S , .TE) without
ending the diversion.
If you look at the actual m s reset macro, you will see that it calls another macro,
named .BG, the very first time it is itself called. The .BG macro removes the macros
associated with the unused Bell Labs technical memorandum formats (because the for-
mat has already been determined at that point). Like .1 Z, the .BG macro is only
called once during a formatting run. In our emulation, we don't make use of the
 0 What‘s in a Macro Package? 0 489

Technical Memorandum macros so we have not implemented the . B G macro. How-

-
ever, one could easily appIy the idea behind the B G macro: to execute a macro before
we begin processing the body of a document. This can be useful if a format requires a
number of preliminary or header macros that supply information about the document.

. PageTransitions
A single page transition macro is the only macro that must be implemented for n r o f f
and t r o f f to produce paged output. An example of this simplest of all possible
macro packages follows.*
.de NP \ ” New P a g e
‘bP
‘sp li
Ins
..
.wh - 1 . 2 5 i NP
-br
.rs
.sp Ili

The page transition is triggered by a trap set 1.25 inches from the bottom of the page.
When output text reaches the trap, the . N P macro is executed, which breaks the page
(but not the line), spaces down 1 inch, and enters no-space mode. The three lines fol-
lowing the macro and trap definition take care of the special case of the first page, for
which the .N P macro is not invoked.
The .w h request, which sets the location of the traps used for page transition,
interprets the value 0 as the top of the page. Negative values are interpreted relative to
the bottom of the page. So, for example, assuming that the page length is 11 inches,
the requests:
.wh 1 O i BT \ ” Bottom Title Macro

and :
.wh -1i BT \ ” Bottom Title Macro
are equivalent. The second form is the most common.
This simple “package” provides only one macro for page transition. The bottom
margin of the text portion of the page is determined by the trap location; the top margin
by a spacing request in the macro executed at the trap. However, it is far more com-
mon to work with at least two page transition macros: one for the page top and one for
the bottom.

*This “package” was contributed by Will Hopkins of VenturCom, Inc.

490 0 UNlX Text Processing 0

An example of a two-trap, two-macro macro package is given below:

.wh 0 NP
.wh -1i FO
. d e NP \"New Page
'sp li
.tl * T o p o f P a g e \\n%"' \".tl does not cause break
'sp 12i
' ns
..
. d e FO \"Page F o o t e r
'sp .25i
.tl " P a g e Bottom"
' bP
..
A trap is set at the top of the page (. wh 0) to execute the .NP macro. This macro
provides a top margin and outputs a title in that space. The formatter begins processing
lines of text until the bottom of the page trap is encountered. It invokes the . F O
macro, which supplies a footer margin and outputs a centered title. The . F O macro
then causes a page break, which in turn invokes .NP at the top of the new page. It is
important that both of these macros avoid causing a break, so that text in fill mode will
continue smoothly onto the next page.
By setting traps for both the top and bottom of a page you have more control over
the size of the bottom and top margins, the placement of headers and footers, and
advanced features like footnotes and multiple-column processing.
Take some time to experiment with this bare bones macro package. If you place
it in a file, such as pagemacs, you can use it to format text files, as in the following
example:
$ nroff pageanacs t e x t

No-Space Mode in Page Transitions

No-space mode is often used in a page transition macro to keep space from being output
at the top of a page. It is standard page makeup for the top line of each page to begin
at the same point. Without no-space mode, a spacing request (such as prespacing in a
paragraph macro) that falls just before the page transition would result in space being
output at the top of the page, causing uneven positioning of the top line of the page.
Any output text lines restore space mode, so you don't have to explicitly turn it
back on. However, if you explicitly want to put space at the top of the page (to paste in
a figure, for example), use .rs (restore spacing) before the spacing request. The fol-
lowing sequence can be used to start a new page and space down 2 inches below the
top margin:
 0 What?sin a Macro Package? 0 49 1

?
bP
.r s
?sp 2 i

This works in all cases, except on the first page. You must force a break on the
first page before you can restore spacing. An .f 1 request will do the trick:
.fl
-rs
.sp 3 i
-ce
A T i t l e on a T i t l e Page
- bP
The .f 1 request is useful when you want to flush the line buffer and cause a break.

The First Page

As you might expect from the previous example, the first page is unlike others that fol-
low it. That is because there is no automatic transition to the first page. To get around
this, the formatter causes a ?pseudo-page transition? when it first encounters a break or
begins processing text outside a diversion.
For the top of page trap to be executed on the first page, you must set the trap and
define the top of page macro before specifying any request that causes a break or ini-
tiates processing. You can test this with the sample macros by putting an explicit .b r
request before the .NP macro definition. After that test, try replacing .br with a
.t 1 request. Even though this request does not cause a break, it does initiate process-
ing of text, and so the .NP macro is not executed.

Page Transitions in m s
Let?s take a closer look now at the trap positions we set in the initialization sequence
for our ms-like package, together with the definitions of the macros placed at those
positions:
.de I2

\? S e t Page Transition Traps

. w h 0 NP
. w h -\\n(FMu FO
. w h -\\n(FMu/2u BT
\I* D e f i n e Page T r a n s i t i o n Macros
..
.de NP \ ?I New P a g e M a c r o
?sp \\n (HMu/2u
.PT
492 UNlX Text Processing 0

'sp I\\n(HMu
'ns
-.
.de F O \" Footer Macro
'bP
..
.de P T \" P a g e T o p T i t l e Macro
.ti ' \ \ * ( L H ' \ \ * (a!'\\*(RH'
-.
.de B T \" Bottom T i t l e Macro
.tl ' \ \ * (LF'\\* (CF'\\* (RF'
'sp .5i
..
You'll notice a couple of differences from our earlier example. Instead of specifying
"hard-coded'' values for trap locations, we have set up a top margin value in the regis-
ter HM (header margin) and a bottom margin value in F M (footer margin).
Now we have three trap locations and four page transition macros. In the simpli-
fied form shown here, you may wonder why so many macros are used for this simple
task. We'll look at that later, as we show some of the additional things that are done in
these macros. But for the moment, let's focus on what these macros are. Their trap
locations are shown in Figure 16-1.

.NP (new page) is invoked by a trap at the top of each page ( .wh 0 NP). It
spaces down I/z the distance specified in the HM register, calls the P T macro,
and then spaces down the full distance specified by the header margin.
.P T (page title) prints out a three-part title consisting of user-definable strings
LH, CH, and RH (left header, center header, and right header).
.FO (footer) is invoked by a trap at the distance from the bottom of the page
specified by the FM register (.wh -\\n (FMu FO). This macro causes a
break to a new page. Note the use of ' bp rather than .bp so that any par-
tially filled line is not output, but is held till the next page.
.BT (bottom title) is invoked by a trap at I/z the distance from the bottom of
the page specified by the F M register ( .wh - \ \ n ( FMu / 2 u BT).

Although this sequence is different than our earlier example, it is about as easy to
understand. The main difference, however, is that there are two traps at the bottom of
the page. The first (FO) causes a page break, and the second (BT) places the footer.
Even though the first trap caused a page break, the formatter keeps going till it reaches
the true bottom of the page specified by the page length. On its way, it passes the
second trap that invokes .B T .
The use of the four page transition macros is slightly inconsistent in m s ; .P T i s
invoked from .N P , but - BT, which could just as well be invoked by .FO, is instead
invoked by a trap.
 0 What's in a Macro Package? 0 493

NP
----------- PT -----------
- ---------- text starts -----------

Headers and Footers

Most books, and many shorter documents, include headers and footers. In books,
headers often include the title of the book on the left-hand page, and the title of the
chapter on the right. The footer typically includes the page number, either centered or
in the lower outside comer, alternating from left to right. (Although all three elements
are usually present, they can be in different positions depending on the book design.)
As previously mentioned, the t 1 request was designed specifically for produc-
ing the three-part titles used in headers and footers. The m s package uses this request
in both the P T and B T macros, filling the three fields with symmetrically named
string invocations. If the string is undefined, the field is blank.
The macro package itself may define one or more of the fields. The .IZ macro
from ms contains this piece of code:
. i f "\\*(CH"" . d s CH "- \ \ \ \ n ( P N -
.if n .ds CF " \ \ * ( D Y

The m s macros define the center header as the page number set off by hyphens. For
nrof f only, the center footer i s set to the current date. (An n r o f f-formatted docu-
ment is assumed to be a draft, but a t ro f f -formatted document is assumed to be final
camera-ready copy.)
494 0 UNlX Text Processing 0

The m s macros transfer the page number from the % register to one called PN.
Note the number of backslashes required to get the page number output at the proper
time-not in the string definition, nor in the macro definition, but at the time the title is
output.
If you don’t like this arrangement, you can simply redefine the strings (including
redefining them to nothing if you want nothing to be printed). As a developer of mac-
ros built on top of m s , you could, for example, have a chapter heading macro automati-
cally set the chapter title into one of these strings. (More on this later.)
Headers and footers are often set in a different type and size than the body of the
book. If you are using a standard macro package, font and size changes can simply be
embedded in the header or footer strings:
.ds LH “\fIAn Introduction t o Text Processing\fP

Or, if you are writing your own macros or redefining an underlying package like m s ,
you can embed the changes directly into the t 1request: .
.ti \s-2\\* (LF’\ \ * (cF’\\* (RF\sO’

Another point: it is often desirable to alternate headers and footers on odd and even
pages. For example, if you want to put a book title at the outside upper comer of a
left-hand (even) page, and the chapter title at the outside upper comer of a right-hand
(odd) page, you can’t really work within the structure m s provides.
To do this properly, you could use a construct like the following within your
.P T macro:
.if e . t l ‘\\*(TI”’
.if o .if \\n%-l . t l “ ‘ \ \ * ( C H ‘

where the string TI holds the title of the book, and CH holds the title of the chapter.
If it’s an odd page, we also test that it’s not the first page. By invoking specific strings,
you do lose the generality of the mechanism provided by m s .

Page Numbers in Three-Part Titles

Inasmuch as the chief application of three-part titles is in producing header and footer
lines from within page transition macros, there is a special syntax for including page
numbers. A % character anywhere in a title will be replaced with the current page
number. This saves the trouble of having to do a proper number register interpolation
and makes it easier for unsophisticated users of m s or mm to include page numbers in
header and footer strings.
Whenever nro f f or t r o f f makes use of a character in a special way, you
can be sure there is a back door that allows you to change that character. The - p c
@age character) request allows you to specify that some other character than % should
perform this function:
- PC A
\ ” Use A instead of % t o p r i n t page # i n .tl

This does not change the name of the % number register, which also contains the page
number.
 0 What's in a Macro Package? 0 495

Title Length
The other thing you can adjust is the length of the three-part title produced by .tl.
Usually it is set to the same length as a text line, but this need not be so. Title length is
specified independently of line length with the .It (length [ofl title) request. For
example:
.It 6 . 5 i

The title length is not independent of the page offset, so if you want a title that is longer
than the line length, yet centered on the page, you will need to muck with the page
offset as well. (Note that this is most likely to occur within a page transition macro.)
- POli \" Page Layout Defaults
.I1 6 . 5 i

-It 7 i
.tl ' A l c u i n User' s Guide' '%' \ " T i t l e w i l l e x t e n d 1/2 i n c h
\'r past r i g h t margin

-PO -.
25i
-It 7 i
.tl ' A l c u i n User's Guide' '%* \" T i t l e w i l l e x t e n d 1/4 i n c h
-PO + . 2 5 i \" on e i t h e r s i d e
An - 1t request without an argument will restore the previous title length.

Some Extensions to the Basic Package .

Thus far, we've looked at what it will take to implement a small ms-like macro pack-
age. Now let's look at some extensions to the basic structure of the package that will
make it more flexible. These extensions could be added to the minimal package shown
earlier in this chapter, or they could be added to a full m s package, as we have done at
our own site.

Changing Page Size

As mentioned earlier, the initialization sequence usually sets up default values for line
length, page offset, and the placement of the top and bottom traps. In the standard m s
package, all of these values are set up to produce an g1/2-by-l1 inch page.
This is fine for n r o f f, but with t r o f f , one might well want to produce a dif-
ferent page size. For example, many books are typeset for a 5I/2-by-g1/2inch page.
The most obvious move is to change the page length:
496 0 UNlX Text Processing 0

.pl 8.5i

and then reset the line length, title length, and page offset using the standard registers
m s provides.
This may not work if your output device uses continuous-roll paper, such as a
typesetter. However, in n r o f f,or when using t r o f f with a sheet-fed laser printer,
this may split your formatted output pages across physical sheets of paper. (Some dev-
ices translate a .bp into a page eject code or formfeed;others simply add blank lines
to push text onto the next physical page. For this reason, it i s perhaps preferable to
think of .p l as the paper length rather than the page length.)
In addition, when you are printing a small page, it is nice to print cut marks to
show the location of the page boundaries. If you change the page length, any cut marks
you place will be off the page that t r o f f knows about, and will not be printed.
For both of these reasons, we took a different approach. We modified the m s
.I Z macro so that changing the header and footer margins would effectively change
the page size, instead of just the margins. (In standard ms, you can change the size of
the top and bottom margins, but this doesn't change the page size, because the place-
ment of the footers is fixed after the initialization macro has been called. The trap posi-
tion for F O is reset at the top of every page, thus taking into account changes in the
value of the F M register. But the trap position for B T is never touched after .I Z has
been executed.)
In our package, we decided to set up some standard page sizes as part of .I Z.
In our business, writing and producing technical manuals, we often print books in both
sizes. Early drafts are printed on the laser printer in 8 1 / 2 by 1 1 format; later drafts and
final camera-ready copy are produced in by 8'12 format. We also produce quick-
reference documents in a narrow 6-panel card or pamphlet. The user selects the size by
specifying a command-line switch. This approach has the advantage of letting the user
change all of the parameters associated with a given size with one easy command.
The .I z macro in our mini-ms package now looks like this:
.de IZ \ " Initialization macro
\ " Initialize Number Registers
\ " Quick Reference Card size
.if \ \nS=2 I t \
nr pw 3.5i \" P a g e Width
nr t H 1.25i \" Trim Height adjustment
nr LL 2.8i \" Line Length
nr LT 2.8i\} \" Title Length
\" 5 1/2 b y 8 1/2 size
.ie \\nS=l \t\
nr pw 5.5i \v' Page Width
nr t H 1.25i \" Trim Height adjustment
nr LL 4.253. \" Line Length
nr LT 4.25i\) \" Title Length
\" 8 1 / 2 by 11 size
.el \ I \
nr \ " Page Width
 0 What's in a Macro Package? 0 497

nr tH 0 \ " Trim Height adjustment

nr LL 6i \ " Line Length
nr LT 6i\} \ " Title Length
\"Values independent of page size
.nr FM li \ '' Footer Margin
.nr HM li \ " Header Margin
.nr PO li \ " Page Offset
.nr PS 10 \ " Point Size
.nr VS 12 \ " Vertical Spacing
.nr FL \\n(LLu*11/12 \ " Footnote Length
.nr PI 5n \ " Paragraph Indent
.nr QI 5n \ " Quoted Paragraph Indent
.nr PD 0 . 3 ~ \ '' Interparagraph Spacing
\ " Set Page Dimensions through requests
-ps \\n(PS
.vs \\n(vs
.PO \\n(POu
- 1 1 \\n(LLu
.It \\n(LTu
.ft 1
.hy 14 \ " Specify hyphenation rules
\ " Set Page Transition Traps
.wh 0 NP \ " Top of page
.wh - (\\n(FMu+\\n(tHu) FO \ " Footer
.wh - ( (\\n(FMu/2u)+\\n (tHu) BT \ " Bottom titles
.if \\nS .wh -\\n(tHu CM \ " Position of bottom mark
.-
The .NP macro has been modified as follows:
.de NP \ " New Page Macro
'sp \\n(tHu \ I' Space down by trim height
.ie \\nS \ { \
CM \ If small format, print cut mark
I
SP \\n (HMu/2u-l~\}\"Correct baseline spacing
.el 'sp \\n(HMu/2u \ " Space down by half HM
.PT
' SP I \\n (HMu+\\n(tHu \ " Space to HM plus adjustment
' ns
. *

By simply setting the S (size) register from the command line, the user can choose
from one of three different sizes. For example:
$ d i t r o f f -Tps -rS1 textfile I devps 1 lp
page size.
will choose the 5L/2-by-81/2

k
498 0 UNlX Text Processing 0

What we've done here is to assume that the paper size is still 8 l / 2 by 1 1 . We've
defined a fudge factor, which we've called the trim height adjustment, and stored it in a
register called tH. If the user has set the size register from the command line, we use
this adjustment factor to:

shift the location of the footer trap:

.wh - (\\n (FMu+\\n (tHu) FO

. shift the location of the bottom title trap:

.wh - ( (\\n (FMu/2u) +\\n ( t H u ) BT

. place a new trap to print cut marks at the true bottom of the page:
.if \\nS . w h -\\n(tHu CM

. space down at the start of the

' s p \\n(tHu
.NP macro:
. i e \\ns \ { \
CM
I sp \\n (HMu/2u-lv\}
. e l 'sp \\n(HMu/2u
- PT
' sp I \\n (HMu+\\n ( t H u

Note that in .NP we need to adjust for the extra line spacing that occurs as a result of
printing the cut marks. Otherwise, the .PT macro would be invoked one line lower on
a page with cut marks than on one without.

Cut Marks
We've mentioned that if you are producing typeset or laser-printed copy on less than an
8]/z by 1 1 page, it is usually desirable to place marks showing the actual page boun-
dary. The paper i s then cut on these marks in preparation for pasteup on camera-ready
boards.
A s you've seen in the preceding discussion, we print the cut mark at the top of
the page from the .NP macro, after spacing down by the desired trim height. The cut
marks at the bottom of the page are printed by calling the cut mark macro with a trap
placed at the trim height from the bottom of the page.
As you'll notice, the execution of the cut mark macro is conditioned on the pres-
ence of the S register, which indicates that the user has requested a small page.
Here's a simple version of the actual cut mark macro:
.de CM \" C u t Mark m a c r o
'PO - (\\n (pWu-\\n ( L L u / 2 u ) \" C e n t e r c u t mark a r o u n d t e x t
.It \\n(pWu \" S e t t i t l e l e n g t h f o r c u t mark
'tl '+"+' \" P r i n t c u t mark
 0 What's in a Macro Package? 0 499

.It \ \ n ( L T u \" Reset t i t l e length

'po + (\\n (pWu-\\n (LLu/2u) \ " R e s e t page o f f s e t
..
As with all activity that takes place during the page transition, it is very important
that nothing in the cut mark macro causes a break. For this reason, all break causing
requests are started with the no-break control character ('), and the cut marks them-
selves are printed with .t 1, which doesn't cause a break. (The other way to avoid
breaks is to do all of your page transition work in a different environment, but doing
this uses up one of the environments, which might be better used for another purpose.)
We've specified the width of the page in the pW register. To center the cut
marks around the text, we adjust the page offset by the difference between the page
width and half the line length. Then we set the title length to the page width, and actu-
ally print the cut marks with .t 1. Then, of course, we reset the original page offset
and title length.
In the implementation shown, we use simple plus signs to create the cut marks.
This creates a slight inaccuracy, because the page width will be from end to end of the
plus signs, and the height from baseline to baseline, rather from the center of the plus as
we'd like.
There are two ways that we could deal with this. One is to fudge the height and
the width to account for the character widths. The other is to use a specially drawn
mark that will put the actual cut lines at the edge rather than the center of the figure.
A very simple way to do this is to use the box rule, the root-en, and the underrule.
Because the cut marks are no longer symmetrical, though, we'll need to give the cut
mark macro an argument to specify whether we're at the top or the bottom of the page:
.de CM \" Cut Mark macro
'PO - (\\n (pWu-\\n ( L L u / 2 u ) \" C e n t e r c u t mark a r o u n d t e x t
'It \\n(pWu \" S e t t i t l e l e n g t h f o r c u t mark
. i e n \ \ $ l n T " ' t l ' \ (br\ (rn"\ ( r n \ (br' \ " P r i n t c u t mark
. e l ' t l ' \ (br\ (ul"\ ( u l \ (br'
'It \ \ n ( L T u \" Reset t i t l e length
'PO +(\\n(pWu-\\n(LLu/2u) \" Reset page offset
..
When we invoke .CM from within .NP,we'll just have to add the argument T to
specify we're at the top.
The cut marks will look like this:

r 1
L J
500 UNlX Text Processing 0

Other Exercises in Page Transition

We've looked at the basic mechanism for page transition, and shown one way to extend
that mechanism to allow the user to select different page sizes. We have not exhausted
the topic of page transition, however. Before we begin to discuss the development of
macros that prescribe document formats, rather than basic page formatting, we will
briefly consider these topics:

Footnotes
Multicolumn processing
Page top resets
Handling widows and orphans

Footnotes
Footnotes make page transition an even more complex exercise. Anyone who has typed
footnotes on a typewriter knows the problem. Because the presence of a footnote shor-
tens the space available on the page for regular text, you need to know the size of the
footnote before you know if its reference will fit on the bottom of the current page, or
will be pushed to the top of the next. There is always the possibility of a classic
Catch-22: a footnote whose reference falls at the bottom of the page only if the foot-
note itself isn't printed there.
Let's look first at a very simple footnote mechanism-ne that has a reasonable
chance of failure in the face of heavy demand, but nonetheless illustrates the basic
mechanism at work.
The first thing we need to know i s the position of the page bottom trap for a nor-
mal p a g e u n e without any footnotes. For example, in m s , we know that its location
is -\\n (FMu. (Now m s has a perfectly good footnote mechanism, but for purposes
of argument, we're going to pretend we need to add one.)
All we really need to do, on the simplest level, is to save footnotes in a diversion,
measure them, then move the footer trap back up the page by a distance equal to the
size of the diversion.
In the new page macro, we initialize (reset to 0) a counter (fC)that will tell us if
there are any footnotes on the page and how many. (We want to handle the first foot-
note differently than others on that page.) We also initialize a bottom position for print-
ing footnotes (Fb) and initialize it with the value of the standard footer margin. (This
will be the starting point that will be decremented as footnotes are encountered.) Last,
we provide a reset that restores the page footer trap at the standard footer margin if it
has been changed because of footnotes on a previous page.
\ " Add t o .NP
.nr f C 0 1 \" I n i t i a l i z e footnote counter
.n r F b 0-\\n (FMu \r' I n i t i a l i z e footnote position
-c h FO -\\n ( F M u \ " R e s e t normal footer l o c a t i o n
 0 What's in a Macro Package? D 501

Now, a pair of foomote macros are required to open and close a diversion:
.de FS \" Footnote Start
.nr fC 1 \" Set flag that there are footnotes
.ev 1 \" Use environment 1
.da FN \" Divert text of footnote
.if \\n(fC=l \ { \ \" If first footnote
\1' li' \" Print 1 inch line before it
.br\ I
..
.de F E \ " Footnote End
.br
.di \ " End diversion
.ev \ " Restore environment
.nr Fb -\\n(dn \ " Decrement footnote position by
\ " size o f diversion;
\ " note that Fb is already negative.
\ " Reset footer t r a p
.ie (\\n (nl+lv)> (\\n ( .p+\\n (Fb) .ch FO \\n (nlu+lvu
.el .ch F O -\\n(Fb

The footnotes are processed in a separate environment. This environment needs to be

initialized, perhaps as part of the .Iz macro, or as part of the .FS macro the very
first time it is called. The latter method makes it easier for users to change settings for
this environment. It is recommended that you preserve a separate environment (either 1
or 2) for footnote processing. Here is a sample initialization sequence:
.ev 1 \ " Initialize first environment for footnotes
.ps 8
.vs 1 0
- 1 1 \\n(FLu \ " FL was initialized t o 11/12 of LL
. ev

.
The F S macro opens a diversion (. da F N ) into which we append the text of the
footnote. Before the first footnote on a page, the .FS macro adds a one-inch reference
line to mark the beginning of footnotes. After we have closed the diversion in the .FE
macro, we obtain the size of it from the read-write register .dn. This amount is used
to increase Fb (two negatives amounts are added together) and change the location of
the footer trap further up the page.
Before changing that trap, the footnote end macro has to find out if the new footer
trap will be placed above or below the current location. If the new trap location is
below where we are, all is well; the page trap is moved up to that location. However, if
the current footnote places the location above the current position, there's going to be
trouble. In this case, we need to execute the footer macro immediately.
The t rof f formatter keeps the current page position in the nl register, and the
page length in the register .p. A s a result, we can set the trap position based on a
conditional:
502 0 UNlX Text Processing 0

- ie(\\n(nl+lv)> (\\n(.p+\\n (Fb) .ch FO \\n (nlu+lvu

.el .ch FO -\\n(Fb
If the footnote won't fit, this code puts the trap one line below the current position; oth-
erwise, the footer trap location is moved up the page.
Now we'll have to redefine the footer macro to print the diverted footnotes, if
there are any:
.de FO \ " Redefine FO
.if \\n (fC\{ \
.evl \" Invoke first environment
.nf \" Good practice when outputting diversions
.FN \" Print diversion
.rm FN \" Remove printed diversion
.ev\ }
* bP \ " Now break page
..
Because the footnote macros are complicated, it might be a useful aside to look at
the process of debugging these macros. We used several .t m requests to report (to
standard error) on the sequence of events during a formatting run of a file that included
footnotes. What we wanted to know was the location of the footer trap and when it was
sprung. Inside the .F E macro, we inserted .t m requests to show which of the condi-
tional .c h requests were executed.
.ie (\\n(nl+lv)> (\\n(.p+\\n (Fb) \ { \
.tm ! ! ! ! ! ! FE: Change trap to current location (\\n(nl+lv)
.ch FO \\n(nlu+lvu \ )
.el \ I \
.tm ! ! ! ! ! ! FE: Move trap up the page (\\n(Fbu)
.ch FO -\\n(Fb \ )
Then, inside the .F O macro, we inserted messages to locate two positions on the page:
where the footer macro is invoked by the trap and where the footnotes have been out-
put.
.de FO
.tm ! ! ! ! FO: position is \\n(nl (\\n(.p+\\n(Fb) BEFORE

.tm ! ! ! ! FO: position is \\n(nl AFTER footnotes

* bP
..
To see these terminal messages without the formatted text, we invoke n r o f f
and redirect output to / d e v / n u l l . (tmacpack is a small macro package used for
testing these macros.)
 0 What’s in a Macro Package? a 503

$ nroff tmacpack textfile > /dev/null

! ! ! ! ! ! F E : M o v e t r a p u p t h e page (-36Ou)
! ! ! ! ! ! F E : M o v e t r a p u p t h e page (-44Ou)
! ! ! ! ! ! F E : M o v e t r a p u p t h e page (-52Ou)
! ! ! ! ! ! F E : M o v e t r a p u p t h e p a g e (-68Ou)
! ! ! ! F O : p o s i t i o n i s 1 9 8 0 (2640+-680) BEFORE
! ! ! ! FO: p o s i t i o n i s 2 4 2 0 AFTER f o o t n o t e s
! ! ! ! ! ! FE: M o v e t r a p u p t h e page (-36Ou)
! ! ! ! ! ! F E : Move t r a p u p t h e p a g e (-44Ou)
! ! ! ! ! ! FE: Move trap u p t h e page (-52011)
! ! ! ! ! ! FE: C h a n g e t r a p t o c u r r e n t l o c a t i o n ( 2 1 0 O + l v )
! ! ! ! FO: p o s i t i o n i s 2 1 4 0 (2640+-640) BEFORE
! ! ! ! FO: p o s i t i o n i s 2 5 8 0 AFTER f o o t n o t e s
! ! ! ! ! ! F E : M o v e t r a p u p t h e p a g e (-32Ou)
! ! ! ! FO: p o s i t i o n i s 2 3 2 0 (2640+-320) BEFORE
! ! ! ! F O : p o s i t i o n i s 2 4 0 0 AFTER f o o t n o t e s

Part of the reason for making this aside is the difficulty of writing effective footnote
macros. It requires a fair amount of testing to make sure they work in all cases. When
we spring the footer trap for the second time, the messages alert us to a problem-the
Catch-22 we mentioned earlier. The formatter encountered a footnote on the last input
line. The only way to fit both the footnote reference and the footnote on the same page
was to ignore the footer margin and let the footnote run into it.
Standard m s provides a better way of handling this overflow. In addition, the
Nroff/Troff User’s Manual describes a similar mechanism. Our simplified version, ade-
quate only for demonstration of this mechanism, will borrow from both of these
sources. (It might be said that a “working” version requires several empirically
discovered fudge factors or, as Joseph Ossanna called them, “uncertainty corrections”.)
The problem is how to split the footnote overflow if it extends beyond where we
want the bottom of the page to be. The solution is to put two trap-invoked macros at
the original (standard) page bottom location. The trap mechanism in t r o f f allows
only one macro to be executed by a trap at a given location. If you write:
.wh -\\n(FMu M1 \ “ P l a c e f i r s t macro
. w h -\\n(FMu M2 \“Overwrite f i r s t macro a t t h i s l o c a t i o n

all you will succeed in doing i s wiping out the first placement with the second.
However, you can move a trap location to an occupied position. The second trap
“hides” the first and renders it ineffective, but the first is still in place and is restored ifs
the second subsequently moves off the spot.
So here’s what we do in our trap initialization:
.wh 1 6 i F O \“ P u t r e g u l a r f o o t e r o u t o f t h e way
\“ (way o f f t h e page)
.wh -\\n(FMu FX \“ P l a c e f o o t n o t e o v e r f l o w macro
. c h FO -\\n(FMu \“ H i d e f o o t n o t e o v e r f l o w macro

The . F X (footnote o v e g o w ) macro will be invoked only if the F O trap is moved (as
it will be whenever there are footnotes on the page). In FX, all we do is start another
504 UNlX Text Processing 0

diversion, so that excess footnote text that would overflow at the bottom of the page is
saved for the next:
.de F X \ " F o o t n o t e overflow
.if \\n(fC .di eF \" Divert e x t r a f o o t n o t e
..
(We'll explain the reason for the test in a moment.)
Odd as it may seem, this diversion can be terminated from the footer macro .FO,
even though that macro is invoked before the footnote overflow macro! Because the
. F N diversion inside the . F O macros springs the footnote overflow trap and starts the
overflow diversion, we can close that diversion by a request in . F O following the
diversion.
The code in .FO now looks like this:
. n r dn 0 \ " R e s e t d i v e r s i o n s i z e register
.if \\n(fC \ t \ \" If there are footnotes
.ev 1
- nf
. FN
.rm FN
. .
i f ' \\n ( z'eF' .d i \ " E n d d i v e r s i o n o p e n e d by F X
. ev
.nr f C 0 \ } \ " Done w i t h f o o t n o t e s
' bP
There are several things here that need further explanation. The number register z
always contains the name of the last completed diversion. (Don't ask us how they
manage to put a string into a number register!) If our overflow diversion was this last
completed diversion, we terminate it:
. i f '\\n( .z'eF' .di

Then, we must take care of another eventuality. If we get this far without trigger-
ing the overflow trap-that is, if .F N did fit on the page-we want to disable the over-
flow macro, which we can do by zeroing our count register f C.
Now on the next page we have to handle any footnote overflow. We write a new
macro that invokes .FS and . F E to output the overflow diversion (.eF) into the
normal footnote diversion ( .FN).
.de F x \" P r o c e s s e x t r a f o o t n o t e
.F S
-n f \ " N o - f i l l mode
- eF \" O v e r f l o w d i v e r s i o n
.fi
FE
.rm eF
..
In the new page macro, we add a test to check if the last diversion amounted to any-
thing, and if it did, we invoke the .Fx macro.
 0 What's in a Macro Package? 0 505

\ " added t o .NP

.if \\n(dn .Fx

To test this new feature, we might add messages inside . F X , the macro invoked by a
hidden trap to open a diversion that collects any footnote overflow, and inside . F x ,
the macro that redirects the overflow back into the normal footnote diversion. You
should be able to accomplish this part on your own, as well as to venture into areas that
we did not cover (such as automatic numbering or marking of footnotes.) Before imple-
menting a footnote mechanism, we urge you to study the mechanisms in one of the
existing macro packages. However, following the chain of events from when a footnote
is encountered to when it is output in the footer macr-n the current page or on the
next-may seem like a t r o f f exercise equivalent to what Alfred Hitchcock called a
MacGuffin: a hopelessly complicated plot not meant to be figured out but that supplies
a reason for many entertaining scenes.

Multicolumn Processing
While we're still on the subject of page transition, we should look briefly at how multi-
column processing works.
Multiple columns are generally produced by using the mark and return
mechanism-. mk and .rt-and by manipulating the line length and page offset for
each successive column. The basic trick is to have the page bottom macro check if
multiple columns are in effect, and if so, whether or not the current column is the last
one.
A simple macro to initiate two-column processing might look like this*:
.de 2C
.mk \" Mark t o p position
.nr CL 0 1 /" Initialize column count flag
.ie \\$I .nr C w \\$I \" Test arg 1 for Column Width
.el .nr CW 2.75i \" or set default CW
.ie \\$2 .nr GW \\$2 \" Test arg 2 for Gutter Width
.el .nr GW .5i \" or set default GW
\" Save current one-column settings
.nr p 0 \\n(.o \" Save current page offset
.nr 1L \ \ n ( L L u \" Save original line length
.nr LL \\n(Cwu \" Set line length t o Column Width
- 1 1 \\n(LLu \" Set line length t o Column Width
.-

*Despite similar macro and number register names, this i s not the two-column macro used in m s . The ms
package provides a more general multiple column macro, - MC, of which 2c is a specialized call.
506 0 UNlX Text Processing 0

(We must save the default line length in a new register and redefine LL, or else a para-
graph macro, or any other macro that calls .RT, will interfere with two-column pro-
cessing.)
The page footer needs to include the following requests:
.de FO \" New footer macro
.ie \\n+(CL<2\{\ \" If incremental column count < 2
'Po+ (\\n(CWu+\\n(GWu) \" then increase page offset
' rt \ '' Return to mark
'ns \ I \" Enter no-space mode
.el \ t \ \ '' Otherwise
' P O \\n(pOu \ 'I Restore original page offset
'bp \ I \" Start a new page
..
Because two-column processing is likely to continue beyond a single page, we need to
modify the page top macro to mark the top of the page and initialize (set to zero) the
column count register. The two requests at the bottom of the definition have been
added:
.de NP \"New Page Macro
'sp \\n (HMu/2u
.PT
'sp I\\n(HMu
' ns
' mk \"Mark top of page
.if \\n(CL .nr CL 0 1 \"Reset autoincrementing column count
-.
After the C L register has been created by . 2 C , it can also be used as a flag that two-
column processing is in effect. The page top resets it to 0 to start the first column on a
new page.
The macro to return to single-column processing looks like this:
.de 1C
.rr CL \I' Remove column count register
.PO \\n(POu \ " Reset original page offset
.nr LL \\n (1Lu
- 1 1 \\n(LLu \ " and line length
- bP \ " Start a new page
. *

The column count register is removed, and the original page offset and line length are
restored. Unfortunately, using this mechanism, you cannot return to single-column
mode on the same page, without resorting to extensive use of diversions. If the first
column has already gone to the bottom of the page, there is no way for a prematurely
terminated second column to "go back" and fit the text into two even-sized columns on
the same page.
 0 What's in a Macro Package? 0 507

Page Top Resets

We've already discussed the use of a reset macro from within paragraphs to deal with
common errors. Page transitions are also a convenient place to put some different kinds
of resets. Like paragraphs, you can rely on their regular occurrence and can therefore
trap certain conditions.
In particular, you can use them when you want an effect to take place for only
one page and then stop. For example, in our business, we are often required to produce
not just complete manuals, but replacement pages to be inserted into an existing
manual. Sometimes the update page will be exactly the same size as the original, but
often it is longer, and requires additional space.
To avoid changing the numbering on subsequent pages, additional full or partial
pages are inserted with a special numbering scheme. For example, if a page is num-
bered 3-4 (section 3, page 4), and changes to that page run on to an additional page, the
new page will be numbered 3-4a.
In this situation, we need to temporarily change the way page numbers are han-
dled, then change back when the page is done. We've defined a macro called .UN,
which looks like this:
.de UN \ " Update page numbering macro
.nr Un 1 \ " Set flag t o test on page break
.nr % -1
.ie ! ''\\$lwT''
.as N N \\$l
.el .as NN a
_.
Our extended ms macro package normally puts the section number (sE) and the page
number (PN), separated by a hyphen, into the string NN. In this macro, we simply
append a letter to that smng. By default we add the letter a, but we give the user the
option to specify another letter as an argument to the macro, so pages can be numbered
3-4, 3-4a, 3-4b, and so on. To use the macro, the user simply enters it anywhere on the
update page. VoilB! The page number now has an a on the end.
Notice that the original page number register (%) was first decremented, so that
this new page will have the same number as the previous one. More to the point of this
discussion, notice that the macro sets the U n number register to 1 as a flag that update
numbering is in effect.
This flag is tested in the page top macro for the next page, and if it is set, the ori-
ginal page numbering scheme is restored as follows:
.if \\n(Un=l \ [ \
ds NN \\\\n (sE-\\\\n (PN
nr Un O \ }

(Note that four backslashes are required in the number register interpolations used in
defining NN because the string definition will be interpreted twice, once when the
macro is defined, and once when it is executed.)
508 UNlX Text Processing 0

Keep this trick in mind because there are many cases in which you can use the
page bottom or page top macro to reset conditions that you don't want to carry across
more than one page. We'll see another in just a moment.

