Basic Reference Guide
Basic Reference Guide
Version 8 Release 1
LC18-9897-01
IBM WebSphere DataStage
Version 8 Release 1
LC18-9897-01
Note
Before using this information and the product that it supports, read the information in “Notices” on page 503.
Contents v
REMOVE function . . . . . . . . . . . . 289 TIME function . . . . . . . . . . . . . 354
REMOVE statement . . . . . . . . . . . 290 TIMEDATE function . . . . . . . . . . . 354
REPEAT statement . . . . . . . . . . . 292 TIMEOUT statement . . . . . . . . . . . 355
REPLACE function . . . . . . . . . . . 292 TPARM function . . . . . . . . . . . . 356
RETURN statement . . . . . . . . . . . 294 TPRINT statement . . . . . . . . . . . . 358
RETURN (value) statement . . . . . . . . . 295 TRANS function . . . . . . . . . . . . 359
REUSE function . . . . . . . . . . . . 295 TRANSACTION statements . . . . . . . . 360
REVREMOVE statement . . . . . . . . . . 296 TRANSACTION ABORT statement . . . . . . 360
REWIND statement . . . . . . . . . . . 297 TRANSACTION COMMIT statement . . . . . 362
RIGHT function . . . . . . . . . . . . 298 TRANSACTION START statement . . . . . . 362
RND function . . . . . . . . . . . . . 298 TRIM function . . . . . . . . . . . . . 362
ROLLBACK statement . . . . . . . . . . 299 TRIMB function . . . . . . . . . . . . 363
RPC.CALL function . . . . . . . . . . . 299 TRIMBS function . . . . . . . . . . . . 364
RPC.CONNECT function . . . . . . . . . 300 TRIMF function . . . . . . . . . . . . 364
RPC.DISCONNECT function . . . . . . . . 301 TRIMFS function . . . . . . . . . . . . 365
SADD function . . . . . . . . . . . . . 302 TRIMS function . . . . . . . . . . . . 365
SCMP function . . . . . . . . . . . . . 303 TTYCTL statement . . . . . . . . . . . 365
SDIV function . . . . . . . . . . . . . 303 TTYGET statement . . . . . . . . . . . 366
SEEK statement . . . . . . . . . . . . 304 TTYSET statement . . . . . . . . . . . . 372
SEEK (ARG.) statement . . . . . . . . . . 305 UNASSIGNED function . . . . . . . . . . 374
SELECT statements . . . . . . . . . . . 306 UNICHAR function . . . . . . . . . . . 374
SELECTE statement . . . . . . . . . . . 307 UNICHARS function . . . . . . . . . . . 375
SELECTINDEX statement . . . . . . . . . 308 UNISEQ function . . . . . . . . . . . . 375
SELECTINFO function . . . . . . . . . . 308 UNISEQS function . . . . . . . . . . . 375
SEND statement . . . . . . . . . . . . 309 UNLOCK statement . . . . . . . . . . . 376
SENTENCE function . . . . . . . . . . . 310 UPCASE function . . . . . . . . . . . . 376
SEQ function . . . . . . . . . . . . . 310 UPRINT statement . . . . . . . . . . . 377
SEQS function . . . . . . . . . . . . . 311 WEOF statement . . . . . . . . . . . . 377
SET TRANSACTION ISOLATION LEVEL WEOFSEQ statement . . . . . . . . . . . 378
statement . . . . . . . . . . . . . . . 311 WRITE statements . . . . . . . . . . . . 379
SETLOCALE function . . . . . . . . . . 313 WRITEBLK statement . . . . . . . . . . 383
SETREM statement . . . . . . . . . . . 314 WRITELIST statement . . . . . . . . . . 384
SIN function. . . . . . . . . . . . . . 315 WRITESEQ statement . . . . . . . . . . 384
SINH function . . . . . . . . . . . . . 315 WRITESEQF statement . . . . . . . . . . 386
SLEEP statement . . . . . . . . . . . . 315 WRITET statement . . . . . . . . . . . 387
SMUL function . . . . . . . . . . . . . 316 WRITEU statement . . . . . . . . . . . 389
SOUNDEX function . . . . . . . . . . . 316 WRITEV statement . . . . . . . . . . . 389
SPACE function . . . . . . . . . . . . 317 WRITEVU statement . . . . . . . . . . . 389
SPACES function . . . . . . . . . . . . 317 XLATE function . . . . . . . . . . . . 389
SPLICE function . . . . . . . . . . . . 318 XTD function . . . . . . . . . . . . . 390
SQRT function . . . . . . . . . . . . . 318
SQUOTE function . . . . . . . . . . . . 319 Chapter 7. Quick Reference . . . . . 391
SSELECT statement . . . . . . . . . . . 319 Compiler Directives . . . . . . . . . . . 391
SSUB function . . . . . . . . . . . . . 321 Declarations . . . . . . . . . . . . . . 392
STATUS function . . . . . . . . . . . . 321 Assignments. . . . . . . . . . . . . . 392
STATUS statement . . . . . . . . . . . . 325 Program Flow Control . . . . . . . . . . 393
STOP statement . . . . . . . . . . . . 328 File I/O . . . . . . . . . . . . . . . 394
STORAGE statement . . . . . . . . . . . 328 Sequential File I/O . . . . . . . . . . . 396
STR statement . . . . . . . . . . . . . 329 Printer and Terminal I/O . . . . . . . . . 397
STRS function . . . . . . . . . . . . . 329 Tape I/O . . . . . . . . . . . . . . . 398
SUBR function . . . . . . . . . . . . . 330 Select Lists . . . . . . . . . . . . . . 399
SUBROUTINE statement . . . . . . . . . 331 String Handling . . . . . . . . . . . . 399
SUBS function . . . . . . . . . . . . . 331 Data Conversion and Formatting . . . . . . . 402
SUBSTRINGS function . . . . . . . . . . 332 NLS . . . . . . . . . . . . . . . . 403
SUM function . . . . . . . . . . . . . 333 Mathematical Functions . . . . . . . . . . 404
SUMMATION function . . . . . . . . . . 333 Relational Functions . . . . . . . . . . . 407
SYSTEM function . . . . . . . . . . . . 334 System . . . . . . . . . . . . . . . 407
TABSTOP statement . . . . . . . . . . . 337 Remote Procedure Calls . . . . . . . . . . 408
TAN function . . . . . . . . . . . . . 337 Miscellaneous . . . . . . . . . . . . . 408
TANH function . . . . . . . . . . . . . 338
TERMINFO function . . . . . . . . . . . 338
Chapter 8. ASCII and Hex Equivalents 409
vi BASIC Reference Guide
Chapter 9. Correlative and Conversion !ERRNO subroutine . . . . . . . . . . . 473
Codes . . . . . . . . . . . . . . 413 !FCMP subroutine . . . . . . . . . . . . 474
Algebraic Functions . . . . . . . . . . . 414 !GET.KEY subroutine . . . . . . . . . . . 474
Concatenation . . . . . . . . . . . . . 417 !GET.PARTNUM subroutine . . . . . . . . 475
Date Conversion . . . . . . . . . . . . 418 !GET.PATHNAME subroutine . . . . . . . . 477
ICONV and OCONV Differences . . . . . . 423 !GET.PU subroutine . . . . . . . . . . . 477
International Date Conversion . . . . . . . . 424 Equate Names for Keys . . . . . . . . . 478
Extended Character Set Conversion . . . . . . 424 The PU$SPOOLFLAGS Key . . . . . . . 479
Fcode . . . . . . . . . . . . . . . . 424 !GET.USER.COUNTS subroutine . . . . . . . 480
Group Extraction . . . . . . . . . . . . 426 !INLINE.PROMPTS subroutine . . . . . . . 480
Length Function . . . . . . . . . . . . 427 !INTS subroutine . . . . . . . . . . . . 482
Masked Character Conversion . . . . . . . . 427 !MAKE.PATHNAME subroutine . . . . . . . 482
Masked Decimal Conversion . . . . . . . . 428 !MATCHES subroutine . . . . . . . . . . 483
Formatting Numbers . . . . . . . . . . . 431 !MESSAGE subroutine . . . . . . . . . . 484
Monetary Conversion . . . . . . . . . . 433 !PACK.FNKEYS . . . . . . . . . . . . 485
Packed Decimal Conversion . . . . . . . . 434 !REPORT.ERROR subroutine . . . . . . . . 488
Time Conversion . . . . . . . . . . . . 434 Printing on the Last Line and Printing a
Radix Conversion . . . . . . . . . . . . 436 Heading . . . . . . . . . . . . . . 490
ASCII Conversion . . . . . . . . . . . . 436 !SET.PU subroutine . . . . . . . . . . . 490
Map conversion . . . . . . . . . . . . 437 Equate Names for Keys . . . . . . . . . 491
NLS Map Name . . . . . . . . . . . . 437 The PU$SPOOLFLAGS Key . . . . . . . 491
NR Code . . . . . . . . . . . . . . . 437 Equate Names for Return Code . . . . . . 492
P code . . . . . . . . . . . . . . . . 438 Printing on the Last Line and Printing a
Exponential Notation . . . . . . . . . . . 438 Heading . . . . . . . . . . . . . . 492
Range Function . . . . . . . . . . . . . 439 !TIMDAT subroutine . . . . . . . . . . . 493
S code . . . . . . . . . . . . . . . . 440 !USER.TYPE subroutine . . . . . . . . . . 494
Substitution code . . . . . . . . . . . . 440 !VOC.PATHNAME subroutine . . . . . . . . 494
Text extraction . . . . . . . . . . . . . 441
File Translation . . . . . . . . . . . . . 441 Product documentation . . . . . . . 497
International Time Conversion. . . . . . . . 442 Contacting IBM . . . . . . . . . . . . 497
Chapter 10. BASIC Reserved Words 443 How to read syntax diagrams . . . . 499
Contents vii
viii BASIC Reference Guide
Chapter 1. Introduction to WebSphere DataStage BASIC
WebSphere DataStage BASIC is a business-oriented programming language designed to work efficiently
with the WebSphere DataStage environment. It is easy for a beginning programmer to use yet powerful
enough to meet the needs of an experienced programmer.
The power of WebSphere DataStage BASIC comes from statements and built-in functions that take
advantage of the extensive database management capabilities of WebSphere DataStage. These benefits
combined with other BASIC extensions result in a development tool well-suited for a wide range of
applications.
BASIC Terminology
WebSphere DataStage BASIC programmers should understand the meanings of the following terms:
v BASIC program
v Source code
v Object code
v Variable
v Function
v Keyword.
BASIC Program
A BASIC program is a set of statements directing the computer to perform a series of tasks in a specified
order. A BASIC statement is made up of keywords and variables.
Source Code
Source code is the original form of the program written by the programmer.
Object code is compiler output, which can be executed by the WebSphere DataStage RUN command or
called as a subroutine.
Variable
A variable is a symbolic name assigned to one or more data values stored in memory. A variable’s value
can be numeric or character string data, the null value, or it can be defined by the programmer, or it can
be the result of operations performed by the program. Variable names can be as long as the physical line,
but only the first 64 characters are significant. Variable names begin with an alphabetic character and can
include alphanumeric characters, periods ( . ), dollar signs ( $ ), and percent signs ( % ). Upper- and
lowercase letters are interpreted as different; that is, REC and Rec are different variables.
Function
Keyword
A BASIC keyword is a word that has special significance in a BASIC program statement. The case of a
keyword is ignored; for example, READU and readu are the same keyword. For a list of reserved words
see Appendix A.
Subroutines
A subroutine is a set of instructions that perform a specific task. It is a small program that can be
embedded in a program and accessed with a GOSUB statement, or it can be external to the program and
accessed with a CALL statement. Common processes are often kept as external subroutines. This lets the
programmer access them from many different programs without having to rewrite them.
When a GOSUB statement or CALL statement statement is encountered, program control branches to the
referenced subroutine. An internal subroutine must begin with a statement label. An external subroutine
must begin with a SUBROUTINE statement statement.
A RETURN statementstatement can be used at the end of a subroutine to return program flow to the
statement following the last referenced GOSUB or CALL statement. If there is no corresponding CALL or
GOSUB statement, the program halts and returns to the WebSphere DataStage command level. If an
external subroutine ends before a RETURN statement is encountered, a RETURN is provided
automatically.
Note: If an ABORT statement, STOP statement, or CHAIN statement statement is encountered during
subroutine execution, program execution aborts, stops, or chains to another BASIC program and control
never returns to the calling program.
One or more arguments separated by commas can be passed to the subroutine as an argument list. An
argument can be a constant, variable, array variable, or expression, each representing an actual value. The
Source Syntax
A BASIC source line has the following syntax:
[ label ] statement [ ; statement ... ] <Return>
You can put more than one statement on a line. Separate the statements with semicolons.
A BASIC source line can begin with a statement label. It always ends with a carriage return (Return). It
can contain up to 256 characters and can extend over more than one physical line.
Statement Types
BASIC statements can be used for any of the following purposes:
v Input and output control
v Program control
v Assignment (assigning a value to a variable)
v Specification (specifying the value of a constant)
v Documentation
Input statements indicate where the computer can expect data to come from (for example, the keyboard, a
particular file, and so on). Output statements control where the data is displayed or stored.
In general, BASIC statements are executed in the order in which they are entered. Control statements alter
the sequence of execution by branching to a statement other than the next statement, by conditionally
executing statements, or by passing control to a subroutine.
Assignment statements assign values to variables, and specification statements assign names to constants.
Any text that appears between a comment symbol and a carriage return is treated as part of the
comment. Comments cannot be embedded in a BASIC statement. If you want to put a comment on the
same physical line as a statement, you must end the statement with a semicolon ( ; ), then add the
comment, as in the following example:
IF X THEN
A = B; REM correctly formatted comment statement
B = C
END
You cannot put comments between multiple statements on one physical line. For example, in the second
line of the following program the statement B = C is part of the comment and is not executed:
IF X THEN
A = B; REM The rest of this line is a comment; B = C
END
However, you can put comments in the middle of a statement that occupies more than one physical line,
as in the following example:
Statement Labels
A statement label is a unique identifier for a program line. A statement label consists of a string of
characters followed by a colon. The colon is optional when the statement label is completely numeric.
Like variable names, alphanumeric statement labels begin with an alphabetic character and can include
periods ( . ), dollar signs ( $ ), and percent signs ( % ). Upper- and lowercase letters are interpreted as
different; that is, ABC and Abc are different labels. Statement labels, like variable names, can be as long
as the length of the physical line, but only the first 64 characters are significant. A statement label can be
put either in front of a BASIC statement or on its own line. The label must be first on the line-that is, the
label cannot begin with a space.
Spaces or Tabs
In a program line, spaces or tabs that are not part of a data item are ignored. Therefore you can use
spaces or tabs to improve the program’s appearance and readability.
WebSphere DataStage BASIC handles this difference transparently in nearly every case, but in a few
instances the operating system differences become apparent. If you want your program to work on
different operating systems, watch sequential file I/O (that is, writing to or reading from type 1 and type
19 files, line by line or in blocks of data). In particular, be aware of the potential differences that occur:
v When moving a pointer through a file
v When reading or writing blocks of data of a specified length
Special Characters
The WebSphere DataStage BASIC character set comprises alphabetic, numeric, and special characters. The
alphabetic characters are the upper- and lowercase letters of the alphabet. The numeric characters are the
digits 0 through 9. The special characters are as follows. Most of the special characters are not permitted
in a numeric constant or a variable name.
Space
Tab
= Equal sign or assignment symbol
+ Plus sign
- Minus sign
* Asterisk, multiplication symbol, or nonexecutable comment
** Exponentiation
/ Slash or division symbol
Storing Programs
BASIC programs are stored as records in type 1 or type 19 files. The program file must exist before you
invoke an editor to create a new record to hold your program. Record IDs must follow the conventions
for type 1 and type 19 files.
Editing Programs
You can use the WebSphere DataStage Editor or any suitable editor, such as vi on UNIX or edit on
Windows NT, to write your programs. You can edit programs in the WebSphere DataStage environment
or at the operating system level.
pathname is the relative or absolute path name of the program you want to edit. For example, the
program PAYROLL is stored as a record in the file BP. To edit it with vi, enter:
>VI BP/PAYROLL
If you want to use vi, or any other editor, directly from WebSphere DataStage, you can create a VOC
entry that invokes your chosen editor. For example, this VOC entry calls edit from WebSphere DataStage
on Windows NT:
EDIT
001 V
002 \win25\edit.com
003 PR
Getting Started
To create and use a BASIC program, follow these steps:
1. Use the CREATE.FILE command to create a type 1 or type 19 WebSphere DataStage file to store your
BASIC program source. The RUN command uses the filename BP if a filename is not specified, so
many people use BP as the name of their general BASIC program file.
2. Use the WebSphere DataStage Editor or some other editor to create the source for your BASIC
program as a record in the file you created in step 1.
3. Once you have created the record containing your BASIC program source statements, use the BASIC
command to compile your program. The BASIC command creates a file to contain the object code
output by the compiler. You do not have to know the name of the object file because the program is
always referred to by the source filename.
4. If the BASIC compiler detects any errors, use the Editor to correct the source code and recompile
using the BASIC command.
5. When your program compiles without any errors, execute it using the RUN command. Use the RAID
command to debug your program.
Types of Data
Although many program languages distinguish different types of data, the WebSphere DataStage BASIC
compiler does not. All data is stored internally as character strings, and data typing is done contextually
at run time. There are three main types of data: character string, numeric, and unknown (that is, the null
value).
In NLS mode all data is held in the WebSphere DataStage internal character set. In all WebSphere
DataStage I/O operations, data is converted automatically by applying the map specified for a file or a
device. One character can be more than one byte long and can occupy zero or more positions on the
screen. WebSphere DataStage BASIC provides functions so that programs can determine what these
characteristics are. For more information about character sets, see WebSphere DataStage National Language
Support Guide.
The beginning and terminating marks enclosing character string data must match. In other words, if you
begin a string with a single quotation mark, you must end the string with a single quotation mark.
If you use either a double or a single quotation mark within the character string, you must use the
opposite kind to begin and end the string. For example, this string should be written:
"It’s a lovely day."
The empty string is a special instance of character string data. It is a character string of zero length. Two
adjacent double or single quotation marks, or backslashes, specify an empty string:
In your source code you can use any ASCII character in character string constants except ASCII character
0 (NUL), which the compiler interprets as an end-of-string character, and ASCII character 10 (linefeed),
which separates the logical lines of a program. Use CHAR(0) and CHAR(10) to embed these characters in
a string constant.
Numeric Data
All numeric data is represented internally either as floating-point numbers with the full range of values
supported by the system’s floating-point implementation, or as integers. On most systems the range is
from 10-307 through 10+307 with 15 decimal digits of precision.
Numeric Constants
Numeric constants can be represented in either fixed-point or floating-point form. Commas and spaces
are not allowed in numeric constants.
Fixed-Point Constants.
Fixed-point form consists of a sequence of digits, optionally containing a decimal point and optionally
preceded by a plus ( + ) or minus ( - ) sign. Some examples of valid fixed-point constants are:
12
-132.4
+10428
Floating-Point Constants.
Floating-point form, which is similar to scientific notation, consists of a sequence of digits, optionally
preceded by a plus ( + ) or minus ( - ) sign representing the mantissa. The sequence of digits is followed
by the letter E and digits, optionally preceded by a minus sign, representing the power of 10 exponent.
The exponent must be in the range of -307 through +307. Some examples of valid floating-point constants
are:
1.2E3
-7.3E42
-1732E-4
Use the PRECISION statement to set the maximum number of fractional digits that can result from
converting numbers to strings.
Note: Do not confuse the null value with the empty string. The empty string is a character string of zero
length which is known to have no value. Unlike null, whose value is defined as unknown, the value of
the empty string is known. You cannot use the empty string to represent the null value, nor can you use
the null value to represent ″no value.″
Like all other data in WebSphere DataStage BASIC, the null value is represented internally as a character
string. The string is made up of the single byte CHAR(128). At run time when explicit or implicit
dynamic array extractions are executed on this character, it is assigned the data type ″null.″ WebSphere
DataStage BASIC programs can reference the null value using the system variable @NULL. They can test
whether a value is the null value using the ISNULL function and ISNULLS function.
Here is an example of the difference between an empty string and the null value. If you concatenate a
string value with an empty string, the string value is returned, but if you concatenate a string value with
the null value, null is returned.
A = @NULL
B = ""
C = "JONES"
X = C:B
Y = C:A
The resulting value of X is ″JONES″, but the value of Y is the null value. When you concatenate known
data with unknown data, the result is unknown.
Programmers should also note the difference between the null value, a special constant whose type is
″null,″ and the stored representation of the null value, the special character CHAR(128) whose type is
″string.″ BASIC programs can reference the stored representation of null using the system variable
@NULL.STR instead of @NULL.
Constants
Constants are data that do not change in value, data type, or length during program execution. Constants
can be character strings or numeric strings (in either integer or floating-point form). A character string of
no characters-the empty string-can also be a constant.
Variables
Variables are symbolic names that represent stored data values. The value of a variable can be:
v Unassigned
v A string, which can be an alphanumeric character string, a number, or a dynamic array
v A number, which can be fixed-point (an integer) or floating-point
v The null value
v A dimensioned array (that is, a vector or matrix)
v A subroutine name
v A file
v A select list
The value of a variable can be explicitly assigned by the programmer, or it can be the result of operations
performed by the program during execution. Variables can change in value during program execution. At
the start of program execution, all variables are unassigned. Any attempt to use an unassigned variable
produces an error message.
A variable name must begin with an alphabetic character. It can also include one or more digits, letters,
periods, dollar signs, or percent signs. Spaces and tabs are not allowed. A variable name can be any
length up to the length of the physical line, but only the first 64 characters are significant. A variable
name cannot be any of the BASIC Reserved Words listed in Appendix A. In WebSphere DataStage, upper-
and lowercase characters in a variable name are interpreted differently.
WebSphere DataStage BASIC also provides a set of system variables called @Variables. Many of these are
read-only variables. Read-only @variables cannot be changed by the programmer.
In NLS mode you can include characters outside the ASCII character set only as constants defined by the
$DEFINE statement and EQUATE statement statements, or as comments. Everything else, including
variable names, must use the ASCII character set. For more information about character sets, see
WebSphere DataStage National Language Support Guide.
Array Variables
An array is a variable that represents more than one data value. There are two types of array:
dimensioned and dynamic. Dimensioned arrays can be either standard or fixed. Fixed arrays are
provided in PICK, IN2, and REALITY flavor accounts for compatibility with other Pick systems.
Dimensioned Arrays
Each value in a dimensioned array is called an element of the array. Dimensioned arrays can be one- or
two-dimensional.
A one-dimensional array is called a vector. Its elements are arranged sequentially in memory. An element
of a vector is specified by the variable name followed by the index of the element enclosed in
parentheses. The index of the first element is 1. The index can be a constant or an expression. Two
examples of valid vector element specifiers are:
A(1)
COST(35)
A two-dimensional array is called a matrix. The elements of the first row are arranged sequentially in
memory, followed by the elements of the second row, and so on. An element of a matrix is specified by
the variable name followed by two indices enclosed in parentheses. The indices represent the row and
column position of the element. The indices of the first element are (1,1). Indices can be constants or
expressions. The indices used to specify the elements of a matrix that has four columns and three rows
are illustrated by the following:
1,1 1,2 1,3 1,4
2,1 2,2 2,3 2,4
3,1 3,2 3,3 3,4
Vectors are treated as matrices with a second dimension of 1. COST(35) and COST(35,1) are equivalent
specifications and can be used interchangeably.
Both vectors and matrices have a special zero element that is used in MATPARSE statement, MATREAD
statements, and MATWRITE statements. The zero element of a vector is specified by vector.name(0), and
the zero element of a matrix is specified by matrix.name(0,0). Zero elements are used to store fields that do
not fit in the dimensioned elements on MATREAD or MATPARSE statements.
Dimensioned arrays are allocated either at compile time or at run time, depending on the flavor of the
account. Arrays allocated at run time are called standard arrays. Arrays allocated at compile time are
called fixed arrays. Standard arrays are redimensionable; fixed arrays are not redimensionable and do not
have a zero element. All arrays are standard unless the program is compiled in a PICK, IN2, or REALITY
flavor account, in which case they are fixed arrays. To use fixed arrays in PIOPEN, INFORMATION and
Dynamic Arrays
Dynamic arrays map the structure of WebSphere DataStage file records to character string data. Any
character string can be a dynamic array. A dynamic array is a character string containing elements that are
substrings separated by delimiters. At the highest level these elements are fields separated by field marks
( F ) (ASCII 254). Each field can contain values separated by value marks ( V ) (ASCII 253). Each value
can contain subvalues separated by subvalue marks ( S ) (ASCII 252).
A common use of dynamic arrays is to store data that is either read in from or written out to a
WebSphere DataStage file record. However, WebSphere DataStage BASIC includes facilities for
manipulating dynamic array elements that make dynamic arrays a powerful data type for processing
hierarchical information independently of WebSphere DataStage files.
The number of fields, values, or subvalues in a dynamic array is limited only by the amount of available
memory. Fields, values, and subvalues containing the empty string are represented by two consecutive
field marks, value marks, or subvalue marks, respectively.
and:
JONESVSMITH
Conceptually, this dynamic array has an infinite number of fields, all of which are empty except the first
two. References made to the third or fourth field, for example, return an empty string.
and:
BETTYSSUESMARY
The first value has three subvalues: TOM, DICK, and HARRY. The second value also has three subvalues:
BETTY, SUE, and MARY.
The second field has two values: JONES and SMITH. Each value has one subvalue: JONES and SMITH.
can be considered a dynamic array containing one field, which has one value, which has one subvalue,
all of which are: NAME AND ADDRESS.
The following character string can be considered a dynamic array containing two fields:
JONESVSMITHVBROWNF$1.23VV$2.75
The first field has three values: JONES, SMITH, and BROWN. The second field has three values: $1.23, an
empty string, and $2.75
You can create a dynamic array in two ways: by treating it as a concatenation of its fields, values, and
subvalues; or by enclosing the elements of the dynamic array in angle brackets, using the syntax:
array.name < field# , value# , subvalue# >
The example has two fields. The first field has two values, and the second field has two values. The first
value of the second field has two subvalues, and the second value of the second field also has two
subvalues.
You must use the following statements to declare that the first field contains the two values JONES and
SMITH:
A = ""
A<1,1> = "JONES"
A<1,2> = "SMITH"
The statement:
A = ""
A<1> = "JONES"
declares that the first field contains only JONES with no other values or subvalues. Similarly, the
statement:
A<2,1> = 1.23
declares that the first value of the second field is 1.23 with no subvalues. The statements:
A<2,2,1> = 2.50
A<2,2,2> = 10
declare that the second value of the second field has two subvalues, 2.50 and 10, respectively.
File Variables
A file variable is created by a form of the OPEN statement. Once opened, a file variable is used in I/O
statements to access the file. There are two types of file variable: hashed file variable and sequential file
variable. File variables can be scalars or elements of a dimensioned array.
Expressions
An expression is part of a BASIC statement. It can comprise:
v A string or numeric constant
v A variable
v An intrinsic function
v A user-defined function
v A combination of constants, variables, operators, functions, and other expressions
Format Expressions
A format expression formats variables for output. It specifies the size of the field in which data is
displayed or printed, the justification (left, right, or text), the number of digits to the right of the decimal
point to display, and so on. Format expressions work like the FMT function. The syntax is:
variable format
Either width or mask can specify the size of the display field.
background specifies the character used to pad the field (Space is the default padding character).
You must specify justification as left, right, or text (text left-justifies output, but breaks lines on spaces
when possible).
edit specifies how to format numeric data for output, including such things as the number of digits to
display to the right of the decimal point, the descaling factor, whether to round or truncate data, and
how to indicate positive and negative currency, handle leading zeros, and so on.
If a format expression is applied to the null value, the result is the same as formatting an empty string.
This is because the null value has no printable representation.
You can use the STATUS function to determine the result of the format operation. The STATUS function
returns the following after a format operation:
0 The format operation is successful.
1 The variable is invalid.
2 The format expression is invalid.
In NLS mode, the FMT function formats an expression in characters; the FMTDP function formats it in
display positions. The effect of the format mask can be different if the data contains double-width or
multibyte characters. For more information about display length, see WebSphere DataStage National
Language Support Guide.
Arithmetic Operators
Arithmetic operators combine operands comprising one or more variables, constants, or intrinsic
functions. Resulting arithmetic expressions can be combined with other expressions almost indefinitely.
The syntax of arithmetic expressions is:
expression operator expression
** X ** Y
* Multiplication X*Y
/ Division X/Y
+ Addition X+Y
- Subtraction X-Y
You can use parentheses to change the order of evaluation. Operations on expressions enclosed in
parentheses are performed before those outside parentheses.
The result of any arithmetic operation involving the null value is the null value. Since the null value is
unknown, the result of combining it with anything must also be unknown. So in the following example,
B is the null value:
A = @NULL
B = 3 + A
If a character string variable containing only numeric characters is used in an arithmetic expression, the
character string is treated as a numeric variable. That is, the numeric string is converted to its equivalent
internal number and then evaluated numerically in the arithmetic expression. For example, the following
expression is evaluated as 77:
55 + "22"
In a BASIC program compiled in an IDEAL, PICK, PIOPEN, REALITY, or IN2 flavor account, arithmetic
can be performed on string data only if the string can be interpreted as a single-valued number. The
previous statement successfully compiles in PICK, PIOPEN, IN2, REALITY, and IDEAL flavor accounts
but causes a run-time error. The REUSE function converts 2 to a string which then has to be converted
back to a number for the arithmetic operation. This is harmless. The multivalued string cannot be
converted to a number and causes a nonnumeric data warning.
The IDEAL flavor uses single-valued arithmetic because of the performance penalty incurred by
multivalued arithmetic. To perform multivalued arithmetic in IDEAL, PICK, PIOPEN, IN2, and REALITY
flavor accounts, use the VEC.MATH option of the $OPTIONS statement.
String Operators
The concatenation operator ( : or CAT) links string expressions to form compound string expressions, as
follows:
’HELLO. ’ : ’MY NAME IS ’ : X : ". WHAT’S YOURS?"
or:
’HELLO. ’ CAT ’MY NAME IS ’ CAT X CAT ". WHAT’S
YOURS?"
If, for instance, the current value of X is JANE, these string expressions both have the following value:
"HELLO. MY NAME IS JANE. WHAT’S YOURS?"
Multiple concatenation operations are performed from left to right. Parenthetical expressions are
evaluated before operations outside the parentheses.
With the exception of the null value, all operands in concatenated expressions are considered to be string
values, regardless of whether they are string or numeric expressions. However, the precedence of
arithmetic operators is higher than the concatenation operator. For example:
"THERE ARE " : "2" + "2" : "3" : " WINDOWS."
Substring Operator
A substring is a subset of contiguous characters of a character string. For example, JAMES is a substring
of the string JAMES JONES. JAMES JON is also a substring of JAMES JONES.
You can specify a substring as a variable name or an array element specifier, followed by two values
separated by a comma and enclosed in square brackets. The two values specify the starting character
position and the length of the substring. The syntax is:
expression [ [ start, ] length ]
The bold brackets are part of the syntax and must be typed.
If start is 0 or a negative number, the starting position is assumed to be 1. If start is omitted, the starting
position is calculated according to the following formula:
string.length - substring.length + 1
This lets you specify a substring consisting of the last n characters of a string without having to calculate
the string length. So the following substring specification:
"1234567890" [5]
Another syntax for removing substrings from a string, similar to the previous syntax, is:
expression [ delimiter, occurrence, fields ]
The bold brackets are part of the syntax and must be typed. Use this syntax to return the substring that is
located between the stated number of occurrences of the specified delimiter. fields specifies the number of
successive fields after the specified occurrence of the delimiter that are to be returned with the substring.
The delimiter is part of the returned value when successive fields are returned. This syntax performs the
same function as the FIELD function.
All substring syntaxes can be used with the assignment operator ( = ) to replace the value normally
returned by the [ ] operator with the value assigned to the variable. For example:
A=’12345’
A[3]=1212
PRINT "A=",A
Assigning the three-argument syntax of the [ ] operator provides the same functionality as the
FIELDSTORE function.
Relational Operators
Relational operators compare numeric, character string, or logical data. The result of the comparison,
either true ( 1 ) or false ( 0 ), can be used to make a decision regarding program flow (see the IF
statement). Table 2 lists the relational operators.
Table 2. Relational Operators
Operator Relation Example
EQ or = Equality X=Y
NE or # Inequality X#Y
>< or <> Inequality X <> Y
LT or < Less than X<Y
GT or > Greater than X>Y
LE or <= or =< or #> Less than or equal to X <= Y
GE or >= or => or #< Greater than or equal to X >= Y
When arithmetic and relational operators are both used in an expression, the arithmetic operations are
performed first. For example, the expression:
X + Y < (T - 1) / Z
is true if the value of X plus Y is less than the value of T minus 1 divided by Z.
String comparisons are made by comparing the ASCII values of single characters from each string. The
string with the higher numeric ASCII code equivalent is considered to be greater. If all the ASCII codes
are the same, the strings are considered equal.
If the two strings have different lengths, but the shorter string is otherwise identical to the beginning of
the longer string, the longer string is considered greater.
Note: An empty string is always compared as a character string. It does not equal numeric zero.
A space is evaluated as less than zero. Leading and trailing spaces are significant. If two strings can be
converted to numeric, then the comparison is always made numerically.
The results of any comparison involving the null value cannot be determined-that is, the result of using a
relational operator to compare any value to the null value is unknown. You cannot test for the null value
using the = (equal) operator, because the null value is not equal to any value, including itself. The only
way to test for the null value is to use the function ISNULL function or ISNULLS function.
The pattern is a general description of the format of the string. It can consist of text or the special
characters X, A, and N preceded by an integer used as a repeating factor. X stands for any characters, A
stands for any alphabetic characters, and N stands for any numeric characters. For example, 3N is the
pattern for character strings made up of three numeric characters. If the repeating factor is zero, any
number of characters will match the string. For example, 0A is the pattern for any number of alphabetic
characters, including none. If an NLS locale is defined, its associated definitions of alphabetic and numeric
determine the pattern matching.
An empty string matches the following patterns: ″0A″, ″0X″, ″0N″, ″...″, ″″, ’’, or \\.
BASIC uses characters rather than bytes to determine string length. In NLS mode, MATCHES works in
the same way for multibyte and single-byte character sets. For more information about NLS and character
sets see WebSphere DataStage National Language Support Guide.
IF Operator
The IF statement operator lets you indicate a value conditional upon the truth of another value. The IF
operator has the following syntax:
variable = IF expression THEN expression ELSE expression
variable is assigned the value of the THEN expression if the IF expression is true, otherwise it is assigned
the value of the ELSE expression. The IF operator is similar to the IF statement, but it can sometimes be
more efficient.
Logical Operators
Numeric data, string data, and the null value can function as logical data. Numeric and string data can
have a logical value of true or false. The numeric value 0 (zero), is false; all other numeric values are true.
Character string data other than an empty string is true; an empty string is false. The null value is neither
true nor false. It has the special logical value of null.
Logical operators perform tests on logical expressions. Logical expressions that evaluate to zero or an
empty string are false. Logical expressions that evaluate to null are null. Expressions that evaluate to any
other value are true.
The operands of the logical operators are considered to be logical data types. Tables Table 3 on page 19,
Table 4 on page 19, and Table 5 on page 19 show logical operation results.
Arithmetic and relational operations take precedence over logical operations. WebSphere DataStage
logical operations are evaluated from left to right (AND statements do not take precedence over OR
statements).
Note: The logical value NULL takes the action of false, because the condition is not known to be true.
Assignment Operators
Assignment operators are used in WebSphere DataStage BASIC assignment statements to assign values to
variables. Table 6 shows the operators and their uses.
Table 6. Assignment Operators
Operator Syntax Description
= variable = expression Assigns the value of expression to
variable without any other operation
specified.
+= variable += expression Adds the value of expression to the
previous value of variable and assigns
the sum to variable.
-= variable -= expression Subtracts the value of expression from
the previous value of variable and
assigns the difference to variable.
:= variable := expression Concatenates the previous value of
variable and the value of expression to
form a new value for variable.
Vector Functions
Vector functions process lists of data rather than single values. By using the VEC.MATH (or V) option of
the $OPTIONS statement, the arithmetic operators ( +, -, *, / ) can also operate on dynamic arrays as lists
of data.
The operations performed by vector functions or operators are essentially the same as those performed
by some standard functions or operators. The difference is that standard functions process all variables as
single-valued variables, treating delimiter characters as part of the data. On the other hand, vector
functions recognize delimiter characters and process each field, value, and subvalue individually. In fact,
vector functions process single-valued variables as if they were dynamic arrays with only the first value
defined.
Vector functions have been implemented as subroutines for compatibility with existing WebSphere
DataStage BASIC programs. Each subroutine is assigned a name made up of the function’s name
preceded by a hyphen. For example, the name of the subroutine that performs the ADDS function is
-ADDS. Because the subroutines are cataloged globally, they can be accessed using the method described
in the CALL statement.
The first column of Table 8 shows the functions for manipulating dynamic arrays that are available with
WebSphere DataStage BASIC. The second column shows the corresponding instructions to use for
single-valued variables. In this table, m1 and m2 represent dynamic arrays; s1 and s2 represent
single-valued variables; p1, p2, and so on, represent single-valued parameters. The value of the function is
the resulting dynamic array.
Table 8. Vector Functions
Vector Function Corresponding Instruction for Single-Valued Field
ADDS (m1, m2) s1 + s2
ANDS (m1, m2) s1 AND s2
CATS (m1, m2) s1 : s2
CHARS (m1) CHAR (s1)
COUNTS (m1, p1) COUNT (s1, p1)
When a function or operator processes two dynamic arrays, it processes the lists in parallel. In other
words, the first value of field A and the first value of field B are processed together, then the second
value of field A and the second value of field B are processed together, and so on.
First, the function processing isolates the first field of each dynamic array, which can be written as:
A <1> = 123V456S7890S2468V10
B <1> = 13V57S912
Then the first subvalues of the first values of the first fields are isolated and added:
A <1, 1, 1> = 123
B <1, 1, 1> = 13
This produces the first subvalue of the first value of the first field of the result:
X <1, 1, 1> = 136
Since there are no more subvalues in the first value of either first field, the second values of the first
fields are isolated:
A <1, 2> = 456S7890S2468
B <1, 2> = 57S912
The first subvalues of the second values of the first fields are isolated and added:
A <1, 2, 1> = 456
B <1, 2, 1> = 57
This produces the first subvalue of the second value of the first field of the result:
X <1, 2, 1> = 513
Since B <1, 2, 3> does not exist, it is equal to an empty string. In arithmetic expressions an empty string
equals zero.
Since there are no more subvalues in either second value of the first fields, these values are isolated:
A <1, 3> = 10
B <1, 3> = ""
Since there are no more subvalues or values in either first field, the second fields of each dynamic array
are isolated and the process repeats down to the subvalue levels. The second fields can be written as
follows:
A <2> = 3691V33S12
B <2> = 1234V8
Then the first subvalues of the first values of the second fields are isolated and added:
A <2, 1, 1> = 3691
B <2, 1, 1> = 1234
This produces the first subvalue of the first value of the second field of the result:
X <2, 1, 1> = 4925
Then the first subvalues of the second values of the second fields are isolated and added:
A <2, 2, 1> = 33
B <2, 2, 1> = 8
This produces the first subvalue of the second value of the second field of the result:
X <2, 2, 1> = 41
Then the second subvalues of the second values of the second fields are isolated and added:
A <2, 2, 2> = 12
B <2, 2, 2> = ""
This produces the second subvalue of the second value of the second field of the result:
X <2, 2, 2> = 12
Since there are no more elements in either dynamic array, the result is:
X <1, 1, 1> = 136
X <1, 2, 1> = 513
X <1, 2, 2> = 8802
X <1, 2, 3> = 2468
X <1, 3, 1> = 10
X <2, 1, 1> = 4925
X <2, 2, 1> = 41
X <1, 2, 2> = 12
These elements are put into the resultant dynamic array, separated by the delimiter mark corresponding
to the highest levels that are different (for example, X<1,1,1> and X<1,2,1> have different value levels, so
they are separated by a value mark). This yields the following:
X = 136V513S8802S2468V10F4925V41S12
REUSE Function
If two dynamic arrays are processed by the vector functions described in the preceding section, and they
contain unequal numbers of fields, values, or subvalues, then zeros or empty strings are added to the
shorter list until the two lists are equal.
When you use the REUSE function, the last value in the shorter list is reused until all the elements in the
longer list are exhausted or until the next higher delimiter is encountered.
X is the null value, Y is a dynamic array containing only the character used to represent the null value
(CHAR(128)), and Z is a dynamic array containing two values, CHAR(128) and A, separated by a value
mark.
If you extract all or part of the dynamic array from X, you get the null value in all cases:
X<1> is l
X<2> is l
X<1,2> is l
But if you extract all or part of the dynamic array from Y or Z, you get the null value only when the
extract operator specifically references that element of the array:
Y<1> is l
Y<2> = ""
Y<1,2> is l
Z<1> = ^128VA
Z<2> = ""
Z<1,1> is l
Z<1,2> = A
When the dynamic array extraction finds a string made up of only CHAR(128) between two system
delimiters or between a system delimiter and an end-of-string character, CHAR(128) is converted to the
null value before any other operation is performed.
See the EXTRACT, INSERT, REPLACE, DELETE, REMOVE, and REUSE functions, and the “INS
statement” on page 208, DEL statement, and REMOVE function for details about how BASIC dynamic
array operations handle the null value.
Use the WebSphere DataStage command CREATE.FILE to create a type 1 or type 19 file in which to store
the source code of your BASIC programs. You can create and edit the source code with an operating
system editor (such as vi), the WebSphere DataStage Editor, or a combination of the two.
BASIC
filename is the name of the type 1 or type 19 file containing the BASIC programs to be compiled. You can
compile more than one program at a time if you put all the programs in the same file.
compiles all the programs in BP and puts the output in a record named BASIC_tttt_dddd in the &PH&
file (tttt and dddd are a time and date stamp).
BASIC Options
You can use the following options with the BASIC command:
+$option
Turns on the specified $OPTIONS option, or defines a flavor.
-$option
Turns off the specified $OPTIONS option or flavor.
-I Suppresses execution of RAID or VLIST on a compiler or a BASIC program.
-LIST or -L
Generates a listing of the program.
-XREF or -X
Generates a cross-reference table of statement labels and variable names used in the program.
-SPOOL or -S
Generates a listing of the program and spools it directly to the printer rather than to a file.
-T Suppresses the symbol and line number tables that are usually appended to the end of the object
file, for run-time error messages.
The +$ Option
The +$ option specifies the $OPTIONS options you want to turn on, or the flavor you want the program
to use. See the $OPTIONS statement for the list of options and flavors. You must specify all options you
want to turn on before the options you want to turn off.
The -$ Option
The -$ option specifies the $OPTIONS options, or the flavor, you want to turn off. See the $OPTIONS
statement for the list of options and flavors. You must specify all options you want to turn off after the
options you want to turn on.
The -I Option
The -I option inhibits the execution of RAID or VLIST on your BASIC program. This lets you bypass
subroutines already debugged and provides security to your subroutines.
Compilation Complete.
>ED BP.L DATE.INT13 lines long.
The listing shows three columns: Variable, Type, and References. Variable is the name of the variable or
symbol. Type is one of the following symbol types:
References shows the numbers of all lines in the program that refer to the symbol. Each line number can
have a symbol after it to indicate what the line contains:
* Definition of symbol
= Assignment of variable
! Dimension of array
@ Argument to CALL statement
The -SPOOL option is useful when the listing is very long and you need to look at different parts of the
program simultaneously.
The -T Option
The -T option suppresses the table of symbols and the table of line numbers that are appended to the end
of object files. These tables are used for handling run-time error messages. Suppressing them results in
somewhat smaller object files, but run-time error messages do not know the line number or variable
involved in the error.
Compiler Directives
Compiler directives are BASIC statements that direct the behavior of the compiler. Functions performed
by compiler directives include: inserting source code from one program into another program during
compilation, setting compile-time compatibility with another flavor, and specifying a condition for
compiling certain parts of a program. Most compiler directive statements are prefixed by a dollar sign ( $
).
If you do not specify filename, the included program must be in the same file as the program you are
compiling.
If program is in a different file, you must specify filename in the $INCLUDE statement. filename must be
defined in the VOC file.
The $INSERT statement is included for compatibility with Prime INFORMATION programs. $INSERT is
used, like $INCLUDE, to insert other code in your program during compilation, returning afterward to
compile the next statement in your program.
The PRIMOS path name is converted to a valid filename using the following conversion rules:
Table 9. Conversion rules
Character Converts to Characters
/ converts to ?\
? converts to ??
An ASCII NUL converts to ?0
An initial . converts to ?.
Any leading *> is ignored. If a full path name is specified, the > between directory names changes to a /
to yield the following:
[ pathname/ ] program
$INSERT uses the transformed argument directly as a filename of the file containing the source to be
inserted. It does not use the VOC file.
Use the $UNDEFINE statement to remove an identifier defined by a previous $DEFINE statement from
the symbol table. You can also specify conditional compilation with the $UNDEFINE statement.
Conditional Compilation
You can specify the conditions under which all or part of a BASIC program is to be compiled, using:
v A modified version of the IF statement
v $IFDEF statement
v $IFNDEF statement
Conditional compilation with the modified IF statement is useful for customizing large programs that are
to be used by more than one kind of user. It can also reduce the size of the object code and increase
program efficiency.
You can use the compiler directives $IFDEF and $IFNDEF to control whether or not sections of a
program are compiled. Both of these compiler directives test a given identifier to see if it is currently
defined (that is, has appeared in a $DEFINE statement compiler directive and has not been undefined). If
the identifier that appears in a $IFDEF is defined, all the program source lines appearing between the
$IFDEF compiler directive and the closing $ENDIF compiler directive are compiled. If the identifier is not
defined, all the lines between the $IFDEF compiler directive and the $ENDIF compiler directive are
ignored.
The $IFNDEF compiler directive is the complement to the $IFDEF compiler directive. The lines following
the $IFNDEF compiler directive are included in the compilation if the identifier is not defined. If the
identifier is defined, all lines between the $IFNDEF compiler directive and the $ENDIF compiler directive
are ignored. $IFDEF and $IFNDEF compiler directives can be nested up to 10 deep.
IF Statements
The syntax of the conditional compilation statement is the same as that of the IF statement with the
exception of the test expression, which must be one of the following: $TRUE, $T, $FALSE, or $F. The
syntaxes are as follows:
IF $TRUE THEN statements ELSE statements
IF $T THEN statements ELSE statements
IF $FALSE THEN statements ELSE statements
IF $F THEN statements ELSE statements
The conditional compilation statement can specify a variable name rather than one of the test
specifications listed previously. If it does, an EQUATE statement equating the variable to the test
specification must appear at the beginning of the source code. For example:
EQUATE USER.A LIT "$T"
IF USER.A THEN statements ELSE statements
Consider a program that contains debugging statements in its source code. Using the conditional
compilation statement, you could create two versions of the object code from the same source: a test
version that contains the debugging statements, and a release version without the debugging statements.
The following steps produce the two required versions of the program:
1. Include a conditional debugging statement throughout the source code:
IF TEST PRINT X,Y
If the compile-time symbol is not defined and the $ELSE clause is included, only the statements between
$ELSE and $ENDIF are compiled.
The $ELSE compiler directive introduces the alternative clause of an $IFDEF compiler directive. The
$ENDIF compiler directive marks the end of a conditional compilation block.
In the following example, identifier is not defined so the statements following the $ELSE compiler
directive are compiled. All the statements up to the $ENDIF compiler directive are compiled.
