Enterprise COBOL Concepts

Dr. David Woolbright

[email protected]
2013

Why Cobol?
Billions of lines of existing code with more
added each year
Designed for business
Great compilers
Runs fast
Relatively simple to learn
The language keeps evolving

Enterprise Cobol for z\OS

IBM COBOL:
http://www-01.ibm.com/software/awdtools/cobol/zos/
IBM COBOL Documentation:

http://www-01.ibm.com/software/awdtools/cobol/zos/library/

Languagage Reference
http://publibfp.boulder.ibm.com/epubs/pdf/igy3lr50.pdf
Programming Guide
http://publibfp.boulder.ibm.com/epubs/pdf/igy3pg50.pdf

Program Organization

Program Organized like a book

Division Identification, Environment, Data, Procedure
Section
Paragraph
Sentence
Clause
Phrase
Word

Grammatical Hierarchy

The grammatical hierarchy follows this form:

Identification division

Paragraphs

Entries

Environment division

Sections

Paragraphs

Entries
Clauses
Phrases

Data division

Sections

Entries

Clauses

Clauses
Phrases

Procedure division

Sections

Paragraphs

Sentences
Statements
Phrases

Coding Rules

Cols 1-6 left blank. Compiler fills in with sequence numbers

Col 7 Usually blank,* means comment line, - is continuation, D for debugging lines
Cols 8-11 A margin or Area A
Cols 12-72 B margin or Area B
Cols 73-80 unused
1 2 3 4 5 6|7| 8 9 10 11|12 13 71 71 |
Seq Nos | | Area A | Area B
|

Continuation of Statements

Statements can be continued on the next

line in Area B

Continuation of Literals

Continue the constant through column 71

Put a - in column 7
Continue constant with a OR
Continue constant in area B

Things That Go in Area A

Area A items:

Division headers
Section headers
Paragraph headers or paragraph names
Level indicators or level-numbers (01 and 77)
DECLARATIVES and END DECLARATIVES
End program, end class, and end method markers

Things That Go in Area B

Area B items:
Entries, sentences, statements, and clauses
Continuation lines

Things That Go in A or B

Area A or B
Level-numbers
Comment lines
Compiler-directing statements
Debugging lines
Pseudo-text

Structure of a Program

IDENTIFICATION DIVISION
IDENTIFICATION DIVISION.
PROGRAM-ID. HELLO.
AUTHOR. JOE SMITH.
INSTALLATION. TSYS.
DATE-WRITTEN. 12/03/2011.
DATE-COMPILED. 12/03/2011.

Only PROGRAM-ID is required

Some interesting parms can be coded on the
PROGRAM-ID

ENVIRONMENT DIVISION
This division connects external DD file names with internal file names.
ENVIRONMENT DIVISION.
INPUT-OUTPUT SECTION.
FILE-CONTROL.
SELECT MSTRFILE ASSIGN TO MSTRFILE
SELECT CUSTOMER-FILE
ASSIGN TO CUSTMAST
ORGANIZATION IS INDEXED
ACCESS MODE IS RANDOM
RECORD KEY IS COSTOMER-KEY
FILE STATUS IS CUSTOMER-FILE-STATUS.

Internal File
Name
External DD
File Name

DATA DIVISION
Used to create variables and constant fields
Only three data types
numeric
PIC 99999.
alphanumeric (text/string) PIC XXX.
alphabetic
PIC AAA.
Level numbers indicate subordination of fields. Use
levels 01-49
Alphabetic is seldom used

Level Numbers

Group item a subdivided field

Elementary item a non-subdivided field
01 Group or independent item
Higher numbers indicate subordinate fields

Level Numbers

66, 77, 88 have special significance

66 Used to rename (no longer used)
77 An independent item (choose 01)
88 Condition name

Level Numbers
01

77

XXX.
05 YYY.
10
10
05 ZZZ
AAA

AAA PIC X.
BBB PIC X.
PIC X(20).
PIC 999V99.

Condition Names
01

TRAN-CODE
PIC X.
88
GOOD
88
BAD
88
INDIFFERENT

SET GOOD TO TRUE

Equivalent to
IF (GOOD)

VALUE G.
VALUE B.
VALUE I.

MOVE G TO TRAN-CODE

Equivalent to IF TRAN-CODE = G

Level 88

Condition Names

Picture Clauses
Picture clause values usually use 9, X, V,
S, A
9 a decimal digit
X any alphanumeric character
V an implied decimal point
S a sign
A A-Z, and blank

Picture Clauses

PIC 9(6)
PIC 9(6)V99
PIC 999999V99
PICTURE X(10)
PIC XXXXXXXXXX
PIC S9(4)V9(4)
PIC S9999V9999
PIC 9(18)

Numeric Edited Fields

XXXBXXBXXXX
99/99/99
ZZ,ZZZ.99DB
***,***.99
----.99
$$$9.99
99999.99

USAGE Clause
Specifies the format in which data is stored in
memory
Normally, the phrase USAGE IS is omitted
01 COST
01 COST

USAGE IS PACKED-DECIMAL PIC S9(5).

PACKED-DECIMAL PIC S9(5).

01 FIRST-NAME
01 FIRST-NAME

USAGE IS DISPLAY PIC X(20).

PIC X(20).

DATA DIVISION
We define data used in input-output operations.
FILE SECTION.
FD CUSTOMER-FILE.
01 CUSTOMER-MASTER.
05 CUST-NUM
05 CUST-FNAME
05 CUST-LNAME
FD SALES-REPORT.
01
REPORT-AREA
PIC

PIC 9(2).
PIC X(20).
PIC X(20).
X(132).

Data Formats
Older terms:

Modern terms:

BINARY
BINARY
FLOATING POINT
FLOATING POINT
PACKED-DECIMAL
BINARY
BINARY (NATIVE)
PACKED-DECIMAL.
BINARY.

COMPUTATIONAL
COMP
COMP-1
COMP-2
COMP-3
COMP-4
COMP-5
05 XDATA PIC S9(5)
05 YDATA PIC S9(4)

EBCDIC
EBCDIC is an IBM format for storing alphanumeric
characters
A - XC1
J - XD1
B - XC2
K - XD2
S XE2
C - XC3
L - XD3
T XE3
D - XC4
M - XD4
U XE4
E - XC5
N - XD5
V XE5
F - XC6
O - XD6
W XE6
G - XC7
P - XD7
X XE7
H - XC8
Q - XD8
Y XE8
I - XC9
R - XD9
Z XE9

EBCDIC
EBCDIC is an IBM format for storing alphanumeric
characters
0 - XF0
SPACE X40
1 XF1
.
- X4B
2 - XF2
,
- X6B
3 XF3
*
- X5C
4 - XF4
- X60
5 XF5
6 - XF6
7 XF7
8 XF8
9 XF9

BINARY DATA
Stored in 2s Complement format
Leftmost bit is a sign ( 0 +, 1 - )
If the number is positive, interpret it as plain binary
01011 = 8 + 2 + 1 = + 11
If the number is negative, compute the complement
Invert. (Change all 1s to 0s and 0s to 1s.) Add 1. The
result is the additive complement

BINARY DATA
10011 is a negative number.
Inverting we have 01100.
Adding 1 we have 01100 + 1 = 01101. This is a positive
number. 01101 8 + 4 + 1 = 13, so the original number
is -13.

BINARY DATA
Declaring a data field as BINARY causes the
data to be stored in 2s complement format.
01

X-FIELD

PIC S9(4) BINARY VALUE -1.

X-FIELD will contain

1111111111111111 = XFFFF.
Binary data is processed in a register

PACKED-DECIMAL DATA

Defining a field to be PACKED-DECIMAL or Computational-3

causes the data to be stored internally in a packed decimal format.

There are 2 decimal digits stored in each byte. A sign is stored in

the rightmost nibble. (C +, D -)

Y-FIELD PIC S999 VALUE -23 PACKED-DECIMAL.

produces a 2 byte field containing X023D

Most business arithmetic occurs in packed decimal.

Packed No-Sign Data

Packed no-sign data is a non-native data
type that was created to save space on a
disk when storing dates
If the digits in a date like 10/23/89 were
stored in a packed field, the field would
require 4 bytes: 01|02|38|9C
By removing the sign, the date fits in 3
bytes: 10|23|89

Packed No-Sign Data

These type fields require special handling
in Cobol
Program PKNOSIGN illustrates how a
displayable date can be recovered from a
packed no-sign field.

ZONED-DECIMAL DATA
A numeric field which is described as DISPLAY, or in
which the usage clause is omitted, is stored in a zoned
decimal format.
In zoned decimal, each digit takes up one byte, and a
sign is stored in the zone portion of the rightmost byte of
the field.
Z-FIELD PIC S999 VALUE -32
produces a 3 byte field containing XF0F3D2.