Handling Widows and Orphans

Widows and orphans are the bane of any markup language-the one real advantage of
current wysiwyg systems. A widow is a single or partial line from the end of a para-
graph left over at the start of the next page. An orphan is a single line from the start of
a paragraph left alone at the bottom of a page. Both of these are considered poor page
layout.
As we've discussed, a macro package can take care of orphans simply by includ-
ing an .n e request in the paragraph macro. Widows are much harder to take care of,
because you don't know where the end of the paragraph will fall until you reach it.
In n ro f f and t ro f f, the only way you can handle this problem i s to process
each paragraph in a diversion, find out how long it was, then go back and break it up if
necessary. This greatly increases processing time, and is probably not worth the effort.
You could limit the extra work by testing the position on the page and only
diverting paragraphs that occur within range of the page bottom. However, even so,
this is a difficult problem you may not want to attempt.
It may be satisfactory to give users an increased capability for dealing with
widows when they do occur. Normally, the solution is to print out the document, find
any offending widow lines, then go back and manually break the pages a line earlier.
However, sometimes it is inconvenient to break the paragraph earlier-it would be
better to add the line to the bottom of the current page.
In standard m s , the location of the footer trap is reset to -\n (FMu in the .N P
macro at the top of every page. The user can get extra length on a page just by chang-
ing the value of F M on the preceding page.
We could also write a macro that would let the user make the change on the
offending page. For example, in m s :
.de EL \ " Extra Line macro
.nr e L 1 \ " Set flag
.ch F O -(\\n(FMu-1v)u \ " Put trap one line lower

All tbe user has to do is to introduce this macro anywhere on the page to be affected. It
is your job as macro developer to reset the normal page length-and the most likely
place is in the page top macro for the next page:
.if \\n(eL=l \ { \
.ch FO -\\n(FMu \ " Reset t o normal location for m s
.nr eL O \ } \*' Clear flag
 C H A P T E R

’ 17

An Extended ms Macro Package

In the previous chapter, we’ve looked at some of the essential elements of a macro
package-the innards that make it tick. However, few people will write a macro pack-
age just because they think they can do a better job at the basics than m s or mm.
More often, users who need specific formatting effects will build a macro set to achieve
those effects.
The macros used to produce this book are a good example of a custom macro
package. They were developed to create a distinctive and consistent style for a series of
books on UNIX by different authors. Although this macro package must of course do
all of the basics we’ve talked about, many of its macros provide solutions to more
specific problems. For example, there are macros for showing side-by-side before and
after screens for v i and macros for inserting italicized commentary in examples.
To illustrate more concretely the issues that force you to create or significantly
extend a macro package, this chapter will look at a set of extended m s macros for
typesetting technical manuals. Extensions built into this package fall into two major
categories:

Extensions that make it easier to control the appearance of a document, particu-

larly the page size (described in the last chapter) and the style of section head-
ings, tables, and figures.
. Extensions that address needs of books, manuals, and other documents larger
than the technical papers that m s and mm were originally designed for. These
extensions include improved methods for handling tables of contents and
indexes.

One of the chief weaknesses of the m s and mm packages is that they were
designed for smaller documents. For example, m s does not provide table of contents
generation, and the approach used by mm is suitable only for short documents. Neither
package supports automatic index generation. In this chapter and the next, we will also
look at ways to redress these problems.

9 509
510 0 UNlX Text Processing 0

Creating a Custom Macro Package

In this chapter, we will present an extended macro package designed for technical docu-
mentation. Based on the m s macro package, these extensions were originally
developed by Steve Talbott of Masscomp; they have been extended and altered during
several years of use in our technical writing and consulting business. Because we
needed to produce technical manuals for a number of different clients, we needed a
macro package that allowed us the flexibility to achieve a variety of document formats.
An important step in implementing this package was to establish the relation of
new and redefined macros to the original m s package. We wanted to read in the stan-
dard tmac .s package, and then simply overwrite or remove unwanted macros. Then
we organized our extensions into three groups: redefinitions of standard ms macros,
common macros we added to provide specific features or capabilities for all documents,
and format macros that were most often used to control the appearance or structure of a
document.
The format macros can be modified for the specifications of a unique document
format. Each format design has its own file, and the user only needs to specify which
of these formats are to be read in during the formatting run.
Following is a summary of the steps we followed to implement our m S macro
package. While describing this implementation, we don’t pretend that it is unique or
right for all uses; we do hope that it suggests ways to set up your own custom package.

1. Create a new directory to store the macro files.

2. Make a working copy of tmac .s and any subordinate files it reads in,
moving them to a new directory.
3. Create the tmac .Sredef s file to contain definitions of standard m s
macros that we’ve redefined, such as .IZ.
4. Create the tmac. Scommon file to contain utility and feature macros
available in all formats. The list macros described in this chapter are kept
here.
5. Create separate files containing definitions for unique document formats.
6. Set up tmac .S to control which files are read in and to handle certain
parameters that might be set from the command line.
7. Put tmac.S in /usr/lib/tmac, either by placing the file in that
directory or by creating a tmac .S file that sources the tmac .S file in
the macro directory.

The master file of this package is tmac . S, although it does not contain any macro
definitions. It allows users to set some parameters from the command line, and then it
reads in the standard m s macro package and the two files that contain redefinitions and
common macros. Last, it checks the value of a number register (v) to determine which
group of format macros are to be read in.
 1
0 An Extended m s Macro Package 0 51 1

Here's what our t m a c .S file looks like:

- \ I ' tmac.S - the main format macro package

.s o /work/macros/tmac.s \ " Read in standard ms

.s o /work/macros/tmac.Sredefs \ " Redefinitions of macros
- so /work/macros/tmac.Scommon \ " Common utility macros
\ " Check register v for version
\" and read in special format macros
.ie \nv \ I \
.if \nv=9 . s o /work/macros/tmac.Stest
.if \nv=8 . s o /work/macros/tmac.Squickref
-if \nv=7 . s o /work/macros/tmac.Slarge
-if \nv=6 . s o /work/macros/overheads
-if \nv=5 . s o /work/macros/tmac.Straining
.if \nv=4 . s o /work/macros/tmac.Sprime
.if \nv=3 . s o /work/macros/tmac.Scogx
-if \nv=2 . s o /work/macros/tmac.Smanuals
-if \nv=l . s o /work/macros/tmac.Snutshell\)
-el . s o /work/macros/tmac.Sstandard

The -r option to n r o f f and t rof f is used to select a particular version of the for-
mat macros. For instance, the first set of format macros is designed for producing our
Nutshell Handbooks. To format a document using the macros defined in
t m a c . Snutshell, a user would enter:
$ ditroff -Tps -mS -rvl chOl I devps I lp

One of the files, t m a c .S t e s t , is available for use during the development and test-
ing of new versions of the macros. We'll look at some of the different formats later in
this chapter.
A few other details about this implementation may help you customize a package.
Both m s and mm include a number of Bell-specific macros that are not very useful for
users outside of AT&T. For example, it is unlikely that you will require the various
styles of technical memoranda used internally at Bell Labs. Unused macro definitions
need not get in your way, but they do use up possible names and number registers that
may conflict with what you want to do. The . r n macro allows you to rename a
macro; .r m will remove the definition of a macro.
You may want to remove selected macros. For example, you might want to start
the modifications to a macro package built on m s with the following request:
.rm TM I M MF MR EG OK RP TR S2 S 3 SG IE [ I I [ [ . .I [ o \
[c [ 5 [ 4 [ 3 r 2 [ O [< I < [> I > [ - 1-

(Note the use of the backslash to make this apparent two-line request into a single long
line.)
There is a slight performance loss in reading in a large macro package, and then
removing a number of the macros. For efficiency, you'd be better off removing the
undesirable macros from your copy of the m s source file.
51 2 UNlX Text Processing 0

Reading in tmac .Sredefs after t m a c - s overwrites some of the standard

m s macros with our own definitions. The standard versions are thus not available. If
you want to retain a standard macro definition, you can make it available under a dif-
ferent name. Use the .r n request to rename the standard macro before overwriting its
definition.
As discussed in the previous chapter, we redefined the .IZ macro to allow the
setting of various page sizes. Because the standard .IZ macro is invoked from
t m a c - s at the start of the formatting run, we can’t simply overwrite its definition. We
must either delete the standard .IZ macro definition or comment out its invocation.
Then the new 1 2 macro in t m a c . Sredef s will be executed.
As you develop your own set of extensions, you will undoubtedly consider addi-
tional modifications. Appendix F lists the set of extended macros that we use. You
may not need many of the specialized macros provided in this package. But it will
show you how to build on an existing package and how easy it is to modify the appear-
ance of a document.

Structured Technical Documents

The m s and mm packages provide a number of macros to produce title pages,
abstracts, and so on for technical memoranda. Subsections can be numbered or unnum-
bered.
Anyone who has used the UNrX Programmers’ Manual is familiar with the output
of these packages. The technical papers collected in that volume bear superficial resem-
blance to the chapters of a book. However, they lack continuity-section, figure, and
table numbers, where present, are relative only to the current section, not to the entire
volume.
A macro package designed for producing technical books or manuals may need at
least some modification to produce section headings. Chapter and section headings
should make the structure of a document visible. In a nontechnical book, chapters are
often the only major structural element. They divide the book into major topics, and
give readers stopping points to digest what they have read.
Chapters are usually distinguished from a formatting point of view by a page
break and some kind of nonstandard typesetting. For example, a chapter number and
title may be set in large type, and the text may begin lower on the page.
In technical books and manuals, which are often not read straight through as much
as they are used for reference, frequent section headings within a chapter give the reader
guideposts. There are often several levels of heading-more or less depending on
whether the book is intended primarily for reading or for reference. This book uses
three levels of headings within a chapter, one for major changes in topic, the others for
less significant changes.
Section headings can be distinguished merely by type font and size changes, as in
this book, or by section numbering as well. Properly used, section numbers can be very
helpful in a technical manual. They allow detailed cross references to different parts of
the book without using page numbers. Referencing by page numbers can result in errors
because page numbers are not fixed until the book is done.
 0 An Extended ms Macro Package 0 513

Detailed breakdown of a chapter into subsections can also help the writer of a
technical manual. Because a manual (unlike an essay or other free-form work of non-
fiction) has definite material that must be covered, it can be written successfully from
an outline. It i s often possible to write technical material by entering the outline in the
form of section and subsection headings and then filling in the details.
In this approach, numbered sections also have a place because they make the out-
line structure of the document more visible. In reviewing technical manuals, we can
often identify many weaknesses simply by looking at the table of contents. Sections in
a technical manual should be hierarchical, and the table of contents should look effec-
tive as afi outline. For example, a chapter in our hypothetical Alcuin User’s Guide
might look like this:
Chapter Two: Getting Started with Alcuin

2.1 Objectives o f this Session

2.2 Starting Up the System

2.2.1 Power-up Procedure
2.2.2 Software Initialization

2.3 Creating Simple Glyphs

2.3.1 Opening Font Files
2.3.2 Using the Bit Pad
2.3.2.1 The Cell Coordinate System
2.3.2.2 Pointing and Clicking

How much easier it is to see the structure than in a case where the proper hierarchica1
arrangement of topics has not been observed. How often have you seen a “flat” table
of contents like this:
Chapter Two: Using Alcuin

2.0 Starting Up the System

2.1 Power-up Procedure
2.2 Software Initialization
2.3 Creating Simple Glyphs
2.4 Opening Font Files
2.5 Using the Bit Pad
2.6 The Cell Coordinate System
2.7 Pointing and Clicking
514 UNlX Text Processing 0

Even when numbered section headings are not appropriate, they can be a useful tool for
a writer during the draft stage, because they indicate where the organization has not
been properly thought through. For example, we often see manuals that start with a
general topic and then describe details, without a transitional overview.
A macro package should allow the writer to switch between numbered and
unnumbered headings easily. Both mm and m s do provide this capability, and we
want to include it in our macros. However, we also want to include more flexibility
than either of these packages to define the format of headings.
Because headings are the signposts to the book’s structure, changing their appear-
ance can make a big difference in how the book is read. Different levels of headings
need to stand out from the text to a greater or lesser degree, so that readers can easily
scan the text and find the topic that they want.
The mechanisms for emphasis (in t r o f f ) are font and size changes, and the
amount of space before and after a heading. Underlining and capitalization can also be
used (especially in n r o f f but also in t r o f f) for alternate or additional emphasis.
In our package, we include five levels of heading: a chapter-level heading and
four levels of numbered or unnumbered subsection headings.
As described in the previous section, our custom macro package incorporates
several different versions of the basic macros required to produce technical documents.
In each version, the name of the heading macro is the same, but its definition is modi-
fied slightly to produce a different appearance. These different versions help us con-
form to the document styles used by our clients. Whenever we have a client who needs
a new format, we customize the macro definitions, rather than add new macros.
The beauty of this approach is that the input macros the user needs to enter in a
document are identical, or nearly so. Thus, we don’t increase the number of new mac-
ros that our users must learn, and it eliminates the recoding of existing documents to
achieve a new format.
This approach is also useful when you support different types of output devices.
Originally, our designs were developed for the HP LaserJet printer, which supports a
limited set of fonts and sizes. When we purchased an Apple Laserwriter and Lino-
tronic L 100 typesetter, our formatting options increased, making available multiple
fonts and variable point sizes. In an environment supporting multiple types of printers,
you might want to adapt formats for specific printers.

The Chapter Heading

The chapter heading is in a class by itself, because it requires more emphasis than sub-
section headings, and because the macro that produces it may need to initialize or reset
certain registers used within the chapter (such as section, figure, or table numbers).
In an arbitrary reversal of terminology, we call our chapter macro .Se (section).
It could just as well be called .CH for chapter, but we use .C h for a subsection head-
ing (as we’ll see in a moment) and want to avoid confusion. In addition, this macro can
be used for appendices as well as chapters, so the more general name seems appropriate.
The chapter heading has three major parts:
 An Extended m s Macro Package 0 515

chapter-specific register initialization, including registers for section number-

ing, table and figure numbering, and page numbering
appearance of the actual chapter break
table of contents processing

Because this is a long macro definition, let's look at it in sections.

.de Se \ " section; $ 1 = number: $ 2 = name:
\ " $3 = type (Chapter, Appendix, etc)
\"
\ " 1. Number Register Initialization
\"
.ie !"\\$l"" \ { \ " Test for sect number
nr sE \\$I \ " Assign to register SE
if !\\n(sE \ { . \ " Test if not a numeric
.af sE A \ " Handle appendices
if "\\$l"A" .nr sE 1
if "\\$l"B" .nr sE 2
if "\\$l"C" .nr sE 3
if "\\$l"D'' .nr sE 4
if "\\$l''E" .nr sE 5
if " \ \ $ l " F " .nr sE 6
if "\\$l"G" .nr sE 7
if "\\$l"H" .nr sE 8
if "\\$l"I'* .nr sE 9
if "\\$l"J" .nr s E l o \ } \ }
\ " Only go as far as J
.el \ ( \
nr s E 0
tm Preface or if Appendix past letter J:
tm Set number register s E to position
tm of that letter in the alphabet
tm and alter register format:
tm For Appendix K, enter:
tm .Se K "Title"
tm .nr sE 11
tm .af sE A
.\I
.if \\n%>l .bp \ " Check if consecutive sections
\ " in same file and break page
.nr $ 1 \ " N o w reset page number
.nr PN 1
.a f PN 1
.ie !"\\$l"*' \I. \ " Test for sect number
\ " to set page number type
516 0 UNlX Text Processing

ds NN \\\\n (sE-\\\\n (PN

ds H1 \ \ n ( s E \ " Set f o r subsection numbering
\I
.el \ {
ds NN \\\\n(PN
n r sE O \ }
.ds RF \\\\*(NN \ " Assign page number t o footer
.nr fG 0 \ " Initialize figure counter
.nr t B 0 \ " Initialize table counter
The macro first initializes a number of registers. Chapters are usually numbered on the
first page, along with the title. If subsections are to be numbered, the chapter number is
the root number for all headings. We need to take this number as an argument, and
store it into a register for later use.
Because appendices are usually lettered rather than numbered, we also need to
consider the special case of appendices. (This could be done with a separate macro;
however, this package uses a single multipurpose macro.) The code for this is quite
cumbersome, but works nonetheless: if the first argument to the macro is non-numeric,
it is tested to see if it is one of the first ten letters in the alphabet. If so, a number is
stored into the register, but the output format is changed to alphabetic.
If the argument is not a letter between A and J, a message is printed. This mes-
sage is more verbose than you would generally want to use, but it is included to make
the point that you can include detailed messages.
The macro next sets up the special page numbering scheme used in many com-
puter manuals-the chapter number is followed by a hyphen and the page number (e.g.,
1-1). This numbering scheme makes it easier to make last minute changes without
renumbering and reprinting the entire book.
Finally, the macro initializes counters for automatically numbering figures and
tables. We'll see how these are used in a few pages.
The next portion of the macro is the part that is most variable-it controls the
actual appearance of the chapter heading. This is the part of the macro that has led us
to develop several different versions.
In designing chapter headings, let your imagination be your guide. Look at books
whose design you like, and work from there. Three different designs we used on the
HP LaserJet are shown in Figure 17-1. (These designs are a compromise between
aesthetics and the capabilities of the output device.) This book is another model.
The macro for the first heading in Figure 17-1 is used as follows:
.Se 2 "Getting Started with Alcuin"

or :
- Se A "Summary of Alcuin Drawing Primitives" "Appendix"

The heading starts on a new page. If a third argument is not present, it is assumed that
the section type is Chapter, and the section is labeled accordingly. An alternate section
type can be specified in the optional third argument. This argument is usually Appendix
but can be any string the user wants printed before the section number.
 ~

0 An Extended ms Macro Package 0 517

CHAPTER 2
GETTING STARTED WITH ALCUIN

2
Getting Started with Alcuin

Chapter 2
Getting Started with Alcuin

Fig. Z7-Z. Some Different Styles of Chapter Heading

518 0 UNlX Text Processing 0

The portion of the macro definition that creates the first heading in Figure 17-1
follows:
.\" P a r t 2 of Se Macro: Output c h a p t e r heading
.RT
.in 0
.lg 0 \ " D i s a b l e l i g a t u r e before . t r
\" Translate t i t l e t o uppercase
. t r aAbBcCdDeEfFgGhHiIjJkKlLmMnNoOpPqQrRsStTuUvVwWxXyYz2
- SP
- na
\ " T e s t for s e c t i o n t y p e a r g u m e n t
. i e !"\\$3"" .ds c H \ \ $ 3
. e l .ds c H C h a p t e r \" D e f a u l t i s chapter
\ " I f s e c t i o n number s u p p l i e d
\ " o u t p u t s e c t i o n number a n d t y p e
\" i n 1 4 p t . bold.
. i f !"\\$1"" \ ( \
\s14\f3\\*(cH \ \ $ l \ f l \ s O
\I
\ " If n o s e c t i o n number b u t
\" t h e r e i s a t y p e (i.e., P r e f a c e )
\" then output s e c t i o n type
. i f "\\$l"" . i f !"\\$3"" \{\
\ s 1 4 \ f 3 \ \ * ( c H \ f 1\ S O
\I
-SP 5 P
\" Test for s e c t i o n t i t l e
\ " P r i n t it i n 1 4 p t . b o l d
.if ! "\\$2"" \{\
\s14\f3\\$2\fl\sO
\}
- S P 6P
.ad b
. H1 \ " Draw l i n e
\ " R e t r a n s l a t e arguments
. t r aabbccddeeffgghhiijjkkllmmnnooppqqrrssttuuvvwwxxyyzz
.sp 3
.n s \ " E n a b l e no-space mode

There are a couple of points you may want to note about this code:

The actual section title, as specified in the second argument, is forced to all
uppercase using the .t r request.
The horizontal line under the title is drawn using a utility macro called .H1
(horizontal line), which simply draws a line the width of the page, less any
indent that is in effect:
 0 An Extended m s Macro Package 0 519

.de H1 \ " Horizontal line. $1 = underline char

- br
\1' \\n ( - lu-\\n ( .iu\&\\$l'
- br
..
No-space mode is turned on at the end of the macro, to inhibit inconsistent
spacing caused by users placing spacing requests or paragraph macros after the
.S e macro. All of the heading macros use this technique because inconsistent
spacing around headings will give the page an uneven look.

An alternate definition for this section of the macro follows. This code produces
the second heading shown in Figure 17-1.
. \ " Part 2 of S e Macro (Alternate):
.ad r \ " Right justified
. fl
.rs
.sp .75i \ " Move down from top
\ " Section number in 24 pt. bold
.if !"\\$I"" \ { \
\s24\f3\\$l\fl\sO\}
.sp 12p
\ " Section title in 20 pt. bold
-if ! "\\$2"" \s20\f3\\$2\fP\s10
-sp 12p
\ " Optional 2nd line of title
.if ! "\\$3"" \s20\f3\\$3\fP\s10
-sp 3
.ad b
.ns
This version i s much simpler; it doesn't print the section type at all, just the number or
letter. However, because it prints a right-justified title, we have given the user the
option of splitting a long title into two parts.
The final part of the macro (in either version) adds the section title to the table of
contents. As was the case with .H1, this is done by an internal utility routine that is
defined elsewhere. We'll discuss how this works later.
\ " Last Part of Se Macro
\ " Now do toc
.tC \\$I \\$2 \\$3
..
520 0 UNlX Text Processing 0

A Mechanism for Numbered Headings

Before we describe the lower-level headings used within a chapter, we need to explore
how to generate automatically numbered sections. We have defined a version of the
m s .NH macro that is called internally by our own heading macros. It has the same
name and uses the same internal registers as the m s macro, but the font and spacing
requests specified in the m s .NH macro are removed. All that this macro now does is
generate the section number string.
.de NH \ " r e d e f i n e from -MS
. n r NS \\$l I'\ S e t NS t o a r g 1
. i f !\\n(.$ . n r NS 1 \" S e t NS t o 1 if n o a r g
. i f !\\n(NS . n r NS 1 \" o r NS i s n u l l o r n e g a t i v e
. n r H\\n(NS +1 \" Increment Heading level register
\" T e s t which level i s i n effect
.if !\\n(NS-4 . n r H5 0 \ " t h e n reset l o w e r l e v e l s t o 0
.if !\\n(NS-3 . n r H4 0
.if !\\n(NS-2 . n r H3 0
.if !\\n(NS-1 . n r H2 0
\ " P u t t o g e t h e r s e c t i o n number
.if ! \ \ $ l .if \\n(.$ .nr H1 1 \ " S e t f i r s t level
.ds SN \ \ n ( H l \ " B e g i n b u i l d i n g SN
. i e \\n(NS-1 . a s SN .\\n(H2 \n == 1.1 2 n d l e v e l
. e l . a s SN . . \ 'I or == 1.
. i f \\n(NS-2 .as SN .\\n(H3 \-1 == 1.1.1 3rd
. i f \\n(NS-3 .as SN .\\n(H4 \ " == 1.1.1.1 4th
. i f \\n(NS-4 .as SN .\\n(H5 \ - == 1.1.1.1.1 5 t h
'ti \ \ n ( . i u
\ \ * (SN \ " O u t p u t SN s t r i n g
..
This macro repays study, because it shows several clever ways to use number registers.
First, the argument to the macro i s placed into a number register. This register is then
used to select which of a series of further registers will be incremented:
. n r NS \\$l

.nr H\\n(lJS +1

If the macro is called as .N H 1, register H1 will be incremented; if the call is .NH

2, register H 2 will be incremented, and so on. Then, depending on the value of that
same NS register, the appropriate register value will be appended to the section number
string S N .
 0 An Extended m s Macro Package 0 52 1

Subsection Headings
In our package, we allow four levels of subsection headings, created by macros called
.A h (A head) through - Dh (D head). The macros for all four levels have the same
essential structure; they differ only in the appearance of the printed text. Again, we
have different styles for different clients.
The distinction between levels of headings in one of those styles is as follows:

The A head prints the heading in 14-point bold type, all uppercase, with 2 6
points of space above the heading and 18 points below.
The B head prints the heading in 14-point bold type, mixed case, with 23
points of space above the heading and 15.5 points below.
The C head prints the heading in 12-point bold type, mixed case, with 18
points of space above the heading and 12 points below.
The D head prints the heading in 10-point bold type, mixed case, with 18
points of space above the heading and none below. The heading actually runs
into the text and is separated from it only by a period.

All levels of headings can be either numbered or unnumbered, depending on the state of
a number register called nH. If nH is 0, headings are unnumbered; if it is 1, they are
numbered.
Here i s one version of the .A h macro. From this example, you should be able to
build the lower-level headings as well.
.de Ah \ " A-heading ; $1 = title
-sp 26p
.RT
.ne 8 \ " Need room on page
.ps 14 \'' 1 4 pt. on 16 pt, heading
.vs 16
-1g 0
.tr aAbBcCdDeEfFgGhHiIjJkKlLmMnNoOpPqQrRsStTuUvVwWxXyYzZ
.bd I 4 \ " Embolden italic font (optional)
\f3\c \ " Bold font; concatenate next input
.if \\n (nH \ { \ " i f producing numbered heads
ie \ \ n ( s E .NH 2 \ " If chapter (Se macro) is
\ " numbered, then 2nd level
el .NH l \ ) \ " I f not, 1st level head
\&\\Sl\fl \ " Output title
.LP 0 \ " Paragraph reset; (0 = no space)
\" RT resets default point size
.bd I \ " Turn o f f emboldening
.tr aabbccddeeffgghhiijjkkllmmnnooppqqrrssttuuvvwwxxyyzz
-
1g
.sp 18p
.ns
522 0 UNlX Text Processing 0

.tC \\*(SN \\$l A h \ " Output TOC info

__
Some pointers: First, whenever you force capitalization with .t r, be sure to turn off
ligatures, because they do not capitalize. Second, when you boldface a user-supplied
string, it is wise to artificially embolden italics as well, in case the user embeds an italic
font switch in the heading. Third, don't forget to enter no-space mode to ensure con-
sistent spacing following the heading.
As you can see, the - N H macro is called to generate a section heading only if the
nH register has been set. In addition, the macro checks to make sure that a major sec-
tion number has been specified by the .S e macro. As you may recall, .S e sets the
first number in the numbered heading string (Hl). If .S e has been called, the subsec-
tion headings start at level 2, otherwise they start from the top.
To make it very easy for even novice users to specify whether they want num-
bered or unnumbered headings, the package includes a macro called .Nh (numbered
headings) that turns numbering on or o f f
.de Nh \ " Numbered headings; $1 = turn o n (1) or off ( 0 )
\ " $ 1 = 2 will cause only A heads t o be numbered
.nr nH \\$l
-.
This is a matter of macro package style, as mentioned earlier. Steve Talbott's style,
when he initially developed this package, was to code everything as macros, even where
the macro simply sets a number register or defines a string. This makes the package
very easy to learn, because you can give a new user a concise, unambiguous list of mac-
ros to enter into a file.
Other examples of this style include the .Ti and .St (title and subtitle) mac-
ros, described in Appendix F, which simply define the m s RF and LF strings for run-
ning footers. Because of the mnemonically named macros, new users don't have to
remember whether the title goes in the right footer or the left, and so on. They simply
enter the title of the book and chapter as arguments to the respective macros. The
disadvantage is that users are insulated from an understanding of what is really going
on, which may be an obstacle to learning more advanced skills.

An Alternate Definition
To give you an idea of how easy it is to change the look of a document by redefining a
few macros, let's look at how we could redefine the heading for this section. One
popular layout style in technical manuals uses a very wide left margin in which only the
headings are printed, as follows.

An Alternate Definition To give you an idea of how

easy it is to change the
look of a document ...
 0 An Extended m s Macro Package 0 523

Here's the modified macro to produce this heading:

.de Ah \ rr A-heading; alternate version
\" Requires resetting default page
\" (PO) to allow for extra offset.
\ 'I .nr PO 2.5i for 1.5 extra offset
.nr Po 1.5i \" Set amount of extra offset
. n r Gw .2i \" Set width of gutter
- mk \" Mark vertical position
-PO -1.5i \ '' Set new page offset
- 1 1 \\n (Pou-\\nGwu
.ps 12 \ " Set 12 pt. on 14 pt.
.vs 14
\&\f3\\$l\fl \" Output header in bold
- rt \" Return to vertical position
.PO \\n(~Ou \" Reset default page offset
.LP 0 \" Reset point size and line length
.ns
.tC \\*(SN \\$l Ah \ " Output TOC info
..

Figure and Table Headings

In technical manuals, it is common to number and title all figures and tables, both for
easy reference from within the text, and for collection into lists of figures and tables
that will appear in the table of contents.
These macros are easy to construct and, apart from whatever appearance you
decide to give them, nearly identical in content. There i s a "start" macro and an
"end" macro:
.de Fs \ " Start figure; $1= reserved space;
\ '' $2= F, floating figure
.RT
.if "\\$2"F" \ I . \ " Figure can float
nr kF 1
KF\ 1
.if \\$l \ t - \ " Specify amount of space
ne \\$I \I' required for paste-up
fl
rs
sp \\$l\)
__
.de Fe \ " Figure end; $1 = title
-SP
.bd I 3
.nr fG +1 \ " Increment Figure counter
524 0 UNlX Text Processing 0

\'Ithen determine format

.ie \\n(Se .ds fG \\*(Hl-\\n(fG
.el .ds fG \\n(fG
. ce \'* Output centered figure
\f3Figure \\*(fG. \\$l\fl
. t C I.\\* ( f G " "\\$1" "Figure"
.bd I
- SP
.if \\n(kF=l .KE \ " End keep if in effect
. t C " \ \ * (fG" "\\$1" "Figure" \ " Output TOC info
..
A s you can see, the .F s (figure start) macro allows the user to reserve space for a fig-
ure to be pasted in, and for it to float to a new page, using the m s "floating keep"
mechanism.
Neither of these options are necessary. The macro can simply bracket a figure
created with pic, for example, in which case all that the macro provides is a con-
sistent amount of space before the figure starts.
The .F e vzgure end) macro does most of the work. If a keep is in effect, .F e
terminates it. In addition, it prints the figure caption below the figure and adds a con-
sistent amount of space below the caption. The figure is automatically numbered with
the section number, and a figure number that is incremented each time the macro is
called. As you may remember, this figure number register, f G, was initialized to 0 in
.Se.
To give the user some options with figure numbering, a second argument allows
the user to turn it off entirely. In addition, if the section is unnumbered, the section
number and hyphen will be omitted. To accomplish this involves a little juggling of
strings and number registers (which is something you should plan to get used to when
you write macros). Notice that we use the string H1 for the section number rather than
the section number register itself (sE),because we went to some trouble in the .S e
macro to handle lettered appendices as well as numbered chapters.
You could easily add optional appearance features to this macro. For example, in
one implementation, we draw a horizontal line above and below the figure, and print the
caption left justified and in italics below the bottom line.
The figure end macro also calls the table of contents macro, which will be
described later.
The macros for labeling tables are very simple, because the standard .T S and
.TE macros do everything necessary except providing consistent pre- and post-spacing
and printing the caption. In this case, the caption i s at the top:
.de Ts \ " Table start; $1 = title
.nr t B +1 \ " Increment Table counter
\" Determine format
.ie \\n(Se .ds t B \\*(Hl-\\n(tB \ " Section Table
.el .ds t B \\n(tB
- SP
.ce 2 \ " Output label and
 0 An Extended ms Macro Package 0 525

\f3Table \\*(tB. \ " title on 2 lines

\&\\Sl\fl
.tC "\\*(tB" "\\$1" "Table" \ " Output TOC info
.bd I
* LP \ " Paragraph reset
..
.de Te \ " Table end -- no arguments
.RT \ '' Reset
- SP
..

Lists, Lists, and More Lists

One of the significant features lacking in the m s macros is the ability to generate
automatically numbered or lettered lists. You can use the .IP macro and number or
letter a list yourself-but what good is a computer if it can't handle a task like this?
One of the nicest features of Steve Talbott's extended m s package is its set of
comprehensive, general-purpose list generation macros. There are three macros: .L s
(list start), .Li (list item), and .L e (list end). Unlike mm, in which different types
of lists must be specified using different macros, here you request a different type of list
by giving an argument to the .Ls macro. You can request any of the types of lists in
Table 17-1.