$DEFINE identifier .
.
.
$UNDEFINE identifier$IFDEF identifier [ statements ]
$ELSE
[ statements ]
$ENDIF
If the compile-time symbol is not defined and the $ELSE clause is included, only the statements between
$IFNDEF and $ELSE are compiled.
The $ELSE compiler directive introduces the alternative clause of an $IFNDEF compiler directive. The
$ENDIF compiler directive marks the end of a conditional compilation block.
In the following example the $IFNDEF compiler directive determines that identifier is defined so that the
compiler is forced to compile the statements following the $ELSE compiler directive:
$DEFINE identifier .
.
.
$IFNDEF identifier [ statements ]
$ELSE
[ statements ]
$ENDIF
During compilation, the compiler displays an asterisk on the screen for every 10 lines of source code
successfully compiled. You can monitor the progress of the compilation process by counting the number
of asterisks on the screen at any time. If an error is encountered, a question mark ( ? ) rather than the
asterisk ( * ) is displayed for those 10 lines.
Successful Compilation
When all errors in the source code are corrected, the compiler successfully completes the compilation by
producing an object code record. The object code record is stored in a file whose name is made up of the
source filename suffixed with .O (sourcename.O). The object code record ID is the same as the source file
record ID (program name).
For example, if source code record MAIN is stored in a file called BP, executing the following compile
statement:
>BASIC BP MAIN
compiles the source code in record MAIN in file BP, producing object code that is stored in record MAIN
in file BP.O . The compiler creates the object code file if it does not exist.
The RUN command appends .O to filename and executes the record containing object code in filename.O .
If filename is omitted, the default file, BP, is assumed.
program is the name of the source code of the program. options can be one or more of the following:
NO.WARN
Suppresses all warning (nonfatal) messages.
NO.PAGE
Turns off automatic paging. Programs that position the cursor with @ functions do not need to
disable pagination.
LPTR Spools program output to the printer rather than to the terminal screen.
KEEP.COMMON
If the program is executed from within a chain, links the unnamed common.
TRAP Causes the program to enter the interactive debugger, RAID, whenever a nonfatal error occurs.
For example, the following command executes the record MAIN in the BP.O file:
>RUN BP MAIN
Note: Cataloged programs are considered executable files (that is, the RUN command is not required). To
run a cataloged program, enter its catalog name at the system prompt.
Catalog Space
There are three ways to catalog a program: locally, normally (standard), and globally. Each has different
implications. There is no one best way of cataloging.
Local Cataloging
Local cataloging creates a VOC entry for the program. This entry is a verb that points to the file and
record containing the object code for the cataloged program. A locally cataloged program can be accessed
only from the account in which it was cataloged, unless you copy the VOC entry for the catalog name to
another account.
Since cataloging a program locally only creates a VOC entry pointing to the object file, you need not
catalog the program again every time you recompile it.
Normal Cataloging
Normal cataloging copies the specified object record to the system catalog space and gives it a name of
the form:
*account*catalog.name
Normal cataloging also creates a VOC entry for the program. This entry is a verb that contains the name
*account*catalog in field 2. A normally cataloged program can be accessed only from the account in which
it was cataloged, unless you copy the VOC entry for the catalog name to another account or specify the
full catalog name, including the account prefix.
Since cataloging a program normally copies the object code to the system catalog space, you must catalog
the program again every time you recompile it.
Global Cataloging
Global cataloging copies the specified object record into the system catalog space and gives it a name in
one of the following forms:
*catalog.name-catalog.name$catalog.name!catalog.name
VOC entries are not created for globally cataloged programs. They are available to all accounts on the
system as soon as they are cataloged. The WebSphere DataStage command processor looks in the system
catalog space for verbs or external subroutines that have an initial *. The run machine looks in the system
catalog space for verbs or subroutines whose names begin with *, -, $, or ! .
Because cataloging a program globally copies the object code to the system catalog space, you must
catalog the program again every time you recompile it.
If you simply enter CATALOG at the system prompt, you are prompted for the argument values, one at a
time:
>CATALOGCatalog name or LOCAL =LOCALFile name =BPProgram name =MONTHLY.SALES
If you press Return at any of the prompts, CATALOG terminates without cataloging anything. FORCE or
NOXREF cannot be specified at a prompt. You can specify the LOCAL keyword in response to the
prompt for catalog.name.
If you do not specify catalog.name, CATALOG uses the program name as the catalog name.
DELETE.CATALOG
The DELETE.CATALOG command removes locally, normally, or globally cataloged programs. It has the
following syntax:
DELETE.CATALOG catalog.name
catalog.name is used to determine if the program is either globally or normally cataloged. If it is, the
program is removed from the system catalog. If the program is not in the system catalog, the VOC file is
searched for a local catalog entry. If the program is locally cataloged, only the VOC entry is deleted, not
the object code.
If a program is running when you try to delete it, the deletion does not take effect until the program
terminates.
DECATALOG
The DECATALOG command removes a locally cataloged program. It deletes the object code and removes
the catalog entry from the user’s VOC file. It has the following syntax:
DECATALOG [filename [ [program ] ]
This command deletes the object code of program from the ″.O″ portion of filename. Use an asterisk ( * ) in
place of program to indicate all records in the file. This command can also be executed after building a
select list of programs to be uncataloged.
For example, if 21 users are running the same BASIC program at the same time without catalog shared
memory, and the program code requires 50 kilobytes of memory, the total amount of memory used by
everyone running that program is 21*50, or 1050, kilobytes. On the other hand, if the program is loaded
into catalog shared memory, all 21 users can run one copy of the program, which uses only 50 kilobytes
of memory. In this example, catalog shared memory saves 1000 kilobytes, or one megabyte, of memory.
Before users can have access to programs running from catalog shared memory, the system administrator
must explicitly choose the programs and load them into memory.
Locks
WebSphere DataStage locks control access to records and files among concurrent users. To provide this
control, WebSphere DataStage supports the following two levels of lock granularity:
v Fine granularity (record locks)
v Coarse granularity (file locks)
The level at which you acquire a lock is known as granularity. Record locks affect a smaller component
(the record) and provide a fine level of granularity, whereas file locks affect a larger component (the file)
and provide a coarse level of granularity.
Lock compatibility determines what your process can access when other processes have locks on records or
files. Record locks allow more compatibility because they coexist with other record locks, thus allowing
more transactions to take place concurrently. However, these ″finer-grained″ locks provide a lower
isolation level. File locks enforce a higher isolation level, providing more concurrency control but less
compatibility.
Lock compatibility decreases and isolation level increases as strength and granularity increase. This can
increase the possibility of deadlocks at high isolation levels. Within each granularity level, the strength of
the lock can vary. WebSphere DataStage supports the following locks in order of increasing strength:
v Shared record lock
v Update record lock
v Shared file lock
v Intent file lock
v Exclusive file lock
The locks become less compatible as the granularity, strength, and number of locks increase. Therefore the
number of lock conflicts increase, and fewer users can access the records and files concurrently. To
maximize concurrency, you should acquire the minimum lock required to perform a BASIC statement for
the shortest period of time. The lock can always be promoted to a lock of greater strength or escalated to
a coarser level of granularity if needed.
In WebSphere DataStage BASIC a shared record lock can be acquired with a MATREADL statement,
READL statement, READVL statement, or “RECORDLOCK statements” on page 284, and released with a
CLOSE statement, RELEASE statement, or STOP statement.
In WebSphere DataStage BASIC an update record lock can be acquired or escalated from a shared record
lock with a MATREADU, READU, READVU, or RECORDLOCK statement, and released with a CLOSE,
DELETE, MATWRITE, RELEASE, STOP, WRITE, or The WRITEV statement.
Note: A shared file lock you own is incompatible with an update record lock you own. Be sure to use a
LOCKED clause to avoid deadlocks.
In WebSphere DataStage BASIC a shared file lock can be acquired or promoted with a FILELOCK
statement and released with a CLOSE, Syntax, RELEASE, or STOP statement.
In WebSphere DataStage BASIC, an intent file lock can be acquired or promoted from a shared file lock
with a FILELOCK statement, and released with a CLOSE, Syntax, RELEASE, or STOP statement.
In WebSphere DataStage BASIC an exclusive file lock can be acquired from a shared file lock with a
FILELOCK statement and released with a CLOSE, FILEUNLOCK, RELEASE, or STOP statement.
Deadlocks
Deadlocks occur when two users who acquire locks incrementally try to acquire a lock that the other user
owns, and the existing lock is incompatible with the requested lock. The following situations can lead to
deadlocks:
v Lock promotion from a shared record or shared file lock to a stronger lock
v Lock escalation to file locks when two users try to escalate at the same time
You can configure WebSphere DataStage to automatically identify and resolve deadlocks as they occur
through the Deadlock Daemon Administration menu, or you can manually fix a deadlock by selecting
Transactions
A transaction is a group of logically related operations on the database. In a transaction either the entire
sequence of operations or nothing at all is applied to the database. For example, a banking transaction
that involves the transfer of funds from one account to another involves two logically related operations:
a withdrawal and a deposit. Both operations, or neither, must be performed if the accounts concerned are
to remain reconciled.
Active Transactions
WebSphere DataStage BASIC supports nested transactions. Any transaction can include:
v Read and write operations
v Other transactions or subtransactions that can contain other operations or other transactions
When a transaction begins, it is active. If a second transaction begins before the first transaction is
committed or rolled back, the new (child) transaction becomes the active transaction while the first
(parent) transaction continues to exist but inactively. The child transaction remains active until:
v It is committed or rolled back, when the parent transaction becomes active again
v Another transaction (child) begins and becomes the active transaction
Only one transaction can be active at any time, although many transactions can exist concurrently. Only
one transaction can exist at each transaction nesting level. The top-level transaction is at nesting level 1.
When no transactions exist, the nesting level is 0.
When a read operation for a record occurs, the most recent image of the record is returned to the user.
This image is retrieved from the active transaction cache. If it is not found there, the parent transactions
are then searched. Finally, if it is not found in the parent transactions, the image is retrieved from the
database file.
When a child transaction is committed, the operations are adopted by the parent transaction. When it is
rolled back, the operations are discarded and do not affect the database or the parent transaction.
Transaction Properties
Each transaction must possess properties commonly referred to as the ACID properties:
v Atomicity
v Consistency
v Isolation
v Durability
In nested transactions, child transactions exhibit atomicity, consistency, and isolation properties. They do
not exhibit the durability property since the parent adopts its operation when it is committed. The
Atomicity
Either all the actions of a transaction occur successfully or the transaction is nullified by rolling back all
operations. The transaction system ensures that all operations performed by a successfully committed
transaction are reflected in the database, and the effects of a failed transaction are completely undone.
Consistency
A transaction moves the database from one valid state to another valid state, and if the transaction is
prematurely terminated, the database is returned to its previous valid state.
Isolation
The actions carried out by a transaction cannot become visible to another transaction until the transaction
is committed. Also, a transaction should not be affected by the actions of other concurrent transactions.
WebSphere DataStage provides different isolation levels among concurrently executing transactions.
Durability
Once a transaction completes successfully, its effects cannot be altered without running a compensating
transaction. The changes made by a successful transaction survive subsequent failures of the system.
Serializability
In addition to the ACID properties, SQL standards stipulate that transactions be serializable. Serializability
means that the effects of a set of concurrent transactions should produce the same results as though the
individual transactions were executed in a serial order, and as if each transaction had exclusive use of the
system. In WebSphere DataStage, serializability can be achieved by using isolation level 4 for all
transactions.
If an isolation level is not specified for a child transaction, the isolation level is inherited from the parent
transaction. A specified isolation level should be the same or higher than the parent’s isolation level.
Isolation levels that are lower than the parent transaction’s isolation level are ignored, and the parent’s
If one of these statements is omitted or out of order, the program does not compile.
The run machine also enforces the use of well-formed transactions. A transaction starts when BEGIN
TRANSACTION is executed and ends when COMMIT or ROLLBACK is executed. Program execution
then continues at the statement following the next END TRANSACTION statement.
@Variables
You can use the following @Variables to track transaction activity:
v @ISOLATION
v @TRANSACTION
v @TRANSACTION.ID
v @TRANSACTION.LEVEL
@ISOLATION indicates the current transaction isolation level (0, 1, 2, 3, or 4) for the active transaction, or
the current default isolation level if no transaction exists.
@TRANSACTION is a numeric value that indicates transaction activity. Any nonzero value indicates that
a transaction is active. 0 indicates that no active transaction exists.
@TRANSACTION.ID indicates the transaction number of the active transaction. An empty string
indicates that no transaction exists.
@TRANSACTION.LEVEL indicates the transaction nesting level of the active transaction. 0 indicates that
no transaction exists.
You cannot use the following statements while a transaction is active. Doing so causes a fatal error.
v CLEARFILE statement
v SET TRANSACTION ISOLATION LEVEL statement
You cannot use the EXECUTE or PERFORM statements in a transaction to execute most WebSphere
DataStage commands and SQL statements. However, you can use EXECUTE statement and PERFORM
statement to execute the following WebSphere DataStage commands and SQL statements within a
transaction:
Table 15. Commands permitted in a transaction
CHECK.SUM INSERT SEARCH SSELECT
COUNT LIST SELECT (RetrieVe) STAT
DELETE (SQL) LIST.ITEM SELECT (SQL) SUM
DISPLAY LIST.LABEL SORT UPDATE
ESEARCH RUN SORT.ITEM
GET.LIST SAVE.LIST SORT.LABEL
If a BASIC statement in a transaction has an ON ERROR clause and a fatal error occurs, the ON ERROR
clause is ignored.
Isolation Levels
isolation level
Setting a transaction’s isolation level helps avoid various data anomalies. WebSphere DataStage BASIC
lets you set different isolation levels depending on which data anomalies you want to avoid. Your
transaction runs at the specified isolation level because the transaction subsystem verifies that you have
acquired the required locks for that isolation level. If you have not done so, the program fails.
You can specify isolation levels with the following BASIC statements:
v BEGIN TRANSACTION
v SET TRANSACTION ISOLATION LEVEL
You can use the LOGIN entry in the VOC file to set the isolation level for a session. Do this by including
a SET.SQL command that sets the isolation level. This sets the default isolation level for all transactions
that occur during that session, including WebSphere DataStage commands and SQL statements. For
example, the program might include the statement SET.SQL ISOLATION 2 to set the isolation level to 2
each time a user logs in to the account. This affects all SQL statements and BASIC transactions that occur
during this session.
Each level provides a different degree of protection against the anomalies described in the following
section. Only level 4 provides true serializability. However, due to performance issues, for a large system
we recommend that you use level 2 rather than 3 or 4 for most programs. WebSphere DataStage provides
a default of 0 for backward compatibility.
Data Anomalies
Isolation levels provide protection against the following data anomalies or conflicts that can occur when
two processes concurrently access the data:
v Lost updates occur when two processes try to update an object at the same time. For example, Process
A reads a record. Process B reads the same record, adds 10, and rewrites it. Process A adds 20 to the
record that it previously read, and rewrites the record. Thus, Process B’s update is lost.
v Dirty reads occur when one process modifies a record and a second process reads the record before the
first is committed. If the first process terminates and is rolled back, the second process has read data
that does not exist.
v Nonrepeatable reads occur when a process is unable to ensure repeatable reads. For example, this can
occur if a transaction reads a record, another transaction updates it, then the first transaction rereads it,
and gets a different value the second time.
v Phantom writes occur when a transaction selects a set of records based on selection criteria and
another process writes a record that meets those criteria. The first process repeats the same selection
and gets a different set of records.
Table 16 lists the data anomalies and the isolation levels at which they can occur.
Table 16. Levels at Which Anomalies Can Occur
Anomaly Level 0 Level 1 Level 2 Level 3 Level 4
Lost update No1 No No No No
Dirty read Yes Yes No No No
Nonrepeatable Yes Yes Yes No No
read
Phantom write Yes Yes Yes Yes No
1
Lost updates cannot occur if ISOMODE is set to 2 or 1. If ISOMODE is 0, it is possible for a process
running at isolation level 0 to cause a lost update in your process.
The default ISOMODE setting 1 ensures that BASIC transactions obey the locking rules for isolation level
1, as described in “Isolation Levels and Locks.” This means a record cannot be written or deleted in a
transaction unless the record or file is locked for update. A write or delete of a locked record is known as
a well-formed write.
ISOMODE setting 0 provides compatibility with earlier releases that did not enforce the requirement for
well-formed writes in transactions. Since transactions should always use well-formed writes, we
recommend that you modify any transactions that do not follow this rule as soon as possible, so that you
can set ISOMODE to 1.
Setting ISOMODE to 2 enforces all writes and deletes in BASIC to be well-formed. This mode is available
so that when converting an application to use transactions, you can determine whether any programs
have not yet been converted. You should not use ISOMODE 2 permanently since many system programs
are not (and need not be) transactional.
The BASIC run machine checks that the user has acquired the necessary locks to perform a BASIC
statement. If the minimum locks for the current isolation level are not held by the user, a fatal error
results.
The minimum locks required in a transaction to achieve isolation levels that ensure successful file
operation are listed in Table 17.
Table 17. Isolation Level Locking Requirements
Operation Isolation Level Minimum Lock
Read 0 NO.ISOLATION None
1 READ.UNCOMMITTED None
2 READ.COMMITTED Shared record lock (RL)
3 REPEATABLE.READ Shared record lock (RL)
4 SERIALIZABLE Shared file lock (FS)
1
Different ISOMODE settings affect the locking rules for isolation level 0.
Example
The following example illustrates how isolation levels affect transactions. A transaction running at
isolation level 2 deletes records for Customer 100 from the file CUST. The transaction scans the file
ORDERS for all orders placed by this customer and deletes each order. The part of the transaction that
deletes the orders does not want to lock the ORDERS file unnecessarily.
Note: You cannot run RAID or VLIST on programs compiled with the -I option.
RAID
You can use RAID with your WebSphere DataStage BASIC programs. RAID is both an object code and a
source code debugger-a powerful tool for detecting errors in WebSphere DataStage BASIC code. RAID
lets you do the following:
v Set and delete breakpoints. You can suspend execution of the program at specified lines by setting
breakpoints in the source code. Once RAID suspends program execution, you can examine the source
code, change or display variable values, set additional breakpoints, or delete breakpoints.
v Set watchpoints. You can keep track of changes to variable values by setting watchpoints in the source
code. When a variable’s value changes, RAID can print both the old and new values and the source
instruction that caused the change.
v Step through and display source code, line by line or in segments.
v Examine object addresses.
v Display and modify variables.
v Display all the elements of an array in succession.
You can invoke RAID from the command processor, from within a BASIC program, or by pressing the
Break key while your BASIC program is executing.
filename is the name of the file in which the source code is stored. RAID appends ″.O″ to filename in order
to locate and operate on the object code. If you do not specify filename, RAID assumes the BP file by
default.
program is the name of the record containing the source code of the program.
Use RAID the same way you use RUN. This causes RAID to be invoked just before program execution.
For example, the following command executes the file BP.O/MAIN using the RAID debugger:
>RAID BP MAIN
When you invoke RAID from the command processor, RAID displays the first executable source code
instruction, followed by a double colon ( :: ). Enter a RAID command at the :: prompt. To run the
program, enter R at the :: prompt. To quit RAID, enter Q. RAID commands are discussed in detail in
″RAID Commands″.
The DEBUG statement takes no arguments. When the run machine encounters this statement, the
program stops execution, displays a double colon ( :: ), and prompts you to enter a RAID command.
You can also enter the debugger while a BASIC program is running by pressing the Break key and then
selecting the break option D.
In addition to regular variable names, you can reference some special ″register″ variables with RAID.
WebSphere DataStage BASIC object code is executed by the run machine. When you assign a new
variable, the run machine allocates memory for that variable and creates a pointer to that memory
location. There are special variables that the run machine uses to hold temporary information. These are
the ″registers″ that can be referenced by $R0 through $Rn, and a matrix address variable referenced by
$MATRIX. An arbitrary number of these registers is available, depending on the needs of your program.
The appropriate amount is always made available. You never have more than you need.
Note: Variables that are not referenced are not carried in the symbol table of BASIC object code.
Therefore, RAID can only display the contents of variables referenced in the current subroutine. RAID
ignores all variables that are not referenced and treats them as unknown symbols.
Registers hold intermediate values in a program. For example, for the following statement the sum of 3
and 4 is evaluated and placed in $R0:
A=B:3+4
The reference to the third element of array B is put in the $MATRIX pointer. The object code evaluates
the statement as follows:
REMOVE A FROM $MATRIX SETTING C
RAID Commands
You can enter any RAID command from the double colon ( :: ) prompt. RAID commands have the
following general syntax:
position command qualifier
Table 18. General Syntax of RAID Commands
Argument Explanation Note
position
Tells where and how often to execute
the RAID command in the program.
You can provide one of the following:
line The decimal number of a line of the
source code.
address The hexadecimal address of an object
code instruction, indicated by a
leading 0X.
procedure The name of a procedure in the
source code.
variable The name of a variable in the source
code. You must specify the variable
exactly as it appears in the source
code. Variable names are
case-sensitive, so ″A″ is not the same
as ″a″. Subscript variable to indicate
an element of an array. For example,
A[1,2].
n Indicates the number of times to
execute the command.
qualifier
Can be either of the following:
string A string of characters to search for or
to replace the value of a variable.
* Indicates a special form of the
specified command.
The search begins at the line following the current line. If string is found, RAID sets the current line to
the line containing string and displays that line. If you issue the / command without specifying string,
RAID searches for the last-searched string. This is an empty string if you have not yet specified a string
during this RAID session. If RAID encounters the end of file without finding string, it starts at the first
line and continues the search until it finds string or returns to the current line.
You can set a RAID breakpoint at the current line, an object code address, a BASIC source line number,
the beginning of a specified procedure, or a BASIC source line number within a specified procedure.
RAID recognizes lines in called subroutines. RAID executes the program up to the breakpoint and then
stops and issues the :: prompt. At that point you can issue another RAID command.
The following example sets breakpoints at line 30 and line 60 of the source code, then runs the program.
The program stops executing and displays the RAID prompt when it reaches line 30, and again when it
reaches line 60.
::30B::60B::R
D: Deleting Breakpoints
Use the D command to delete RAID breakpoints. There are two syntaxes:
[ address | line ] D
D*
You can delete a RAID breakpoint at the current line, an object code address, or a BASIC source line
number. If you use the * option, this command deletes all breakpoints. Some BASIC statements produce
multiple run machine statements. If you delete a breakpoint by line number, RAID tries to match the
starting address of the BASIC source number. A breakpoint set at anything other than the starting address
of a BASIC source line must be deleted by its address.
If you use the n option, RAID displays and executes the next n instructions.
M: Setting Watchpoints
Use the M command to set watchpoints. The syntaxes are as follows:
variable M [ ; [ variable ] M ...]
variable =VALUE M
The second syntax lets you set a watchpoint for a variable set to a specific value.
A watchpoint condition occurs when RAID monitors a variable until the variable’s value changes. The
program then suspends operation and displays the variable’s old value, its new value, and the source
instruction that caused the change to occur. If no change occurs, no display appears. This command
accepts up to eight variables. To continue monitoring a variable after RAID has already displayed a
change to that variable’s value, you need only enter M again, not variable M.
Q: Quitting RAID
Use the Q command to quit RAID.
If the line includes a subroutine call, RAID steps into the subroutine. If you use the n option, RAID steps
through the next n lines. If you use the * option, RAID steps around any subroutine call, essentially
treating the entire subroutine as a single line. That is, the S* command instructs RAID to display and
execute a source line. If the line includes a subroutine call, RAID executes the subroutine and displays the
first source line occurring after the subroutine returns.
A window comprises a group of 10 lines of source code, centered around the current line. For example, if
the current line is 4 when you issue the W command, RAID displays the first 10 lines. The W command
by itself does not change the current line. If you use the line option, RAID changes the current line to line
and displays a window centered around that line. For example, if the current line is 14, RAID displays
the lines 9-18. Note that this command affects only the current display lines; it does not affect the
executable lines. That is, if you are stepping through the code using the S command, RAID considers the
current display line to be the last line displayed before issuing the first S command.
The second syntax lets you set a variable to the empty string. The value of X! is displayed with a carriage
return immediately following.
RAID displays the contents. The run machine variables are the following:
Inmat Value set by the INMAT function
Col1 Value set by the COL1 function
Col2 Value set by the COL2 function
Tab Value set by the TABSTOP statement
Precision
Value set by the PRECISION statement
Printer
Printer channel, set by the PRINTER statement
Psw Value set of the last internal comparison
Lsw Value set of the last lock test
Status Value set by the STATUS function
The current window changes each time this command is used, since the last line printed becomes the
current line. For example, if the current line is 14 when you issue the Z command, RAID displays lines
14-23. The current line becomes 23.
If you use the line option, the current line becomes line, and RAID displays a window with the specified
line first. Regardless of the syntax used, the last line printed becomes the current line once you issue this
command. Note that this command affects only the current display lines. It does not affect the executable
lines.
If you use the n option, RAID adds n to the current line. However, the command only displays valid
lines. For example, if you start at line 4 and use the command 3+, the command moves to line 7 only if
line 7 is a valid line for the code. If line 7 is invalid, RAID moves to the first valid line after line 7. Note
that this command does not affect the currently executed line. That is, if you are stepping through the
code using the S command, RAID considers the current line to be the line you displayed before issuing
the first S command.
If you use the n option, RAID subtracts n from the current line. However, the command only displays
valid lines. For example, if you start at line 14 and use the command 3-, the command moves to line 11
only if line 11 is a valid address for the code. If line 11 is invalid, RAID moves backward to the first valid
address before address 11.
To change an array element, you must subscript the variable. This option is not available for debugging
programs in shared memory.
VLIST
Use VLIST to display a listing of the object code of a BASIC program. The syntax for VLIST is as follows:
VLIST [ filename ] program [ R ]
filename
The name of the file containing the source code of the BASIC program. The default filename is
BP.
program
The name of the program to list.
R Displays internal reference numbers for variables and constants rather than source code names
and values.
VLIST displays each line of source code followed by the lines of object code it generated. VLIST also
displays program statistics.
>VLIST BP TO.LISTMain Program "BP.O/TO.LIST"
Compiler Version: 7.3.1.1
Object Level : 5
Machine Type : 1
Local Variables : 1
Subroutine args : 0
Unnamed Common : 0
Named Common Seg: 0
Object Size : 34
Source lines : 4
0001: FOR I = 1 TO 10
0001 0000 : 0F8 move 0 => I
0001 0006 : 098 forincr I 10 1 0020:
0002: PRINT I
0002 0014 : 130 printcrlf I
0003: NEXT I
0003 001A : 0C2 jump 0006:
0004: END
0004 0020 : 190 stop
! statement
Syntax
! [comment.text]
Description
Use the ! statement to insert a comment in a BASIC program. Comments explain or document various
parts of a program. They are part of the source code only and are nonexecutable. They do not affect the
size of the object code.
A comment must be a separate BASIC statement and can appear anywhere in a program. A comment
must begin with one of the following comment designators:
REM * ! $*
Any text that appears between a comment designator and the end of a physical line is treated as part of
the comment, not as part of the executable program. If a comment does not fit on one physical line, you
can continue it on the next physical line only by starting the new line with a comment designator. If a
comment appears at the end of a physical line containing an executable statement, you must put a
semicolon ( ; ) before the comment designator.
Example
The PRINT statement at the end of the third line is not executed because it follows the exclamation point
on the same line and is treated as part of the comment. Lines 4, 5, and 6 show how to include a comment
in the same sequence of executable statements.
001: PRINT "HI THERE"; ! Anything after the ! is a
comment.
002: ! This line is also a comment and does not
print.
003: IF 5<6 THEN PRINT "YES"; ! A comment; PRINT
"PRINT ME"
004: IF 5<6 THEN
005: PRINT "YES"; ! A comment
006: PRINT "PRINT ME"
007: END
Description
Use the #INCLUDE statement to direct the compiler to insert the source code in the record program and
compile it with the main program. The #INCLUDE statement differs from the $CHAIN statement in that
the compiler returns to the main program and continues compiling with the statement following the
#INCLUDE statement.
When program is specified without filename, program must be a record in the same file as the program
containing the #INCLUDE statement.
If program is a record in a different file, the filename must be specified in the #INCLUDE statement,
followed by the name of the program. The filename must specify a type 1 or type 19 file defined in the
VOC file.
The #INCLUDE statement is a synonym for the $INCLUDE and INCLUDE statements.
Example
PRINT "START"
#INCLUDE END
PRINT "FINISH"
When this program is compiled, the #INCLUDE statement inserts code from the program END (see the
example on the END statement page). This is the program output:
START
THESE TWO LINES WILL PRINT ONLY
WHEN THE VALUE OF ’A’ IS ’YES’.
$* statement
Syntax
$* [comment.text]
Description
Use the $* statement to insert a comment in a BASIC program. Comments explain or document various
parts of a program. They are part of the source code only and are nonexecutable. They do not affect the
size of the object code.
A comment must be a separate BASIC statement and can appear anywhere in a program. A comment
must begin with one of the following comment designators:
REM * ! $*
Any text that appears between a comment designator and the end of a physical line is treated as part of
the comment, not as part of the executable program. If a comment does not fit on one physical line, you
can continue it on the next physical line only by starting the new line with a comment designator. If a
Example
The PRINT statement at the end of the third line is not executed because it follows the exclamation point
on the same line and is treated as part of the comment. Lines 4, 5, and 6 show how to include a comment
in the same sequence of executable statements.
001: PRINT "HI THERE"; $* Anything after the $* is a
comment.
002: $* This line is also a comment and does not
print.
003: IF 5<6 THEN PRINT "YES"; $* A comment; PRINT
"PRINT ME"
004: IF 5<6 THEN
005: PRINT "YES"; $* A comment
006: PRINT "PRINT ME"
007: END
$CHAIN statement
Syntax
$CHAIN [filename] program
Description
Use the $CHAIN statement to direct the compiler to read source code from program and compile it as if it
were part of the current program. The $CHAIN statement differs from the $INCLUDE statement,
#INCLUDE, and INCLUDE statements in that the compiler does not return to the main program. Any
statements appearing after the $CHAIN statement are not compiled or executed.
When the program name is specified without a filename, the source code to insert must be in the same
file as the current program.
If the source code to insert is in a different file, the $CHAIN statement must specify the name of the
remote file followed by the program name. filename must specify a type 1 or type 19 file defined in the
VOC file.
When statements in program generate error messages, the messages name the program containing the
$CHAIN statement.
Example
PRINT "START"
$CHAIN END
PRINT "FINISH"
When this program is compiled, the $CHAIN statement inserts code from the program END (see the
example on the END statement page). This is the program output:
$COPYRIGHT statement
Syntax
$COPYRIGHT "copyright.notice"
Description
Use the $COPYRIGHT statement to specify the copyright information that is inserted in the copyright
field of the object header code.
The copyright field in the object code header is set to the empty string at the beginning of compilation. It
remains empty until it encounters a $COPYRIGHT statement in the program.
If more than one $COPYRIGHT statement is included in the program, only the information included in
the last one encountered is inserted in the object code header.
$DEFINE statement
Syntax
$DEFINE identifier [replacement.text]
Description
Use the $DEFINE statement to define identifiers that control program compilation. $DEFINE has two
functions:
v Defining an identifier
v Supplying replacement text for an identifier
replacement.text is a string of characters that the compiler uses to replace identifier everywhere it appears
in the program containing the $DEFINE statement.
When used as a replacement text supplier, $DEFINE adds the specified identifier and its associated
replacement.text to the symbol table. Each time identifier is found in the program following the $DEFINE
statement in which its value was set, it is replaced by replacement.text. If replacement.text is not specified,
identifier is defined and has a null value.
Separate replacement.text from identifier with one or more blanks. Every character typed after this blank is
added to replacement.text up to, but not including, the Return character that terminates the
replacement.text.
Note: Do not use comments when supplying replacement.text because any comments after replacement.text
are included as part of the replacement text. Any comments added to replacement.text might cause
unexpected program behavior.
Conditional Compilation
You can use $DEFINE with the $IFDEF statement or $IFNDEF statement to define an identifier that
controls conditional compilation. The syntax is as follows:
$DEFINE identifier [replacement.text]
.
.
.
{ $IFDEF | $IFNDEF } identifier [ statements ]
$ELSE
[ statements ]
$ENDIF
The $IFDEF or $IFNDEF statement that begins the conditional compilation block tests identifier to
determine whether it is defined by a $DEFINE statement. If you use $IFDEF and identifier is defined, the
statements between the $IFDEF and the $ELSE statements are compiled. If identifier is not defined, the
statements between the $ELSE and $ENDIF statements are compiled.
If you use $IFNDEF, on the other hand, and identifier is defined, the statements between $ELSE and
$ENDIF are compiled. If identifier is not defined, the statements between the $IFDEF and $ELSE
statements are compiled.
Example
In this example the identifier NAME.SUFFIX is defined to have a value of PROGRAM.NAME[5]. When
the compiler processes the next line, it finds the symbol NAME.SUFFIX, substitutes PROGRAM.NAME[5]
in its place and continues processing with the first character of the replacement text.
$DEFINE NAME.SUFFIX PROGRAM.NAME[5]
IF NAME.SUFFIX = ’.B’ THEN
.
.
.
END
.
.
.
$EJECT statement
Syntax
$EJECT
Description
Use the $EJECT statement to begin a new page in the listing record.
$IFDEF statement
Syntax
$IFDEF identifier [ statements ]
[ [ $ELSE ]
[ statements ] ]
$ENDIF
Use the $IFDEF statement to test for the definition of a compile-time symbol. $IFDEF tests to see if
identifier is currently defined (that is, has appeared in a $DEFINE statement and has not been undefined).
If identifier is currently defined and the $ELSE clause is omitted, the statements between the $IFDEF and
$ENDIF statements are compiled. If the $ELSE clause is included, only the statements between $IFDEF
and $ELSE are compiled.
If identifier is not defined and the $ELSE clause is omitted, all the lines between the $IFDEF and $ENDIF
statements are ignored. If the $ELSE clause is included, only the statements between $ELSE and $ENDIF
are compiled.
Example
$IFNDEF statement
Syntax
$IFNDEF identifier [ statements ]
[ [ $ELSE ]
[ statements ] ]
$ENDIF
Description
Use the $IFNDEF statement to test for the definition of a compile-time symbol. The $IFNDEF statement
complements the $IFDEF statement.
If identifier is currently not defined and the $ELSE clause is omitted, the statements between the $IFNDEF
and $ENDIF statements are compiled. If the $ELSE clause is included, only the statements between
$IFNDEF and $ELSE are compiled.
If identifier is defined and the $ELSE clause is omitted, all the lines between the $IFNDEF and $ENDIF
statements are ignored. If the $ELSE clause is included, only the statements between $ELSE and $ENDIF
are compiled.
Example
Description
Use the $INCLUDE statement to direct the compiler to insert the source code in the record program and
compile it with the main program. The $INCLUDE statement differs from the $CHAIN statement in that
the compiler returns to the main program and continues compiling with the statement following the
$INCLUDE statement.
When program is specified without filename, program must be a record in the same file as the program
currently containing the $INCLUDE statement.
If program is a record in a different file, the filename must be specified in the $INCLUDE statement,
followed by the name of the program. The filename must specify a type 1 or type 19 file defined in the
VOC file.
The $INCLUDE statement is a synonym for the #INCLUDE and INCLUDE statements.
Example
PRINT "START"
$INCLUDE END
PRINT "FINISH"
When this program is compiled, the $INCLUDE statement inserts code from the program END (see the
example on the END statement page). This is the program output:
START
THESE TWO LINES WILL PRINT ONLY
WHEN THE VALUE OF ’A’ IS ’YES’.
$INSERT statement
Syntax
$INSERT primos.pathname
Description
Use the $INSERT statement to direct the compiler to insert the source code contained in the file specified
by primos.pathname and compile it with the main program. The difference between $INSERT and
$INCLUDE statement (and its synonyms #INCLUDE and INCLUDE) is that $INSERT takes a PRIMOS
path name as an argument, whereas $INCLUDE takes a WebSphere DataStage filename and record ID.
The PRIMOS path name is converted to a path name; any leading *> is ignored.
$INSERT is included for compatibility with Prime INFORMATION programs; the $INCLUDE statement
is recommended for general use.
If primos.pathname is the name of the program only, it is interpreted as a relative path name. In this case,
the program must be a file in the same directory as the program containing the $INSERT statement.
primos.pathname is converted to a valid path name using the following conversion rules:
/ is converted to ?\
? is converted to ??
ASCII CHAR 0 (NUL) is converted to ?0
. (period) is converted to ?.
If you specify a full path name, the > between directory names changes to a / to yield:
[pathname/] program
$INSERT uses the transformed argument directly as a path name of the file containing the source to be
inserted. It does not use the file definition in the VOC file.
Example
PRINT "START"
$INSERT END
PRINT "FINISH"
When this program is compiled, the $INSERT statement inserts code from the program END (see the
example on the END statement page). This is the program output:
START
THESE TWO LINES WILL PRINT ONLY
WHEN THE VALUE OF ’A’ IS ’YES’.
$MAP statement
Syntax
$MAP mapname
Description
In NLS mode, use the $MAP statement to direct the compiler to specify the map for the source code. Use
the $MAP statement if you use embedded literal strings that contain non-ASCII characters.
mapname must be the name of a map that has been built and installed.
Note: You can execute programs that contain only ASCII characters whether NLS mode is on or off. You
cannot execute programs that contain non-ASCII characters that were compiled in NLS mode if NLS
mode is switched off.
For more information, see WebSphere DataStage National Language Support Guide.
Example
The following example assigns a string containing the three characters alpha, beta, and gamma to the
variable GREEKABG:
$OPTIONS statement
Syntax
$OPTIONS [flavor] [options]
Description
Use the $OPTIONS statement to set compile-time emulation of any flavor. This does not allow object
code compiled in one flavor to execute in another flavor. You can select individual options in a program
to override the default setting.
For instance, the following statement instructs the compiler to treat all BASIC syntax as if it were running
in a PICK flavor account:
$OPTIONS PICK
Another way to select compile-time emulation is to specify one of the following keywords in field 6 of
the VOC entry for the BASIC command:
INFORMATION.FORMAT
PICK.FORMAT
REALITY.FORMAT
IN2.FORMAT
PIOPEN.FORMAT
By default the VOC entry for the BASIC command corresponds with the account flavor specified when
your account was set up.
options are specified by the keywords listed in following table. To turn off an option, prefix it with a
minus sign ( - ).
Table 20. Options for the $OPTIONS Statement
Option Name Option Letter Description
CASE none Differentiates between uppercase and
lowercase identifiers and keywords.
You can also set individual options by using special versions of some statements to override the current
setting. These are listed as follows:
Table 21. Override Versions for Statements
Statement Equal to...
ABORTE ABORT statement with $OPTIONS STOP.MSG
ABORTM ABORT with $OPTIONS -STOP.MSG
HEADINGE HEADING statement with $OPTIONS HEADER.EJECT
HEADINGN HEADING with $OPTIONS -HEADER.EJECT
SELECTV SELECT statements with $OPTIONS VAR.SELECT
SELECTN SELECT with $OPTIONS -VAR.SELECT
STOPE STOP statement with $OPTIONS STOP.MSG
STOPM STOP with $OPTIONS -STOP.MSG
The default settings for each flavor are listed in the following table:
Table 22. Default Settings of $OPTIONS Options
IDEAL PICK INFO. REALITY IN2 PIOPEN
CASE 3
COMP.PRECISION
COUNT.OVLP 3 3 3
END.WARN 3 3 3
EXEC.EQ.PERF 3 3
EXTRA.DELIM 3 3 3
Compilation Complete.
>ED BP OPT4 lines long.
----: P0001: $OPTIONS PICK
0002: A=’12’
0003: B=’14’
0004: PRINT A,B
Bottom at line 4
----: Q>BASIC BP OPTCompiling: Source = ’BP/OPT’, Object = ’BP.O/OPT’
Compilation Complete.
$PAGE statement
The $PAGE statement is a synonym for the $EJECT statement.
$UNDEFINE statement
Syntax
$UNDEFINE identifier
Description
Use the $UNDEFINE statement to remove the definition of identifiers set with the $DEFINE statement.
The $UNDEFINE statement removes the definition of identifier from the symbol table if it appeared in a
previous $DEFINE statement. If the identifier was not previously defined, $UNDEFINE has no effect.
identifier is the identifier whose definition is to be deleted from the symbol table.
You can use $UNDEFINE with the $IFDEF statement or $IFNDEF statement to undefine an identifier that
controls conditional compilation. The syntax is as follows:
$UNDEFINE identifier .
.
.
{ $IFDEF | $IFNDEF } identifier [ statements ]
$ELSE
[ statements ]
$ENDIF
The $IFDEF statement that begins the conditional compilation block tests identifier to determine whether
it is currently defined. Using this syntax, the $UNDEFINE statement deletes the definition of identifier
from the symbol table, and the statements between the $ELSE and the $ENDIF statements are compiled.
If you use the $IFNDEF statement, on the other hand, and identifier is undefined, the statements between
$IFDEF and $ENDIF are compiled. If identifier is not defined, the statements between $IFDEF and $ELSE
are compiled.
Description
Use the * statement to insert a comment in a BASIC program. Comments explain or document various
parts of a program. They are part of the source code only and are nonexecutable. They do not affect the
size of the object code.
A comment must be a separate BASIC statement, and can appear anywhere in a program. A comment
must begin with one of the following comment designators:
REM * ! $*
Any text that appears between a comment designator and the end of a physical line is treated as part of
the comment, not as part of the executable program. If a comment does not fit on one physical line, you
can continue it on the next physical line only by starting the new line with a comment designator. If a
comment appears at the end of a physical line containing an executable statement, you must put a
semicolon ( ; ) before the comment designator.
Example
The PRINT statement at the end of the third line is not executed because it follows the asterisk on the
same line and is treated as part of the comment. Lines 4, 5, and 6 show how to include a comment in the
same sequence of executable statements.
001: PRINT "HI THERE"; * Anything after the * is a
comment
002: * This line is also a comment and does not
print.
003: IF 5<6 THEN PRINT "YES"; * A comment; PRINT
"PRINT ME"
004: IF 5<6 THEN
005: PRINT "YES"; * A comment
006: PRINT "PRINT ME"
007: END
Description
Use the < > operator (angle brackets) to extract or replace elements of a dynamic array.
Angle brackets to the right of an assignment operator indicate that an EXTRACT function is to be
performed. For examples, see the EXTRACT function.
@ function
Syntax
@ (column [,row])
@ (-code [ ,arg ])
Description
Use the @ function with the PRINT statement to control display attributes, screen display, and cursor
positioning.
Note: You can save processing time by assigning the result of a commonly used @ function, such as @
(-1), to a variable, rather than reevaluating the function each time it is used.
-code is the terminal control code that specifies a particular screen or cursor function.
arg specifies further information for the screen or cursor function specified in - code.
Cursor Positioning
You position the cursor by specifying a screen column and row position using the syntax @ (column [ ,row
]). If you do not specify a row, the current row is the default. The top line is row 0, the leftmost column
is column 0. If you specify a column or row value that is out of range, the effect of the function is
undefined.
If you use the @ function to position the cursor, automatic screen pagination is disabled.
You can use the @ function with terminal control codes to specify various cursor and display operations
using the syntax @ (-code [ ,arg ]).
If you want to use mnemonics rather than the code numbers, you can use an insert file of equate names
by specifying either of the following options when you compile your program:
$INCLUDE UNIVERSE.INCLUDE ATFUNCTIONS.H
$INCLUDE SYSCOM ATFUNCTIONS.INS.IBAS (PIOPEN flavor
only)
Note: Not all terminal control codes are supported by all terminal types. If the current terminal type does
not support the code you specified, the function returns an empty string. You can use this to test whether
your program operates correctly on a particular terminal, and whether you need to code any alternative
actions.
Note: If you issue multiple video attributes (such as blink and reverse video) at the same time, the result
is undefined.
Clears the screen and positions the cursor in the upper-left corner.
Clears the current screen line starting at the position under the cursor to the end of that line and clears
all lines below that line. The cursor does not move.
Clears the current screen line starting at the position under the cursor to the end of that line. The cursor
does not move.
Causes any printable characters that are subsequently displayed to blink. If you move the cursor before
issuing the stop blink function, @(IT$EBLINK), the operation of the @(IT$SBLINK) code is undefined.
Stops blink mode. If a start blink function, @(IT$SBLINK), was not transmitted previously, the effect of
this sequence is undefined.
Protects all printable characters that are subsequently displayed from update until the characters are
erased by one of the clear functions @(IT$CS), @(IT$CLEOS), or @(IT$CLEOL). If you move the cursor
before issuing the stop protect function, @(IT$EPA), the operation of this code is undefined. The start
protect function is useful only for terminals that are in block mode.
Stops the protect mode. If a start protect string was not previously transmitted, the effect of this sequence
is undefined. The stop protect function is useful only for terminals that are in block mode.
Moves the cursor one position to the left without deleting any data. For m greater than 0, the function
@(IT$CUB, m) moves the cursor m positions to the left. In moving to the left, the cursor cannot move
beyond the start of the line.
Moves the cursor up one line toward the top of the screen. For m greater than 0, the function @(IT$CUU,
m) moves the cursor up m lines. The cursor remains in the same column, and cannot move beyond the
top of the screen.
Causes all printable characters that are subsequently displayed to be displayed at reduced intensity. If a
cursor-positioning sequence is used before the stop half-intensity function, @(IT$EHALF), the operation of
this function is undefined.
Terminates reverse video mode. If a start reverse video function, @(IT$SREV), was not previously
transmitted, the effect of this sequence is undefined.
Terminates the underlining mode established by a start underlining function, @(IT$SUL). The effect of
this sequence is unspecified if a start underlining string was not previously transmitted.
Inserts a blank line at the current cursor position. For m greater than 0, the function @(IT$IL, m) inserts m
blank lines at the current cursor position. If m is omitted, the default is 1. The effect when m is less than 1
is undefined. All lines from the current cursor position to the end of the screen scroll down. The bottom
m lines on the screen are lost.
Deletes the line at the current cursor position; the function @(IT$DL, 1) has the same effect. For m greater
than 1, the lines above the current line are deleted until m minus 1 lines have been deleted or the top of
the file has been reached, whichever occurs first. All lines below the current cursor position scroll up. The
last lines on the screen are cleared.
Inserts a space at the current cursor position. All characters from the cursor position to the right edge of
the screen are shifted over one character to the right. Any character at the rightmost edge of the screen is
lost. For m greater than 0, the function @(IT$ICH, m) inserts m spaces at the current cursor position,
shifting the other characters accordingly.
Starts insert character mode. Characters sent to the terminal screen are inserted at the current cursor
position instead of overwriting the character under the cursor. The characters under and to the right of
the cursor are shifted over one character to the right for each character transmitted, and any character at
the rightmost edge of the screen is lost.
Turns off insert character mode. Characters sent to the terminal screen overwrite the characters at the
current cursor position.
Deletes the character at the current cursor position. All characters to the right of the cursor move one
space to the left, and the last character position on the line is made blank. For m greater than 1, the
function @(IT$DCH, m) deletes further characters, to the right of the original position, until m characters
have been deleted altogether or until the end of the display has been reached, whichever occurs first.
Enables the auxiliary (printer) port on the terminal. All characters sent to the terminal are displayed on
the screen and also copied to the auxiliary port.
Disables the auxiliary (printer) port on the terminal, and stops the copying of the character stream to the
auxiliary port.
Places the auxiliary (printer) port on the terminal in transparent mode. All characters sent to the terminal
are sent only to the auxiliary port and are not displayed on the terminal screen.