ZONED-DECIMAL DATA
Z-FIELD PIC S999 VALUE -32.
produces a 3 byte field containing XF0F3D2.
Z-FIELD PIC S999 VALUE 32.
produces a 3 byte field containing XF0F3C2.
W-FIELD PIC 999 VALUE 0.
MOVE -32 TO W-FIELD
produces a 3 byte field containing XF0F3C2.

DATA DIVISION
Define the data needed for internal processing in the
WORKING-STORAGE SECTION.
Storage is statically allocated and exists for the life of the
run unit.
WORKING-STORAGE SECTION.
01 TOTAL-FIELDS.
05 CUST-TOTAL
05 COST-TOTAL
01 DATE-AND-TIME.
05 CD-YEAR
05 CD-MONTH

PIC S9(7)V99 VALUE 0.

PIC S9(7)V99 VALUE 0.
PIC 9999.
PIC 99.

DATA RELATIONSHIPS
BINARY

PACKEDDECIMAL

CHARACTER or
ALPHANUMERIC

ZONEDDECIMAL

DATA DIVISION

Describe data that exists in another program,

or storage you want to associate with a
symbolic name in the LINKAGE SECTION.
LINKAGE SECTION.
01 LK-DATA-AREA
05
NAME
05
AGE

PIC X(40).
PIC 999.

DATA DIVISION
The LOCAL-STORAGE SECTION is used to have
storage allocated each time a program is
entered, and deallocated on return from the
program. Used for compatibility with C or Java.
LOCAL-STORAGE SECTION.
01 CUST-NO
PIC X(3).
01 COST
PIC 9(5)V99.

Initialization of Storage
WORKING-STORAGE for programs is allocated at
the start of the run unit.
Any data items with VALUE clauses are
initialized to the appropriate value at that time.

Initialization of Storage
For the duration of the run unit, WORKING-STORAGE
items persist in their last-used state. Exceptions are:
1) A program with INITIAL specified in the PROGRAMID paragraph In this case, WORKING-STORAGE data
items are reinitialized each time the program is entered.
IDENTIFICATION DIVISION.
PROGRAM-ID. MAIN IS INITIAL.
...

Initialization of Storage
For the duration of the run unit, WORKING-STORAGE
items persist in their last-used state. Exceptions are:

2) A subprogram that is dynamically called and then

canceled In this case, WORKING-STORAGE data items are
reinitialized on the first reentry into the program following
the CANCEL.
MOVE PROG23 TO PROGID
CALL PROGID
CANCEL PROGID
CALL PROGID

Group and Data Items

01 Customer-Record.
05 Customer-Name.
10 Last-Name Pic x(17).
10 Filler Pic x.
10 Initials Pic xx.
05 Part-Order.
10 Part-Name Pic x(15).
10 Part-Color Pic x(10).

REDEFINES
01 MONTH-NAMES.
05 STRING-1 PIC X(15)
VALUE JANFEBMARAPRMAY.
05 MONTH REDEFINES STRING-1
OCCURS 5 TIMES PIC XXX.
MOVE MONTH(3) TO MONTH-OUT

REDEFINES
05
05

10
10

AMOUNT
AMOUNTX

XFIELD
YFIELD
20 A
20 B

PIC ZZ9.9-.
REDEFINES AMOUNT
PIC X(6).
PIC 9(5).
REDEFINES XFIELD.
PIC X(3).
PIC X(2).

Literals
String Literals enclosed in quotes (single
or double)
MOVE "INVALID" To CUST-NAME

Numeric literals without quotes

MOVE 19 TO CUST-AGE

Literals
Hexadecimal literals with X
MOVE XAF3B TO CUST-CODE

Constants
A constant is a data item that has only one value and it
can never change
Unfortunately, COBOL does not define a construct
specifically for constants
Moral: All values are subject to change
Data Division.
01 Report-Header pic x(50)
value "Company Report".
01 Interest
pic 9v9999
value 1.0265.

Figurative Constants
There are some figurative constants supplied by the
language:

ZERO
- an appropriate form of 0
SPACE
- x40
HIGH-VALUES - binary 1s
LOW-VALUES - binary 0s
QUOTE
- a single quote
NULL
- binary 0s used for pointers
ALL - Technically not a figurative constant:
X PIC X(5) VALUE ALL 3.

Tables (Arrays)
A table is a set of logically consecutive data items that
you define in the DATA DIVISION by using the
OCCURS clause.
01 TABLE.
05 TABLE-ENTRY OCCURS 10 TIMES.
10
NUM
PIC 99.
10
NAME
PIC X(30).
10
ITEM
PIC X(5) OCCURS 3 TIMES.

Referencing a Table
01 TABLE.
05 TABLE-ENTRY
10
NUM
10
NAME
10
ITEM

OCCURS 10 TIMES.
PIC 99.
PIC X(30).
PIC X(5) OCCURS 3 TIMES.

COBOL tables are 1-indexed

Some valid References with subscripts:
TABLE-ENTRY(SUB)
TABLE
NUM(SUB)
NAME(SUB)
ITEM(SUB1,SUB2)

Subscripts vs Indexes
Subscripts are defined separately from the table
definitions.
01 MYTABLE.
05 ITEM PIC X(3) OCCURS 10 TIMES.
01 I
PIC 9(4) BINARY.
...
MOVE 1 TO I
MOVE ABC TO ITEM(I)
Subscripts are numeric fields choose BINARY fields for
efficiency, although packed and zoned fields also work

Subscripts vs Indexes
Subscripts can be easily printed
01
01

MYTABLE.
05 ITEM PIC X(3) OCCURS 10 TIMES.
I
PIC 9(4) BINARY.
...
MOVE 1 TO I
MOVE ABC TO ITEM(I)
DISPLAY I

Subscripts vs Indexes
Subscripts represent an occurrence number, 1 is
the first occurrence, 2 is the second,
01
01

MYTABLE.
05 ITEM PIC X(3) OCCURS 10 TIMES.
I
PIC 9(4) BINARY.
...
PERFORM VARYING I FROM 1 BY 1 UNTIL I > 10
DISPLAY ITEM(I)
END-PERFORM

Subscript Program

Subscripts vs Indexes
Indexes are created when a table is defined
01 MYTABLE.
10
10

LETTERVALS
PIC X(10) VALUE 'ABCDEFGHIJ'.
LETTER REDEFINES LETTERVALS
PIC X(1) OCCURS 10 TIMES
INDEXED BY I.

Indexes are manipulated with SET statements and

automatically altered by PERFORM statements
SET I TO 1
MOVE LETTER(I) TO CHAROUT

Subscripts vs Indexes
Indexes are generally more efficient than subscripts
Indexes represent offsets from the beginning of the table
SET I TO 1
Causes I to have the binary value 0 internally.
It takes a bit of work to print them

Index Program

Tables with Two Dimensions

01 TABLE.
05 TABLE-ENTRY OCCURS 10 TIMES
INDEXED BY SUB1.
10
NUM
PIC 99.
10
NAME
PIC X(30).
10
ITEM
PIC X(5) OCCURS 3 TIMES
INDEXED BY SUB2.
Valid References with subscripts:
TABLE-ENTRY(SUB1)
TABLE
NUM(SUB1)
NAME(SUB1)
ITEM(SUB1,SUB2)

Subscripts and Indexes

In a two-dimensional table, the two subscripts
correspond to the row and column numbers.
In a three-dimensional table, the three
subscripts correspond to the depth, row, and
column numbers.
Indexes use address computation to efficiently
reference items in a table.

A Testy Program

This program displays You Chose Paper.

Why?

PROCEDURE DIVISION
The PROCEDURE DIVISION is where you code the
executable statements in your COBOL program
Divided into Paragraphs (terminated with periods):
100-MAIN.
DISPLAY HELLO
PERFORM 200-SUB
GOBACK
.
200-SUB.
DISPLAY WORLD!
.

The Evils of Coding Periods

Periods have proven troublesome in the
procedure division, and a new style of coding
has evolved to deal with them
The period is a statement terminator but we
want to think of it as a paragraph terminator only
We will code in a period-less style
Only use periods to end a paragraph name

PROCEDURE DIVISION
To resolve ambiguity caused by not using
periods, we will use statement delimiters:
END-IF
END-PERFORM
END-COMPUTE
...