TABLE 17-1. List Types

Argument List Type

A Alphabetic with uppercase letters
a Alphabetic with lowercase letters
B Bulleted with by default
N Numbered with arabic numerals
R Numbered with uppercase roman numerals
K Numbered with lowercase roman numerals

The bulleted list uses the bullet character ( 0 ) by default. However, as you will
see, the macro allows you to specify an alternate bullet using an optional third argu-
ment. This "bullet" could be a dash, a box ( \ (sq), a checkmark ( \ (sr), or any
other character.
Lists can be nested, and there is a default list type for each level of nesting, so the
type argument does not really need to be specified.
Here's the list start macro:
526 0 UNlX Text Processing 0

.nr 10 0 1 \ " Initialize nested list level counter

.de Ls
-\" list start; $1 = A ( L P H A ) , a(alpha), B(ullet), N(umeric),
.\ " R(oman), oman an); $2 = indent
.\ " $3 = alternate bullet character
- br
.if !"\\$1"A*' .if !"\\$1"B" .if !"\\$l"N" .if ! " \ \ $ l " R " \
if !"\\$l"r" .if ! " \ \ $ l " a " .if !"\\$l"" \
tm L s : Need A a B N R r or null as list type
.nr l\\n+(10 0 1
.ie "\\.$l"" \I\ \"Set defaults
if "\\n(10"1" .af l\\n(10 1 \"Numeric at 1st level
if "\\n(10"2" .af l\\n(10 a \"1c alpha at 2nd level
if "\\n(10"3" .af l\\n(10 i \"1c roman at 3rd level
if "\\n ( l O " 4 " .ds l\\n (10 \ (bu\"Bullet at 4th level
if "\\n (lO"5" .ds l\\n (10 \f3\-\fl\"Dash at 5th level
if \\n(10-5 .ds l\\n(lO\(bu \"Bullet above 5th level
if \\n(10-3 .nr l\\n(10 0-1 \ }
.el \ I \
if " \ \ $ l " A " . a f l\\n(10 A
if "\\$l"a" . a f l\\n(10 a
if "\\$l"B"\{\
if "\\$3'"' .ds l\\n (10 \ (bu
if ! "\\$3'"' .ds l\\n(10 \\$3
nr l\\n(10 0-1\}
if " \ \ $ l " R " .af l\\n(10 I
if " \ \ $ l " r " . a f l\\n(10 i \ }
.ie ! "\\$2"" .nr i\\n(lO \ \ $ 2 \ " List indent
.el .nr i\\n(lO 5 \ " Default indent
.RS
..
When you first look at this macro, you may be a little overwhelmed by the complex
number register names. In fact, there is not much to it.
One number register, 1 0 , is used as a counter for nested lists. As you can see,
this register is initialized to 0 outside of the list macro definition itself. Then, when the
.Ls macro is called, this register is autoincremented at the same time as it is used to
define the name of another number register:
.nr l\\n+(10 0 1
It is this second number register interpolation-1 \ \ n+ ( 10-that is actually used to
number the list. This is a technique we promised to show you back when we were first
describing number registers. We create a series of related number register names by
interpolating the value of another register as one character in the name.
Think this through for a moment. The first time .Ls is called, the request:
 0 An Extended m s Macro Package 0 527

.nr l\\n+(10 0 1

defines a number register that is actually called 1 1 (the letter 1 followed by the value
of number register 10-which is 1). A second call to .Ls without closing the first
list (which, as we shall see, bumps the counter back one) will define number register
1 2 , and so on.
In a similar way, another series of number registers ( i\ \n ( 10) allows a dif-
ferent indent to be specified for each nested level, if the user so desires.
With the exception of the bulleted list, all of the different list types are numbered
using the same number register ( I n , where n is the nesting depth). The different types
of lists are created simply by changing the output format of this register using the .a f
request.
Here's the .Li macro:
. d e Li \ " List item; $1 = 0 n o blank line b e f o r e item
.br
.if "\\$l"O" .ns
.ie "\\n (l\\n ( 1 0 1 1 - 1 " .IP I#\\* (l\\n (10" "\\n (i\\n (10''
.el \{\
.nr l\\n(10 +1
.IP 'I \ \n (l\\n (10.'I "\\n (i\\n ( 1 0 " \ )

The actual list counter itself (as opposed to the nesting counter) is incremented, and the
appropriate value printed.
The number and the associated text is positioned with the standard m s - I P
macro. If you don't have access to the m s macros, you could simulate the action of
the - IP macro as follows:
. d e IP
.nr Ip 1
. s p \\n(PDu
.in \\$2u
.ti -\\$2u
.ta \\$2u
\\Sl\t\c

r
-.
However, there is one drawback to using an IP-style macro as the basis of the list.

.I P macro puts its argument at the left margin, as was done with this
sentence.
Instead, we'd like something that puts the mark in the middle of the indent, as
was done with this sentence.
Here's the macro that produced the second example:
528 0 UNlX Text Processing 0

.de I P
. n r Ip 1
. sp \\n (PDu
. i n \\$2u
. n r i l \\$2/2u+\w'\\$l' \ " Amount t o move l e f t
. n r i 2 \\$2-\w' \\$I' \*I Amount t o move back
. t a \\n(i2u
.ti -\\n(ilu
\\$l\t\c
..
This version of the macro places the mark not just at a position half the depth of the
indent, but exactly in the middle of the indent by adjusting the indent by the width of
the mark argument. Number registers are used for clarity, to avoid nesting the various
constructs too deeply.
(Note that this simplified .IP macro lacks some of the functionality of the m s
.IP macro, which saves the current indent and therefore allows you to nest indents by
using the .R S and .RE macros.)
If you are using m s , and you want to create a macro that puts the mark in the
center of the indent, be sure to name this macro something other than .IP, so that you
don't conflict with the existing macro of that name.
Here's the list end:
.de L e \" L i s t end; $1=0 n o b l a n k l i n e f o l l o w i n g l a s t i t e m
-br
.rr l \ \ n ( 1 0
.rr i \ \ n ( l O
.rm l\\n(10
.nr 1 0 -1
.RE
.ie !\\n(10 \ { \
ie "\\$l"On .LP 0
e l .LP\}
.el . i f !"\\$1"0" .sp \\n(PDu
..
This macro removes the list numbering registers an^ strings, decrements the nested list
counter, and calls the m s .RE macro to "retreat" back to the left (if necessary
because of a nested loop). Finally, it leaves a blank line following the end of the list.
(As you might remember, PD is the m s register containing the paragraph distance-
0 . 3 by
~ default.)

Source Code and Other Examples

In a technical manual, there are often further issues brought out by the need to show
program source code or other material that loses essential formatting if it is set with
proportional rather than monospaced type.
 An Extended ms Macro Package 0 529

As previously discussed, the basic trick in d i t r o f f is to use the Cw font. If

you are using o t r o f f , you will need to use the c w preprocessor (see your UNIX
manual for details) or some other type of workaround. (When we were using
o t rof f, our print driver allowed font substitutions based on size. We told the driver
to use the printer's constant-width font whenever t r o f f used a point size of 1 1 .
Then, we wrote a macro that changed the point size to 1 1 , but used .cs to set the
character spacing to the actual size for the printer's constant-width font. This was not a
very elegant solution, but it worked-so if you are stuck with o t r o f f , don't despair.
Put your ingenuity to work and you should come up with something.)
Besides the change to the CW font, though, there are several other things we'd
like to see in a macro to handle printouts of examples. We'd like examples to be con-
sistently indented, set off by a consistent amount of pre- and post-line spacing, and set
in no-fill mode.
Here's an example of a pair of macros to handle this situation:
.de Ps\" P r i n t o u t s t a r t ; $1 = indent ( d e f a u l t is 5 s p a c e s )
.br
.sp \\n(PDu
-ns
.nr pS \\n(.s \" Save c u r r e n t point s i z e
.nr vS \\n(.v \" Save current vertical spacing
.nr pF \ \ n ( . f \'l Save current f o n t
.nr PI \\n ( i - \" Save current indent
*ps 8
.vs 10
.ft cw
.ie ! "\\$1"" . i n +\\$In
.el . i n +5n
.n f
* .

.de P e \" P r i n t o u t end; $1 n o n - n u l l , no concluding

- br
.if "\\$1"" .sp \ \ n ( P D u
.ps \ \ n ( p S u
.vs \\n(vSu
. f t \\n(pF
. i n \\n(pIu
. r r pS
.rr V S
- r r pF
.rr P I
.fi

The trick of saving the current environment in temporary registers is a useful one. The
alternative is to use a separate environment for the printouts, but this assumes that the
available environments are not already in use for some other purpose. You could also
530 0 UNlX Text Processing 0

call a reset macro to restore the default state-but this may not actually be the state that
was in effect at the time.
In addition, you shouldn't rely on t r o f f's ability to return to the previous set-
ting by making a request like - 1 1without any argument. If you do so, an error might
result if the user has himself made an - 1 1 request in the interim.
In short, you should either save registers or use a different environment whenever
you change formatting parameters in the opening macro of a macro pair. Then restore
them in the closing macro of the pair.

Notes, Cautions, and Warnings

Another important macro for technical manuals is one that gives a consistent way of
handling notes, cautions, and warnings. (Traditionally, a note gives users important
information that they should not miss, but will not cause harm if they do. A caution is
used for information that, if missed or disregarded, could lead to loss of data or damage
to equipment. A warning is used for information that is critical to the user's life or
limb.)
Obviously, this is a simple macro-all that is required is some way of making the
note, caution, or warning stand out from the body of the text. You could redefine the
macro shown here in any number of ways depending on the style of your publications.
.de Ns \ ' I note/caution/warning; $1 = type "N", "C", "W"
.sp 2
.ne 5
.ce
.if !"\\$l"N" .if !"\\$l"C*' .if ! " \ \ $ l " W " \{\
. tm "Need N, C, or W as argument for Ns macro-using N"
\f3NOTE\fl\}
.if "\\$l"N" \f3NOTE\fl
.if " \ \ $ l " C " \f3CAUTION\fl
.if "\\$l"W" \f3WARNING\fl
- SP
.ns
.nr nI \\n(.iu \ " Save current indent, if any
.nr nL \\n ( .lu \ " Save current line length
.ie \\nS>O .nr IN 5n\" Make indent less if in small format
.el .nr IN 10n \ " Larger indent for full-size page
.in +\\n(INu \ '' Indent specified amount
- 1 1 -\\n(INu \ " Decrement line length same amount
-.
.de Ne \ I ' "note end"; no args
.in \\n(nIu \ " Restore previous indent
- 1 1 \\n(nLu \ " Restore previous line length
.rr nI \ " Remove temporary registers
.rr nL
.sp 2
..
 An Extended m s Macro Package 0 53 1

A warning looks like this:

WARNING

You should be careful when reading books on t r o f f, because they can be damag-
ing to your health. Although escape sequences are allowed, they are not exactly
high adventure.

A different version of a caution macro is shown below. It uses a graphic symbol

to mark a caution statement.

CAUTION
One client had a convention of marking a
caution statement with a large diamond in a
square. These diamonds will appear in a
second color in the printed book.

To produce the escape sequences to draw the symbol, we used p i c , processing the
description and capturing it in a file. Then we read it into our macro definition. (We
could also have produced the escape sequences to draw the symbol without pic's
help; this would result in much more compact code.) The drawing of the symbol does
take up most of the .Gc macro definition. Before we actually output the symbol, the
current vertical position is marked. After it is output, we mark its bottom position.
Then we return to the top before placing the warning label and processing the text.
After the caution statement is output, the closing macro, GE, checks the current verti-
cal position against the bottom position of the symbol.
.de Gc \"Graphic Caution Macro
.ne 10
.mk a \ " Mark current t o p position
.
br \ " pic output belongs here
\v' 7 2 0 ~ '\D' 10u -7201.1'
.sp -1
\D'1720u Ou'
.sp -1
\h' 720u' \D' 1Ou 720u'
.sp -1
\h' 7 2 0 ~ '\v' 7 2 0 ~ '\D' 1 - 7 2 0 ~OU'
.sp -1
\h' 360u' \D' 1360u 360u'
.sp -1
532 0 UNlX Text Processing 0

\v' 3 6 0 ~ '\D' 1 - 3 6 0 ~3601.1'

\h' 7 2 0 ~ '
.sp -1
\h' 36011'\v' 7 2 0 ~ \D'' 1-360~-36011'
.sp -1
\~'360~'\D'1360~-360~'
.sp -1
.sp 1+72Ou \ I mEnd of pic output
- SP
.mk q \" Mark bottom of symbol
- sp 1 \\nau \" Move back to top (.mk a )
.in +1.5i \" Indent to right of symbol
- 1 1 -.5i \" Reduce line length
- s p .5v
- ce
\f3CAUTION\fl \ " Output Caution label
.sp .3v

.de GE \ " Graphic Caution end

.br
-SP
.in \ " Reset previous settings
-11
\ " If bottom of symbol (.mk q )
\ " is below current vertical position
\ " then move to that position
.if \\nqu>\\n(nlu+\\n(.vu .sp I\\nqu
.sp .3v

Table of Contents, Index, and Other End Lists

Here's the part you've all been waiting for. One of the nicest things a formatter can do
for a writer i s automatically generate lists such as a table of contents and an index.
These are very time consuming to produce manually, and subject to error. There are
basically two ways to do the trick, and both apply to an index as well as a table of con-
tents, endnotes, and other collected lists.
The technique used by mm, which generates an automatic table of contents at the
end of each formatting run, is to collect headings into a diversion using the .da
request. This diversion is then output from within a special macro called the "end
macro," which we have yet to discuss.
The second technique is to use the .t m request to write the desired information
to standard error output. Then that output i s redirected to capture the messages in a file,
where they can be edited manually or automatically processed by other programs.
 a An Extended m s Macro Package 0 533

The advantage of the first approach is that it is clean and simple, and entirely
internal to the formatter. However, it is really suitable only for short documents. A
long document such as a book is not normally formatted in a single pass, but chapter by
chapter. It is not desirable to format it all at once just to get the table of contents at the
end. In addition, a large document generally will end up creating a large diversion-
often one that is too large for t r o f f to handle.
The second approach, on the other hand, opens up all kinds of possibilities for
integration with other tools in the UNIX environment. The output can be saved, edited,
and processed in a variety of ways. As you can imagine from our philosophy of letting
the computer do the dirty work, this is the approach we prefer.
However, there is still a place for diversions, so we'll take a close look at both
approaches in the sections that follow.

Diverting to the End

Although we prefer to create our major end lists-the able of con :n 5 and index-by
writing to stderr,we find it very useful to use diversions for another type of list.
We've added a couple of special macros that allow a writer to insert remarks
intended specifically for the reviewers of a draft document or for personal use. Because
technical reviewers frequently miss questions embedded in the text, we designed the
. R n macro to highlight notes. This macro makes these remarks stand out in the text
and then collects them for output again at the end of the document.
.de Rn \ " Note t o reviewers : $1 = Note
\ " Print note in text and at end
\ " Output note first
- SP
\f3Note t o reviewers:\fP \\$l
- SP
.ev 2
.da rN \ " Then append into diversion
.sp 0.2v
.in 0
.ie "\\*(NN"" \ ( s q Page \\n(PN: \\$l
.el \(sq Page \\*(NN: \\$l
.br
.da
.nr RN 1 \ " Flag it for EM
.ev
..
Another macro, .Pn, is used to collect a list of personal notes or reminders and output
them on a page at the end. These notes do not appear in the body of the text.
.de P n \ " Personal Note; $1= note
\ " Note listed at end, b u t not in text
.ev2
.if \ \ n ( ~ n < l.nr P n 0 1 \ " Set up autoincrement counter
534 0 UNlX Text Processing 0

.da pN
-br
.IP " \ \ n + ( P n . " 5n
\\$I
ie " \ \ * (NN"" (Page \ \ n (PN)
.el (Page \\*(NN)
.br
- da
. n r pN 1 \ " Flag it for EM
- ev
..
Only the .Rn macro produces output in the body of the document, but both macros
append the notes into a diversion that we can process at the end of the document. The
divert and append (.da) macro creates a list of notes that can be output by invoking
the macro created by the diversion.
For each macro, we format the lists slightly differently. In the .Rn macro, we
print a box character ( 0 ) (to give the feeling of a checklist), then the page number on
which the review note occurred. This allows the reviewer or the writer to easily go
back and find the note in context. In the .P n macro, we use an autoincrementing
counter to number personal notes; this number is output through .IP. It is followed
by the note and the page reference in parentheses.
The formatting of text inside a diversion can be tricky. The text could be format-
ted twice: when it i s read into the diversion, and when the diversion is output. The one
thing to keep in mind is that you don't want line filling to be in effect both times. If
line filling is in effect when the text i s read into the diversion, you should turn it off
when the diversion is output. You can also use transparent output (\ !) to hide macros
or requests so that they will be executed only at the time the diversion i s output. We
have also taken the precaution of processing the diversion in a separate environment.
Now what about printing the list at the end? Well, as it turns out, nroff and
t rof f include a special request called .e m that allows you to supply the name of a
macro that will be executed at the very end of the processing run, after everything else
is finished.
The .e m request allows you to define the name of a macro that will be executed
when all other input has been processed. For example, the line:
.em EM

placed anywhere in a file or macro package, will request that the macro .EM be exe-
cuted after everything else has been done. The definition of .EM is up to you.
The m s macros already have specified the name of this macro as .EM, the end
macro. In its usual obscure way, mm calls its end macro .) q. If you are writing your
own package, you can call it anything you like. You can either edit the existing end
macro, or simply add to it using the .a m (appendto mucro) request.
All that ms does with this macro is to process and output any diversions that
have not been properly closed. (This might happen, for example, if you requested a
floating keep, but its contents had not yet been printed out.)
 0 An Extended m s Macro Package 535

The end macro is a good place to output our own special diversions that we've
saved for the end. What we need to do now is to add some code for processing our list
of review notes:

.de EM
-br
.if \\n (RN=l \ { \
\&\C

'b P

.ce
\f3NOTES TO REVIEWERS\fl
-sp 2
Reviewers, please address t h e following questions:
- SP
.ev 2
.nf
.rN
.ev
.\I
.if \\n(pN=l \ { \
.br
\&\C
'bP

.
ce
\f3Notes To Myself:\fl
.sp 2
.ev 2
-
nf
- PN
- ev
-\I
..

(Note: we have found that to print anything from the .EM macro in the standard m s
package, it i s necessary to invoke . N P explicitly following a page break. However,
when using our simplified version of this package as shown in the last chapter, our
.EM does not need a .NP.) The list collected by the .Rn macro is printed on a new
page, looking something like this:
536 UNIX Text Processing 0

NOTES TO REVIEWERS

Reviewers, please address the following questions:

0 Page 3-1: Why can't I activate the bit pad before opening a font file?

0 Page 3-7: I s there a size restriction on illuminated letters?

A Diverted Table of Contents

Given the preceding discussion, it should be easy for you to design a diverted table of
contents. The magic .
t C macro we kept invoking from our headings might look
something like this:
.de t C \** t a b l e of contents; $ l = s e c t number;
$2=title; $3=type
. i f "\\$3"\\* (cH"\ { \
. d a SL \ " Divert and append t o s e c t i o n l i s t
.sp 3
\ \ * ( c H \\$I: \ \ $ 2
. s p 1.5
.d a
.\I
.if "\\$3"Ah"\{ \
.da SL \ " D i v e r t and append t o s e c t i o n list
.b r
\\$l \\$2\\a\\t\\* (NN
.b r
.d a
.\I
.if "\\$3"Eh"\ { \
. d a sL \ " Divert and append t o s e c t i o n l i s t
- br
\\$l \ \ $ 2 \ \ a \ \ t \ \ * (NN
.b r
- da
-\I
-if \ \ $ 3 "Figure " \ { \
r'

. d a fL \ " D i v e r t a n c ~ a p p e n d t o figure ist

\\$l \ \ $ 2 \ \ a \ \ t \ \ * (NN
- da
-\I
-if "\\$3"Table" \{ \
.da tL \ " Divert and append t o t a b l e l i s t
\\$l \\$2\\a\\t\\* (NN
 0 An Extended m s Macro Package 0 537

.da
- \ }
__
The diversion S L i s set up to handle the main heading (chapter, appendix, unit, or sec-
tion) and two levels o f subheadings (A-heads or B-heads). The diversions fL and t L
are set up to compile lists of figures and tables, respectively.
In the end macro, to print the table of contents, you have to cause a break to a
new page, print: introductory captions, and so on, and then follow by outputting the
collected diversion of each type. The following example shows the code to print:
.br \ " Automatically invoke diverted toc
\&\C \ " by including these lines in EM macro
'bP \ " Or place in own macro
.ta \\n (LLu-5n \\n (LLuR
.ce
\f3Table of Contents\fR
.sp 2
.nf \ " Process in no-fill mode
\\t\f3~age\fP
.SL
.rm s L \ " Clear diversion
\ " Add code here t o output figure
\ " and table list diversions
We set two tab stops based on the default line length (\n (LLu). The second tab stop
is used to set a right-adjusted page number in the right margin. The first tab stop is
used to run a leader from the entry to the page number. The escape sequences that out-
put the leader and tab (\a and \t)were specified in the .t C macros. (And to protect
the escape sequence inside a diversion an extra backslash was required.)
Now we can obtain a table of contents each time we format the document. The
format of the table of contents shows the hierarchical structure of the document:

Table of Contents

Page

Chapter Two: Getting Started with Alcuin

2.1 Objectives of this Session ................................................................ 2-1

2.2 Starting Up the System .................................................................... 2-2
2.2.1 Power-up Procedure .................................................................... 2-2
2.2.2 Software Initialization ................................................................. 2-3
2.3 Creating Simple Glyphs ................................................................... 2-4
538 0 UNlX Text Processing 0

When Diversions Get Too Big

One of the major problems with collecting a table of contents in a diversion is that, with
a large document, the diversions quickly grow too large for the formatter to handle. It
will abort with a message like “Out of temp file space.”
The solution is to break up your diversions based on the number of entries they
contain. One way to do this is to base the name of the diversion on a number register,
and do some arithmetic to increment the name when the diversion has been added to a
certain number of times.
For example, instead of just diverting to a macro called .sL, we could divert to
one called xn, where n is a number register interpolation generated as follows:
.de t C

.nr X X +1
.nr x0 \\n(xX/100+1
.da x\\n(xO

Each time .t c is called, register xx is incremented by 1, and its value, divided by

100, is placed into another register, x0. Until the value of register xx exceeds 100-
that is, until .tC has been called 99 times-xO will be equal to 1. From 100 to 199,
x0 will be equal to 2, and so on.
Accordingly, the actual macro into which output is diverted-represented as
x\\n (x0-will first be x l , then x2, and so on.
When it comes time to output the collected entries, instead of calling a single
diversion, we call the entire series:
.xl
.x 2
.x 3
* x4
Here, we are assuming that we will have no more than 400 entries. If there are fewer
entries, one or more of these diverted macros may be empty, but there’s no harm in
that. If there are more than 400, the contents of .x5 (et al) would still have been col-
lected, but we would have failed to print them out. We have the option of adding
another in the series of calls in the end macro, or rebuking the user for having such a
large table of contents!

Writing to Standard Error

Although we’ve answered one of the objections to a diverted table of contents by the
register arithmetic just shown, there is another, more compelling reason for not using
this approach for large documents: there is no way to save or edit the table of contents.
 o An Extended m s Macro Package 0 539

It is produced on the fly as part of the processing run and must be recreated each time
you print the document.
For a very large document, such as a book, this means you must format the entire
book, just to get the table of contents. It would be far preferable to produce the table of
contents in some form that could be saved, so the tables from each chapter could be
assembled into a single large table of contents for the entire book.
(Incidentally, producing a table of contents for a large document introduces some
other issues as well. For example, you may want to have an overall table of contents
that shows only top-level headings, and individual chapter table of contents that give
more detail. Working out the macros for this approach is left as an exercise for the
reader.)
The best way to produce a table of contents for a large book is simply to write the
entries to standard error using .t m , and rely on an external program to capture and
process the entries.
In d i t r o f f , you can instead use the .s y request to execute the echo com-
mand and redirect the entries to a file. An example of this method might be:
.sy echo \\$l \\$2\a\t\\*(NN >> toc$$
However, this approach causes additional system overhead because it spawns echo
subprocesses. Also, because it does not work with o t r o f f, we have used the more
general approach provided by .t m .
Our .t C macro might look like this:

.de tC \ " Standard error; table of contents;

\ " $l=sect number; $2=title; $3=type
-if "\\$3"\\* (cH"\{\
.tm ><CONTENTS:.sp 3
tm ><CONTENTS:\\*(cH \\$1\\$2
-tm ><CONTENTS:.sp 1.5
.\I
.if "\\$3"Ah" -
tm ><CONTENTS: \\$I \\$2\a\t\\* (NN
.if "\\$3"Bh" .tm ><CONTENTS:\\$l \\$2\a\t\\* (NN
.if "\\$3"Figure" .tm ><FIGURE:\\$l \\$2\a\t\\*(NN
.if "\\$3"Table" .tm ><Table: \\$l \\$2\a\t\\* (NN

Instead of diverting the section lists to separate macros from the lists of figures and
tables, we send all entries out to standard error.
To capture this output in a file, we simply need to redirect the error output:
$ ditroff -Tps ... 2> toc

To do this, we will use our format shell script, which was introduced in Chapter 12,
and will be revisited in the next (and final) chapter.
Because actual error messages might be present in the output, we prefix a label
indicating the type of entry, for example:
540 0 UNlX Text Processing 0

><CONTENTS:
><FIGURE:
><TABLE:

It will be up to some outside program to separate the different groups of entries and
subject them to further processing. We'll use a sed script to separate the entries in
the table of contents from the figure lists, table lists, and index entries. (In the next
chapter, we'll look at the post-processing of these entries.) Now let's look at a macro
to generate index entries that will also be written to standard error.

Indexes
A simple index can be handled in much the same way as a table of contents. A macro
for a simple index might look like this:
.de XX
\ " Section-page number set up
\ " by Se macro in string NN
.tm INDEX:\\$l\t\\*(NN
..
You might also want to have a macro that doesn't print the page number, but is
just used for a cross-reference:
.de XN \ " Cross-reference Index entry, no page number
.tm INDEX:\\$l
..
You might also want a macro pair that will index over several different pages:
.de IS Index macro
\Iv

\ " Interpolate % for page number

.ie \\n(.$=l .tm INDEX:\\$l, \\n%
.el \I\
.nr X\\$2 \\n%
.ds Y\\$2 \\$l \ I
.if \\n(.t<=lP .tm *\\$l* near end of page
.if \\n(nl<l.2i .tm *\\$I* near top of page
..
.de IE \ " Index end macro
.ie !\\n(.$=l .tm IE needs an argument!
.el .tm INDEX:\\*(Y\\$l, \\n(X\\$l-\\n%
.if \\n(.t<=lP .tm *\\*(Y\\$l* near end of page
.if \\n(nl<l.2i .tm *\\*(Y\\$l* near top of page
_.
The .I S macro prints out an entry, just like .XX. However, in addition, it saves the
argument into a string, and takes a letter or digit as an optional second argument. This
second argument is used to define a number register and string that will be saved, and
 0 An Extended ms Macro Package 0 54 1

not printed until the index and macro is called with the same argument. The index and
macro print the starting number, followed by a hyphen and the current page number.
All of this discussion still avoids one major issue. The real trick of indexing is
what you do with the raw output after you have it, because a great deal of sorting, con-
catenation, and reorganization is required to rearrange the entries into a meaningful
order. Fortunately or unfortunately, this topic will have to wait until the next chapter.
 A P E
C
.
H T R

18
m

Putting It All Together

Before returning to the topic of table of contents and index processing, using shell tools
that we will build, let’s review what we’ve covered so far.
We started with a promise to show you how the UNIX environment could support
and enhance the writing process. To do that, we’ve had to delve into many details and
may have lost the big picture.
Let’s return to that big picture here. First, UNIX provides what any computer
with even rudimentary word-processing capabilities provides: the ability to save and
edit text. Few of us write it perfectly the first time, so the ability to rewrite the parts of
a document we don’t like without retyping the parts we want to keep is a major step
forward.
However, no one will argue that UNIX offers better tools at this simple level than
those available in other environments. The v i editor is a good editor, but it is not the
easiest to learn and lacks many standard word-processing capabilities.
Where UNIX’s editing tools excel is in performing complex or repetitive edits. A
beginner may have little rlse for pattern matching, but an advanced user cannot do
without it. Few, if any, microcomputer-based or standalone word processors can boast
the sophisticated capabilities for global changes that UNIX provides in even its most
primitive editors.
When you go beyond v i , and begin to use programs such as e x , sed, and
awk, you have unmatched text-editing capabilities-power, if you will, at the expense
of user friendliness.
Second, UNIX’s hierarchical file system, multiuser capabilities, and ample disk
storage capacity make it easy to organize large and complex writing jobs-especially
ones involving the efforts of more than one person. This can be a major advantage of
UNIX over microcomputer-based or dedicated word processors.
Anyone who has tried to write a multiauthor work on a floppy-based system
knows how easy it is to lose track of the latest version of a file, and to get lost among a
multitude of disks. UNIX makes it easy to share files, and to set up a consistent frame-
work for managing them.

. 542
 0 Putting It All Together 0 543

In addition to storing multiple versions of documents on line, you can use the file
system to set up specific environments for writing. For example, a separate .e x r c file
in each directory can define abbreviations and command maps specific to a book or sec-
tion.
Third, UNIX provides a wide range of formatting tools. Using t r o f f , pic,
t b l , and eqn, you can easily typeset books. This is not as unique and powerful a
capability as it was even two or three years ago. The advent of low-cost laser printers
and wysiwyg “desktop publishing” tools like Microsoft WORD, MacWrite, and Aldus
Pagemaker allow PC users to do typesetting as well.
However, despite the glamor of desktop publishing, and the easy-to-use appeal of
products for the Macintosh, the UNIX typesetting facilities offer many advantages.
Chief among these advantages is the very feature in which t r o f f at first seems much
weaker than its low-end competitors, namely, the use of embedded codes to control for-
matting.
Wysiwyg systems are easy for beginners to use, and they are very satisfying
because you can immediately see what you are going to get on the printed page. But
have you ever tried to make a global font change in MacWrite? Or had to make a
change to a document after it was “pasted up” with Pagemaker? Or had to wait end-
lessly while Microsoft WORD reformats an entire document after you change the mar-
gins?
Because t r o f f codes can be edited, just like any other text in a file, it is very
easy to change your mind about formatting and make global changes. And after you
have mastered the art of writing macros, it i s even easier to change formats simply by
changing macro definitions. And because the editing and formatting functions are
separate, you don’t have to wait for the computer while you are making those
changes-that happens while you print.
This is not to say that t r o f f is superior to the best possible wysiwyg system.
High-end systems from companies like Interleaf, Xyvision, and Texet offer power,
speed, and ease of use all at once. Unfortunately, the software is costly, and requires
the use of high-cost bit-mapped workstations. This can lead to a bottleneck in docu-
ment production unless you have enough money to spend on hardware. Because
t r o f f requires only a standard alphanumeric terminal, it provides much more “bang
for the buck.”
There is no question that the publishing system of the future will be a wysiwyg
system. But for now, a low-cost UNIX system with v i and t r o f f is still one of the
most cost-effective publishing systems around.
This brings us to the final strength of UNIX-its extensibility. More than an
operating system or a collection of programs, UNIX is a philosophy of computing.
Let’s consider an analogy. The Volkswagen beetle was a unique automobile of the six-
ties and seventies. Its simple design was one of the reasons that made it popular; the
“bug” was user-maintainable. V W owners (“users”) could tinker with their cars, per-
forming such tasks as changing spark plugs by hand. They scoffed at owners of other
cars who depended upon mechanics. It is perhaps this same feeling of independence-
let me do it myself-that the UNIX environment fosters in its users. There are many
quite capable software environments that are packaged to keep users out. In some
ways, the secret of UNIX is that its working parts are visible. The UNIX environment,
544 UNIX Text Processing

like the VW beetle, is designed so that users can take it apart and put it back together.
UNIX is a philosophy of computing. As we’ve stressed again and again, UNIX pro-
vides general-purpose tools, all of which are designed to work together.
No single program, however well thought out, will solve every problem. There is
always a special case, a special need, a situation that runs counter to the expected. But
UNIX is not a single program: it is a collection of hundreds. And with these basic
tools, a clever or dedicated person can devise a way to meet just about any text-
processing need.
Like the fruits of any advanced system, these capabilities don’t fall unbidden into
the hands of new users. But they are there for the reaching. And over time, even writ-
ers who want a word processor they don’t have to think about will gradually reach out
for these capabilities. Faced with a choice between an hour spent on a boring, repetitive
task and an hour putting together a tool that will do the task in a flash, most of us will
choose to tinker.
The index and table of contents mechanism in this chapter is a good example of
putting together individual UNIX tools to do a job that no one of them can easily do
alone. Its explanation is a fitting end to this book, which has tried throughout to put the
UNIX text-processing tools in a wider context.