Disables the auxiliary (printer) port on the terminal and enables the display of the character stream on
the terminal screen.
Sets a time, in milliseconds, that an application should pause after enabling or disabling the auxiliary
port. The value of this function is an integer in the range 0 through 32,767. The function is used in
conjunction with the !SLEEP$ subroutine; for example:
PRINT @(IT$AUXON):;CALL !SLEEP$(@(IT$AUXDLY))
Copies the contents of the screen to the auxiliary port. The function does not work for some terminals
while echo delay is enabled.
Starts 80-column mode. On some terminals it can also clear the screen.
Starts 132-column mode. On some terminals it can also clear the screen.
Moves the cursor down one line. For m greater than 0, the function @(IT$CUD, m) moves the cursor
down m lines. The cursor remains in the same column, and cannot move beyond the bottom of the
screen.
Moves the cursor to the right by one character position without overwriting any data. For m greater than
0, the function @(IT$CUF, m) moves the cursor m positions to the right. The cursor cannot move beyond
the end of the line.
Is an implementation of the ANSI X3.64-1979 and ISO 6429 standards for the video attribute portion of
Select Graphic Rendition. It always carries an argument m that is an additive key consisting of one or
more of the following video attribute keys:
In this example, m is set to 74 (2 + 8 + 64) for half-intensity underline display in reverse video. Bold,
italic, fast blink, and concealed are not supported on all terminals. To set the video attributes
half-intensity and underline, specify the following:
@(-35,10)
Sets the background and foreground colors to a combination that you have previously defined in your
system terminfo file.
Sets the color that is used to display characters on the screen. @(IT$FCOLOR, arg) always takes an
argument that specifies the foreground color to be chosen, as follows:
Table 24. Foreground colors
Value Name Description
0 IT$63 Black
1 IT$RED Red
2 IT$GREEN Green
3 IT$YELLOW Yellow
4 IT$BLUE Blue
5 IT$MAGENTA Magenta
6 IT$CYAN Cyan
7 IT$WHITE White
8 IT$DARK.RED Dark red
9 IT$CERISE Cerise
10 IT$ORANGE Orange
11 IT$PINK Pink
12 IT$DARK.GREEN Dark green
13 IT$SEA.GREEN Sea green
14 IT$LIME.GREEN Lime green
15 IT$PALE.GREEN Pale green
16 IT$BROWN Brown
17 IT$CREAM Cream
18 IT$DARK.BLUE Dark blue
19 IT$SLATE.BLUE Slate blue
20 IT$VIOLET Violet
21 IT$PALE.BLUE Pale blue
22 IT$PURPLE Purple
The color attributes are not additive. Only one foreground color at a time can be displayed. If a terminal
does not support a particular color, a request for that color should return an empty string.
Sets the background color that is used to display characters on the screen. The @(IT$BCOLOR, arg)
function always has an argument that specifies the background color to be chosen. (See ″Set Foreground
Color @(IT$FCOLOR)″ on page 1-42 for a list of available colors.)
Switches on the line graphics mode for drawing boxes or lines on the screen.
Specifies the line graphics character required. The argument can be one of the following:
Table 25. Graphics characters
Value Token Description
0 IT$GRFX.CROSS Cross piece
1 IT$GRFX.H.LINE Horizontal line
2 IT$GRFX.V.LINE Vertical line
3 IT$GRFX.TL.CORNER Top-left corner
4 IT$GRFX.TR.CORNER Top-right corner
5 IT$GRFX.BL.CORNER Bottom-left corner
6 IT$GRFX.BR.CORNER Bottom-right corner
7 IT$GRFX.TOP.TEE Top-edge tee piece
8 IT$GRFX.LEFT.TEE Left-edge tee piece
9 IT$GRFX.RIGHT.TEE Right-edge tee piece
10 IT$GRFX.BOTTOM.TEE Bottom-edge tee piece
Blanks the terminal’s display. Subsequent output to the screen is not visible until the unblank screen
function, @(IT$UBS), is used.
Restores the terminal’s display after it was blanked. The previous contents of the screen, and any
subsequent updates, become visible.
Scroll Up @(IT$SU)
Moves the entire contents of the display up one line. For m greater than 0, the function @(IT$SU, m)
moves the display up m lines or until the bottom of the display is reached, whichever occurs first. For
each line that is scrolled, the first line is removed from sight and another line is moved into the last line.
This function works only if the terminal is capable of addressing character positions that do not all fit on
the screen, such that some lines are not displayed. This normally requires the terminal to be set to vertical
two-page mode in the initialization string. The effect of attempting to scroll the terminal too far is
undefined.
Moves the entire contents of the display down one line. For m greater than 0, the function @(IT$SD, m)
moves the display down m lines or until the top of the display is reached, whichever occurs first. For
each line that is scrolled, the last line is removed from sight and another line is moved into the top line.
This function works only if the terminal is capable of addressing character positions that do not all fit on
the screen, such that some lines are not displayed. This normally requires the terminal to be set to vertical
two-page mode in the initialization string. The effect of attempting to scroll the terminal too far is
undefined.
Moves the entire contents of the display one column to the right. For m greater than 0, the function
@(IT$SR, m) moves the display m columns to the right or until the left edge of the display is reached,
whichever occurs first. For each column scrolled, the rightmost column is removed from sight and
another leftmost column appears. This function works only if the terminal is capable of addressing
character positions that do not fit on the screen, such that some columns are not displayed. This normally
requires the terminal to be set to horizontal two-page mode in the initialization string. The effect of
attempting to scroll the terminal too far is undefined.
Moves the entire contents of the display one column to the left. For m greater than 0, the function
@(IT$SL, m) moves the display m columns to the left or until the right edge of the display is reached,
whichever happens first. For each column scrolled, the leftmost column is removed from sight and
another rightmost column appears. This function works only if the terminal is capable of addressing
character positions that do not fit on the screen, such that some columns are not displayed. This normally
requires the terminal to be set to horizontal two-page mode in the initialization string. The effect of
attempting to scroll the terminal too far is undefined.
Makes the cursor stay in the last position on the line when characters are printed past the last position.
Makes the cursor move to the first position on the next line down when characters are printed past the
last position.
Sets keys on the numeric keypad to the labelled functions instead of numbers.
Starts bold mode; subsequently, any characters entered are shown more brightly on the screen.
Characters entered in this setting are not shown on the screen. This function can be used when entering
passwords, for example.
Switches on start protect mode. Characters entered in this mode are not removed when the screen is
cleared.
Redisplays the user-defined characters that were sent by the system line set function, @(IT$SLS). The
system line is defined as an extra line on the terminal display but is addressable by the normal cursor
positioning sequence. On most terminals the system line normally contains a terminal status description.
The number of usable lines on the screen does not change.
Removes from the display the characters that were set by the @(IT$SLS) function and replaces them with
the default system status line. The number of usable lines on the screen does not change.
Displays the user-defined status line, and positions the cursor at the first column of the status line.
Subsequent printing characters sent to the terminal are displayed on the status line. Issuing a system line
reset function, @(IT$SLR), terminates printing on the status line, and leaves the cursor position undefined.
Positions the cursor at column m of the current line. If m is omitted, the default is 0. The @(IT$CHA, m)
function must have the same effect as @(m).
Erases the character under the cursor and replaces it with one or more spaces, determined by the
argument m. If you do not specify m, or you specify a value for m that is less than 2, only the character
under the cursor is replaced. If you specify an argument whose value is greater than 1, the function
replaces the character under the cursor, and m -1 characters to the right of the cursor, with spaces. The
cursor position is unchanged.
Locks line n of the screen display (top line is 0). The line cannot be modified, moved, or deleted from the
screen until it is unlocked.
Returns the characters used to display WebSphere DataStage delimiters on screen. From left to right, the
delimiters are: item, field, value, subvalue, and text.
These functions are available for any additional terminal definition strings that you require.
Terminals whose video attributes are described as embedded or on-screen use a character position on the
terminal screen whenever a start or stop video attribute is received. Programs driving such terminals
must not change an attribute in the middle of a contiguous piece of text. You must leave at least one
blank character position at the point where the attribute changes. The field in the terminal definition
record called xmc is used to specify the number of character positions required for video attributes. A
program can examine this field, and take appropriate action. To do this, the program must execute
GET.TERM.TYPE and examine the @SYSTEM- .RETURN.CODE variable, or use the definition
VIDEO.SPACES from the TERM INFO.H file.
Many terminals do not clear video attributes automatically when the data on a line is cleared or deleted.
The recommended programming practice is to reposition to the point at which a start attribute was
emitted, and overwrite it with an end attribute, before clearing the line.
Note: Where possible, you should try to ensure that any sequences that clear data also clear video
attributes. This might not be the case for all terminal types. An exception is @(IT$CS) clear screen. The
sequence associated with this function should always clear not only all data on the screen but also reset
any video attributes to normal.
Examples
In the next example, the PRINT statement positions the cursor to home, at the top-left corner of the
screen, and clears the screen:
PRINT @(IT$CS):
The $INCLUDE statement is used to include the ATFUNCTIONS insert file of equate names. Assignment
statements are used to assign the evaluated @ functions to variables. The variables are used in PRINT
statements to produce code that clears the screen and returns the cursor to its original position; positions
the cursor at column 5, line 20; turns on the reverse video mode; prints the string; and turns off the
reverse video mode.
$INCLUDE UNIVERSE.INCLUDE ATFUNCTIONS.H
CLS = @(IT$CS)
REVERSE.ON = @(IT$SREV)
REVERSE.OFF = @(IT$EREV)
.
.
.
PRINT CLS: @(5,20):
PRINT REVERSE.ON:"THIS IS REVERSE VIDEO":REVERSE.OFF
The next example gives a yellow foreground, not a green foreground, because color changes are not
additive:
PRINT @(IT$FCOLOR, IT$BLUE):@(IT$FCOLOR, IT$YELLOW)
If you have a terminal that supports colored letters on a colored background, the next example displays
the text ″Hello″ in yellow on a cyan background. All subsequent output is in yellow on cyan until
another color @ function is used. If your color terminal cannot display colored foreground on colored
background, only the last color command is used, so that this example displays the text ″Hello″ in yellow
on a black background.
PRINT
@(IT$BCOLOR,IT$CYAN):@(IT$FCOLOR,IT$YELLOW):"Hello"
If your color terminal cannot display colored foreground on colored background, the previous example
displays the text ″Hello″ in black on a cyan background.
In the last example, the screen is cleared, the cursor is positioned to the tenth column in the tenth line,
and the text ″Hello″ is displayed in foreground color cyan. The foreground color is then changed to white
for subsequent output. This sequence of display instructions can be executed again, whenever it is
required, by a further PRINT SCREEN statement.
SCREEN =
@(IT$CS):@(10,10):@(IT$FCOLOR,IT$CYAN):"Hello"
SCREEN = SCREEN:@(IT$FCOLOR,IT$WHITE)
PRINT SCREEN
[ ] operator
Syntax
expression [ [ start, ] length ]
expression [ delimiter, occurrence, fields ]
Description
Use the [ ] operator (square brackets) to extract a substring from a character string. The bold brackets are
part of the syntax and must be typed.
start is an expression that evaluates to the starting character position of the substring. If start is 0 or a
negative number, the starting position is assumed to be 1. If you omit start, the starting position is
calculated according to the following formula:
string.length - substring.length + 1
This lets you specify a substring consisting of the last n characters of a string without having to calculate
the string length.
If start exceeds the number of characters in expression, an empty string results. An empty string also
results if length is 0 or a negative number. If the sum of start and length exceeds the number of characters
in the string, the substring ends with the last character of the string.
Use the second syntax to return a substring located between the specified number of occurrences of the
specified delimiter. This syntax performs the same function as the FIELD function.
delimiter can be any string, including field mark, value mark, and subvalue mark characters. It delimits
the start and end of the substring (all that appears within the two delimiters). If delimiter consists of more
than one character, only the first character is used.
occurrence specifies which occurrence of the delimiter is to be used as a terminator. If occurrence is less
than 1, 1 is assumed.
fields specifies the number of successive fields after the delimiter specified by occurrence that are to be
returned with the substring. If the value of fields is less than 1, 1 is assumed. The delimiter is part of the
returned value in the successive fields.
If expression is the null value, any substring extracted from it will also be the null value.
Examples
In the following example (using the second syntax) the fourth # is the terminator of the substring to be
extracted, and one field is extracted:
A="###DHHH#KK"
PRINT A["#",4,1]
All substring syntaxes can be used in conjunction with the assignment operator ( = ). The new value
assigned to the variable replaces the substring specified by the [ ] operator. For example:
A=’12345’
A[3]=1212
PRINT "A=",A
A[3] replaces the last three characters of A (345) with the newly assigned value for that substring (1212).
The FIELDSTORE function provides the same functionality as assigning the three-argument syntax of the
[ ] operator.
ABORT statement
Syntax
ABORT [expression ...]
ABORTE [expression ...]
ABORTM [expression ...]
Description
Use the ABORT statement to terminate execution of a BASIC program and return to the command
prompt. ABORT differs from STOP in that a STOP statement returns to the calling environment (for
example, a menu, a paragraph, another BASIC program following an EXECUTE statement, and so on),
whereas ABORT terminates all calling environments as well as the BASIC program. You can use it as part
of an IF...THEN statement to terminate processing if certain conditions exist.
If expression is used, it is printed when the program terminates. If expression evaluates to the null value,
nothing is printed.
Chapter 6. BASIC Statements and Functions 87
The ABORTE statement is the same as the ABORT statement except that it behaves as if $OPTIONS
STOP.MSG were in force. This causes ABORT to use the ERRMSG file to produce error messages instead
of using the specified text. If expression in the ABORTE statement evaluates to the null value, the default
error message is printed:
Message ID is NULL: undefined error
For information about the ERRMSG file, see the ERRMSG statement.
The ABORTM statement is the same as the ABORT statement except that it behaves as if $OPTIONS
-STOP.MSG were in force. This causes ABORT to use the specified text instead of text from the ERRMSG
file.
Example
PRINT "DO YOU WANT TO CONTINUE?":
INPUT A
IF A="NO" THEN ABORT
ABS function
Syntax
ABS (expression)
Description
Use the ABS function to return the absolute value of any numeric expression. The absolute value of an
expression is its unsigned magnitude. If expression is negative, the value returned is:
-expression
If expression is positive, the value of expression is returned. If expression evaluates to the null value, null is
returned.
Example
Y = 100
X = ABS(43-Y)
PRINT X
ABSS function
Syntax
ABSS (dynamic.array)
Description
Use the ABSS function to return the absolute values of all the elements in a dynamic array. If an element
in dynamic.array is the null value, null is returned for that element.
ACOS function
Syntax
ACOS (expression)
Description
Use the ACOS function to return the trigonometric arc-cosine of expression. expression must be a numeric
value. The result is expressed in degrees. If expression evaluates to the null value, null is returned. The
ACOS function is the inverse of the COS function.
Example
PRECISION 5
PRINT "ACOS(0.707106781) = ":ACOS(0.707106781):"
degrees"
ADDS function
Syntax
ADDS (array1, array2)
CALL -ADDS (return.array, array1, array2)
CALL !ADDS (return.array, array1, array2)
Description
Use the ADDS function to create a dynamic array of the element-by-element addition of two dynamic
arrays.
Each element of array1 is added to the corresponding element of array2. The result is returned in the
corresponding element of a new dynamic array. If an element of one array has no corresponding element
in the other array, the existing element is returned. If an element of one array is the null value, null is
returned for the sum of the corresponding elements.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A = 2:@VM:4:@VM:6:@SM:10
B = 1:@VM:2:@VM:3:@VM:4
PRINT ADDS(A,B)
Description
Use the ALPHA function to determine whether expression is an alphabetic or non-alphabetic string. If
expression contains the characters a through z or A through Z, it evaluates to true and a value of 1 is
returned. If expression contains any other character or an empty string, it evaluates to false and a value of
0 is returned. If expression evaluates to the null value, null is returned.
If NLS is enabled, the ALPHA function uses the characters in the Alphabetics field in the NLS.LC.CTYPE
file. For more information, see WebSphere DataStage National Language Support Guide.
Example
PRINT "ALPHA(’ABCDEFG’) = ":ALPHA(’ABCDEFG’)
PRINT "ALPHA(’abcdefg’) = ":ALPHA(’abcdefg’)
PRINT "ALPHA(’ABCDEFG.’) = ":ALPHA(’ABCDEFG.’)
PRINT "ALPHA(’SEE DICK’) = ":ALPHA(’SEE DICK’)
PRINT "ALPHA(’4 SCORE’) = ":ALPHA(’4 SCORE’)
PRINT "ALPHA(’’) = ":ALPHA(’’)
ANDS function
Syntax
ANDS (array1, array2)
CALL -ANDS (return.array, array1, array2)
CALL !ANDS (return.array, array1, array2)
Description
Use the ANDS function to create a dynamic array of the logical AND of corresponding elements of two
dynamic arrays.
Each element of the new dynamic array is the logical AND of the corresponding elements of array1 and
array2. If an element of one dynamic array has no corresponding element in the other dynamic array, a
false (0) is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
If both corresponding elements of array1 and array2 are the null value, null is returned for those elements.
If one element is the null value and the other is 0 or an empty string, a false is returned for those
elements.
Example
A = 1:@SM:4:@VM:4:@SM:1
B = 1:@SM:1-1:@VM:2
PRINT ANDS(A,B)
ASCII function
Syntax
ASCII (expression)
Description
Use the ASCII function to convert each character of expression from its EBCDIC representation value to its
ASCII representation value. If expression evaluates to the null value, null is returned.
The ASCII function and the EBCDIC function perform complementary operations.
Example
X = EBCDIC(’ABC 123’)
Y = ASCII(X)
PRINT "EBCDIC", "ASCII", " Y "
PRINT "------", "-----", "---"
FOR I = 1 TO LEN (X)
PRINT SEQ(X[I,1]) , SEQ(Y[I,1]),Y[I,1]
NEXT I
ASIN function
Syntax
ASIN (expression)
Description
Use the ASIN function to return the trigonometric arc-sine of expression. expression must be a numeric
value. The result is expressed in degrees. If expression evaluates to the null value, null is returned. The
ASIN function is the inverse of the SIN function.
Example
PRECISION 5
PRINT "ASIN(0.707106781) = ":ASIN(0.707106781):"
degrees"
Description
Use the ASSIGNED function to determine if variable is assigned a value. ASSIGNED returns 1 (true) if
variable is assigned a value, including common variables and the null value. It returns 0 (false) if variable
is not assigned a value.
PICK Flavor
When you run in a PICK flavor account, all common variables are initially unassigned. ASSIGNED
returns 0 (false) for common variables until the program explicitly assigns them a value.
Example
A = "15 STATE STREET"
C = 23
X = ASSIGNED(A)
Y = ASSIGNED(B)
Z = ASSIGNED(C)
PRINT X,Y,Z
assignment statements
Syntax
variable = expression
variable += expression
variable -= expression
variable := expression
Description
Use assignment statements to assign a value to a variable. The variable can be currently unassigned (that
is, one that has not been assigned a value by an assignment statement, a statement, or any other
statement that assigns values to variables) or have an old value that is to be replaced. The assigned value
can be a constant or an expression. It can be any data type (that is, numeric, character string, or the null
value).
Use the operators += , -= , and := to alter the value of a variable. The += operator adds the value of
expression to variable. The -= operator subtracts the value of expression from variable. The := operator
concatenates the value of expression to the end of variable.
Use the system variable @NULL to assign the null value to a variable:
variable = @NULL
Use the system variable @NULL.STR to assign a character string containing only the null value (more
accurately, the character used to represent the null value) to a variable:
variable = @NULL.STR
ATAN function
Syntax
ATAN (expression)
Description
Use the ATAN function to return the trigonometric arc-tangent of expression. expression must be a numeric
value. The result is expressed in degrees. If expression evaluates to the null value, null is returned. The
ATAN function is the inverse of the TAN function.
Examples
The following example prints the numeric value 135 and the angle, in degrees, that has an arc-tangent of
135:
PRINT 135, ATAN(135)
AUTHORIZATION statement
Syntax
AUTHORIZATION "username"
Description
Use the AUTHORIZATION statement to specify or change the effective run-time user of a program. After
an AUTHORIZATION statement is executed, any SQL security checking acts as if username is running the
program.
username is a valid login name on the machine where the program is run. username must be a constant.
username is compiled as a character string whose user identification (UID) number is looked up in the
/etc/passwd file at run time.
An AUTHORIZATION statement changes only the user name that is used for SQL security checking
while the program is running. It does not change the actual user name, nor does it change the user’s
You can change the effective user of a program as many times as you like. The username specified by the
most recently executed AUTHORIZATION statement remains in effect for subsequent EXECUTE
statement and PERFORM statements as well as for subroutines.
When a file is opened, the effective user’s permissions are stored in the file variable. These permissions
apply whenever the file variable is referenced, even if a subsequent AUTHORIZATION statement
changes the effective user name.
A program using the AUTHORIZATION statement must be compiled on the machine where the program
is to run. To compile the AUTHORIZATION statement, SQL DBA privilege is required. If the user
compiling the program does not have DBA privilege, the program will not be compiled. You cannot run
the program on a machine different from the one where it was compiled. If you try, the program
terminates with a fatal error message.
Example
AUTHORIZATION "susan"
OPEN "","SUES.FILE" TO FILE.S ELSE PRINT "CAN’T OPEN
SUES.FILE"
AUTHORIZATION "bill"
OPEN "","BILLS.FILE" TO FILE.B ELSE PRINT "CAN’T OPEN
BILLS.FILE"
FOR ID = 5000 TO 6000
READ SUE.ID FROM FILE.S, ID THEN PRINT ID ELSE
NULL
READ BILL.ID FROM FILE.B, ID THEN PRINT ID ELSE
NULL
NEXT ID
AUXMAP statement
Syntax
AUXMAP { ON | OFF | expression }
Description
In NLS mode, use the AUXMAP statement to associate an auxiliary device with a terminal.
AUXMAP ON causes subsequent PRINT statements directed to print channel 0 to use the auxiliary map.
If no auxiliary map is defined, the terminal map is used. AUXMAP OFF causes subsequent PRINT
statements to use the terminal map. OFF is the default. If expression evaluates to true, AUXMAP is turned
on. If expression evaluates to false, AUXMAP is turned off.
A program can access the map for an auxiliary device only by using the AUXMAP statement. Other
statements used for printing to the terminal channel, such as CRC32 function, PRINT, or INPUTERR
statement, use the terminal map.
If NLS is not enabled and you execute the AUXMAP statement, the program displays a run-time error
message. For more information, see WebSphere DataStage National Language Support Guide.
Description
The ISOLATION LEVEL clause sets the transaction isolation level for the duration of that transaction. The
isolation level reverts to the original value at the end of the transaction.
1
Lost updates are prevented if the ISOMODE configurable parameter is set to 1 or 2.
Examples
BITAND function
Syntax
BITAND (expression1, expression2)
Description
Use the BITAND function to perform the bitwise AND comparison of two integers specified by numeric
expressions. The bitwise AND operation compares two integers bit by bit. It returns a bit of 1 if both bits
are 1; otherwise it returns a bit of 0.
Note: Differences in hardware architecture can make the use of the high-order bit non-portable.
BITNOT function
Syntax
BITNOT (expression [,bit#])
Description
Use the BITNOT function to return the bitwise negation of an integer specified by any numeric
expression.
bit# is an expression that evaluates to the number of the bit to invert. If bit# is unspecified, BITNOT
inverts each bit. It changes each bit of 1 to a bit of 0 and each bit of 0 to a bit of 1. This is equivalent to
returning a value equal to the following:
(-expression)-1
If expression evaluates to the null value, null is returned. If bit# evaluates to the null value, the BITNOT
function fails and the program terminates with a run-time error message.
Values that are not integers are truncated before the operation is performed.
Note: Differences in hardware architecture can make the use of the high-order bit non-portable.
Example
PRINT
BITNOT(6),BITNOT(15,0),BITNOT(15,1),BITNOT(15,2)
BITOR function
Syntax
BITOR (expression1, expression2)
Description
Use the BITOR function to perform the bitwise OR comparison of two integers specified by numeric
expressions. The bitwise OR operation compares two integers bit by bit. It returns the bit 1 if the bit in
either or both numbers is 1; otherwise it returns the bit 0.
Values that are not integers are truncated before the operation is performed.
Example
PRINT BITOR(6,12)
* Binary value of 6 = 0110
* Binary value of 12 = 1100
BITRESET function
Syntax
BITRESET (expression, bit#)
Description
Use the BITRESET function to reset to 0 the bit number of the integer specified by expression. Bits are
counted from right to left. The number of the rightmost bit is 0. If the bit is 0, it is left unchanged.
If expression evaluates to the null value, null is returned. If bit# evaluates to the null value, the BITRESET
function fails and the program terminates with a run-time error message.
Values that are not integers are truncated before the operation is performed.
Example
PRINT BITRESET(29,0),BITRESET(29,3)
* The binary value of 29 = 11101
* The binary value of 28 = 11100
* The binary value of 21 = 10101
PRINT BITRESET(2,1),BITRESET(2,0)
* The binary value of 2 = 10
* The binary value of 0 = 0
BITSET function
Syntax
BITSET (expression, bit#)
Description
Use the BITSET function to set to 1 the bit number of the integer specified by expression. The number of
the rightmost bit is 0. If the bit is 1, it is left unchanged.
If expression evaluates to the null value, null is returned. If bit# evaluates to the null value, the BITSET
function fails and the program terminates with a run-time error message.
Values that are not integers are truncated before the operation is performed.
BITTEST function
Syntax
BITTEST (expression, bit#)
Description
Use the BITTEST function to test the bit number of the integer specified by expression. The function
returns 1 if the bit is set; it returns 0 if it is not. Bits are counted from right to left. The number of the
rightmost bit is 0.
If expression evaluates to the null value, null is returned. If bit# evaluates to null, the BITTEST function
fails and the program terminates with a run-time error message.
Values that are not integers are truncated before the operation is performed.
Example
PRINT
BITTEST(11,0),BITTEST(11,1),BITTEST(11,2),BITTEST(11,
3)
* The binary value of 11 = 1011
BITXOR function
Syntax
BITXOR (expression1, expression2)
Description
Use the BITXOR function to perform the bitwise XOR comparison of two integers specified by numeric
expressions. The bitwise XOR operation compares two integers bit by bit. It returns a bit 1 if only one of
the two bits is 1; otherwise it returns a bit 0.
Values that are not integers are truncated before the operation is performed.
Note: Differences in hardware architecture can make the use of the high-order bit non-portable.
BREAK statement
Syntax
BREAK [KEY] { ON | OFF | expression }
Description
Use the BREAK statement to enable or disable the Intr, Quit, and Susp keys on the keyboard.
When the BREAK ON statement is in effect, pressing Intr, Quit, or Susp causes operations to pause.
When the BREAK OFF statement is in effect, pressing Intr, Quit, or Susp has no effect. This prevents a
break in execution of programs that you do not want interrupted.
When expression is used with the BREAK statement, the value of expression determines the status of the
Intr, Quit, and Susp keys. If expression evaluates to false (0, an empty string, or the null value), the Intr,
Quit, and Susp keys are disabled. If expression evaluates to true (not 0, an empty string, or the null
value), the Intr, Quit, and Susp keys are enabled.
A counter is maintained for the BREAK statement. It counts the number of executed BREAK ON and
BREAK OFF commands. When program control branches to a subroutine, the value of the counter is
maintained; it is not set back to 0. For each BREAK ON statement executed, the counter decrements by 1;
for each BREAK OFF statement executed, the counter increments by 1. The counter cannot go below 0.
The Intr, Quit, and Susp keys are enabled only when the value of the counter is 0. The following
example illustrates the point:
Table 27. Counters and the BREAK Statement
Statement from Command Counter Key Status
- - 0 ON
Main program BREAK OFF +1 OFF
Subroutine BREAK OFF +2 OFF
Subroutine BREAK ON +1 OFF
Main program BREAK ON 0 ON
Examples
The following example disables the Intr, Quit, and Susp keys if QTY is false, 0, an empty string, or the
null value; it enables them if QTY is true, not 0, not an empty string, or not the null value:
BREAK QTY ;*
Description
Use the BSCAN statement to scan the leaf nodes of a B-tree file (type 25) or of a secondary index. The
record ID returned by the current scan operation is assigned to ID.variable. If you specify rec.variable, the
contents of the record whose ID is ID.variable is assigned to it.
file.variable specifies an open file. If file.variable is not specified, the default file is assumed (for more
information on default files, see the OPEN statement). If the file is neither accessible nor open, the
program terminates with a run-time error message.
record is an expression that evaluates to a record ID of a record in the B-tree file. If the USING clause is
used, record is a value in the specified index. record specifies the relative starting position of the scan.
record need not exactly match an existing record ID or value. If it does not, the scan finds the next or
previous record ID or value, depending on whether the scan is in ascending or descending order. For
example, depending on how precisely you want to specify the starting point at or near the record ID or
value SMITH, record can evaluate to SMITH, SMIT, SMI, SM, or S.
If you do not specify record, the scan starts at the leftmost slot of the leftmost leaf, or the rightmost slot of
the rightmost leaf, depending on the value of the seq expression. The scan then moves in the direction
specified in the BY clause.
indexname is an expression that evaluates to the name of a secondary index associated with the file.
RESET resets the internal B-tree scan pointer. If the scanning order is ascending, the pointer is set to the
leftmost slot of the leftmost leaf; if the order is descending, the pointer is set to the rightmost slot of the
rightmost leaf. If you do not specify seq, the scan is done in ascending order. If you specify record in the
FROM clause, RESET is ignored.
seq is an expression that evaluates to A or D; it specifies the direction of the scan. ″A″, the default,
specifies ascending order. ″D″ specifies descending order.
If the BSCAN statement finds a valid record ID, or a record ID and its associated data, the THEN
statements are executed; the ELSE statements are ignored. If the scan does not find a valid record ID, or
if some other error occurs, any THEN statements are ignored, and the ELSE statements are executed.
Any file updates executed in a transaction (that is, between a BEGIN TRANSACTION statement and a
COMMIT statement) are not accessible to the BSCAN statement until after the COMMIT statement has
been executed.
The STATUS function returns the following values after the BSCAN statement is executed:
0 The scan proceeded beyond the leftmost or rightmost leaf node. ID.variable and rec.variable are set
to empty strings.
1 The scan returned an existing record ID, or a record ID that matches the record ID specified by
record.
2 The scan returned a record ID that does not match record. ID.variable is either the next or the
previous record ID in the B-tree, depending on the direction of the scan.
If NLS is enabled, the BSCAN statement retrieves record IDs in the order determined by the active
collation locale; otherwise, BSCAN uses the default order, which is simple byte ordering that uses the
standard binary value for characters; the Collate convention as specified in the NLS.LC.COLLATE file for
the current locale is ignored. For more information about collation, see WebSphere DataStage National
Language Support Guide.
Example
The following example shows how you might indicate that the ELSE statements were executed because
the contents of the leaf nodes were exhausted:
BSCAN ID,REC FROM FILE,MATCH USING "PRODUCT" BY "A"
THEN
PRINT ID,REC
END ELSE
ERR = STATUS()
BEGIN CASE
CASE ERR = 0
PRINT "Exhausted leaf node contents."
CASE ERR = 3
PRINT "No active indices, or file is not
type 25."
CASE ERR = 4
PRINT "Index name does not exist."
CASE ERR = 5
PRINT "Invalid BY clause value."
CASE ERR = 6
PRINT "Index must be built."
CASE ERR = 10
PRINT "Internal error detected."
END CASE
GOTO EXIT.PROGRAM:
END
BYTE function
Syntax
BYTE (expression)
Description
In NLS mode, use the BYTE function to generate a byte from the numeric value of expression. BYTE
returns a string containing a single byte.
If expression evaluates to a value in the range 0 to 255, a single-byte character is returned. If expression
evaluates to a value in the range 0x80 to 0xF7, a byte that is part of a multibyte character is returned.
If NLS is not enabled, BYTE works like the CHAR function. For more information, see WebSphere
DataStage National Language Support Guide.
When NLS is enabled, the BYTE and CHAR functions return the following:
BYTE(32)
Returns a string containing a single space.
CHAR(32)
Returns a string containing a single space.
BYTE(230)
Returns a string containing the single byte 0xe6.
CHAR(230)
Returns a string containing the multibyte characters æ (small ligature Æ).
BYTELEN function
Syntax
BYTELEN (expression)
Description
In NLS mode, use the BYTELEN function to generate the number of bytes contained in the ASCII string
value in expression.
The bytes in expression are counted, and the count is returned. If expression evaluates to the null value,
null is returned.
If NLS is not enabled, BYTELEN works like the LEN function. For more information, see WebSphere
DataStage National Language Support Guide.
BYTETYPE function
Syntax
BYTETYPE (value)
Description
In NLS mode, use the BYTETYPE function to determine the function of a byte in value.
If value is from 0 to 255, the BYTETYPE function returns a number that corresponds to the following:
-1 value is out of bounds
0 Trailing byte of a 2-, 3-, or > 3-byte character
1 Single-byte character
2 Leading byte of a 2-byte character
3 Leading byte of a 3-byte character
4 Reserved for the leading byte of a 4-byte character
5 System delimiter
BYTETYPE behaves the same whether NLS is enabled or not. For more information, see WebSphere
DataStage National Language Support Guide.
Description
In NLS mode, use the BYTEVAL function to examine the bytes contained in the internal string value of
expression. The BYTEVAL function returns a number from 0 through 255 as the byte value of n in
expression. If you omit n, 1 is assumed.
If an error occurs, the BYTEVAL function returns a number that corresponds to the following conditions:
-1 expression is the empty string or expression has fewer than n bytes or n is less than 1.
null value
expression is the null value.
BYTEVAL behaves the same whether NLS is enabled or not. For more information, see WebSphere
DataStage National Language Support Guide.
CALL statement
Syntax
CALL name [ ( [MAT] argument [ , [MAT] argument ...] ) ]
variable = ’name’
CALL @variable [ ( [MAT] argument [ , [MAT] argument ...] ) ]
Description
Use the CALL statement to transfer program control from the calling program to an external subroutine
or program that has been compiled and cataloged.
Locally cataloged subroutines can be called directly. Specify name using the exact name under which it
was cataloged. For more details, see The CATALOG Command on page 1-15.
External subroutines can be called directly or indirectly. To call a subroutine indirectly, the name under
which the subroutine is cataloged must be assigned to a variable or to an element of an array. This
variable name or array element specifier, prefixed with an at sign (@), is used as the operand of the
CALL statement.
The first time a CALL is executed, the system searches for the subroutine in a cataloged library and
changes a variable that contains the subroutine name to contain its location information instead. This
procedure eliminates the need to search the catalog again if the same subroutine is called later in the
program. For indirect calls, the variable specified in the CALL as the @variable is used; for direct calls, an
internal variable is used. With the indirect method, it is best to assign the subroutine name to the variable
only once in the program, not every time the indirect CALL statement is used.
arguments are variables, arrays, array variables, expressions, or constants that represent actual values. You
can pass one or more arguments from the calling program to a subroutine. The number of arguments
passed in a CALL statement must equal the number of arguments specified in the SUBROUTINE
statement that identifies the subroutine. If multiple arguments are passed, they must be separated by
commas. If an argument requires more than one physical line, use a comma at the end of the line to
indicate that the list continues.
If argument is an array, it must be preceded by the MAT keyword, and the array should be named and
dimensioned in both the calling program and the subroutine before using this statement. If the array is
The actual values of arguments are not passed to the subroutine. Instead, a pointer to the location of each
argument is passed. Passing a pointer instead of the values is more efficient when many values need to
be passed to the subroutine. This method of passing arguments is called passing by reference; passing
actual values is called passing by value.
All scalar and matrix variables are passed to subroutines by reference. If you want to pass variables by
value, enclose them in parentheses. When data is passed by value, the contents of the variable in the
main program do not change as a result of manipulations to the data in the subroutine. When data is
passed by reference, the memory location of the variable is changed by manipulations in both the main
program and the subroutines. Constants are passed to subroutines by value.
When an array is passed to an external subroutine as an argument in a CALL statement, any dimensions
assigned to the array in the subroutine are ignored. The dimensions of the original array as it exists in the
calling program are maintained. Therefore, it is a common and acceptable practice to dimension the array
in the subroutine with subscripts or indices of one. For example, you could dimension the arrays in the
subroutine as follows:
DIM A (1), B (1, 1), C (1, 1)
When the corresponding array arguments are passed from the calling program to the subroutine at run
time, arrays A, B, and C inherit the dimensions of the arrays in the calling program. The indices in the
DIMENSION statement are ignored.
A better way to declare array arguments in a subroutine is to use the MAT keyword of the SUBROUTINE
statement in the first line of the subroutine. The following example tells the subroutine to expect the three
arrays A, B, and C:
SUBROUTINE X(MAT A, MAT B, MAT C)
When a RETURN statement is encountered in the subroutine, or when execution of the subroutine ends
without encountering a RETURN statement, control returns to the statement following the CALL
statement in the calling program. For more details, see the RETURN statement.
Examples
The following example calls the local subroutine SUB. It has no arguments.
CALL SUB
The following example calls the local subroutine QTY.ROUTINE with three arguments:
CALL QTY.ROUTINE(X,Y,Z)
The following example calls the subroutine cataloged as *PROGRAM.1 with six arguments. The argument
list can be expressed on more than one line.
AAA="*PROGRAM.1"
CALL @AAA(QTY,SLS,ORDER,ANS,FILE.O,SEQ)
The following example calls the subroutine *MA with three arguments. Its index and three arguments are
passed.
STATE.TAX(1,2)=’*MA’
CALL @STATE.TAX(1,2)(EMP.NO,GROSS,NET)
The following example calls the subroutine cataloged as *SUB and two matrices are passed to two
subroutine matrices. A third, scalar, argument is also passed.
The following example shows the SUBROUTINE statement in the subroutine SUB that is called by the
preceding example. The arrays Q and P need not be dimensioned in the subroutine.
SUBROUTINE SUB( MAT Q,MAT P,C )
CASE statement
Syntax
BEGIN CASE
CASE expression statements [ CASE expression statements .
.
. ]
END CASE
Description
Use the CASE statement to alter the sequence of instruction execution based on the value of one or more
expressions. If expression in the first CASE statement is true, the following statements up to the next
CASE statement are executed. Execution continues with the statement following the END CASE
statement.
If the expression in a CASE statement is false, execution continues by testing the expression in the next
CASE statement. If it is true, the statements following the CASE statement up to the next CASE or END
CASE statement are executed. Execution continues with the statement following the END CASE
statement.
If more than one CASE statement contains a true expression, only the statements following the first such
CASE statement are executed. If no CASE statements are true, none of the statements between the BEGIN
CASE and END CASE statements are executed.
If an expression evaluates to the null value, the CASE statement is considered false.
Use the ISNULL function with the CASE statement when you want to test whether the value of a
variable is the null value. This is the only way to test for the null value since null cannot be equal to any
value, including itself. The syntax is:
CASE ISNULL (expression)
Use an expression of the constant ″1″ to specify a default CASE to be executed if none of the other CASE
expressions evaluate to true.
Examples
In the following example NUMBER is equal to 3. CASE 1 is always true, therefore control is transferred
to subroutine 30. Once the subroutine RETURN is executed, control proceeds to the statement following
the END CASE statement.
NUMBER=3
BEGIN CASE
CASE NUMBER=1
GOTO 10
CASE 1
GOSUB 30
CASE NUMBER<3
GOSUB 20
END CASE
In the following example, control proceeds to the statement following the END CASE because ’NAME’
does not meet any of the conditions:
NAME="MICHAEL"
BEGIN CASE
CASE NAME[1,2]=’DA’
PRINT NAME
GOTO 10
CASE NAME[1,2]=’RI’
PRINT NAME
GOSUB 20
CASE NAME[1,2]=’BA’
PRINT NAME
GOSUB 30
END CASE
PRINT ’NO MATCH’
STOP
CATS function
Syntax
CATS (array1, array2)
CALL -CATS (return.array, array1, array2)
CALL !CATS (return.array, array1, array2)
Description
Use the CATS function to create a dynamic array of the element-by-element concatenation of two
dynamic arrays.
Each element of array1 is concatenated with the corresponding element of array2. The result is returned in
the corresponding element of a new dynamic array. If an element of one dynamic array has no
corresponding element in the other dynamic array, the existing element is returned. If an element of one
dynamic array is the null value, null is returned for the concatenation of the corresponding elements.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A="A":@VM:"B":@SM:"C"
B="D":@SM:"E":@VM:"F"
PRINT CATS(A,B)
CHAIN statement
Syntax
CHAIN command
Description
Use the CHAIN statement to terminate execution of a BASIC program and to execute the value of
command. command is an expression that evaluates to any valid WebSphere DataStage command. If
command evaluates to the null value, the CHAIN statement fails and the program terminates with a
run-time error message.
Local variables belonging to the current program are lost when you chain from one program to another.
Named and unnamed common variables are retained.
CHAIN differs from the EXECUTE statement or PERFORM statements in that CHAIN does not return
control to the calling program. If a program chains to a procedure, any calling procedures that are nested
are removed.
Unnamed common variables are lost when a chained program is invoked in a PICK, IN2, or REALITY
flavor account. If you want to save the values of variables in unnamed common, use the
KEEP.COMMON keyword to the RUN command at execution.
Example
The following program clears the screen, initializes the common area, and then runs the main application:
PRINT @(-1)
PRINT "INITIALIZING COMMON, PLEASE WAIT"
GOSUB INIT.COMMON
CHAIN "RUN BP APP.MAIN KEEP.COMMON"
CHANGE function
Syntax
CHANGE (expression, substring, replacement [ ,occurrence [ ,begin] ]
)
Description
Use the CHANGE function to replace a substring in expression with another substring. If you do not
specify occurrence, each occurrence of the substring is replaced.
occurrence specifies the number of occurrences of substring to replace. To change all occurrences, specify
occurrence as a number less than 1.
begin specifies the first occurrence to replace. If begin is omitted or less than 1, it defaults to 1.
If substring is an empty string, the value of expression is returned. If replacement is an empty string, all
occurrences of substring are removed.
The CHANGE function behaves like the EREPLACE function except when substring evaluates to an
empty string.
Example
A = "AAABBBCCCDDDBBB"
PRINT CHANGE (A,"BBB","ZZZ")
PRINT CHANGE (A,"","ZZZ")
PRINT CHANGE (A,"BBB","")
CHAR function
Syntax
CHAR (expression)
Description
Use the CHAR function to generate an ASCII character from the numeric value of expression.
If expression evaluates to the null value, null is returned. If expression evaluates to 128, CHAR(128) is
returned, not the null value. CHAR(128) is the equivalent of the system variable @NULL.STR.
If NLS mode is enabled, and if expression evaluates to a number from 129 through 247, the CHAR
function generates Unicode characters from x0081 through x00F7. These values correspond to the
equivalent ISO 8859-1 (Latin 1) multibyte characters. The evaluation of numbers from 0 through 127, 128,
and 248 through 255 remains the same whether NLS is enabled or not.
The UNICHAR function is the recommended method for generating Unicode characters. For more
information, see WebSphere DataStage National Language Support Guide.
Note: In order to run programs using the CHAR function in NLS mode, you must first recompile them
in NLS mode.
Example
X = CHAR(38)
Y = CHAR(32)
PRINT X:Y:X
Description
Use the CHARS function to generate a dynamic array of ASCII characters from the decimal numeric
value of each element of dynamic.array.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
If any element in the dynamic array is the null value, null is returned for that element. If any element in
the dynamic array evaluates to 128, CHAR(128) is returned, not the null value. CHAR(128) is the
equivalent of the system variable @NULL.STR.
If NLS mode is enabled, and if any element in the dynamic array evaluates to a number from 129
through 247, the CHARS function generates Unicode characters from x0081 through x00F7. These values
correspond to the equivalent ISO 8859-1 (Latin 1) multibyte characters. The evaluation of numbers from 0
through 127, 128, and 248 through 255 remains the same whether NLS is enabled or not.
The UNICHARS function is the recommended method for generating a dynamic array of Unicode
characters. For more information, see WebSphere DataStage National Language Support Guide.
Example
X = CHARS(38:@VM:32:@VM:38)
PRINT X
The dynamic array X comprises three elements: CHAR(38) (an ampersand ( & )), CHAR(32) (a space),
and another CHAR(38). The program prints a dynamic array of these elements separated by value marks:
&V V&
CHECKSUM function
Syntax
CHECKSUM (string)
Description
Use the CHECKSUM function to return a cyclical redundancy code (a checksum value).
Example
A = "THIS IS A RECORD TO BE SENT VIA SOME PROTOCOL"
REC = A:@FM:CHECKSUM(A)
PRINT REC
Description
Use the CLEAR statement at the beginning of a program to set all assigned and unassigned values of
variables outside of the common area of the program to 0. This procedure avoids run-time errors for
unassigned variables. If you use the CLEAR statement later in the program, any values assigned to
non-common variables (including arrays) are lost.
Use the COMMON option to reset the values of all the variables in the unnamed common area to 0.
Variables outside the common area or in the named common area are unaffected.
Example
A=100
PRINT "The value of A before the CLEAR statement:"
PRINT A
CLEAR
PRINT "The value of A after the CLEAR statement:"
PRINT A
PRINT
*
COMMON B,C,D
D="HI"
PRINT "The values of B, C, and D"
PRINT B,C,D
CLEAR COMMON
PRINT B,C,D
CLEARDATA statement
Syntax
CLEARDATA
Description
Use the CLEARDATA statement to flush all data that has been loaded in the input stack by the DATA
statement. No expressions or spaces are allowed with this statement. Use the CLEARDATA statement
when an error is detected, to prevent data placed in the input stack from being used incorrectly.
Example
The following program is invoked from a paragraph. A list of filenames and record IDs is passed to it
from the paragraph with DATA statements. If a file cannot be opened, the CLEARDATA statement clears
the data stack since the DATA statements would no longer be valid to the program.
TEN:
INPUT FILENAME
IF FILENAME="END" THEN STOP
OPEN FILENAME TO FILE ELSE
CLEARFILE statement
Syntax
CLEARFILE [file.variable] [ON ERROR statements] [LOCKED
statements]
Description
Use the CLEARFILE statement to delete all records in an open dictionary or data file. You cannot use this
statement to delete the file itself. Each file to be cleared must be specified in a separate CLEARFILE
statement.
file.variable specifies an open file. If file.variable is not specified, the default file is assumed (for more
information on default files, see the OPEN statement).
The CLEARFILE statement fails and the program terminates with a run-time error message if:
v The file is neither accessible nor open.
v file.variable evaluates to the null value.
v A distributed file contains a part file that cannot be accessed, but the CLEARFILE statement clears
those part files still available.
v A transaction is active. That is, you cannot execute this statement between a BEGIN TRANSACTION
(or TRANSACTION START) statement and the COMMIT (or TRANSACTION START) or ROLLBACK
statement that ends the transaction.
The ON ERROR clause is optional in the CLEARFILE statement. The ON ERROR clause lets you specify
an alternative for program termination when a fatal error is encountered during processing of the
CLEARFILE statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number. If a
CLEARFILE statement is used when any portion of a file is locked, the program waits until the file is
released. The ON ERROR clause is not supported if the CLEARFILE statement is within a transaction.
The LOCKED clause handles a condition caused by a conflicting lock (set by another user) that prevents
the CLEARFILE statement from processing. The LOCKED clause is executed if one of the following
conflicting Locks exists:
v Exclusive file lock
v Intent file lock
v Shared file lock
v Update record lock
v Shared record lock
If the CLEARFILE statement does not include a LOCKED clause, and a conflicting lock exists, the
program pauses until the lock is released.