DISPLAY

DISPLAY
Handy for debugging and simple report
creation
DISPLAY X Y Z
DISPLAY "A = " A
Data is written to SYSOUT

MOVE

MOVE STATEMENT
Used to copy data from one field to
another
Example MOVE X-FIELD TO Y-FIELD Z-FIELD

Data is copied from the sending field to

the receiving field

MOVE STATEMENT
To move data from one field to another field, the two
fields should be compatible but dont have to be
identically pictured
Alphanumeric - PIC X(10)
Numeric
- PIC 999v99
Numeric-Edited - PIC 999.99Compatible moves:
-Alphanumeric to Alphanumeric
-Numeric to Numeric
-Numeric to Numeric edited

MOVE STATEMENT
Compatible moves:
-Alphanumeric to Numeric if the sending field is an
unsigned integer
-Alphanumeric to Numeric edited if the sending field is an
unsigned integer
-Numeric to Alphanumeric if the sending field is an
unsigned integer

MOVE STATEMENT
Compatible moves:
-Numeric edited fields can be sent to Numeric and
Numeric edited fields this causes a de-edit process to
occur
-Numeric edited fields can be sent to Alphanumeric and
Alphanumeric edited fields this causes a de-edit
process to occur

MOVE STATEMENT
Moving data can cause data conversions
to occur.
01
01

X
Y

PIC S9(4) BINARY.

PIC
S9(5) PACKED-DECIMAL.

MOVE X TO Y

MOVE STATEMENT
If the receiving field is larger than the sending
field, the receiving field is filled with leading 0s
in a numeric move:
01
01

X
Y

PIC S9(3) VALUE 123.

PIC S9(5) VALUE 0.
MOVE X TO Y
RESULT:
Y = XF0F0F1F2C3

MOVE STATEMENT
If the receiving field is larger than the sending
field, the receiving field is filled with trailing
spaces in a alphanumeric move.
01
01

X
Y

PIC X(3) VALUE ABC.

PIC X(5) VALUE SPACES.
MOVE X TO Y
RESULT:
Y = XC1C2C34040

MOVE STATEMENT
If the receiving field is smaller than the sending
field, data will be truncated on the left for
numeric moves and on the right for
alphanumeric moves
01
01
01
01

X
Y
A
B

PIC S9(5) VALUE 12345.

PIC S9(3) VALUE 0.
PIC X(5) VALUE ABCDE
PIC X(3) VALUE SPACES.
MOVE X TO Y
MOVE A TO B
RESULT:
Y = XF3F4F5
B = XC1C2C3

Use Move for Assignment

01

WORK.
05
A-FIELD
05
B-FIELD

PIC X(3).
PIC S999V99.

MOVE ABC TO A-FIELD

MOVE 123.45 TO B-FIELD
MOVE LOW-VALUE TO WORK

Assignment Can Cause

Conversions
Consider the following move:
01
01
01

A
B
C

PIC
PIC
PIC

MOVE A TO B
MOVE A TO C

S999V99.
ZZ9.99-.
S9(5)V9999 PACKED-DECIMAL.
Zoned to Numeric-edited
Zoned to Packed-decimal

MOVE CORRESPONDING
MOVE CORRESPONDING
identifier-1 TO identifier-2
CORR
Usually a bad idea
Both identifiers must name group items.
Elementary items with the same name are moved.

01

A-GROUP.
01 B-GROUP.
05 W
PIC X(3).
05
05 X
PIC X(2).
05
05 Y
PIC 999.
05
MOVE CORRESPONDING A-GROUP TO B-GROUP

W to W, X to X, Y to Y

W
X
Y

PIC X(3).
PIC X(2).
PIC 999.

MOVE CORRESPONDING
Subordinate items must not be identified by the keyword
FILLER
No reference modification for either identifier
Subordinate items must not include a REDEFINES,
RENAMES, OCCURS, INDEX or POINTER description
01

A-GROUP.
01 B-GROUP.
05 W
PIC X(3).
05
05 X
PIC X(2).
05
05 Y
PIC 999.
05
MOVE CORRESPONDING A-GROUP TO B-GROUP

W to W, X to X

P
X
W

PIC X(3).
PIC X(2).
PIC 999.

INITIALIZE

INITIALIZE
SPACE is the implied sending item for
receiving items of category alphabetic,
alphanumeric, alphanumeric-edited,
DBCS, national, or national-edited.
ZERO is the implied sending item for
receiving items of category numeric or
numeric-edited.

INITIALIZE
01 PRICE-FIELDS.
05 UNIT-PRICE
PIC 9(5)V9(2) PACKED-DECIMAL.
05 DISCOUNT
PIC V9(2).
05 UNIT-CODE
PIC XX.
05 SALES-PRICE
PIC S9(4) BINARY.
. . .
INITIALIZE PRICE-FIELDS

ADD

ADD Semantics
All identifiers or literals that precede the keyword
TO are added together, and this sum is added to
and stored in identifier-2. This process is
repeated for each successive occurrence of
identifier-2 in the left-to-right order in which
identifier-2 is specified.
ADD X Y Z TO P Q
Before X=1, Y=2, Z=3, P=4, Q=6
After X=1, Y=2, Z=3, P=10, Q=12

ADD EXAMPLES

ADD
ADD
ADD
ADD
ADD
ADD

X
X
X
1
X
X

TO Y
Y Z TO P
Y TO P Q
TO Z
TO Y ROUNDED
TO Y
ON SIZE ERROR
DISPLAY ADD ERROR
END-ADD

ADD GIVING

ADDGIVING Semantics
All identifiers or literals that precede the keyword TO are
added together, and this sum is added to identifier-2 to
obtain a temporary sum. (Identifier-2 is unchanged)
The the temporary sum is moved to identifier-3.

ADD X Y Z TO V GIVING P
Before X=1, Y=2, Z=3, V=4, P=6
After X=1, Y=2, Z=3, V=4, P=10

SUBTRACT

SUBTRACT
All identifiers or literals preceding the keyword FROM are
added together and their sum is subtracted from and
stored immediately in identifier-2. This process is
repeated for each successive occurrence of identifier-2,
in the left-to-right order in which identifier-2 is specified.

SUBTRACT X Y FROM P Q
Before: X=1,Y=2, P=3,Q=4
After: X=1,Y=2, P=0,Q=1

SUBTRACT

SUBTRACT Semantics
All identifiers or literals preceding the keyword FROM are
added together and their sum is subtracted from
identifier-2 to obtain a temporary value which is moved
to identifier-3.

SUBTRACT X Y FROM P GIVING Q

Before: X=1,Y=2,P=5,Q=6
After: X=1,Y=2,P=5,Q=2

MULTIPLY

MULTIPLY Semantics
In format 1, the value of identifier-1 or literal-1 is
multiplied by the value of identifier-2; the product is then
placed in identifier-2. For each successive occurrence of
identifier-2, the multiplication takes place in the left-toright order in which identifier-2 is specified.

MULTIPLY X BY P Q
Before: X=2,P=4,Q=5
After: X=2,P=8,Q=10

MULTIPLY

MULTIPLY
In format 2, the value of identifier-1 or literal-1 is
multiplied by the value of identifier-2 or literal-2. The
product is then stored in the data items referenced by
identifier-3. Identifier-2 is unchanged.

MULTIPLY X BY Y GIVING Z
Before: X=2, Y=3, Z=4
After: X=2, Y=3, Z=6

DIVIDE

DIVIDE
In format 1, the value of identifier-1 or literal-1 is divided
into the value of identifier-2, and the quotient is then
stored in identifier-2. For each successive occurrence of
identifier-2, the division takes place in the left-to-right
order in which identifier-2 is specified.
DIVIDE X INTO Y Z
Before: X=3, Y=7, Z=12
After: X=3, Y=2, Z=4

DIVIDE

DIVIDE
In format 2, the value of identifier-1 or literal-1 is divided
into the value of identifier-2 or literal-2. The value of the
quotient is stored in each data item referenced by
identifier-3.

DIVIDE X INTO Y GIVING Z

Before: X = 2, Y = 13, Z = 1
After: X = 2, Y = 13, Z = 6

DIVIDE

DIVIDE
In format 3, the value of identifier-1 or literal-1 is
divided by the value of identifier-2 or literal-2.
The value of the quotient is stored in each data
item referenced by identifier-3.
DIVIDE X BY Y GIVING Z
Before: X = 10, Y = 3, Z = 1
After: X = 10, Y = 3, Z = 3

DIVIDE

DIVIDE
In format 4, the value of identifier-1 or literal-1 is divided
into identifier-2 or literal-2. The value of the quotient is
stored in identifier-3, and the value of the remainder is
stored in identifier-4.

DIVIDE X INTO Y
GIVING Z
REMAINDER R
Before: X = 2, Y = 9, Z = 8, R = 7
After: X = 2, Y = 9, Z = 4, R = 1

COMPUTE

COMPUTE

05

COMPUTE can be used to initialize a numeric field

Usually reserved for nontrivial computations. For
simple computations choose ADD, SUBTRACT,
MULTIPLY or DIVIDE
X

PIC

S9(4)V9 PACKED DECIMAL.