Saving an External Table of Contents

A s discussed in the last chapter, t rof f does provide a mechanism (namely diver-
sions) to collect and process a table of contents directly within the formatter. However,
this approach is best suited to short documents, because it requires that the entire docu-
ment be reformatted to produce the table of contents.
Likewise, you could even produce and sort an index entirely within trof f,
though the effort required would be large. (In fact, a recent article on Usenet, the on-
line UNIX news network, described an implementation of a sort algorithm using
t rof f macros. It is painfully slow-it was done just to prove that it could be done,
rather than for practical application.)
The beauty of UNIX, though, is that you don’t have to stretch the limits of
trof f to do everything necessary to produce a book. Just as editing is separated from
formatting, you can separate processing the table of contents and the index from format-
ting the rest of the text.
The trof f formatter provides the basic mechanisms for producing the raw
material-the lists of headings or index terms, accompanied by the page numbers on
which they occur. However, the actual saving and processing of the raw material is
done with make, sed, awk, sort, and the shell.
In Chapter 12, we began to look at how a shell script (which we called format)
could manage the formatting process. We used the programming power of the shell not
only to save the user the trouble of remembering command-line options and compli-
cated postprocessor names, but also to apply the power of sed to various ancillary for-
matting tasks.
 0 Putting It All Together 0 545

The collection of a table of contents and index requires that we first return to this
script. As we left Chapter 17, both the table of contents and the index macros simply
write data to standard error.
A Bourne shell user can redirect this error output to a file using the following
syntax:
$ dittoff file 2> tocfile
The problem is that the table of contents, index entries, and potential formatter error
messages are all captured in the same file. We need a mechanism for parsing this file
into its separate elements. The user could do this manually, but it is far better to let a
program do it.
The first step is to redirect all of the error output from the formatter to a tem-
porary file. After formatting i s done, we can use sed to search for the identifying
strings that we introduced as part of the "error message" and output the matching lines
into separate files. True error messages should be sent back to the screen, and the tem-
porary file removed.
The trick here is naming the files into which the saved data i s stored by s e d . It
is not appropriate simply to append table of contents data to one file, because we are
likely to reformat a document many times while rewriting and revising it. Instead, we
want to have a unique table of contents file and a unique index file for each source file
that we format. The best way to do this without cluttering up the current directory is to
create a subdirectory for each type of data we want to save-toc, i n d e x , and so on.
Let's look at how we did these things in the format script:
roff="ditrof f -Tps"; files=""; options="-mS"
pre=" I ditbl"; post=" I devps ''
sed="l sed -f /work/macros/new/cleanup.sed"
pages=""; toc="2>/tmp$$"; lp=" I lp - s "
if [ ! -d index a ! -d toc 1; then
echo "NO index and toc. Use the buildmake command."
toc= " 2> / dev /nu11''
fi
while [ ~ ~ $ #! =~ - **O " I; do
case $1 in
-?) echo "Format Options are:''
echo "-m* Specify other macro package (-mm)"
echo "--s Use small format (5-1/2 by 8-1/2)"
echo " - 0 Print selected pages"
echo "-cg Format for Compugraphic typesetter"
echo "-E Invoke EQN preprocessor"
echo "-P Invoke PIC preprocessor"
echo " - G Invoke GFZAP & PIC preprocessors"
echo "-x Redirect output to /dev/null"
echo "-y Invoke nroff; pipe output to screen";
echo "-a Set interactive troff -a option"
echo "-* Any troff option"; exit;;
-m*) options="$l";;
546 0 UNlX Text Processing 0

-s) options="$options -rS1 -rvl"; ;

-0) pages="$pages -0$1"; toc="2>/dev/null"; ;
-cg) roff="ditroff -Tcg86"; post=" I ditplus -dtcg86"; ;
-E) pre="$pre I dieqn";;
-P) pre="l pic -T720 -D Spre";;
-G) pre="l grap I pic -T720 -D Spre";;
-x) options="$options - z " ; post=""; lp=""-,, -
- y ) roff="nroff"; post=""; lp="l col I pg";;
-a) post=""; options="$options -a" ,, --
- * ) options="$options $1"; ;
* ) if [ -f $1 1; then
files= " $ files $1"
tx fi1 e= " $1"
if [ -d /print 1; then touch /print/$txfile
else
echo "USAGE: format (options) files"
echo " T O list options, type format - ? "; exit
fi; ;
esac
shift
done
if -n "$files" - 0 ! -t 0 1; then
# Use soelim to expand . s o ' s in input files
# otherwise use cat to send files down pipe.
eval "cat $files $sed $pre I
$roff $options - $toc $post $pages $toc $ 1 ~ "
else echo "fmt: no files specified"; exit
fi
if [ -f tmp$$ 1; then
if [ -d toc 1; then
sed -n -e "S/^><CONTENTS: \ ( . * \ ) /\l/p" tmp$$ > toc/$txfile
fi
if [ -d index ] ; then
sed -n -e " s / ~ > < I N D E X : \ ( . * \ ) / \ ~ / ~tmp$$
" > index/$txfile
fi
if [ -d figlist 1; then
sed -n -e "s/"><FIGURE:\ ( . * \ ) /\l/p" tmp$$ > figlist/$txfile
fi
if [ -d tablist 1; then
sed -n -e lr~/A><TABLE:\(.*\)/\l/p" tmp$$ > tablist/$txfile
fi
sed -n "/"></!p"
rm /tmp$$
fi
exit
 0 Putting It All Together 547

Now, for example, when we format a file called chOl, a file of the same name will be
written in each of the four subdirectories toc, index, f iglist, and tablist.
Each time we reformat the same file, the output will overwrite the previous contents of
each accessory file, giving us the most up-to-date version. When we use the -0 option
for only partial formatting, writing out of these files is disabled by redirecting error out-
put to /dev/null, so that we don't end up with a partial table of contents file.
There's also a -x option, to allow us to format a file to produce the table of con-
tents and index without producing any regular output. This option uses t r o f f ' s - z
option to suppress formatted output, and sets the post and lp shell variables to the
null string.
(You may also notice the -cg option, which specifies a different device to both
t ro f f and the postprocessor-in this case, a Compugraphic typesetter instead of an
Apple Laserwriter. This is included as an aside, to give you an idea of how this is
done.)
The contents of the toc, f iglis t,and tab1i s t directories can be assem-
bled into a complete table of contents, or formatted on the spot for a chapter-level table
of contents. We can use the following simple sequence of commands (which could be
saved into a shell script):
echo .ta \n(LLu-Sn \n(LLuR > book-toc
echo .ce >> book.toc
echo \f3TABLE OF CONTENTS\fP >> book.toc
echo .sp 2 >> book.toc
echo "\t\f3Page\fP" >> book.toc
cat /toc/ch?? /toc/app? >> book.toc
echo .bp >> book.toc
cat /figlist/ch?? /figlist/app? 7 > book.toc
echo -bp >> book-toc
cat /tablist/ch?? /tablist/app? >> book.toc

The resulting book - toc source file looks like this:

.ta \n (LLu-5n \n (LLuR
.ce
\f3TABLE OF CONTENTS\fP
.sp 2
I I\f3Page\fP
.sp 3
Chapter 1 Introduction to Alcuin
.sp 1.5
1.1 A Tradition of Calligraphic Excellence\a\tl-2
1.2 Illuminated Bit-Mapped Manuscripts\a\tl-4
-sp 3
Chapter 2 Getting Started with Alcuin
.sp 1 . 5
2.1 Objectives of this Session\a\t2-1
2.2 Starting Up the System\a\t2-2
548 0 UNlX Text Processing u

2.2.1 Power-up Procedure\a\t2-2

The index will require more serious postprocessing.

. Index Processing
It is relatively simple to assemble the components of a table of contents into sequential
order, but it is much more difficult to process the index entries, because they must be
sorted and manipulated in a variety of ways.
This is one of the most complex tasks presented in this book. So let's start at the
beginning, with the raw data that is output by t rof f, and directed to our index sub-
directory by the format shell script. For illustration, we'll assume a sparse index for
a short book containing only three chapters.
As you may recall, the user creates the index simply by entering macro calls of
the form:
.XX "input devices"
or:
.XX "input devices, mouse"

or:
.XR "mouse (see input devices)"
throughout the text. Both macros write their arguments to standard output; the . X X
macro adds the current page number, but the . X R (cross reference) macro does not.
The user is responsible for using consistent terminology, capitalization, and spelling. A
comma separates optional subordinate entries from the major term.
An index term should be entered on any page that the user wants indexed-at the
start and end of a major topic, at least, and perhaps several in between if the discussion
spans several pages.
In our example, entries are saved into the three files chOl, ch02, and ch03
in the order in which they appear in the respective input files. The indexing term
entered by the user is printed, separated from the current page number by a tab. Certain
cross reference entries do not have a page number. The content of the raw index files
after chapters 1 through 3 have been formatted follows. (Here, and in the following dis-
cussion, a tab is represented by the symbol I 1.)
$ c a t index/ch??
Alcuin, overview of1 11-1
illuminated manuscripts1 11-1
fonts, designing1 11-2
Alcuin, supported input devicesl 11-2
input devices 1 11-2
input devices, mouse I 11-2
input devicesl 11-2
mouse (see input devices)
 0 Putting It All Together 0 549

input devices, bit padl 11-3

bit pad (see input devices)
input devices I 11-3
startup, of systeml 12-1
power, location of main switch1 12-3
power, for graphics display1 12-1
startup, of systeml 12-2
input devices, mouse1 12-2
input devices, bit padl 12-3
fonts, selecting1 13-1
glyphs, designing1 13-2
extra line space1 13-3
symbolic n a m e s ) 13-3
C! operator1 13-4

To create a presentable index from this raw data, we need to do the following:

Sort the entries into dictionary order, and remove duplicates, if any. (Duplicate
entries occur whenever the user enters . X X macros with the same argument
over several input pages, and two or more of those entries fall on the same out-
put page.)
Combine multiple occurrences of the same term, appending a complete list of
page numbers and properly subordinating secondary terms.
Introduce formatting codes, so that the resulting file will have a pleasing, con-
sistent appearance.

Just how complex a task this is may not be immediately apparent, but rest assured that
it takes the combined operation of sort, u n i q , and several different awk and
s e d scripts to do the job properly.
Fortunately, we can hide all of this complexity within a single shell program, so
that all the user needs to type is:
$ cat index/files I indexprog > book.ndx

Sorting the Raw Index

The first part of i n d e x p r o g processes the index entries before they are passed to
awk. The s o r t program prepares a list of alphabetical index entries; u n i q removes
duplicate entries.
s o r t -t\l I -bf +O -1 +In I uniq

The options to the s o r t command specify primary and secondary sort operations,
affecting the first and second fields separately. The -t option specifies that a tab char-
acter separates fields. The primary sort is alphabetic and performed on the indexing
term; the secondary sort is numeric and performed on the page number. The primary
sort is also controlled by the foilowing options: the -b option (ignore leading blanks
550 UNlX Text Processing 0

in making comparisons) is a safety feature; the -f (fold uppercase and lowercase

letters) is more important because the default sort order places all uppercase letters
before all lowercase ones; and +O -1 ensures that the alphabetic sort considers only
the first field. The secondary sort that is performed on the second field (+ln) i s
numeric and ensures that page numbers will appear in sequence.
Now let’s look at the index entries after they have been sorted:
13 operator1 13-4
Alcuin, overview of1 11-1
Alcuin, supported input devicesl 11-2
bit pad (see input devices)
extra line space1 13-3
fonts, designing 1-2
fonts, selecting1 13-1
glyphs, designing1 13-2
illuminated manuscripts1 11-1
input devicesl 11-2
input devicesl 11-3
input devices, bit padl 11-3
input devices, bit padl 12-3
input devices, mouse I 11-2
input devices, mouse I 12-2
mouse (see input devices)
power, f o r graphics display1 12-1
power, location o f main switch1 12-1
startup, of system1 12-1
startup, of system1 12-2
symbolic names1 13-3

Multiple entries that differ only in their page number are now arranged one after the
other.
The sort command is a simple way to obtain a sorted list of entries. However,
sorting can actually be a complicated process. For instance, the simple sort com-
mand that we showed above obviously works fine on our limited sample of entries.
And while it is designed to process entries with section-page numbering (4-1, 4-2, 4-3),
this command also works fine when sorting entries with continuous page numbering ( 1 ,
2.3).
However, section page numbering does present a few additional problems that we
did not encounter here, Two-digit section numbers and page numbers, as well as
appendices (A-1, A-2, A-3) will not be sorted correctly. For instance, this might cause
the indexing program to produce the following entry:
Alcuin, software A-2, 1-1, 1-10, 1-3, 11-5, 2-1

There are two ways to handle this problem. One is to change the indexing macro in
t rof f so that it produces three fields. Then the sorting command can sort on the sec-
tion number independent of the page number. (Because our awk portion of the index-
 Putting It All Together 0 551

ing program is set up to operate on entries with one or two fields, you'd have to change
the program or use a sed script to reduce the number of fields.)
The second method uses sed to replace the hyphen with a tab, creating three
fields. Actually, we run a sed script before the entries are sorted and another one
after that operation to restore the entry. Then s o r t will treat section numbers and
page numbers separately in secondary numeric sort operations, and get them in the right
order.
The only remaining problem is how to handle appendices. What happens is that
when a numeric sort i s performed on section numbers, lettered appendices are sorted to
the top of the list. This requires cloaking the letter in a numeric disguise. Presuming
that we won't have section numbers greater than 99, our sed script prepends the
number 100 to each letter; this number is also removed afterwards.
sed '
s/I I \ ( [O-91 [0-91*\) -/ I 1\11 I/
s/I I\ ( [A-Z] \ ) - / I I lOO\l I I/' I
s o r t -t\ -bf + O -1 + I n +2n I u n i q 1
sed '
s/ I I loo\ ( [A-Zl \I I 111 I \I-/
s/\ ( I I .*\) I I /\l-/'
Now the sorting operation of our index program handles a wider range of entries.

Building the Page Number List

The next step is more complex. We must now combine multiple occurrences of each
term that differ only in the page number, and combine all of the page numbers into a
single list. The awk program is the tool of choice. We can use a script for comparing
and manipulating successive lines similar to the one described in Chapter 13. We begin
by building the page number list for each entry.
awk '
B E G I N { ORS = " " ;F S = ' ' 1 I" 1
NF == 1 { i f (NR == 1) p r i n t f ( " % s " , $0);
else p r i n t f ("\n%s", $0) )
NF > 1 1
i f ($1 == c u r r )
p r i n t f (", % s " , $2)
else (
i f (NR == 1) p r i n t f ("%s", $0)
e l s e p r i n t f ("\n%s", $0)
c u r r = $1
}
1'
First, the program sets the output record separator (ORs)to the nul1 string, rather than
the default newline. This means that output records will be appended to the same line,
unless we specify an explicit newline.
552 UNlX Text Processing

Second, it sets the field separator (FS) to the tab character. This divides each
index entry into two fields: one containing the text, the other containing the page
number. (As you may recall, the page number is separated from the text of the entry by
a tab when it is output from t r o f f .)
Then, if the number of fields (NF) is 1 (that is, if there is no tab-separated page
number, as is the case with cross reference entries generated with .XR), the program
prints out the entire record ($0). If this is not the first line in the file (NR = l), it
precedes the record with an explicit newline ( \ n).
If the number of fields is greater than 1 (which is the case for each line containing
a tab followed by a page number), the program compares the text of the entry in the
first field ( $ 1 ) with its previous value, as stored into the variable c u r r .
The next few lines might be easier to understand if the condition were written in
reverse order:
if ($1 ! = curr)
{ if ( N R == 1) printf ("%is", $0)
else printf ("\n%s", $0)
curr = $ 1
1
else printf (", % s " , $2)

If the first field is not equal to c u r r , then this i s a new entry, so the program
prints out the entire record (again preceding it with an explicit newline if this is not the
first line of the file). The value of c u r r is updated to form the basis of comparison
for the next record.
Otherwise (if the first field in the current record is the same as the contents of the
variable c u r r ) , the program appends a comma followed by the value of the second
field ( $2 ) to the current record.
The output after this stage of the program looks like this:
@ operator1 13-4
Alcuin, overview of1 11-1
Alcuin, supported input devices1 11-2
bit pad (see input devices)
extra line space1 13-3
fonts, designing 1-2
fonts, selecting I 13-1
glyphs, designing1 13-2
illuminated manuscripts1 11-1
input devices I I1-2,1-3
input devices, bit padl I1-3,2-3
input devices, mouse I I 1-2,2-2
mouse (see input devices)
power, for graphics display1 12-1
power, location of main switch1 12-1
startup, of system1 I2-1,2-2
symbolic names1 13-3
 Putting It All Together 0 553

Subordinating Secondary Entries

The next trick i s to subordinate secondary entries under the main entry, without reprint-
ing the text of the main entry. In addition, we want to represent consecutive page
numbers as a range separated by two dots (. .) rather than as a list of individual pages.
We'll show this script in two sections:
1 awk
2 BEGIN { FS = "1 I'I; I
3 I
4 n = split ($1, curentry, " , " )
5 if (curentry[l] == lastentry[l])
6 printf ( " % s " , curentry[2])
7 else {
8 if ( n > 1) printf ("%s\n B s " , curentry[l], curentry[2])
9 else printf ( " % s " , $1)
10 lastentry [ 11 = curentry [l]
11 l
12
This section of the script uses auk's s p l i t function to break the first field into two
parts, using a comma as a separator.
There are several cases that the program has to consider:

The text of the entry does not contain a comma, in which case we can just
print the entire first field. See h e 9: p r i n t f ( r v % ~ ' f , $1).
The entry does contain a comma, in which case we want to see if we have a
new primary term ( c u r e n t r y [ 1] ) or just a new secondary one ( c u r e n -
t r y [ZI).
If the primary term i s the same as the last primary term encountered (and saved
into the variable 1a st e n t ry), we only need to print out the secondary term.
See line 6: p r i n t f ( " % s ", c u r e n t r y [ 2 ] ) .
Otherwise, we want to print out both the primary and secondary terms: See
line 8: p r i n t f ( " % s \ n % s " , c u r e n t r y [ l ] , c u r e n t r y [ 2 ] ) .

For example:
@ operator1 13-4
Alcuin, overview of1 11-1
Alcuin, supported input devices1 11-2

When the first line is processed, the split will return a value of 0, so the entire line will
be output.
When the second line is processed, l a s t e n t r y contains the string @ operator,
c u r e n t r y [ 11 contains Alcuin, and c u r e n t r y [ 2 ] contains overview of. Because
l a s t e n t r y is not the same as c u r e n t r y [ l ] , the program prints out both
c u r e n t r y t l l and c u r e n t r y L 2 1 .
554 0 UNlX Text Processing 0

When the third line is processed, curentry[l) again contains the word
Alcuin, but curentry [ 2 I contains the words supported input devices. In this case,
only curent ry [ 2 3 is printed.
The next part of the script, which follows, looks considerably more complicated,
but uses essentially the same mechanism. It splits the second field on the line (the page
number list) on the hyphens that separate section number from page number. Then, it
compares the various sections it has split to determine whether or not it is dealing with
a range of consecutive pages. If so, it prints only the first and last members of the
series, separating them with the range notation (. .).
If you were able to follow the previous portion of the script, you should be able to
piece this one together as well:
NF == 1{ printf ("\n") 1
(NF > 1 ) & & ($2 ! - / . * - .*/I t
printf ("\t")
n = split ($2, arr, " , " )
printf ( " % s " , arr[l])
split (arr [l], last, 'I-'')

for (i = 2; i <= n; ++il t

split (arr [il , curr, " - " )
i f ((currlll == last[ll)&&(curr[21/1 == last[2]/1+1)) (
if (i ! = n) t
split (arr[i+ll, follow, " - " )
if ((currC11 ! = follow[ll) 1 I (curr[2]/1+1 ! = follow[2]/1))
printf ( " . . % s " , arr[i])
1 e l s e printf ( " - - % s " , arr[i])
1 e l s e printf ('I, %s", arr[i])
last[ll = curr[ll; lastl21 = curr[21
1
printf ("\n")
1'

The output from this awk program (in sequence with the previous ones) now looks
like this:
@ operator1 13-4
Alcuin
overview of1 11-1
supported input devices! 11-2
bit pad (see input devices)
extra line space( 13-3
fonts
des igning 1-2
selecting 13-1
glyphs
des i gn i n g 13-2
illuminated manuscripts1 11-1
input devices I 11-2.. 1-3
bit padl 11-3, 2-3
mouse I 11-2, 2-2
mouse (see input devices)
 0 Putting It All Together 0 555

power
for graphics display! 12-1
location of main switch1 12-1
startup
of system1 12-1 2-2 f f

symbolic names1 13-3

That’s starting to look like an index!

Adding Formatting Codes

We could simply quit here, and let the user finish formatting the index. However, awk
can continue the job and insert formatting codes.
We’d like auk to put in headings and divide the index into alphabetic sections.
In addition, it would be nice to insert indentation requests, so that we can format the
index source file in fill mode so that any long lines will wrap correctly.
Let’s look at the coded output before we look at the script that produces it. Only
the beginning of the output is shown:
.ti -4n
(3 operator1 13-4
. br

.ne 4
.ti -2n
\ fBA\f R
- br
.ne 2
.ti -4n
Alcuin
.br
.ti -4n
overview of1 11-1
.br
.ti -4n
supported input devices1 11-2
.br
.ne 4
.ti -2n
\ f B B \ fR
.br
.ne 2
.ti -4n
bit pad (see input devices)
- br
556 0 UNlX Text Processing 0

.ne 4
.ti -2n
\ fBE\ fR
- br
.ne 2
.ti -4n
extra line space( 13-3
- br

Here's a script that does this part of the job:

awk
BEGIN {OFS = ""

lower = "abcdefghijklmnopqrstuvwxyz"
upper = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
1
N F > O {
if ($0 ! - .*/) {
/ A

n = l
while ((newchar = substr($l,n,l)) ! - /[A-Za-z] / ) {
n = n + l
if (n == 1 0 0 ) { # bad line
newchar = oldchar
break
}
1
if (newchar / [a-z] / ) { -
for (i = 1; i <= 26; ++i) {
if (newchar == substr (lower, i, 1 ) ) [
newchar = substr (upper, i, 1)
break
}
1
1
if (substr ($l,l, 1) -
/ [0-9]/)
newchar =
if (newchar ! = oldchar) {
printf ("\n\n%s\n", .ne 4")
printf ("%s\n", ".ti -2n")
p r int f ( I' % s % s % s \ n " , \ \ f B , newchar , " \ \ fR" )
printf (ll%s\n"," .br")
oldchar = newchar }
printf ('*%s\n", ".ne 2 " )
1
printf ("%s\n", ".ti -4n")
printf ("%s\n", $0)
 0 Putting It All Together 0 557

printf ("%s\n", ".br")

I'
Every line in the input (NF > 1) will be subjected to the last three lines in the pro-
gram. It will be surrounded by formatting codes and printed out.
printf ("%s\n", " . t i -4n")
printf ("%s\n", $0)
printf ("%s\n", ".br")
The rest of the script checks when the initial character of a primary entry changes and
prints a heading.
As you may have noticed, in the output of the previous script, secondary entries
were indented by three leading spaces. They can be excluded from consideration at the
outset by the condition:
if ($0 ! - / A .*/) {

All other lines are checked to determine their initial character. The awk program's
substr function extracts the first letter of each line. Then, much as it did before, the
program compares each entry with the previous until it detects a change.
The program is basically looking for alphabetic characters, but must test (espe-
cially in computer manuals) for strings that begin with nonalphabetic characters. (If it
doesn't do this, it will loop endlessly when it comes to a string that doesn't begin with
an alphabetic character.) If the program loops 100 times on a single line, it assumes
that the character is nonalphabetic, breaks out of the loop, and goes on to the next line.
When the program finds a change in the initial alphabetic character, it prints a
heading consisting of a single capital letter and associated formatting requests.
Primary terms beginning with nonalphabetic characters are output without causing
a change of heading. (Because they are already sorted to the beginning of the file, they
will all be listed at the head of the output, before the A's.)

Final Formatting Touches

Having come this far, it hardly seems fair not to finish the job, and put in the final for-
matting codes that will allow us to format and print the index without ever looking at
the source file (although we should save it to allow manual fine-tuning if necessary).
A simple sed script can be used for these final touches:
sed *'li\\
.Se \'r\'l \"Index\"\\
. i n +4n\\
.MC 3.15i 0.2i\\
.ds RF I n d e x - \\\\\\\\n(PN\\
.ds CF\\
.ds LF\\
.n a
s/ I I/ /"
558 0 UNlX Text Processing 0

Assuming that we’re using our extended ms macros, these initial macros will create the
section heading Index, print the index in two columns, and use a page number of the
form Index - n. (Note how many backslashes are necessary before the number
register invocation for E”. Backslashes must be protected from the shell, sed, and
t rof f. This line will be processed quite a few times, by different programs, before it
is output.)
Finally, the script converts the tab separating the index entry from the first page
number into a pair of spaces.

Special Cases
But our indexing script is not complete. There are a number of special cases still to
consider. For example, what about font changes within index entries? In a computer
manual, it may be desirable to carry through “computer voice” or italics into the index.
However, the t rof f font-switch codes will interfere with the proper sorting of
the index. There is a way around this-awkward, but effective. As you may recall, we
use a sed script named c l e a n u p . sed called from within format. This script
changes double quotation marks to pairs of matched single quotation marks for typeset-
ting, and changes double hyphens to em dashes. W e can also use it to solve our current
problem.
First, we add the following lines to cleanup. sed:
/^\.X[XRI/I
s / \ \ \ (fP\)/%%-/g
s / \ \ \ (fS\)/%%--/g
S/\\\ (fB\)/ % % - - - / g
s / \ \ \(fI\)/ % % - - - - 19
S / \ \ \(fR\)/ % % - - - - - /g
s / \ \ \(f (CW\) / % % - - - - - - /g
1
Within an .X X or .XR macro, the script will change the standard t rof f font-
switch codes into an arbitrary string of nonalphabetic characters.
Then we add the -d option (dictionary order) to our initial sort command in
the index program. This option causes sort to ignore nonalphabetic characters when
making comparisons. (The exception will be lines like @ operator, which contain no
alphabetic characters in the first field. Such lines will still be sorted to the front of the
list.)
Finally, we use the concluding sed script in the indexing sequence to restore the
proper font-switch codes in the final index source file:
s/%%------ / \ \ \ \ € (CW/g
s / % % - - - - - / \ \ \ \ fR/g
s/%%---- /\\\\fI/g
s/%%--- / \ \ \ \ f B / g
s/%%--/\\\\fS/g
s/%%-/\\\\fP/g
 0 Putting It All Together 0 559

We might also want to consider the case in which a leading period (as might occur if
we were indexing t r o f f formatting requests) appears in an index entry. Inserting the
following line one line from the end of the last awk script we created will do the trick.
These lines insulate t r o f f codes in index entries from the formatter when the index
source file is processed by t r o f f for final printing:
if ($0 - / A \ - .* / ) printf ("\\&")
if ($0 - /-%%--*\./) printf ("\\&")

Lines beginning with a . will be preceded with a trof f zero-width character (\&).

The Entire Index Program

We have broken the indexing process into stages to make it easier to understand. How-
ever, there is no need to keep individual awk and sed scripts; they can be combined
into a single shell program simply by piping the output of one portion to another, within
the shell program.
Here's the whole program, as finally assembled:
sed '
s/I I\ ( L0-91 lo-91 * \ )-1 I 1\11 I/
s/I I\ ( I
[A--Zl \)-I I100\1 I I/' I
sort - t \ l -bdf + O -1 + I n +2n 1 u n i q I
sed '
s/I I loo\ ( [ A - Z l \ ) I 111 I \I-/
SI\ ( I I .*\) I l/\l-/' I
awk '
B E G I N { O R S = '*"; FS = "I I" 1
NF == 1 { i f (NR 1) p r i n t f
== ("%s", $ 0 ) ;
e l s e p r i n t f ("\n%s", $0) }
NF > 1{
if ($1 == curr)
printf (", %s", $2)
else {
if (NR == 1) p r i n t f (rr%~", $0)
else p r i n t f ("\n%s", $0)
c u r r = $1
1
}' I awk '
BEGIN { FS = ''1 I"; 1
t
n = s p l i t ($1, c u r e n t r y , " , " )
i f ( c u r e n t r y [ 11 == l a s t e n t r y [ 11 )
p r i n t f (" %s", c u r e n t r y [ 2 ] )
else {
if (n > 1) p r i n t f ("%s\n %s", c u r e n t r y [ l ] , curentry[2])
560 0 UNlX Text Processing 0

else printf ( r r % ~ " ,$1)

lastentry [ 1 I = curentry [ 13
1
1
NF == 1 { printf ("\n") }
(NF > 1) & & ($2 ! - / . * -. * / ) [
printf ("\t")
n = split ($2, arr, " , " )
printf ( " % s " , arr [l])
split (arr[ l ] , last, " - " )
for (i = 2; i <= n; ++i) {
split (arr[i], curr, " - " )
if ((curr[ll == last[ll) & & (curr[2]/1 == last[2]/1+1)) (
if (i != n) {
split (arr[i+l], follow, " - ' I )
if ((currI11 != follow[ll) I I (curr[2]/1+1 != follow[21/1))
printf ( " . . % s " , arr[i])
} else printf ("..%s", arr[i])
} else printf ( " , %s", arr[i])
last[ll = curr[ll; last[2] = curr[2]
j
printf ("\n")
1' I awk
BEGIN { O F S = "'I
lower = "abcdefghijklmnopqrstuvwxyz"
upper = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
1
N F > O {
if ( $ 0 !- /" .*/) {
n = l
while ( (newchar = substr ($l,n,1)) ! - / [A-Za-z] / ) {
n = n + l
if (n == 100) { C bad line
newchar = oldchar
break
1
}
if (newchar -
/[a-z]/) (
for (i = 1; i <= 26; ++i) {
if (newchar == substr (lower, i, 1)) {
newchar = substr (upper, I, 1)
break
}
1
l
 0 Putting It All Together 0 561

if ( s u b s t r ($1,1,1) - / [0-91/)
newchar = ""
if ( n e w c h a r ! = o l d c h a r ) {
printf ("\n\n%s\n", ".ne 4")
printf (*'%s\n", " . t i -2n")
p r i n tf ( " % s% s % s \n" , " \ \ f B" , n e w c h a r , I' \ \ f R" )
printf ("%s\n", " .br")
oldchar = newchar

}
printf ("%s\n", ".ne 2")
1
printf ("%s\n", " . t i -4n")
if ( $ 0 - / A \ -. */) p r i n t f ("\\ti")
i f ($0 - /A%%--*\ -
/) printf ("\\&")
printf ("%s\n", $ 0 )
printf (*'%s\n", " - b r " )
}' 1 sed "li\\
. ~ e\ " \ " \ " I n d e x \ " \ \
.in +4n\\
.MC 3.4i 0 . 2 i \ \
.ds RF I n d e x - \\\\\\\\n(PN\\
.ds CF\\
.ds LF\\
.n a
s/%%------ /\\\\i (Cw/g
s/%%----- / \ \ \ \ f R / g
s/%%---- / \\\\fI/g
s/%%---/\\\\fB/g
s/%%--/\\\\€S/g
-
s / % % / \ \ \ \ fP / g "

The result of all this processing is source text that can be piped directly to the formatter,
saved in a file for later formatting (or perhaps minor editing), or both (using t e e to
"split" standard output into two streams, one of which is saved in a file).
Assuming that the various raw source files produced by t ro f f are stored in a
subdirectory called i n d e x f iles, and that the index script is called i n d e x p r o g ,
we can format and print the index as follows:
$ cat i n d e x f i l e s / * I indexprog I d i t r o f f -mS I ... I lp

The result will look something like this:

562 0 UNIX Text Processing 0

INDEX

@ operator 3-4 P
power
for graphics display 2-1
A location of main switch 2-1
Alcuin
overview of 1-1
supported input devices 1-2 S
startup
of system 2-1-2-2
B symbolic names 3-3
bit pad (see input devices)

1 Let make Remember the Details

Even though we’ve hidden the really complex details of index processing inside a shell
script, and the format shell script itself handles a lot of the dirty work, there is still a lot
for the user to keep track of. The make utility introduced in Chapter 1 1 can take us a
long way towards making sure that everything that needs to happen for final production
of the finished book comes together without a hitch.
Here are some of the things we want to make sure have been done:

All of the relevant sections have been printed in their most up-to-date form.
Odd as it may seem, it is possible to have last minute changes to a file that
never make i t into the printed book.
. The book has been proofed using whatever automatic tools we have provided,
including the proof and double shell scripts (or wwb if you have it).
All “review notes” embedded in the text must also be satisfied and removed.
An updated table of contents and index have been printed.

You can probably think of others as well.

 0 Putting It All Together 0 563

The make utility is the perfect tool for this job. We’ve already seen in Chapter
11 how it can be used to specify the files (and the formatting options) required for each
section of the book. Unfortunately, this part of the job requires that you keep formatted
output files, which are quite large. If disk space is a problem, this drawback might lead
you to think that make isn’t worth the bother.
However, with a little thought, you can get around this restriction. Instead of
keeping the formatted output file, you can keep a zero-length file that you touch
whenever you format the source file. You could add the following line to the end of
the format script:
touch print/$file

Or, if you use make itself to print your document, you could put the touch com-
mand into the makefile. Your makefile might look like this:
book : print/chOl print/ch02 print/ch03 ...
print/chOl : chOl
sh /usr/local/bin/format -mS -rvl -rS2 chOl
touch print/chOl

print/ch02 : ch02
sh /usr/local/bin/format -mS -P -rvl -rS2 ch02
touch print/ch02

Notice that in order to execute the local formatting shell script, it is necessary to exe-
cute s h and specify the complete pathname. The options specified with the format
shell script can be specific to each file that is formatted. However, generally you want
to use the same options to format all the files that make up a particular document.
Using variables, you can create a more generalized makefile that is easier to change.
FORMAT: s h /usr/local/bin/format
OPTIONS: -mS -P -rvl -rS2

book : print/chOl print/ch02 print/ch03 ...

print/chOl : chOl
$ (FORMAT) $ (OPTIONS) chOl
touch print/chOl

print/ch02 : ch02
$ (FORMAT) $ (OPTIONS) ch02
touch print/ch02
564 UNlX Text Processing 0

The variables used by make are set like shell variables. But when they are referenced,
the name of the variable must be enclosed in parentheses in addition to being prefixed
with a dollar sign.
A user can now easily edit the OPTIONS variable to add or remove options. You
could also place additional options on the command for a particular file. This is not
necessary, though, just because some of the files have tables, equations, or pictures and
others don't. Other than the small bit of extra processing it requires, there's no reason
not to run the preprocessors on all files.
Our makefile can be further expanded. To make sure that our index and table of
contents are up-to-date (and to automate the process of creating them out of the indivi-
dual raw output files that the format script creates), we can add the following dependen-
cies and creation instructions:
book: print/chOl ... proof/chOl . _ .book-index book-toc

print/chOl : chOl

book-index : index/chOl index/ch02 -..

cat index/* I sh /usr/local/bin/indexprog > book-index

book.toc : toc/chOl ...figlist/chOl - - - tablis.t/chOl...

echo .ta \n(LLu-5n \n(LLuR > book-toc
echo .ce >> book-toc
echo \f3TABLE OF CONTENTS\fP >> book-toc
echo .sp 2 >> book-toc
echo "\t\f3Page\fP" >> book.toc
cat toc/chOl ...toc/appz >> book-toc
echo ' .bp' >> book-toc
cat figlist/chOl ...figlist/appz >> book-toc
echo ' .bp' >> book-toc
cat tablist/chOl...tablist/appz >> book.toc

toc/chOl : chOl
$ ( F O R M A T ) $(OPTIONS) -X ChOl
toc/ch02 : ch02
$ (FORMAT) $ (OPTIONS) -X ch02

index/chOl : chOl
$ (FORMAT) $ (OPTIONS) -X ChOl
 0 Putting It All Together 0 565

figlist/chOl : chOl
$(FORMAT) $(OPTIONS) -X ChOl

tablist/chOl : chOl
$ (FORMAT) $ (OPTIONS) -X ChOl

3ecause we have directories named t o c and i ndex , we give our source files names
such as book. t o c and book. index.
We can therefore enter:
$ make book.toc

and the table of contents will be compiled automatically. When you enter the above
command, the make program recognizes book. toc as a target. It evaluates the
following line that specifies several dependent components.
book.toc: toc/chOl toc/ch02 toc/ch03

In turn, each of these components are targets dependent on a source file.

toc/ch02: ch02
$ (FORMAT) $ (OPTIONS) -X ch02

What this basically means is that if changes have been made to to ch02 since the file
book t oc was compiled, the source file will be formatted again, producing new toc
entries. The other files, assuming that they have not been changed, will not be re-
formatted as their entries are up-to-date.
We can add other “targets”, for instance, to check whether or not every chapter
in the book has been proofed since it was last edited. Based on when the dependent
components were last updated, you could invoke the p r o o f program on the associated
file, grep for Review Note macros, or just print a message to the user reminding him
or her to proof the file.
To do this, we create a pseudo-target. If no file with the name p r o o f exists, it
can never be up-to-date, so typing:
$ make proof

will automatically force proofing of the document according to the rules you have speci-
fied in the makefile.
The p r i n t directory also serves as a pseudo-target, useful for printing indivi-
dual chapters. Users don’t have to remember the formatting options that must be speci-
fied for a particular file.
And if all these reasons don’t convince you to learn make and begin construct-
ing makefiles for large documents, perhaps this next benefit will. It gives you a simple
two-word command to print an entire book and its apparatus.
566 0 UNlX Text Processing 0

$ make book
When you enter this command, each formatting command as it is being executed will
be displayed on the screen. If you wish to suppress these messages while you do other
work, invoke make with the - s option or place the line .SILENT : at the top of
the makefile.