If a LOCKED clause is used, the value returned by the STATUS function is the terminal number of the
user who owns the conflicting lock.
Example
OPEN "","TEST.FILE" ELSE PRINT "NOT OPEN"
EXECUTE "LIST TEST.FILE"
CLEARFILE
CHAIN "LIST TEST.FILE"
CLEARPROMPTS statement
Syntax
CLEARPROMPTS
CALL !CLEAR.PROMPTS
Description
Use the CLEARPROMPTS statement to clear the value of the in-line prompt. Once a value is entered for
an in-line prompt, the prompt continues to have that value until a CLEARPROMPTS statement is
executed, unless the in-line prompt control option A is specified. CLEARPROMPTS clears all values that
have been entered for in-line prompts.
CLEARSELECT statement
Syntax
CLEARSELECT [ALL | list.number]
Use the CLEARSELECT statement to clear an active select list. This statement is normally used when one
or more select lists have been generated but are no longer needed. Clearing select lists prevents
remaining select list entries from being used erroneously.
Use the keyword ALL to clear all active select lists. Use list.number to specify a numbered select list to
clear. list.number must be a numeric value from 0 through 10. If neither ALL nor list.number is specified,
select list 0 is cleared.
If list.number evaluates to the null value, the CLEARSELECT statement fails and the program terminates
with a run-time error message.
PICK, REALITY, and IN2 flavor accounts store select lists in list variables instead of numbered select lists.
In those accounts, and in programs that use the VAR.SELECT option of the $OPTIONS statement, the
syntax of CLEARSELECT is:
CLEARSELECT [ALL | list.variable]
Example
The following program illustrates the use of CLEARSELECT to clear a partially used select list. The
report is designed to display the first 40-odd hours of lessons. A CLEARSELECT is used so that all the
selected records are not printed. Once the select list is cleared, the READNEXT statement ELSE clause is
executed.
OPEN ’SUN.SPORT’ TO FILE ELSE STOP "CAN’T OPEN FILE"
HOURS=0
*
EXECUTE ’SSELECT SUN.SPORT BY START BY INSTRUCTOR’
*
START:
READNEXT KEY ELSE
PRINT ’FIRST WEEK’, HOURS
STOP
END
READ MEMBER FROM FILE,KEY ELSE GOTO START:
HOURS=HOURS+MEMBER<4>
PRINT MEMBER<1>,MEMBER<4>
IF HOURS>40 THEN
******
CLEARSELECT
******
GOTO START:
END
GOTO START:
END
Description
Use the CLOSE statement after opening and processing a file. Any file locks or record locks are released.
file.variable specifies an open file. If file.variable is not specified, the default file is assumed. If the file is
neither accessible nor open, or if file.variable evaluates to the null value, the CLOSE statement fails and
the program terminates with a run-time error message.
The ON ERROR clause is optional in the CLOSE statement. The ON ERROR clause lets you specify an
alternative for program termination when a fatal error is encountered during processing of the CLOSE
statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
Example
CLEAR
OPEN ’’,’EX.BASIC’ TO DATA ELSE STOP
READ A FROM DATA, ’XYZ’ ELSE STOP
A<3>=’*’
WRITE A ON DATA, ’XYZ’
CLOSE DATA
CLOSESEQ statement
Syntax
CLOSESEQ file.variable [ON ERROR statements]
Description
Use the CLOSESEQ statement after opening and processing a file opened for sequential processing.
CLOSESEQ makes the file available to other users.
file.variable specifies a file previously opened with an OPENSEQ statement. If the file is neither accessible
nor open, the program terminates with a run-time error message. If file.variable is the null value, the
The ON ERROR clause is optional in the CLOSESEQ statement. The ON ERROR clause lets you specify
an alternative for program termination when a fatal error is encountered during processing of the
CLOSESEQ statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
Example
In this example, the CLOSESEQ statement closes FILE.E, making it available to other users:
OPENSEQ ’FILE.E’, ’RECORD1’ TO FILE ELSE ABORT
READSEQ A FROM FILE THEN PRINT A ELSE STOP
CLOSESEQ FILE
END
COL1 function
Syntax
COL1 ( )
Description
Use the COL1 function after the execution of a FIELD function to return the numeric value for the
character position that immediately precedes the selected substring (see the FIELD function). Although
the COL1 function takes no arguments, parentheses are required to identify it as a function.
The value obtained from COL1 is local to the program or subroutine executing the FIELD function.
Before entering a subroutine, the current value of COL1 in the main program is saved. The value of
COL1 in the subroutine is initialized as 0. When control is returned to the calling program, the saved
value of COL1 is restored.
If no FIELD function precedes the COL1 function, a value of 0 is returned. If the delimiter expression of
the FIELD function is an empty string or the null value, or if the string is not found, the COL1 function
returns a 0 value.
The FIELD function in the following example returns the substring CCC. COL1( ) returns 8, the position
of the delimiter ( $ ) that precedes CCC.
SUBSTRING=FIELD("AAA$BBB$CCC",’$’,3)
POS=COL1()
In the following example, the FIELD function returns a substring of 2 fields with the delimiter ( . ) that
separates them: 4.5. COL1( ) returns 6, the position of the delimiter that precedes 4.
SUBSTRING=FIELD("1.2.3.4.5",’.’,4,2)
POS=COL1()
COL2 function
Syntax
COL2 ( )
Description
Use the COL2 function after the execution of a FIELD function to return the numeric value for the
character position that immediately follows the selected substring (see the FIELD function). Although the
COL2 function takes no arguments, parentheses are required to identify it as a function.
The value obtained from COL2 is local to the program or subroutine executing the FIELD function.
Before entering a subroutine, the current value of COL2 in the main program is saved. The value of
COL2 in the subroutine is initialized as 0. When control is returned to the calling program, the saved
value of COL2 is restored.
If no FIELD function precedes the COL2 function, a value of 0 is returned. If the delimiter expression of
the FIELD function is an empty string or the null value, or if the string is not found, the COL2 function
returns a 0 value.
Examples
The FIELD function in the following example returns the substring 111. COL2( ) returns 4, the position of
the delimiter ( # ) that follows 111.
SUBSTRING=FIELD("111#222#3","#",1)
P=COL2()
In the following example, the FIELD function returns a substring of two fields with the delimiter ( & )
that separates them: 7&8. COL2( ) returns 5, the position of the delimiter that follows 8.
SUBSTRING=FIELD("&7&8&B&","&",2,2)
S=COL2()
In the next example, FIELD( ) returns the whole string, because the delimiter ( . ) is not found. COL2( )
returns 6, the position after the last character of the string.
SUBSTRING=FIELD("9*8*7",".",1)
Y=COL2()
In the next example, FIELD( ) returns an empty string, because there is no tenth occurrence of the
substring in the string. COL2( ) returns 0 because the substring was not found.
SUBSTRING=FIELD("9*8*7","*",10)
O=COL2()
Description
Use the COMMIT statement to commit all file I/O changes made during a transaction. The WORK
keyword is provided for compatibility with SQL syntax conventions; it is ignored by the compiler.
A transaction includes all statements between a BEGIN TRANSACTION statement and the COMMIT or
ROLLBACK statement that ends the transaction. Either a COMMIT or a ROLLBACK statement ends the
current transaction.
When a subordinate transaction commits, it makes the results of its database operations accessible to its
parent transaction. The subordinate transaction commits to the database only if all of its predecessors up
to the top-level transaction are committed.
If a top-level transaction succeeds, all changes to files made during the active transaction are committed
to disk.
If a subordinate transaction fails, all its changes are rolled back and do not affect the parent transaction. If
the top-level transaction fails, none of the changes made during the active transaction are committed, and
the database remains unaffected by the failed transaction. This ensures that the database is maintained in
a consistent state.
If the COMMIT statement succeeds, the THEN statements are executed; any ELSE statements are ignored.
If COMMIT fails, any ELSE statements are executed. After the THEN or the ELSE statements are
executed, control is transferred to the statement following the next END TRANSACTION statement.
All Locks obtained during a transaction remain in effect for the duration of the active transaction; they
are not released by a RELEASE statement, WRITE statements, , or MATWRITE statements statement that
is part of the transaction. The parent transaction adopts the acquired or promoted locks. If a subordinate
transaction rolls back, any locks that have been acquired or promoted within that transaction are
demoted or released.
The COMMIT statement that ends the top-level transaction releases locks set during that transaction.
Locks obtained outside the transaction are not affected by the COMMIT statement.
If no transaction is active, the COMMIT statement generates a run-time warning, and the ELSE
statements are executed.
Example
This example begins a transaction that applies locks to rec1 and rec2. If no errors occur, the COMMIT
statement ensures that the changes to rec1 and rec2 are written to the file. The locks on rec1 and rec2 are
released, and control is transferred to the statement following the END TRANSACTION statement.
BEGIN TRANSACTION
READU data1 FROM file1,rec1 ELSE ROLLBACK
READU data2 FROM file2,rec2, ELSE ROLLBACK
.
.
.
The update record lock on rec1 is not released on completion of the first WRITE statements statement but
on completion of the COMMIT statement.
COMMON statement
Syntax
COM[MON] [/name/] variable [ ,variable ...]
Description
Use the COMMON statement to provide a storage area for variables. Variables in the common area are
accessible to main programs and external subroutines. Corresponding variables can have different names
in the main program and in external subroutines, but they must be defined in the same order. The
COMMON statement must precede any reference to the variables it names.
A common area can be either named or unnamed. An unnamed common area is lost when the program
completes its execution and control returns to the WebSphere DataStage command level. A named
common area remains available for as long as the user remains in the WebSphere DataStage environment.
The common area name can be of any length, but only the first 31 characters are significant.
Arrays can be dimensioned and named with a COMMON statement. They can be redimensioned later
with a DIMENSION statement, but the COMMON statement must appear before the DIMENSION
statement. When an array is dimensioned in a subroutine, it takes on the dimensions of the array in the
main program regardless of the dimensions stated in the COMMON statement. For a description of
dimensioning array variables in a subroutine, see the CALL statement.
When programs share a common area, use the $INCLUDE statement to define the common area in each
program.
Example
Program:
COMMON NAME, ADDRESS (15, 6), PHONE
Subroutine:
COMMON A, B (15, 6), C
In this example the variable pairs NAME and A, ADDRESS and B, PHONE and C are stored in the same
memory location.
COMPARE function
Syntax
COMPARE (string1, string2 [ , justification ])
Description
Use the COMPARE function to compare two strings and return a numeric value indicating the result.
If NLS is enabled, the COMPARE function uses the sorting algorithm and the Collate convention
specified in the NLS.LC.COLLATE file in order to compare the strings. For more information about
collation, see WebSphere DataStage National Language Support Guide.
Examples
In the following example, the strings AB99 and AB100 are compared with the right-justified option and
the result displayed. In this case the result displayed is -1.
PRINT COMPARE(’AB99’,’AB100’,’R’)
An example in NLS mode follows. It compares the strings anilno and anillo, returning the result as 1. It
sets the locale to Spanish and compares the strings again. In this case, the result displayed is -1.
$INCLUDE UNIVERSE.INCLUDE UVNLSLOC.H
x=SETLOCALE( UVLC$ALL, ’OFF’ )
PRINT COMPARE( ’anilno’, ’anillo’, ’L’ )
x=SETLOCALE( UVLC$ALL, ’ES-SPANISH’ )
PRINT COMPARE( ’anilno’, ’anillo’, ’L’ )
CONTINUE statement
The CONTINUE statement is a loop-controlling statement. For syntax details, see the FOR statement and
the LOOP statement.
CONVERT function
Syntax
CONVERT (expression1, expression2, variable)
Description
Use the CONVERT function to return a copy of variable with every occurrence of specified characters in
variable replaced with other specified characters. Every time a character to be converted appears in
variable, it is replaced by the replacement character.
If expression2 contains more characters than expression1, the extra characters are ignored. If expression1
contains more characters than expression2, the characters with no corresponding expression2 characters are
deleted from the result.
If variable is the null value, null is returned. If either expression1 or expression2 is the null value, the
CONVERT function fails and the program terminates with a run-time error message.
Example
A="NOW IS THE TIME"
PRINT A
A=CONVERT(’TI’,’XY’,A)
PRINT A
A=CONVERT(’XY’,’T’,A)
PRINT A
CONVERT statement
Syntax
CONVERT expression1 TO expression2 IN variable
Description
Use the CONVERT statement to replace every occurrence of specific characters in a string with other
characters. Every time the character to be converted appears in the string, it is replaced by the
replacement character.
If expression2 contains more characters than expression1, the extra characters are ignored. If expression1
contains more characters than expression2, the characters with no corresponding expression2 characters are
deleted from the variable.
If variable is the null value, null is returned. If either expression1 or expression2 evaluates to the null value,
the CONVERT statement fails and the program terminates with a run-time error message.
Example
A="NOW IS THE TIME"
PRINT A
CONVERT ’TI’ TO ’XY’ IN A
PRINT A
CONVERT ’XY’ TO ’T’ IN A
PRINT A
COS function
Syntax
COS (expression)
Description
Use the COS function to return the trigonometric cosine of an angle. expression is an angle expressed as a
numeric value in degrees. The COS function is the inverse of the ACOS function.
Values outside the range of 0 to 360 degrees are interpreted as modulo 360. Numbers greater than 1E17
produce a warning message and 0 is returned. If expression evaluates to the null value, null is returned.
Example
PRINT "COS(45) = " : COS(45)
END
COSH function
Syntax
COSH (expression)
Description
Use the COSH function to return the hyperbolic cosine of expression. expression must be a numeric value.
Example
PRINT "COSH(2) = ":COSH(2)
COUNT function
Syntax
COUNT (string, substring)
Description
Use the COUNT function to return the number of times a substring is repeated in a string value.
string is an expression that evaluates to the string value to be searched. substring is an expression that
evaluates to the substring to be counted. substring can be a character string, a constant, or a variable.
By default, each character in string is matched to substring only once. Therefore, when substring is longer
than one character and a match is found, the search continues with the character following the matched
substring. No part of the matched string is recounted toward another match. For example, the following
statement counts two occurrences of substring TT and assigns the value 2 to variable C:
C = COUNT (’TTTT’, ’TT’)
In PICK, IN2, and REALITY flavors, the COUNT function continues the search with the next character
regardless of whether it is part of the matched string. For example, the following statement counts three
occurrences of substring TT:
C = COUNT (’TTTT’, ’TT’)
Use the COUNT.OVLP option of the $OPTIONS statement to get this behavior in IDEAL and
INFORMATION flavor accounts.
Example
A=COUNT(’ABCAGHDALL’,’A’)
PRINT "A= ",A
*
Z=’S#FF##G#JJJJ#’
Q=COUNT(Z,’#’)
PRINT "Q= ",Q
*
Y=COUNT(’11111111’,’11’)
PRINT "Y= ",Y
COUNTS function
Syntax
COUNTS (dynamic.array, substring)
CALL -COUNTS (return.array, dynamic.array, substring)
CALL !COUNTS (return.array, dynamic.array, substring)
Description
Use the COUNTS function to count the number of times a substring is repeated in each element of a
dynamic array. The result is a new dynamic array whose elements are the counts corresponding to the
elements in dynamic.array.
substring is an expression that evaluates to the substring to be counted. substring can be a character string,
a constant, or a variable.
Each character in an element is matched to substring only once. Therefore, when substring is longer than
one character and a match is found, the search continues with the character following the matched
substring. No part of the matched element is recounted toward another match.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
In PICK, IN2, and REALITY flavors, the COUNTS function continues the search with the next character
regardless of whether it is part of the matched string. Use the COUNT.OVLP option of the $OPTIONS
statement to get this behavior in IDEAL and INFORMATION flavor accounts.
Example
ARRAY="A":@VM:"AA":@SM:"AAAAA"
PRINT COUNTS(ARRAY, "A")
PRINT COUNTS(ARRAY, "AA")
CREATE statement
Syntax
CREATE file.variable {THEN statements [ELSE statements] | ELSE
statements}
Description
Use the CREATE statement after an OPENSEQ statement to create a record in a type 1 or type 19 file or
to create a UNIX or DOS file. CREATE creates the record or file if the OPENSEQ statement fails. An
OPENSEQ statement for the specified file.variable must be executed before the CREATE statement to
associate the path name or record ID of the file to be created with the file.variable. If file.variable is the null
value, the CREATE statement fails and the program terminates with a run-time error message.
Use the CREATE statement when OPENSEQ cannot find a record or file to open and the next operation
is to be a NOBUF statement, READSEQ statement, or READBLK statement. You need not use the
CREATE statement if the first file operation is a WRITESEQ statement, since WRITESEQ creates the
record or file if it does not exist.
If the record or file is created, the THEN statements are executed, and the ELSE statements are ignored. If
no THEN statements are specified, program execution continues with the next statement.
If the record or file is not created, the ELSE statements are executed; any THEN statements are ignored.
File Buffering
Normally WebSphere DataStage uses buffering for sequential input and output operations. Use the
NOBUF statement after an OPENSEQ statement to turn off buffering and cause all writes to the file to be
performed immediately. For more information about file buffering, see the NOBUF statement.
In the following example, RECORD4 does not yet exist. When OPENSEQ fails to open RECORD4 to the
file variable FILE, the CREATE statement creates RECORD4 in the type 1 file FILE.E and opens it to the
file variable FILE.
OPENSEQ ’FILE.E’, ’RECORD4’ TO FILE
ELSE CREATE FILE ELSE ABORT
WEOFSEQ FILE
WRITESEQ ’HELLO, UNIVERSE’ TO FILE ELSE STOP
CRC32 function
Syntax
CRC32 (string)
Description
Example
This example uses the CRC function to return a number that is a cyclic redundancy code for the specified
string:
MyString = "This is any arbitrary string value"
CheckValue = CRC32(MyString) ;* returns 36235
CRT32 function
Syntax
CRT [print.list]
Description
Use the CRT statement to print data on the screen, regardless of whether a PRINTER ON statement has
been executed. The syntax for print.list is the same as for PRINT statements.
print.list can contain any BASIC expression. The elements of the list can be numeric or character strings,
variables, constants, or literal strings; the null value, however, cannot be output. The list can consist of a
single expression or a series of expressions separated by commas ( , ) or colons ( : ) for output formatting.
If no print.list is designated, a blank line is output.
Expressions separated by commas are printed at preset tab positions. You can use multiple commas
together to cause multiple tabulation between expressions.
Expressions separated by colons are concatenated. That is, the expression following the colon is printed
immediately after the expression preceding the colon. To print a list without a LINEFEED and RETURN,
end the print.list with a colon ( : ).
If NLS is enabled, the CRT statement uses the terminal map in order to print. For more information about
maps and devices, see WebSphere DataStage National Language Support Guide.
DATA statement
Syntax
DATA expression [ ,expression ...]
Description
Use the DATA statement to place values in an input stack. These values can be used as responses to
INPUT statements executed later in the program or in a subroutine (see the INPUT statement). The
values can also serve as responses to WebSphere DataStage commands that request input.
Expressions used in DATA statements can be numeric or character string data. The null value cannot be
stored in the input stack. If expression evaluates to null, the DATA statement fails and the program
terminates with a run-time error message.
Put a comma at the end of each line of a DATA statement to indicate that more data expressions follow
on the next line.
The order in which expressions are specified in the DATA statement is the order in which the values are
accessed by subsequent INPUT statements: first-in, first-out. When all DATA values have been exhausted,
the INPUT statement prompts the user for a response at the terminal.
The DATA statement must be executed before an INPUT statement that is to use expression for input.
You can list the current data in the stack from your program by accessing the @DATA.PENDING variable
with the statement:
PRINT @DATA.PENDING
Example
In the following example, the INPUT NBR statement uses the first value placed in the input stack by the
DATA statement, 33, as the value of NBR. The INPUT DESCR statement uses the second value, 50, as the
value of DESCR. The INPUT PRICE statement uses the third value, 21, as the value of PRICE.
X=33; Y=50; Z=21
DATA X,Y,Z
X=Y+Z
*
INPUT NBR
INPUT DESCR
INPUT PRICE
INPUT QTY
PRINT NBR,DESCR,PRICE,QTY
The value of NBR is the value of X when the DATA statement is executed, not the current value of X
(namely, Y+Z). The INPUT QTY statement has no corresponding value in the input stack, so it prompts
the user for input.
DATE function
Syntax
DATE ( )
Description
Use the DATE function to return the numeric value of the internal system date. Although the DATE
function takes no arguments, parentheses are required to identify it as a function.
The internal format for the date is based on a reference date of December 31, 1967, which is day 0. All
dates thereafter are positive numbers representing the number of days elapsed since day 0. All dates
before day 0 are negative numbers representing the number of days before day 0. For example:
Table 28. Internal Formats for Dates
Date Internal Representation
December 10, 1967 -21
November 15, 1967 -46
February 15, 1968 46
January 1, 1985 6575
Example
PRINT DATE()
PRINT OCONV(DATE(),"D2/")
DCFLUSH function
Syntax
DCFLUSH (file.variable, opt)
Description
Use the DCFLUSH function to flush the disk cache file buffers.
opt is the flush mode. If opt is set to 1, the disk cache file buffer is flushed directly to disk, otherwise it is
flushed from the disk cache to the operating system file buffer for later flushing to disk by the operating
system.
DCOUNT function
Syntax
DCOUNT (string, delimiter)
Description
Use the DCOUNT function to return the number of delimited fields in a data string.
delimiter is an expression that evaluates to the delimiter separating the fields to be counted. delimiter can
be a character string of 0, 1, or more characters.
DCOUNT differs from COUNT in that it returns the number of values separated by delimiters rather
than the number of occurrences of a character string. Two consecutive delimiters in string are counted as
one field. If delimiter evaluates to an empty string, a count of 1 plus the number of characters in the string
is returned. If string evaluates to an empty string, 0 is returned.
If string evaluates to the null value, null is returned. If delimiter evaluates to the null value, the DCOUNT
function fails and the program terminates with a run-time error message.
In PICK, IN2, and REALITY flavors, the DCOUNT function continues the search with the next character
regardless of whether it is part of the matched delimiter string. Use the COUNT.OVLP option of the
$OPTIONS statement to get this behavior in IDEAL and INFORMATION flavor accounts.
Example
REC="88.9.B.7"
Q=DCOUNT(REC,’.’)
PRINT "Q= ",Q
REC=34:@VM:55:@VM:88:@VM:"FF":@VM:99:@VM:"PP"
R=DCOUNT(REC,@VM)
PRINT "R= ",R
DEBUG statement
Syntax
DEBUG
Description
Use the DEBUG statement to invoke RAID, the interactive BASIC debugger. The DEBUG statement takes
no arguments. When this statement is encountered, program execution stops and the double colon ( :: )
prompt appears, waiting for a RAID command. The following table summarizes the RAID commands:
DEFFUN statement
Syntax
DEFFUN function [ ( [MAT] argument [ , [MAT] argument ...] ) ]
[CALLING call.ID]
Description
Use the DEFFUN statement to define a user-written function. You must declare a user-defined function
before you can use it in a program. The DEFFUN statement provides the compiler with information such
as the function name and the number and type of arguments. You can define a user-written function only
once in a program. A subsequent DEFFUN statement for an already defined user-written function causes
a fatal error.
call.ID is an expression that evaluates to the name by which the function is called if it is not the same as
the function name. It can be a quoted string (the call ID itself) or a variable that evaluates to the call ID.
If you do not use the CALLING clause, the user-defined function is presumed to be defined in the VOC
file and cataloged without any prefix.
Examples
The following example defines a user-written function called MYFUNC with the arguments or formal
parameters A, B, and C:
FUNCTION MYFUNC(A, B, C)
Z = ...
RETURN (Z)
END
The next example declares the function MYFUNC. It uses the function with the statement T = MYFUNC
(X, Y, Z). The actual parameters held in X, Y, and Z are referenced by the formal parameters A, B, and C,
so the value assigned to T can be calculated.
DEFFUN MYFUNC(X, Y, Z)
T = MYFUNC(X, Y, Z)
END
DEL statement
Syntax
DEL dynamic.array < field# [ ,value# [ ,subvalue#] ] >
Description
Use the DEL statement to delete a field, value, or subvalue from a dynamic array. The DEL statement
works similarly to the DELETE function.
dynamic.array is an expression that evaluates to a dynamic array. If dynamic.array evaluates to the null
value, null is returned.
field# is an expression that evaluates to the field in dynamic.array. value# is an expression that evaluates to
the value in the field. subvalue# is an expression that evaluates to the subvalue in the value. These
expressions are called delimiter expressions. The numeric values of the delimiter expressions specify
which field, value, or subvalue to delete. The entire position is deleted, including its delimiter characters.
value# and subvalue# are optional. If they are equal to 0, the entire field is deleted. If subvalue# is equal to
0 and value# and field# are greater than 0, the specified value in the specified field is deleted. If all three
delimiter expressions are greater than 0, only the specified subvalue is deleted.
If any delimiter expression is the null value, the DEL statement fails and the program terminates with a
run-time error message.
If a higher-level delimiter expression has a value of 0 when a lower-level delimiter expression is greater
than 0, the 0 delimiter is treated as if it were equal to 1. The delimiter expressions are, from highest to
lowest: field, value, and subvalue.
Examples
In the following examples a field mark is shown by F, a value mark is shown by V, and a subvalue mark
is shown by S.
The next example deletes the first subvalue in field 4 and sets the value of Q to
FLD1FVAL1VSUBV1SSUBV2FFSUBV4:
Q=R
DEL Q<4,1,1>
The next example deletes the second value in field 2 and sets the value of Q to
FLD1FVAL1FFSUBV3SSUBV4:
Q=R
DEL Q<2,2,0>
The next example deletes field 3 entirely and sets the value of Q to
FLD1FVAL1VSUBV1SSUBV2FSUBV3SSUBV4:
Q=R
DEL Q<3,0,0>
The next example deletes the second subvalue in field 4 and sets the value of Q to
FLD1FVAL1VSUBV1SSUBV2FFSUBV3:
Q=R
DEL Q<4,1,2>
DELETE function
Syntax
DELETE (dynamic.array, field#[ ,value#[ ,subvalue#] ] )
Description
Use the DELETE function to erase the data contents of a specified field, value, or subvalue and its
corresponding delimiter from a dynamic array. The DELETE function returns the contents of the dynamic
array with the specified data removed without changing the actual value of the dynamic array.
dynamic.array is an expression that evaluates to the array in which the field, value, or subvalue to be
deleted can be found. If dynamic.array evaluates to the null value, null is returned.
field# is an expression that evaluates to the field in the dynamic array; value# is an expression that
evaluates to the value in the field; subvalue# is an expression that evaluates to the subvalue in the value.
The numeric values of the delimiter expressions specify which field, value, or subvalue to delete. The
entire position is deleted, including its delimiting characters.
If any delimiter expression is the null value, the DELETE function fails and the program terminates with
a run-time error message.
If a higher-level delimiter expression has a value of 0 when a lower-level delimiter is greater than 0, the 0
delimiter is treated as if it were equal to 1. The delimiter expressions are, from highest to lowest: field,
value, and subvalue.
If the DELETE function references a subordinate element of a higher element whose value is the null
value, the dynamic array is unchanged. Similarly, if all delimiter expressions are 0, the original string is
returned.
Examples
In the following examples a field mark is shown by F, a value mark is shown by V, and a subvalue mark
is shown by S.
The next example deletes the first subvalue in field 4 and sets the value of Q to
FLD1FVAL1VSUBV1SSUBV2FFSUBV4:
Q=DELETE (R,4,1,1)
The next example deletes the second value in field 2 and sets the value of Q to
FLD1FVAL1FFSUBV3SSUBV4:
Q=DELETE (R,2,2)
The next example deletes field 3 entirely and sets the value of Q to
FLD1FVAL1VSUBV1SSUBV2FSUBV3SSUBV4:
Q=DELETE (R,3,0,0)
The next example deletes the second subvalue in field 4 and sets the value of Q to
FLD1FVAL1VSUBV1SSUBV2FFSUBV3:
Q=DELETE (R,4,1,2)
DELETE statements
Syntax
DELETE [ file.variable , ] record.ID [ ON ERROR statements ]
[ LOCKED statements ][ THEN statements ] [ ELSE statements ]
DELETEU [ file.variable , ] record.ID [ ON ERROR statements ]
[ LOCKED statements ]
[ THEN statements ] [ ELSE statements ]
Description
Use the DELETE statements to delete a record from a WebSphere DataStage file. If you specify a file
variable, the file must be open when the DELETE statement is encountered (see the OPEN statement).
If the file does not exist or is not open, the program terminates and a run-time error results. If you do not
specify a file variable, the most recently opened default file is used (see the OPEN statement for more
information on default files). If you specify both a file variable and a record ID, you must use a comma
to separate them.
If the file is an SQL table, the effective user of the program must have SQL DELETE privilege to delete
records in the file. For information about the effective user of a program, see the AUTHORIZATION
statement.
The record is deleted, and any THEN statements are executed. If the deletion fails, the ELSE statements
are executed; any THEN statements are ignored.
If a record is locked, it is not deleted, and an error message is produced. The ELSE statements are not
executed.
If either file.variable or record.ID evaluates to the null value, the DELETE statement fails and the program
terminates with a run-time error message.
Use the DELETEU statement to delete a record without releasing the update record lock set by a previous
READU statement (see the statements).
The file must have been previously opened with an OPEN statement. If a file variable was specified in
the OPEN statement, it can be used in the DELETEU statement. You must place a comma between the
file variable and the record ID expression. If no file variable is specified in the DELETEU statement, the
statement applies to the default file. See the OPEN statement for a description of the default file.
The ON ERROR clause is optional in the DELETE statement. Its syntax is the same as that of the ELSE
clause. The ON ERROR clause lets you specify an alternative for program termination when a fatal error
is encountered during processing of the DELETE statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
If the DELETE statement does not include a LOCKED clause, and a conflicting lock exists, the program
pauses until the lock is released.
If a LOCKED clause is used, the value returned by the STATUS function is the terminal number of the
user who owns the conflicting lock.
A record lock held by a DELETEU statement can be released explicitly with a RELEASE statement or
implicitly with a WRITE, WRITEV, MATWRITE, or DELETE statement. The record lock is released when
you return to the command prompt.
Examples
OPEN "","MLIST" TO MALIST ELSE STOP
PRINT "FILE BEFORE DELETE STATEMENT:"
EXECUTE "COUNT MLIST"
PRINT
DELETE MALIST, "JONES"
PRINT "FILE AFTER DELETE STATMENT:"
EXECUTE "LIST MLIST"
3 records listed.
2 records listed.
In the following example, the data portion of the SUBSIDIARIES files is opened to the file variable SUBS.
If the file cannot be opened an appropriate message is printed. The record MADRID is read and then
deleted from the file. An update record lock had been set and is maintained by the DELETEU statement.
OPEN "","SUBSIDIARIES" TO SUBS
READU REC FROM SUBS, ’MADRID’
ELSE STOP ’Sorry, cannot open Subsidiaries file’
DELETEU SUBS, "MADRID"
DELETELIST statement
Syntax
DELETELIST listname
Use the DELETELIST statement to delete a select list saved in the &SAVEDLISTS& file.
or:
record.ID account.name
record.ID is the name of a select list in the &SAVEDLISTS& file. If you specify account.name, the
&SAVEDLISTS& file of the specified account is used instead of the local &SAVEDLISTS& file.
If listname evaluates to the null value, the DELETELIST statement fails and the program terminates with a
run-time error message.
DELETEU statement
Use the DELETEU statement to maintain an update record lock while performing the DELETE statement.
DIMENSION statement
Syntax
DIM[ENSION] matrix (rows, columns) [ , matrix (rows, columns) ...]
DIM[ENSION] vector (subscript) [ , vector (subscript) ...]
Description
Use the DIMENSION statement to define the dimensions of an array variable before referencing the array
in the program. For a matrix (a two-dimensional array), use the DIMENSION statement to set the
maximum number of rows and columns available for the elements of the array. For a vector (a
one-dimensional array), use the DIMENSION statement to set the maximum value of the subscript (the
maximum elements) in the array.
matrix and vector can be any valid variable name. The maximum dimension can be any valid numeric
expression. When specifying the two dimensions of a matrix, you must use a comma to separate the row
and column expressions. These expressions are called indices.
You can use a single DIMENSION statement to define multiple arrays. If you define more than one array
with a DIMENSION statement, you must use commas to separate the array definitions.
The DIMENSION statement declares only the name and size of the array. It does not assign values to the
elements of the array. Assignment of values to the elements is done with the MAT statement, MATPARSE
statement, MATREAD statements, MATREADU statement, and Assignment Statements.
The DIMENSION statement in an IDEAL or INFORMATION flavor account is executed at run time. The
advantage of the way WebSphere DataStage handles this statement is that the amount of memory
allocated is not determined until the DIM statement is executed. This means that arrays can be
redimensioned at run time.
When redimensioning an array, you can change the maximum number of elements, rows, columns, or
any combination thereof. You can even change the dimensionality of an array (that is, from a
one-dimensional to a two-dimensional array or vice versa).
The DIMENSION statement fails if there is not enough memory available for the array. When this
happens, the INMAT function is set to a value of 1.
An array variable that is passed to a subroutine in its entirety as an argument in a CALL statement
cannot be redimensioned in the subroutine. Each array in a subroutine must be dimensioned once. The
dimensions declared in the subroutine DIMENSION statement are ignored, however, when an array is
passed to the subroutine as an argument (for more information, see the CALL statement).
In PICK, IN2, and REALITY flavor accounts, arrays are created at compile time, not run time. Arrays are
not redimensionable, and they do not have a zero element. To get the same characteristics in an
INFORMATION or IDEAL flavor account, use the STATIC.DIM option of the $OPTIONS statement.
Examples
DIM ARRAY(2,2)
ARRAY(1,1)="KK"
ARRAY(1,2)="GG"
ARRAY(2,1)="MM"
ARRAY(2,2)="NN"
In the next example warning messages are printed for the unassigned elements in the matrix. The
elements are assigned empty strings as their values.
DIM ARRAY(2,3)
*
PRINT
FOR X=1 TO 2
FOR Y=1 TO 3
PRINT "ARRAY(":X:",":Y:")", ARRAY(X,Y)
NEXT Y
NEXT X
DIM S(3,2)
S(1,1)=1
S(1,2)=2
S(2,1)=3
S(2,2)=4
S(3,1)=5
S(3,2)=6
In the next example the common elements are preserved. Those elements that cannot be referenced in the
new matrix (S(3,1), S(3,2) ) are lost.
DIM S(2,2)
*
PRINT
FOR X=1 TO 2
FOR Y=1 TO 2
PRINT "S(":X:",":Y:")", S(X,Y)
NEXT Y
NEXT X
ARRAY(2,1) MM
ARRAY(2,2) NN
ARRAY(2,3) Program ’DYNAMIC.DIMENSION’:
Line 12, Variable previously undefined, empty string
used.
S(1,1) 1
S(1,2) 2
S(2,1) 3
S(2,2) 4
DISPLAY statement
Syntax
DISPLAY [print.list]
Description
Use the DISPLAY statement to print data on the screen, regardless of whether a PRINTER statement ON
statement has been executed. The syntax for print.list is the same as for print statements.
The elements of the list can be numeric or character strings, variables, constants, or literal strings; the null
value, however, cannot be output. The list can consist of a single expression or a series of expressions
separated by commas ( , ) or colons ( : ) for output formatting. If no print.list is designated, a blank line is
output.
Expressions separated by commas are printed at preset tab positions. You can use multiple commas
together to cause multiple tabulation between expressions.
Expressions separated by colons are concatenated. That is, the expression following the colon is printed
immediately after the expression preceding the colon. To print a list without a LINEFEED and RETURN,
end the print list with a colon ( : ).
Example
DISPLAY "This can be used to print something on the
DISPLAY "terminal while"
DISPLAY "the PRINTER ON statement is in effect."
DIV function
Syntax
DIV (dividend, divisor)
Description
Use the DIV function to calculate the value of the quotient after dividend is divided by divisor.
Example
X=100; Y=25
Z = DIV (X,Y)
PRINT Z
DIVS function
Syntax
DIVS (array1, array2)
CALL -DIVS (return.array, array1, array2)
CALL !DIVS (return.array, array1, array2)
Description
Use the DIVS function to create a dynamic array containing the result of the element-by-element division
of two dynamic arrays.
Each element of array1 is divided by the corresponding element of array2 with the result being returned
in the corresponding element of a new dynamic array. If elements of array1 have no corresponding
elements in array2, array2 is padded with ones and the array1 elements are returned. If an element of
array2 has no corresponding element in array1, 0 is returned. If an element of array2 is 0, a run-time error
message is printed and a 0 is returned. If either element of a corresponding pair is the null value, null is
returned.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A=10:@VM:15:@VM:9:@SM:4
B=2:@VM:5:@VM:9:@VM:2
PRINT DIVS(A,B)
DOWNCASE function
Syntax
DOWNCASE (expression)
Description
Use the DOWNCASE function to change all uppercase letters in expression to lowercase. If expression
evaluates to the null value, null is returned.
If NLS is enabled, the DOWNCASE function uses the conventions specified by the Ctype category for the
Lowercase field of the NLS.LC.CTYPE file to change the letters in expression. For more information about
the NLS.LC.CTYPE, see WebSphere DataStage National Language Support Guide.
DQUOTE function
Syntax
DQUOTE (expression)
Description
Use the DQUOTE function to enclose an expression in double quotation marks. If expression evaluates to
the null value, null is returned.
Example
PRINT DQUOTE(12 + 5) : " IS THE ANSWER."
END
DTX function
Syntax
DTX (number [ ,size] )
Description
Use the DTX function to convert a decimal integer to its hexadecimal equivalent.
size indicates the minimum size which the hexadecimal character string should have. This field is
supplemented with zeros if appropriate.
If number evaluates to the null value, null is returned. If size is the null value, the DTX function fails and
the program terminates with a run-time error message.
Example
X = 25
Y = DTX (X)
PRINT Y
Y = DTX (X,4)
PRINT Y
END
Description
Use the EBCDIC function to convert each character of expression from its ASCII representation value to its
EBCDIC representation value. The EBCDIC and ASCII functions perform complementary operations. Data
that is not represented in ASCII code produces undefined results.
If expression evaluates to the null value, the EBCDIC function fails and the program terminates with a
run-time error message.
Example
X = ’ABC 123’
Y = EBCDIC(X)
PRINT "ASCII", "EBCDIC", " X "
PRINT "------", "-----", "---"
FOR I = 1 TO LEN (X)
PRINT SEQ(X[I,1]) , SEQ(Y[I,1]),X[I,1]
NEXT I
ECHO statement
Syntax
ECHO {ON | OFF | expression}
Description
Use the ECHO statement to control the display of input characters on the screen.
If ECHO ON is specified, subsequent input characters are displayed, or echoed, on the screen. If ECHO
OFF is specified, subsequent input characters are assigned to the INPUT statement variables but are not
displayed on the screen.
The ability to turn off character display is useful when the keyboard is to be used for cursor movement
or for entering password information. If expression evaluates to true, ECHO is turned ON. If expression
evaluates to false, ECHO is turned OFF. If expression evaluates to the null value, it is treated as false, and
ECHO is turned OFF.
Example
PROMPT ""
ECHO OFF
PRINT "ENTER YOUR PASSWORD"
INPUT PWORD
ECHO ON
END statement
Syntax
END
Description
Use the END statement to terminate a BASIC program or a section of an IF statement,, or OPEN
statement.
An END statement is the last statement in a BASIC program; it indicates the logical end of the program.
When an END statement that is not associated with an IF, READ, or OPEN statement is encountered,
execution of the program terminates. You can use comments after the END statement.
You can also use the END statement with conditional statements in the body of a program. In this case
END indicates the end of a multi-statement conditional clause.
In INFORMATION and REALITY flavors a warning message is printed if there is no final END
statement. The END.WARN option of the $OPTIONS statement prints the warning message in IDEAL,
IN2, PICK, and PIOPEN flavors under the same conditions.
Example
A="YES"
IF A="YES" THEN
PRINT "THESE TWO LINES WILL PRINT ONLY"
PRINT "WHEN THE VALUE OF ’A’ IS ’YES’."
END
*
PRINT
PRINT "THIS IS THE END OF THE PROGRAM"
END ; * END IS THE LAST STATEMENT EXECUTED
Description
Use the ENTER statement to transfer program control from the calling program to an external subroutine
without returning to the calling program. The subroutine must have been compiled and cataloged.
The ENTER statement is similar to the CALL statement, except that with the ENTER statement, program
flow does not return from the entered program to the calling program (see the CALL statement). The
ENTER statement also does not accept arguments.
In the PIOPEN flavor, the ENTER statement is a synonym for the CALL statement. It takes arguments
and returns control to the calling program.
External subroutines can be entered directly or indirectly. To enter a subroutine indirectly, assign the
name of the cataloged subroutine to a variable or to an element of an array. Use the name of this variable
or array element, prefixed with an at sign (@), as the operand of the ENTER statement.
If subroutine evaluates to the null value, the ENTER statement fails and the program terminates with a
run-time error message.
Example
The following program transfers control to the cataloged program PROGRAM2:
ENTER PROGRAM2
EOF(ARG.) function
Syntax
EOF(ARG.)
Description
Use the EOF(ARG.) function to check if the command line argument pointer is past the last command
line argument. ARG. is part of the syntax of the EOF(ARG.) function and must be specified. EOF(ARG.)
returns 1 (true) if the pointer is past the last command line argument, otherwise it returns 0 (false).
The arg# argument of the GET(ARG.) statement and SEEK(ARG.) statements affect the value of the
EOF(ARG.) function.
EQS function
Syntax
EQS (array1, array2)
CALL -EQS (return.array, array1, array2)
CALL !EQS (return.array, array1, array2)
Description
Use the EQS function to test if elements of one dynamic array are equal to the elements of another
dynamic array.
If you use the subroutine syntax, the resulting dynamic array returns as return.array.
Example
A=1:@VM:45:@SM:3:@VM:"one"
B=0:@VM:45:@VM:1
PRINT EQS(A,B)
EQUATE statement
Syntax
EQU[ATE] symbol TO expression [ ,symbol TO expression ...]
EQU[ATE] symbol LIT[ERALLY] string [ ,symbol LIT string ...]
Description
In an EQUATE statement, symbol represents the value of expression or string. You can use the two
interchangeably in the program. When the program is compiled, each occurrence of symbol is replaced by
the value of expression or string. The value is compiled as object code and does not have to be reassigned
each time the program is executed.
You can define multiple symbols in a single EQUATE statement. symbol cannot be a number.
You can define symbol only once. Any subsequent EQUATE state generates a compiler error because the
compiler interprets the symbol before the statement is parsed.
If you use TO as a connector, the object can be any BASIC expression. If you use LIT or LITERALLY as a
connector, the object must be a literal string.
RAID does not recognize EQUATE symbols. You must use the object value in RAID sessions.
There is no limit on the number of EQUATE statements allowed by the BASIC compiler, except that of
memory.
If symbol is the same as the name of a BASIC function, the function is disabled in the program. If a
statement exists with the same name as a disabled function, the statement is also disabled.
Examples
In the next example COST is made equivalent to the value of the expression PRICE*QUANTITY:
EQUATE COST LIT "PRICE * QUANTITY"
PRICE=3;QUANTITY=7
PRINT "THE TOTAL COST IS $": COST
EREPLACE function
Syntax
EREPLACE (expression, substring, replacement [,occurrence [,begin]
] )
Description
Use the EREPLACE function to replace substring in expression with another substring. If you do not
specify occurrence, each occurrence of substring is replaced.
occurrence specifies the number of occurrences of substring to replace. To replace all occurrences, specify
occurrence as a number less than 1.
begin specifies the first occurrence to replace. If begin is omitted or less than 1, it defaults to 1.
If substring is an empty string, replacement is prefixed to expression. If replacement is an empty string, all
occurrences of substring are removed.
If expression evaluates to the null value, null is returned. If substring, replacement, occurrence, or begin
evaluates to the null value, the EREPLACE function fails and the program terminates with a run-time
error message.
The EREPLACE function behaves like the CHANGE function except when substring evaluates to an
empty string.
Example
A = "AAABBBCCCDDDBBB"
PRINT EREPLACE (A,"BBB","ZZZ")
PRINT EREPLACE (A,"","ZZZ")
PRINT EREPLACE (A,"BBB","")
ERRMSG statement
Syntax
ERRMSG message.ID [ ,message.ID ...]
Description
Use the ERRMSG statement to print a formatted error message from the ERRMSG file.
If message.ID evaluates to the null value, the default error message is printed:
Message ID is NULL: undefined error
A standard Pick ERRMSG file is supplied with WebSphere DataStage. Users can construct a local
ERRMSG file using the following syntax in the records. Each field must start with one of these codes
shown in the following table:
Table 30. ERRMSG File Codes
Code Action
A[(n)] Display next argument left-justified; n specifies field
length.
D Display system date.
E [string] Display record ID of message in brackets; string
displayed after ID.
H [string] Display string.
L [(n)] Output a newline; n specifies number of newlines.
R [(n)] Display next argument right-justified; n specifies field
length.
S [(n)] Output n blank spaces from beginning of line.
T Display system time.
Example
>ED ERRMSG 17 lines long.
----: P0001: HBEGINNING OF ERROR MESSAGE
0002: L
0003: HFILE NAMED "
0004: A
0005: H" NOT FOUND.
0006: L
0007: H END OF MESSAGE
Bottom at line 7
----: QOPEN ’SUN.SPORT’ TO test
THEN PRINT "File Opened" ELSE ERRMSG "1", "SUN.SPORT"
Description
Use the EXCHANGE function to replace one character with another or to delete all occurrences of the
specified character.
string is an expression evaluating to the string whose characters are to be replaced or deleted. If string
evaluates to the null value, null is returned.
If yy is FF, all occurrences of xx are deleted. If xx or yy consist of fewer than two characters, no
conversion is done. If xx or yy is the null value, the EXCHANGE function fails and the program
terminates with a run-time error message.
If NLS is enabled, EXCHANGE uses only the first two bytes of xx and yy in order to evaluate the
characters. Note how the EXCHANGE function evaluates the following characters:
Bytes...
Evaluated as...
00 through FF
00 through FF
00 through FA
Unicode characters 0000 through FA
FB through FE
System delimiters
For more information about character values, see WebSphere DataStage National Language Support Guide.
Example
In the following example, 41 is the hexadecimal value for the character A and 2E is the hexadecimal
value for the period character ( . ):
PRINT EXCHANGE(’ABABC’,’41’,’2E’)
EXECUTE statement
Syntax
EXECUTE commands [CAPTURING variable] [PASSLIST
[dynamic.array] ]
[RTNLIST [variable] ] [ {SETTING | RETURNING}
variable]
EXECUTE commands [ ,IN < expression] [ ,OUT > variable]
[ ,SELECT[ (list) ] < dynamic.array] [ ,SELECT[ (list) ] >
Description
Use the EXECUTE statement to execute WebSphere DataStage commands from within the BASIC
program and then return execution to the statement following the EXECUTE statement.
EXECUTE creates a new environment for the executed command. This new environment is initialized
with the values of the current prompt, current printer state, Break key counter, the values of in-line
prompts, KEYEDIT statements, KEYTRAP statements, and KEYEXIT statements. If any of these values
change in the new environment, the changes are not passed back to the calling environment. In the new
environment, stacked @Variables are either initialized to 0 or set to reflect the new environment.
@variables that are not stacked are shared between the EXECUTE and calling environments.
commands can be sentences, paragraphs, verbs, procedures, menus, or BASIC programs. You can specify
multiple commands in the EXECUTE statement in the same way they are specified in a UniVerse
paragraph. Each command or line must be separated by a field mark (ASCII CHAR 254).