COMPUTE X ROUNDED = (A + B) / 2.3

ON SIZE ERROR
DISPLAY X WAS TRUNCATED
END-COMPUTE

Arithmetic Operators
Operation

Operator

Addition

Subtraction

Multiplication

Division

**

Exponentiation

Parentheses provide precedence.

Always parenthesize!
((X + Y) * ( Z ** 3))

Locate Mode I/O

Your Program

Input Buffers

Output Buffers
Region

Move Mode I/O

READ MYFILE INTO
MYREC

WRITE RECOUT
FROM MYREC
01 MYREC PIC X(80).

Input Buffers

Output Buffers
Region

QSAM FILE OPERATIONS

ENVIRONMENT DIVISION.
INPUT-OUTPUT-SECTION.
FILE-CONTROL.
SELECT MY-INPUT-FILE
ASSIGN TO MASTER
FILE STATUS IS MAST-STAT.
MY-INPUT-FILE Internal file name
MASTER External DD name
MAST-STAT A two byte PIC XX field in which the
operating system returns a status code. Consult IBM
Language Reference page 315 for status code details.

QSAM FILE OPERATIONS

ENVIRONMENT DIVISION.
INPUT-OUTPUT-SECTION.
FILE-CONTROL.
SELECT MY-INPUT-FILE
ASSIGN TO MASTER
FILE STATUS IS MAST-STAT.
DATA DIVISION.
FILE SECTION.
FD
MY-INPUT-FILE
01
RECORD-AREA PIC X(80).

QSAM
Queued Sequential Access Method
For input files, records are buffered when the file
is OPENed
For output, records are buffered before being
written
Records are processed from the beginning
record sequentially to the end of the file
Very efficient access method for sequential files
Sometimes referred to as flat files

QSAM FILE OPERATIONS

Every file must be OPENed before it can
be processed.
Opening a QSAM Input file queues
records for subsequent read operations
OPEN INPUT MY-INPUT-FILE
OPEN OUTPUT MY-OUTPUT-FILE

Files should be closed when you have

finished processing the records
CLOSE MY-FILE

OPEN

CLOSE

READ

QSAM Input File Operations

Remember: READ a file, WRITE a record.
READ MY-INPUT-FILE
AT END MOVE NO TO MORE-RECS
END-READ

This is a locate-mode read and the most

efficient way to read records

File Reading
READ MY-INPUT-FILE INTO MY-REC
AT END MOVE NO TO MORE-RECS
END-READ WRITE

This is a move-mode read and the least

efficient way to deliver records

QSAM Output File Operations

Write a record!
WRITE MY-RECORD (locate mode)
WRITE MY-RECORD FROM REC-BUFF
END-WRITE
(move mode)
CLOSE MY-OUTPUT-FILE
CLOSE MY-INPUT-FILE

Sequential File Reading Pattern

READ MYFILE
AT END MOVE N TO MORE-RECS
END-READ
PRIMING READ
PERFORM UNTIL MORE-RECS = N
(process a record code)
READ MYFILE
AT END MOVE N TO MORE-RECS
END-READ
END-PERFORM
CONTINUATION
READ

File Status Codes

00
10
2x
3x
4x
9x

normal
end of file
invalid key
permanent i/o error
logic error
unsuccessful operation

Exercise #1
Create a file of 80 byte records
Each record has 3 fields
AFIELD ZONED DECIMAL
with 4 DIGITS & 2 DECIMALS
BFIELD PACKED DECIMAL
with 7 DIGITS & 3 DECIMALS
CFIELD - PACKED DECIMAL
with 7 DIGITS & 1 DECIMAL
Print a report with a column for each field and a column for the
computed value :
(AFIELD + BFIELD)/ CFIELD
Print the result with 2 decimals rounded.
Total each column.

FLOW OF CONTROL
There is a theoretical result in Computer Science by two
Italian mathematicians, Boehm and Jacopini, that states
that only 3 control structures are required to write any
program:
Sequence - Do this, now do this, now do this,
Selection - If something is true do this, else do that
Repetition While something is true, do this
Practice has shown that being able to create procedures
is helpful in overcoming complexity, but they arent
strictly necessary
One implication of this result is that GO TO statements
arent needed

FLOW OF CONTROL
F
?
T
T

IF

IF
The condition is tested and either the true
or false blocks are selected for execution
Dont use NEXT SENTENCE if you are
using END-IF as the delimiter (and you
should). Use of NEXT SENTENCE causes
execution to continue with the next closest
period, which is probably the end of the
paragraph.

IF Examples
IF

X < Y
ADD 1 TO X
DISPLAY AAA
ELSE
DISPLAY BBB
END-IF
IF

X > Y
DISPLAY X WAS BIGGER
END-IF

NESTED IFs
Each ELSE is matched with the nearest preceding IF
IF X < Y
DISPLAY XXX
IF Y < Z
DISPLAY ZZZ
ELSE
DISPLAY AAA
END-IF
MORAL: Indent properly and terminate all if statements
with END-IF

EVALUATE

EVALUATE
EVALUATE PLANET-NUMBER
WHEN 1 MOVE "Mercury" TO PLANET-NAME
WHEN 2 MOVE "Venus " TO PLANET-NAME
WHEN 3 MOVE "Earth " TO PLANET-NAME
WHEN 4 MOVE "Mars " TO PLANET-NAME
WHEN 5 MOVE "Jupiter" TO PLANET-NAME
WHEN 6 MOVE "Saturn " TO PLANET-NAME
WHEN 7 MOVE "Uranus " TO PLANET-NAME
WHEN 8 MOVE "Neptune" TO PLANET-NAME
WHEN 9 MOVE "Pluto " TO PLANET-NAME
WHEN OTHER MOVE " " TO PLANET-NAME
END-EVALUATE.

EVALUATE
EVALUATE PLANET-NAME
WHEN "Mercury"
WHEN "Venus "
WHEN "Earth "
WHEN "Mars
"
WHEN "Jupiter"
WHEN "Saturn "
WHEN "Uranus "
WHEN "Neptune"
WHEN "Pluto "
WHEN OTHER
END-EVALUATE.

MOVE
MOVE
MOVE
MOVE
MOVE
MOVE
MOVE
MOVE
MOVE
MOVE

1
2
3
4
5
6
7
8
9
0

TO
TO
TO
TO
TO
TO
TO
TO
TO
TO

PLANET-NUMBER
PLANET-NUMBER
PLANET-NUMBER
PLANET-NUMBER
PLANET-NUMBER
PLANET-NUMBER
PLANET-NUMBER
PLANET-NUMBER
PLANET-NUMBER
PLANET-NUMBER

EVALUATE
EVALUATE TRUE
WHEN PLANET-NAME = "Mercury" MOVE 1 TO PLANET-NUMBER
WHEN PLANET-NAME = "Venus " MOVE 2 TO PLANET-NUMBER
WHEN PLANET-NAME = "Earth " MOVE 3 TO PLANET-NUMBER
WHEN PLANET-NAME = "Mars " MOVE 4 TO PLANET-NUMBER
WHEN PLANET-NAME = "Jupiter" MOVE 5 TO PLANET-NUMBER
WHEN PLANET-NAME = "Saturn " MOVE 6 TO PLANET-NUMBER
WHEN PLANET-NAME = "Uranus " MOVE 7 TO PLANET-NUMBER
WHEN PLANET-NAME = "Neptune" MOVE 8 TO PLANET-NUMBER
WHEN PLANET-NAME = "Pluto " MOVE 9 TO PLANET-NUMBER
WHEN OTHER MOVE 0 TO PLANET-NUMBER
END-EVALUATE.

EVALUATE
EVALUATE Qty ALSO TRUE ALSO Member
WHEN 1 THRU 5 ALSO VOP < 501 ALSO "Y"
MOVE 2 TO Discount
WHEN 6 THRU 16 ALSO VOP < 501 ALSO "Y"
MOVE 3 TO Discount
WHEN 17 THRU 99 ALSO VOP < 501 ALSO "Y"
MOVE 5 TO Discount
WHEN 1 THRU 5 ALSO VOP < 2001 ALSO "Y"
MOVE 7 TO Discount
WHEN 6 THRU 16 ALSO VOP < 2001 ALSO "Y"
MOVE 12 TO Discount
WHEN 17 THRU 99 ALSO VOP < 2001 ALSO "Y"
MOVE 18 TO Discount
WHEN 1 THRU 5 ALSO VOP > 2000 ALSO "Y"
MOVE 10 TO Discount
WHEN 6 THRU 16 ALSO VOP > 2000 ALSO "Y"
MOVE 23 TO Discount
END-EVALUATE

EVALUATE
EVALUATE TRUE ALSO Position
WHEN L-Arrow ALSO 2 THRU 10
SUBTRACT 1 FROM Position
WHEN R-Arrow ALSO 1 THRU 9
ADD 1 TO Position
WHEN L-Arrow ALSO 1
MOVE 10 TO Position
WHEN R-Arrow ALSO 10
MOVE 1 TO Position
WHEN DelKey ALSO ANY
PERFORM DeleteChar
WHEN Char ALSO 1 THRU 9
PERFORM InsertChar
ADD 1 TO Position
WHEN Char ALSO 10
PERFORM InsertChar
WHEN OTHER PERFORM
DisplayErrorMessage
END-EVALUATE

PERFORM

PERFORM Paragraph
PERFORM paragraph name
Execute all instructions in the paragraph
Return control to the next instruction after the PERFORM

PERFORM 100-ROUTINE
PERFORM 200-ROUTINE
PERFORM 100-ROUTINE

100-ROUTINE.