Building the Makefile

You are limited only by your imagination and ingenuity in organizing your work with a
makefile. However, the more complex the makefile, the longer it gets-, and the more
difficult for inexperienced users to create.
You can get around this problem too-just write a shell script to build the
makefile, taking as arguments the files that make up the document. Here's such a
script, called buildmake, that will produce a makefile similar to the one just
described. (The make utility requires that the actions to be performed for each target
begin with a tab. Such explicit tabs are shown in the following script by the symbol
I 1.)

if [ $ # -eq 0 1;
then
echo ''USAGE : buildmake files"
echo " ( Y o u must specify the files that make up the book)"
exit
fi
if [ ! -d print 1 ; then
mkdir print
mkdir proof
fi
if [ ! -d index 1; then
mkdir index
fi
if [ ! -d toc 3 ; then
mkdir toc
mkdir figlist
mkdir tablist
fi
for x
do
prifiles="$prifiles print/$x"
profiles="$profiles proof/$x"
tcfiles="$tcfiles toc/$x"
xfiles="$xfiles index/$x"
fg fi1 e s= " $ fg fi1 e s figlist / $ x "
tbfiles="$tbfiles toc/$x"
done
echo ".SILENT:" > makefile
 0 Putting It All Together 0 567

echo "FORMAT = sh /usr/local/bin/format" >> makefile

echo "OPTIONS = -mS" >> makefile
echo "INDEXPROG = sh / u s r / l o c a l / b i n / i n d e x p r o g " > > m a k e f i l e
echo "book : Sprifiles $profiles book-toc book.index">>makefile
echo "book.index : $xfiles/" >> makefile
echo " I lcat Sxfiles I $(INDEXPROG) > book.index">>makefile
echo " I I $ (FORMAT) $ (OPTIONS) book. index" >> makefile
echo "book-toc : Stcfiles" >> makefile
echo " I lecho .ta \n(LLu-5n \n(LLuR > book.toc">>makefile
echo " I lecho .ce >> book.toc" >> makefile
echo " I /echo \f3TABLE OF CONTENTS\fP >> book.toc">>makefile
echo " I lecho .sp 2 >> book.toc" >> makefile
echo '* I \echo "\t\f3Page\fP" >> book-toc" >> makefile
echo " I lcat /work/lib/toc-top > book-toc" >> makefile
echo 'I I lcat Stcfiles >> book.toc" >> makefile
echo '' I echo .bp >> book-toc" >> makefile
echo I' I cat Sfgfiles >> book.toc" >> makefile
echo '* 1 echo .bp >> book-toc" >> makefile
echo I' 1 cat Stbfiles >> book.toc" >> makefile
echo '' I $ (FORMAT) $ (OPTIONS) book-toc" >> makefile
for x
do
echo "print/$x : $x" >> makefile
echo " I I $ (FORMAT) $ (OPTIONS) $x" >> makefile
echo "proof/$x : $x" >> makefile
echo " f lecho $x has n o t been proofed" >> makefile
echo "toc/$x : $x" >> makefile
echo " I I$(FORMAT) $(OPTIONS) -x $x" >> makefile
echo "index/$x : $x" >> makefile
echo " I ($(FORMAT) $(OPTIONS) -x $ x " >> makefile
done

To create a complex makefile, all the user needs to do is type:

$ buildmake files
In addition, the user may want to manually edit the first line of the makefile, which
specifies formatter options.

Where to Go from Here

Large as this book is, it is far from comprehensive. We have covered the basic editing
and formatting tools in some detail, but even there, topics have been glossed over. And
when i t comes to the more advanced tools, programs not explicitly designed for text
processing, much has been left out.
568 0 UNIX Text Processing 0

The sheer size and complexity of UNIX is one of its fascinations. To a beginner,
it can be daunting, but to an advanced user, the unknown has an appeal all its own.
Particularly to a technical writer, for whom the computer is a subject as well as a tool,
the challenge of taking more control over the process of book production can be end-
lessly fascinating. The subject and the method of inquiry become ever more
intertwined, until, in Yeats’s immortal phrase:
How can you know the dancer from the dance?
 A P P E N D 1 X
m

Editor Command Summary

This section is divided into five major parts, describing the commands in the text edi-
tors v i , ex, sed, and a w k , and the pattern-matching syntax common to all of
them.

Pattern-Matching Syntax
A number of UNIX text-processing programs, including ed, e x , v i , sed, and
grep, allow you to perform searches, and in some cases make changes, by searching
for text patterns rather than fixed strings. These text patterns (also called regular
expressions) are formed by combining normal characters with a number of special char-
acters. The special characters and their use are as follows:

Matches any single character except newline.

* Matches any number (including zero) of the single character (includ-
ing a character specified by a regular expression) that immediately
precedes it. For example, because . means “any character,” . *
means “match any number of any characters.”
[..I Matches any one of the characters enclosed between the brackets.
For example, [ A B ] matches either A or B. A range of consecutive
characters can be specified by separating the first and last characters
in the range with a hyphen. For example, [A-Zl matches any
uppercase letter from A to Z, and [0-9] matches any digit from 0
to 9. If a caret (”) is the first character in the brackets, the com-
parison is inverted: the pattern will match any characters except those
enclosed in the brackets.

569 .
570 0 UNlX Text Processing 0

\ { n,m\ } Matches a range of occurrences of the single character (including a

character specified by a regular expression) that immediately precedes
it. n and m are integers between 0 and 256 that specify how many
occurrences to match. \{ n\} matches exactly n occurrences, \{ n,\}
matches at least n occurrences, and \{n,m\) matches any number of
occurrences between n and m. For example, A\ { 2 , 3 \ 1 matches
either A A (as in AARDVARK) or A A A (as in A A A Travel Agency)
but will not match the single letter A. This feature is not supported in
all versions of v i .
h
Requires that the following regular expression be found at the begin-
ning of the line.
$ Requires that the preceding regular expression be found at the end of
the line.
\ Treats the following special character as an ordinary character. For
example, \ . stands for a period and \ * for an asterisk.
\ ( \) Saves the pattern enclosed between \ ( and \ ) in a special holding
space. Up to nine patterns can be saved in this way on a single line.
They can be “replayed” in substitutions by the escape sequences \ 1
to \ 9. This feature is not used in g r e p and egrep.
\n Matches the nth pattern previously saved by \ ( and \ ) , where n is
a number from 0 to 9 and previously saved patterns are counted from
the left on the line. This feature is not used in g r e p and egrep.

The e g r e p and auk programs use an extended set of metacharacters:

regexp+ Matches one or more occurrences of the regular expression

(regexp).
regexp? Matches zero or one Occurrences of the regular expression.
regexp I regexp Matches lines containing either regexp.
( regexp ) Used for grouping in complex regular expressions (e.g., with
I above).

Regular expressions in e x (: commands from vi> offer some different extensions:

\< Constrains the following pattern to be matched only at the beginning of a

word.
\> Constrains the following pattern to be matched only at the end of a word.
\u Appended to the replacement string of a substitute command, converts first
character of replacement string to uppercase.
 0 Editor Command Summary 0 571

\U Appended to the replacement string of a substitute command, converts

entire replacement string to uppercase.
\1 Appended to the replacement string of a substitute command, converts first
character of replacement string to lowercase.
\L Appended to the replacement string of a substitute command, converts
entire replacement string to uppercase.

The vi Editor

Command-Line Syntax
There are two commands to invoke the vi editor:
vi [options] Wle(s)l
or:
vi e w Cfile(s)]
If a file i s not named, vi will open a file that can be given a name with the :f com-
mand or when it is saved using the :w command. If more than one file is named, the
first file is opened for editing and :n is used to open the next file. The view com-
mand opens the firstfile for read-only use; changes cannot be saved.
Options:

-1 Open file for editing LISP programs

-r Recover file
-R Open file in read-only mode; same as using view
-ttag Start at rug
-X Open encrypted file
+ Open file at last line
+n Open file at line n
+ /pattern Open file at first occurrence of pattern
-wn Set window to n lines

Operating Modes
After the file is opened, you are in command mode. From command mode, you can
invoke insert mode, issue editing commands, move the cursor to a different position in
572 UNIX Text Processing 0

the file, invoke e x commands or a UNIX shell, and save or exit the current version of
the file.
The following commands invoke insert mode:
a A i I o O R s S
While in insert mode, you can enter new text in the file. Press the ESCAPE key to exit
insert mode and return to command mode.

Command Syntax
The syntax for editing commands is:
[ n ] operator [n] object

The commands that position the cursor in the file represent objects that the basic editing
operators can take as arguments. Objects represent all characters up to (or back to) the
designated object. The cursor movement keys and pattern-matching commands can be
used as objects. Some basic editing operators are:

c Change
d Delete
y Yankor copy

If the current line is the object of the operation, then the operator is the same as the
object: cc, dd, yy. n is the number of times the operation i s performed or the
number of objects the operation is performed on. If both n’s are specified, the effect is
n times n.
The following text objects are represented:

word Includes characters up to a space or punctuation mark. Capitalized

object is variant form that recognizes only blank spaces.
sentence Up to . ! ? followed by two spaces.
paragraph Up to next blank line or paragraph macro defined by p a r a =
option.
section Up to next section heading defined by sect= option.

Examples:

2cw Change the next two words

dl Delete up to the next paragraph
dA Delete back to the beginning of the line
 0 Editor Command Summary 0 573

5YY Copy the next five lines

3dl Delete three characters to the right of the cursor

Status Line Commands

Most commands are not echoed on the screen as you input them. However, the status
line at the bottom of the screen is used to echo input for the following commands:

/ ? Start pattern-matching search forward ( / ) or backwards (?)

*
* Invoke an e x command
I Invoke a UNIX command that takes as its input an object in the
buffer and replaces it with output from the command

Commands that are input on the status line must be entered by pressing the RETURN
key. In addition, error messages and output from the *G command are displayed on the
status line.

Summary of vi Commands

Repeat last command (insert, change, or delete).

"@ Repeat last command.
@buffer Execute command stored in buffer.
a Append text after cursor.
A Append text at end of line.
"A Unused.
b Back up to beginning of word in current line.
B Back up to word, ignoring punctuation.
"B Scroll backward one window.
C Change operator.
C Change to end of current line.
"C Unused.
d Delete operator.
D Delete to end of current line.
"D Scroll down half-window.
e Move to end of word.
E Move to end of word, ignoring punctuation.
*E Show one more line at bottom of window.
f Find next character typed forward on current line.
F Find next character typed back on current line.
574 0 UNlX Text Processing 0

^F Scroll forward one window.

g Unused.
G Go to specified line or end of file.
^G Print information about file on status line.
h Left arrow cursor key.
H Move cursor to home position.
^H Left arrow cursor key; BACKSPACE key in insert mode.
i Insert text before cursor.
I Insert text at beginning of line.
^I Unused in command mode; in insert mode, same as TAB key.
j Down arrow cursor key.
J Join two lines.
^J Down arrow cursor key.
k Up arrow cursor key.
K Unused.
^K Unused.
1 Right arrow cursor key.
L Move cursor to last position in window.
^L Redraw screen.
m Mark the current cursor position in register (a-z).
M Move cursor to middle position in window.
^M Carriage return.
n Repeat the last search command.
N Repeat the last search command in reverse direction.
^N Down arrow cursor key.
0 Open line below current line.
0 Open line above current line.
^O Unused.
P Put yanked or deleted text after or below cursor.
P Put yanked or deleted text before or above cursor.
^P Up arrow cursor key.
q Unused.
Q Quit v i and invoke ex
^Q Unused in command mode; in input mode, quote next character.
r Replace character at cursor with the next character you type.
R Replace characters.
^R Redraw the screen.
S Change the character under the cursor to typed characters.
S Change entire line.
^S Unused.
t Move cursor forward to character before next character typed.
T Move cursor back to character after next character typed.
^T Unused in command mode; in insert mode, used with autoindent option
set.
 0 Editor Command Summary 0 5 75

u Undo the last change made.

U Restore current line, discarding changes.
^U Scroll the screen upward half-window.
V Unused.
V Unused.
^V Unused in command mode; in insert mode, quote next character.
W M o v e to beginning of next word.
W Move to beginning of next word, ignoring punctuation.
^W Unused in command mode; in insert mode, back up to beginning of
word.
X Delete character under the cursor.
X Delete character before cursor.
^X Unused.
Y Yank or copy operator.
Y Make copy of current line.
^Y Show one more line at top of window.
Z Redraw the screen, repositioning cursor when followed by CR at the top,
. at the middle, and - at the bottom of screen.
22 Exit the editor, saving changes.
^Z Unused.

Characters Not Used in Command Mode

The following characters are unused in command mode and can be mapped as user-
defined commands.

^A g K ^K
^O 9 *T V
V "W *X ^Z
* \ - (underscore)

vi s e t Options

The following options can be specified with the :s e t command.

Option
(Abbreviation) Default Description
autoindent noai Indents each line to the same level as
(ai) the line above. Use with
shi f t w i d t h option.
576 0 UNIX Text Processing 0

aut oprint =P Changes are displayed after each editor

(ap) command. (For global replacement,
last replacement displayed.)
autowrite noaw Automatically writes (saves) file if
(awl changed before opening another file
with :n or before giving UNIX com-
mand with : !.
beaut ify nob f Ignores all control characters during
(bf) input (except tab, newline, or
formfeed).
directory =tmp Names directory in which ex stores
(dir) buffer files. (Directory must be writ-
able.)
edcompatible noed- Uses e d-like features on substitute
compatible commands.
errorbells errorbells Error messages ring bell.
(eb)
hardtabs =8 Defines boundaries for terminal
(ht1 hardware tabs.
ignorecase noic Disregards case during a search.
(ic)
lisp nolisp Indents are inserted in appropriate LISP
format. () { ) [[ and 11 are modified to
have meaning for lisp.
list nolist Tabs print as "1; ends of lines are
marked with $. (Used to tell if end
character is a tab or a space.)
magic magic Wildcard characters . * [ are special in
patterns.
mesg mesg Permits messages to display on termi-
nal while editing in vi.
number nonu Displays line numbers on left of screen
(nu) during editing session.
open open Allows entry to open or visual mode
from ex.
 0 Editor Command Summary 0 577

optimize noopt Deletes carriage returns at the end of

(opt) lines when printing multiple lines;
speeds output on dumb terminals when
printing lines with leading white space
(blanks or tabs).
paragraphs =IPLPPPQP Defines paragraph delimiters for move-
(para) LIPPlPiPbP ment by ( or I. The pairs of characters
in the value are the names of
n r o f f / t r o f f macros that begin
paragraphs .
prompt prompt Sets e x prompt (:).
readon 1y nor0 Any writes (saves) of a file will fail
(ro) unless you use ! after the write
(works with w, ZZ, or autowrite).
redraw noredraw Terminal will redraw the screen when-
(re) ever edits are made (insert mode
pushes over existing characters; deleted
lines immediately close up). Default
depends on line speed and terminal
type. noredraw is useful at slow
speeds on a dumb terminal; deleted
lines show up as @, and inserted text
appears to overwrite existing text until
you press ESC.
remap remap Allows nested map sequences.
report =5 Size of a large edit (i.e., number of
lines affected by a single edit) that will
trigger a warning message on bottom
line of screen.
scroll = [ l/2 window] Amount of screen to scroll.
sections =SHNHH HU Defines section delimiters for { ]
movement. The pairs of characters in
the value are the names of
n r o f f / t r o f f macros that begin
sections.
shell =/bin/sh Pathname of shell used for shell escape
(sh) (: !) and shell command (: sh). Value
is derived from shell environment.
578 0 UNlX Text Processing 0

shiftwidth sw=a Defines number o f spaces to indent

(sw) when using the >> or << commands
in the a u t o i n d e n t option.
showmatch nosm In v i , when ) or ] is entered, cursor
(sm) moves briefly to matching ( or {. (If
match is not on the screen, rings the
error message bell.) Very useful for
programming.
sh o w m o d e nosmd (System V, Release 2 v i only). The
(smd) string Input Mode is printed on the
command line whenever input mode is
entered.
slowopen Holds off display during insert.
(slow) Default depends on line speed and ter-
minal type.
tabstop =8 Sets number of spaces that a TAB
(ts) indents during editing session. (Printer
still uses system tab o f 8.)
ta g l e n g th =O Defines the number of characters that
(HI are significant for tags. Default (zero)
means that all characters are signifi-
cant.
tags =tags Pathname o f files containing tags. (See
/usr/lib/tags the tag(1) command.) B y default,
system searches /usr / 1i b / t ags
and the file t a g s in the current direc-
tory.
term Terminal type.
terse noterse Displays briefer error messages.
ti m e o u t timeout Macros "time out" after 1 second.

ttYtYPe Terminal type.

warn warn Displays N o write since last change as
warning.
window Shows a certain number of lines of the
(W) file on the screen. Default depends on
line speed and terminal type.
wrapscan ws Searches wraparound end of file.
(ws)
 0 Editor Command Summary 0 579

wrapmargin =O Defines right margin. If greater than

(wm) zero, automatically inserts carriage
returns to break lines.
writ eany nowa Allows saving to any file.
(wa)

The e x Editor
The ex editor is a line editor that serves as the foundation for the screen editor, vi.
All e x commands work on the current line or a range of lines in a file. In vi, ex
commands are preceded by a colon and entered by pressing RETURN. In ex itself, the
colon is supplied as the prompt at which you enter commands.
The ex editor can also be used on its own. To enter ex from the UNIX
prompt:
e x filename
Any of the options described for invoking vi may also be used with ex. In addition,
the vi command Q can be used to quit the vi editor and enter ex.
To exit ex:

X Exit, saving changes

9! Quit, without saving changes
vi Enter vi from ex

To enter an ex command from vi:

:address command options
The colon ( :) indicates an ex command. The address is a line number or range of
lines that are the object of the command.
The following options can be used with commands:

1 Indicates a variant form of the command.

parameters Indicates that additional information can be supplied. A parameter
can be the name of a file.
count Is the number of times the command is to be repeated.
flag #, p, and 1 indicate print format.

Unlike vi commands, the count cannot precede the command as it will be taken for
the address. d3 deletes three lines beginning with the current line; 3d deletes line 3.
580 0 UNlX Text Processing 0

As you type the address and command, it is echoed on the status line. Enter the com-
mand by pressing the RETURN key.

Addresses
If no address is given, the current line is the object of the command. If the address
specifies a range of lines, the format is:

where x and y are the first and last addressed lines. x must precede y in the buffer. x
and y may be line numbers or primitives. Using ; instead of , sets the current line to
x before interpreting y (that is, the current position will be at x at the completion of the
command). 1 , $ addresses all lines in the file.
The following address symbols can be used:

Current line
n Absolute line number
$ Last line
% All lines, same as 1, $
x- I +n n line before or after x
-In1 One or n lines previous
+ [nl One or n lines ahead
'X Line marked with x
r r
Previous context
lpatl or ?pat? Ahead or back to line matching pat

e x Commands

abbrev a b [string text]

Define string when typed to be translated into text. If string and text
are not specified, list all current abbreviations.
append [address]a [ !3
text

Append text at specified address, or at present address if none is

specified. With the ! flag, toggle the autoindent setting dur-
ing the input of text.
 0 Editor Command Summary 0 581

args ar
Print the members of the argument list, with the current argument
printed within brackets ([ I ) .
change [address]c [ !1
text

Replace the specified lines with text. With the ! flag, toggle the
a u t o i n d e n t setting during the input of text.

COPY [address]c odestination

Copy the lines included in address to the specified destination
address. The command t is a synonym for copy.
delete [address]d[bufSeer]
Delete the lines included in address. If bufSer is specified, save or
append the text to the named buffer.
edit e[!] [ + n ] f i l e
Begin editing on file. If the ! flag is used, do not warn if the
present file has not been saved since the last change. If the +n
argument is used, begin editing on line n.
file ? Ifilename]
Change the name of the current file tofilename, which is considered
?not edited.? If no filename i s specified, print the current status of
the file.
global [address]g[!] /pattern / [commands]
Execute commands on all lines that contain pattern. If commands
are not specified, print all such lines. If the ! flag is used, execute
commands on all lines not containing pattern.
insert [address]i[ !]
text

Insert text at line before the specified address, or at present address

if none is specified. With the ! flag, toggle the a u t o i n d e n t
setting during the input of text.
join [address]j [count]
Place the text in the specified range on one line, with white space
adjusted to provide two blank characters after a (.), no blank charac-
ters if a ) follows, and one blank character otherwise.
k [address]k char
Mark the given address with char.
 0 UNlX Text Processing

[address11[count]
Print the specified lines in an unambiguous manner.
m a p char commands
Define a macro named char in visual mode with the specified
sequence of commands. char may be a single character, or the
sequence #n, representing a function key on the keyboard.
mark [addresslmachar
Mark the specified line with char, a single lowercase letter. Return
later to the line with ’x .
move [addresslmdestination
Move the lines specified by address to the destination address.
next n [ ! ] [ [+cornmandl filelist]
Edit the next file in the command-line argument list. Use a r g s
for a listing of arguments. Iffilelist is provided, replace the current
argument list with filelist and begin editing on the first file; if com-
mand is given (containing no spaces), execute command after editing
the first such file.
number [addresslnu[count]
Print each line specified by address preceded by its buffer line
number. # may be used as an abbreviation for number as well as
nu.
open [addresslo[/pattern/]
Enter open mode at the lines specified by address, or lines match-
ingpattern. Exit o p e n mode with Q.
preserve P =e
Save the current editor buffer as though the system had crashed.
print [addresslp[count]
Print the lines specified by address with nonprinting characters
printed. P may also be used as an abbreviation.
[addresslpu[char]
Restore previously deleted or yanked lines from named buffer speci-
fied by char to the line specified by address; if char is not specified,
the last deleted or yanked text is restored.
quit s[!l
Terminate current editing session. If the file was not saved since the
last change, or if there are files in the argument list that have not yet
be accessed, you will not be able to quit without the ! flag.
 0 Editor Command Summary 0 583

read [address]r[!] vile]

Copy the text of file at the specified address. If file i s not specified,
the current filename i s used.
read [address]r !command
Read in the output of command into the text after the line specified
by address.
recover rec Vile]
Recoverfile from system save area.
rewind rew[! ]
Rewind argument list and begin editing the first file in the list. The
! flag rewinds without warning if the file has not been saved since
the last change.
set se parameter parameter2 .. .
Set a value to an option with each parameter, or if no parameter is
supplied, print all options that have been changed from their
defaults. For Boolean-valued options, each parameter can be
phrased as option or nooption; other options can be assigned with
the syntax, option=value.
shell sh
Create a new shell, Resume editing when the shell is terminated.
source sofile
Read and execute commands from file.
substitute [address]s [[/pattern/repN]options]
[count]
Replace each instance of pattern on the specified lines with repl. If
pattern and repl are omitted, repeat last substitution. The following
options are supported:
g Substitute all instances of pattern
c Prompt for confirmation before each change
t [address]t destination
Copy the lines included in address to the specified destination
address.
ta [address]t a tag
Switch the focus of editing to tag.
u n a b b r e v i a t e una word
Remove word from the list of abbreviations.
undo U
Reverse the changes made by the last editing command.
584 UNlX Text Processing 0

unmap unm char

Remove char from the list of macros.
V [addresslv/pattern/ [commands]
Execute commands on all lines not containing pattern. If commands
are not specified, print all such lines.
version ve
Print the current version number of the editor and the date the editor
was last changed.
visual [addresslvi [type] [count]
Enter v i s u a l mode at the line specified by address. Exit with
Q. type is either -, ", or . (see the z command). count speci-
fies an initial window size.
write [address]w[! ] [[>>]file]
Write lines specified by address tofile, or full contents of buffer if
address is not specified. Iffile is also omitted, save the contents of
the buffer to the current filename. If >> file is used, write contents
to the end of the specifiedfile. The ! flag forces the editor to write
over any current contents offile.
write [addresslw !command
Write lines specified by address to command through a pipe.
"q w q [ !1
Write and quit the file in one movement.
xit X
Write file if changes have been made to the buffer since last write,
then quit.
yank [address]y a [char][count]
Place lines specified by address in named buffer indicated by char.
If no char is specified, place in general buffer.
[address]z [type][count]
Print a window of text with line specified by address at the top.
type is as follows:
+ Place specified line at the top of the window (default)
- Place specified line at bottom of the window
* Print the window before the window associated with type. -
= Place specified line in the center of the window and leave the
current line at this line
count specifies the number of lines to be displayed.
 0 Editor Command Summary 0 585

[address]! command
Execute command in a shell. If address is specified, apply the lines
contained in address as standard input to command, and replace the
lines with the output.
[address]=
Print the line number of the line indicated by address.
[address]< [count]
or [address] > [count]
Shift lines specified by address in specified direction. Only blanks
and tabs are shifted in a left shift (<).
address address
Print the lines specified in address.
RETURN RETURN
Print the next line in the file.
& [address]& [options][count]
Repeat the previous substitute command.
-
[address] [count]
Replace the previous regular expression with the previous replace-
ment pattern from a s u b st it u t e command.

The sed Editor

sed [options]file(s)

The following options are recognized:

-n Only print lines specified with the p command, or the p flag of the
s command
-e cmd Next argument is an editing command
-€file Next argument is a file containing editing commands

All sed commands have the general form:

[address][, address][ !]command [arguments]
The sed editor copies each line of input into a pattern space. sed instructions con-
sist of addresses and editing commands. If the address of the command matches the
line in the pattern space, then the command is applied to that line. If a command has
no address, then it is applied to each input line. It is important to note that a command
affects the contents of the space; subsequent command addresses attempt to match the
line in the pattern space, not the original input line.
586 UNlX Text Processing 0

Pattern Addressing

In a sed command, an address can either be a line number or a pattern, enclosed in

slashes (/pattern/). Address types cannot be mixed when specifying two addresses.
Patterns can make use of regular expressions, as described at the beginning of this
appendix. Additionally, \n can be used to match any newline in the pattern space
(resulting from the N command), but not the newline at the end of the pattern space. If
no pattern is specified, command will be applied to all lines. If only one address is
specified, the command will be applied to all lines between the first and second
addresses, inclusively. Some commands can only accept one address.
The ! operator following a pattern causes sed to apply the command to all
lines that do not contain the pattern.
A series of commands can be grouped after one pattern by enclosing the com-
mand list in curly braces:
[ /pattern / 1[ , /pattern / 1{
command1
command2
1

Alphabetical List of Commands

:label
Specify a label to be branched to by b or t. label may contain up
to eight characters.
[/pattern/]=
Write to standard output the line number of each line addressed by
pattern.
[addressla\
text
Append text following each line matched by address. If text goes
over more than one line, newlines must be “hidden” by preceding
them with a backslash. The insertion will be terminated by the first
newline that is not hidden in this way. The results of this command
are read into the pattern space (creating a multiline pattern space)
and sent to standard output when the list of editing is finished or a
command explicitly prints the pattern space.
b [addressl][, address2]b[Zabell
Branch to label placed with : command. If no Zabel, branch to the
end of the script. That is, skip all subsequent editing commands (up
to label) for each addressed line.
 0 Editor Command Summary 0 587

C [addressl 3 E, address21c \
text
Replace pattern space with text. (See a for details on text.)
d [addressl I [, address2ld
Delete line in pattern space. Thus, line is not passed to standard
output and a new line of input is read; editing resumes with first
command in list.
D [address]][address2]D
Delete first part (up to embedded newline) of multiline pattern
created by N command and begin editing. Same as d if N has not
been applied to a line.
g [address]I [ ,address2lg
Copy contents of hold space (see h or H command) into pattern
space, wiping out previous contents.
G [address11[, address23G
Append contents of hold space (see h or H command) to contents
of the pattern space.
h [addressl][,address21h
Copy pattern space into hold space, a special buffer. Previous con-
tents of hold space are obliterated.
H [addressl][,address21H
Append pattern space to contents of the hold space. Previous and
new contents are separated by a newline.
i [addressl]i \
text
Insert text before each line matched by address. (See a for details
on text.)
n [address] I[, address2111
Read next line of input into pattern space. Current line is output but
control passes to next editing command instead of beginning at the
top of the list.
N [address]][,address21N
Append next input line to contents of pattern space; the two lines are
separated by an embedded newline. (This command is designed to
allow pattern matches across two lines.)
P [addressl][, address2lp
Print the addressed line(s). Unless the -n command-line option is
used, this command will cause duplication of the line in the output.
Also used when commands change flow control (d, N, b).
588 0 UNlX Text Processing 0

P [addressl][,address2lP
Print first part (up to embedded newline) o f multiline pattern created
by N command. Same as p if N has not been applied to a line.

4 [address]q
Quit when address is encountered. The addressed line is first writ-
ten to output, along with any text appended to it by previous a or
r commands.
r [address]r file
Read contents o f file and append after the contents of the pattern
space. Exactly one space must separate the r and the filename.
S [addressl ] [,address21 slpattern/replacement/Vlags]
Substitute replacement for pattern on each addressed line. If pattern
addresses are used, the pattern / / represents the last pattern
address specified. The following flags can be specified:
g Replace all instances o f /pattern/ on each addressed
line, not just the first instance.
p Print the line if a successful substitution is done.
If several successful substitutions are done, multiple copies of
the line will be printed.
w file Write the line to a file if a replacement was
done. A maximum of ten different files can be opened.
t [addressl][,address2lt [label]
Test if successful substitutions have been made on addressed lines,
and i f so, branch to label. (See b and :.) If label is not specified,
drop to bottom of list of editing commands.
W [addressl][,address2lw file
Write contents of pattern space tofile. This action occurs when the
command is encountered rather than when the pattern space is out-
put. Exactly one space must separate the w and the filename. A
maximum of ten different files can be opened.
X [addressl][ ,address21 x
Exchange contents of the pattern space with the contents of the hold
space.

awk
An auk program consists of patterns and procedures:
pattern {procedure ]
 a Editor Command Summary 0 509

Both are optional. If pattern is missing, (procedure) will be applied to all lines. If
{procedure) is missing, the line will be passed unaffected to standard output (Le., it
will be printed as is).
Each input line, or record, is divided into fields by white space (blanks or tabs) or
by some other user-definable record separator. Fields are referred to by the variables
$ 1 , $ 2 , . . . , $n. $ 0 refers to the entire record.

Patterns
Patterns can be specified using regular expressions as described at the beginning of this
appendix.
pattern {procedure ]
The following additional pattern rules can be used in awk:

The special pattern B E G I N allows you to specify procedures that will take
place before the first input line is processed. (Generally, you set global vari-
ables here.)
Interrupt place after the last input line is processed.
and $ can be used to refer to the beginning and end of a field, respectively,
rather than the beginning and end of a line.
A pattern can be a relational expression using any of the operators <, <=,
_- , ! =, >=, and >. For example, $ 2 > $ 1 selects lines for which the
second field is greater than the first. Comparisons can be either string or
numeric.
Patterns can be combined with the Boolean operators I I (or), is& (and), and
! (not).
Patterns can include any of the following predefined variables. For example,
N F > 1 selects records with more than one field.

Special Variables
FS Field separator (blank and tab by default)
RS Record separator (newline by default)
OF S Output field separator (blank by default)
ORS Output record separator (newline by default)
NR Number of current record
NF Number of fields in current record
$0 Entire input record
590 0 UNlX Text Processing 0

$1, $ 2 , ..., $n First, second, . . . nth field in current record, where

fields are separated by F S

Procedures
Procedures consist of one or more commands, functions, or variable assignments,
separated by newlines or semicolons, and contained within curly braces. Commands
fall into four groups:

variable or array assignments

printing commands
built-in functions
control flow commands

Variables and Array Assignments

Variables can be assigned a value with an = sign. For example:
)j.S = \ \ I ,

Expressions using the operators +, -, /, and % (modulo) can be assigned to vari-

ables.
Arrays can be created with the s p l i t function (see following auk commands)
or can be simply named in an assignment statement. ++, +=, and -= are used to
increment or decrement an array, as in the C language. Array elements can be sub-
scripted with numbers (array[ 11, . . . ,array[n]) or with names. (For example, to count
the number of occurrences of a pattern, you could use the following program:
/pattern/{ n["/pattern/"]++ }
END ( p r i n t n["/paftern/"] }

a w k Commands

for
for (i=lower; i<=upper; i++)
command
While the value of variable i is in the range between lower and
upper, do command. A series of commands must be put within
braces. <= or any relational operator can be used; ++ or -- can
be used to decrement variable.
 0 Editor Command Summary 0 591

for i i n array
command
For each occurrence of variable i in array, do command. A series of
commands must be put inside braces.
if i f (condition)
command
[else]
[command]
If condition is true, do command(s), otherwise do command in
e 1 s e clause. condition can be an expression using any of the rela-
tional operators <, <=, ==, !=, >=, or >, as well as the
pattern-matching operator -(e.g., i f $ 1 - .
/ [ A a ] */). A
series of commands must be put within braces.
length x = length(arg)
Return the length of arg. If arg is not supplied, $ 0 is assumed.
log x=log (arg)
Return logarithm of arg.
print p r i n t [args]
Print args on output. args is usually one or more fields, but may
also be one or more of the predefined variables. Literal strings must
be surrounded by quotation marks. Fields are printed in the order
they are listed. If separated by commas in the argument list, they
are separated in the output by the character specified by OFS. If
separated by spaces, they are concatenated in the output.
printf p r i n t f "format, "expression(s)
Formatted print statement. Fields or variables can be formatted
according to instructions in the format argument. The number of
arguments must correspond to the number specified in the format
sections.
Format follows the conventions of the C language's p r i n t f state-
ment. Here are a few of the most common formats:
%n.md a floating point number;
n = total number of digits.
m = number of digits after decimal point.
% [ - ] nc n specifies minimum field length for
format type c. - justifies value in
field; otherwise value is right justified.
Format can also contain embedded escape sequences: \n
(newline) or \ t (tab) are the most common.
Spaces and literal text can be placed in the format argument
by surrounding the entire argument with quotation marks. If
there are multiple expressions to be printed, you should
592 UNlX Text Processing 0

specify multiple formats. An example is worth a thousand

words. For an input file containing only the line:
5 5
The program:
{printf ("The sum on line % s is %d \n", NR, $ 1 + $ 2 ) )

will produce:
The sum on line 1 is 10.

followed by a newline.
split x = s p l i t (string, array [, sepl 1
Split string into elements of array
a r r a y [ 13 , . - . ,a r r a y [ n ] . string is split at each occurrence
of separator sep. If sep is not specified, F S is used. The number
of array elements created is returned.
sprint€ x = s p r i n t f ("format", expression)
Return the value of expression(s), using the specified format (see
p r i n t f).

sqrt x = sqrt(arg)
Return square root of arg.
substr x = s u b s t r (string,m, [ n l )
Return substring of string beginning at character position m and con-
sisting of the next n characters. If n is omitted, include all charac-
ters to the end of string.
while w h i l e (condition)
command
Do command while condition is true (see i f for a description of
allowable conditions). A series of commands must be put within
braces.
 A P P E D X
. w
N
w
I
w

Formatter Command Summary

This appendix is divided into ten subsections, each covering a different facet o f the
n r o f f/tr o f f formatting system. These sections are:

n r o f € Jtr o f f command-line syntax

n ro f f/tro f f requests
escape sequences
predefined number registers
special characters
the ms macro package
the mm macro package
the t b 1 preprocessor
the eqn preprocessor
the pic preprocessor

In the following sections, italics are used for values that you supply. Optional
arguments to requests or macros are enclosed in brackets.

593 .
594 0 UNlX Text Processing 0

n r o f f /troff Command-Line Syntax

n r o f f [options] wiles]

-cname Prepend / u s r / l i b / m a c r o s / c m p . n . [ d t ] .name to files

(old versions of nrof f only).
-e Space words equally on the line instead of in full multiples of the
space character.
-h Use tabs in large spaces.
-i Read standard input after files are processed.
-kname Compact macros and output to [ d t ] . n a m e (old versions of n r o f f
only).
-mname Prepend / u s r / 1ib/ t m a c / tmac.name to files.
-nn Number first page n.
-0 list Print only pages contained in list. Individual pages in list should
be separated by commas; a page range is specified by n-m; n-
indicates from page n to the end.
-q Invoke simultaneous input/output of .r d requests.
-ran Set register a to n.
-sn Stop every n pages.
-Tname Output is for device type name. Values are shown in Table B-1.
(Check your manual for other devices, especially those supported
by the mm command.)
-un Embolden characters by overstriking n times.
-2 Throw away output except messages from t m request.