The EXECUTE statement has two main syntaxes. The first syntax requires options to be separated by
spaces. The second and third syntaxes require options to be separated by commas. In the third syntax, the
″//″ preceding the keywords and the periods following them are optional; the compiler ignores these
marks. Except for the slashes and periods, the second and third syntaxes are the same.
In the first syntax the CAPTURING clause assigns the output of the executed commands to variable. The
PASSLIST clause passes the current active select list or expression to the commands for use as select list 0.
The RTNLIST option assigns select list 0, created by the commands, to variable. If you do not specify
variable, the RTNLIST clause is ignored. Using the SETTING or RETURNING clause causes the
@SYSTEM.RETURN.CODE of the last executed command to be placed in variable.
In the second syntax the executed commands use the value of expression in the IN clause as input. When
the IN clause is used, the DATA queue is passed back to the calling program, otherwise data is shared
between environments. The OUT clause assigns the output of the commands to variable. The SELECT
clauses let you supply the select list stored in expression as a select list to the commands, or to assign a
select list created by the commands to variable. If list is not specified, select list 0 is used. The PASSLIST
clause passes the currently active select list to the commands. If you do not specify list, select list 0 in the
current program’s environment is passed as select list 0 in the executed command’s environment. The
STATUS clause puts the @SYSTEM.RETURN.CODE of the last executed command in variable.
The EXECUTE statement fails and the program terminates with a run-time error message if:
v dynamic.array or expression in the IN clause evaluates to the null value.
v The command expression evaluates to the null value.
In transactions you can use only the following WebSphere DataStage commands and SQL statements with
EXECUTE:
CHECK.SUM INSERT SEARCH SSELECT COUNT LIST SELECT (RetrieVe) STAT DELETE (SQL)
LIST.ITEM SELECT (SQL) SUM DISPLAY LIST.LABEL SORT UPDATE ESEARCH RUN SORT.ITEM
In INFORMATION flavor accounts, the EXECUTE statement without any options is the same as the
PERFORM statement. In this case executed commands keep the same environment as the BASIC program
that called them. Use the EXEC.EQ.PERF option of the $OPTIONS statement to cause EXECUTE to
behave like PERFORM in other flavors.
Example
The following example performs a nested SELECT, demonstrating the use of the CAPTURING, RTNLIST,
and PASSLIST keywords:
CMD = "SELECT VOC WITH TYPE = V"
EXECUTE CMD RTNLIST VERBLIST1
CMD = "SELECT VOC WITH NAME LIKE ...LIST..."
EXECUTE CMD PASSLIST VERBLIST1 RTNLIST VERBLIST2
CMD = "LIST VOC NAME"
EXECUTE CMD CAPTURING RERUN PASSLIST VERBLIST2
PRINT RERUN
The program first selects all VOC entries that define verbs, passing the select list to the variable
VERBLIST1. Next, it selects from VERBLIST1 all verbs whose names contain the string LIST and passes
the new select list to VERBLIST2. The list in VERBLIST2 is passed to the LIST command, whose output is
captured in the variable RERUN, which is then printed.
EXIT statement
Syntax
EXIT
Description
Use the EXIT statement to quit execution of a Description loop or a Description loop and branch to the
statement following the NEXT or REPEAT statement of the loop. The EXIT statement quits exactly one
loop. When loops are nested and the EXIT statement is executed in an inner loop, the outer loop remains
in control.
Example
COUNT = 0
LOOP
WHILE COUNT < 100 DO
INNER = 0
LOOP
WHILE INNER < 100 DO
COUNT += 1
INNER += 1
IF INNER = 50 THEN EXIT
REPEAT
PRINT "COUNT = ":COUNT
REPEAT
EXP function
Syntax
EXP (expression)
Use the EXP function to return the value of ″e″ raised to the power designated by expression. The value of
″e″ is approximately 2.71828. expression must evaluate to a numeric value.
If expression is too large or small, a warning message is printed and 0 is returned. If expression evaluates
to the null value, null is returned.
Example
X=5
PRINT EXP(X-1)
EXTRACT function
Syntax
EXTRACT (dynamic.array, field# [,value# [,subvalue#] ] )
variable < field# [ ,value# [ ,subvalue#] ] >
Description
Use the EXTRACT function to access the data contents of a specified field, value, or subvalue from a
dynamic array. You can use either syntax shown to extract data. The first syntax uses the EXTRACT
keyword, the second uses angle brackets.
dynamic.array is an expression that evaluates to the array in which the field, value, or subvalue to be
extracted is to be found. If dynamic.array evaluates to the null value, null is returned.
field# specifies the field in the dynamic array; value# specifies the value in the field; subvalue# specifies the
subvalue in the value. These arguments are called delimiter expressions. The numeric values of the
delimiter expressions determine whether a field, a value, or a subvalue is to be extracted. value# and
subvalue# are optional.
Angle brackets used as an EXTRACT function appear on the right side of an assignment statement. Angle
brackets on the left side of the assignment statement indicate that a REPLACE function is to be
performed (for examples, see the REPLACE function).
The second syntax uses angle brackets to extract data from dynamic arrays. variable specifies the dynamic
array containing the data to be extracted. field#, value#, and subvalue# are delimiter expressions.
Here are the five outcomes that can result from the different uses of delimiter expressions:
Case 1:
If field#, value#, and subvalue# are omitted or evaluate to 0, an empty string is returned.
Case 2:
If value# and subvalue# are omitted or evaluate to 0, the entire field is extracted.
Case 3:
If subvalue# is omitted or specified as 0 and value# and field# evaluate to nonzero, the entire
specified value in the specified field is extracted.
If a higher-level delimiter expression has a value of 0 when a lower-level delimiter is greater than 0, a 1 is
assumed. The delimiter expressions are from highest to lowest: field, value, and subvalue.
If the EXTRACT function references a subordinate element of an element whose value is the null value,
null is returned.
Example
In the following example a field mark is shown by F, a value mark is shown by V, and a subvalue mark
is shown by S:
VAR=1:@FM:4:@VM:9:@SM:3:@SM:5:@FM:1:@VM:0:@SM:7:@SM:3
Z=EXTRACT(VAR,1,0,0)
PRINT Z
*
Z=VAR<1,1,1>
PRINT Z
*
Z=EXTRACT(VAR,2,1,1)
PRINT Z
*
Z=VAR<3,2,3>
PRINT Z
*
Z=EXTRACT(VAR,10,0,0)
PRINT Z
*
Z=EXTRACT(VAR,2,2,0)
PRINT Z
*
FADD function
Syntax
FADD (number1, number2)
CALL !FADD (return.array, number1, number2)
Description
Use the FADD function to perform floating-point addition on two numeric values. If either number
evaluates to the null value, null is returned. If either number1 or number2 evaluates to the null value, null
is returned. return.array equates to number1 plus number2.
This function is provided for compatibility with existing software. You can also use the + operator to
perform floating-point addition.
FDIV function
Syntax
FDIV (number1, number2)
CALL !FDIV (return.array, number1, number2)
Description
Use the FDIV function to perform floating-point division on two numeric values. number1 is divided by
number2. return.array equates to number1 divided by number2. If number2 is 0, a run-time error message is
produced and a 0 is returned for the function. If either number evaluates to the null value, null is
returned.
This function is provided for compatibility with existing software. You can also use the / operator to
perform floating-point division.
Example
PRINT FDIV(.234,.567)
FFIX function
Syntax
FFIX (number)
Description
Use the FFIX function to convert a floating-point number to a numeric string with fixed precision. If
number evaluates to the null value, null is returned.
FFLT function
Syntax
FFLT (number)
Description
Use the FFLT function to round a number to a string with a precision of 13. The number also converts to
scientific notation when required for precision. If number evaluates to the null value, null is returned.
FIELD function
Syntax
FIELD (string, delimiter, occurrence [ ,num.substr] )
Use the FIELD function to return one or more substrings located between specified delimiters in string.
delimiter evaluates to any character, including field mark, value mark, and subvalue marks. It delimits the
start and end of the substring. If delimiter evaluates to more than one character, only the first character is
used. Delimiters are not returned with the substring.
occurrence specifies which occurrence of the delimiter is to be used as a terminator. If occurrence is less
than 1, 1 is assumed.
num.substr specifies the number of delimited substrings to return. If the value of num.substr is an empty
string or less than 1, 1 is assumed. When more than one substring is returned, delimiters are returned
along with the successive substrings.
If either delimiter or occurrence is not in the string, an empty string is returned, unless occurrence specifies
1. If occurrence is 1 and delimiter is not found, the entire string is returned. If delimiter is an empty string,
the entire string is returned.
If string evaluates to the null value, null is returned. If string contains CHAR(128) (that is, @NULL.STR),
it is treated like any other character in a string. If delimiter, occurrence, or num.substr evaluate to the null
value, the FIELD function fails and the program terminates with a run-time error message.
Examples
D=FIELD("###DHHH#KK","#",4)
PRINT "D= ",D
The variable D is set to DHHH because the data between the third and fourth occurrence of the delimiter
# is DHHH.
REC="ACADABA"
E=FIELD(REC,"A",2)
PRINT "E= ",E
Z is set to an empty string since ″?″ does not appear in the string.
Q=FIELD("+1+2+3ABAC","+",2,2)
PRINT "Q= ",Q
Q is set to ″1+2″ since two successive fields were specified to be returned after the second occurrence of
″+″.
Description
Use the FIELDS function to return a dynamic array of substrings located between specified delimiters in
each element of dynamic.array.
delimiter evaluates to any character, including value and subvalue characters. It marks the start and end
of the substring. If delimiter evaluates to more than one character, the first character is used.
occurrence specifies which occurrence of the delimiter is to be used as a terminator. If occurrence is less
than 1, 1 is assumed.
num.substr specifies the number of delimited substrings to return. If the value of num.substr is an empty
string or less than 1, 1 is assumed. In this case delimiters are returned along with the successive
substrings. If delimiter or occurrence does not exist in the string, an empty string is returned, unless
occurrence specifies 1. If occurrence is 1 and the specified delimiter is not found, the entire element is
returned. If occurrence is 1 and delimiter is an empty string, an empty string is returned.
If dynamic.array is the null value, null is returned. If any element in dynamic.array is the null value, null is
returned for that element. If delimiter, occurrence, or num.substr evaluates to the null value, the FIELDS
function fails and the program terminates with a run-time error message.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A="000-P-0":@VM:"-H--O-":@SM:"N-I-T":@VM:"BC":@SM:"-
L-"
PRINT FIELDS(A,"-",2)
FIELDSTORE function
Syntax
FIELDSTORE (string, delimiter, start, n, new.string)
Description
Use the FIELDSTORE function to modify character strings by inserting, deleting, or replacing fields
separated by specified delimiters.
delimiter evaluates to any single ASCII character, including field, value, and subvalue marks.
n specifies the number of fields of new.string to insert in string. n determines how the FIELDSTORE
operation is executed. If n is positive, n fields in string are replaced with the first n fields of new.string. If
n is negative, n fields in string are replaced with all the fields in new.string. If n is 0, all the fields in
new.string are inserted in string before the field specified by start.
If string evaluates to the null value, null is returned. If delimiter, start, n, or new.string is null, the
FIELDSTORE function fails and the program terminates with a run-time error message.
Example
Q=’1#2#3#4#5’
*
TEST1=FIELDSTORE(Q,"#",2,2,"A#B")
PRINT "TEST1= ",TEST1
*
TEST2=FIELDSTORE(Q,"#",2,-2,"A#B")
PRINT "TEST2= ",TEST2
*
TEST3=FIELDSTORE(Q,"#",2,0,"A#B")
PRINT "TEST3= ",TEST3
*
TEST4=FIELDSTORE(Q,"#",1,4,"A#B#C#D")
PRINT "TEST4= ",TEST4
*
TEST5=FIELDSTORE(Q,"#",7,3,"A#B#C#D")
PRINT "TEST5= ",TEST5
FILEINFO function
Syntax
FILEINFO ( file.variable , key)
key is a number that indicates the particular information required. These key numbers are described in
the table ″Keys and Values Supplied to the FILEINFO Function.″
Description
Use the FILEINFO function to return information about the configuration of the specified file, such as its
parameters, its modulus and load, its operating system filename, and its VOC name. The information
returned depends on the file type and the value of the key.
If the first argument is not a file variable, all keys except 0 return an empty string. A warning message is
also displayed. A fatal error results if an invalid key is supplied.
An insert file of equate names is provided to let you use mnemonics rather than key numbers. The insert
file, called FILEINFO.INS.IBAS, is located in the INCLUDE directory in the UV account directory. It is
referenced in PIOPEN flavor accounts through a VOC file pointer called SYSCOM. Use the $INCLUDE
statement to insert this file if you want to use equate names, as shown in the example. The following
table lists the symbolic name, value, and description:
Table 31. Keys and Values Supplied to the FILEINFO Function
Symbolic Name Value Description
FINFO$IS.FILEVAR 0 1 if file.variable is a valid file variable;
0 otherwise.
FINFO$VOCNAME 1 VOC name of the file.
FINFO$PATHNAME 2 Path name of the file.
FINFO$TYPE 3 File type as follows:
1 Static hashed
3 Dynamic hashed
4 Type 1
5 Sequential
If the function executes successfully, the value returned by the STATUS function is 0. If the function fails
to execute, STATUS returns a nonzero value. The following table lists the key, file type, and returned
value for key:
Table 32. FILEINFO Values Returned by File Type
Key Dynamic Directory Distributed Sequential
0 1 = file open 0 = file 1 = file open 0 = file Dynamic array of 1 = file open 0 = file
closed closed codes: 1 = file open 0 closed
= file closed
1 VOC name VOC name VOC name VOC name
2 File’s path name Path name of file Dynamic array of File’s path name
complete path names
in VOC record order
(path name as used
in VOC for
unavailable files)
part file. For example, if the part file is a hashed file, some values, such as minimum modulus, have an empty value in the dynamic array for that part file.
Note: The first time that an I/O operation fails for a part file in a distributed file, the FILEINFO function
returns an error code for that part file. For any subsequent I/O operations on the distributed file with the
same unavailable part file, the FILEINFO function returns -1.
NLS Mode
The FILEINFO function determines the map name of a file by using the value of FINFO$NLSMAP. NLS
uses the insert file called FILEINFO.H. For more information about maps, see WebSphere DataStage
National Language Support Guide.
Examples
In the following example, the file containing the key equate names is inserted with the $INCLUDE
statement. The file FILMS is opened and its file type displayed.
$INCLUDE SYSCOM FILEINFO.INS.IBAS
OPEN ’’,’FILMS’ TO FILMS
ELSE STOP ’CANT OPEN FILE’
PRINT FILEINFO(FILMS,FINFO$TYPE)
In the next example, the file FILMS is opened and its file type displayed by specifying the numeric key:
OPEN ’’,’FILMS’ TO FILMS
ELSE STOP ’CANT OPEN FILE’
PRINT FILEINFO(FILMS,3)
FILELOCK statement
Syntax
FILELOCK [ file.variable ] [ , lock.type ]
[ON ERROR statements] [LOCKED statements]
Description
Use the FILELOCK statement to acquire a lock on an entire file. This prevents other users from updating
the file until the program releases it. A FILELOCK statement that does not specify lock.type is equivalent
to obtaining an update record lock on every record of the file.
file.variable specifies an open file. If file.variable is not specified, the default file is assumed (for more
information on default files, see the OPEN statement). If the file is neither accessible nor open, the
program terminates with a run-time error message. If file.variable evaluates to the null value, the
FILELOCK statement fails and the program terminates with a run-time error message.
The ON ERROR clause is optional in the FILELOCK statement. The ON ERROR clause lets you specify
an alternative for program termination when a fatal error is encountered during processing of the
FILELOCK statement.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number. If a
FILELOCK statement is used when any portion of a file is locked, the program waits until the file is
released.
If the FILELOCK statement does not include a LOCKED clause and a conflicting lock exists, the program
pauses until the lock is released.
If a LOCKED clause is used, the value returned by the STATUS function is the terminal number of the
user who owns the conflicting lock.
Releasing Locks
A shared, intent, or exclusive file lock can be released by a Syntax, RELEASE statement, or STOP
statement.
Locks acquired or promoted within a transaction are not released when previous statements are
processed.
Examples
OPEN ’’,’SUN.MEMBER’ TO DATA ELSE STOP "CAN’T OPEN
FILE"
FILELOCK DATA LOCKED STOP ’FILE IS ALREADY LOCKED’
FILEUNLOCK DATA
FILEUNLOCK statement
Syntax
FILEUNLOCK [file.variable] [ON ERROR statements]
Description
Use the FILEUNLOCK statement to release a file lock set by the FILELOCK statement.
file.variable specifies a file previously locked with a FILELOCK statement. If file.variable is not specified,
the default file with the FILELOCK statement is assumed (for more information on default files, see the
OPEN statement). If file.variable evaluates to the null value, the FILEUNLOCK statement fails and the
program terminates with a run-time error message.
The FILEUNLOCK statement releases only file locks set with the FILELOCK statement. Update record
locks must be released with one of the other unlocking statements (for example, WRITE, WRITEV, and so
on).
The ON ERROR clause is optional in the FILEUNLOCK statement. The ON ERROR clause lets you
specify an alternative for program termination when a fatal error is encountered during processing of the
FILEUNLOCK statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number. The
ON ERROR clause is not supported if the FILEUNLOCK statement is within a transaction.
In the following example, the first FILEUNLOCK statement unlocks the default file. The second
FILEUNLOCK statement unlocks the file variable FILE.
OPEN ’’,’SUN.MEMBER’ ELSE STOP "CAN’T OPEN
SUN.MEMBER"
FILELOCK
.
.
.
FILEUNLOCK
OPEN ’EX.BASIC’ TO FILE ELSE STOP
FILELOCK FILE
.
.
.
FILEUNLOCK FILE
FIND statement
Syntax
FIND element IN dynamic.array [ ,occurrence] SETTING fmc [ ,vmc
[ ,smc] ]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the FIND statement to locate an element in dynamic.array. The field, value, and subvalue positions of
element are put in the variables fmc, vmc, and smc respectively.
element evaluates to a character string. FIND succeeds only if the string matches an element in its entirety.
If element is found in dynamic.array, any THEN statements are executed. If element is not found, or if
dynamic.array evaluates to the null value, fmc, vmc, and smc are unchanged, and the ELSE statements are
executed.
If occurrence is unspecified, it defaults to 1. If occurrence is the null value, the FIND statement fails and the
program terminates with a run-time error message.
Example
A="THIS":@FM:"IS":@FM:"A":@FM:"DYNAMIC":@FM:"ARRAY"
FIND "IS" IN A SETTING FM,VM,SM ELSE ABORT
PRINT "FM=",FM
PRINT "VM=",VM
PRINT "SM=",SM
FINDSTR statement
Syntax
FINDSTR substring IN dynamic.array [ ,occurrence]
SETTING fmc [ ,vmc [ ,smc] ]
{THEN statements [ELSE statements] | ELSE statements}
Use the FINDSTR statement to locate substring in dynamic.array. The field, value, and subvalue positions
of substring are placed in the variables fmc, vmc, and smc respectively.
FINDSTR succeeds if it finds substring as part of any element in dynamic array. If substring is found in
dynamic.array, any THEN statements are executed. If substring is not found, or if dynamic.array evaluates
to the null value, fmc, vmc, and smc are unchanged, and the ELSE statements are executed.
If occurrence is unspecified, it defaults to 1. If occurrence is the null value, FINDSTR fails and the program
terminates with a run-time error message.
Example
A="THIS":@FM:"IS":@FM:"A":@FM:"DYNAMIC":@FM:"ARRAY"
FINDSTR "IS" IN A SETTING FM,VM,SM ELSE ABORT
PRINT "FM=",FM
PRINT "VM=",VM
PRINT "SM=",SM
FIX function
Syntax
FIX (number [ ,precision [ ,mode ] ] )
Description
Use the FIX function to convert a numeric value to a floating-point number with a specified precision.
FIX lets you control the accuracy of computation by eliminating excess or unreliable data from numeric
results. For example, a bank application that computes the interest accrual for customer accounts does not
need to deal with credits expressed in fractions of cents. An engineering application needs to throw away
digits that are beyond the accepted reliability of computations.
precision is an expression that evaluates to the number of digits of precision in the floating-point number.
If you do not specify precision, the precision specified by the PRECISION statement is used. The default
precision is 4.
mode is a flag that specifies how excess digits are handled. If mode is either 0 or not specified, excess
digits are rounded off. If mode is anything other than 0, excess digits are truncated.
Examples
FLUSH statement
Syntax
FLUSH file.variable {THEN statements [ELSE statements] | ELSE
statements}
Description
The FLUSH statement causes all the buffers for a sequential I/O file to be written immediately. Normally,
sequential I/O uses UNIX ″stdio″ buffering for input/output operations, and writes are not performed
immediately.
file.variable specifies a file previously opened for sequential processing. If file.variable evaluates to the null
value, the FLUSH statement fails and the program terminates with a run-time error message.
After the buffer is written to the file, the THEN statements are executed, and the ELSE statements are
ignored. If THEN statements are not present, program execution continues with the next statement.
If the file cannot be written to or does not exist, the ELSE statements are executed; any THEN statements
are ignored.
See the OPENSEQ statement and WRITESEQ statements for more information on sequential file
processing.
Example
OPENSEQ ’FILE.E’, ’RECORD1’ TO FILE THEN
PRINT "’FILE.E’ OPENED FOR SEQUENTIAL PROCESSING"
END ELSE STOP
WEOFSEQ FILE
*
WRITESEQ ’NEW LINE’ ON FILE THEN
FLUSH FILE THEN
PRINT "BUFFER FLUSHED"
END ELSE PRINT "NOT FLUSHED"
ELSE ABORT
*
CLOSESEQ FILE
END
FMT function
Syntax
FMT (expression, format)
expression format
Description
Use the FMT function or a format expression to format data for output. Any BASIC expression can be
formatted for output by following it with a format expression.
format is an expression that evaluates to a string of formatting codes. The syntax of the format expression
is:
[width] [fill] justification [edit] [mask]
The format expression specifies the width of the output field, the placement of background or fill
characters, line justification, editing specifications, and format masking.
If expression evaluates to the null value, null is returned. If format evaluates to null, the FMT function and
the format operation fail.
width is an integer that specifies the size of the output field in which the value is to be justified. If you
specify mask, you need not specify width. If you do not specify mask, width is required.
fill specifies the character to be used to pad entries when filling out the output field. fill is specified as a
single character. The default fill character is a space. If you want to use a numeric character or the letter
L, R, T, or Q as a fill character, you must enclose it in single quotation marks.
Decimal notation:
L Left justification - Break on field length.
R Right justification - Break on field length.
T Text justification - Left justify and break on space.
U Left justification - Break on field length.
Exponential notation:
Q Right justification - Break on field length.
QR Right justification - Break on field length.
QL Left justification
Note: The E, M, C, D and N options define numeric representations for monetary use, using prefixes or
suffixes. In NLS mode, these options override the Numeric and Monetary categories.
mask lets literals be intermixed with numerics in the formatted output field. The mask can include any
combination of literals and the following three special format mask characters:
#n Data is displayed in a field of n fill characters. A blank is the default fill character. It is used if the
format string does not specify a fill character after the width parameter.
%n Data is displayed in a field of n zeros.
*n Data is displayed in a field of n asterisks.
If you want to use numeric characters or any of the special characters as literals, you must escape the
character with a backslash ( \ ).
A #, %, or * character followed by digits causes the background fill character to be repeated n times.
Other characters followed by digits cause those characters to appear in the output data n times.
mask can be enclosed in parentheses ( ) for clarity. If mask contains parentheses, you must include the
whole mask in another set of parentheses. For example:
((###) ###-####)
You must specify either width or mask in the FMT function. You can specify both in the same function.
When you specify width, the string is formatted according to the following rules:
If string is larger than width n, a text mark (CHAR(251)) is inserted every nth character and each field is
padded with the fill character to width.
:
0 The edit code is successful.
1 The string expression is invalid.
2 The edit code is invalid.
REALITY Flavor
In REALITY flavor accounts, you can use conversion codes in format expressions.
Examples
Format Expression
Formatted Value
Z=FMT(″236986″,″R##-##-##″)
Z= 23-69-86
X=″555666898″ X=FMT(X,″20*R2$,″)
X= *****$555,666,898.00
Y=″DAVID″ Y=FMT(Y,″10.L″)
Y= DAVID.....
V=″24500″ V=FMT(V,″10R2$Z″)
V= $24500.00
R=FMT(77777,″R#10″)
R= 77777
B=″0.12345678E1″ B=FMT(B,″9*Q″)
B= *1.2346E0
PRINT 233779 ″R″
233779
PRINT 233779 ″R0″
233779
PRINT 233779 ″R00″
2337790000
PRINT 233779 ″R2″
233779.00
PRINT 233779 ″R20″
2337790000.00
PRINT 233779 ″R24″
233779.00
PRINT 233779 ″R26″
2337.79
PRINT 2337.79 ″R″
2337.79
PRINT 2337.79 ″R0″
2338
PRINT 2337.79 ″R00″
23377900
PRINT 2337.79 ″R2″
2337.79
PRINT 2337.79 ″R20″
23377900.00
PRINT 2337.79 ″R24″
2337.79
FMTDP function
Syntax
FMTDP (expression, format [ , mapname ] )
Description
In NLS mode, use the FMTDP function to format data for output in display positions rather than
character lengths.
expression evaluates to the numeric or string value to be formatted. Any characters that cannot be mapped
in expression are assumed to have a display length of 1.
format is an expression that evaluates to a string of formatting codes. The syntax of the format expression
is:
[width] [fill] justification [edit] [mask]
The format expression specifies the width of the output field, the placement of background or fill
characters, line justification, editing specifications, and format masking. For complete syntax details, see
the FMT function.
If format has a display length greater than 1, and there is only one display position left to fill, FMTDP
enters the extra fill character. The returned string can occupy more display positions than you intended.
mapname is the name of an installed map. If mapname is not installed, the display positions of the
characters in expression are used. If any characters that cannot be mapped exist in expression, the display
length is 1, that is, the unmapped character displays as a single character that cannot be mapped. If
mapname is omitted, the map associated with the channel activated by the PRINTER statement ON
statement is used; otherwise, the map associated with the terminal channel (or print channel 0) is used.
You can also specify mapname as CRT, AUX, LPTR, and OS. These use the maps associated with the
terminal, auxiliary printer, print channel 0, or the operating system, respectively. If you specify mapname
as NONE, the string is not mapped.
If you execute FMTDP when NLS is disabled, the behavior is the same as for FMT. For more information
about display length, see WebSphere DataStage National Language Support Guide.
FMTS function
Syntax
FMTS (dynamic.array, format)
CALL -FMTS (return.array, dynamic.array, format)
CALL !FMTS (return.array, dynamic.array, format)
Description
Use the FMTS function to format elements of dynamic.array for output. Each element of the array is acted
upon independently and is returned as an element in a new dynamic array.
format is an expression that evaluates to a string of formatting codes. The syntax of the format expression
is:
[width] [background] justification [edit] [mask]
If dynamic.array evaluates to the null value, null is returned. If format evaluates to null, the FMTS function
fails and the program terminates with a run-time error message.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
FMTSDP function
Syntax
FMTSDP (dynamic.array, format [ , mapname ] )
Description
In NLS mode, use the FMTSDP function to format elements of dynamic.array for output in display
positions rather than character lengths. Each element of the array is acted upon independently and is
returned as an element in a new dynamic array. Any characters that cannot be mapped in dynamic.array
are assumed to have a display length of 1.
format is an expression that evaluates to a string of formatting codes. The syntax of the format expression
is:
[width] [background] justification [edit] [mask]
The format expression specifies the width of the output field, the placement of background or fill
characters, line justification, editing specifications, and format masking. For complete syntax details, see
the FMT function.
If format has a display length greater than 1, and there is only one display position left to fill, FMTSDP
enters the extra fill character. The returned string can occupy more display positions than you intend.
mapname is the name of an installed map. If mapname is not installed, the display positions of the
characters in dynamic.array are used. If any characters that cannot be mapped exist in dynamic.array, the
display length is 1, that is, the unmapped character displays as a single character that cannot be mapped.
If mapname is omitted, the map associated with the channel activated by the PRINTER statement ON
statement is used; otherwise, the map associated with the terminal channel (or print channel 0) is used.
You can also specify mapname as CRT, AUX, LPTR, and OS. These use the maps associated with the
terminal, auxiliary printer, print channel 0, or the operating system, respectively. If you specify mapname
as NONE, the string is not mapped.
If dynamic.array evaluates to the null value, null is returned. If format evaluates to null, the FMTSDP
function fails and the program terminates with a run-time error message.
Note: If you execute FMTSDP when NLS is disabled, the behavior is the same as for FMTS function.
For more information about display length, see WebSphere DataStage National Language Support Guide.
FMUL function
Syntax
FMUL (number1, number2)
CALL !FMUL (return.array, number1, number2)
Use the FMUL function to perform floating-point multiplication on two numeric values. If either number
evaluates to the null value, null is returned. return.array equates to number1 multiplied by number2.
This function is provided for compatibility with existing software. You can also use the * operator to
perform floating-point multiplication.
Example
PRINT FMUL(.234,.567)
FOLD function
Syntax
FOLD (string, length )
CALL !FOLD (subdivided.string, string, length)
Description
Use the FOLD function to divide a string into a number of substrings separated by field marks.
string is separated into substrings of length less than or equal to length. string is separated on blanks, if
possible, otherwise it is separated into substrings of the specified length.
If string evaluates to the null value, null is returned. If length is less than 1, an empty string is returned. If
length is the null value, the FOLD function fails and the program terminates with a run-time error
message.
Examples
PRINT FOLD("THIS IS A FOLDED STRING.",5)
In the following example, the blanks are taken as substring delimiters, and as no substring exceeds the
specified length of six characters, the output would be:
RED FM ORANGEFM YELLOWFM GREENFMBLUEFMINDIGOFMVIOLET
FOLDDP function
Syntax
FOLDDP (string, length [ , mapname ] )
In NLS mode, use the FOLDDP function to divide a string into a number of substrings separated by field
marks. The division is in display positions rather than character lengths.
string is separated into substrings of display length less than or equal to length. string is separated on
blanks, if possible, otherwise it is separated into substrings of the specified length.
If string evaluates to the null value, null is returned. If length is less than 1, an empty string is returned. If
length is the null value, the FOLDDP function fails and the program terminates with a run-time error
message.
If you execute FOLDDP when NLS is disabled, the behavior is the same as for FOLD function. For more
information about display length, see WebSphere DataStage National Language Support Guide.
FOOTING statement
Syntax
FOOTING [ON print.channel] footing
Description
Use the FOOTING statement to specify the text and format of the footing to print at the bottom of each
page of output.
The ON clause specifies the logical print channel to use for output. print.channel is an expression that
evaluates to a number from -1 through 255. If you do not use the ON clause, logical print channel 0 is
used, which prints to the user’s terminal if PRINTER OFF is set (see the PRINTER statement). Logical
print channel -1 prints the data on the screen, regardless of whether a PRINTER ON statement has been
executed.
footing is an expression that evaluates to the footing text and the control characters that specify the
footing’s format. You can use the following format control characters, enclosed in single quotation marks,
in the footing expression:
C[n] Prints footing line centered in a field of n blanks. If n is not specified, centers the line on the
page.
D Prints current date formatted as dd mmm yyyy.
G Inserts gaps to format footings.
I Resets page number, time, and date for PIOPEN flavor only.
Q Allows the use of the ] ^ and \ characters.
R[n] Inserts the record ID left-justified in a field of n blanks.
S Left-justified, inserted page number.
T Prints current time and date formatted as dd mmm yyyy hh:mm:ss. Time is in a 12-hour format
with ″am″ or ″pm″ appended.
\ Prints current time and date formatted as dd mmm yyyy hh:mm:ss. Time is in a 12-hour format
with ″am″ or ″pm″ appended. Do not put the backslash inside single quotation marks.
L Starts a new line.
] Starts a new line. Do not put the right bracket inside single quotation marks.
P[n] Prints current page number right-justified in a field of n blanks. The default value for n is 4.
Two single quotation marks ( ’ ’ ) print one single quotation mark in footing text.
When the program is executed, the format control characters produce the specified results. You can
specify multiple options in a single set of quotation marks.
If either print.channel or footing evaluates to the null value, the FOOTING statement fails and the program
terminates with a run-time error message.
Pagination begins with page 1 and increments automatically on generation of each new page or upon
encountering the PAGE statement.
Output to a terminal or printer is paged automatically. Use the N option in either a HEADING statement
or a FOOTING statement to turn off automatic paging.
The characters ] ^ and \ are control characters in headings and footings. To use these characters as
normal characters, you must use the Q option and enclose the control character in double or single
quotation marks. You only need to specify Q once in any heading or footing, but it must appear before
any occurrence of the characters ] ^ and \.
The control character G (for gap) can be used to add blanks to text in footings to bring the width of a line
up to device width. If G is specified once in a line, blanks are added to that part of the line to bring the
line up to the device width. If G is specified at more than one point in a line, the blank characters are
distributed as evenly as possible to those points.
See the following examples, in which the vertical bars represent the left and right margins:
Specification
Result
"Hello there"
|Hello there |
"’G’Hello there"
| Hello there|
"’G’Hello there’G’"
| Hello there |
"Hello’G’there"
|Hello there|
"’G’Hello’G’there’G’"
| Hello there |
The minimum gap size is 0 blanks. If a line is wider than the device width even when all the gaps are 0,
the line wraps, and all gaps remain 0.
If NLS is enabled, FOOTING calculates gaps using varying display positions rather than character
lengths. For more information about display length, see WebSphere DataStage National Language Support
Guide.
The control character S (for sequence number) is left-justified at the point where the S appears in the line.
Only one character space is reserved for the number. If the number of digits exceeds 1, any text to the
right is shifted right by the number of extra characters required.
INFORMATION Flavor
In an INFORMATION flavor account the default width of the page number field is the length of the page
number. Use the n argument to P to set the field width of the page number. You can also include multiple
P characters to specify the width of the page field, or you can include spaces in the text that immediately
precedes a P option. For example, ’PPP’ prints the page number right-justified in a field of three blanks.
Note: In all other flavors, ’PPP’ prints three identical page numbers, each in the default field of four.
Date Format
In an INFORMATION flavor account the default date format is mm-dd-yy, and the default time format is
24-hour style.
In PICK, IN2, REALITY, and IDEAL flavor accounts, use the HEADER.DATE option of the $OPTIONS
statement to cause HEADING statement, FOOTING, and PAGE statements to behave as they do in
INFORMATION flavor accounts.
PIOPEN Flavor
The control character P (for page) is right-justified at the point at which the P appears in the line. Only
one character space is reserved for the number. If the number of digits exceeds 1, literal characters to the
left of the initial position are overwritten. Normally you must enter a number of spaces to the left of the
P to allow for the maximum page number to appear without overwriting other literal characters. For
example, the statement:
FOOTING "This is page ’P’ of 100000"
The control character I (for initialize) resets the page number to 1, and resets the date.
Description
Use the FOR statement to create a FOR...NEXT program loop. A program loop is a series of statements
that execute repeatedly until the specified number of repetitions have been performed or until specified
conditions are met.
variable is assigned the value of start, which is the initial value of the counter. end is the end value of the
counter.
The loop.statements that follow the FOR clause execute until the NEXT statement is encountered. Then the
counter is adjusted by the amount specified by the STEP clause.
At this point a check is performed on the value of the counter. If it is less than or equal to end, program
execution branches back to the statement following the FOR clause and the process repeats. If it is greater
than end, execution continues with the statement following the NEXT statement.
The WHILE condition specifies that as long as the WHILE expression evaluates to true, the loop
continues to execute. When the WHILE expression evaluates to false, the loop ends, and program
execution continues with the statement following the NEXT statement. If a WHILE or UNTIL expression
evaluates to the null value, the condition is false.
The UNTIL condition specifies that the loop continues to execute only as long as the UNTIL expression
evaluates to false. When the UNTIL expression evaluates to true, the loop ends and program execution
continues with the statement following the NEXT statement.
expression can also contain a conditional statement. As expression you can use any statement that takes a
THEN or an ELSE clause, but without the THEN or ELSE clause. When the conditional statement would
execute the ELSE clause, expression evaluates to false; when the conditional statement would execute the
THEN clause, expression evaluates to true. The LOCKED clause is not supported in this context.
You can use multiple WHILE and UNTIL clauses in a FOR...NEXT loop.
Use the CONTINUE statement within FOR...NEXT to transfer control to the next iteration of the loop,
from any point in the loop.
Use the EXIT statement within FOR...NEXT to terminate the loop from any point within the loop.
If STEP is not specified, increment is assumed to be 1. If increment is negative, the end value of the counter
is less than the initial value. Each time the loop is processed, the counter is decreased by the amount
specified in the STEP clause. Execution continues to loop until the counter is less than end.
The body of the loop is skipped if start is greater than end, and increment is not negative. If start, end, or
increment evaluates to the null value, the FOR statement fails and the program terminates with a run-time
error message.
You can nest FOR...NEXT loops. That is, you can put a FOR...NEXT loop inside another FOR...NEXT
loop. When loops are nested, each loop must have a unique variable name as its counter. The NEXT
statement for the inside loop must appear before the NEXT statement for the outside loop.
If you omit the variables in the NEXT statement, the NEXT statement corresponds to the most recent
FOR statement. If a NEXT statement is encountered without a previous FOR statement, an error occurs
during compilation.
INFORMATION Flavor
In an INFORMATION flavor account the FOR variable is checked to see if it exceeds end before increment
is added to it. That means that the value of the FOR variable does not exceed end at the termination of
the loop. In IDEAL, PICK, IN2, and REALITY flavors the increment is made before the bound checking.
In this case it is possible for variable to exceed end. Use the FOR.INCR.BEF option of the $OPTIONS
statement to get IDEAL flavor behavior in an INFORMATION flavor account.
Examples
In the following example, the loop is executed 100 times or until control is transferred by one of the
statements in the loop:
FOR VAR=1 TO 100
NEXT VAR
FORMLIST statement
Syntax
FORMLIST [variable] [TO list.number] [ON ERROR statements]
Description
FSUB function
Syntax
FSUB (number1, number2)
CALL !FSUB (result, number1, number2)
Description
Use the FSUB function to perform floating-point subtraction on two numeric values. number2 is
subtracted from number1. If either number evaluates to the null value, null is returned. result equates to
number1 minus number2.
This function is provided for compatibility with existing software. You can also use the - operator to
perform floating-point subtraction.
Example
PRINT FSUB(.234,.567)
FUNCTION statement
Syntax
FUNCTION [name] [ ( [MAT] variable [ , [MAT] variable ...] ) ]
Use the FUNCTION statement to identify a user-written function and to specify the number and names
of the arguments to be passed to it. The FUNCTION statement must be the first line that is not a
comment within the user-written function. A user-written function can contain only one FUNCTION
statement.
name is specified for documentation purposes; it need not be the same as the function name or the name
used to reference the function in the calling program. name can be any valid variable name.
variable is an expression that passes values between the calling programs and the function. variables are
the formal parameters of the user-written function. When actual parameters are specified as arguments to
a user-written function, the actual parameters are referenced by the formal parameters so that calculations
performed in the user-written function use the actual parameters.
Separate variables by commas. Up to 254 variables can be passed to a user-written function. To pass an
array, you must precede the array name with the keyword MAT. When a user-written function is called,
the calling function must specify the same number of variables as are specified in the FUNCTION
statement.
An extra variable is hidden so that the user-written function can use it to return a value. An extra
variable is retained by the user-written function so that a value is returned by the RETURN statement
(value) statement. This extra variable is reported by the MAP and MAKE.MAP.FILE commands. If you
use the RETURN statement in a user-written function and you do not specify a value to return, an empty
string is returned by default.
The program that calls a user-written function must contain a DEFFUN statement that defines the
user-written function before it uses it. The user-written function must be cataloged in either a local
catalog or the system catalog, or it must be a record in the same object file as the calling program.
If the user-defined function recursively calls itself within the function, a DEFFUN statement must precede
it in the user-written function.
Examples
The following user-defined function SHORT compares the length of two arguments and returns the
shorter:
FUNCTION SHORT(A,B)
AL = LEN(A)
BL = LEN(B)
IF AL < BL THEN RESULT = A ELSE RESULT = B
RETURN(RESULT)
The following example defines a function called MYFUNC with the arguments or formal parameters A,
B, and C. It is followed by an example of the DEFFUN statement declaring and using the MYFUNC
function. The actual parameters held in X, Y, and Z are referenced by the formal parameters A, B, and C
so that the value assigned to T can be calculated.
FUNCTION MYFUNC(A, B, C)
Z = ...
RETURN (Z)
.
.
.
END
DEFFUN MYFUNC(X, Y, Z)
T = MYFUNC(X, Y, Z)
END
Description
Use the GES function to test if elements of one dynamic array are greater than or equal to corresponding
elements of another dynamic array.
Each element of array1 is compared with the corresponding element of array2. If the element from array1
is greater than or equal to the element from array2, a 1 is returned in the corresponding element of a new
dynamic array. If the element from array1 is less than the element from array2, a 0 is returned. If an
element of one dynamic array has no corresponding element in the other dynamic array, the undefined
element is evaluated as empty, and the comparison continues.
If either element of a corresponding pair is the null value, null is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
GET statements
Syntax
GET[X] read.var[ , length ] [ SETTING read.count ] FROM device
[ UNTIL eop.char.list ] [ RETURNING last.char.read ]
[ WAITING seconds ] [ THEN statements ] [ ELSE
statements ]
Description
Use GET statements to read a block of data from an input stream associated with a device, such as a
serial line or terminal. The device must be opened with the OPENDEV statement or OPENSEQ
statement. Once the device has been opened, the GET statements read data from the device. The GET
statements do not perform any pre- or postprocessing of the data stream, nor do they control local echo
characteristics. These aspects of terminal control are handled either by the application or by the device
driver. The behavior of certain devices can be managed through the TTYSET statement/TTYGET
statement interface.
Use the GETX statement to return the characters in ASCII hexadecimal format. For example, the sequence
of 8-bit character ″abcde″ is returned as the character string ″6162636465″. However, the value returned in
the last.char.read variable is in standard ASCII character form.
read.var is the variable into which the characters read from device are stored. If no data is read, read.var is
set to the empty string.
length is the expression evaluating to the number of characters read from the data stream; if length and
timeout are not specified, the default length is 1. If length is not specified, but an eop.char.list value is
included, no length limit is imposed on the input.
read.count is the variable that records the actual count of characters read and stored in read.var. This might
differ from length when a timeout condition occurs or when a recognized end-of-packet character is
detected.
eop.char.list is an expression that evaluates to a recognized end-of-packet delimiters. The GET operation
terminates if a valid end-of-packet character is encountered in the data stream before the requested
number of characters is read.
last.char.read is a variable that stores the last character read by the GET operation. If no data is read,
read.var is set to the empty string. If the input terminated due to the maximum number of characters
being read or because of a timeout condition, an empty string is returned.
seconds specifies the number of seconds the program should wait before the GET operation times out.
Terminating Conditions
GET statements read data from the device’s input stream until the first terminating condition is
encountered. The following table lists the possible terminating conditions:
Table 33. GET Statements Terminating Conditions
Condition Description
Requested read length has been satisfied The read is fully satisfied. read.var contains the characters
read, and last.char.read contains an empty string. Program
control passes to the THEN clause if present. The default
requested read length is one character unless an
end-of-packet value has been selected (in which case, no
length limit is used).
Recognized end-of-packet character has been processed The read is terminated by a special application-defined
character in the data stream. The data read to this point,
excluding the end-of-packet character, is stored in
read.var. The end-of-packet character is stored in
last.char.read. Program control passes to the THEN clause
if present. This terminating condition is only possible if
the UNTIL clause has been specified. If there is no
UNTIL clause, no end-of-packet characters are
recognized.
Timeout limit has expired The read could not be satisfied within the specified time
limitation. If no characters have been read, read.var and
last.char.read are set to the empty string, and read.count is
set to 0. The system status code is set to 0 and might be
checked with the STATUS function. Control passes to the
ELSE clause if present. This condition is only possible if
the WAITING clause is specified. In the absence of a
WAITING clause, the application waits until one of the
other terminating conditions is met.
Device input error An unrecoverable error occurred on the device.
Unrecoverable errors can include EOF conditions and
operating system reported I/O errors. In this case, the
data read to this point is stored in read.var, and the
empty string is stored in last.char.read. If no characters
have been read, read.var and last.char.read are set to the
empty string, and read.count is set to 0. The system status
code is set to a nonzero value and might checked with
the STATUS function. Control passes to the ELSE clause
if present.
For GET statements, the THEN and ELSE clauses are optional. They have different meanings and
produce different results, depending on the conditions specified for terminal input.
The following rules apply only if the THEN or ELSE clauses are specified:
v If the UNTIL clause is used without a WAITING clause or an expected length, the GET statement
behaves normally. The program waits indefinitely until a termination character is read, then executes
the THEN clause. The ELSE clause is never executed.
v If the WAITING clause is used, the GET statement behaves normally, and the ELSE clause is executed
only if the number of seconds for timeout has elapsed. If the input terminates for any other reason, it
executes the THEN clause.
v If the WAITING clause is not used and there is a finite number of characters to expect from the input,
then only the type-ahead buffer is examined for input. If the type-ahead buffer contains the expected
number of characters, it executes the THEN clause; otherwise it executes the ELSE clause. If the
type-ahead feature is turned off, the ELSE clause is always executed.
v In a special case, the ELSE clause is executed if the line has not been attached before executing the GET
statement.
In summary, unless the WAITING clause is used, specifying the THEN and ELSE clauses causes the GET
statement to behave like an INPUTIF statement...FROM statement. The exception to this is the UNTIL
clause without a maximum length specified, in which case the GET statement behaves normally and the
ELSE clause is never used.
Example
The following code fragment shows how the GET statement reads a number of data buffers representing
a transaction message from a device:
DIM SAVEBUFFER(10)
SAVELIMIT = 10
OPENDEV "TTY10" TO TTYLINE ELSE STOP "CANNOT OPEN
TTY10"
I = 1
LOOP
GET BUFFER,128 FROM TTYLINE UNTIL CHAR(10)
WAITING 10
ELSE
IF STATUS()
THEN PRINT "UNRECOVERABLE ERROR DETECTED ON
DEVICE,
"IM SAVEBUFFER(10)
SAVELIMIT = 10
OPENDEV "TTY10" TO TTYLINE ELSE STOP "CANNOT OPEN
TTY10"
I = 1
LOOP
GET BUFFER,128 FROM TTYLINE UNTIL CHAR(10)
WAITING 10
ELSE
IF STATUS()
THEN PRINT "UNRECOVERABLE ERROR DETECTED ON
DEVICE,":
ELSE PRINT "DEVICE TIMEOUT HAS OCCURRED, ":
PRINT "TRANSACTION CANNOT BE COMPLETED."
STOP
END
GETX statement
Use the GETX statement to read a block of data from an input stream and return the characters in ASCII
hexadecimal format. For details, see the GET statement.
GET(ARG.) statement
Syntax
GET(ARG. [ ,arg#] ) variable [THEN statements] [ELSE statements]
Description
Use the GET(ARG.) statement to retrieve the next command line argument. The command line is
delimited by blanks, and the first argument is assumed to be the first word after the program name.
When a cataloged program is invoked, the argument list starts with the second word in the command
line.
Blanks in quoted strings are not treated as delimiters and the string is treated as a single argument. For
example, ″54 76″ returns 54 76.
arg# specifies the command line argument to retrieve. It must evaluate to a number. If arg# is not
specified, the next command line argument is retrieved. The retrieved argument is assigned to variable.
THEN and ELSE statements are both optional. The THEN clause is executed if the argument is found.