200-ROUTINE.

300-ROUTINE.

PERFORM THRU

PERFORM paragraph name THRU paragraph name

PERFORM 100-XXX THUR 100-XXX-EXIT

100-XXX.
DISPLAY IN 100-XXX.
100-XXX-EXIT.
EXIT.
There is an implicit EXIT in every paragraph so why do I need to code it explicitly?

Perform Thru (Let it Die)

I view use of PERFORMTHRU as an older,
unnecessary style, caused by a compiler problem that
existed many years ago. Using PERFORM THRU was a
way to address this ancient problem. The problem has
been fixed for many years now, yet old habits die hard.

PERFORM WITH TIMES

PERFORM x TIMES
MOVE 5 TO COUNT
PERFORM COUNT TIMES
DISPLAY XXX
END-PERFORM
PERFORM 100-DISPLAY COUNT TIMES

PERFORM UNTIL

PERFORM UNTIL
MOVE 0 TO X
PERFORM UNTIL X > 10
MOVE X TO X-EDITED
DISPLAY X-EDITED
ADD 1 TO X
END-PERFORM
PERFORM X-PARA UNTIL X > 10
PERFORM X-PARA WITH TEST AFTER
UNTIL X > 10

PERFORM VARYING

Inline Perform
PERFORM VARYING X FROM 1 BY 1
UNTIL X > 100
DISPLAY X
END-PERFORM
PRINTS:
1
2
3

100

Inline PERFORM
PERFORM VARYING X FROM 5 BY -1
UNTIL X =0
DISPLAY X
END-PERFORM
PRINTS:
5
4
3
2
1
0

Inline PERFORM
MOVE 10 TO X
PERFORM WITH TEST AFTER
UNTIL
DISPLAY X
SUBTRACT 1 FROM X
END-PERFORM

X = 0

PERFORM PARAGRAPH
PERFORM 100-RTN
WITH TEST AFTER
VARYING X FROM 1 BY 1
UNTIL X = 100

100-RTN.
.

Inline PERFORM
MOVE ZERO TO Y
PERFORM UNTIL X = 0
READ AT END MOVE 0 TO X
ADD X TO Y
DISPLAY Y
END-PERFORM

Alternate PERFORM
PERFORM 100-PARA VARYING I FROM 1 BY 1 UNTIL I > 5
AFTER J FROM 1 BY 1 UNTIL J > 3
END-PERFORM
100-PARA.
DISPLAY I J
.
1 1
1 2
1 3
2 1
2 2
2 3
3 1
3 2
3 3
4 1

Table Processing with Subscripts

01 TOT
PIC S9(8) PACKED DECIMAL.
01 SUB1
PIC 99.
01 TEMP-REC.
05 TEMP OCCURS 12 TIMES
PIC S9(4).

MOVE 0 TO TOT
PERFORM VARYING SUB1 FROM 1 BY 1
UNTIL SUB1 > 12
ADD TEMP(SUB1) TO TOT
ENDPERFORM

Table Processing with Indexes

01 TOT
01 TEMP-REC.
05 TEMP

PIC S9(8) PACKED DECIMAL.

OCCURS 12 TIMES INDEXED BY K

PIC S9(4).

MOVE 0 TO TOT
PERFORM VARYING K FROM 1 BY 1
UNTIL K > 12
ADD TEMP(K) TO TOT
END-PERFORM

Manipulating Indexes
Indexes cant be manipulated with
ordinary arithmetic commands. Instead
use SET.

SET
SET
SET
SET

K
K
K
K

TO 3
UP BY 1
UP BY 2
DOWN BY 3

CONTINUE

Used to indicate that no operation is present.

EXIT

Sometimes used to provide an explicit exit from a

paragraph or a program.
There is an implicit exit at the end of each paragraph
100-PARA.
DISPLAY X
EXIT
.

STOP

Usually used to permanently halt execution of a

program
Can be used to temporarily halt execution and
require operator intervention

GOBACK

Functions like an EXIT PROGRAM when coded

at the end of a called program
Functions like STOP RUN when coded in a main
program
I prefer coding this in place of STOP RUN

GO TO

Causes an unconditional jump in program execution to the

procedure that is named.
This statement should be used only in very special situations, for
instance, to branch to an error routine that terminates the program
from a deeply nested area of your program.
Overuse (any?) of this statement is unnecessary and leads to
spaghetti code
Dont even think of using the alternate forms of GO TO !

Exercise #2

Create a file of 80 byte records

Each record has 3 fields
CUSTLNAME CHARACTER 15
CUSTFNAME - CHARACTER 15
BALANCE - PACKED DECIMAL 5 BYTES 2 DECS
Create a table of the file data using subscripts.
Sort the table with a bubble sort on customer name.
Print the sorted table:
Last Name First Name Balance

Exercise #3
Repeat Exercise #2 using indexes

SEQUENTIAL SEARCH

SEARCH
Search performs a sequential search with an index
Rule of thumb: Use SEARCH for tables with 20 items or less
DEPENDING ON field must contain the number of table entries
01
01

RECCOUNT
PIC 9(2).
SALES-TAX.
05 TAB-ENTRIES OCCURS 20 TIMES
DEPENDING ON RECCOUNT
INDEXED BY K.
10 ZIPCODE
PIC 9(5).
10 RATE
PIC V999.

SET K TO 1
SEARCH TAB-ENTRIES
AT END MOVE 0 TO TAX
WHEN ZIPIN = ZIPCODE(K)
COMPUTE TAX = RATE(K) * AMOUNT
END-SEARCH

Exercise #4
Read the file from exercise #3
Store the data in a table
Read the file BCST.SICCC01.PDSLIB(DAT4EXER).
Each record has a LAST NAME, FIRST NAME field:
Last name columns 1 15
First name columns 16 30
Code a Sequential Search statement to find each name
in the file. Print each name, Found or Not Found, and
the customer balance if found.

SEARCH ALL

SEARCH ALL

01
01

SEARCH ALL performs a binary search with an index

ENTRIES MUST BE IN ORDER BY KEY!
No SET necessary (the whole table is searched)
RECCOUNT must contain the number of table entries.
RECCOUNT
PIC 999 PACKED-DECIMAL.
SALES-TAX.
05 TAB-ENTRIES OCCURS 100 TIMES
DEPENDING ON RECCOUNT
ASCENDING KEY ZIPCODE
INDEXED BY K.
10 ZIPCODE
PIC 9(5).
10 RATE
PIC V999.

SEARCH ALL TAB-ENTRIES

AT END MOVE 0 TO TAX
WHEN ZIPCODE(K) = ZIPIN
COMPUTE TAX = RATE(K) * AMOUNT
END-SEARCH

SEARCH ALL CONSTRAINTS

The condition following WHEN must test for equality

Compound conditions with ANDs not ORs
Only one WHEN clause
VARYING not allowed
OCCURS item and its index must appear on the left of the
equal sign
WHEN TEMP(K) = 80

SEARCH ALL Constraints

Table must indicate ASCENDING or
DESCENDING KEY
01 TABLE.
05 CUST-REC OCCURS 40 TIMES
DEPENDING ON RECCOUNT
ASCENDING KEY CUST
INDEXED BY K.
10 CUST
PIC 9(4).
10 RATE
PIC V999.

Exercise #5
Read the file from exercise #3
Store the data in a table
Read the file BCST.SICCC01.PDSLIB(DAT4EXER).
Each record has a LAST NAME , FIRST NAME fields
Last name columns 1 15
First name columns 16 30
Code a binary search statement to find each name in the
file. Print each name, Found or Not Found, and the
customer balance if found.