TABLE B-1. Device Names for n r o f f

Abbreviation Used for

37 TELETYPE Model 37 terminal (default for n ro f f )
450 DASI 450 terminal (default for mm)
tn300 GE TermiNet 300 printer
300 DASI 300 terminal
832 Anderson Jacobson 832 printer
2631 Hewlett-Packard 263 1
4000a Trendata 4000a
8510 C. Itoh printer
1P ASCII line printer
X EBCDIC line printer
 Formatter Command Summary 0 595

t r o f f Options
t r o f f [options] Fles]

-a Send printable ASCII approximation to standard output. ot r-

o f f sends its output directly to a connected typesetter unless the
- t or -a option is specified, in which case it is sent to standard
output. d i t r o f f always writes to standard output.
-b Report phototypesetter status (otr o f f only).
-cname Prepend / u s r / l i b / m a c r o s / c m p .t . [ d t 1 .name t o files
( o t r o f f only).
-f Do not stop the phototypesetter when the formatting run is done
( o t r o f f only).
-Fdir Format output for device name using the font tables in directory dir
instead of / u s r / l i b / f o n t ( d i t r o f f only).
-i Read standard input after files.
-kname Compact macros and output to [dt].name ( o t r o f f only).
-mname Prepend / u s r / l i b / t m a c / t m a c .name tofiles.
-nn Number first page n.
-olist Print only pages contained in list. Individual pages in list should
be separated by commas. A page range is specified by n-m; n-
indicates from page n to the end.
-Pn Print all characters in point size n, but retain motions for sizes
specified in document (ot r o f f only).
-q Do not echo .r d requests.
-ran Assign value n to register a.
-sn Stop every n pages.
-t Send output to standard output instead of directly to the photo-
typesetter ( o t r o f f only).
-Tname Format output for device name using the device description and
font width tables in / u s r / l i b / f o n t / d e v n a m e ( d i t r o f f
only).
-W If the phototypesetter is busy, wait until it is free ( o t r o f f only).

n r o f f / t r o f f Requests

ab [text] Abort and print text as message. If text is not specified, the
message User Abort i s printed.
. a d [c] Adjust one or both margins if filling i s in effect (see .f i).
c can be:
b or n Adjust both margins
C Center all lines
596 UNlX Text Processing 0

1 Adjust left margin only

K Adjust right margin only
.a f rc Assign format c to register r. c can be:
1 0, 1, 2, etc.
001 000, 001, 002, etc.
i Lowercase roman
I Uppercase roman
a Lowercase alphabetic
A Uppercase alphabetic
.am xx y y Append to macro xx; end append at call of y y (default y y =
. .).
.a s xx string Append to string xx.
.bdfn Overstrike characters in fontf, n times.
.b d f s n Overstrike special font s, n times when font f is in effect.
- b p [nl Begin new page. Number next page n.
.br Break to a new line (output partial lines).
.c2 c Set no-break control character to c (default ? ).
.cc c Set control character to c (default .).
.c f file Copy contents of file into output, uninterpreted ( d i t ro f f
only).
.c e [n] Center next n lines; if n is 0, stop centering (default n = 1).
.c h xx [n] Change trap position for macro xx to n. If n is absent,
remove the trap.
.csfnm Use constant character spacing for font f of n/36 ems. If m
is given, the em is taken to be m points.
.cu [n] Continuous underline (including interword spaces) on next
n lines. If n is 0, stop underlining. Italicize in t r o f f .
(See . u l . )
.da [xu] Divert following text, appending it to macro xx. If no argu-
ment, end diversion.
Define macro xx. End definition at .yy (default . y y =
. -).
.di [xx] Divert following text to newly defined macro xx. If no
argument, end diversion.
.ds xx string Define xx to contain string.
.dt n x x Install diversion trap to invoke macro x x at position n .
.e c [c] Set escape character to c (default \).
.e l anything Else portion of i f - e l s e . See - i e .
.e m xx Set end macro to xx.
.eo Turn escape character mechanism off. See .e c.
.ev [n] Change environment to n. If no argument, restore previous
environment (OIn12 = initial value 0).
.ex Exit from formatter.
.fcab Set field delimiter to a and pad character to b.
 0 Formatter Command Summary 0 597

.f i Turn on fill mode (default: fill is on).

.fl Flush output buffer.
-fPnf Assign font f to position n.
. f tf Change font tof.
.hc[c] Change hyphenation-indication character used with .hw to
c (default -).
.hw words Specify hyphenation points for words (e.g., .hw spe-
c i-f y ) .
.h y n Turn hyphenation on ( n 2 l ) or off (n=O).
n= 1 Hyphenate whenever necessary
n=2 Don't hyphenate last word in page or diversion
n=4 Don't split off first two characters of word
n=8 Don't split off last two characters of word
n=14 Use all three restrictions
.i e c anything If portion of if-else. See .el.
.if ! c anything If condition c is false, do anything.
.if n anything If expression n >o, do anything.
.if ! n anything If expression n SQ, do anything.
.i f ' stringl ' string2 ' anything
If stringl and string2 are identical, do anything.
i f ! ' stringl ' string2 anything
If stringl and string2 are not identical, do anything.
-ig YY .
Ignore following text, up to line beginning with y y .
.i n [+][n] Set indent to n or increment indent by +n. If no argument,
restore previous indent.
.it nxx Set input line count trap to invoke macro xx after n lines of
input text have been read.
.IC c Set leader repetition character to c. (See .tc.) Leaders are
invoked by \a.
.lg n Turn ligature mode on if n is absent or nonzero.
- 1 1[+][n] Set line length to n or increment line length by +n. If no
argument, restore previous line length (default 6.5 inches).
.Is n Set line spacing to n. If no argument, restore previous line
spacing (initial value 1 ) .
.It n Set title length to n. If no argument, restore previous value.
- m c IC1 [nl Set margin character to c, and place it n spaces to the right
of margin. If c is missing, turn margin character off.
Default for n is 0 . 2 inches in nrof f and 1 em in
trof f .
.mk [r] Mark current vertical place in register r. Return to mark
with .rt, or .s p I \nr.
.n a Do not adjust margins. (See .ad.)
.ne n If n lines do not remain on this page, start new page.
.nf No filling or adjusting of output lines. (See .ad and
.fi.)
598 0 UNIX Text Processing 0

.nh Turn hyphenation off. (See .hy.)

.nm [n m s i] Number output lines (n20) or turn numbering off (n=O). f n
sets initial line number: m sets numbering interval; s sets
separation of numbers and text; i sets indent of text.
.nn n Do not number next n lines, but keep track of numbering
sequence, which can be resumed with .nm+ 0.
.n r r n [m] Assign the value n to number register r and optionally set
autoincrement to m.
- ns Turn no-space mode on. (See .rs.)
.n x file Switch tofile and do not return to current file. (See .so.)
.os Output saved space specified in previous .sv request.
.pc c Set page number character to c.
. p i cmd Pipe output of t r o f f to cmd instead of to standard out-
put.
Set page length to n or increment page length by +n. If no
argument, restore default (default 1 1 inches).
Print names and sizes of all defined macros.
Set next page number to n, or increment page number by
b.
Offset text a distance of n from left edge of page, or incre-
ment the current offset by +n. If no argument, restore pre-
vious offset.
.ps n Set point size to n (t r o f f only). (Default 10 points.)
-r d [prompt] Read input from terminal, after printing optional prompt.
.r m xx Remove macro or string XT.
.r n xx yy Rename request, macro, or string xx to y y .
.rr r Remove register r.
.rsxxyy Restore spacing. (Turn no-space mode off; see .n s .)
.r t [b] Return (upward only) to marked vertical place, or to f n
from top of page or diversion. (See .mk.)
.so file Switch out tofile, then return to current file. (See .nx.)
- sp n Leave n blank lines (default 1).
.ss n Space character size set to n/36 em (no effect in n r o f f).
.sv n Save n lines of space; output such space with .o s .
.sy cmd [args] Execute UNIX command cmd with optional arguments
( d i t r o f f only).
.ta n[t] m [ t ] Set tab stop at positions n, m, etc. If t is not given, tab is
left adjusting; if t is:
R Right adjust
C Center
.tc c Define tab character as c (e.g., .t c . will draw a string of
dots to tab position).
.ti [k][n] Indent next output line n spaces, or increment the current
indent by f n for the next output line.
 0 Formatter Command Summary 0 599

.t l ''I c' r' Specify left (I), centered (c), right ( r ) title.
.t m text Terminal message. (Print text on standard error.)
.tr ab Translate character a to b.
.uf f Underline font set to f (to be switched to by - u l ) .
.u 1 [n] Underline (italicize in t r o f f ) next n input lines. Do not
underline interword spaces.
.vs [n] Set vertical line spacing to n. If no argument, restore previ-
ous spacing (default 1/6 inch in n r o f f , 12 points in
troff).
.whnxx When position n is reached, execute macro xr; negative
values of II are with respect to page bottom.

. Escapesequences
\ To prevent or delay the interpretation of \ .
\e Printable version of the current escape character.
Y ' (acute accent); equivalent to yaa.
Y ' (grave accent); equivalent to yga.
\- - Minus sign in the current font.
\. Period (dot). (See de.)
\ (space) Unpaddable space-size space character.
\O Digit width space.
\I 1/6-em narrow space character (zero width in n r o f f).
\A 1/12-em half-narrow space character (zero width in
n r o f f).
\& Nonprinting, zero-width character.
\! Transparent line indicator.
\" Beginning of comment.
\\SN Interpolate argument 9.
\% Default optional hyphenation character.
\ (xx Character named xx.
\*x, \ * ( x u Interpolate string x or xr.
\a Noninterpreted leader character for use in macros.
\ b 'abc...' Bracket building function-stack abc... vertically.
\C Interrupt text processing.
\d Downward 1/2-em vertical motion (1/2line in n r o f f).
\D '1 X , y ' Draw a line from current position to coordinates x y
( d i t r o f f only).
\D'c d' Draw circle of diameter d with left edge at current posi-
tion ( d i t r o f f only).
600 0 UNlX Text Processing 0

\ D ' e dl d2' Draw ellipse with horizontal diameter dl and vertical

diameter d2, with left edge at current position ( d i t r -
off only).
\D 'a x l y l x2 y2' Draw arc counterclockwise from current position, with
center at x l ,yl and endpoint at XI +x2 ,yl +y2 ( d i t r o f f
only).
\D ' - XI y l x2 y x . . .' Draw spline from current position through the specified
coordinates (ditroff only).
\ fx,\ f (AX,\ fN Change to font named x or xx or to position N.
\h'N' Local horizontal motion; move right N (negative left).
\H 'n' Set character height to n points, without changing width
( d i t r o f f only).
\jx Mark horizontal place on output line in register x.
\kx Mark horizontal place on input line in register x .
\ 1'Nc' Horizontal line drawing function (optionally with c,
default - ).
\L'Nc' Vertical line drawing function (optionally with c, default
1).
\nx,\n ( x x Interpolate number register x or xx.
\ o'abc ...' Overstrike characters a, b, c...
\P Break and spread output line.
\r Reverse 1-em vertical motion (reverse line in n r o f f).
\sN,\sW Point-size change function.
\ S 'n' Slant output n degrees to the right ( d i t r o f f only).
Negative values slant to the left. A value of zero turns
off slanting.
\t Noninterpreted horizontal tab.
\u Reverse (up) 1/2-em vertical motion (1/2 line in
n r o f f).
\v 'N' Local vertical motion; move down N (negative up).
\w 'string Interpolate width of string.
\x'N' Extra line-space function (negative before, positive after).
\ zc Print c with zero width (without spacing).
\I Begin conditional input.
\} End conditional input.
vnewline) Concealed (ignored) newline.
\x X, any character not listed above.
 I
0 Formatter Command Summary 601

Predefined Number Registers

Read-only Registers

.$ Number of arguments available at the current macro level.

-$$ Process ID of t r o f f process ( d i t r o f f only).
.A Set to 1 in t r o f f , if -a option used; always 1 in n r o f f .
.H Available horizontal resolution in basic units.
.T In n r o f f , set to 1 if -T option used; in t r o f f , always 0; in
d i t r o f f , you can print the value of -T with the string \ * ( .T.
.v Available vertical resolution in basic units.
.a Extra line space most recently utilized using \x’N’.
-c Number of lines read from current input file.
.d Current vertical place in current diversion; equal to n l if no diver-
sion.
.f Current font as physical quadrant ( 1 to 4 in o t r o f f ; no limit in
d i t r o f f).
.h Text baseline high-water mark on current page or diversion.
.i Current indent.
-j Current adjustment type (O=. ad1 or .na; 1=. adb; 3=. adc;
5=. adr).
.1 Current line length.
.n Length of text portion on previous output line.
-0 Current page offset.
-P Current page length.
.s Current point size.
.t Distance to the next trap.
.u Equal to 1 in fill mode and 0 in no-fill mode.
.v Current vertical line spacing.
.w Width of previous character.
.x Reserved version-dependent register.
-Y Reserved version-dependent register.
.z Name of current diversion.

ReadJWrite Registers
9- Current page number.
ct Character type (set by width function).
dl Width (maximum) of last completed diversion.
dn Height (vertical size) of last completed diversion.
dw Current day of the week ( 1 to 7).
602 0 UNlX Text Processing 0

dY Current day of the month (1 to 3 1).

hP Current horizontal place on input line.
In Output line number.
mo Current month ( 1 to 12).
nl Vertical position of last printed text baseline.
sb Depth of string below baseline (generated by w i d t h function).
st Height of string above baseline (generated by w i d t h function).
Yr Last two digits of current year.

Special Characters

On the Standard Fonts

The following special characters are usually found on the standard fonts:

close quote fi fi ligature

open quote fl fl ligature
314 em dash ff E ligature
hyphen ffi ffi ligature
hyphen ffl ffl ligature
current font minus sign degree
bullet t dagger
square I
foot mark
rule t cent sign
114 @ registered trademark
112 0 copyright
314

On the Special Font

The following characters are usually found on the special font except for the uppercase
Greek letter names followed by t which are mapped into uppercase English letters in
whatever font is mounted on font position one (default is Times Roman).
 0 Formatter Command Summary 0 603

Miscellaneous Characters

8 \ (sc section 1 \ (da downarrow

' \ (aa acute accent 1 \(br boxrule
' \ (ga grave accent 4 \ (dd double dagger
- \ (ul underrule v \ (rh righthand
+ \ (-> right arrow a \ (lh left hand
t \ (<- left arrow 0 \ (ci circle
t \(ua uparrow

Mathematic Symbols

+ \(Pl math plus U cup (union)

- \(mi math minus n cap (intersection)
= \(eq math equals c subset of
* \(** math star I> superset of
f \(sl slash (matching backslash) G improper subset
4
- \(sr square root a improper superset
\ (rn root en extender m infinity
2 \(>= greater than or equal to a partial derivative
I \(<= less than or equal to V gradient
-
= \ ( = = identically equal 1 not
# \(-= approx equal I integral sign
- \(ap approximates oc proportional to
# \(!= not equal 0 empty set
x \(mu multiply E member of
+ \(di divide I or
* \(+- plus-minus

Bracket Building Symbols

\ (It left top o f large curly bracket

\ (lk left center of large curly bracket
\ (lb left bottom o f large curly bracket
\ (rt right top o f large curly bracket
\ (rk right center o f large curly bracket
\ (rb right bottom o f large curly bracket
\ (IC left ceiling (top) of large square bracket
\ (bv bold vertical
\ (If left floor (bottom) o f large square bracket
\ (rc right ceiling (top) o f large square bracket
\ (rf right floor (bottom) o f large square bracket
604 0 UNlX Text Processing 0

Greek Characters

alpha A \(*A Alphat

beta B \(*B Beta?
gamma r \(*G Gamma
delta A \(*D Delta
epsilon E \(*E Epsilon?
zeta z \(*Z Zeta?
eta H \(*Y Eta?
theta 0 \(*H Theta
iota 1 \ ( * I Iota?
kappa K \(*K Kappa?
lambda A \ ( * L Lambda
mu M \(*M Mu?
nu N \(*N Nu?
xi E \(*C Xi
omicron 0 \(*O Omicront
Pi n \(*P Pi
rho p \(*R Rho?
sigma z \(*S Sigma
terminal sigma
tau T \ (*T Taut
upsilon Y \ ( * u Upsilon
phi Q, \ ( * F Phi
chi X \ (*x Chit
psi Y \ ( * Q Psi
omega R \ (*W Omega

Thems Macros

Summary of ms Macros

- 1c Return to single-column format.

- 2c Start two-column format.
.AB Begin abstract.
.AE End abstract.
.AI name Name of author’s institution (used in cover sheet).
.AU name Author’s name (used in cover sheet).
.B [text] Print text in boldface. If text is missing, equivalent to .f t 3.
 Formatter Command Summary 0 605

.B1 Enclose following text in a box.

.B2 End boxed text.
.BX word Surround word in a box.
.DA Print date on each page.
.DS Start displayed text.
.DSB Start left-justified block, centered.
.DSC Start centered display.
.DSL Start left-centered display.
-DE End displayed text.
-EQ Begin equation.
.EN End equation.
.FS Start footnote.
.FE End footnote.
- I [text] Print text in italics. If text is missing, equivalent to .ft 2.
.IP label n Indent paragraph n spaces with hanging label.
.KS Start keep.
.KE End of keep or floating keep.
.KF Begin floating keep.
.L G Increase type size by two points ( t r o f f only).
.LP Start block paragraph.
.N D Change or omit date.
.NH n Numbered section heading, level n.
.NL Restore default type size ( t r o f f only).
.PP Start indented paragraph.
.R [text] Print text in roman. If text is missing, equivalent to . f t 1 .
.RP Initiate title page for a “released paper.”
.RS Increase relative indent one level. Use with IP.
.RE End one level of relative indent.
.SG Signature line.
.SH Unnumbered section heading.
SM Decrease type size by two points (tr o f f only).
.T L Title line.
.TS [HI Start table. H will put table header on all pages. Use this
option with following .TH.
.TH Table header ends. Must be used with .T S H.
.TE End table.
.U L Underline following text, even in t r o f f.
606 UNlX Text Processing 0

Internal Macros Worth Knowing About

.IZ Basic initialization; executed automatically before any text is

processed. It is then removed, and cannot be invoked again.
.RT Reset. Invoked by all paragraph macros, plus . R S , .RE,
.TS, .TE, SH, and .NH. Resets various values to defaults
I

stored in number registers listed below.

BG Prints cover sheet, if any. Also performs some special first page
initialization. Invoked once by the very first . R T in a docu-
ment.
I NP New page. Invoked at the top of each page. Performs various
page top resets, and calls .PT.
.P T Page titles. Contains running headers. Can be redefined.
Invoked by .NP at \n ( HMu from the top of the page.
.B T Bottom titles. Continuous running footers. Invoked by trap at
\ n ( FMu / 2 u from the bottom of the page.
.F O Footer. The bottom of the text on the page. Invoked by trap at
\n (FMu.

Number Registers Containing Page Layout Defaults

cw Column width (default 7/15 of line length).

FL Footnote length (default 11/12 of line length).
FM Bottom margin (default 1 inch).
GW Intercolumn gap width for multiple columns (default 1/15 of line
length).
HM Top margin (default 1 inch).
LL Line length (default 6 inches).
LT Title length (default 6 inches).
PD Paragraph spacing (default 0.3 of vertical spacing).
PI Paragraph indent (default 5 ens).
PO Page offset (default 26/27 inches).
PS Point size (default 10 points).
vs Vertical line spacing (default 12 points).
 0 Formatter Command Summary 0 607

Predefined and User-Definable Strings

DY The current date.

LH Left header, printed by .
t l ' \ \ * (LH' \\*CH' \ \ * (RH' in
P T macro. Null unless user-defined.
CH Center header, printed by . t l ' \ \ * (LH'\\*CH'\\*(RH'
in PT macro. Null unless user-defined.
RH Right header, printed by .t l ' \ \ * (LH' \\*CH' \ \ * (RH' in
P T macro. Null unless user-defined.
LF Left footer, printed by . t l ' \ \ * (LH'\\*CH'\\ (RH' in
BT macro. Null unless user-defined.
CF Center footer, printed by . t l '\\*(LH'\\*CH'\\RH' in
BT macro. Contains PN by default.
RF Right footer, printed by .t l ' \ \ * (LH' \\*cH' \\RH' in
BT macro. Contains PN by default.

Reserved Macro and String Names

The following macro and string names are used by the ms package. Avoid using these
names for compatibility with the existing macros. An italicized n means that the name
contains a numeral (generally the interpolated value of a number register).

I AX DA FL KJ OD RT TR
.I B DW FN KS OK so TS
B1 DY FO LB PP s2 TT
1. B2 EE FS LG PT s3 TX
[C BB EG FV LP PY SG UL
[O BG EL FX LT QE SH us
h
BT EM FY MC QF SM ux
1
BX EN HO ME QP SN WB
- C En I MF QS SY WH
1c c1 EQ IE MH R TA WT
2c c2 EZ IH MN R3 TC XF
AB CA FA IM MO RA TD XK
AE cc FE In MR RC TE XP
AI CF FF IP ND RE TH
An CH FG I Z NH Rn TL
AT CM FJ KD NL RP TM
AU CT FK KF NP RS TQ

The following number register names are used by the m s package. An italicized
n means that the name contains a numeral (generally the interpolated value of another
number register).
608 UNlX Text Processing 0

#T EF H5 IX MF OJ QP TV
AJ FC HM I# MG PD RO TY
AV FL HT J# ML PE SJ TZ
BC FM IO KG MM PF ST vs
BD FP IF KI MN PI T. WF
BE GA IK KM NA PN TB xx
BH GW IM L1 NC PO TC YE
BI HI IP LE ND PQ TD YY
BQ H2 IR LL NQ PS TK ZN
BW H3 IS LT NS PX TN
cw H4 IT MC NX QI TQ

Note that with the exception of [ c and [ 0,none of the number register, macro, or
string names contain lowercase letters, so lowercase or mixed case names are a safe bet
when you’re writing your own macros.

T h e m Macros

Summary of mm Macros

.1c Return to single-column format.

-2c Start two-column format.
-AS [xl[nl Start abstract type x, indent n spaces. (Used with .TM and
.RP only.) (Types: l=abstract on cover sheet and first page:
2=abstract only on cover sheet; 3=abstract only on Memorandum
for File cover sheet.) End with .AE.
.AE End abstract. Begin with .AS.
.AF [company name]
Alternate format for first page. Change first page
“Subject/Date/From” format. If argument i s given, other head-
ings are not affected. No argument suppresses company name
and headings.
-AL[xl[nl Start list type x ( 1 , A, a, I, or i), indent n spaces. If third argu-
ment is 1, don’t put a blank line between items. Default i s num-
bered listing, indented 5 spaces.
.AT title Author’s title follows.
.AU name Author’s name and other information follows.
.AV name Approval signature line for name.
.B [w][XI. . . Set w in bold (underline in nrof f ) and x in previous font; up
to six arguments.
.BS Begin block of text to be printed at bottom of page, after foot-
notes (if any), but before footer.
 o Formatter Command Summary 0 609

.BE End bottom block and print after footnotes (if any), but before
footer.
- BI [wl [XI Set w in bold (underline in n r o f f ) and x in italics; up to 6
arguments.
* B L [nl 111 Start bullet list and indent text n spaces. If second argument is
1, don’t put a blank line between items.
- B R[WI txl Set w in bold (underline in nrof f) and x in roman; up to six
arguments.
. C S [pgs] [other] [tot] F g s ] [tbls] [ref3
Cover sheet numbering information.
- D F bf bl Start floating display of type x and mode y, with indent a.
(Default i s no indent, no-fill mode.) End with . D E . x is: L
(no indent), 1 (indent standard amount), C (center each line
individually), or CB (center as a block). y is: N (no-fill mode)
or F (fill mode).
.D S Start floating or static display of type x and mode y, with indent
n. Type and mode are as in .DF. End with .DE.
.DE End floating or static display started with .D S or .D F .
. D L [ H I [I1 Start dashed list and indent text n spaces. If second argument i s
1, no space between items.
.EC [caption] [n] v]
Equation caption. Arguments optionally override default
numbering, where flag f determines use of number n. If e 0
(default), n is a prefix to number; if f=1, n is a suffix; if 6 2 , n
replaces number.
. E F [text] Print text as the footer on all even pages. text has the format:
’left’ center ’right’.
.EH [text] Print text as the heading on all even pages. text has the format:
’left’ center ’right’.
. EQ [text] Start equation display using text as label.
. EN End equation display.
. E X [caption] [ n ] v]
Exhibit caption. Arguments optionally override default number-
ing, where flag f determines use of number n. I f e O (default), n
is a prefix to number; if el,a is a suffix; if e 2 , n replaces
number.
.FC [text] Use text for formal closing.
.FD [O-II] Setup default footnote format.
.FS [c] Start footnote using c for indicator. Default i s numbered foot-
note.
.F E End footnote.
F G [title] Figure title follows.
.Hn [heading] Numbered heading level n follows.
.HC [ c ] Use c as hyphenation indicator.
610 UNlX Text Processing 0

.H M [mark] Heading mark style follows arabic ( I or OOl), roman (i or I ) ,

or alphabetic (a or A).
.HU heading Unnumbered heading follows.
.H X User-supplied exit macro before printing heading.
.HY User-supplied exit macro in middle of printing heading.
.HZ User-supplied macro after heading.
- 1 [wl [XI Set w in italics (underline in n r o f f) and x in previous font.
Up to six arguments.
- IB [wl 1x1 Set w in italics (underline in n r o f f ) and x in bold. Up to
six arguments.
-IR IWI [XI Set w in italics (underline in n r o f f ) and x in roman.
.L B n m pad type [mark] [LI-space] [LB-space]
List beginning. Allows complete control over list format. It
takes the following arguments:
n - Text indent.
m - Mark indent.
pad - Padding associated with mark.
type - If 0, use the specified mark. If nonzero, and mark
i s 1, A , a, I, i, list will be automatically numbered or
alphabetically numbered or alphabetically sequenced. In
this case, type controls how the mark will be displayed.
For example, if mark is 1, type will have the following
results:
Type Format

mark - The symbol or text that will be used to start each

list entry. mark can be null (creates hanging indent), a text
string, or 1, A, a, I, or i to create an automatically num-
bered or lettered list. Format of the mark will be affected
by type.
LI-space - The number of blank lines to be output
between each following .L I macro (default 1).
LB-space - The number of blank lines to be output by the
L B macro itself (default 0).
.LC [n] Clear list level n.
.L E End list.
. L I [mark] Item in list and specify mark.
.ML mark [n] [ I ]
Start marked list, indent n spaces. If third argument is 1 , no
space between items in list.
 Formatter Command Summary 0 61 1

.MT [type] [title]

Specify memorandum type and title. type is:
11 11 = No type

0 = No type
1 = Memorandum for file (default)
2 = Programmer?s notes
3 = Engineer?s notes
4 = Released paper
5 = External letter
string = string is printed.
title i s user-supplied text prefixed to page number.
ND date New date. Change date to date.
nP Double-line indent on paragraph start.
.NS [type] Notation start. Specify notation type. type is:
11 I1 = copy to

0 = copy to
1 = Copy (with att.) to
2 = Copy (no att.) to
3 = Att.
4 = Atts.
5 = Enc.
6 = Encs.
7 = Under Separate Cover
8 = Letter to
9 = Memorandum to
1 0 = Copy (with atts.) to
1 1 = Copy (without atts.) to
1 2 = Abstract Only to
1 3 = Complete Memorandum to
string = Copy string to
- NE Notation end.
.OF [text] Print text as the footer on all odd pages. text has the format:
?left?center ?right?.
.OH [text] Print text as the heading on all odd pages. text has the format:
?left?center ?right?.
.OK [topic] Other keywords. Specify topic for TM cover sheet.
.OP Force an odd page.
.p [type] Start paragraph type. type is: 0 = left justified (default), 1 =
indented, 2 = indented except after .H, .LC, .DE.
.P F [text] Print text as the page footer on all pages. text has the format:
?left?center ?right?.
P H [text] Print text as the page heading on all pages. text has the format:
?left?center ?right?.
.PM [type] Proprietary marking on each page (type: P=PRIVATE;
N=NOTICE).
612 0 UNlX Text Processing 0

*PX Page-heading user exit.

.R Return to roman font (end underlining in nrof f).
- R B [WI [XI Set w in roman and x in bold.
.RD [input] Read input from terminal.
.RI [WI 1x1 Set w in roman and x in italics.
. R S [arg] Start automatically numbered reference. arg manually specified
reference number.
.RF End of reference text.
- RL [nl 111 Start reference listing, indent text n spaces. If second argument
is 1, no space between list items.
.RP Produce reference page.
- s [nl [ml Set point size to n and vertical spacing to m (trof f only)
(defaults: 10 or 12). Alternatively, either argument can be speci-
fied as +n/m to increment/decrement current value, D to use
default, C to use current value, P to use previous value.
.SA [n] Set right margin justification to n. n is: 0 = no justification or
1 = justification. (Defaults: no justification for nrof f,justifi-
cation for t rof f .)
.S G [name] Use name for signature line.
.S K n Skip n pages.
- SM xLYlIz1 Reduce string x by one point. If strings x,y, and z are specified,
y is reduced by one point.
.SP [n] Leave n blank vertical spaces.
.TB [title][n]W Supply table title. Arguments optionally override default
numbering, where flag f determines use of number n. If 6 0
(default), n is a prefix to number; i f e l , n is a suffix; i f f i 2 , n
replaces number.
. T S [HI Start table. H will put table header on all pages. Use this option
with following .T H .
.TH N Table header ends. Must be used with TS H . N = only print
I

table headers on new page.

.TE End table.
.T C [level] [level] [tab] [head]]. . .
Generate table of contents.
.TL Title of memorandum follows on next line.
.TM [n] Number a technical memorandum n. (Up to nine may be speci-
fied.)
.TP Top-of-page macro.
.TX User-supplied exit for table-of-contents titles.
.T Y User-supplied exit for table-of-contents header.
. V L n [m] [ I ] Start variable item list. Indent text n spaces and mark m spaces.
If third argument is 1, no space between list items.
.VM [n] [m] Add n lines to top margin and m lines to bottom.
 0 Formatter Command Summary 0 613

.wc [XI Change column or footnote width to x. x is:

FF All footnotes same as first
-FF Turn off FF mode
N Normal default mode
WD Wide displays
-WD Use default column mode
WF Wide footnotes
-WF Turn off W F mode

Predefined String Names

BU Bullet; same as \ (bu.

C i List of indents for table of contents levels.
DT Current date, unless overridden. Month, day, year (e.g., July 28,
1986).
EM Em dash string (em dash in t r o f f and a double hyphen in
n r o f f).
F Footnote number generator.
HF Fonts used for each level of heading (l=roman, 2=italic, 3=bold).
HP Point size used for each level of heading.
Le Title set for List Of Equations.
Lf Title set for List Of Figures.
Lt Title set for List Of Tables.
Lx Title set for List Of Exhibits.
RE SCCS Release and Level of mm.
Rf Reference number generator.
RP Title for references.
TM Trademark string. Places the letters TM one-half line above the
text that it follows.

Number Registers Used in mm

A dagger (t)next to a register name indicates that the register can only be set from the
command line or before the mm macro definitions are read by the formatter. Any regis-
ter that has a single-character name can be set from the command line.
614 UNlX Text Processing 0

At If set to 1 , omits technical memorandum headings and provides

spaces appropriate for letterhead. See .A F macro.
Au Inhibits author information on first page. See .AU macro.
ct Flag indicating type of copy (original, draft, etc.).
c1 Level of headings saved for table of contents (default 2). See
TC macro.
CP If set to 1, list of figures and tables appear on same page as table
of contents. Otherwise, they start on a new page. (Default is 1.)
Dt If set to 1, sets debug mode (default 0). If set, mm will continue
even when it encounters normally fatal errors.
De If set to 1 , ejects page after each floating display. (Default is 0.)
Df Format of floating displays. See .D F macro.
Ds Sets the pre- and post-space used for static displays.
Et Font for the Subject/Date/From: O=bold; l=roman. (Default is
0.1
Ec Equation counter, incremented for each .EC macro.
E j Heading level for page eject before headings. (Default is 0, no
eject.)
Eq If set to 1, places equation label at left margin. (Default is 0.)
Ex Exhibit counter, incremented for each .EX macro.
Fg Figure counter, incremented for each .F G macro.
Fs Vertical spacing between footnotes.
H1-H7 Heading counters for levels 1-7, incremented by the .H macro
of corresponding level or the .HU macro if at level given by the
Hu register. The H2-H7 registers are reset to 0 by any .H (or
.HU) macro at a lower-numbered level.
Hb Level of heading for which break occurs before output of body
text (default 2 lines).
Hc Level of heading for which centering occurs (default 0).
Hi Indent type after heading. (Default 1 =paragraph indent.) Legal
values are: 0 left justified, 1 indented, 2 indented except after
.H, .LC, -DE. (Default is 0.)
Hs Level of heading for which space after heading occurs. (Default
= 2; .H2.)
Ht Numbering type of heading: single ( 1 ) or concatenated (0).
(Default is 0.)
Hu Sets level of numbered heading that unnumbered heading resem-
bles. (Default = 2; .H2.)
HY Sets hyphenation. If set to 1 , H y enables hyphenation. (Default
is 0.)
Lt Sets length of page. (Default is 66v.)
Le Flag for list of equations following table of contents. 0 = do not
print; 1 = print. (Default is 0.)
 Formatter Command Summary 0 61 5

Lf Flag for list of figures following table of contents. 0 = do not

print; 1 = print. (Default is 0.)
Li Default indent of lists. (Default is 5.)
Ls List spacing between items by level. (Default = 6, spacing
between all levels of list.)
Lt Flag for list of tables following table of contents. 0 = do not
print; 1 = print (Default is 0.)
Lx Flag for list of exhibits following table of contents. 0 = do not
print; 1 = print {Default is 0.)
Nt Page numbering style. O=header on all pages; l=header printed
as footer on page 1 ; 2=no header on page 1 ; 3=section page as
footer; 4=no header unless .PH defined; 5=section page and
section figure as footer. (Default i s 0.)
NP Numbering style for paragraphs. 0 = unnumbered; 1 = num-
bered.
0 Offset of page. For n r o f f , this value is an unscaled number
representing character positions. Default i s 9 (7.5). For
t r o f f , this value is scaled. Default is S i .
oc Table of contents page numbering style. O=lowercase roman;
l=arabic. (Default is 0.)
Of Figure caption style. O=period separator; 1 =hyphen separator.
(Default is 0.)
P Current page number.
Pi Amount of indent for paragraph. (Default i s 5 for n r o f f , 3
for t r o f f . )
Ps Amount of spacing between paragraphs. (Default is 3v.)
Pt Paragraph type. Legal values are: 0 left justified, 1 indented, 2
indented except after .H, .LC, .DE. (Default is 0.)
Pv Inhibits “PRIVATE” header. See .P V macro for values.
Rf Reference counter, incremented for each .R S .
St Default point size for trof f. Default is 10. (Vertical spacing
is \n5+2.)
Si Standard indent for displays. (Default is 5 for n r o f f , 3 for
troff.)
Tt Type of n r o f f output device. Causes register settings for
specific devices.
Tb Table counter, incremented for each .TB.
U” Underlying style ( n r o f f ) for .H and -HU. If not set, use
continuous underline; otherwise, don’t underline punctuation and
white space. (Default is 0.)
wt Width of page (line and title length). (Default is 6 i . )
616 UNlX Text Processing 0

Other Reserved Macro and String Names

In mm, the only macro and string names you can safely use are names consisting of a
single lowercase letter, or two character names whose first character is a lowercase
letter and whose second character is anything but a lowercase letter. Of these, c2 and
nP are already used.

t b l Command Characters and Words

.TS Start table.