The ELSE clause is executed if the argument is not found. If the argument is not found and no ELSE
clause is present, variable is set to an empty string.
If no arg# is specified or if arg# evaluates to 0, the argument to the right of the last argument retrieved is
assigned to variable. The GET statement fails if arg# evaluates to a number greater than the number of
command line arguments or if the last argument has been assigned and a GET with no arg# is used. To
move to the beginning of the argument list, set arg# to 1.
If arg# evaluates to the null value, the GET statement fails and the program terminates with a run-time
error message.
Example
If the command line is changed to RUN PROG, the system looks in the file PROG for the program with
the name of the first argument. If PROG is a cataloged program, the command line would have to be
changed to PROG ARG1 ARG2 ARG3 to get the same results.
GETLIST statement
Syntax
GETLIST listname [TO list.number] [SETTING variable]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the GETLIST statement to activate a saved select list so that a READNEXT statement can use it.
or:
record.ID account.name
record.ID is the record ID of a select list in the &SAVEDLISTS& file. If account.name is specified, the
&SAVEDLISTS& file of that account is used instead of the one in the local account.
If listname evaluates to the null value, the GETLIST statement fails and the program terminates with a
run-time error message.
The TO clause puts the list in a select list numbered 0 through 10. If list.number is not specified, the list is
saved as select list 0.
The SETTING clause assigns the count of the elements in the list to variable. The system variable
@SELECTED is also assigned this count whether or not the SETTING clause is used. If the list is retrieved
successfully, even if the list is empty, the THEN statements execute; if not, the ELSE statements execute.
PICK, REALITY, and IN2 flavor accounts store select lists in list variables instead of numbered select lists.
In those accounts, and in programs that use the VAR.SELECT option of the $OPTIONS statement, the
syntax of the GETLIST statement is:
GETLIST listname [TO list.variable] [SETTING variable]
{THEN statements [ELSE statements] | ELSE statements}
Description
In NLS mode use the GETLOCALE function to return the names of specified categories of the current
locale. The GETLOCALE function also returns the details of any saved locale that differs from the current
one.
category is one of the following tokens that are defined in the WebSphere DataStage include file
UVNLSLOC.H:
UVLC$ALL
The names of all the current locale categories as a dynamic array. The elements of the array are
separated by field marks. The categories are in the order Time, Numeric, Monetary, Ctype, and
Collate.
UVLC$SAVED
A dynamic array of all the saved locale categories.
UVLC$TIME
The setting of the Time category.
UVLC$NUMERIC
The setting of the Numeric category.
UVLC$MONETARY
The setting of the Monetary category.
UVLC$CTYPE
The setting of the Ctype category.
UVLC$COLLATE
The setting of the Collate category.
If the GETLOCALE function fails, it returns one of the following error tokens:
LCE$NO.LOCALES
WebSphere DataStage locales are not enabled.
LCE$BAD.CATEGORY
Category is invalid.
For more information about locales, see WebSphere DataStage National Language Support Guide.
GETREM function
Syntax
GETREM (dynamic.array)
Description
Use the GETREM function after the execution of a REMOVE statement, a REMOVE function, or a
REVREMOVE statement, to return the numeric value for the character position of the pointer associated
with dynamic.array.
Example
DYN = "THIS":@FM:"HERE":@FM:"STRING"
REMOVE VAR FROM DYN SETTING X
PRINT GETREM(DYN)
GOSUB statement
Syntax
GOSUB statement.label [ : ]
GO SUB statement.label [ : ]
Description
Use the GOSUB statement to transfer program control to an internal subroutine referenced by
statement.label. A colon ( : ) is optional in GOSUB statements, even though it is required after nonnumeric
statement labels at the beginning of program lines.
Use the RETURN statement at the end of the internal subroutine referenced by the GOSUB statement, to
transfer program control to the statement following the GOSUB statement.
Use the RETURN TO statement at the end of an internal subroutine to transfer control to a location in the
program other than the line following the GOSUB statement. In this case, use statement.label to refer to the
target location.
Be careful with the RETURN TO statement, because all other GOSUBs or CALLs active when the GOSUB
is executed remain active, and errors can result.
A program can call a subroutine any number of times. A subroutine can also be called from within
another subroutine; this process is called nesting subroutines. You can nest up to 256 GOSUB calls.
Subroutines can appear anywhere in the program but should be readily distinguishable from the main
program. To prevent inadvertent entry into the subroutine, precede it with a STOP statement, END, or
GOTO statement that directs program control around the subroutine.
Example
VAR=’ABKL1234’
FOR X=1 TO LEN(VAR)
Y=VAR[X,1]
GOSUB 100
NEXT X
STOP
100*
IF Y MATCHES ’1N’ THEN RETURN TO 200
PRINT ’ALPHA CHARACTER IN POSITION ’,X
RETURN
200*
PRINT ’NUMERIC CHARACTER IN POSITION ’,X
STOP
GOTO statement
Syntax
GO[TO] statement.label [ : ]
GO TO statement.label [ : ]
Description
Use the GOTO statement to transfer program control to the statement specified by statement.label. A colon
( : ) is optional in GOTO statements.
If the statement referenced is an executable statement, that statement and those that follow are executed.
If it is a nonexecutable statement, execution proceeds at the first executable statement encountered after
the referenced statement.
Example
X=80
GOTO 10
STOP
*
10*
IF X>20 THEN GO 20 ELSE STOP
*
20*
PRINT ’AT LABEL 20’
GO TO CALCULATE:
STOP
*
CALCULATE:
PRINT ’AT LABEL CALCULATE’
GROUP function
Syntax
GROUP (string, delimiter, occurrence [ ,num.substr] )
Description
Use the GROUP function to return one or more substrings located between specified delimiters in string.
delimiter evaluates to any character, including field mark, value mark, and subvalue marks. It delimits the
start and end of the substring. If delimiter evaluates to more than one character, only the first character is
used. Delimiters are not returned with the substring.
occurrence specifies which occurrence of the delimiter is to be used as a terminator. If occurrence is less
than 1, 1 is assumed.
If either delimiter or occurrence is not in the string, an empty string is returned, unless occurrence specifies
1. If occurrence is 1 and delimiter is not found, the entire string is returned. If delimiter is an empty string,
the entire string is returned.
If string evaluates to the null value, null is returned. If string contains CHAR(128) (that is, @NULL.STR),
it is treated like any other character in a string. If delimiter, occurrence, or num.substr evaluates to the null
value, the GROUP function fails and the program terminates with a run-time error message.
Examples
D=GROUP("###DHHH#KK","#",4)
PRINT "D= ",D
The variable D is set to DHHH because the data between the third and fourth occurrence of the delimiter
# is DHHH.
REC="ACADABA"
E=GROUP(REC,"A",2)
PRINT "E= ",E
Z is set to an empty string since ″?″ does not appear in the string.
Q=GROUP("+1+2+3ABAC","+",2,2)
PRINT "Q= ",Q
Q is set to ″1+2″ since two successive fields were specified to be returned after the second occurrence of
″+″.
GROUPSTORE statement
Syntax
GROUPSTORE new.string IN string USING start, n [ ,delimiter]
Description
Use the GROUPSTORE statement to modify character strings by inserting, deleting, or replacing fields
separated by specified delimiters.
start evaluates to a number specifying the starting field position. Modification begins at the field specified
by start. If start is greater than the number of fields in string, the required number of empty fields is
generated before the GROUPSTORE statement is executed.
n specifies the number of fields of new.string to insert in string. n determines how the GROUPSTORE
operation is executed. If n is positive, n fields in string are replaced with the first n fields of new.string. If
n is negative, n fields in string are replaced with all the fields in new.string. If n is 0, all the fields in
new.string are inserted in string before the field specified by start.
If string evaluates to the null value, null is returned. If new.string, start, n, or delimiter is null, the
GROUPSTORE statement fails and the program terminates with a run-time error message.
Example
Q=’1#2#3#4#5’
GROUPSTORE "A#B" IN Q USING 2,2,"#"
PRINT "TEST1= ",Q
*
Q=’1#2#3#4#5’
GROUPSTORE "A#B" IN Q USING 2,-2,"#"
PRINT "TEST2= ",Q
*
Q=’1#2#3#4#5’
GROUPSTORE "A#B" IN Q USING 2,0,"#"
PRINT "TEST3= ",Q
*
Q=’1#2#3#4#5’
GROUPSTORE "A#B#C#D" IN Q USING 1,4,"#"
PRINT "TEST4= ",Q
*
Q=’1#2#3#4#5’
GROUPSTORE "A#B#C#D" IN Q USING 7,3,"#"
PRINT "TEST5= ",Q
fGTS function
Syntax
GTS (array1, array2)
CALL -GTS (return.array, array1, array2)
CALL !GTS (return.array, array1, array2)
Description
Use the GTS function to test if elements of one dynamic array are greater than elements of another
dynamic array.
Each element of array1 is compared with the corresponding element of array2. If the element from array1
is greater than the element from array2, a 1 is returned in the corresponding element of a new dynamic
array. If the element from array1 is less than or equal to the element from array2, a 0 is returned. If an
If either of a corresponding pair of elements is the null value, null is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
HEADING statement
Syntax
HEADING [ON print.channel] heading
HEADINGE [ON print.channel] heading
HEADINGN [ON print.channel] heading
Description
Use the HEADING statement to specify the text and format of the heading to print at the top of each
page of output.
The ON clause specifies the logical print channel to use for output. print.channel is an expression that
evaluates to a number from -1 through 255. If you do not use the ON clause, logical print channel 0 is
used, which prints to the user’s terminal if PRINTER OFF is set (see the PRINTER statement). Logical
print channel -1 prints the data on the screen, regardless of whether a PRINTER ON statement has been
executed.
heading is an expression that evaluates to the heading text and the control characters that specify the
heading’s format. You can use the following format control characters, enclosed in single quotation
marks, in the heading expression:
C[n] Prints heading line centered in a field of n blanks. If n is not specified, centers the line on the
page.
D Prints current date formatted as dd mmm yyyy.
T Prints current time and date formatted as dd mmm yyyy hh:mm:ss. Time is in 12-hour format with
″am″ or ″pm″ appended.
\ Prints current time and date formatted as dd mmm yyyy hh:mm:ss. Time is in 12-hour format with
″am″ or ″pm″ appended. Do not put the backslash inside single quotation marks.
G Inserts gaps to format headings.
I Resets page number, time, and date for PIOPEN flavor only.
Q Allows the use of the ] ^ and \ characters.
R[n] Inserts the record ID left-justified in a field of n blanks.
L Starts a new line.
] Starts a new line. Do not put the right bracket inside single quotation marks.
N Suppresses automatic paging.
P[n] Prints current page number right-justified in a field of n blanks. The default value for n is 4.
S Left-justified, inserted page number.
^ Prints current page number right-justified in a field of n blanks. The default value for n is 4. Do
not put the caret inside single quotation marks.
Two single quotation marks ( ’ ’ ) print one single quotation mark in heading text.
If either print.channel or heading evaluates to the null value, the HEADING statement fails and the
program terminates with a run-time error message.
Pagination begins with page 1 and increments automatically on generation of each new page or upon
encountering the PAGE statement.
Output to a terminal or printer is paged automatically. Use the N option in either a HEADING or a
FOOTING statement to turn off automatic paging.
The HEADINGN statement is the same as the HEADING statement with the $OPTIONS
-HEADER.EJECT selected. HEADINGN suppresses a page eject with the HEADING statement. The page
eject is suppressed in IDEAL, PICK, REALITY, and IN2 flavor accounts.
The characters ] ^ and \ are control characters in headings and footings. To use these characters as
normal characters, you must use the Q option and enclose the control character in double or single
quotation marks. You only need to specify Q once in any heading or footing, but it must appear before
any occurrence of the characters ] ^ and \.
The control character G (for gap) can be used to add blanks to text in headings to bring the width of a
line up to device width. If G is specified once in a line, blanks are added to that part of the line to bring
the line up to the device width. If G is specified at more than one point in a line, the space characters are
distributed as evenly as possible to those points. See the following examples, in which the vertical bars
represent the left and right margins:
Specification
Result
"Hello there"
|Hello there |
"’G’Hello there"
| Hello there|
"’G’Hello there’G’"
| Hello there |
"Hello’G’there"
|Hello there|
"’G’Hello’G’there’G’"
| Hello there |
The minimum gap size is 0 blanks. If a line is wider than the device width even when all the gaps are 0,
the line wraps, and all gaps remain 0.
The control character S (for sequence number) is left-justified at the point where the S appears in the line.
Only one character space is reserved for the number. If the number of digits exceeds 1, any text to the
right is shifted right by the number of extra characters required. For example, the statement:
HEADING "This is page ’S’ of 100000"
INFORMATION Flavor
In an INFORMATION flavor account the default width of the page number field is the length of the page
number. Use the n argument to P to set the field width of the page number. You can also include multiple
P characters to specify the width of the page field, or you can include blanks in the text that immediately
precedes a P option. For example, ’PPP’ prints the page number right-justified in a field of three blanks.
Note: In all other flavors, ’PPP’ prints three identical page numbers, each in the default field of four.
Date Format
In an INFORMATION flavor account the default date format is mm-dd-yy, and the default time format is
24-hour style.
In PICK, IN2, REALITY, and IDEAL flavor accounts, use the HEADER.DATE option of the $OPTIONS
statement to cause HEADING, FOOTING statement, and PAGE statements to behave as they do in
INFORMATION flavor accounts.
PIOPEN Flavor
The control character P (for page) is right-justified at the point at which the P appears in the line. Only
one character space is reserved for the number. If the number of digits exceeds 1, literal characters to the
left of the initial position are overwritten. Normally you must enter a number of blanks to the left of the
P to allow for the maximum page number to appear without overwriting other literal characters. For
example, the statement:
HEADING "This is page ’P’ of 100000"
The control character I (for initialize) resets the page number to 1, and resets the date.
HUSH statement
Syntax
HUSH { ON | OFF | expression} [ SETTING status ]
Description
Use the HUSH statement to suppress the display of all output normally sent to a terminal during
processing. HUSH also suppresses output to a COMO file or TANDEM display.
SETTING status sets the value of a variable to the value of the HUSH state before the HUSH statement
was executed. It can be used instead of the STATUS function to save the state so that it can be restored
later. STATUS has a value of 1 if the previous state was HUSH ON or a value of 0 if the previous state
was HUSH OFF.
You might use this statement when you are transmitting information over phone lines or when you are
sending data to a hard-copy terminal. Both these situations result in slower transmission speeds. The
unnecessary data display makes the task even slower.
HUSH acts as a toggle. If it is used without a qualifier, it changes the current state.
Do not use this statement to shut off output display unless you are sure the display is unnecessary. When
you use HUSH ON, all output is suppressed including error messages and requests for information.
The previous state is returned by the STATUS function. If terminal output was suppressed prior to
execution of the HUSH statement, the STATUS function returns a 1. If terminal output was enabled
before execution of the HUSH statement, the STATUS function returns a 0.
Example
In the following example, terminal output is disabled with the HUSH statement and the previous state
was saved in the variable USER.HUSH.STATE.
After executing some other statements, the program returns the user’s process to the same HUSH state as
it was in previous to the execution of the first HUSH statement:
ICHECK function
Syntax
ICHECK ( dynamic.array [ , file.variable ] , key [ , column# ] )
Description
Use the ICHECK function to check if data you intend to write to an SQL table violates any SQL integrity
constraints. ICHECK verifies that specified data and primary keys satisfy the defined SQL integrity
constraints for an SQL table.
dynamic.array is an expression that evaluates to the data you want to check against any integrity
constraints.
file.variable specifies an open file. If file.variable is not specified, the default file variable is assumed (for
more information on default files, see the OPEN statement).
key is an expression that evaluates to the primary key you want to check against any integrity constraints.
column# is an expression that evaluates to the number of the column in the table whose data is to be
checked. If you do not specify column#, all columns in the file are checked. Column 0 specifies the
primary key (record ID).
If dynamic.array, file.variable, key, or column# evaluates to the null value, the ICHECK function fails and the
program terminates with a run-time error message.
You might use the ICHECK function to limit the amount of integrity checking that is done and thus
improve performance. If you do this, however, you are assuming responsibility for data integrity. For
example, you might want to use ICHECK with a program that changes only a few columns in a file. To
do this, turn off the OPENCHK configurable parameter, open the file with the OPEN statement rather
than the OPENCHECK statement, and use the ICHECK function before you write the updated record to
verify, for each column you are updating, that you are not violating the table’s integrity checks.
If the ON UPDATE clause of a referential constraint specifies an action, ICHECK always validates data
being written to the referenced table; it does not check the referencing table. Therefore, ICHECK can
succeed, but when the actual write is done, it can have a constraint failure while attempting to update
the referencing table. If the referential constraint does not have an ON UPDATE clause, or if these clauses
specify NO ACTION, the referencing table is checked to ensure that no row in it contains the old value of
the referenced column.
ICHECK does not check triggers when it checks other SQL integrity constraints. Therefore, a write that
fires a trigger can fail even if the ICHECK succeeds.
If the record violates more than one integrity constraint, ICHECK returns a dynamic array only for the
first constraint that causes a failure.
The ICHECK function is advisory only. That is, if two programs try to write the same data to the same
column defined as UNIQUE (see error 5), an ICHECK in the first program might pass. If the second
program writes data to the file before the first program writes the data that passed ICHECK, the first
program’s write fails even though the ICHECK did not fail.
ICONV function
Syntax
ICONV (string, conversion)
Use the ICONV function to convert string to a specified internal storage format. string is an expression
that evaluates to the string to be converted.
conversion is an expression that evaluates to one or more valid conversion codes, separated by value
marks (ASCII 253).
string is converted to the internal format specified by conversion. If multiple codes are used, they are
applied from left to right. The first conversion code converts the value of string. The second conversion
code converts the output of the first conversion, and so on.
If string evaluates to the null value, null is returned. If conversion evaluates to the null value, the ICONV
function fails and the program terminates with a run-time error message.
For information about converting strings to an external format, see the OCONV function.
Examples
ICONVS function
Syntax
ICONVS (dynamic.array, conversion)
CALL -ICONVS (return.array, dynamic.array, conversion)
CALL !ICONVS (return.array, dynamic.array, conversion)
Description
Use the ICONVS function to convert each element of dynamic.array to a specified internal storage format.
conversion is an expression that evaluates to one or more valid conversion codes, separated by value
marks (ASCII 253).
Each element of dynamic.array is converted to the internal format specified by conversion and is returned
in a dynamic array. If multiple codes are used, they are applied from left to right. The first conversion
code converts the value of each element of dynamic.array. The second conversion code converts the value
of each element of the output of the first conversion, and so on.
If dynamic.array evaluates to the null value, null is returned. If an element of dynamic.array is the null
value, null is returned for that element. If conversion evaluates to the null value, the ICONV function fails
and the program terminates with a run-time error message.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
For information about converting elements in a dynamic array to an external format, see the OCONVS
function.
Description
Use the IF statement to determine program flow based on the evaluation of expression. If the value of
expression is true, the THEN statements are executed. If the value of expression is false, the THEN
statements are ignored and the ELSE statements are executed. If expression is the null value, expression
evaluates to false. If no ELSE statements are present, program execution continues with the next
executable statement.
The IF statement must contain either a THEN clause or an ELSE clause. It need not include both.
Use the ISNULL function with the IF statement when you want to test whether the value of a variable is
the null value. This is the only way to test for the null value since null cannot be equal to any value,
including itself. The syntax is:
IF ISNULL (expression) ...
You can write IF...THEN statements on a single line or separated onto several lines. Separating statements
onto several lines can improve readability. Either way, the statements are executed identically.
You can nest IF...THEN statements. If the THEN or ELSE statements are written on more than one line,
you must use an END statement as the last statement of the THEN or ELSE statements.
Conditional Compilation
You can specify the conditions under which all or part of a BASIC program is to be compiled, using a
modified version of the IF statement. The syntax of the conditional compilation statement is the same as
that of the IF statement except for the test expression, which must be one of the following: $TRUE, $T,
$FALSE, or $F.
Example
X=10
IF X>5 THEN PRINT ’X IS GREATER THAN 5’;Y=3
*
IF Y>5 THEN STOP ELSE Z=9; PRINT ’Y IS LESS THAN 5’
*
IF Z=9 THEN PRINT ’Z EQUALS 9’
ELSE PRINT ’Z DOES NOT EQUAL 9’ ; STOP
*
IF Z=9 THEN
GOTO 10
END ELSE
STOP
END
*
10*
IF Y>4
THEN
IFS function
Syntax
IFS (dynamic.array, true.array, false.array)
CALL -IFS (return.array, dynamic.array, true.array, false.array)
CALL !IFS (return.array, dynamic.array, true.array, false.array)
Description
Use the IFS function to return a dynamic array whose elements are chosen individually from one of two
dynamic arrays based on the contents of a third dynamic array.
Each element of dynamic.array is evaluated. If the element evaluates to true, the corresponding element
from true.array is returned to the same element of a new dynamic array. If the element evaluates to false,
the corresponding element from false.array is returned. If there is no corresponding element in the correct
response array, an empty string is returned for that element. If an element is the null value, that element
evaluates to false.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
ILPROMPT function
Syntax
ILPROMPT (in.line.prompt)
Description
in.line.prompt is an expression that evaluates to a string containing in-line prompts. In-line prompts have
the following syntax:
<< [ control, ] ... text [ , option ] >>
control is an option that specifies the characteristics of the prompt. Separate multiple control options with
commas.
Table 35. Possible Control Options
Control Explanation
A Always prompts when the sentence containing the
control option is executed. If this option is not specified,
the input value from a previous execution of this prompt
is used.
Cn Uses the word in the nth position in the command line
as the input value. (The verb is in position 1.)
text is the prompt text to display. If you want to include quotation marks (single or double) or
backslashes as delimiters within the prompt text, you must enclose the entire text string in a set of
delimiters different from the delimiters you are using within the text string. For example, to print the
following prompt text:
’P’RINTER OR ’T’ERMINAL
or
"’P’RINTER OR ’T’ERMINAL"
option can be any valid ICONV function conversion or matching pattern (see the MATCH operator). A
conversion must be in parentheses.
If in.line.prompt evaluates to the null value, the ILPROMPT function fails and the program terminates
with a run-time error.
If the in-line prompt has a value, that value is substituted for the prompt. If the in-line prompt does not
have a value, the prompt is displayed to request an input value when the sentence is executed. The value
entered at the prompt is then substituted for the in-line prompt.
Example
A="This is your number. - <<number>>"
PRINT ILPROMPT(A)
PRINT ILPROMPT("Your number is <<number>>, and your
letter is
<<letter>>.")
INCLUDE statement
Syntax
INCLUDE [filename] program
INCLUDE program FROM filename
Description
Use the INCLUDE statement to direct the compiler to insert the source code in the record program and
compile it along with the main program. The INCLUDE statement differs from the $CHAIN statement in
that the compiler returns to the main program and continues compiling with the statement following the
INCLUDE statement.
When program is specified without filename, program must be a record in the same file as the program
currently containing the INCLUDE statement.
If program is a record in a different file, the name of the file in which it is located must be specified in the
INCLUDE statement, followed by the name of the program. The filename must specify a type 1 or type
19 file defined in the VOC file.
The INCLUDE statement is a synonym for the $INCLUDE and #INCLUDE statements.
Example
PRINT "START"
INCLUDE END
PRINT "FINISH"
When this program is compiled, the INCLUDE statement inserts code from the program END (see the
example on the END statement page). This is the program output:
START
THESE TWO LINES WILL PRINT ONLY
WHEN THE VALUE OF ’A’ IS ’YES’.
Description
Use the INDEX function to return the starting character position for the specified occurrence of substring
in string.
string is an expression that evaluates to any valid string. string is examined for the substring expression.
When substring is found and if it meets the occurrence criterion, the starting character position of the
substring is returned. If substring is an empty string, 1 is returned. If the specified occurrence of the
substring is not found, or if string or substring evaluate to the null value, 0 is returned.
If occurrence evaluates to the null value, the INDEX function fails and the program terminates with a
run-time error message.
In PICK, IN2, and REALITY flavor accounts, the search continues with the next character regardless of
whether it is part of the matched substring. Use the COUNT.OVLP option of the $OPTIONS statement to
get this behavior in IDEAL and INFORMATION flavor accounts.
Example
Q=’AAA11122ABB1619MM’
P=INDEX(Q,1,4)
PRINT "P= ",P
*
X=’XX’
Y=2
Q=’P1234XXOO1299XX00P’
TEST=INDEX(Q,X,Y)
PRINT "TEST= ",TEST
*
Q=INDEX("1234",’A’,1)
PRINT "Q= ",Q
* The substring cannot be found.
*
POS=INDEX(’222’,’2’,4)
PRINT "POS= ",POS
* The occurrence (4) of the substring does not exist.
INDEXS function
Syntax
INDEXS (dynamic.array, substring, occurrence)
CALL -INDEXS (return.array, dynamic.array, substring, occurrence)
CALL !INDEXS (return.array, dynamic.array, substring, occurrence)
Use the INDEXS function to return a dynamic array of the starting column positions for a specified
occurrence of a substring in each element of dynamic.array.
When substring is found, and if it meets the occurrence criterion, the starting column position of the
substring is returned. If substring is an empty string, 1 is returned. If the specified occurrence of substring
cannot be found, 0 is returned.
If dynamic.array evaluates to the null value, 0 is returned. If any element of dynamic.array is null, 0 is
returned for that element. If occurrence is the null value, the INDEXS function fails and the program
terminates with a run-time error message.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
INDICES function
Syntax
INDICES (file.variable [ ,indexname])
Description
Use the INDICES function to return information about the secondary key indexes in a file.
If only file.variable is specified, a dynamic array is returned that contains the index names of all secondary
indexes for the file. The index names are separated by field marks. If file.variable has no indexes, an
empty string is returned.
If indexname is specified, information is returned in a dynamic array for indexname. Field 1 of the dynamic
array contains the following information:
Table 36. Field 1 of Dynamic Arrays
Value Value can be... Description
Value 1 D Data descriptor index.
I I-descriptor index.
A A-descriptor index.
S S-descriptor index.
R Right-justified.
Value 8 N Not unique.
U Unique.
Value 9 part numbers Sub-valued list of distributed file part
numbers.
Value 10 1 Index needs building Sub-valued list corresponding to
subvalues in Value 9.
empty No build needed
Value 11 1 Empty strings not indexed Sub-valued list corresponding to
subvalues in Value 9.
empty Empty strings indexed
Value 12 1 Updating enabled Sub-valued list corresponding to
subvalues in Value 9.
empty Updating disabled
Value 13 index path names Sub-valued list of path names for
indexes on distributed file part files,
corresponding to subvalues in Value
9.
Value 14 1 Updates pending Sub-valued list corresponding to
subvalues in Value 9.
empty No updates pending
Value 15 L Left-justified Sub-valued list corresponding to
subvalues in Value 9.
R Right-justified
Value 16 N Not unique Sub-valued list corresponding to
subvalues in Value 9.
U Unique
Value 17 collate name Name of the Collate convention of
the index.
If Value 1 of Field 1 is D, A, or S, Field 2 contains the field location (that is, the field number), and Field 6
contains either S (single-valued field) or M (multi-valued field).
If Value 1 of Field 1 is I or SQL, the other fields of the dynamic array contain the following information,
derived from the I-descriptor in the file dictionary:
Field Value can be...
If Value 1 of Field 1 is C, the other fields of the dynamic array contain the following information, derived
from the A- or S-descriptor in the file dictionary:
Field Value can be...
Field 2
Field number (location of field)
Field 3
Column heading
Field 4
Association code
Fields 5-6
Empty
Field 7
Output conversion code
Field 8
Correlative code
Field 9
L or R (justification)
Field 10
Width of display column
If either file.variable or indexname is the null value, the INDICES function fails and the program terminates
with a run-time error message.
Any file updates executed in a transaction (that is, between a BEGIN TRANSACTION statement and a
COMMIT statement) are not accessible to the INDICES function until after the COMMIT statement has
been executed.
INMAT function
Syntax
INMAT ( [array] )
Description
Use the INMAT function to return the number of array elements that have been loaded after the
execution of a MATREAD statements, MATREADL statement, MATREADU statement, or MATPARSE
statement, or to return the modulo of a file after the execution of an OPEN statement. You can also use
the INMAT function after a DIM statement to determine whether the DIM statement failed due to lack of
available memory. If a preceding DIM statement fails, INMAT returns a value of 1.
If the matrix assignment exceeds the number of elements specified in its dimensioning statement, the zero
element is loaded by the MATREAD, MATREADL, MATREADU, or MATPARSE statement. If the array
dimensioning is too small and the zero element has been loaded, the INMAT function returns a value of
0.
If array is specified, the INMAT function returns the current dimensions of the array. If array is the null
value, the INMAT function fails and the program terminates with a run-time error message.
Example
DIM X(6)
D=’123456’
MATPARSE X FROM D,’’
Y=INMAT()
PRINT ’Y= ’:Y
*
DIM X(5)
A=’CBDGFH’
MATPARSE X FROM A,’’
C=INMAT()
PRINT ’C= ’:C
*
OPEN ’’,’VOC’ TO FILE ELSE STOP
T=INMAT()
PRINT ’T= ’:T
INPUT statement
Syntax
INPUT variable [ ,length] [ : ] [ _ ]
INPUT @ (col, row) [ , | : ] variable [ ,length] [ : ] [format] [ _ ]
INPUTIF @ (col, row) [ , | : ] variable [ ,length] [ : ] [format] [ _ ]
[THEN statements] [ELSE statements]
Use the INPUT statement to halt program execution and prompt the user to enter a response. Data
entered at the terminal or supplied by a DATA statement in response to an INPUT statement is assigned
to variable. Input supplied by a DATA statement is echoed to the terminal. If the response is a RETURN
with no preceding data, an empty string is assigned to variable.
The INPUT statement has two syntaxes. The first syntax displays a prompt and assigns the input to
variable. The second syntax specifies the location of the input field on the screen and lets you display the
current value of variable. Both the current value and the displayed input can be formatted.
Use the INPUTIF statement to assign the contents of the type-ahead buffer to a variable. If the type-ahead
buffer is empty, the ELSE statements are executed, otherwise any THEN statements are executed.
Use the @ expression to specify the position of the input field. The prompt is displayed one character to
the left of the beginning of the field, and the current value of variable is displayed as the value in the
input field. The user can edit the displayed value or enter a new value. If the first character typed in
response to the prompt is an editing key, the user can edit the contents of the field. If the first character
typed is anything else, the field’s contents are deleted and the user can enter a new value. Editing keys
are defined in the terminfo files; they can also be defined by the KEYEDIT statement. Calculations are
based on display length rather than character length.
col and row are expressions that specify the column and row positions of the input prompt. The prompt is
positioned one character to the left of the input field. Because the prompt character is positioned to the
left of the col position, you must set the prompt to the empty string if you want to use column 0.
Otherwise, the screen is erased before the prompt appears.
length specifies the maximum number of characters allowed as input. When the maximum number of
characters is entered, input is terminated. If the @ expression is used, the newline is suppressed.
If length evaluates to less than 0 (for example, -1), the input buffer is tested for the presence of characters.
If characters are present, variable is set to 1, otherwise it is set to 0. No input is performed.
If you use the underscore ( _ ) with the length expression, the user must enter the RETURN manually at
the terminal when input is complete. Only the specified number of characters is accepted.
Use a format expression to validate input against a format mask and to format the displayed input field.
The syntax of the format expression is the same as that for the FMT function. If you specify a length
expression together with a format expression, length checking is performed. If input does not conform to
the format mask, an error message appears at the bottom of the screen, prompting the user for the correct
input.
The colon ( : ) suppresses the newline after input is terminated. This allows multiple input prompts on a
single line.
The default prompt character is a question mark. Use the PROMPT statement to reassign the prompt
character.
The INPUT statement prints only the prompt character on the screen. To print a variable name or prompt
text along with the prompt, precede the INPUT statement with a PRINT statement.
The INPUT statement lets the user type ahead when entering a response. Users familiar with a sequence
of prompts can save time by entering data at their own speed, not waiting for all prompts to be
displayed. Responses to a sequence of INPUT prompts are accepted in the order in which they are
entered.
If NLS is enabled, INPUT @ displays the initial value of an external multibyte character set through the
mask as best as possible. If the user enters a new value, mask disappears, and an input field of the
approximate length (not including any inserted characters) is entered. For details about format and mask,
see the FMTDP function.
Only backspace and kill are supported for editing functions when using a format mask with input. When
the user finishes the input, the new value is displayed again through the mask in the same way as the
original value. For more information about NLS in BASIC programs, seeWebSphere DataStage National
Language Support Guide.
PICK Flavor
In a PICK flavor account, the syntax of the INPUT and INPUT @ statements includes THEN and ELSE
clauses:
INPUT variable [ ,length] [ : ] [ _ ] [THEN statements] [ELSE
statements]
INPUT @ (col, row) [ , | : ] variable [ ,length] [ : ] [format] [ _ ]
[THEN statements] [ELSE statements]
To use THEN and ELSE clauses with INPUT statements in other flavors, use the INPUT.ELSE option of
the $OPTIONS statement.
Examples
In the following examples of program output, bold type indicates words the user types. In the first
example the value entered is assigned to the variable NAME:
Source Lines
Program Output
INPUT NAME PRINT NAME
? DaveDave
In the next example the value entered is assigned to the variable CODE. Only the first seven characters
are recognized. A RETURN and a LINEFEED automatically occur.
Source Lines
Program Output
INPUT CODE, 7 PRINT CODE
? 12345671234567
In the next example the user can enter more than two characters. The program waits for a RETURN to
end input, but only the first two characters are assigned to the variable YES.
Source Lines
Program Output
INPUT YES, 2_ PRINT YES
? 123412
In the next example the input buffer is tested for the presence of characters. If characters are present,
VAR is set to 1, otherwise it is set to 0. No input is actually done.
Source Lines
Program Output
INPUT VAR, -1 PRINT VAR
0
In the next example the PRINT statement puts INPUT NAME before the input prompt:
Source Lines
Program Output
PRINT ″INPUT NAME″: INPUT NAME PRINT NAME
INPUT NAME? DaveDave
In the next example the contents of X are displayed at column 5, row 5 in a field of 10 characters. The
user edits the field, replacing its original contents (CURRENT) with new contents (NEW). The new input
is displayed. If the PRINT statement after the INPUT statement were not used, X would be printed
immediately following the input field on the same line, since INPUT with the @ expression does not
execute a LINEFEED after a RETURN.
Source Lines
Program Output
PRINT @(-1) X = ″CURRENT″ INPUT @(5,5) X,10 PRINT PRINT X
?NEW_______
NEW
INPUTCLEAR statement
Syntax
INPUTCLEAR
Description
Use the INPUTCLEAR statement to clear the type-ahead buffer. You can use this statement before input
prompts so input is not affected by unwanted characters.
Example
PRINT "DO YOU WANT TO CONTINUE (Y/N)?"
INPUTCLEAR
INPUT ANSWER, 1
INPUTDISP statement
Syntax
INPUTDISP [@(col, row) [ , | : ] ] variable [format]
Use the INPUTDISP statement with an @ expression to position the cursor at a specified location and
define a format for the variable to print. The current contents of variable are displayed as the value in the
defined field. Calculations are based on display length rather than character length.
col specifies the column position, and row specifies the row position.
format is an expression that defines how the variable is to be displayed in the output field. The syntax of
the format expression is the same as that for the FMT function.
Example
PRINT @(-1)
X = "CURRENT LINE"
INPUTDISP @(5,5),X"10T"
INPUTDP statement
Syntax
INPUTDP variable [ , length] [ : ] [ _ ] [THEN statements] [ELSE
statements]
Description
In NLS mode, use the INPUTDP statement to let the user enter data. The INPUTDP statement is similar
to the INPUT statement, INPUTIF statement, and INPUTDISP statement, but it calculates display
positions rather than character lengths.
length specifies the maximum number of characters in display length allowed as input. INPUTDP
calculates the display length of the input field based on the current terminal map. When the specified
number of characters is entered, an automatic newline is executed.
The colon ( : ) executes the RETURN, suppressing the newline. This allows multiple input prompts on a
single line.
If you use the underscore ( _ ), the user must enter the RETURN manually when input is complete, and
the newline is not executed.
For more information about display length, see WebSphere DataStage National Language Support Guide.
INPUTERR statement
Syntax
INPUTERR [error.message]
Use the INPUTERR statement to print a formatted error message on the bottom line of the terminal.
error.message is an expression that evaluates to the error message text. The message is cleared by the next
INPUT statement or is overwritten by the next INPUTERR or PRINTERR statement. INPUTERR clears
the type-ahead buffer.
error.message can be any BASIC expression. The elements of the expression can be numeric or character
strings, variables, constants, or literal strings. The null value cannot be output. The expression can be a
single expression or a series of expressions separated by commas ( , ) or colons ( : ) for output formatting.
If no error message is designated, a blank line is printed. If error.message evaluates to the null value, the
default error message is printed:
Message ID is NULL: undefined error
Expressions separated by commas are printed at preset tab positions. The default setting of the tab stop is
10 characters. For information about changing the default setting, see the TABSTOP statement. Multiple
commas can be used together to cause multiple tabulations between expressions.
Expressions separated by colons are concatenated: that is, the expression following the colon is printed
immediately after the expression preceding the colon.
INPUTIF statement
Use the INPUTIF statement to assign the contents of the type-ahead buffer to a variable. For details, see
the INPUT statement.
INPUTNULL statement
Syntax
INPUTNULL character
Description
Use the INPUTNULL statement to define a character to be recognized as an empty string when it is input
in response to an INPUT statement. If the only input to the INPUT statement is character, that character is
recognized as an empty string. character replaces the default value of the INPUT variable with an empty
string. If character evaluates to the null value, the INPUTNULL statement fails and the program
terminates with a run-time error message.
You can also assign an empty string to the variable used in the INPUT statement @ statement before
executing the INPUT @. In this case entering a RETURN leaves the variable set to the empty string.
Note: Although the name of this statement is INPUTNULL, it does not define character to be recognized
as the null value. It defines it to be recognized as an empty string.
INPUTTRAP statement
Syntax
INPUTTRAP [trap.chars] {GOTO | GOSUB} label [ ,label ...]
Use the INPUTTRAP statement to branch to a program label or subroutine when a trap character is
input. Execution is passed to the statement label which corresponds to the trap number of the trap
character. If the trap number is larger than the number of labels, execution is passed to the statement
specified by the last label in the list.
trap.chars is an expression that evaluates to a string of characters, each of which defines a trap character.
The first character in the string is defined as trap one. Additional characters are assigned consecutive trap
numbers. Each trap character corresponds to one of the labels in the label list. If trap.chars evaluates to the
null value, the INPUTTRAP statement fails and the program terminates with a run-time error message.
Using GOTO causes execution to be passed to the specified statement label. Control is not returned to the
INPUTTRAP statement except by the use of another trap. Using GOSUB causes execution to be passed to
the specified subroutine, but control can be returned to the INPUTTRAP statement by a RETURN (value)
statement. Control is returned to the statement following the INPUTTRAP statement, not the INPUT
statement @ statement that received the trap.
INS statement
Syntax
INS expression BEFORE dynamic.array < field# [ ,value# [ ,subvalue#]
] >
Description
Use the INS statement to insert a new field, value, or subvalue into the specified dynamic.array.
field#, value#, and subvalue# specify the type and position of the new element to be inserted and are called
delimiter expressions.
There are three possible outcomes of the INS statement, depending on the delimiter expressions specified.
Case 1:
If both value# and subvalue# are omitted or are 0, INS inserts a new field with the value of
expression into the dynamic array.
v If field# is positive and less than or equal to the number of fields in dynamic.array, the value of
expression followed by a field mark is inserted before the field specified by field#.
v If field# is -1, a field mark followed by the value of expression is appended to the last field in
dynamic.array.
v If field# is positive and greater than the number of fields in dynamic.array, the proper number of
field marks followed by the value of expression are appended so that the value of field# is the
number of the new field.
Case 2:
If value# is nonzero and subvalue# is omitted or is 0, INS inserts a new value with the value of
expression into the dynamic array.
v If value# is positive and less than or equal to the number of values in the field, the value of
expression followed by a value mark is inserted before the value specified by value#.
v If value# is -1, a value mark followed by the value of expression is appended to the last value in
the field.
In IDEAL, PICK, PIOPEN, and REALITY flavor accounts, if expression is an empty string and the new
element is appended to the end of the dynamic array, the end of a field, or the end of a value, the
dynamic array, field, or value is left unchanged. Additional delimiters are not appended. Use the
EXTRA.DELIM option of the $OPTIONS statements to make the INS statement append a delimiter to the
dynamic array, field, or value.
If expression evaluates to the null value, null is inserted into dynamic.array. If dynamic.array evaluates to the
null value, it remains unchanged by the insertion. If the INS statement references a subordinate element
of an element whose value is the null value, the dynamic array is unchanged.
If any delimiter expression is the null value, the INS statement fails and the program terminates with a
run-time error message.
In INFORMATION and IN2 flavor accounts, if expression is an empty string and the new element is
appended to the end of the dynamic array, the end of a field, or the end of a value, a delimiter is
appended to the dynamic array, field, or value. Use the -EXTRA.DELIM option of the $OPTIONS
statement to make the INS statement work as it does in IDEAL, PICK, and REALITY flavor accounts.
Examples
In the following examples a field mark is shown by F, a value mark is shown by V, and a subvalue mark
is shown by S.
The first example inserts the character # before the first field and sets Q to #FFF1V2V3S6F9F5F7V3:
R=@FM:@FM:1:@VM:2:@VM:3:@SM:6:@FM:9:@FM:5:@FM:7:@VM:3
Q=R
INS "#" BEFORE Q<1,0,0>
The next example inserts a # before the third value of field 3 and sets the value of Q to
FF1V2V#V3S6F9F5F7V3:
Q=R
INS "#" BEFORE Q<3,3,0>
The next example inserts a value mark followed by a # after the last value in the field and sets Q to
FF1V2V3S6F9V#F5F7V3:
The next example inserts a # before the second subvalue of the second value of field 3 and sets Q to
FF1V2S#V3S6F9F5F7V3:
Q=R
INS "#" BEFORE Q<3,2,2>
INSERT function
Syntax
INSERT (dynamic.array, field#, value#, subvalue#, expression)
INSERT (dynamic.array, field# [ ,value# [ ,subvalue#] ] ; expression)
Description
Use the INSERT function to return a dynamic array that has a new field, value, or subvalue inserted into
the specified dynamic array.
field#, value#, and subvalue# specify the type and position of the new element to be inserted and are called
delimiter expressions. value# and subvalue# are optional, but if either is omitted, a semicolon ( ; ) must
precede expression, as shown in the second syntax line.
There are three possible outcomes of the INSERT function, depending on the delimiter expressions
specified.
Case 1:
If both value# and subvalue# are omitted or are 0, INSERT inserts a new field with the value of
expression into the dynamic array.
v If field# is positive and less than or equal to the number of fields in dynamic.array, the value of
expression followed by a field mark is inserted before the field specified by field#.
v If field# is -1, a field mark followed by the value of expression is appended to the last field in
dynamic.array.
v If field# is positive and greater than the number of fields in dynamic.array, the proper number of
field marks followed by the value of expression are appended so that the value of field# is the
number of the new field.
Case 2:
If value# is nonzero and subvalue# is omitted or is 0, INSERT inserts a new value with the value of
expression into the dynamic array.
v If value# is positive and less than or equal to the number of values in the field, the value of
expression followed by a value mark is inserted before the value specified by value#.
v If value# is -1, a value mark followed by the value of expression is appended to the last value in
the field.
v If value# is positive and greater than the number of values in the field, the proper number of
value marks followed by the value of expression are appended to the last value in the specified
field so that the number of the new value in the field is value#.
Case 3:
If field#, value#, and subvalue# are all specified, INSERT inserts a new subvalue with the value of
expression into the dynamic array.
In IDEAL, PICK, PIOPEN, and REALITY accounts, if expression is an empty string and the new element is
appended to the end of the dynamic array, the end of a field, or the end of a value, the dynamic array,
field, or value is left unchanged. Additional delimiters are not appended. Use the EXTRA.DELIM option
of the $OPTIONS statement to make the INSERT function append a delimiter to the dynamic array, field,
or value.
If expression evaluates to the null value, null is inserted into dynamic.array. If dynamic.array evaluates to the
null value, it remains unchanged by the insertion. If any delimiter expression is the null value, the
INSERT function fails and the program terminates with a run-time error message.
In INFORMATION and IN2 flavor accounts, if expression is an empty string and the new element is
appended to the end of the dynamic array, the end of a field, or the end of a value, a delimiter is
appended to the dynamic array, field, or value. Use the -EXTRA.DELIM option of the $OPTIONS
statement to make the INSERT function work as it does in IDEAL, PICK, and REALITY flavor accounts.
Examples
In the following examples a field mark is shown by F, a value mark is shown by V, and a subvalue mark
is shown by S.
The first example inserts the character # before the first field and sets Q to #FFF1V2V3S6F9F5F7V:
R=@FM:@FM:1:@VM:2:@VM:3:@SM:6:@FM:9:@FM:5:@FM:7:@VM:3
Q=INSERT(R,1,0,0,"#")
The next example inserts a # before the third value of field 3 and sets the value of Q to
FF1V2V#V3S6F9F5F7V3:
Q=INSERT(R,3,3;"#")
The next example inserts a value mark followed by a # after the last value in the field and sets Q to
FF1V2V3S6F9V#F5F7V3:
Q=INSERT(R,4,-1,0,"#")
The next example inserts a # before the second subvalue of the second value of field 3 and sets Q to
FF1V2S#V3S6F9F5F7V3:
Q=INSERT(R,3,2,2;"#")
INT function
Syntax
INT (expression)
expression must evaluate to a numeric value. Any arithmetic operations specified are calculated using the
full accuracy of the system. The fractional portion of the value is truncated, not rounded, and the integer
portion remaining is returned.
Example
PRINT "123.45 ", INT(123.45)
PRINT "454.95 ", INT(454.95)
ISNULL function
Syntax
ISNULL (variable)
Description
Use the ISNULL function to test whether a variable is the null value. If variable is the null value, 1 (true)
is returned, otherwise 0 (false) is returned. This is the only way to test for the null value since the null
value is not equal to any value, including itself.
Example
X = @NULL
Y = @NULL.STR
PRINT ISNULL(X), ISNULL(Y)
ISNULLS function
Syntax
ISNULLS (dynamic.array)
CALL -ISNULLS (return.array, dynamic.array)
Description
Use the ISNULLS function to test whether any element of dynamic.array is the null value. A dynamic
array is returned, each of whose elements is either 1 (true) or 0 (false). If an element in dynamic.array is
the null value, 1 is returned, otherwise 0 is returned. This is the only way to test for the null value since
the null value is not equal to any value, including itself.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
ITYPE function
Syntax
ITYPE (i.type)
Description
Use the ITYPE function to return the value resulting from the evaluation of an I-descriptor expression in
a WebSphere DataStage file dictionary.
i.type is an expression evaluating to the contents of the compiled I-descriptor. The I-descriptor must have
been compiled before the ITYPE function uses it, otherwise you get a run-time error message.
i.type can be set to the I-descriptor to be evaluated in several ways. One way is to read the I-descriptor
from a file dictionary into a variable, then use the variable as the argument to the ITYPE function. If the
I-descriptor references a record ID, the current value of the system variable @ID is used. If the
I-descriptor references field values in a data record, the data is taken from the current value of the system
variable @RECORD.
To assign field values to @RECORD, read a record from the data file into @RECORD before invoking the
ITYPE function.
If i.type evaluates to the null value, the ITYPE function fails and the program terminates with a run-time
error message.