STRING

STRING
Used to build string expressions by
concatenation (blanks not stored)
STRING FNAME DELIMITED BY
MNAME DELIMITED BY
LNAME DELIMITED BY
INTO NAME-OUT

STRING
Blanks stored
DELIMITED BY SIZE means include the entire literal
or variable contents
STRING FNAME DELIMITED BY
DELIMITED BY SIZE
MNAME DELIMITED BY
DELIMITED BY SIZE
LNAME DELIMITED BY
DELIMITED BY SIZE
INTO NAME-OUT

STRING

UNSTRING

UNSTRING
Separates a string into several component
strings
Sending field must be alphanumeric
UNSTRING NAME
DELIMITED BY ,
INTO LNAME
FNAME
MI
END-UNSTRING

UNSTRING
UNSTRING JCL-DATA
DELIMITED BY ALL SPACES OR ALL ','
INTO
WS-DATE-REQUESTED
WS-DATE1
WS-DATE2
END-UNSTRING

Exercise #6
Read the file
BCST.SICCC01.PDSLIB(STRINGS)
Print the first name, middle initial, and last
names in columns

Exercise #7
Read the file
BCST.SICCC01.PDSLIB(STRINGS1)
Print the digits followed by the letters. Use
/ as the delimiter of the two fields.

CALL

Static and Dynamic Call

Programs A and B that are linked together
prior to being loaded are statically linked.
If A calls B, the call is static
If programs A and B are separately
compiled and linked, A can call B
dynamically:
01 PGMNAME
PIC X(8) VALUE B.

CALL PGMNAME

Calling Other Programs Statically

CALL literal program name
USING identifier-1,
Examples
CALL CUST1030 USING X,Y
CALL PROG1000

CALLING ANOTHER PROGRAM

CALL PROGXXXX USING A,B
IDENTIFICATION DIVISION.
PROGRAM-ID. PROGXXXX.

LINKAGE SECTION.
01 X
PIC
X(5).
01 Y
PIC
999V99.
PROCEDURE DIVISION USING X,Y.

GOBACK
.

Static and Dynamic CALLS

Two methods for generating dynamic

calls
1. CALL MYPROG USING X,Y,Z
(Using an identifier insures a dynamic
call.)
2. DYNAM/NODYNAM compiler option
determines whether a static or dynamic
call occurs
3. At TSYS all calls are dynamic.

Exercise #8
Write a main program that,
1) Prints I am in the main program,
2) Calls your subprogram,
3) Prints I am back in the main program
Write a subprogram that prints I am in the
subprogram.
Compile and link the programs, execute
the main program

Exercise #9
Write a main program that passes 3
packed decimal numbers to a subprogram
and a fourth variable in which to receive a
result.
Write a subprogram that accepts the 3
integers, computes their sum and returns
the answer to the main
Have the main print the result

Exercise #10
Write a main program that passes a
variable X by reference and Y by content
Have the subprogram add one to both
numbers
Have the main program print X and Y after
returning from the subprogram

NUMERIC Class Test

Before using a suspect field that has a PIC of
9s, use the NUMERIC class test to verify the
field before doing arithmetic
IF AMOUNT-IN IS NUMERIC
ADD 1 TO AMOUNT-IN
ELSE
DISPLAY AMOUNT IS NOT NUMERIC
END-IF

Sign Test
Numeric data can be tested for positive, negative, and
zero values
IF AMOUNT-IN IS POSITIVE
ADD 1 TO AMOUNT-IN
END-IF
IF AMOUNT-IN IS NEGATIVE
DISPLAY AMOUNT-IN
END-IF
IF AMOUNT-IN IS ZERO
DISPLAY THE FIELD IS ZERO
END-IF

INSPECT (TALLYING)

INSPECT (TALLYING)
INSPECT MYLINE
TALLYING ECOUNT FOR ALL E
AFTER INITIAL START"
BEFORE INITIAL END
END-INSPECT
INSPECT WORK TALLYING
COUNT1 FOR LEADING *
COUNT2 FOR CHARACTERS
END-INSPECT
INSPECT WORK TALLYING
COUNT1 FOR ALL * BEFORE .
COUNT2 FOR ALL CHARACTERS AFTER .
END-INSPECT

INSPECT (REPLACING)

INSPECT (REPLACING)
INSPECT MYDATA REPLACING ALL X" BY Y
AFTER INITIAL A"
BEFORE INITIAL Z
INSPECT MYDATA REPLACING LEADING " BY +
INSPECT MYDATA REPLACING ALL A" BY +
AFTER INITIAL X"
INSPECT MYDATA REPLACING FIRST A" BY +"
AFTER INITIAL A"
BEFORE INITIAL Z
INSPECT MYDATA REPLACING ALL AAAA" BY ZZZZ"

INSPECT (CONVERTING)

INSPECT (CONVERTING)
INSPECT TEXTLINE CONVERTING
"abcdefghijklmnopqrstuvwxyz" TO
"ABCDEFGHIJKLMNOPQRSTUVWXYZ"
INSPECT FIELDA CONVERTING
1234567890 TO ABCDEFGHIJ

REFERENCE MODIFICATION

05

Allows you to process PIC 9 and PIC X fields

as if they were an array of characters.
FIELDNAME(start pos : [ length])
FIELDA

PIC X(7) VALUE ABCDEFG.

FIELDA(1:3) ABC
FIELDA(2:2) BC
FIELDA(4: ) DEFG

Qualification of Names

COBOL allows the same variable name to be used to

define fields in different records or group items.

Duplicate names must be qualified when they are

referenced
01 XFIELD.
05 YFIELD.
10 ZFIELD PIC X(3).
If ZFIELD is a duplicate name it can be qualified in two
ways: ZFIELD OF YFIELD
ZFIELD OF XFIELD

Intrinsic Functions
COBOL does not permit user-defined
functions or procedures
Intrinsic (built-in) Functions can be used in
your programs
Three broad categories of intrinsic
functions: date functions, numeric
functions and string functions.

Intrinsic Functions
Intrinsic Function values are replaced in the position where
they occur by the function result.
In COBOL, an Intrinsic Function is a temporary data item
whose value is determined at the time the function is
executed.
Functions that return a number value (numeric & integer)
are always considered to be signed.
A function that returns a number value can be used only in
an arithmetic expression or as the source of a MOVE
statement.

Intrinsic Functions
Intrinsic function pattern:
FUNCTION FunctionName(Parameters)
FunctionName is the name of the function and
Parameters is one or more parameters supplied
to the function.
COMPUTE NUM = FUNCTION RANDOM(99)
MOVE FUNCTION REVERSE(ABCD) TO NAME

Sample Intrinsic Functions

CHAR(PosInt) AlphanumericReturns the character at
ordinal position PosInt of the collating sequence.
ORD(Alph) IntegerReturns the ordinal position of
character Alph.
ORD-MAX({Any}...)
IntegerReturns the ordinal position of whichever of the
parameters has the highest value. All parameters must
be of the same type. The parameter list may be replaced
by an array.
ORD-MIN({Any}...) IntegerReturns the ordinal
position of whichever of the parameters has the lowest
value. All parameters must be of the same type.

Intrinsic Functions
REVERSE(Alph) Alphanumeric Returns a
character string with the characters in Alph
reversed.
LOWER-CASE(Alph) Alphanumeric Returns a
character string with the characters in Alph
changed to their lower case equivalents.
UPPER-CASE(Alph) Alphanumeric Returns a
character string with the characters in Alph
changed to their upper case equivalents

Date Intrinsic Functions

CURRENT-DATE - Returns a 21 character
string representing the current date and
time
DATE-OF-INTEGER(PosInt) - Returns
the yyyymmdd (standard date) equivalent
of the integer date - PosInt. The integer
date is the number of days that have
passed since Dec 31st 1600 in the
Gregorian Calendar.

Math Intrinsic Functions

MAX({Any}...)Return type on type of Any.
Takes a parameter list and returns the content of
whichever parameter contains the maximum
value.
The returned type depends upon the parameter
types as follows;
Alphanumeric if parameters are Alphabetic or
Alphnumeric.
Integer if all are integer.
Numeric if all are Numeric.
An array may be used instead of the parameter
list.

Math Intrinsic Functions

SQRT(Num) Returns an approximation of
the square root of Num.
STANDARD-DEVIATION({Num}...)
Returns an approximation of the standard
deviation of its parameters.
And many others

Condition Names
01 NO-OF-NEIGHBORS
PIC 9.
88 JUST-RIGHT VALUE 2 THRU 3.
88 TOO-FEW
VALUE 0 THRU 1.
88 TOO-MANY
VALUE 4 THRU 8.
01 MARITAL-STATUS PIC X.
88 VALID-STATUS
VALUE S M D W.

Variable Length Tables

01 TCOUNT
PIC S9(3) PACKED-DECIMAL.
01 STATE-TABLE.
05 ST-GROUP OCCURS 1 TO 100 TIMES
DEPENDING ON TCOUNT
ASCENDING KEY IS ZIP
INDEXED BY NDX.
10
ZIP
PIC X(5).
10
RATE PIC V9999.