.TE End table.
.TS H Used when the table will continue onto more than one page.
Used with . T H to define a header that will print on every
page.
.TH With .T S H, ends the header portion of the table.
.T& Continue table after changing format line.

Options
Options affect the entire table. The options should be separated by commas, and the
option line must be terminated by a semicolon.

center Center with current margins.

expand Flush with current right and left margins.
(blank) Flush with current left margin (default).
box Enclose table in a box.
doublebox Enclose table in two boxes.
allbox Enclose each table entry in a box.
t a b (x) Define the tab symbol as x .
linesize ( n ) Set lines or rules (e.g., from box) to n point type.
delim (xy) Recognize x and y as the eqn delimiters.
 Formatter Command Summary 0 617

Format
The format line affects the layout of individual columns and rows of the table. Each
line contains a key letter for each column of the table. The column entries should be
separated by spaces, and the format section must be terminated by a period. Each line
of format corresponds to one line of the table, except for the last, which corresponds to
all following lines up to the next .T & , if any.

Key letters

C Center.
1 Left justify.
r Right justify.
n Align numerical entries.
a Align alphabetic subcolumns.
S Horizontally span previous column entry across this column.
h
Vertically continue entry from previous row down through this
row.

Other choices (must follow a key letter)

b Boldface. Must be followed by a space.

i Italics. Must be followed by a space.
Pn Point size n.
t Begin any corresponding vertically spanned table entry at the
top line of its range.
Equal width columns.
Minimum column width. Also used with text blocks. n can
be given in any acceptable t r o f f units.
VI1 Vertical line spacing. Used only with text blocks.
n Amount of separation between coIumns (default is 3n).
I Single vertical line. Typed between key letters.
II Double vertical line. Typed between key letters.
Single horizontal line. Used in place of a key letter.
Double horizontal line. Used in place of a key letter.
618 UNlX Text Processing 0

Data
The data portion includes both the heading and text of the table. Each table entry must
be separated by a tab symbol.

.xx trof f commands may be used (such as .s p #

and .ce
#). Do not use macros, unless you know what you’re doing.
\ A s last character in a line, combine following line with current
line (\ is hidden).
\ n Vertically spanned table entry. Span table entry immediately
above over this row.
- or = A s the only character in a line, extend a single or double hor-
izontal line the full width of the table.
\$ - or \ $ = Extend a single or double horizontal line the full width of the
column.
\- Extend a single horizontal line the width of the contents of the
column.
\Rx Print x’s as wide as the contents of the column.
...T{ Start text block as a table entry. Must be used with wn,
column width option.
...T 1 End text block.

eqn Command Characters

- EQ Start typesetting mathematics

.EN End typesetting mathematics

Character Translations
The following character sequences are recognized and translated as shown.
 0 Formatter Command Summary 619

>= approx
<= nothing
_
___ cdot
I= times X
+- de 1 V
-> grad V
<- -.. ...
<< I - - - I ,... ,
>> sum
inf
partial
int
prod
r”n
half union
prime
v
inter n
Digits, parentheses, brackets, punctuation marks, and the following words are converted
to roman font when encountered:

s i n cos t a n s i n h c o s h t a n h a r c
max min l i n l o g I n e x p
R e I m a n d if f o r d e t

Greek letters can be printed in uppercase or lowercase. To obtain Greek letters,

simply spell them out in the case you want:

alpha a sigma
beta P tau
gamma Y u p s i 1o n
delta 6 phi
epsilon E chi
zeta
eta
r psi
omega
rl
theta e GAMMA
iota 1 DELTA
kappa K THETA
lambda h LAMBDA
mu P XI
nu V PI
xi 5 SIGMA
omicron 0 UPSILON
Pi 7t PHI
rho P PSI
OMEGA

The following words translate to marks on the tops of characters.

620 o UNlX Text Processing 0

x dot X x vec
..
x dotdot X x dyad
x hat 13 x bar
x tilde
-
X x under

Words Recognized By eqn

above Separate the pieces of a pile or matrix column.

back n Move backwards horizontally n 1/1OO’s of an em.
bold Change to bold font.
ccol Center a column of a matrix.
col??? Used with a preceding 1 or r to left or right adjust the columns
of the matrix.
cpile Make a centered pile (same as pile).
define Create a name for a frequently used string.
delim Define two characters to mark the left and right ends of an eqn
equation to be printed in line.
down n Move down n 1/1OO’s of an em.
fat Widen the current font by overstriking it.
font x Change to font x, where x is the one-character name or the number
of a font.
from Used in summations, integrals, and similar constructions to signify
the lower limit.
fwd n Move forward horizontally n 1/1OO’s of an em.
gfont x Set a global font x for all equations.
gsize n Set a global size for all equations.
UP n Move up n 1/1OO’s of an em.
italic Change to italic font.
lcol Left justify a column of a matrix.
left Create large brackets, braces, bars, etc.
lineup Line up marks in equations on different lines.
lpile Left justify the elements of a pile.
mark Remember the horizontal position in an equation. Used with
lineup.
matrix Create a matrix.
ndef ine Create a definition which only takes effect when neqn is running.
over Make a fraction.
pile Make a vertical pile with elements centered above one another.
rcol Right adjust a column of a matrix.
right Create large brackets, braces, bars, etc.
roman Change to roman font.
 0 Formatter Command Summary 0 621

rpile Right justify the elements of a pile.

size n Change the size of the font to n.
sqrt Draw a square root sign.
sub Start a subscript.
SUP Start a superscript.
tdefine Make a definition that will apply only for eqn.
to Used in summations, integrals, and similar constructions to signify
the upper limit.
Force extra space into the output.
-
Force a space one half the size of the space forced by .
{ I Force eqn to treat an element as a unit.
I
* . .
I
A string within quotation marks is not subject to alterations by
eqn.

Precedence
If you don’t use braces, eqn will do operations in the order shown in the following
list.

dyad v e c under bar t i l d e hat dot dotdot

fwd back down up
fat roman i t a l i c bold size
sub sup sqrt over
from t o

These operations group to the left:

over sqrt left right

All others group to the right.

The pic Preprocessor

In p i c there are often dozens of ways to draw a picture, not only because of the many
permissible abbreviations, but because p i c combines the language of geometry with
English. You can specify a line, for example, with direction, magnitude, and starting
point, yet often achieve the same effect by simply stating, “from there to there.”
Full descriptions of primitive objects in p i c can be ended by starting another
line, or by the semicolon character (;). A single primitive description can be continued
on the next line, however, by ending the first with a backslash character (\). Com-
ments may be placed on lines beginning with #.
622 UNlX Text Processing 0

pic Macros
The following macros are used to delimit p i c input from the body of the source file.
Only text within these macros will be processed by p i c .

.PS [ h [w]] Start p i c description. h and w, if specified, are the desired

height and width of the picture; the full picture will expand or
contract to fill this space.
.PS <file Read contents of file in place of current line. -
.PE End p i c description.
IPF End p i c description and return to vertical position before
matching PS.

Declarations
At the beginning of a p i c description, you may declare a new scale, and declare any
number of variables.
p i c assumes you want a 1-to-1 scale, with 1 = one inch. You can declare a dif-
ferent scale, say 1 = one-nth of an inch, by declaring, s c a l e = n.
p i c takes variable substitutions for numbers used in the description. Instead of
specifying, l i n e r i g h t n, you may use a lowercase character as a variable, for
example, a, by declaring at the top of the description:
a = n
You may then write l i n e r i g h t a.

Primitives
Primitives may be followed by relevant options. Options are discussed later in this sec-
tion.

a r c [cw] [options] [ ' 'text' '3

A fraction of a circle. (Default = 1/4 of a circle.) The c w option specifies
a clockwise arc; default is counterclockwise.
a r r o w [options] ["text' 'I [ t h e n . . . I
Draw an arrow. Essentially the same as l i n e ->.
box [options] [ ' 'text' ']
Draw a box.
c i r c l e [options] [ ' 'text' 'I
Draw a circle.
 0 Formatter Command Summary 0 623

ellipse [options] [ ’ ’text’ ’1

Draw an ellipse.
line [options] [ ’ ’text’ ‘1 [then . . .1
Draw a line.
move [options] [ ‘ ’text‘ ‘1
A move of position in the drawing. (Essentially, an invisible line.)
spline [options] [ ’ ’text‘ ‘I [then . . .3
A line, with the feature that a “then” results in a gradual (sloped) change in
direction.
‘ ‘text’
Text centered at current point.

Options

right [n] Specifies direction of primitive; default is direction in which

left [n] the previous description had been heading. Diagonals result by
UP rn1 using two directions on the option line. Each direction can be
down [n] followed by a specified length n.
rad n Specifies a primitive to have radius n (or diameter n).
diam n
ht n Specifies the height or width of the primitive to be n. For an
wid n arrow, line, or spline, refers to size of arrowhead.
same Specifies a primitive of the same dimensions of the most recent
matching primitive.
a t point Specifies primitive to be centered at point.
with .position at point
Specifies the designated position of the primitive to be at point.
f r o m pointl t o point2
Specifies the primitive to be drawn from pointl to point2.
Points may be expressed as Cartesian coordinates or in respect
to previous objects.
-> Specify the arrowhead to be directed forwards.
<- Specify the arrowhead to be directed backwards.
<-> Specify the arrowhead to be directed both ways.
chop n m Chop off n from beginning of primitive, and m from end. With
only one argument, the same value will be chopped from both
ends.
624 UNlX Text Processing 0

dotted Specifies the primitive to be drawn dotted, dashed, or to be

dashed invisible. Default is solid line.
invis

then.. . Continue primitive in a new direction. Relevant only to lines,

splines, moves, and arrows.

Text
Place text within quotation marks. To break the line, break into two (or more) sets of
quotation marks. Text always appears centered within the object, unless given one of
the following arguments:

1j u s t Text appears left justified to the center.

rj u s t Text appears right justified to the center.
above Text appears above the center.
below Text appears below the center.

Object Blocks
A complex object that is the combination o f several primitives (for example, an octa-
gon) can be treated as a single object by declaring it as a block:
Object: [
description

1
Brackets are used as delimiters. Note that the object is declared as a proper noun, hence
it should begin with a capital letter.

Macros
The same sequence o f commands can be repeated by using macros. The syntax is:
de € i n e sequence %
description

%
 0 Formatter Command Summary 0 625

In this example, we have used the percent sign (YO)as the delimiter, but any character
that is not in the description may be used.
Macros can take variables, expressed in the definition as “$1” through “$9”.
Invoke the macro with the syntax: sequence(vuluel,vulueZ,. . .)

Positioning
In a pic description, the first action will begin at (O,O), unless otherwise specified
with coordinates. Thus, the point (0,O) will move down and right on the drawing, as
objects are placed above and to the left of the first object.
All points are ultimately translated by the formatter into x- and y-coordinates.
You may therefore refer to a specific point in the picture by incrementing or decrement-
ing by coordinates, Le., 2nd ellipse - (3,1).
You may refer to the x- and y-coordinates of an object by placing .x or .y at
the end. For example, last box. x will refer to the x-coordinate of the most recent
box drawn. Some of the physical attributes of the object may also be referred to simi-
larly, as follows:
.x X-coordinate of object’s center.
-Y Y-coordinate of object’s center.
.h t Height of object.
.wid Width of object.
.rad Radius of object.

Unless otherwise positioned, each object will begin at the point where the last
object left off. If a command (or sequence of commands) is set off by braces ( 1 I),
however, pic will then return to the point before the first brace.

Positioning between Objects

When referring to a previous object, you must use proper names. This can be done two
ways:

By refemng to it by order, e.g., 1st box, 3rd box, last box, 2nd
last box, etc.
By declaring it with a name, in initial caps, on its declaration line, e.g.,
Linel: line 1.5 right from last box.sw

To refer to a point between two objects, or between two points on the same object, you
may write: fraction of the way between jht.position and second.position or
(abbreviated) fraction<first.position,secondgosition>
626 UNlX Text Processing 0

Corners

When you refer to a previous object, p i c will assume that you mean the center of the
object, unless you use a corner to specify a particular point on the object. The syntax
is:
.corner of object
for example, .sw o f last box. You can also use an abbreviated syntax:
object. corner
for example, last box. s w .
These corners may be:
n North (same as t)
S South (same as b)
e East (same as r)
W West (same as 1)
ne Northeast
nu Northwest
se Southeast
SW Southwest
t Top (same as n)
b Bottom (same as s )
r Right (same as e)
1 Left (same as w)
start Point where drawing of object began
end Point where drawing of object ended

You may also refer to the upper right, upper left, lower right, and
lower left of an object.

Numerical Operators
Several operators are functional in pic. These are:
+ Addition
- Subtraction
* Multiplication
/ Division
% Modulo
I
1

0 Forrnatter Command Summary 627

Default Values
arcrad 0.25 ellipsewid 0.75
arrowwid 0.05 linewid 0.5
arrowht 0.1 lineht 0.5
boxwid 0.75 movewid 0.5
boxht 0.5 moveht 0.5
circlerad 0.25 scale 1
dashwid 0.05 textht 0
ellipseht 0.5 textwid 0
 A P P E N D I X
a

C
Shell Command Summary

This section describes the syntax of the Bourne Shell. It lists special characters, vari-
ables, and built-in programming commands used by the shell.

Special Files

$HOME/.profile Executed at shell startup.

Special Characters for Filename Generation

* Match any string of characters.

? Match any single character.
[ . ..] Match any of the enclosed characters. A pair of characters separated
by a minus will match any character lexically between the pair.

Special Characters for Control Flow

I Perform pipeline (use output of preceding command as input of fol-

lowing command, e.g., c a t file I Ipr).
I Separate sequential commands on the same line.
& Run command in background (e.g., l p r file&).
&& Execute command if previous command was successful (e.g., grep
stringfile && lpr file).
 0 Shell Command Summary 0 629

It Execute command if previous command was unsuccessful (e.g.,

grep string1 file I I grep string2 file).
0 Execute commands enclosed in ( ) in a subshell; output from the
entire set can then be redirected as a unit or placed in the background.
I I
... Take all characters between single quotation marks literally. (Don’t
allow special character meaning.)
\ Take following character literally.
I’
... I#
Take enclosed characters literally but allow variable and command
substitution.
‘cmd‘ Use output of cmd as argument to another command.
# Begin a comment in a shell file.
<file Take input fromfile.
<<string Read standard input up to a line identical to string.
>file Redirect output to file (overwrite).
>>file Redirect output to end offile (append).
>&digit Duplicate standard input from digit e.g., 2 > c 1.
<&- Close standard input.
>&- Close standard output.

Variable Substitution

variable=value Set variable to value.

$variable Use value of variable.
$ variable-value Use variable if set; otherwise set to value. For example:
$ variable[:]-value TERM=$ { 1 :-$TERM} will set the TERM variable to the
value of the first argument to a shell script, if given, or else
to the existing (default) value of TERM.

$ variable=value Use variable if not set; otherwise set to value.

$ variable[:]=value
$variable?value Use variable if set; otherwise print value then exit.
$variable[:]?value
$ variable+value Use value if variable is set; otherwise nothing.
$variuble[:]+value

If the colon (:) is included in these expressions, a test is performed to see if the variable
is non-null as well as set.
630 UNlX Text Processing 0

Shell Parameters Set by the Shell under Execution

$# Number of command-line arguments.

$- Options supplied in invocation or by the s e t command.
$? Return value of last executed command.
$$ Return process number of current process.
$! Return process number of last background command.

Shell Variables Initially Set By p r o f i l e

$HOME Default (home directory) value for the cd command.

$IFS Internal field separators.
$MAIL Default mail file.
$PATH Default search path for commands.
$PS1 Primary prompt string; default is $.
$PS2 Secondary prompt string; default is >.
$TERM Specifies the type of terminal.

Shell Functions

name ( ) {commandl; . . . ; commandn)

Create a function called name that consists of the commands enclosed
in braces. The function can be invoked by name within the current
script.

Built-in Commands

file file
Execute contents offile.
break b r e a k [n]
Exit from a for, w h i l e , or u n t i l loop in n levels.
 0 Shell Command Summary 0 631

case case value i n

patternl ) commands: ;

-
patternn ) commands ;;
esac
For each item in list that matches pattern, execute command.
cd cd [ d i r ]
Change current directory to dir.
continue c o n t i n u e [n]
Resume nth iteration of a f o r , w h i l e , or u n t i l loop.
echo e c h o args
Print args on standard output.
eval e v a l [arg . . .I
Evaluate arguments, then execute results.
exec exec [cmd]
Execute cmd in place of current shell.
exit e x i t [n]
Exit the shell with exit status n , e.g., e x i t 1 .
export e x p o r t [var . ..]
Export variable var to environment.
for f o r variable [ i n list . ..]
do
commands
done
For variable x (in optional list) do commands.
if i f condition
t h e n commands
[e 1 i f condition2
t h e n commands;?] . . .
[e 1s e commands31
f i
If condition is met, do list of commands, or else if condition2 is met,
do commands2, otherwise do commands3. (See t e s t for a list of
conditions.)
hash h a s h cmds
Temporarily add cmds to search path.
632 0 UNlX Text Processing 0

login l o g i n [user . . . ]
Log in as another user.
newgrp newgrp [group . . . ]
Change your group ID to group; if no argument, change back to your
default group.

PWd PWd
Print current working directory.
read r e a d [var . . . ]
Read value of var from standard input.
readonly r e a d o n l y [var . . .I
Mark variable vur as read only.
return return
Stop execution o f current shell function and return to calling level.
set set [t] [ o p t i o n s ] [urg . , ,I
With no arguments, s e t prints the values o f all variables known to
the current shell. The following options can be enabled (-option) or
disabled (+option).

Don’t treat subsequent arguments beginning with - as

options.
-a Automatically export all subsequently defined variables.
-e Exit shell if any command has a nonzero exit status.
-k Put keywords in an environment for a command.
-n Read but do not execute commands.
-t Exit after one command is executed.
-U Treat unset variables as an error.
-V Print commands as they are executed.
-X Turn on t r a c e mode in current shell (echo com-
mands in scripts as they are executed).
. .. Assigned in order to $ 1 , $ 2 , . . . $ 9 .

shift shift
Perform a shift for arguments, e.g., $ 2 becomes $1.
test t e s t exp I [exp]
Evaluate the expression exp. An alternate form o f the command uses
[ ] rather than the word test. The following primitives are used to
construct expression.

-b file True i f file exists and is a block special file.

-c file True iffile exists and is a character special file.
 1

0 Shell Command Summary 0 633

-d file True if file exists and is a directory.

-€ file True if file exists and is a regular file.
-g file True if file exists and its set-group-id bit i s set.
-k file True iffile exists and its sticky bit is set.
-n sl True if the length of string sl is nonzero.
-r file True if file exists and is readable.
--s file True iffile exists and has a size greater than zero.
-t In1 True if the open file whose file descriptor number is
n (default is 1 ) is associated with a terminal device.
-u file True iffile exists and its set-user-id bit is set.
-w file True if file exists and is writable.
-x fife True iffile exists and is executable.
-z s l True if the length of string s l i s zero.
SI = s2 True if strings sl and s2 are identical.
SI != s2 True if strings sl and s2 are not identical.
SI True if string SI is not the null string.
nl -eq n2 True if the integers nl and n2 are algebraically
equal. Any of the comparisons -ne, -gt, -ge,
-It, and -le may be used in place of -eq.

times times
Print accumulated process times.
trap t r a p [cmd] [nl
Execute cmd if signal n is received. Useful signals include:
0 Successful exit of command.
1 Hangup of terminal line.
2 Interrupt.
15 Process is killed.
type type commands
Print information about commands.
until un t i 1condition
[do commands]
done
Until condition is met, do commands (see t e s t for conditions).
ulimit u 1 imi t [size]
Set maximum size of file that can be created to size; if no arguments,
print current limit.
umask uma s k [nnn]
Set file creation mask to octal value nnn.
634 UNlX Text Processing 0

unset unset vars ...

Remove definitions for variable var.
wait w a i t [n]
Wait for specified process with identification number ( n ) to terminate
and report its status.
while w h i 1e condition
[do commands]
done
While condition is met, do commands (see test for conditions).
filename filename
Read and execute commands from executable file JiZename.
 A P P E N I X
8 8
D
8 .
D

Format of t r o f f Width Tables

As discussed in Chapter 4, trof f uses width tables stored in the directory

/usr/lib/font to determine how to place text on the page. To do this, it needs to
know how wide each character is.
For each type of trof f output device supported by your system, there should
be a directory called /us r / 1ib / font / devu, where XIC is the name of the device.
For example, on our system:
$ 1s -F /usr/lib/font
devl j/
devps/

Within each of these directories resides an overall device description file, called DESC,
and individual font files for the fonts on your system. These files exist both in ASCII
and binary form. The binary files are created from the ASCII versions using a utility
called makedev, and have the suffix - out.
On our system, here’s what the font directory for the HP Laserjet contains:
$ 1s /usr/lib/font/devl
B DESC I S
B.out DESC-outI.out S.out
cw HB R TY
CW.out HB.out R.out TY.out

TheDESCFile
The DESC file contains an overall description of the output device, including its reso-
lution in dots per inch, the paper size, the fonts that will be mounted by default, the
available point sizes, and a complete list of all the t r o f f special character names
supported on that device.

635 8
636 0 UNlX Text Processing 0

A DESC file might look something like the following example:

# HP L a s e r J e t
f o n t s 6 R I B HB CW S
s i z e s 7 8 1 0 1 2 1 4 17 22 27 0
res 3 0 0
hor 1
vert 1
u n i t w i d t h 12
p a p e r w i d t h 2400
p a p e r l e n g t h 3300

charset
\I \ A \-
f i fl f f Fi F 1
b r vr u l r u
b u s q e m h y 1 4 1 2 3 4 aa g a

sc g r n o i s p t es m o
dd r h l h b s o r c i
It l b r t r b lk r k b v I f r f IC r c
The following keywords are used in the D E S C file:

fonts The number of fonts to be mounted for the device, followed

by a list of the font names (maximum is ten). The user can
request other fonts from within a document. However, the
fonts listed here will be “mounted” (by analogy with the
CAT typesetter), and can by referenced by position (\ f 1,
\ f 2 . . .) as well as by name.
sizes The sizes in which the various fonts are available.
res The resolution of the output device, in dots per inch.
hor The minimum number of units of resolution that the device
can move in a horizontal direction.
vert The minimum number of units of resolution that the device
can move in a vertical direction.
unitwidth The point size at which character widths are specified in the
other files.
paperwidth The width of the page in units of resolution (e.g., 8 inches x
300 = 2400, the width for the LaserJet, because it forces a ‘ / 2 -
inch margin).
 0 Format of trof f Width Tables 0 637

paperlength The length of the page in units of resolution (e.g., 1 1 inches x

300 = 3300, the length for the LaserJet).
biggest font The maximum number of characters in a font.
charset The list of character names that are supported on this output
device. The keyword should be on a line by itself; the list of
characters starts on the next line.
# Begins a comment.

Font Description Files

For each font listed on the fonts line of the DESC file, there should be a font file
with the same name. The font file contains a list of all the characters in the font, along
with the width and other associated information.
A font file looks like this:
name R
internalname Roman

charset
4 0 0
8 0 0
VII 0 3 13
ru 25 0 17

A 42 2 65
B 35 2 66
C 37 2 67

W 40 0 119
X 28 0 120
Y 28 1 121
z 25 0 122

Four columns, separated by tabs, are listed for each character.

638 0 UNlX Text Processing 0

The first column lists the character name-either the letter, digit, or symbol, or a
two-character t r o f f special character name defined in the c h a r s e t section of
DESC.
The second column contains the width of the character in output device units.
The width is the width of the character at the point size specified by the u n i t w i d t h
keyword in DESC. For example, if unitwidth is 12, then from the portion of the
table just shown, we know that a 12-point A in the roman font is 42 units wide. The
t r o f f formatter determines the width at other point sizes by scaling the
u n i t w i d t h size.
The third column describes the character type-that is, whether it is an descender
(l), ascender (2), both (3), or neither (0).
The fourth column contains the typeset code for the character. This code is the
value that the output device will recognize to generate the character. This information
is obtained from the typesetter or laser printer vendor. The code can be in decimal or
octal form. (Octal is specified by a leading zero.)
In general, whomever supplied the driver for the output device will provide you
with appropriate width tables for the supported fonts. However, you may have access
to other public domain fonts for output devices that support downloadable raster fonts.
In this case, you may need to build your own tables.
In addition, you may want to “tune” tables by adjusting the widths slightly if you
find that the character spacing is poor. Creating a font table from scratch requires a
magnifying glass, a micrometer, a good eye, and a lot of patience.

Compiling Font Files

After you are satisfied with your width tables, they need to be compiled using the
m a k e d e v utility:
$ makedev DESC Compile all fonts in DESC

Running m a k e d e v on DESC will compile all of the fonts listed on the f o n t s line
in that file. You can compile a font that is not included in DESC by specifying its
name on the command line:
$ makedev B Compile the bold font

Font Usage Limitations

The user is not restricted to using the “mounted” fonts that have been listed in DESC.
Any font supported by the output device, and for which a compiled width table exists,
can be referred to from within a document. For example, if you had a Palatino font
family named PA, PB, and P I , there should be files called:
 0 Format of t r o f f Width Tables 639

PA. o u t PB. out PI. o u t

One problem that i s sometimes encountered is that t rof f has problems if a font that
is used in this way is larger (in absolute file size) than the largest of the mounted fonts
specified in DESC. The t ro f f forrnatter only aIlocates enough memory for the larg-
est font in DESC. If you encounter this problem, you can either strip unneeded charac-
ters out of the font, pad a font in DESC, or add the large font that i s giving you trouble
to DESC.
 A P P E N D I X
8 8 8 8

Comparing m m and IClS

If you have both ms and mm on your system, you may be interested in looking at both
packages, perhaps evaluating features. In general, ms has many of the same capabili-
ties as mm. However, it lacks some essential features, such as automatically numbered
lists and table o f contents generation. On the other hand, it is much easier to learn the
internals of ms, and therefore easier to extend it with your own macros.

. Paragraphs
The basic paragraph types are block and indented.
ms mm Description

.P .LP Begin a block paragraph.

.P 1 .PP Begin a paragraph with indented first line.

In mm, the default paragraph type can be changed from block to indented by setting the
number register Pt to 1 or 2. The ms macros lack this generalizing mechanism.

Justification 8

When using the nrof f formatter, mm does not justify the right margin. .
SA 1
turns on justification for both formatters. .
SA 0 turns it off.
The ms macros do not provide a macro for inhibiting the normal justification o f
.
paragraphs. However, the n a request can be used to do this.

640 8
 0 Cornparingmmandms 0 64 1

Displays =
Displays are produced in a very similar way in both macro packages, using the DS/DE
pair of macros. In mm, display are left justified; in m s , displays are indented. The
options that allow you to change the placement of the display are basically the same.
The mm macros provide for static and floating displays (. DF). In ms, this is
done with a separate pair of keep macros (KS/KF and KE).
In mm,you can turn on fill mode within the display and specify an indent from
the right margin. This is used for quoted material and has its equivalent in m s with the
QP or the QS/QE pair.
In addition, the same set of delimiter pairs for tbl, eqn, and pic are avail-
able in both packages.

Formatting Lists
The mm macros have sophisticated list formatting macros that are lacking in m s . The
.IP macro in m s produces the equivalent of a variable-item list in mm. In other
words, you can get a numbered list by specifying the number as a label to an indented
paragraph, but you cannot get an automatically numbered list.

ChangeFont
The .B (change to bold), .I (change to italic), and .R (change to roman) macros used
for changing fonts are the same. The mm macros allow up to seven arguments for
alternating with the previous font, but m s is limited to two.

Change Point Size

Both packages allow you to change point size. In mm, .S specifies a new point size
and .SM reduces point size relative to the current size.
When you change the point size using ms macros, it is always done relative to
the current point size. The .LG and .S M macros increase and decrease the current
point size by 2 points. The .NL macro restores the default point size.
642 UNlX Text Processing 0

= Headers and Footers

The mm macros provide macros for specifying a delimited string that will appear left
justified, centered, and right justified in a page header or footer. The .PH macro
.
defines a page header and PF defines a page footer. In addition, mm provides varia-
tions o f these macros for specifying headers and footers for odd and even pages.
The m s macros handle this through setting individual strings. To define a string
that appears left justified in a header, use:
.ds L H string
The other strings for the header are CH and RH; other strings for the footer are LF,
C F , and RF.

Section Headings
Numbered and unnumbered section headings are available in both packages. The .SH
and .NH macros are used in ms. The .H and .HU macros are used in mm. The
main difference is where you specify the heading string. In mm,it is the first argument
on the line with the macro. In ms, it follows on the line after the macro and continues
up to the first paragraph macro.

Footnotes
. .
The pair of macros used for footnotes is the same ( F S and FE), although automatic
numbering of footnotes is provided in mm. One difference is that in m m the footnote
at the bottom of the page is printed in 8 points. The mm macros also provide a pair of
.
macros ( .RF and RE) for collecting a page o f references.
 A P P E N X
I . D I
I

The format Macros

Throughout this book, we’ve made extensive references to portions of the extended m s
macro package that we use in our technical writing business. These macros are used in
conjunction with the format shell script to provide a complete document formatting
environment.
This package was originally developed by Steve Talbott of Massachusetts Com-
puter Corp. (MASSCOMP). W e have extended and generalized it to meet the docu-
ment design needs of many different clients.
The purpose of this appendix is to summarize, in one place, the function of the
macros that the package contains. We have found that this set of macros covers the
basic needs of people involved in the development of technical books and manuals.
The package relies on the existence of the underlying m s macros. In this sense,
it is not a complete package. However, it is possible to define a simple subset of the
m s macros to cover the basics if the full implementation of m s is not available.
For more information on the full implementation of these macros, please feel free
to contact us in care of the publisher.

Summary of the Macros

The following list summarizes the user-callable macros in the format macro pack-
age.

.[ABCDIh A-level head, B-level head, and so on.

.Dr Specify whether the current version is a draft. (Drafts are
dated.)
.Fs Start a figure.

643
644 UNlX Text Processing 0

.F e title Figure end. Figures are automatically numbered, and

given the specified title.
.H1 [ C ] Print a horizontal line the width of the page, using charac-
ter c. (Default is underscore.)
.I O C [strings] Start an interoffice memo.
.TO List of names following . T O
will b e placed
I”’

in separate distribution list.

.TO name name is addressee. Maximum of five such .TO lines.
.DA date date is date of the memo; will be included in page footer.
.ND date date is date of the memo; will be omitted from page
footer.
.FR name name is sender. Maximum of five such .FR lines.
.CC name name is person to receive copy of memo. Maximum of
five .CC lines.
.SU subject subject is subject of the memo.
.IP label indent [O] Begin paragraph with “hanging indent.” Following text
is indented, while label remains at the margin.
.LP [O] Start a (left-justified) paragraph. 0 suppresses blank line.
.Ls [type] [indent] [buEletl
Start a (possibly nested) list. type is N (number), A
(alphabetical uppercase), a (alphabetical lowercase), I
(Roman numeral uppercase), i (Roman numeral lower-
case), B (bullet). Default indent is 5. bullet is alternative
bullet string (null string is acceptable).
.Li [O] List item. 0 suppresses preceding blank line.
.Le [O] End of innermost list. 0 suppresses preceding blank line.
.Lt [l 12) Enter address blocks and date (l), and salutation (2) of a
letter.
.Nd n Need n lines. If n lines do not remain on the page, eject
new page. Unlike .ne, .Nd causes a break.
.Nh [1 Io] Enable/disable numbered headings (enabled by default).
.NS type Start a NOTE of type N (Note), C (Caution), w (Warn-
ing), R (Review Note), or P (Private Note). Review
notes are printed in the text and summarized in a list at the
end. Private notes appear only in the end list.
 The format Macros 0 645

.Ne End a note.

.OB string Print an overbar (over a string).
.P s [indent] Start a “printout” (display). Text is printed in the CW
font and preserved as is-there is no filling.
.P e End a printout. See .Ps .
.Rh [O I 11 [desc] h e a d . . .
Create reference page header.
.Se [number] [title] Start a section (chapter). This sets up many defaults, and
i s desirable to use for most documents.

.SE Screen end. End a computer screen illustration begun with

.ss.
- S S [width] [height] Start a screen illustration (box with curved comers). If
width and height are not specified, scale to size of con-
tents.
.T c level Specify what level of heading will be saved in the table of
contents (Ah to D h )
.Ti text Title-goes in left page footer.
.St text Subtitle-goes in right page footer.
.Ts title Start a table with given caption. Tables are automatically
numbered.
.Te End a table. (Output a blank line.)
. X X text Make an index entry out of text, with automatic addition
of a page number.
.XN text Make an index cross-reference out of text (no page
number).
 A P P E N D I X
8 8 8 8

Selected Readings

The following books may be helpful either when you’re starting out, or when you’re
ready to go on to more advanced topics.

Introductory UNIX Texts

Kochan, Steven G. and Patrick H. Wood. Exploring the UNiX System, Hasbrouck
Heights, NJ: Hayden Book Co., 1984. A comprehensive introduction to the UNIX
system. (371 pp.)
Todino, Grace. Learning the UNIX Operating System, Newton, MA: O’Reilly and
Associates, Inc., Nutshell Handbooks, 1985. A brief introduction to essential UNIX
skills, designed to be read and mastered in one or two sessions. (73 pp.)

Advanced Topics
Kemighan, Brian and Rob Pike. The UNIX Programming Environment, Englewood
Cliffs, NJ: Prentice-Hall, 1984. The best introduction to the practical philosophy of
UNIX programming. (240 pp.)
Kochan Steven G. and Patrick H. Wood. UNIX Shell Programming, Hasbrouck
Heights, NJ: Hayden Book Co., 1985. A comprehensive and readable discussion of
shell programming. (422 pp.)
Talbott, Steve. Managing Projects with Make, Newton, MA: O’Reilly and Associates
Inc., Nutshell Handbooks, 1985. A concise but thorough description of the UNIX
make utility. (63 pp.)