Example
KEYEDIT statement
Syntax
KEYEDIT (function, key) [ , (function, key) ] ...
Description
Use the KEYEDIT statement to assign specific keyboard keys to the editing functions of the INPUT
statement @ statement, and to the !EDIT.INPUT and !GET.KEY subroutines. KEYEDIT supports the
following editing functions:
v Left arrow (<-)
v Enter (Return)
v Back space
v Right arrow (->)
v Insert character
v Delete character
v Insert mode on
v Insert mode off
v Clear field
v Erase to end-of-line
v Insert mode toggle
In addition to the supported editing functions, two codes exist to designate the Esc and function keys.
function is an expression that evaluates to a numeric code assigned to a particular editing function.
Code Function
1 Function key
2 Left arrow (<-)
3 Return key
4 Back space
key is an expression evaluating to a decimal value that designates the keyboard key to assign to the
editing function. There are three key types, described in the following table:
Table 37. Key Types
Type Decimal Value Description
Control 1 through 31 Single character control codes ASCII
1 through 31.
Escape 32 through 159 Consists of the characters defined by
the Esc key followed by the ASCII
value 0 through 127 (see ″Defining
Escape Keys″).
Function 160 through 2,139,062,303 Consists of the characters defined by
the FUNCTION key followed by the
ASCII value 0 through 127. You can
specify up to four ASCII values for
complex keys (see ″Defining Function
Keys″).
If either function or key evaluates to the null value or an empty string, the KEYEDIT statement fails, the
program terminates, and a run-time error message is produced.
To define key, you must know the ASCII value generated by the keyboard on the terminal being used.
Once you know the ASCII code sequence generated by a particular keyboard key, you can use one of the
following three methods for deriving the numeric key value.
A control key is one whose ASCII value falls within the range of 1 through 31. Generally keys of this
type consist of pressing a keyboard key while holding down the Ctrl key. The key value is the ASCII code
value, i.e., Ctrl-A is 1, Ctrl-M is 13, etc.
An escape key is one which consists of pressing the Esc key followed by a single ASCII value. The Esc
key can be defined by issuing a KEYEDIT statement using a function value of 5 and the ASCII value of
the escape character for the key parameter, e.g., KEYEDIT (5,27).
The key value for an escape key is derived by adding the ASCII value of the character following the Esc
key and 32. The constant 32 is added to ensure that the final key value falls within the range of 32 to 159,
i.e., Esc-a is 33 (1+32), Esc-z is 122 (90+32), Esc-p is 144 (112+32), and so on.
A function key is similar to an escape key but consists of a function key followed by one or more ASCII
values. The function key can be defined by issuing a KEYEDIT statement using a function value of 1 and
the ASCII value of the function character for the key parameter, e.g., KEYEDIT(1,1).
Deriving the key value for a function key depends on the number of characters in the sequence the
keyboard key generates. Because the KEYEDIT statement recognizes function keys that generate character
sequences up to five characters long, the following method can be used to derive the key value.
This character sequence is to be assigned to the Clear Field functionality of the INPUT @ statement. It can
be broken into five separate characters, identified as follows:
Table 38. Description of the Sample Clear Field Functionality
Character ASCII Value Meaning
Ctrl-A 1 The preamble character (defines the
function key)
] 93 The first character
6 54 The second character
~ 126 The third character
<Return> 10 The fourth character
First you define the function key value. Do this by issuing the KEYEDIT statement with a function value
of 1 and with a key value defined as the ASCII value of the preamble character, i.e., KEYEDIT (1, 1).
Once you define the function key, the following formula is applied to the remaining characters in the
sequence:
ASCII value * ( 2(8*(character position - 1)
)
The results of each calculation are then added together. Finally, the constant 160 is added to insure that
the final key parameter value falls within the range of 160 through 2,139,062,303. For our example above,
Historically, key values falling in the range of 160 through 287 included an implied Return, as there was
no method for supporting multiple character sequences. With the support of multiple character
sequences, you must now include the Return in the calculation for proper key recognition, with one
exception. For legacy key values that fall within the range of 160 through 287, a Return is automatically
appended to the end of the character sequence, yielding an internal key parameter of greater value.
Before supporting multiple character sequences, this function key would have been defined as:
KEYEDIT (1, 1), (11, 225)
(1,1) defined the preamble of the function key, and (11, 225) defined the Clear-to-end-of-line key. The 225
value was derived by adding 160 to B (ASCII 65). The <Return> (ASCII 10) was implied. This can be
shown by using the SYSTEM(1050) function to return the internal trap table contents:
Table 40. Contents of Internal Trap Table
# Type Value Key
0 1 3 10
1 1 3 13
2 1 1 1
3 1 11 2785
You can use the KEYEDIT statement to designate keys that are recognized as unsupported by the
!EDIT.INPUT subroutine. When the !EDIT.INPUT subroutine encounters an unsupported key, it sounds
the terminal bell.
Define an unsupported key by assigning any negative decimal value for the function parameter.
The SYSTEM function(1050) function returns a dynamic array of defined KEYEDIT, KEYEXIT statement
and KEYTRAP statement keys. Field marks (ASCII 254) delimit the elements of the dynamic array. Each
field in the dynamic array has the following structure:
function.parameter and key.parameter are the values passed as parameters to the associated statement,
except for the INPUTNULL value.
Example
The following example illustrates the use of the KEYEDIT statement and the SYSTEM(1050) function:
KEYEDIT (1,1), (2,21), (3,13), (4,8), (6,6),
(12,176043773)
KEYTRAP (1,2)
keys.dfn=SYSTEM(1050)
PRINT "#","Type","Value","Key"
XX=DCOUNT(keys.dfn,@FM)
FOR I=1 TO XX
print I-1,keys.dfn<I,1>,keys.dfn<I,2>,keys.dfn<I,3>
NEXT I
KEYEXIT statement
Syntax
KEYEXIT (value, key) [ , (value, key) ] ...
Description
Use the KEYEXIT statement to specify exit traps for the keys assigned specific functions by the KEYEDIT
statement. When an exit trap key is typed, the variable being edited with the INPUT statement@
statement or the !EDIT.INPUT subroutine remains in its last edited state. Use the KEYTRAP statement to
restore the variable to its initial state.
value is an expression that specifies a user-defined trap number for each key assigned by the KEYEDIT
statement.
key is a decimal value that designates the specific keyboard key assigned to the editing function. There
are three key types, described in the following table:
See the KEYEDIT statement for how to derive the decimal value of control, escape, and function keys.
If either the value or key expression evaluates to the null value or an empty string, the KEYEXIT
statement fails, the program terminates, and a run-time error message is produced.
KEYEXIT sets the STATUS function to the trap number of any trap key typed by the user.
Examples
The following example sets up Ctrl-B as an exit trap key. The STATUS function is set to 1 when the user
types the key.
KEYEXIT (1,2)
The next example sets up Ctrl-K as an exit trap key. The STATUS function is set to 2 when the user types
the key.
KEYEXIT (2,11)
KEYIN function
Syntax
KEYIN ( )
Description
Use the KEYIN function to read a single character from the input buffer and return it. All special
character handling (such as case inversion, erase, kill, and so on) is disabled. UNIX special character
handling (processing of interrupts, XON/XOFF, conversion of CR to LF, and so on) still takes place.
No arguments are required with the KEYIN function; however, parentheses are required.
KEYTRAP statement
Syntax
KEYTRAP (value, key) [ , (value, key) ] ...
Use the KEYTRAP statement to specify traps for the keys assigned specific functions by the KEYEDIT
statement. When a trap key is typed, the variable being edited with the INPUT statement @ statement or
the !EDIT.INPUT subroutine is restored to its initial state. Use the KEYEXIT statement to leave the
variable in its last edited state.
value is an expression that evaluates to a user-defined trap number for each key assigned by the
KEYEDIT statement.
key is a decimal value which designates the specific keyboard key assigned to the editing function. There
are three key types, described in the following table:
Table 42. Key Types
Type Decimal Value Description
Control 1 through 31 Single character control codes ASCII
1 through 31.
Escape 32 through 159 Consists of the characters defined by
the Esc key followed by the ASCII
value 0 through 127.
Function 160 through 2,139,062,303 Consists of the characters defined by
the function key followed by the
ASCII value 0 through 127. A
maximum of four ASCII values might
be specified for complex keys.
See the KEYEDIT statement for how to derive the decimal value of control, escape, and function keys.
If either the value or key expression evaluates to the null value or an empty string, the KEYEXIT
statement fails, the program terminates, and a run-time error message is produced.
KEYTRAP sets the STATUS function to the trap number of any trap key typed by the user.
Examples
The following example sets up Ctrl-B as a trap key. The STATUS function is set to 1 when the user types
the key.
KEYTRAP (1, 2)
The next example defines function key values for the F1, F2, F3, and F4 keys on a Wyse 50 terminal:
KEYEDIT (1,1)
KEYTRAP (1,224), (2,225), (3,226), (4,227)
PRINT @(-1)
VALUE = "KEY"
INPUT @ (10,10):VALUE
X=STATUS()
BEGIN CASE
CASE X = 1
PRINT "FUNCTION KEY 1"
CASE X =2
PRINT "FUNCTION KEY 2"
CASE X =3
PRINT "FUNCTION KEY 3"
CASE X =4
PRINT "FUNCTION KEY 4"
LEFT function
Syntax
LEFT (string, n)
Description
Use the LEFT function to extract a substring comprising the first n characters of a string, without
specifying the starting character position. It is equivalent to the following substring extraction operation:
string [ 1, length ]
If string evaluates to the null value, null is returned. If n evaluates to the null value, the LEFT function
fails and the program terminates with a run-time error message.
Example
PRINT LEFT("ABCDEFGH",3)
LEN function
Syntax
LEN (string)
Description
Use the LEN function to return the number of characters in string. Calculations are based on character
length rather than display length.
string must be a string value. The characters in string are counted, and the count is returned.
The LEN function includes all blank spaces, including trailing blanks, in the calculation.
If NLS is enabled, use the LENDP function to return the length of a string in display positions rather
than character length. FFor more information about display length, see WebSphere DataStage National
Language Support Guide.
Example
P="PORTLAND, OREGON"
PRINT "LEN(P)= ",LEN(P)
*
NUMBER=123456789
PRINT "LENGTH OF NUMBER IS ",LEN(NUMBER)
Description
In NLS mode, use the LENDP function to return the number of display positions occupied by string
when using the specified map. Calculations are based on display length rather than character length.
mapname is the name of an installed map. If mapname is not installed, the character length of string is
returned.
If mapname is omitted, the map associated with the channel activated by PRINTER statement ON is used,
otherwise it uses the map for print channel 0. You can also specify mapname as CRT, AUX, LPTR, and OS.
These values use the maps associated with the terminal, auxiliary printer, print channel 0, or the
operating system, respectively. If you specify mapname as NONE, the string is not mapped.
The LENDP function includes all blank spaces, including trailing blanks, in the calculation.
If you use the LENDP function with NLS disabled, the program behaves as if the LEN function is used.
See the LEN function to return the length of a string in character rather than display positions.
For more information about display length, see WebSphere DataStage National Language Support Guide.
LENS function
Syntax
LENS (dynamic.array)
CALL -LENS (return.array, dynamic.array)
CALL !LENS (return.array, dynamic.array)
Description
Use the LENS function to return a dynamic array of the number of display positions in each element of
dynamic.array. Calculations are based on character length rather than display length.
Each element of dyamic.array must be a string value. The characters in each element of dynamic.array are
counted, and the counts are returned.
The LENS function includes all blank spaces, including trailing blanks, in the calculation.
If dynamic.array evaluates to the null value, 0 is returned. If any element of dynamic.array is null, 0 is
returned for that element.
If NLS is enabled, use the LENSDP function to return a dynamic array of the number of characters in
each element of dynamic.array in display positions. For more information about display length, see
WebSphere DataStage National Language Support Guide.
Description
In NLS mode, use the LENSDP function to return a dynamic array of the number of display positions
occupied by each element of dynamic.array. Calculations are based on display length rather than character
length.
Each element of dynamic.array must be a string value. The display lengths of each element of
dynamic.array are counted, and the counts are returned.
mapname is the name of an installed map. If mapname is not installed, the character length of string is
returned.
If mapname is omitted, the map associated with the channel activated by PRINTER statement ON is used,
otherwise it uses the map for print channel 0. You can also specify mapname as CRT, AUX, LPTR, and OS.
These values use the maps associated with the terminal, auxiliary printer, print channel 0, or the
operating system, respectively. If you specify mapname as NONE, the string is not mapped.
The LENSDP function includes all blank spaces, including trailing blanks, in the calculation.
If dynamic.array evaluates to the null value, 0 is returned. If any element of dynamic.array is null, 0 is
returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
If you use the LENSDP function with NLS disabled, the program behaves as if the LENS function is
used. See the LENS function to return the length of a string in character length rather than display
length.
For more information about display length, see WebSphere DataStage National Language Support Guide.
LES function
Syntax
LES (array1, array2)
CALL -LES (return.array, array1, array2)
CALL !LES (return.array, array1, array2)
Description
Use the LES function to test if elements of one dynamic array are less than or equal to the elements of
another dynamic array.
Each element of array1 is compared with the corresponding element of array2. If the element from array1
is less than or equal to the element from array2, a 1 is returned in the corresponding element of a new
dynamic array. If the element from array1 is greater than the element from array2, a 0 is returned. If an
element of one dynamic array has no corresponding element in the other dynamic array, the undefined
element is evaluated as empty, and the comparison continues.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
LET statement
Syntax
[LET] variable = expression
Description
Use the LET statement to assign the value of expression to variable. See assignment statements for more
information about assigning values to variables.
Example
LET A=55
LET B=45
LET C=A+B
LET D="55+45="
LET E=D:C
PRINT E
LN function
Syntax
LN (expression)
Description
Use the LN function to calculate the natural logarithm of the value of an expression, using base ″e″. The
value of ″e″ is approximately 2.71828. expression must evaluate to a numeric value greater than 0.
If expression is 0 or negative, 0 is returned and a warning is printed. If expression evaluates to the null
value, null is returned.
Example
PRINT LN(6)
LOCALEINFO function
Syntax
LOCALEINFO (category)
Description
In NLS mode, use the LOCALEINFO function to retrieve the settings of the current locale.
category is one of the following tokens that are defined in the WebSphere DataStage include file
UVNLSLOC.H:
If the specified category is set to OFF, LOCALEINFO returns the string OFF.
If the LOCALEINFO function fails to execute, LOCALEINFO returns one of the following:
LCE$NO.LOCALES
NLS locales are not in force.
LCE$BAD.CATEGORY
Category is invalid.
For more information about locales, see WebSphere DataStage National Language Support Guide.
Example
The following example shows the contents of the multi valued DAYS field when the locale FR-FRENCH
is current. Information for LCT$DAYS is contained in the UVNLSLOC.H file in the INCLUDE directory
in the UV account directory.
category.info = LOCALEINFO(LC$TIME)
PRINT category.info<LCT$DAYS>
LOCATE statement
Syntax (IDEAL, REALITY)
LOCATE expression IN dynamic.array [ < field# [ ,value#] > ] [
,start] [BY seq] SETTING variable {THEN statements [ELSE statements] | ELSE statements}
Syntax (PICK)
LOCATE (expression, dynamic.array [ ,field# [ ,value#] ] ; variable [
;seq] )
{THEN statements [ELSE statements] | ELSE statements}
Syntax (INFORMATION)
LOCATE expression IN dynamic.array < field# [ ,value# [ ,subvalue#]
] >
[BY seq] SETTING variable {THEN statements [ELSE statements] | ELSE statements}
Description
Use a LOCATE statement to search dynamic.array for expression and to return a value indicating one of the
following:
v Where expression was found in dynamic.array
v Where expression should be inserted in dynamic.array if it was not found
expression evaluates to the string to be searched for in dynamic.array. If expression or dynamic.array evaluate
to the null value, variable is set to 0 and the ELSE statements are executed. If expression and dynamic.array
both evaluate to empty strings, variable is set to 1 and the THEN statements are executed.
start evaluates to a number specifying the field, value, or subvalue from which to start the search.
In the IDEAL and PICK syntaxes the delimiter expressions specify the level of the search, and start
specifies the starting position of the search. In INFORMATION syntax the delimiter expressions specify
the starting position of the search.
If any delimiter expression or start evaluates to the null value, the LOCATE statement fails and the
program terminates with a run-time error message.
variable stores the index of expression. variable returns a field number, value number, or a subvalue
number, depending on the delimiter expressions used. variable is set to a number representing one of the
following:
v The index of the element containing expression, if such an element is found
v An index that can be used in an INSERT function to create a new element with the value specified by
expression
During the search, fields are processed as single-valued fields even if they contain value or subvalue
marks. Values are processed as single values, even if they contain subvalue marks.
If the elements to be searched are sorted in one of the ascending or descending ASCII sequences listed
below, you can use the BY seq expression to end the search. The search ends at the place where expression
should be inserted to maintain the ASCII sequence, rather than at the end of the list of specified elements.
Use the following values for seq to describe the ASCII sequence being searched:
″AL″ or ″A″
Ascending, left-justified (standard alphanumeric sort)
″AR″ Ascending, right-justified
″DL″ or ″D″
Descending, left-justified (standard alphanumeric sort)
″DR″ Descending, right-justified
seq does not reorder the elements in dynamic.array; it specifies the terminating conditions for the search. If
a seq expression is used and the elements are not in the sequence indicated by seq, an element with the
value of expression might not be found. If seq evaluates to the null value, the statement fails and the
program terminates.
Use the INFO.LOCATE option of the $OPTIONS statement to use the INFORMATION syntax in other
flavors.
If NLS is enabled, the LOCATE statement with a BY seq expression uses the Collate convention as
specified in the NLS.LC.COLLATE file to determine the sort order for characters with ascending or
descending sequences. The Collate convention defines rules for casing, accents, and ordering. For more
information about how NLS calculates the order, see WebSphere DataStage National Language Support Guide.
The following sections describe the three possible outcomes from the different uses of delimiter
expressions in the IDEAL, REALITY, and PICK versions of the LOCATE statement.
Case 1:
If field# and value# are omitted, the search starts at the first field in dynamic.array.
Case 2:
If only field# is specified and it is greater than 0, the search starts at the first value in the field
indicated by field#. If field# is less than or equal to 0, both field# and value# are ignored.
Case 3:
If both field# and value# are specified, the search starts at the first subvalue in the value specified
by value#, in the field specified by field#. If field# is greater than 0, but value# is less than or equal
to 0, LOCATE behaves as though only field# is specified.
If a field, value, or subvalue containing expression is found, variable returns the index of the located field,
value, or subvalue relative to the start of dynamic.array, field#, or value#, respectively (not the start of the
search). If a field, value, or subvalue containing expression is not found, variable is set to the number of
fields, values, or subvalues in the array plus 1, and the ELSE statements are executed.
INFORMATION Syntax
When you use the INFORMATION flavor syntax of LOCATE, three outcomes can result depending on
how the delimiter expressions are used. The results are described as case 1, case 2, and case 3.
Case 1:
If both value# and subvalue# are omitted or are both less than or equal to 0, the search starts at the
field indicated by field#.
Case 2:
If subvalue# is omitted or is less than or equal to 0, the search starts at the value indicated by
value#, in the field indicated by field#. If field# is less than or equal to 0, field# defaults to 1.
Case 3:
If field#, value#, and subvalue# are all specified and are all nonzero, the search starts at the
subvalue indicated by subvalue#, in the value specified by value#, in the field specified by field#. If
field# or value# are less than or equal to 0, they default to 1.
If a field, value, or subvalue containing expression is found, variable is set to the index of the located field
relative to the start of dynamic.array, the field, or the value, respectively (not the start of the search).
If no field containing expression is found, variable is set to the number of the field at which the search
terminated, and the ELSE statements are executed. If no value or subvalue containing expression is found,
variable is set to the number of values or subvalues plus 1, and the ELSE statements are executed.
Examples
The examples show the IDEAL and REALITY flavor LOCATE statement. A field mark is shown by F, a
value mark is shown by V, and a subvalue mark is shown by S.
Q=’X’:@SM:"$":@SM:’Y’:@VM:’Z’:@SM:4:@SM:2:@VM:’B’:@VM
PRINT "Q= ":Q
LOCATE "$" IN Q <1> SETTING WHERE ELSE PRINT ’ERROR’
PRINT "WHERE= ",WHERE
LOCATE "$" IN Q <1,1> SETTING HERE ELSE PRINT ’ERROR’
PRINT "HERE= ", HERE
NUMBERS=122:@FM:123:@FM:126:@FM:130:@FM
PRINT "BEFORE INSERT, NUMBERS= ",NUMBERS
NUM= 128
LOCATE NUM IN NUMBERS <2> BY "AR" SETTING X ELSE
NUMBERS = INSERT(NUMBERS,X,0,0,NUM)
PRINT "AFTER INSERT, NUMBERS= ",NUMBERS
END
LOCK statement
Syntax
LOCK expression [THEN statements] [ELSE statements]
Description
Use the LOCK statement to protect specified user-defined resources or events against unauthorized use or
simultaneous data file access by different users.
There are 64 public semaphore locks in the WebSphere DataStage system. They are task synchronization
tools but have no intrinsic definitions. You must define the resource or event associated with each
semaphore, ensuring that there are no conflicts in definition or usage of these semaphores throughout the
entire system.
expression evaluates to a number in the range of 0 through 63 that specifies the lock to be set. A program
can reset a lock any number of times and with any frequency desired. If expression evaluates to the null
value, the LOCK statement fails and the program terminates with a run-time error message.
If program B tries to set a lock already set by program A, execution of program B is suspended until the
first lock is released by program A; execution of program B then continues.
The ELSE clause provides an alternative to this procedure. When a LOCK statement specifies a lock that
has already been set, the ELSE clause is executed rather than program execution being suspended.
Program termination does not automatically release locks set in the program. Each LOCK statement must
have a corresponding UNLOCK statement. If a program locks the same semaphore more than once
during its execution, a single UNLOCK statement releases that semaphore.
Alternatively, locks can be released by logging off the system or by executing either the QUIT command
or the CLEAR.LOCKS command.
You can check the status of locks with the LIST.LOCKS command; this lists the locks on the screen. The
unlocked state is indicated by 0. The locked state is indicated by a number other than 0 (including both
positive and negative numbers). The number is the unique signature of the user who has set the lock.
Note: The LOCK statement protects user-defined resources only. The READL statement,, READL
statement, “READ statements” on page 273, MATREADL statement, and MATREADU statements use a
different method of protecting files and records.
Example
The following example sets lock 60, executes the LIST.LOCKS command, then unlocks all locks set by the
program:
LOCK 60 ELSE PRINT "ALREADY LOCKED"
EXECUTE "LIST.LOCKS"
UNLOCK
LOOP statement
Syntax
LOOP
[loop.statements]
[CONTINUE | EXIT]
[{WHILE | UNTIL} expression [DO] ]
[loop.statements]
[CONTINUE | EXIT]
REPEAT
Description
Use the LOOP statement to start a LOOP...REPEAT program loop. A program loop is a series of
statements that executes for a specified number of repetitions or until specified conditions are met.
Use the WHILE clause to indicate that the loop should execute repeatedly as long as the WHILE
expression evaluates to true (1). When the WHILE expression evaluates to false (0), repetition of the loop
stops, and program execution continues with the statement following the REPEAT statement.
Use the UNTIL clause to put opposite conditions on the LOOP statement. The UNTIL clause indicates
that the loop should execute repeatedly as long as the UNTIL expression evaluates to false (0). When the
UNTIL expression evaluates to true (1), repetition of the loop stops, and program execution continues
with the statement following the REPEAT statement.
If a WHILE or UNTIL expression evaluates to the null value, the condition is false.
You can use multiple WHILE and UNTIL clauses in a LOOP...REPEAT loop. You can also nest
LOOP...REPEAT loops. If a REPEAT statement is encountered without a previous LOOP statement, an
error occurs during compilation.
Use the CONTINUE statement within LOOP...REPEAT to transfer control to the next iteration of the loop
from any point in the loop.
Use the EXIT statement within LOOP...REPEAT to terminate the loop from any point within the loop.
Although it is possible to exit the loop by means other than the conditional WHILE and UNTIL
statements (for example, by using GOTO or GOSUB in the DO statements), it is not recommended. Such
a programming technique is not in keeping with good structured programming practice.
Examples
Source Lines
Program Output
X=0 LOOP UNTIL X>4 DO PRINT ″X= ″,X X=X+1 REPEAT
X= 0
X= 1
X= 2
X= 3
X= 4
A=20 LOOP PRINT ″A= ″, A A=A-1 UNTIL A=15 REPEAT
A= 20
A= 19
A= 18
A= 17
A= 16
Q=3 LOOP PRINT ″Q= ″,Q WHILE Q DO Q=Q-1 REPEAT
Q= 3
Q= 2
Q= 1
Q= 0
EXECUTE ″SELECT VOC FIRST 5″ MORE=1 LOOP READNEXT ID ELSE MORE=0 WHILE MORE
DO PRINT ID REPEAT
5 record(s)
selected to
SELECT list #0.
LOOP
HASH.TEST
QUIT.KEY
P
CLEAR.LOCKS
EXECUTE ″SELECT VOC FIRST 5″ LOOP WHILE READNEXT ID DO PRINT ID REPEAT
5 record(s)
selected to
SELECT list #0.
LOOP
HASH.TEST
QUIT.KEY
P
CLEAR.LOCKS
Description
Use the LOWER function to return a value equal to expression, except that system delimiters which
appear in expression are converted to the next lower-level delimiter: field marks are changed to value
marks, value marks are changed to subvalue marks, and so on. If expression evaluates to the null value,
null is returned.
PIOPEN Flavor
In PIOPEN flavor, the delimiters that can be lowered are CHAR(255) through CHAR(252). All other
characters are left unchanged. You can obtain PIOPEN flavor for the LOWER function by:
v Compiling your program in a PIOPEN flavor account
v Specifying the $OPTIONS INFO.MARKS statement
Examples
In the following examples an item mark is shown by I, a field mark is shown by F, a value mark is
shown by V, a subvalue mark is shown by S, and a text mark is shown by T. CHAR(250) is shown as Z.
LTS function
Syntax
LTS (array1, array2)
CALL -LTS (return.array, array1, array2)
CALL !LTS (return.array, array1, array2)
Use the LTS function to test if elements of one dynamic array are less than elements of another dynamic
array.
Each element of array1 is compared with the corresponding element of array2. If the element from array1
is less than the element from array2, a 1 is returned in the corresponding element of a new dynamic
array. If the element from array1 is greater than or equal to the element from array2, a 0 is returned. If an
element of one dynamic array has no corresponding element in the other dynamic array, the undefined
element is evaluated as an empty string, and the comparison continues.
If either of a corresponding pair of elements is the null value, null is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
MAT statement
Syntax
MAT array = expression
MAT array1 = MAT array2
Description
Use the MAT statement to assign one value to all of the elements in the array or to assign all the values
of one array to the values of another array.
Use the first syntax to assign the same value to all array elements. Use any valid expression. The value of
expression becomes the value of each array element.
Use the second syntax to assign values from the elements of array2 to the elements of array1. Both arrays
must previously be named and dimensioned. The dimensioning of the two arrays can be different. The
values of the elements of the new array are assigned in consecutive order, regardless of whether the
dimensions of the arrays are the same or not. If array2 has more elements than in array1, the extra
elements are ignored. If array2 has fewer elements, the extra elements of array1 are not assigned.
Note: Do not use the MAT statement to assign individual elements of an array.
Examples
Source Lines
Program Output
DIM ARRAY(5) QTY=10 MAT ARRAY=QTY FOR X=1 TO 5 PRINT ″ARRAY(″:X:″)=″,ARRAY(X) NEXT
X
ARRAY(1)= 10
ARRAY(2)= 10
ARRAY(3)= 10
ARRAY(4)= 10
ARRAY(5)= 10
DIM ONE(4,1) MAT ONE=1 DIM TWO(2,2) MAT TWO = MAT ONE FOR Y=1 TO 4 PRINT
″ONE(″:Y:″,1)=″,ONE(Y,1) NEXT Y
ONE(1,1)= 1
ONE(2,1)= 1
ONE(3,1)= 1
ONE(4,1)= 1
The following example sets all elements in ARRAY to the empty string:
MAT ARRAY=’’
MATBUILD statement
Syntax
MATBUILD dynamic.array FROM array [ ,start [ ,end] ] [USING
delimiter]
Description
Use the MATBUILD statement to build a dynamic array from a dimensioned array.
dynamic.array is created by concatenating the elements of array beginning with start and finishing with
end. If start and end are not specified or are out of range, they default to 1 and the size of the array
respectively.
array must be named and dimensioned in a DIMENSION statement or COMMON statement before it is
used in this statement.
delimiter specifies characters to be inserted between fields of the dynamic array. If delimiter is not
specified, it defaults to a field mark. To specify no delimiter, specify USING without delimiter.
If an element of array is the null value, the dynamic array will contain CHAR(128) for that element. If
start, end, or delimiter is the null value, the MATBUILD statement fails and the program terminates with a
run-time error.
Overflow Elements
PICK, IN2, and REALITY flavor dimensioned arrays contain overflow elements in the last element.
INFORMATION and IDEAL flavor dimensioned arrays contain overflow elements in element 0.
In PICK, IN2, and REALITY flavor accounts, if end is not specified, dynamic.array contains the overflow
elements of array. In IDEAL and INFORMATION flavor accounts, to get the overflow elements you must
specify end as less than or equal to 0, or as greater than the size of array.
REALITY flavor accounts use only the first character of delimiter, and if USING is specified without a
delimiter, delimiter defaults to a field mark rather than an empty string.
MATCH operator
Syntax
string MATCH[ES] pattern
Description
Use the MATCH operator or its synonym MATCHES to compare a string expression with a pattern.
The following table lists the pattern codes and their definitions:
Table 44. Pattern Matching Codes
Pattern Definition
... Any number of any characters (including none).
0X Any number of any characters (including none).
nX n number of any characters.
0A Any number of alphabetic characters (including none).
nA n number of alphabetic characters.
0N Any number of numeric characters (including none).
nN n number of numeric characters.
’text’ Exact text; any literal string (quotation marks required).
″text″ Exact text; any literal string (quotation marks required).
If n is longer than nine digits, it is used as text in a pattern rather than as a repeating factor for a special
character. For example, the pattern ″1234567890N″ is treated as a literal string, not as a pattern of
1,234,567,890 numeric characters.
If the string being evaluated matches the pattern, the expression evaluates as true ( 1 ); otherwise, it
evaluates as false ( 0 ). If either string or pattern is the null value, the match evaluates as false.
A tilde ( ~ ) placed immediately before pattern specifies a negative match. That is, it specifies a pattern or
a part of a pattern that does not match the expression or a part of the expression. The match is true only
if string and pattern are of equal length and differ in at least one character. An example of a negative
match pattern is:
"’A’~’X’5N
This pattern returns a value of true if the expression begins with the letter A, which is not followed by
the letter X, and which is followed by any five numeric characters. Thus AB55555 matches the pattern,
but AX55555, A55555, AX5555, and A5555 do not.
You can specify multiple patterns by separating them with value marks (ASCII CHAR(253) ). The
following expression is true if the address is either 16 alphabetic characters or 4 numeric characters
followed by 12 alphabetic characters; otherwise, it is false:
ADDRESS MATCHES "16A": CHAR(253): "4N12A"
An empty string matches the following patterns: ″0A″, ″0X″, ″0N″, ″...″, ″″, ’’, or \\.
If NLS is enabled, the MATCH operator uses the current values for alphabetic and numeric characters
specified in the NLS.LC.CTYPE file. For more information about the NLS.LC.CTYPE file, see WebSphere
DataStage National Language Support Guide.
MATCHFIELD function
Syntax
MATCHFIELD (string, pattern, field)
Use the MATCHFIELD function to check a string against a match pattern (see the MATCH operator for
information about pattern matching).
field is an expression that evaluates to the portion of the match string to be returned.
If string matches pattern, the MATCHFIELD function returns the portion of string that matches the
specified field in pattern. If string does not match pattern, or if string or pattern evaluates to the null value,
the MATCHFIELD function returns an empty string. If field evaluates to the null value, the
MATCHFIELD function fails and the program terminates with a run-time error.
pattern must contain specifiers to cover all characters contained in string. For example, the following
statement returns an empty string because not all parts of string are specified in the pattern:
MATCHFIELD ("XYZ123AB", "3X3N", 1)
To achieve a positive pattern match on string above, the following statement might be used:
MATCHFIELD ("XYZ123AB", "3X3N0X", 1)
Examples
Source Lines
Program Output
Q=MATCHFIELD(″AA123BBB9″,″2A0N3A0N″,3) PRINT ″Q= ″,Q
Q= BBB
ADDR=’20 GREEN ST. NATICK, MA.,01234’ ZIP=MATCHFIELD(ADDR,″0N0X5N″,3) PRINT ″ZIP=
″,ZIP
ZIP= 01234
INV=’PART12345 BLUE AU’ COL=MATCHFIELD(INV,″10X4A3X″,2) PRINT ″COL= ″,COL
COL= BLUE
In the following example the string does not match the pattern:
Source Lines
Program Output
XYZ=MATCHFIELD(’ABCDE1234’,″2N3A4N″,1) PRINT ″XYZ= ″,XYZ
XYZ=
In the following example the entire string does not match the pattern:
Source Lines
Program Output
ABC=MATCHFIELD(’1234AB’,″4N1A″,2) PRINT ″ABC= ″,ABC
ABC=
MATPARSE statement
Syntax
MATPARSE array FROM dynamic.array [ ,delimiter]
MATPARSE array [ ,start [ ,end] ] FROM dynamic.array [USING
delimiter]
[SETTING elements]
Use the MATPARSE statement to separate the fields of dynamic.array into consecutive elements of array.
array must be named and dimensioned in a DIMENSION statement or COMMON statement before it is
used in this statement.
start specifies the starting position in array. If start is less than 1, it defaults to 1.
end specifies the ending position in array. If end is less than 1 or greater than the length of array, it
defaults to the length of array.
delimiter is an expression evaluating to the characters used to delimit elements in dynamic.array. Use a
comma or USING to separate delimiter from dynamic.array. delimiter can have no characters (an empty
delimiter), one character, or more than one character with the following effects:
v An empty delimiter (a pair of quotation marks) parses dynamic.array so that each character becomes
one element of array (see the second example). The default delimiter is a field mark. This is different
from the empty delimiter. To use the default delimiter, omit the comma or USING following
dynamic.array.
v A single character delimiter parses dynamic.array into fields delimited by that character by storing the
substrings that are between successive delimiters as elements in the array. The delimiters are not stored
in the array (see the first example).
v A multicharacter delimiter parses dynamic.array by storing as elements both the substrings that are
between any two successive delimiters and the substrings consisting of one or more consecutive
delimiters in the following way: dynamic.array is searched until any of the delimiter characters are
found. All of the characters up to but not including the delimiter character are stored as an element of
array. The delimiter character and any identical consecutive delimiter characters are stored as the next
element. The search then continues as at the start of dynamic.array (see the third example).
v If delimiter is a system delimiter and a single CHAR(128) is extracted from dynamic.array, the
corresponding element in array is set to the null value.
The characters in a multicharacter delimiter expression can be different or the same. A delimiter
expression of /: might be used to separate hours, minutes, seconds and month, day, year in the formats
12:32:16 and 1/23/85. A delimiter expression of two spaces ″ ″ might be used to separate tokens on a
command line that contain multiple blanks between tokens.
The SETTING clause sets the variable elements to the number of elements in array. If array overflows,
elements is set to 0. The value of elements is the same as the value returned by the INMAT function after a
MATPARSE statement.
If all the elements of array are filled before MATPARSE reaches the end of dynamic.array, MATPARSE puts
the unprocessed part of dynamic.array in the zero element of array for IDEAL, INFORMATION, or
PIOPEN flavor accounts, or in the last element of array for PICK, IN2, or REALITY flavor accounts.
Use the INMAT function after a MATPARSE statement to determine the number of elements loaded into
the array. If there are more delimited fields in dynamic.array than elements in array, INMAT returns 0;
otherwise, it returns the number of elements loaded.
If start is greater than end or greater than the length of array, no action is taken, and INMAT returns 0.
If start, end, dynamic.array, or delimiter evaluates to the null value, the MATPARSE statement fails and the
program terminates with a run-time error message.
MATREAD statements
Syntax
MATREAD array FROM [file.variable,] record.ID [ON ERROR
statements]
{THEN statements [ELSE statements] | ELSE statements}
{ MATREADL | MATREADU } array FROM [file.variable,]
record.ID
[ON ERROR statements] [LOCKED statements]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the MATREAD statement to assign the contents of the fields of a record from a WebSphere DataStage
file to consecutive elements of array. The first field of the record becomes the first element of array, the
second field of the record becomes the second element of array, and so on. The array must be named and
dimensioned in a DIMENSION statement or COMMON statement before it is used in this statement.
file.variable specifies an open file. If file.variable is not specified, the default file is assumed (for more
information about default files, see the OPEN statement). If the file is neither accessible nor open, the
program terminates with a run-time error message.
If record.ID exists, array is set to the contents of the record, and the THEN statements are executed; any
ELSE statements are ignored. If no THEN statements are specified, program execution continues with the
next sequential statement. If record.ID does not exist, the elements of array are not changed, and the ELSE
statements are executed; any THEN statements are ignored.
If the file is an SQL table, the effective user of the program must have SQL SELECT privilege to read
records in the file. For information about the effective user of a program, see the AUTHORIZATION
statement.
A MATREAD statement does not set an update record lock on the specified record. That is, the record
remains available for update to other users. To prevent other users from updating the record until it is
released, use a MATREADL or MATREADU statement.
If the number of elements in array is greater than the number of fields in the record, the extra elements in
array are assigned empty string values. If the number of fields in the record is greater than the number of
elements in the array, the extra values are stored in the zero element of array for IDEAL or
INFORMATION flavor accounts, or in the last element of array for PICK, IN2, or REALITY flavor
accounts. The zero element of an array can be accessed with a 0 subscript as follows:
MATRIX (0)
or:
MATRIX (0, 0)
Use the INMAT function after a MATREAD statement to determine the number of elements of the array
that were actually used. If the number of fields in the record is greater than the number of elements in
the array, the value of the INMAT function is set to 0.
If NLS is enabled, MATREAD and other BASIC statements that perform I/O operations always map
external data to the internal character set using the appropriate map for the input file. For details, see the
statement.
The ON ERROR clause is optional in MATREAD statements. Its syntax is the same as that of the ELSE
clause. The ON ERROR clause lets you specify an alternative for program termination when a fatal error
is encountered during processing of the MATREAD statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
The LOCKED clause is optional, but recommended. Its syntax is the same as that of the ELSE clause.
The LOCKED clause handles a condition caused by a conflicting lock (set by another user) that prevents
the MATREAD statement from processing. The LOCKED clause is executed if one of the following
conflicting locks exists:
Table 45. Conflicting Locks
In this statement... This requested lock... Conflicts with these locks...
MATREADL Shared record lock Exclusive file lock, Update record
lock
MATREADU Update record lock Exclusive file lock, Intent file lock,
Shared file lock, Update record lock,
Shared record lock
If a MATREAD statement does not include a LOCKED clause, and a conflicting lock exists, the program
pauses until the lock is released.
If a LOCKED clause is used, the value returned by the STATUS function is the terminal number of the
user who owns the conflicting lock.
Releasing Locks
A shared record lock can be released with a CLOSE statement, RELEASE statement, or STOP statement.
An update record lock can be released with a CLOSE, DELETE, MATWRITE statements, RELEASE, STOP,
, or statement.
Locks acquired or promoted within a transaction are not released when the previous statements are
processed.
Use the MATREADL syntax to acquire a shared record lock and then perform a MATREAD. This lets
other programs read the record with no lock or a shared record lock.
Use the MATREADU syntax to acquire an update record lock and then perform a MATREAD. The
update record lock prevents other users from updating the record until the user who set it releases it.
An update record lock can be acquired when no shared record lock exists, or promoted from a shared
record lock owned by you if no other shared record locks exist.
Example
DIM ARRAY(10)
OPEN ’SUN.MEMBER’ TO SUN.MEMBER ELSE STOP
MATREAD ARRAY FROM SUN.MEMBER, 6100 ELSE STOP
*
FOR X=1 TO 10
PRINT "ARRAY(":X:")",ARRAY(X)
NEXT X
*
PRINT
*
DIM TEST(4)
OPEN ’’,’SUN.SPORT’ ELSE STOP ’CANNOT OPEN SUN.SPORT’
MATREAD TEST FROM 851000 ELSE STOP
MATREADL statement
Use the MATREADL statement to set a shared record lock and perform the MATREAD statement. For
details, see the MATREAD statements.
MATREADU statement
Use the MATREADU statement to set an update record lock and perform the MATREAD statement. For
details, see the MATREAD statements statement.
MATWRITE statements
Syntax
MATWRITE[U] array ON | TO [file.variable,] record.ID[ON ERROR statements] [LOCKED statements]
[THEN statements] [ELSE statements]
Description
Use the MATWRITE statement to write data from the elements of a dimensioned array to a record in a
WebSphere DataStage file. The elements of array replace any data stored in the record. MATWRITE strips
any trailing empty fields from the record.
file.variable specifies an open file. If file.variable is not specified, the default file is assumed (for more
information on default files, see the OPEN statement). If the file is neither accessible nor open, the
program terminates with a run-time message, unless ELSE statements are specified.
If the file is an SQL table, the effective user of the program must have SQL INSERT and UPDATE
privileges to read records in the file. For information about the effective user of a program, see the
AUTHORIZATION statement.
If the OPENCHK configurable parameter is set to TRUE, or if the file is opened with the OPENCHECK
statement, all SQL integrity constraints are checked for every MATWRITE to an SQL table. If an integrity
check fails, the MATWRITE statement uses the ELSE clause. Use the ICHECK function to determine what
specific integrity constraint caused the failure.
If NLS is enabled, MATWRITE and other BASIC statements that perform I/O operations always map
internal data to the external character set using the appropriate map for the output file. For details, see
the WRITE statements statement. For more information about maps, see WebSphere DataStage National
Language Support Guide.
The ON ERROR clause is optional in the MATWRITE statement. The ON ERROR clause lets you specify
an alternative for program termination when a fatal error is encountered while the MATWRITE is being
processed.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
The LOCKED clause handles a condition caused by a conflicting lock (set by another user) that prevents
the MATWRITE statement from processing. The LOCKED clause is executed if one of the following
conflicting locks exists:
v Exclusive file lock
v Intent file lock
v Shared file lock
v Update record lock
v Shared record lock
If the MATWRITE statement does not include a LOCKED clause, and a conflicting lock exists, the
program pauses until the lock is released.
When updating a file, MATWRITE releases the update record lock set with a MATREADU statement. To
maintain the update record lock set with the MATREADU statement, use MATWRITEU instead of
MATWRITE.
The new values are written to the record, and the THEN clauses are executed. If no THEN statements are
specified, execution continues with the statement following the MATWRITE statement.
Use the MATWRITEU statement to update a record without releasing the update record lock set by a
previous MATREADU statement (see the MATREAD statements statement). To release the update record
lock set by a MATREADU statement and maintained by a MATWRITEU statement, you must use a
RELEASE or MATWRITE statement. If you do not explicitly release the lock, the record remains locked
until the program executes the STOP statement. When more than one program or user could modify the
same record, use a MATREADU statement to lock the record before doing the MATWRITE or
MATWRITEU.
In IDEAL and INFORMATION flavor accounts, if the zero element of the array has been assigned a value
by a MATREAD or MATREADU statement, the zero element value is written to the record as the n+1
field, where n is the number of elements dimensioned in the array. If the zero element is assigned an
empty string, only the assigned elements of the array are written to the record; trailing empty fields are
ignored. The new record is written to the file (replacing any existing record) without regard for the size
of the array.
It is generally good practice to use the MATWRITE statement with arrays that have been loaded with
either a MATREAD or a MATREADU statement.
After executing a MATWRITE statement, you can use the STATUS function to determine the result of the
operation as follows (see the STATUS function for more information):
0 The record was locked before the MATWRITE operation.
-2 The record was unlocked before the MATWRITE operation.
-3 The record failed an SQL integrity check.
Example
DIM ARRAY(5)
OPEN ’EX.BASIC’ TO EX.BASIC ELSE STOP ’CANNOT OPEN’
MATREADU ARRAY FROM EX.BASIC, ’ABS’ ELSE STOP
ARRAY(1)=’Y = 100’
MATWRITE ARRAY TO EX.BASIC, ’ABS’
PRINT ’STATUS()= ’,STATUS()
MATWRITEU statement
Use the MATWRITEU statement to maintain an update record lock and perform the MATWRITE
statement. For details, see the MATWRITE statements statement.
MAXIMUM function
Syntax
MAXIMUM (dynamic.array)
CALL !MAXIMUM (result, dynamic.array)
Use the MAXIMUM function to return the element with the highest numeric value in dynamic.array.
Nonnumeric values, except the null value, are treated as 0. If dynamic.array evaluates to the null value,
null is returned. Any element that is the null value is ignored, unless all elements of dynamic.array are
null, in which case null is returned.
result is the variable that contains the largest element found in dynamic.array.
Examples
A=1:@VM:"ZERO":@SM:20:@FM:-25
PRINT "MAX(A)=",MAXIMUM(A)
In the following example, the !MAXIMUM subroutine is used to obtain the maximum value contained in
array A. The nonnumeric value, Z, is treated as 0.
A=1:@VM:25:@VM:’Z’:@VM:7
CALL !MAXIMUM (RESULT,A)
PRINT RESULT
MINIMUM function
Syntax
MINIMUM (dynamic.array)
CALL !MINIMUM (result, dynamic.array)
Description
Use the MINIMUM function to return the element with the lowest numeric value in dynamic.array.
Nonnumeric values, except the null value, are treated as 0. If dynamic.array evaluates to the null value,
null is returned. Any element that is the null value is ignored, unless all elements of dynamic.array are
null, in which case null is returned.
result is the variable that contains the smallest element found in dyamic.array.
Examples
A=1:@VM:"ZERO":@SM:20:@FM:-25
PRINT "MIN(A)=",MINIMUM(A)
In the following example, the !MINIMUM subroutine is used to obtain the minimum value contained in
array A. The nonnumeric value, Q, is treated as 0.
A=2:@VM:19:@VM:6:@VM:’Q’
CALL !MINIMUM (RESULT,A)
PRINT RESULT
MOD function
Syntax
MOD (dividend, divisor)
Description
Use the MOD function to calculate the value of the remainder after integer division is performed on the
dividend expression by the divisor expression.
The MOD function calculates the remainder using the following formula:
MOD (X, Y) = X - (INT (X / Y) * Y)
dividend and divisor can evaluate to any numeric value, except that divisor cannot be 0. If divisor is 0, a
division by 0 warning message is printed, and 0 is returned. If either dividend or divisor evaluates to the
null value, null is returned.
Example
X=85; Y=3
PRINT ’MOD (X,Y)= ’,MOD (X,Y)
MODS function
Syntax
MODS (array1, array2)
CALL -MODS (return.array, array1, array2)
CALL !MODS (return.array, array1, array2)
Description
Use the MODS function to create a dynamic array of the remainder after the integer division of
corresponding elements of two dynamic arrays.