Loading a Variable Length Table

PERFORM WITH TEST AFTER
VARYING NDX FROM 1 BY 1
UNTIL PTABLE-EOF OR
NDX = 100
PERFORM FILE-READ
IF NOT PTABLE-EOF
MOVE ZIP-IN TO ZIP(NDX)
MOVE RATE-IN TO RATE(NDX)
ELSE
SET NDX DOWN BY 1
SET TCOUNT TO NDX
END-IF
END-PERFORM

Exercise #11

Read the file BCST.SICCC01.PDSLIB(STATES)

Store the data in a variable length table

Read the file BCST.SICCC01.PDSLIB(COBDATA4). Each record has a 2 digit

state code and a 5 byte zip code. For each record in this file print a record on a
report. The report will have the following format:
GA
CA
ZD

31907
90003
30002

GEORGIA - VALID ZIP AND STATE

CALIFORNIA - INVALID ZIP
INVALID STATE

Capitalize the state name found in the table. Print message next to each
state name separated by a dash.

Use the SEARCH command perform a sequential search of the table for each look
up. After the program is working, modify it to perform a binary search with SEARCH
ALL

Files with Multiple Record Types

FD TransFile.
01 InsertRec.
02 RECI
02 STUDENTIDI
02 STUDENTNAME.
03 SURNAME
03 INITIALS
02 DOB.
03 YOBIRTH
03 MOBIRTH
03 DOBIRTH
02 COURSECODE
02 GENDER
01 DELETEREC.
02 RECD
02 STUDENTIDD
01 UpdateRec.
02 STUDENTIDU
02 NEWCOURSECODE

PIC X.
PIC 9(7).
PIC X(8).
PIC XX.
PIC
PIC
PIC
PIC
PIC

9(4).
99.
99.
X(4).
X.

PIC X.
PIC 9(7).
PIC 9(7).
PIC X(4).

Multiple O1 File Descriptions

Any number of 01 record descriptions can be
coded with the FD
Only one buffer is used no matter how many
record descriptions have been coded
Record fields cant be referenced before the file
is opened or after it is closed
With multiple record formats, there needs to be
a fixed field to indicate the record type
Value clauses are only used for 88 level items

Writing With Carriage Control

Variable Length Records

The RECORD IS VARYING IN SIZE clause
specifies a file containing variable length
records.

Variable Length Records

The RECSIZE number in the DEPENDING
ON phase must be an elementary
unsigned integer data-item declared in the
WORKING-STORAGE SECTION.
FD TRANFILE
RECORD IS VARYING IN SIZE
FROM 1 TO 80 CHARACTERS
DEPENDING ON RECSIZE.

Variable Length Record Processing

When writing a variable length record, the size of
the record must be placed in the RECSIZE
variable before the WRITE is issued.
When reading a variable length record, the
length of the record is delivered into the
RECSIZE variable.
The 01 Record description must be long enough
to accommodate the largest record

Variable length Record Processing

FD TRANFILE
RECORD IS VARYING IN SIZE
FROM 1 TO 80 CHARACTERS
DEPENDING ON RECSIZE.
01 TRANREC PIC X(80).
88 END-OF-RECS VALUE HIGH-VALUES.
WORKING-STORAGE SECTION.
01 RECSIZE PIC 99.
READ TRANFILE AT END
DISPLAY TRANREC(1:RECSIZE)

Writing Variable Length Records

Many variable length records have a fixed front end, and
varying numbers of segments
FD

CUST
RECORD IS VARYING IN SIZE
FROM 28 TO 408 CHARACTERS
DEPENDING ON RECORD-LEN.
01 CUST-REC.
05 ROOT-SEG.
10 CUST-NO
PIC X(6).
10 INVOICE-COUNT
PIC S99.
05 INVOICE-SEGMENT OCCURS 20 TIMES INDEXED BY NDX
10 INVOICE-DATE PIC X(8).
10 INVOICE-NO
PIC X(5).
10 INVOICE-AMT
pIC S9(5)V99.
You must set the RECORD-LEN before writing the record!

Alternative
FD CUSTFILE
RECORD CONTAINS 20 TO 80 CHARACTERS
01 REC.
05 FIXED-PART PIC X(20).
10
05 VARY-PART OCCURS 1 TO 6 TIMES
DEPENDING ON COUNT
INDEXED BY NDX
COUNT has to be initialized at the time of writing the record

Reading Variable Length Records

FD

CUST
RECORD IS VARYING IN SIZE
FROM 28 TO 408 CHARACTERS.
01 CUST-REC.
05 ROOT-SEG.
10 CUST-NO
PIC X(6).
10 INVOICE-COUNT PIC S99.
05 INVOICE-SEGMENT OCCURS 20 TIMES INDEXED BY NDX
10 INVOICE-DATE PIC X(8).
10 INVOICE-NO
PIC X(5).
10 INVOICE-AMT PIC S9(5)V99.

PERFORM

VARYING NDX FROM 1 BY 1

UNTIL NDX > INVOICE-COUNT
ADD INVOICE-AMT(NDX) TO AMT-OWED
END-PERFORM

EXERCISE #12
Each A record for a given customer is
followed by one to five B records for that
customer.
For each A record, write out one variable
length record that contains the A record as
the fixed part and the associated B
records as the variable parts

EXERCISE #13
BCST.SICCC01.PDSLIB(COBDATA5) CONTAINS TWO
TYPES OF 80 BYTE RECORDS :
RECORD TYPE A
1 BYTE TYPE CODE PIC X CONTAINING A
5 BYTE CUSTOMER ID PIC X(5)

RECORD TYPE B
1 BYTE TYPE CODE PIC X CONTAINING B
5 BYTE PART NUMBER PIC X(5)
6 BYTE COST PIC 9(4)V99

EXERCISE #14
Read the variable length records you created in exercise #5.
Produce a report similar to the one below:
CUSTOMER ID PART #
10030
22322
23444
50043
TOTAL
20030
22322
23444
50043
TOTAL

COST
1,333.34
3.44
98.77
1435.55
1,333.34
3.44
98.77
1435.55

VSAM File Processing

Virtual Storage Access Method

VSAM File Types

ESDS Entry Sequenced Data Set
Allows sequential processing

RRDS Relative Record Data Set

Allows sequential or random access by
relative record number

KSDS Key-Sequenced Data Set

Allows sequential, skip sequential, and
random processing by key

VSAM
VSAM data sets are known as Clusters
For ESDS or RRDS the cluster consists of
a data component
For KSDS the cluster consists of a data
component and an index component
VSAM data is stored on DASD in control
intervals which are grouped into control
areas

VSAM
The Control Interval (CI) is the unit of data
that transfers between the disk and virtual
storage
CI sizes are multiples of 2K with 4k being
common
CIs can be constructed with free space to
accommodate additions to the file
Control Areas (CA) can be constructed
with free space to accommodate additions

VSAM
VSAM dynamically manages the file by
maintaining information in each CI and CA
When a CI becomes too full the data it
contains is split into two CIs
When a CA becomes too full the data it
contains is split into two CAs
VSAM tries to keep records that are
logically close together, physically close as
well

VSAM Indexes

VSAM Components

Access Method Services (AMS)

AMS is a VSAM utility that provides numerous
options

DEFINE CLUSTER
PRINT
REPRO
LISTCAT
DELETE
DEFINE ALTERNATEINDEX
DEFINE PATH
BLDINDEX

VSAM JCL
Unlike QSAM files, VSAM files are usually
allocated in a separate job step before
data can be written to the file
A VSAM cluster is usually created by
deleting and then defining the cluster
After the cluster is defined, a job can run
which writes data to the file

VSAM JCL
Parameters:
INDEXED KSDS
NONINDEXED ESDS
NUMBERED RRDS
KEYS ( len off) primary key info
CISZ (size) control interval size
FREESPACE (ci ca) free space %s

MAKEKSDS

000100
000200
000300
000400
000500
000600
000700
000800
000900
001000
001100
001200
001210
001220
001230
001240
001250
001260
001270
001280

//TSYSAD2C JOB 'YOUR NAME',USER=TSYSAD2,REGION=2048K,MSGCLASS=V

//*MAIN CLASS=TSYSC,USER=TSYSAD2
//DEFINE
EXEC PGM=IDCAMS
//SYSPRINT DD SYSOUT=*
//SYSIN
DD *
DELETE TSYSAD2.PAYROLL.MASTER
DEFINE CLUSTER
(NAME(TSYSAD2.PAYROLL.MASTER)
INDEXED
RECORDSIZE(31 31)
KEYS(5 0)
MGMTCLAS(STANDARD)
FREESPACE(0 0)
SHAREOPTIONS (3 3))
DATA
(NAME(TSYSAD2.PAYROLL.MASTER.DATA)
TRK(1 1)
CONTROLINTERVALSIZE(4096))
INDEX (NAME(TSYSAD2.PAYROLL.MASTER.INDEX)
TRK(1 1))
/*