8 646
 D E X

. 1 C macro (ms) 125-126 awkronym script 390, 392-393, 396-397,

- 2C macro (ms) 125-126 399-400
basic operations 387
BEGIN procedures 392
A built-in functions 408
capabilities of 387
- A Bmacro (ms) 122- 123 changing the field separator 391
.ab request 440 command summary 5 8 7 , 5 8 9 , 5 9 1
absolute motions 448, 464 conditional statements 394
.ad request 71-72,485 debugging 410-41 1
- AE macro (ms) 122- 123 dividing input into records and fields 389
.AI macro (ms) 122- 123 error handling in 404
. A L macro (mm) 156-158 f o r loops 3 9 3 , 4 0 9
alias command 294 f o r loops with arrays 396
alphabetic lists formatted print statement 402
mm macros 156-158 invoking 388-389
.am request 4 1 3 , 5 3 4 passing parameters from a shell script 390
arguments pattern matching 388, 588
in macros 416 scripts for order tracking 405
in shell scripts 23 similarities to sed 387
to UNIX commands 13 subdividing a field 407, 553
arithmetic expressions substr function 557
in t r o f f 419 syntax o f procedures 589
.a s request 430 system variables 392, 588
.AU macro (ms) 122-1 23 testing fields 390
autoincrementing number registers 428 used for indexing 551-557
awk 342, 387-411, 551-557 variables 395, 589
arrays 3 9 5 , 4 0 9 , 5 8 9 while loops 393

. 647
648 0 UNlX Text Processing 0

B .b r request 432,69
bracket-building characters 466, 603
.B macro (mm) 141-142 breaks
.B macro (ms) 114-115 importance of 432,72
.B1 macro (ms) 124 no-break control character 73
.B2 macro (ms) 124 t r o f f requests which cause 73
background processing 20 .BT macro ms) 492
backing up files 336-338 .BX macro (ms) 123
baseline spacing 98-99
.B D macro (ms) I19
- bd request 463 C
b d i f f command 319
- B G macro (ms) 488 C shell 13-14,294, 343
.BI macro (mm) 143 c2 request 435
.BL macro (mm) 151, 154-155 case shell command 347-349
boldfacing c a t command 129, 14-15,298
by overstriking 463 C / W typesetter 64,66,76
Bourne shell 13-14, 343-354 .cc request 435
background processing 20 cd command 16
breaking out of a script ( e x i t ) 352 .CD macro (ms) 117, 119
cancelling commands 13 CDPATH environment variable 294
CDPATH environment variable 294 .ce request 80-82
command summary 625-634 .c f request 460
conditional execution 347-349 .ch request 503
discarding and shifting arguments 349 change bars 476
export command 352, 51 change pages 507
HOME variable 17 checkeq command 251
interrupting commands 13 checkmm command 154
long command lines 22 chmod command 19,23,298
PATH variable 344 chown command 298
prompt 13 c o l c o m m a n d 86
repetitive execution 350, 355 comm command 327-328
resuming commands 13 commands
secondary prompt 22 cancelling 13
shell functions 630 interrupting 13
special characters 629 resuming 13
t e s t options 347 comments
variable substitution 17, 629 in t r o f f 412
variables set by . p r o f i l e 630 conditional execution
variables set by shell under execution 630 in shell scripts 347-349
- bp request 86, 89-91 in t r o f f 418-423
- BR macro (mm) 143 in awk 394
 0 Index 649

constant spacing 47 1,529 directories

copy mode 417 changing 16
core dumps 41 1 creating 17
cover sheet macros 122-123, 127 home 16
cpcommand 17 listing contents of 16
c p i o command 336-338,340 printing current 16
using with f i n d 337-338 public 339
c r y p t command 335 root 15
- cs request 471 sub- 15
c s h command 343 displays
.c s h r c file 294 fill options (mm) 146
c s p l i t command 333-335 floating (mm) 147
CTRL key 42 floating (ms) 119
c u command 339 in mm macros 145-146
.c u request 83-84 in m s macros 117-120
cursor movement 28 labels (mm) 148
customizing v i 51 mmvs. m s 640
c u t command 328, 331 static(mm) 147
cut marks 498-499 static (ms) 119
diversions 431,451, 533, 536-538
appending to 433
D closing 431
creating 431
.da request 433, 534 naming by number register interpolation 538
d a t e c o m m a n d 13 splitting 538
DDL 67 table o f contents 532
.D E macro (mm) 138 used for footnotes 504
- D E macro (ms) 117-120 .DL macro (mm) 154-155
.de request 99, 413 Documenter’s Workbench 64
debugging dot-matix printers 8
in awk 410-411 drawing (see also p i c preprocessor)
in p i c 281 boxes 123,451
in t r o f f 436-438 changing line weight 454
d e r o f f command 322-323 curves 455
D E S C file 635-636 including Macintosh illustrations 460
devps postprocessor 66-67,460 lines 449
.DF macro (mm) 147-148 sample figures 455
.d i request 431, 451 .DS macro (mm) 138, 145-146
dial-up line 339 - D S macro (ms) 117-120
d i f f command 312-314, 316-317, 357 - d s request 101,429
d i f f 3 command 314, 317-318
d i f f m k command 477
650 0 UNIX Text Processing 0

E problem checklist 251

problems with .s o request 480
.EC macro (mm) 133, 148-150 quotation marks 237,242,248
.ec request 435 simple example 233
echo command 14 spaces in equations 236
.E F macro (mm) 133 special character names 232, 239, 241
egrep command 303 square root signs 242
.EH macro (mm) 133 subscripts 237-238
.el request 418-423 summary of command characters 618-62 1
e 1if shell command 348 summations 242
em (defined) 76 superscripts 237-238
emdash 364 syntax 233
.em request 534 tabs within equations 237
emacs editor 5, 24 using braces for grouping 238
en (defined) 76 using mm with 235
end macro 534-536 using n r o f f with 234
environment variables using t b l with 205,234
E X I N I T 51 vertical piles (columns) 245
HOME 17 vertical spacing 232, 250
PATH 344 errors
path 345 in mm 130
setting 20-2 1 in m s 105
TERM 20 in t r o f f macros 439
environments 433, 529 messages from UNIX commands 14
- eo request 435 escape sequences
eqn preprocessor 86, 232-252 in t r o f f 443-444
abbreviating a string 247 .ev request 433
arrays 244 ex editor 177-202, 342-346, 355-357
braces and brackets 243 @ functions 201
diacritical marks 232, 246 : prompt 27
displayed equations 234 abbreviating recurring phrases 198
fonts 232, 249 address symbols 579
fractions 244 appending to existing file 191
Greek alphabet 232,239 command mode 179
grouping items 250 command summary 578-584
horizontal spacing 232 confirming replacements 182
inline expressions 235 copying lines 180
integrals 242 creating a subshell 192
invoking 233 current line 178
limits 242 deleting lines 180
lining up equations 248 differences from s e d 361
matrices 244 editing multiple files 195
point sizes 232, 249 executing from vi 180,27
precedence of operations 250, 621 executing UNIX commands from 192
 0 Index 651

EXINIT variable 51 .exrc file 50,52

exiting 190, 578 extended m s macros 509-541,643-645
- e x r c file 50.52 chapter headings 514,518-519
filtering text through a UNIX command 193 drawing horizontal lines 5 18
global search and replace 183, 186 figure numbering 523
insert mode 179 headers and footers 522
invoking 178,578 invoking 511
leaving insert mode 179 lists 525-528
limiting search to complete words 186 notes 530-531
line addressing in 181, 579 numbered headings 520
mapping commands to keys 198, 200 section headings 521-522
moving lines 180 structure of 5 1 1
moving text blocks by patterns 187 summary o f 643-645
pattern matching 184, 188 table numbering 523
printing line(s) 178 table of contents 532
quitting without saving edits 191
range of lines 179
reading in a file 192 F
reading in result o f UNM command 192
renaming the buffer 191 .F C macro (mm) 133
saving files 190 - f c request 469
saving part of a file 191 .FD macro (mm) 132, 171
scripts 342-346, 355-357 .FE macro (mm) 170-171
search and replace 181-182 .FE macro (ms) 124-125
search for general classes of words 186 .F G macro (nun) 148-150, 169-170
substitute command 179- 181 f g r e p command 303
switching files 196 .f i request 69
syntax of commands 178, 342 fields
using current and alternate filenames 197 in auk 389-392
yanking text from one file to another 197 in c u t a n d p a s t e 328
.EX macro (mm) 133, 148-150, 169-170 in s o r t 325-326
.ex request 478 in t r o f f 469
ex scripts 354-355 file management 3
built by d i f f 357 file system 15
executing with :so 354 files
EXINIT variable 5 1 backing up 336-338
e x i t shell command 352 characteristics 298
export command 51 copying 17
expr command 388 counting characters in 21
expression operators date and time last modified 298
in awk 393 editing multiple 195
in p i c 625 locating 296
in t r o f f 420 metacharacters 13, 19, 295, 627
moving 17
 0 UNIX Text Processing 0

naming restrictions 25 .f p request 92,95

organizing 293 .FS macro (mm) 170-171
permissions 18-19, 298 .FS macro (ms) 124-125
renaming 17 . f t request 92, 95
searching within (see also grep) 301 function keys (mapping) 199
size in bytes 298
tracking changes to (SCCS) 319
transferring to other systems 340 G
viewing contents 298
filling (definition of) 60 g e t o p t command 349
filters 21-22, 298, 360 graphics (see also p i c preprocessor) 253
final book production 562 Greek characters 232, 239, 604
f i n d command 296-298,302 g r e p command 22, 301-302
using with c p i o 337-338 using with find 302
.f 1request 127,438, 491
flushing output buffer (see .f 1)
.F O macro 490,492 H
fonts 62, 66
boldfacing by overstriking 463 -H macro (mm) 162-166
changing 92-94 hanging indents 79-80, 110- 111
changing (eqn) 249 head command 300
changing (ms) 114-115 headers 126, 133, 493
changing (tbl) 216 headings
constant width 47 1,529 in extended ms 5 14-522
contents of font files 66, 635-639 in mm 160-168
downloadable 66 in m s 120-122
four standard 64 in wide margin 522
mounted 92,636 here documents 357
special 64, 96 -HM macro (mm) 167-168
footers 126, 133, 493 home directory 16
footnotes 500-501, 641 HOME variable 17
mm macros 170-171 .HU macro (mm) 162-166
m s macros 124-125 .h w request 73-74
for shell command 355 .hy request 74-75.485
form letters 477 hyphen command 307,74
format shell script 364-365, 369,539, 542, hyphenation 61
545,559 checking for correcmess 74
formatting enabling in trof f 74
with a markup language 6 in mm 101, 137
with a word processor 6 rules for 73
formatting defaults
mm (see also mm macros) 130
m s (see also m s macros) 106
 0 Index 653

I K

. I macro (mm) 141-142 - KE macro (ms) 120

.I macro (ms) 114-115 keep and release (see also displays) 120, 431
.I B macro (mm) 143 Kernighan and Pike
.I D macro ( m s ) 117, 119 UNIX Programming Environment 11, 395
.i e request 418-423 kerning 61
.i f request 418-423 .KS macro (ms) 120
i f shell command 347-349
.ig request 438
.i n request 79 L
indents 79-80, 135
indexing 540,548 laser printers 9
adding formatting codes 555 .IC request
468
building the page number list 55 1 .LDmacro(ms) 117, 119
final formatting codes 557 . L E macro (mm) 151-152
form o f user entries 548 leaders 468-469
sorting raw entries 549 leading 98
special formatting problems 558 letter-quality printers 8
subordinating secondary entries 553 .LG macro (ms) 116-1 17
integrals (see eqn) 242 .1g request 474-475
Interpress 67 .LI macro (mm) 151-152
interrupted lines (in t r o f f 423 ligatures 473, 475, 522
line numbers (in v i ) 47
inverse conditions 423 lists
- IP macro 527 alphabetic (mm) 156-158
.IP macro ( m s ) 106, 110-1 13 alphabetic (mS) 525-526
- I R macro (mm) 143 bulleted (ms) 525-526
.IZ macro 486 extended in m S 525, 528
in m s 113
marked (mm) 154
J mm macros 150-151
mmvs. m s 640
join command 325-327 nested (mm) 153
justification 60,73, 137 nested (ms) 525-526
definition of 60 numbered (mm) 156-158
mm macros 137 numbered (mS) 525-526
m s macros 107 reference (mm) 158
nroff vs. t r o f f 61 user-supplied marks (mm) 156
types of 7 1-72 variable-item (mm) 158, 161-162
. l l request 78,485
local horizontal motions 447
 0 UNlX Text Processing 0

local vertical motions 445 marking a horizontal position 464

.login file 343 marking a position
lpcomrnand 129 in troff 461
.LP macro (ms) 106-107, 121 in vi 57
Is command 13, 16 .MC macro (ms) 125-126
.Is request 8 6 , 9 9 .mc request 476
. It request 495 me macros 88, 104
. m k request 4 6 1 , 5 0 5
mkdir command 17
M .ML macro (mm) 154, 156
mmcommand 129
Macintosh mm macros 88
illustrations 460 .AL macro 156-158
word processing on 9 alphabetic lists 156- 158
macros (see also mm and ms) altering heading style 164-168
appending to 413 .B macro 141-142
arguments 416 .BI macro 143
comparing mmand m s 640-642 .BL macro 151, 154-155
copy mode in 417 bold font 141
defining 127, 173, 413 bottom-of-page processing 176
developing a package 481 .BR macro 143
initializing 486 changing fonts 141
listing existing names 414 changing point sizes 143-144
naming conventions 414 changing reference defaults 172
nested 418 changing the heading mark 167-168
new or extended? 482 comparedto ms 640-642
page transition 489 .DE macro 138
removing 413 default formatting 130
renaming 415 - DF macro 147-148
reset 487 display fill options 146
setting default values 485 display formatting options 145
structure of package 48 1 , 4 8 3 display labels 148
style 441, 522 displays 138, 145-146, 148
tabs and leaders in 470 .DL macro 154-155
/usr / 1 ib/tmac 483 - DS macro 138, 145-146
mail command 21 .EC macro 133, 148-150
make command 320-321 .EF macro 133
coordinating final book production with 562 .EH macro 133
building makefile with a shell script 566 errors 130
makedev command 637 - E X macro 133, 148-150, 169-170
man macros 88 extensions to 173
mapping function keys 199 .FC macro 133
.F D macro 132, 171
.FE macro 170-171
 0 Index 0 655

.F G macro 148-150, 169-170 reserved macro and string names 6 15

floating displays 147 .RF macro 171-172
footers 133 .RI macro 143
footnotes 170-171 .RL macro 158
.F S macro 170-171 roman font 141
.H macro 162-166 .RP macro 132, 172-173
headers 133 .RS macro 171-172
heading number registers 165 - S macro 143-144
heading strings 165 . S A macro 137
headings 165, 176 .SK macro 150
.HM macro 167-168 .SM macro 144-145
.HU macro 162-166 .SP macro 136-137
hyphenation 101, 137 spacing between paragraphs 136
- I macro 141-142 static displays 147
.I B macro 143 strings 175
indented paragraphs 135 summary of macros 608-613
invoking 129 summary o f number registers 613
.I R m a c r o 143 table of contents 168-170, 176
italic font 141 .T B macro 133, 169-170
justification 137 .TC macro 168
.LE macro 151-152 top-of-page processing 176
.L1 macro 151-152 unnumbered headings 162-166
lists 150-156 user exit 176
marked lists 154 user-supplied list marks 156
.ML macro 154, 156 variable-item lists 158, 161-162
modifying 173 vertical margins 176
nested lists 153 vertical spacing 136- 137, 143-144
number registers 174 .vL macro 158, 161-162
numbered headings 162-166 mmt command 129
numbered lists 156-158 modem 339
.P macro 135-136 more command 1 2 9 , 2 9 8
page break 150 mptx macros 88
page layout 132, 134 m S macros (see extended m s macros)
page numbering styles 132 m s macros 104-127
page transition 91, 176 - 1c macro 125
paragraphs 135 .AB macro 122
paragraphs indented with exceptions 136 .AE macro 122
.PF macro 133 . A I macro 122
- P H macro 133 .AU macro 122
predefined string names 612 .B macro 114-1 15
.R macro 141-142 .B1 macro 124
.RB macro 143 . B 2 macro 124
reference lists 158 . B D macro 119
references 17 1- 172 .BG macro 488
 I
656 0 UNlX Text Processing 0

.BT macro 492 page transition 91,483-508

.BXmacro 123 paragraphs 106- 110
.CD macro 117, 119 point size 116-117
changing bottom margin 508 .PP macro 106-107, 121
comparedto mm 640-642 predefined and user-definable strings 607
date string 126 problems on first page 127
. D E macro 117-120 .PT macro 492
displays 117-118 -QE macro 110
drawing a box 123 -QP macro 106-107, 110
.DS macro 117-120 -QS macro 110
error handling 105 quoted paragraphs 107
extensions to (see extended ms macros) .R macro 114-115
- FE macro 124-125 .RE macro 111
.F O macro 490,492 redefining header or fmter 494
fonts 114-115 reserved macro and string names 607
footers 126, 493 reset macro 106, 487-488
footnotes 124-125, 500 . R S macro 111
- FS macro 124-125 .RT macro 106,487-488
headers 126,493 .s H macro 120-122
headings 120-122 .SM macro 116-1 17
- I macro 114-115 spacing between paragraphs 107
- I D macro 117, 119 summary of macros 604-606
indented paragraphs 110 .TL macro 122-123
initialization sequence 486 two-column processing 125-126, 505
internal macros 606 .uL macro 116
internal number register names 606 underlining 116
invoking 105 vertical spacing 107
.I P macro 106, 110-1 13, 527 multi-column processing 125- 126,505-506
.KE macro 120 mvcommand 17
.KS macro 120
labeled item lists 110
.LD macro 117, 119 N
.LGmacro 116-117
.LP macro 106-107, 121 - na request 71
.MC macro 125-126 .n e request 90,508
multi-column processing 125- 126, 505-506 .nf request 69
.NH macro 120-122, 520 .NH macro (ms) 120-122, 520
.NL macro 116-117 .NL macro (ms) 116-117
.NP macro 127, 490,492 -nm request 475
numbered lists 1 I3 .nn request 476
number register default values 485 no-break control character 73
page layout 106,606 no-fill mode 60,69, 83, 146
page layout defaults 106
page size 495
 Index 657

no-space mode 4 9 0 , 5 1 9 .nx request 478

notes, cautions and warnings 530-531
.NP macro (ms) 1 2 7 , 4 9 0 , 4 9 2
.n r request 10 1, 4 19-424 0
n r o f f formatter (see also t r o f f )
command line options 593, 63 options to UNIX commands 13
default line length 78 .os request 90-91
device units 76 output redirection 14-15, 21, 298
inability to use p i c with 253 appending to a file 15
interword spacing 62 overstriking 462, 465
invoking 593, 63
sample output from 68
submitting documents coded for t r o f f 67 P
summary of requests 594-598
units of measure 76 .P macro (nun) 135-1 36
using e q n with 234 pack command 338
.n s request 127,490, 90-91 page breaks 508
number registers 100-101, 419-425, 485, 526, nun macros 150
529 without lie breaks 89
altering output format 428 page description languages 67
as global variables 425 page layout 492
as nested list counter 526 in nun 132, 134
autoincrementing 428, 526 in m s 106
finding names of existing 415 page number
in m s 100 in mm 130, 132-133
interpolating 424 in three-part titles 494
mm 174 setting from command line 133
m s default values 485 page offset 78, 485, 523
naming 424 page size 495
naming by interpolation 526 page top resets 507
predefined 426 page transition 91, 176, 483-508
read-only 427 paragraphs
removing 429 indented (mm) 135
scaled units 424 indented (ms) 110
setting default values with 485 indented with exceptions (mm) 136
setting from command line 101 mm macros 135
substituting for environment switch 529 ms macros 106
used as flags 425 quoted (ms) 107
used to generalize macros 425 spacing between (mm) 136
numbered headings 5 14 spacing between (ms) 107
numbered lists p a s t e command 328,331
mm macros 156-158 P A T H variable 344
m S macros 525-528
 I
658 0 UNIX Text Processing 0

pathname 16 locating objects using Cartesian coordinates

.pc request 494 282
pcat command 338 locating specific points 266
.PF macro (mm) 133 macros 289-290
pg command 129,298-299 movement from a referenced object 265
help screens 299 naming an object 273
-P H macro (mm) 133 place and position notations 272
pic preprocessor 253-292.531 placing objects 264
adjusting drawing motion 260 placing text in a drawing 270, 274
adjusting label placement 257 positioning object blocks 277, 625
a r c 263,283 problems with . s o request 480
arrow 262 programming drawings 28 1
as a programming language 28 1, 29 1 reading description from remote file 289
automatic scaling 279 redefining standard dimensions 278
basic figures (graphics primitives) 255 relational operators for i f statements 292
changing direction of drawing 260 reusing dimensions 256
controlling the dimensions of a drawing 279 scaling 282
copy facility 289 specifying dimensions 255
debugging 28 1 specifying size of graphics primitives 255
declarations 622 s p l i n e 263
default dimensions of standard objects 278 spline 283
defining macros 285, 624 start and end macros 253
defining object blocks 276, 624 s t a r t and end of an object 261,267
delimiters 622 summary of graphics primitives 622
describing single objects 255 turning a comer 266
diagonal lines 262 typical figure description 254
dimension variables 278 units of measure 255
double-headed arrow 267 use of object blocks 277
drawing in clockwise direction 263 used with t r o f f 253, 257
drawing motion 258 user-defined variables 285
enhancements to 291 using bit-mapped input 292
executing UNIX commands from 291 pica (defined) 76
expressions 284, 626 pipes 21-22
f o r loops 291 .pl request 86, 88, 496
functions 284 - pm request 414
height of object 256 .pn request 9 1
i f conditional statements 29 1 .POrequest 78,485
inability to use with n r o f f 253 point size 62, 76, 97-98, 116-1 17
invisible reference object 268 changing (eqn) 249
labeling objects 257, 624 changing (mm) 143-144
language of 253 changing (ms) 116-117
leaving space between objects 259 changing ( t b l ) 216
library of frequently used objects 289
line 257
 0 Index 0 659

postprocessors 460 .RF macro (mm) 171-172

PostScript 67, 460, 474 . R I macro (mm) 143
.P P macro ( m s ) 106-107, 121 .RL macro (mm) 158
p r command 299 .r m request 413, 51 1
predefined conditions - r n request 415, 512
in t r o f f 4 1 9 root directory 15
printers (types of) 8-9 .RP macro (mm) 132, 172-173
problems on first page (ms) 127 .R S macro (mm) 171-172
. p r o f i l e 343 .RS macro (ms) 1 I 1
p r o o f shell script 304, 380-386 . r s request 90, 127, 490
proofreading .RT macro (ms) 106, 487
d o u b l e auk script 395 . r t request 461, 505
shell script for 380-386
.p s request 97, 9 9
pseudo-page transition 49 1 S
- P T macro (ms) 492
pwdcommand 16 . S macro (mm) 143-144
. SA macro (mm) 137
SCCS (Source Code Control System) 319
Q s c r i p t command 341
scrolling 4, 42-43
.QE macro ( m s )
110 s d i f f command 312, 315-316, 318-319
-QP macro (ms) 106-107, 110 search (see also g r e p ) 22, 45, 181, 301
.QS macro (ms) 110 search path 344
section headings 512, 521
sed editor 5, 342, 360-380, 585-587
R addressing 361-362, 585
branching to parts of script 366, 369
.R macro (mm) 141-142 command summary 585-587
.R macro (ms) 114-1 15 command syntax 360,584
- RB macro (mm) 143 differences from ex 361
.r d request 477 excluding lines from editing 365-366
.RE macro (ms) 1 1 1 hold space 375-379
read shell command 400 in f o r m a t script 364
recommended readings 646 inserting lines of text 369
records (in a w k ) 389 invoking 360-361, 584
redefining control and escape characters 435 matching patterns across 2 lines 374-379
reference lists (mm macros) 158 pattern space 375-379
regular expressions 184-188, 568 print command 370
reset macro 106, 487 quit command 373
returning to a horizontal position 464 script for extracting information from a file
returning to a marked position 370
in t r o f f 461 substitute command 363
in v i 57 used in f o r loop 363
660 0 UNlX Text Processing 0

used in indexing script 557-558 defining 429

s e t command 2 0 in t r o f f 101
s h command 343 interpolating 102, 430
. S H macro (ms) 120-122 mm 175
shell scripts 23, 343-346, 354 multiline 102
breaking out of 352 naming 430
conditional execution 347-349 subdirectory 15
discarding and shifting arguments 349 subscripts (see also e q n ) 237
looping 355 superscripts (see also e q n ) 2 3 7 , 4 4 6
number of arguments ( $ # ) 350 super-user 298
repetitive execution 350, 355 .s v request 90-9 I
arguments to 23 .s y request 480, 539
C shell search p a t h 345
definition of 343
e x p o r t command 352 T
initializing variables 352
making executable 344 .t a request 467, 82-83
passing arguments to 345-346 table of contents 532, 544
proofreading script 380, 386 created by diversion 532, 536
putting in path 344 mm macros 168-170
reading data interactively 400 written to standard error 532, 538
setting default values 352 tabs (in t r o f f ) 82-83, 467
t e s t command in 347-349 t a i l command 300
s h i f t shell command 349 t a r command 341
. S K macro (mm) 150 .TB macro (mm) 133. 169-170
. S M macro (mm) 144-145 t b l preprocessor 203-231
. S M macro (ms) 116-1 17 alphabetic data columns 2 12
. s o request 103, 438, 479, 484 breaking up long tables 224
s o e l i m command 480 changing format within table 219
sort command 22, 323, 3 2 5 , 5 4 9 column format options 209, 6 I7
.SP macro (mm) 136- 137 column width 217, 219
.s p request 73. 84-85 complex table example 227
space at top o f page 490 data 205-206, 6 18
s p e l l command 296, 304, 380 describing column formats 209
s p l i t command 333 drawing lines within tables 2 14
S Q t r o ff 439 equations within tables 2 I2
square root signs (see e q n ) fonts 216
.s s request 47 1.472 format options 205-207
standard error 14, 538 global format options 205-208, 616
standard input 2 1 headers 210
standard output 14, 2 I horizontally spanning headers 21 1
strings invoking 204
appending to 430 numeric data columns 2 I2
comparing 4 2 I point sizes 216
 1
Index 661

problems with .so request 480 .as request 430

putting text blocks in a column 221 autoincrementing number registers 428
repeating table headers 224 basic assumptions 59
simple table example 206 .b d request 463
spacing within tables 21 1 boldfacing fonts by overstriking 463
staggered columns 218 .bp request 86, 89-91
summary of commands 616-618 .br request 432, 69
table end macro 205 bracket-building characters 466, 602
table formatting checklist 226 .c2 request 435
table specifications 204 . cc request 435
table start macro 205 .ce request 80-82
titling tables 225 .cf request 460
using eqn with 205, 234 - c h request
503
vertical spacing within data blocks 218 change bars 476
vertically spanned columns 213 changing page size 495
- T C macro (mm) 168 character output translations 473
- t c request 469 command line options 594
TERM variable 20, 27 comments 412
terminal messages from t r o f f 438, 440 comparing strings 421
terminal type 20, 27, 33 compiling font files 638
test command 347-349 conditional execution 418
.t i request 79 constant spacing 471
t i p command 339 copy mode 417
title length 495 .cs request 471
.T L macro (ms) 122-123 .cu request 83-84
.t 1request 493 cut marks 498-499
.t m request 438,440,502,532, 539 .da request 433, 534
t p l u s postprocessor 471 .de request 413, 99
t r command 332 debugging 436-438
.t r request 436, 473-474 default units 77
transparent output 460,534 defining macros 413, 87, 99
traps 89, 485,489, 491-492,495, 503 defining strings 429
trof f formatter 58-103, 412-480 device units 76
.a b request 440 - di request 431, 451
aborting 440 diversions 45 1
absolute motions 448, 464 double or triple spacing 86
.ad request 485, 71-72 downloadable fonts 66
adjusting title length 495 drawing 449-455
aligning numeric data 83 .ds request IO 1,429
.am request 4 13 .ec request 435
appending to a diversion 433 .el request 418-423
appending to a macro 4 13 .em request 534
appending to a string 430 environment switching 433
arithmetic expressions 419 .eo request 435
662 0 UNlX Text Processing 0

error handling 439 - 1s request 86, 99

escape sequences 67,444, 598-599 It request 495
.ev request 433 macro arguments 100,416
.ex request 478 macro names 414
executing system commands from 480 macro style 441
expression operators 420 marking a horizontal position 464
.f c request 469 marking a vertical position 461
.f i request 69 mathematic symbols 602
fields 469 .mc request 476
.fl request 127,438,491 - m k request 461, 505
flushing output buffer 438 multi-column processing 461, 505-506
fonts 65, 92-93 multiline conditions 421
footnotes 500-501 .na request 71
form letters 477 names of existing macros 414
.fp request 92, 95 names of existing number registers 415
.ft request 92, 95 ne request 508, 90
Greek characters 604 negative vertical motions 85
headers and footers 493 nested macros 41 8
horizontal spacing 76 .nf request 69
- hw request 73-74 .nm request 475
- h y request 7475,485 .nn request 476
hyphenation 73-75,485 .n r request 101,424
.ie request 418-423 .ns request 127, 490, 90-91
.if request 418-423 number registers 100-101, 419-425, 485,
.ig request 438 526,529
ignoring input 438 numeric expressions 77
- in request 79 .nx request 478
including Macintosh illustrations 460 .o s request 90-91
interactive use 437 output line numbering 475
interrupted lines 423 overstriking 462, 465
interword spacing 62 page breaks 508
inverse conditions 423 page breaks without line breaks 89
invoking 66, 594 page layout 15,492
justification using m m 137 page length 86, 88
keeping text block together 43 1 page numbering 91
- IC request 468 page numbers 494
leaders 468-469 page offset 523
.1g request 474-475 page top resets 507
ligatures 473, 475 page transition 86-87, 483-508
line drawing 449 .pc request 494
line weight 454 .pl request 496, 86, 88
- 1 1 request 78, 485 .pm request 414
local horizontal motions 447 .pn request 91
local vertical motions 445 - P Orequest 485, 78
 0 Index 663

point size 65, 97-98 .t l request 493

postprocessors 460, 66-67 - t m request438,440,502,532
predefined conditions 4 1 9 .t r request 436, 473-474
predefined number register names 426,599- transparent output 460
600 traps 87,485,489, 491-492, 495, 503
- p srequest 97, 99 two-column processing 461
pseudo-page transition 49 1 .ul request 83-84
- r d request 477 underlining 449-450
reading standard input 477 units of measure 76-77
read-only number registers 427, 599 used with laser printers 9
redefining control and escape characters 435 using with m s 105
removing macros 4 13 using p i c with 253, 257
removing number registers 429 versions of 64, 66
renaming macros 415 vertical spacing 6 5 , 7 6 , 84, 98
returning to a horizontal position 464 vertically stacked labels 467
returning to a vertical position 46 1 .vs request 98-99
.r m request 413, 51 1 -wh request 485,489,503, 87
.r n request 415, 5 1 2 widows and orphans 508
.r s request 127, 490,90 width function 464
.r t request 461, 505 two-column processing,
selecting output pages from command line m s macros 125- 126
92 simple macro for 505-506
setting page number from command line 91
.s o request 103,438,479,484
.s p request 73, 84-85 U
space at top of page 490
space size 472 .U L macro (ms) 116
spacing to an absolute position 85 .u l request 83-84
special characters 96, 602-604 underlining 1 15
.ss request 47 1-472 in t r o f f 449-450
stacking up characters 466 ms macros 116
summary of requests 594-598 uniq command 323, 325
superscripts 446 UNIX
suspending line numbering 476 syntax of commands 13
- sv request 90-91 system fundamentals 12
.s y request 480 version used for this book 12
syntax of requests 67 unpack command 338
.t a request 82-83, 467 unpaddable space 112-1 13,474, 73
tabs and leaders 467 user exit macros (mm) 176
tabs in macros 470 / u s r / l i b / f o n t 63, 65, 93, 635-639
.t c request 469 DESC file 635-637
terminal messages 438 font description files 637-638
three-part titles 494
. t i request 7 9
664 0 UNIX Text Processing 0

/usr/lib/tmac 103 marking place in file 57

u u c p command 339 movement by line number 48
u u n a m e command 339 moving by screenfuls 42
moving cursor by single lines 30
moving cursor by spaces 30
V moving cursor by text blocks 31
moving text 32, 38, 54
variable-item lists named buffers 55-56
mm macros 158, 161-162 numbered buffers 54-55
vertical spacing 76, 84, 98, 107, 136, 143 numeric arguments to commands 30
vertically stacked labels 467 numeric prefixes to commands 53
v i editor 24-54, 177-202 on a dumb terminal 33.37
abbreviations 198 opening a file 25-26
alternative insert commands 54 opening a file to a specific place 49
appendtext 54 opening a new line for insertion 54
appending text 54 pattern matching characters 222
appending to named buffers 56 prompt line 26
changing text 34-35 quitting 27
characters not used in command mode 574 quitting without saving edits 41
command line options 49 read-only mode 50
command mode 27 recovering a buffer 50
command summary 572,578 recovering deletions 55
command syntax 27, 571 repeat last search 46
copying text 39, 54 repeating last command 40
current and alternate filenames 197 replacing characters 35, 54
cursor movement 28, 30 returning to a position 48
cursor movement by line numbers 47 saving a file 28
cursor movement by text blocks 44 screen lines vs. logical lines 30
cursor movement with numeric argument 45 scrolling 42-43
cursor movement within lines 44 search for pattern 45-47
cursor movement within screen 43 search options 53
deleting single characters 37-38 search within current line 47
deleting text 32, 35-37 setting options 50-51, 53, 574
displaying line numbers 3 1, 48 shiftwidth 53
editing multiple files 195 showing contents of numbered buffers 55
errors when opening 27 size of window 51
ex commands in (see also e x ) 180 status line 572
filtering text through a UNIX command 193 summary of options 570
ignoring case during searches 5 1 undoing last change 40, 55
insert mode 24, 27 v i e w m o d e 50
inserting text 27, 32 wrapmargin 28, 53
joining lines 41 v i e w command (see v i editor) 50
leaving insert mode 27, 33
mapping command sequences 198
 0 Index 665

. V L macro (mm)158, 161-162

.vs request 98-99

wccommand 21
.w h request 485, 489, 503, 87
w h i l e shell command 350
who command 22
widows and orphans 508
width function 464
word processors
characteristics of 2-5
command mode vs. insert mode 4
influence on writing process 1-2
limitations of 3, 7
vs. text editors 3
Writer’s Workbench 308
analyze style/readability (style) 308
explain diction errors (explain) 31 1
miscellaneous programs 3 1 1
search for poor phrasing ( d i c t i o n ) 309
wysiwyg defined 6

xargs command 303

 O’REILLY & ASSOCIATES, INC.
981 Chestnut Street + Newton, Massachusetts 02164 + 617-527-4210

X WINDOW
~
SYSTEM PROGRAMMING MANUAL
+ Volume 1: Xlib Programming Tutorial with Example Programs
4 Volume 2: Revised & expanded Xlib man pages with permuted
index and quick reference aids
4 Troff source files available for licensing; generic books for sale
+ Licensed by Stellar, Masscomp, Silicon Graphics, Motorola

NUTSHELL
-
- HANDBOOKS
+ Get New Users Up to Speed
Learning the UNlX Operating System
Learning the Vi Editor

+ Single-session mastery of advanced UNlX topics

Reading and Writing Termcap Entries
Programming with Curses
Managing Projects with Make

4 UUCP Communications for administrators and users

Managing UUCP 24 Usenet
Using UUCP & Usenet

4 UNlX in a Nutshell-a complete desktop quick reference

Specify System V or BSD version

+ DOS Meets UNlX

DOS/UNIX Merge’”, PC-Interface’” , VP/iT, MKS Toolkit TM,

and more. How DOS and UNIX can work together.

POSTSCRIPT TYPESETTING SERVICES

Our Linotronic L100 (1270 Ipi) supports PostScript, the device-

independent page description language. We can accept PostScript
files via a uucp link or on PC or Macintosh disks.

TO ORDER: 1-800-338-NUTS