IDCAMS PRINT
000100 //TSYSAD2P JOB
'A.STUDENT',USER=TSYSAD2,REGION=2048K,MSGCLASS=V
000200 //*MAIN CLASS=TSYSC,USER=TSYSAD2
000210 //* THIS IS AN IDCAMS PRINT
000220 //PRINT
EXEC PGM=IDCAMS
000230 //SYSPRINT DD SYSOUT=*
000240 //SYSIN
DD *
000250
PRINT INFILE(IFILE) 000251
DUMP
000252 /*
000253 //IFILE
DD DSN=TSYSAD2.PAYROLL.MASTER,DISP=SHR
000254 //

IDCAMS REPRO

000100
000200
000210
000220
000230
000240
000250
000251
000252
000253
000254
000255
000256
000257
000258

//TSYSAD2R JOB 'A.STUDENT',USER=TSYSAD2,REGION=2048K,MSGCLASS=V

//*MAIN CLASS=TSYSC,USER=TSYSAD2
//* THIS AN IDCAMS REPRO
//REPRO
EXEC PGM=IDCAMS
//FILEIN
DD DSN=TSYSAD2.PGM1.RESULTS,DISP=SHR
//FILEOUT DD DSN=TSYSAD2.I10.PGM1.RESULTS,DISP=(NEW,CATLG,DELETE),
//
UNIT=SYSDA,DCB=(RECFM=FB,LRECL=80),
//
SPACE=(TRK,(1,1),RLSE)
//SYSIN
DD *
REPRO INFILE(FILEIN) OUTFILE(FILEOUT)
/*
//AMSDUMP DD SYSOUT=*
//

Creating a VSAM File

000100
000200
000300
000400
000500
000600
000610
000620
000630
000640
000700

IDENTIFICATION DIVISION.
PROGRAM-ID. VSAM1.
ENVIRONMENT DIVISION.
INPUT-OUTPUT SECTION.
FILE-CONTROL.
SELECT PAYROLL-MASTER-OUT ASSIGN TO PAYMASTO
ORGANIZATION IS INDEXED
ACCESS IS SEQUENTIAL
RECORD KEY IS ID-OUT
FILE STATUS IS PM-STATUS.
SELECT PAYROLL-MASTER-IN ASSIGN TO PAYMASTI.

Creating a VSAM File

004410 01 PM-STATUS.
004430
05 PM-STAT1
PIC X.
004440
05 PM-STAT2
PIC X.
004441 PROCEDURE DIVISION.
004450
OPEN INPUT PAYROLL-MASTER-IN
004460
OPEN OUTPUT PAYROLL-MASTER-OUT
004461
IF PM-STATUS NOT = '00'
004462
PERFORM 300-PRINT-STATUS
004463
END-IF
004470
PERFORM UNTIL ARE-THERE-MORE-RECORDS = 'NO '
004480
READ PAYROLL-MASTER-IN
004490
AT END
004500
MOVE 'NO ' TO ARE-THERE-MORE-RECORDS
004600
NOT AT END
004700
PERFORM 200-READ-MODULE
004800
END-READ
004900
END-PERFORM
005000
CLOSE PAYROLL-MASTER-IN
005100
PAYROLL-MASTER-OUT
005110
GOBACK

Creating a VSAM File

005130
005410
005420
005430
005440
005500
005510
005520
005530
005600
005700
005800
005900
006000

200-READ-MODULE.
MOVE ID-IN TO ID-OUT
MOVE NAME-IN TO NAME-OUT
MOVE HOURS-IN TO HOURS-OUT
MOVE RATE-IN TO RATE-OUT
WRITE MASTER-REC-OUT
IF PM-STATUS NOT = '00'
PERFORM 300-PRINT-STATUS
END-IF
.
300-PRINT-STATUS.
DISPLAY 'FILE STATUS CODE:' PM-STATUS
GOBACK
.

VSAM Error Strategy

VSAM returns a status code after each
operation
It is imperative that you check each status
code after each operation to insure that
the program is proceeding normally
The status code is a two byte field

OPEN
OPEN
OPEN
OPEN
OPEN

INPUT file-name
OUTPUT file-name
I-O file-name
EXTEND file-name
For EXTEND, access mode must be
sequential

Reading for Sequential Access

READ file-name [NEXT] [RECORD]
[INTO data-name]
[AT END imperative stmt]
[NOT AT END imperative stmt]
[END-READ]
Specify NEXT if access is DYNAMIC and you want
sequential processing
Can be omitted when access is SEQUENTIAL
INTO provides move mode I/O
Omitting INTO provides locate mode I/O

Reading for Random Access

READ file-name [RECORD]
[INTO data-name]
[INVALID KEY imperative stmt]
[NOT INVALID KEY imperative
stmt]
[END-READ]
Be sure to set the key of the record you wish
to read beforehand

Writing
WRITE record-name [FROM dataname]
[INVALID KEY imperative stmt]
[NOT INVALID KEY imperative
stmt]
[END-WRITE]

KSDS FILE CREATION

SELECT INVMAST
ASSIGN TO INVMAST
ORGANIZATION IS INDEXED
ACCESS IS SEQUENTIAL
RECORD KEY IS ITEM-NO
FILE STATUS IS FILE-STAT.
FD INVMAST
01 RECORD-AREA.
05 ITEM-NO PIC X(5)
05
PIC X(75).
WORKING-STORAGE SECTION
01 FILE-STAT
PIC X(2).

KSDS File Commands

OPEN INPUT file-name
OPEN OUTPUT file-name
READ file-name [NEXT] RECORD
[INTO data-area]
[AT END imperative]
[NOT AT END imperative]
[END-READ]

KSDS File Commands

WRITE record [FROM]] data-name]
[INVALID KEY imperative]
[NOT INVALID KEY imperative]
[END-WRITE]
REWRITE record [FROM]] data-name]
[INVALID KEY imperative]
[NOT INVALID KEY imperative]
[END-REWRITE]
CLOSE file-name

KSDS File Statements

START file-name
KEY IS = data-name
>
>=
[INVALID KEY imperative]
[NOT INVALID KEY imperative]
[END-START]

KSDS File Statements

DELETE file-name RECORD
[INVALID KEY imperative]
[NOT INVALID KEY imperative]
[END-DELETE]

Exercise #15

Read BCST.SICCC01.PDSLIB(COBDATA6)
COL 1-5 KEY
COL 6-25 NAME
Allocate a KSDS with record size 25 and a key
field in cols 1-5
Write out a KSDS record for each record in the
file. Write out the records sequentially.

Exercise #16
Read
BCST.SICCC01.PDSLIB(COBDATA7)
COL 1-5 KEY
Read a KSDS record (randomly) for each
record in the file. Write out the names you
find sequentially. If the record doesnt
exist, print a message Not Found

Nested Programs

1)
2)
3)
4)

COBOL programs can be nested. There are many

advantages for doing this:
The monolithic working storage of most COBOL
programs leads to difficulty in debugging because all
data is global
Nested programs break the working storage into
smaller areas that can only be accessed by programs
that need access.
Nested programs provide for parameter passing by
techniques found in all modern languages (by value, by
reference)
There is no execution degradation because of nested
programs. In fact, calling a nested program is more
efficient than calling a separately compiled program.

Nested Programs
Calling a nested program is as efficient as
performing a paragraph
Nested programs provide design flexibility and
encourage good program design
A nested program would be called a function or
subroutine in any other language
Nested programs unleash the power of COBOL
pointers and allow COBOL programmers to
design data structures that encourage efficient
programming techniques

Nested Programs
PROGRAM-ID. MAIN.
PROGRAM-ID. SUB1.
END PROGRAM
PROGRAM-ID.
END PROGRAM
END PROGRAM

SUB1.
SUB2
SUB2.
MAIN.

COBOL Pointers
05 PTR1
USAGE IS POINTER.
05 PTR2-P USAGE IS POINTER VALUE NULL.
SET PTR1 TO ADDRESS OF LINKAGE-THING
SET PTR2 TO PTR1
LINKAGE SECTION.
01 NAME-STRUCTURE.
05 FIRST-NAME PIC X(18).
05 LAST-NAME
PIC X(26).
SET ADDRESS OF NAME-STRUCTURE TO EXAMPLE-P.

COBOL Pointers
IF PTR1 NOT = NULL AND
PTR1 NOT = PTR2
PERFORM 2730-SOMETHING
END-IF

SYNCHRONIZED
The SYNCHRONIZED clause is sometimes used
with USAGE IS COMP or USAGE IS INDEX
items. It is used to optimize speed of processing
but it does so at the expense of increased
storage requirements.
The word SYNC can be used instead of
SYNCHRONIZED
SYNCHRONIZED causes slack bytes to be
generated when needed.