The MODS function calculates each element according to the following formula:
XY.element = X - (INT (X / Y) * Y)
X is an element of array1 and Y is the corresponding element of array2. The resulting element is returned
in the corresponding element of a new dynamic array. If an element of one dynamic array has no
corresponding element in the other dynamic array, 0 is returned. If an element of array2 is 0, 0 is
returned. If either of a corresponding pair of elements is the null value, null is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A=3:@VM:7
B=2:@SM:7:@VM:4
PRINT MODS(A,B)
MULS function
Syntax
MULS (array1, array2)
CALL -MULS (return.array, array1, array2)
CALL !MULS (return.array, array1, array2)
Description
Use the MULS function to create a dynamic array of the element-by-element multiplication of two
dynamic arrays.
Each element of array1 is multiplied by the corresponding element of array2 with the result being
returned in the corresponding element of a new dynamic array. If an element of one dynamic array has
no corresponding element in the other dynamic array, 0 is returned. If either of a corresponding pair of
elements is the null value, null is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A=1:@VM:2:@VM:3:@SM:4
B=4:@VM:5:@VM:6:@VM:9
PRINT MULS(A,B)
NAP statement
Syntax
NAP [milliseconds]
Description
Use the NAP statement to suspend the execution of a BASIC program, pausing for a specified number of
milliseconds.
milliseconds is an expression evaluating to the number of milliseconds for the pause. If milliseconds is not
specified, a value of 1 is used. If milliseconds evaluates to the null value, the NAP statement is ignored.
NEG function
Syntax
NEG (number)
Description
Use the NEG function to return the arithmetic inverse of the value of the argument.
In the following example, A is assigned the value of 10, and B is assigned the value of NEG(A), which
evaluates to -10:
A = 10
B = NEG(A)
NEGS function
Syntax
NEGS (dynamic.array)
CALL -NEGS (return.array, dynamic.array)
Description
Use the NEGS function to return the negative values of all the elements in a dynamic array. If the value
of an element is negative, the returned value is positive. If dynamic.array evaluates to the null value, null
is returned. If any element is null, null is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
NES function
Syntax
NES (array1, array2)
CALL -NES (return.array, array1, array2)
CALL !NES (return.array, array1, array2)
Description
Use the NES function to test if elements of one dynamic array are equal to the elements of another
dynamic array.
Each element of array1 is compared with the corresponding element of array2. If the two elements are
equal, a 0 is returned in the corresponding element of a new dynamic array. If the two elements are not
equal, a 1 is returned. If an element of one dynamic array has no corresponding element in the other
dynamic array, a 1 is returned. If either of a corresponding pair of elements is the null value, null is
returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
NEXT statement
Syntax
NEXT [variable]
Description
Use the NEXT statement to end a FOR...NEXT loop, causing the program to branch back to the FOR
statement and execute the statements that follow it.
Each FOR statement must have exactly one corresponding NEXT statement.
variable is the name of the variable given as the index counter in the FOR statement. If the variable is not
named, the most recently named index counter variable is assumed.
NOBUF statement
Syntax
NOBUF file.variable {THEN statements [ELSE statements] | ELSE
statements}
Description
Use the NOBUF statement to turn off buffering for a file previously opened for sequential processing.
Normally WebSphere DataStage uses buffering for sequential input and output operations. The NOBUF
statement turns off this buffering and causes all writes to the file to be performed immediately. It
eliminates the need for FLUSH operations but also eliminates the benefits of buffering. The NOBUF
statement must be executed after a successful OPENSEQ statement or CREATE statement and before any
input or output operation is performed on the record.
If the NOBUF operation is successful, the THEN statements are executed; the ELSE statements are
ignored. If THEN statements are not present, program execution continues with the next statement.
If the specified file cannot be accessed or does not exist, the ELSE statements are executed; the THEN
statements are ignored. If file.variable evaluates to the null value, the NOBUF statement fails and the
program terminates with a run-time error message.
Example
In the following example, if RECORD1 in FILE.E can be opened, buffering is turned off:
OPENSEQ ’FILE.E’, ’RECORD1’ TO DATA THEN NOBUF DATA
ELSE ABORT
NOT function
Syntax
NOT (expression)
Description
Use the NOT function to return the logical complement of the value of expression. If the value of
expression is true, the NOT function returns a value of false (0). If the value of expression is false, the NOT
function returns a value of true (1).
A numeric expression that evaluates to 0 is a logical value of false. A numeric expression that evaluates to
anything else, other than the null value, is a logical true.
An empty string is logically false. All other string expressions, including strings that include an empty
string, spaces, or the number 0 and spaces, are logically true.
NOTS function
Syntax
NOTS (dynamic.array)
CALL -NOTS (return.array, dynamic.array)
CALL !NOTS (return.array, dynamic.array)
Description
Use the NOTS function to return a dynamic array of the logical complements of each element of
dynamic.array. If the value of the element is true, the NOTS function returns a value of false (0) in the
corresponding element of the returned array. If the value of the element is false, the NOTS function
returns a value of true (1) in the corresponding element of the returned array.
A numeric expression that evaluates to 0 has a logical value of false. A numeric expression that evaluates
to anything else, other than the null value, is a logical true.
An empty string is logically false. All other string expressions, including strings which consist of an
empty string, spaces, or the number 0 and spaces, are logically true.
If any element in dynamic.array is the null value, null is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
X=5; Y=5
PRINT NOTS(X-Y:@VM:X+Y)
NULL statement
Syntax
NULL
Description
Use the NULL statement when a statement is required but no operation is to be performed. For example,
you can use it with the ELSE clause if you do not want any operation performed when the ELSE clause
is executed.
NUM function
Syntax
NUM (expression)
Description
Use the NUM function to determine whether expression is a numeric or nonnumeric string. If expression is
a number, a numeric string, or an empty string, it evaluates to true and a value of 1 is returned. If
expression is a nonnumeric string, it evaluates to false and a value of 0 is returned.
A string that contains a period used as a decimal point ( . ) evaluates to numeric. A string that contains
any other character used in formatting numeric or monetary amounts, for example, a comma ( , ) or a
dollar sign ( $ ) evaluates to nonnumeric.
If NLS is enabled, NUM uses the Numeric category of the current locale to determine the decimal
separator. For more information about locales, see WebSphere DataStage National Language Support Guide.
Example
X=NUM(2400)
Y=NUM("Section 4")
PRINT "X= ",X,"Y= ",Y
NUMS function
Syntax
NUMS (dynamic.array)
CALL -NUMS (return.array, dynamic.array)
CALL !NUMS (return.array, dynamic.array)
Description
Use the NUMS function to determine whether the elements of a dynamic array are numeric or
nonnumeric strings. If an element is numeric, a numeric string, or an empty string, it evaluates to true,
and a value of 1 is returned to the corresponding element in a new dynamic array. If the element is a
nonnumeric string, it evaluates to false, and a value of 0 is returned.
The NUMS of a numeric element with a decimal point ( . ) evaluates to true; the NUMS of a numeric
element with a comma ( , ) or dollar sign ( $ ) evaluates to false.
If dynamic.array evaluates to the null value, null is returned. If an element of dynamic.array is null, null is
returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
OCONV function
Syntax
OCONV (string, conversion)
Description
Use the OCONV function to convert string to a specified format for external output. The result is always
a string expression.
conversion must evaluate to one or more conversion codes separated by value marks (ASCII 253).
If multiple codes are used, they are applied from left to right as follows: the leftmost conversion code is
applied to string, the next conversion code to the right is then applied to the result of the first conversion,
and so on.
If string evaluates to the null value, null is returned. If conversion evaluates to the null value, the OCONV
function fails and the program terminates with a run-time error message.
For information about converting strings to an internal format, see the ICONV function.
Examples
OCONVS function
Syntax
OCONVS (dynamic.array, conversion)
CALL -OCONVS (return.array, dynamic.array, conversion)
CALL !OCONVS (return.array, dynamic.array, conversion)
Description
Use the OCONVS function to convert the elements of dynamic.array to a specified format for external
output.
The elements are converted to the external output format specified by conversion and returned in a
dynamic array. conversion must evaluate to one or more conversion codes separated by value marks
(ASCII 253).
If multiple codes are used, they are applied from left to right as follows: the leftmost conversion code is
applied to the element, the next conversion code to the right is then applied to the result of the first
conversion, and so on.
If dynamic.array evaluates to the null value, null is returned. If any element of dynamic.array is null, null is
returned for that element. If conversion evaluates to the null value, the OCONVS function fails and the
program terminates with a run-time error message.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
For information about converting elements in a dynamic array to an internal format, see the ICONVS
function.
ON statement
Syntax
ON expression GOSUB statement.label [ : ] [,statement.label [ : ] ... ]
ON expression GO[TO] statement.label [ : ] [,statement.label [ : ] ... ]
Description
Use the ON statement to transfer program control to one of the internal subroutines named in the
GOSUB clause or to one of the statements named in the GOTO clause.
Use ON GOSUB to transfer program control to one of the internal subroutines named in the GOSUB
clause. The value of expression in the ON clause determines which of the subroutines named in the
GOSUB clause is to be executed.
During execution, expression is evaluated and rounded to an integer. If the value of expression is 1 or less
than 1, the first subroutine named in the GOSUB clause is executed; if the value of expression is 2, the
second subroutine is executed; and so on. If the value of expression is greater than the number of
subroutines named in the GOSUB clause, the last subroutine is executed. If expression evaluates to the null
value, the ON statement fails and the program terminates with a run-time error message.
statement.label can be any valid label defined in the program. If a nonexistent statement label is given, an
error message is issued when the program is compiled. You must use commas to separate statement
labels. You can use a colon with the statement labels to distinguish them from variable names.
A RETURN statement in the subroutine returns program flow to the statement following the ON GOSUB
statement.
The ON GOSUB statement can be written on more than one line. A comma is required at the end of each
line of the ON GOSUB statement except the last.
If the value of expression is less than 1, the next statement is executed; if the value of expression is greater
than the number of subroutines named in the GOSUB clause, execution continues with the next statement
rather than the last subroutine. To get this characteristic in other flavors, use the ONGO.RANGE option
of the $OPTIONS statement.
Use ON GOTO to transfer program control to one of the statements named in the GOTO clause. The
value of expression in the ON clause determines which of the statements named in the GOTO clause is to
be executed. During execution, expression is evaluated and rounded to an integer.
If the value of expression is 1 or less than 1, control is passed to the first statement label named in the
GOTO clause; if the value of expression is 2, control is passed to the second statement label; and so on. If
the value of expression is greater than the number of statements named in the GOTO clause, control is
passed to the last statement label. If expression evaluates to the null value, the ON statement fails and the
program terminates with a run-time error message.
statement.label can be any valid label defined in the program. If a nonexistent statement label is given, an
error message is issued when the program is compiled. You must use commas to separate statement
labels. You can use a colon with the statement labels to distinguish them from variable names.
If the value of expression is less than 1, control is passed to the next statement; if the value of expression is
greater than the number of the statements named in the GOTO clause, execution continues with the next
statement rather than the last statement label. To get this characteristic with other flavors, use the
ONGO.RANGE option of the $OPTIONS statement.
Examples
Source Lines
Program Output
OPEN statement
Syntax
OPEN [dict,] filename [TO file.variable] [ON ERROR statements]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the OPEN statement to open a WebSphere DataStage file for use by BASIC programs. All file
references in a BASIC program must be preceded by either an OPEN statement or an OPENCHECK
statement for that file. You can open several WebSphere DataStage files at the same point in a program,
but you must use a separate OPEN statement for each file.
dict is an expression that evaluates to a string specifying whether to open the file dictionary or the data
file. Use the string DICT to open the file dictionary, or use PDICT to open an associated Pick-style
dictionary. Any other string opens the data file. By convention an empty string or the string DATA is
used when you are opening the data file. If the dict expression is omitted, the data file is opened. If dict is
the null value, the data file is opened.
filename is an expression that evaluates to the name of the file to be opened. If the file exists, the file is
opened, and the THEN statements are executed; the ELSE statements are ignored. If no THEN statements
are specified, program execution continues with the next statement. If the file cannot be accessed or does
not exist, the ELSE statements are executed; any THEN statements are ignored. If filename evaluates to the
null value, the OPEN statement fails and the program terminates with a run-time error message.
Use the TO clause to assign the opened file to file.variable. All statements that read, write to, delete, or
clear that file must refer to it by the name of the file variable. If you do not assign the file to a file
variable, an internal default file variable is used. File references that do not specify a file variable access
the default file variable, which contains the most recently opened file. The file opened to the current
default file variable is assigned to the system variable @STDFIL.
Default file variables are not local to the program from which they are executed. When a subroutine is
called, the current default file variable is shared with the calling program.
When opening an SQL table, the OPEN statement enforces SQL security. The permissions granted to the
program’s effective user ID are loaded when the file is opened. If no permissions have been granted, the
OPEN statement fails, and the ELSE statements are executed.
Use the INMAT function after an OPEN statement to determine the modulo of the file.
The ON ERROR clause is optional in the OPEN statement. Its syntax is the same as that of the ELSE
clause. The ON ERROR clause lets you specify an alternative for program termination when a fatal error
is encountered while the OPEN statement is being processed.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
The file type is returned if the file is opened successfully. If the file is not opened successfully, the
following values might return:
Value Description
-1 Filename not found in the VOC file.
1
-2 Null filename or file.
-3 Operating system access error that occurs when you do not have permission to access a
WebSphere DataStage file in a directory. For example, this might occur when trying to access a
type 1 or type 30 file.
-41 Access error when you do not have operating system permissions or if DATA.30 is missing for a
type 30 file.
-5 Read error detected by the operating system.
-6 Unable to lock file header.
-7 Invalid file revision or wrong byte-ordering for the platform.
-81 Invalid part file information.
1
-9 Invalid type 30 file information in a distributed file.
-10 A problem occurred while the file was being rolled forward during warm-start recovery.
Therefore, the file is marked ″inconsistent.″
-11 The file is a view, therefore it cannot be opened by a BASIC program.
Examples
OPEN "SUN.MEMBER" TO DATA ELSE STOP "CAN’T OPEN
SUN.MEMBER"
OPEN "FOOBAR" TO FOO ELSE STOP "CAN’T OPEN FOOBAR"
PRINT "ALL FILES OPEN OK"
The following example opens the same file as in the previous example. The OPEN statement includes an
empty string for the dict argument.
OPEN "","SUN.MEMBER" TO DATA ELSE STOP "CAN’T OPEN
SUN.MEMBER"
OPEN "","FOO.BAR" TO FOO ELSE STOP "CAN’T OPEN
FOOBAR"
PRINT "ALL FILES OPEN OK"
OPENCHECK statement
Syntax
OPENCHECK [dict,] filename [TO file.variable]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the OPENCHECK statement to open an SQL table for use by BASIC programs, enforcing SQL
integrity checking. All file references in a BASIC program must be preceded by either an OPENCHECK
statement or an OPEN statement for that file.
The OPENCHECK statement works like the OPEN statement, except that SQL integrity checking is
enabled if the file is an SQL table. All field integrity checks for an SQL table are stored in the security
and integrity constraints area (SICA). The OPENCHECK statement loads the compiled form of these
integrity checks into memory, associating them with the file variable. All writes to the file are subject to
SQL integrity checking.
The file type is returned if the file is opened successfully. If the file is not opened successfully, the
following values might return:
Value Description
-1 Filename not found in the VOC file.
-21 Null filename or file.
-3 Operating system access error that occurs when you do not have permission to access a
WebSphere DataStage file in a directory. For example, this might occur when trying to access a
type 1 or type 30 file.
-41 Access error when you do not have operating system permissions or if DATA.30 is missing for a
type 30 file.
OPENDEV statement
Syntax
OPENDEV device TO file.variable [LOCKED statements]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the OPENDEV statement to open a device for sequential processing. OPENDEV also sets a record
lock on the opened device or file. See the READSEQ statement and WRITESEQ statements for more
details on sequential processing.
device is an expression that evaluates to the record ID of a device definition record in the &DEVICE& file.
If device evaluates to the null value, the OPENDEV statement fails and the program terminates with a
run-time error message. For more information, see ″Devices on Windows NT.″
The TO clause assigns a file.variable to the device being opened. All statements used to read to or write
from that device must refer to it by the assigned file.variable.
If the device exists and is not locked, the device is opened and any THEN statements are executed; the
ELSE statements are ignored. If no THEN statements are specified, program execution continues with the
next statement.
If the device is locked, the LOCKED statements are executed; THEN statements and ELSE statements are
ignored.
If the device does not exist or cannot be opened, the ELSE statements are executed; any THEN statements
are ignored. The device must have the proper access permissions for you to open it.
If NLS is enabled, you can use OPENDEV to open a device that uses a map defined in the &DEVICE&
file. If there is no map defined in the &DEVICE& file, the default mapname is the name in the
NLSDEFDEVMAP parameter in the uvconfig file. For more information about maps, see WebSphere
DataStage National Language Support Guide.
On Windows NT systems, you might need to change to block size defined for a device in the &DEVICE&
file before you can use OPENDEV to reference the device. On some devices there are limits to the type of
sequential processing that is available once you open the device. The following table summarizes the
limits:
Table 46. Limits to the Type of Sequential Processing with Opened Devices
Device Type Block Size Processing Available
4 mm DAT drive No change needed. No limits.
8 mm DAT drive No change needed. No limits.
1/4-inch cartridge drive, 60 MB or Specify the block size as 512 bytes or Use READBLK and WRITEBLK to
150 MB a multiple of 512 bytes. read or write data in blocks of 512
bytes. Use SEEK only to move the
file pointer to the beginning or the
end of the file. You can use WEOF to
write an end-of-file (EOF) mark only
at the beginning of the data or after a
write.
1/4-inch 525 cartridge drive No change needed. No limits.
Diskette drive Specify the block size as 512 bytes or Use SEEK only to move the file
a multiple of 512 bytes. pointer to the beginning of the file.
Do not use WEOF.
The LOCKED clause handles a condition caused by a conflicting lock (set by another user) that prevents
the OPENDEV statement from processing. The LOCKED clause is executed if one of the following
conflicting locks exists:
v Exclusive file lock
v Intent file lock
v Shared file lock
v Update record lock
v Shared record lock
If the OPENDEV statement does not include a LOCKED clause, and a conflicting lock exists, the program
pauses until the lock is released.
Example
The following example opens TTY30 for sequential input and output operations:
OPENDEV ’TTY30’ TO TERM THEN PRINT ’TTY30 OPENED’
ELSE ABORT
Description
The OPENPATH statement is similar to the OPEN statement, except that the path name of the file is
specified. This file is opened without reference to the VOC file. The file must be a hashed WebSphere
DataStage file or a directory (types 1 and 19).
pathname specifies the relative or absolute path name of the file to be opened. If the file exists, it is
opened and the THEN statements are executed; the ELSE statements are ignored. If pathname evaluates to
the null value, the OPENPATH statement fails and the program terminates with a run-time error
message.
If the file cannot be accessed or does not exist, the ELSE statements are executed; any THEN statements
are ignored.
Use the TO clause to assign the file to a file.variable. All statements used to read, write, delete, or clear
that file must refer to it by the assigned file.variable name. If you do not assign the file to a file.variable, an
internal default file variable is used. File references that do not specify file.variable access the most recently
opened default file. The file opened to the default file variable is assigned to the system variable
@STDFIL.
The ON ERROR clause is optional in the OPENPATH statement. Its syntax is the same as that of the
ELSE clause. The ON ERROR clause lets you specify an alternative for program termination when a fatal
error is encountered during processing of the OPENPATH statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
You can use the STATUS function after an OPENPATH statement to find the cause of a file open failure
(that is, for an OPENPATH statement in which the ELSE clause is used). The following values can be
returned if the OPENPATH statement is unsuccessful:
Value Description
Example
The following example opens the file SUN.MEMBER. The path name specifies the file.
OPENPATH ’/user/members/SUN.MEMBER’ ELSE ABORT
OPENSEQ statement
Syntax
OPENSEQ filename, record.ID TO file.variable [USING
dynamic.array]
[ON ERROR statements] [LOCKED statements]
{THEN statements [ELSE statements] | ELSE statements}
OPENSEQ pathname TO file.variable [USING dynamic.array]
[ON ERROR statements] [LOCKED statements]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the OPENSEQ statement to open a file for sequential processing. All sequential file references in a
BASIC program must be preceded by an OPENSEQ or OPENDEV statement for that file. Although you
can open several files for sequential processing at the same point in the program, you must issue a
separate OPENSEQ statement for each. See the READSEQ statement and WRITESEQ statements for more
details on sequential processing.
Note: Performing multiple OPENSEQ operations on the same file results in creating only one update
record lock. This single lock can be released by a CLOSESEQ or RELEASE statement.
filename specifies the name of the type 1 or type 19 file containing the record to be opened.
record.ID specifies the record in the file to be opened. If the record exists and is not locked, the file is
opened and the THEN statements are executed; the ELSE statements are ignored. If no THEN statements
are specified, program execution continues with the next statement. If the record or the file itself cannot
be accessed or does not exist, the ELSE statements are executed; any THEN statements are ignored.
pathname is an explicit path name for the file, pipe, or device to be opened. If the file exists and is not
locked, it is opened and the THEN statements are executed; the ELSE statements are ignored. If the path
name does not exist, the ELSE statements are executed; any THEN statements are ignored.
If the file does not exist, the OPENSEQ statement fails. The file can also be explicitly created with the
CREATE statement.
OPENSEQ sets an update record lock on the specified record or file. This lock is reset by a CLOSESEQ
statement. This prevents any other program from changing the record while you are processing it.
If filename, record.ID, or pathname evaluate to the null value, the OPENSEQ statement fails and the
program terminates with a run-time error message.
The TO clause is required. It assigns the record, file, or device to file.variable. All statements used to
sequentially read, write, delete, or clear that file must refer to it by the assigned file variable name.
If NLS is enabled, you can use the OPENSEQ filename, record.ID statement to open a type 1 or type 19 file
that uses a map defined in the .uvnlsmap file in the directory containing the type 1 or type 19 file. If there
is no .uvnlsmap file in the directory, the default mapname is the name in the NLSDEFDIRMAP parameter
in the uvconfig file.
Use the OPENSEQ pathname statement to open a UNIX pipe, file, or a file specified by a device that uses
a map defined in the .uvnlsmap file in the directory holding pathname. If there is no .uvnlsmap file in the
directory, the default mapname is the name in the NLSDEFSEQMAP parameter in the uvconfig file, or you
can use the SET.SEQ.MAP command to assign a map.
For more information about maps, see WebSphere DataStage National Language Support Guide.
File Buffering
Normally WebSphere DataStage uses buffering for sequential input and output operations. Use the
NOBUF statement after an OPENSEQ statement to turn off buffering and cause all writes to the file to be
performed immediately. For more information about file buffering, see the NOBUF statement.
You can optionally include the USING clause to control whether the opened file is included in the
rotating file pool. The USING clause supplements OPENSEQ processing with a dynamic array whose
structure emulates an &DEVICE& file record. Field 17 of the dynamic array controls inclusion in the
rotating file pool with the following values:
v Y removes the opened file.
v N includes the opened file.
The ON ERROR clause is optional in the OPENSEQ statement. Its syntax is the same as that of the ELSE
clause. The ON ERROR clause lets you specify an alternative for program termination when a fatal error
is encountered while the OPENSEQ statement is being processed.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
The LOCKED clause is optional, but recommended. Its syntax is the same as that of the ELSE clause. The
LOCKED clause handles a condition caused by a conflicting lock (set by another user) that prevents the
OPENSEQ statement from processing. The LOCKED clause is executed if one of the following conflicting
locks exists:
v Exclusive file lock
v Intent file lock
v Shared file lock
v Update record lock
v Shared record lock
If the OPENSEQ statement does not include a LOCKED clause, and a conflicting lock exists, the program
pauses until the lock is released.
Use the STATUS function after an OPENSEQ statement to determine whether the file was successfully
opened.
The file type is returned if the file is opened successfully. If the file is not opened successfully, the
following values might return:
Value Description
-1 Filename not found in the VOC file.
1
-2 Null filename or file.
-3 Operating system access error that occurs when you do not have privileges to access a
WebSphere DataStage file in a directory. For example, this might occur when trying to access a
type 1 or type 30 file.
Examples
The following example reads RECORD1 from the file FILE.E, which is not hashed:
OPENSEQ ’FILE.E’, ’RECORD1’ TO FILE THEN
PRINT "’FILE.E’ OPENED FOR PROCESSING"
END ELSE ABORT
READSEQ A FROM FILE THEN PRINT A ELSE STOP
The next example writes the record read from FILE.E to the file /usr/depta/file1:
OPENSEQ ’/usr/depta/file1’ TO OUTPUT THEN
PRINT "usr/depta/file1 OPENED FOR PROCESSING"
END ELSE ABORT
WRITESEQ A ON OUTPUT ELSE PRINT "CANNOT WRITE TO
OUTPUT"
.
.
.
CLOSESEQ FILE
CLOSESEQ OUTPUT
END
The next example includes the USING clause to remove an opened file from the rotating file pool:
DEVREC = "1"@FM
FOR I = 2 TO 16
DEVREC = DEVREC:I:@FM
NEXT I
DEVREC=DEVREC:’Y’
*
OPENSEQ ’SEQTEST’, ’TESTDATA’ TO TESTFILE USING
ORS function
Syntax
ORS (array1, array2)
CALL -ORS (return.array, array1, array2)
CALL !ORS (return.array, array1, array2)
Description
Use the ORS function to create a dynamic array of the logical OR of corresponding elements of two
dynamic arrays.
Each element of the new dynamic array is the logical OR of the corresponding elements of array1 and
array2. If an element of one dynamic array has no corresponding element in the other dynamic array, a
false is assumed for the missing element.
If both corresponding elements of array1 and array2 are the null value, null is returned for those elements.
If one element is the null value and the other is 0 or an empty string, null is returned. If one element is
the null value and the other is any value other than 0 or an empty string, a true is returned.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A="A":@SM:0:@VM:4:@SM:1
B=0:@SM:1-1:@VM:2
PRINT ORS(A,B)
PAGE statement
Syntax
PAGE [ ON print.channel ] [ page# ]
Description
Use the PAGE statement to print headings, footings, and page advances at the appropriate places on the
specified output device. You can specify headings and footings before execution of the PAGE statement
(see the HEADING statement and FOOTING statements). If there is no heading or footing, PAGE clears
the screen.
The ON clause specifies the logical print channel to use for output. print.channel is an expression that
evaluates to a number from -1 through 255. If you do not use the ON clause, logical print channel 0 is
used, which prints to the user’s terminal if PRINTER OFF is set (see the PRINTER statement). Logical
print channel -1 prints the data on the screen, regardless of whether a PRINTER ON statement has been
executed.
If either print.channel or page# evaluates to the null value, the PAGE statement fails and the program
terminates with a run-time error message.
Example
In the following example the current value of X provides the next page number:
PAGE ON 5 X
PERFORM statement
Syntax
PERFORM command
Description
Use the PERFORM statement to execute a WebSphere DataStage sentence, paragraph, menu, or command
from within the BASIC program, then return execution to the statement following the PERFORM
statement. The commands are executed in the same environment as the BASIC program that called them;
that is, unnamed common variables, @variables, and in-line prompts retain their values, and select lists
and the DATA stack remain active. If these values change, the new values are passed back to the calling
program.
You can specify multiple commands in the PERFORM statement in the same way you specify them in the
body of a WebSphere DataStage paragraph. Each command or line must be separated by a field mark
(ASCII CHAR(254)).
If command evaluates to the null value, the PERFORM statement fails and the program terminates with a
run-time error message.
You cannot use the PERFORM statement within a transaction to execute most WebSphere DataStage
commands and SQL statements. However, you can use PERFORM to execute the following WebSphere
DataStageWebSphere DataStage commands and SQL statements within a transaction:
CHECK.SUM INSERT SEARCH SSELECT COUNT LIST SELECT (RetrieVe) STAT DELETE (SQL)
LIST.ITEM SELECT (SQL) SUM DISPLAY LIST.LABEL SORT UPDATE ESEARCH RUN SORT.ITEM
GET.LIST SAVE.LIST SORT.LABEL
REALITY Flavor
In a REALITY flavor account PERFORM can take all the clauses of the EXECUTE statement. To get these
PERFORM characteristics in other flavor accounts, use the PERF.EQ.EXEC option of the $OPTIONS
statement.
Example
PRECISION statement
Syntax
PRECISION expression
Description
Use the PRECISION statement to control the maximum number of decimal places that are output when
the system converts a numeric value from internal binary format to an ASCII character string value.
expression specifies a number from 0 through 9. Any fractional digits in the result of such a conversion
that exceed the precision setting are rounded off.
If you do not include a PRECISION statement, a default precision of 4 is assumed. Precisions are stacked
so that a BASIC program can change its precision and call a subroutine whose precision is the default
unless the subroutine executes a PRECISION statement. When the subroutine returns to the calling
program, the calling program has the same precision it had when it called the subroutine.
Trailing fractional zeros are dropped during output. Therefore, when an internal number is converted to
an ASCII string, the result might appear to have fewer decimal places than the precision setting allows.
However, regardless of the precision setting, the calculation always reflects the maximum accuracy of
which the computer is capable (that is, slightly more than 17 total digits, including integers).
If expression evaluates to the null value, the PRECISION statement fails and the program terminates with
a run-time error message.
Example
A = 12.123456789
PRECISION 8
PRINT A
PRECISION 4
PRINT A
PRINT statement
Syntax
PRINT [ON print.channel] [print.list]
Description
Use the PRINT statement to send data to the screen, a line printer, or another print file.
The ON clause specifies the logical print channel to use for output. print.channel is an expression that
evaluates to a number from -1 through 255. If you do not use the ON clause, logical print channel 0 is
used, which prints to the user’s terminal if PRINTER OFF is set (see the PRINTER statement). If
You can specify HEADING statement, FOOTING statement, PAGE statement, and PRINTER statement
CLOSE statements for each logical print channel. The contents of the print files are printed in order by
logical print channel number.
print.list can contain any BASIC expression. The elements of the list can be numeric or character strings,
variables, constants, or literal strings; the null value, however, cannot be printed. The list can consist of a
single expression or a series of expressions separated by commas ( , ) or colons ( : ) for output formatting.
If no print.list is designated, a blank line is printed.
Expressions separated by commas are printed at preset tab positions. The default tab stop setting is 10
characters. Calculations for tab characters are based on character length rather than display length. For
information about changing the default setting, see the TABSTOP statement. Use multiple commas
together for multiple tabulations between expressions.
Expressions separated by colons are concatenated. That is, the expression following the colon is printed
immediately after the expression preceding the colon. To print a list without a LINEFEED and RETURN,
end print.list with a colon ( : ).
If NLS is enabled, calculations for the PRINT statement are based on character length rather than display
length. If print.channel has a map associated with it, data is mapped before it is output to the device. For
more information about maps, see WebSphere DataStage National Language Support Guide.
Examples
A=25;B=30
C="ABCDE"
PRINT A+B
PRINT
PRINT "ALPHA ":C
PRINT "DATE ":PRINT "10/11/93"
*
PRINT ON 1 "FILE 1"
* The string "FILE 1" is printed on print file 1.
The following example prints the letter X at location column 10, row 5:
PRINT @(10,5) ’X’
PRINTER statement
Syntax
PRINTER { ON | OFF | RESET }
PRINTER CLOSE [ON print.channel]
Use the PRINTER statement to direct output either to the screen or to a printer. By default, all output is
sent to the screen unless a PRINTER ON is executed or the P option to the RUN command is used. See
the SETPTR command for more details about redirecting output.
PRINTER ON sends output to the system line printer via print channel 0. The output is stored in a buffer
until a PRINTER CLOSE statement is executed or the program terminates; the output is then printed (see
the PRINTER statement CLOSE statement).
PRINTER OFF sends output to the screen via print channel 0. When the program is executed, the data is
immediately printed on the screen.
The PRINTER ON or PRINTER OFF statement must precede the PRINT statement that starts the print
file.
Use the PRINTER RESET statement to reset the printing options. PRINTER RESET removes the header
and footer, resets the page count to 1, resets the line count to 1, and restarts page waiting.
Note: Use TPRINT statement to set a delay before printing. See also the TPARM function statement.
Use the PRINTER CLOSE statement to print all output data stored in the printer buffer.
You can specify print channel -1 through 255 with the ON clause. If you omit the ON clause from a
PRINTER CLOSE statement, print channel 0 is closed. Only data directed to the printer specified by the
ON clause is printed. Therefore, there must be a corresponding PRINTER CLOSE ON print.channel for
each ON clause specified in a PRINT statement. All print channels are closed when the program stops.
Logical print channel -1 prints the data on the screen, regardless of whether a PRINTER ON statement
has been executed.
If print.channel evaluates to the null value, the PRINTER CLOSE statement fails and the program
terminates with a run-time error message.
In PICK, IN2, and REALITY flavor accounts, the PRINTER CLOSE statement closes all print channels.
Example
PRINTER ON
PRINT "OUTPUT IS PRINTED ON PRINT FILE 0"
PRINTER OFF
PRINT "OUTPUT IS PRINTED ON THE TERMINAL"
*
PRINT ON 1 "OUTPUT WILL BE PRINTED ON PRINT FILE 1"
PRINT ON 2 "OUTPUT WILL BE PRINTED ON PRINT FILE 2"
PRINTERR statement
Syntax
PRINTERR [error.message]
Use the PRINTERR statement to print a formatted error message on the bottom line of the terminal. The
message is cleared by the next INPUT statement @ statement or is overwritten by the next PRINTERR or
INPUTERR statement. PRINTERR clears the type-ahead buffer.
error.message is an expression that evaluates to the error message text. The elements of the expression can
be numeric or character strings, variables, constants, or literal strings. The null value cannot be an
element because it cannot be output. The expression can be a single expression or a series of expressions
separated by commas ( , ) or colons ( : ) for output formatting. If no error message is designated, a blank
line is printed. If error.message evaluates to the null value, the default message is printed:
Message ID is NULL: undefined error
Expressions separated by commas are printed at preset tab positions. The default tab stop setting is 10
characters. For information about changing the default setting, see the TABSTOP statement. Use multiple
commas together to cause multiple tabulations between expressions.
Expressions separated by colons are concatenated: that is, the expression following the colon is printed
immediately after the expression preceding the colon.
REALITY Flavor
In a REALITY flavor account the PRINTERR statement prints a formatted error message from the
ERRMSG file on the bottom line of the terminal. REALITY syntax is:
PRINTERR [dynamic.array] [FROM file.variable]
dynamic.array must contain a record ID and any arguments to the message, with each element separated
from the next by a field mark. If dynamic.array does not specify an existing record ID, a warning message
states that no error message can be found.
If dynamic.array evaluates to the null value, the default error message is printed:
Message ID is NULL: undefined error
The FROM clause lets you read the error message from an open file. If file.variable evaluates to the null
value, the PRINTERR statement fails and the program terminates with a run-time error message.
This statement is similar to the STOP statement on a Pick system except that it does not terminate the
program upon execution. You can use it wherever you can use a STOP or ABORT statement.
To use the REALITY version of the PRINTERR statement in PICK, IN2, INFORMATION, and IDEAL
flavor accounts, use the USE.ERRMSG option of the $OPTIONS statement.
WebSphere DataStage provides a standard Pick ERRMSG file. You can construct a local ERRMSG file
using the following syntax in the records. Each field must start with one of these codes, as shown in the
following table:
Table 47. ERRMSG File Codes
Code Action
A[(n)] Display next argument left-justified; n specifies field
length.
D Display system date.
PROCREAD statement
Syntax
PROCREAD variable{THEN statements [ELSE statements] | ELSE statements}
Description
Use the PROCREAD statement to assign the contents of the primary input buffer to a variable. Your
BASIC program must be called by a procedure. If your program was not called from a procedure, the
ELSE statements are executed; otherwise the THEN statements are executed.
If variable evaluates to the null value, the PROCREAD statement fails and the program terminates with a
run-time error message.
PROCWRITE statement
Syntax
PROCWRITE string
Description
Use the PROCWRITE statement to write string to the primary input buffer. Your program must be called
by a procedure.
If string evaluates to the null value, the PROCWRITE statement fails and the program terminates with a
run-time error message.
PROGRAM statement
Syntax
PROG[RAM] [name]
Description
Use the PROGRAM statement to identify a program. The PROGRAM statement is optional; if you use it,
it must be the first line in the program that is not a comment line.
name can be specified for documentation purposes; it need not be the same as the actual program name.
Example
PROGRAM BYSTATE
Description
Use the PROMPT statement to specify the character to be displayed on the screen when user input is
required. If no PROMPT statement is issued, the default prompt character is the question mark ( ? ).
If character evaluates to more than one character, only the first character is significant; all others are
ignored.
The prompt character becomes character when the PROMPT statement is executed. Although the value of
character can change throughout the program, the prompt character remains the same until a new
PROMPT statement is issued or the program ends.
Generally, data the user enters in response to the prompt appears on the screen. If the source of the input
is something other than the keyboard (for example, a DATA statement), the data is displayed on the
screen after the prompt character. Use PROMPT ″ ″ to prevent any prompt from being displayed.
PROMPT ″ ″ also suppresses the display of input from DATA statements.
Examples
Source Lines
Program Output
PROMPT ″HELLO″ PRINT ″ENTER ANSWER ″: INPUT ANS PROMPT ″-″ PRINT ″ENTER ANSWER
″: INPUT ANS PROMPT ″″ PRINT ″ENTER ANSWER ″: INPUT ANS END
ENTER ANSWER HANSWERENTER ANSWER -YESENTER ANSWER NO
PWR function
Syntax
PWR (expression, power)
Description
Use the PWR function to return the value of expression raised to the power specified by power.
The PWR function operates like exponentiation (that is, PWR(X,Y) is the same as X**Y).
A negative value cannot be raised to a power that is not represented by an integer. If it is, the result of
the function is PWR(-X,Y) and an error message is displayed.
Example
A=3
B=PWR(5,A)
PRINT "B= ",B
QUOTE function
Syntax
QUOTE (expression)
Description
Use the QUOTE function to enclose an expression in double quotation marks. If expression evaluates to
the null value, null is returned, without quotation marks.
Example
PRINT QUOTE(12 + 5) : " IS THE ANSWER."
END
RAISE function
Syntax
RAISE (expression)
Description
Use the RAISE function to return a value equal to expression, except that system delimiters in expression
are converted to the next higher-level delimiter: value marks are changed to field marks, subordinate
value marks are changed to value marks, and so on. If expression evaluates to the null value, null is
returned.
PIOPEN Flavor
In PIOPEN flavor, the delimiters that can be raised are CHAR(255) through CHAR(252). All other
characters are left unchanged. You can obtain PIOPEN flavor for the RAISE function by:
v Compiling your program in a PIOPEN flavor account
v Specifying the $OPTIONS INFO.MARKS statement
In the following examples an item mark is shown by I, a field mark is shown by F, a value mark is
shown by V, and a subordinate value mark is shown by S.
RANDOMIZE statement
Syntax
RANDOMIZE (expression)
Description
Use the RANDOMIZE statement with an expression to make the RND function generate the same
sequence of random numbers each time the program is run. If no expression is supplied, or if expression
evaluates to the null value, the internal time of day is used (the null value is ignored). In these cases the
sequence is different each time the program is run.
Example
RANDOMIZE (0)
FOR N=1 TO 10
PRINT RND(4):’ ’:
NEXT N
PRINT
*
RANDOMIZE (0)
FOR N=1 TO 10
PRINT RND(4):’ ’:
NEXT
PRINT
*
RANDOMIZE (3)
FOR N=1 TO 10
PRINT RND(4):’ ’:
NEXT N
PRINT
READ statements
Syntax
READ dynamic.array FROM [file.variable,] record.ID [ON ERROR
statements]
{THEN statements [ELSE statements] | ELSE statements}
{READL | READU} dynamic.array FROM [file.variable,] record.ID [ON ERROR statements] [LOCKED statements]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use READ statements to assign the contents of a record from a WebSphere DataStage file to
dynamic.array.
Table 49. Uses for READ Statements
Use this statement... To do this...
READ Read a record.
READL Acquire a shared record lock and read a record.
READU Acquire an update record lock and read a record.
READV Read a field.
READVL Acquire a shared record lock and read a field.
READVU Acquire an update record lock and read a field.
file.variable specifies an open file. If file.variable is not specified, the default file is assumed (for more
information on default files, see the OPEN statement). If the file is neither accessible nor open, the
program terminates with a run-time error message.
If record.ID exists on the specified file, dynamic.array is set to the contents of the record, and the THEN
statements are executed; any ELSE statements are ignored. If no THEN statements are specified, program
execution continues with the next statement. If record.ID does not exist, dynamic.array is set to an empty
string, and the ELSE statements are executed; any THEN statements are ignored.
If file.variable, record.ID, or field# evaluate to the null value, the READ statement fails and the program
terminates with a run-time error message.
Tables
If the file is a table, the effective user of the program must have SQL SELECT privilege to read records in
the file. For information about the effective user of a program, see the AUTHORIZATION statement.
Distributed Files
If the file is a distributed file, use the STATUS function after a READ statement to determine the results
of the operation, as follows:
-1 The partitioning algorithm does not evaluate to an integer.
-2 The part number is invalid.
NLS Mode
If NLS is enabled, READ and other BASIC statements that perform I/O operations map external data to
the internal character set using the appropriate map for the input file.
If the file contains characters that cannot be mapped, the ELSE statements are executed.
For more information about maps, see WebSphere DataStage National Language Support Guide.
The ON ERROR clause is optional in the READ statement. Its syntax is the same as that of the ELSE
clause. The ON ERROR clause lets you specify an alternative for program termination when a fatal error
is encountered during processing of the READ statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
You can use the LOCKED clause only with the READL, READU, READVL, and READVU statements. Its
syntax is the same as that of the ELSE clause.
The LOCKED clause handles a condition caused by a conflicting lock (set by another user) that prevents
the READ statement from being processed. The LOCKED clause is executed if one of the following
conflicting locks exists:
Table 50. Conditions Causing Execution of LOCKED Clause
In this statement... This requested lock... Conflicts with...
READL READVL Shared record lock Exclusive file lock Update record lock
READU READVU Update record lock Exclusive file lock Intent file lock
Shared file lock Update record lock
Shared record lock
If a READ statement does not include a LOCKED clause, and a conflicting lock exists, the program
pauses until the lock is released.
Releasing Locks
A shared record lock can be released with a CLOSE, RELEASE, or STOP statement. An update record
lock can be released with a CLOSE, DELETE, MATWRITE, RELEASE, STOP, WRITE, or WRITEV
statement.
Locks acquired or promoted within a transaction are not released when the previous statements are
processed.
All record locks are released when you return to the command prompt.
Use the READL syntax to acquire a shared record lock and then read the record. This allows other
programs to read the record with no lock or a shared record lock.
Use the READU statement to acquire an update record lock and then read the record. The update record
lock prevents other users from updating the record until the user who owns it releases it.
An update record lock can only be acquired when no shared record lock exists. It can be promoted from
a shared record lock owned by the user requesting the update record lock if no shared record locks exist.
To prevent more than one program or user from modifying the same record at the same time, use
READU instead of READ.
Use the READV statement to assign the contents of a field in a WebSphere DataStage file record to
dynamic.array.
Use the READVL statement to acquire a shared record lock and then read a field from the record. The
READVL statement conforms to all the specifications of the READL and READV statements.
Use the READVU statement to acquire an update record lock and then read a field from the record. The
READVU statement conforms to all the specifications of the READU and READV statements.
You can specify field# only with the READV, READVL, and READVU statements. It specifies the index
number of the field to be read from the record. You can use a field# of 0 to determine whether the record
exists. If the field does not exist, dynamic.array is assigned the value of an empty string.
In PICK, IN2, and REALITY flavor accounts, if record.ID or field# does not exist, dynamic.array retains its
value and is not set to an empty string. The ELSE statements are executed; any THEN statements are
ignored. To specify PICK, IN2, and REALITY flavor READ statements in an INFORMATION or IDEAL
flavor account, use the READ.RETAIN option of the $OPTIONS statement.
Examples
OPEN ’’,’SUN.MEMBER’ TO FILE ELSE STOP
FOR ID=5000 TO 6000
READ MEMBER FROM FILE, ID THEN PRINT ID ELSE NULL
NEXT ID
OPEN ’’,’SUN.SPORT’ ELSE STOP ’CANT OPEN "SUN.SPORT"’
READ ID FROM "853333" ELSE
The next example locks the record N in the file SUN.MEMBER, reads field 3 (STREET) from it, and prints
the value of the field:
OPEN ’’,’SUN.MEMBER’ TO FILE ELSE STOP
FOR N=5000 TO 6000
READVU STREET FROM FILE,N,3 THEN PRINT STREET ELSE
NULL
RELEASE
NEXT
OPEN "DICT","MYFILE" TO DICT.FILE ELSE STOP
OPEN "","MYFILE" ELSE STOP ; *USING DEFAULT FILE
VARIABLE
READU ID.ITEM FROM DICT.FILE,"@ID" ELSE
PRINT "NO @ID"
STOP
END
1
4 is returned only if the NLSREADELSE parameter is set to 1. If NLSREADELSE is 0, no value is
returned, data is lost, and you see a run-time error message.
READBLK statement
Syntax
READBLK variable FROM file.variable, blocksize{ THEN statements [ ELSE statements ] | ELSE statements }
Description
Use the READBLK statement to read a block of data of a specified length from a file opened for
sequential processing and assign it to a variable. The READBLK statement reads a block of data
beginning at the current position in the file and continuing for blocksize bytes and assigns it to variable.
The current position is reset to just beyond the last byte read.
If the data can be read from the file, the THEN statements are executed; any ELSE statements are
ignored. If the file is not readable or if the end of file is encountered, the ELSE statements are executed
and the THEN statements are ignored. If the ELSE statements are executed, variable is set to an empty
string.
If either file.variable or blocksize evaluates to the null value, the READBLK statement fails and the program
terminates with a run-time error message.
Note: A newline in UNIX files is one byte long, whereas in Windows NT it is two bytes long. This means
that for a file with newlines, the same READBLK statement might return a different set of data
depending on the operating system the file is stored under.
The difference between the READSEQ statement and the READBLK statement is that the READBLK
statement reads a block of data of a specified length, whereas the READSEQ statement reads a single line
of data.
On Windows NT systems, if you use READBLK to read data from a 1/4-inch cartridge drive (60 or 150
MB) that you open with the OPENDEV statement, you must use a block size of 512 bytes or a multiple of
512 bytes.
For more information about sequential file processing, see the OPENSEQ statement, READSEQ statement,
and WRITESEQ statements.
If NLS is enabled and file.variable has a map associated with it, the data is mapped accordingly. For more
information about maps, see WebSphere DataStage National Language Support Guide.
Example
OPENSEQ ’FILE.E’, ’RECORD4’ TO FILE ELSE ABORT
READBLK VAR1 FROM FILE, 15 THEN PRINT VAR1
PRINT
READBLK VAR2 FROM FILE, 15 THEN PRINT VAR2
ND LINE
THIRD L
READL statement
Use the READL statement to acquire a shared record lock and perform the READ statement. For details,
see the statement.
READLIST statement
Syntax
READLIST dynamic.array [FROM list.number]
{ THEN statements [ ELSE statements ] | ELSE statements }
Description
Use the READLIST statement to read the remainder of an active select list into a dynamic array.
list.number is an expression that evaluates to the number of the select list to be read. It can be from 0
through 10. If you do not use the FROM clause, select list 0 is used.
READLIST reads all elements in the active select list. If a READNEXT statement is used on the select list
before the READLIST statement, only the elements not read by the READNEXT statement are stored in
dynamic.array. READLIST empties the select list.
If one or more elements are read from list.number, the THEN statements are executed. If there are no
more elements in the select list or if a select list is not active, the ELSE statements are executed; any
THEN statements are ignored.
In IDEAL and INFORMATION flavor accounts, use the VAR.SELECT option of the $OPTIONS statement
to get READLIST to behave as it does in PICK flavor accounts.
In PICK, REALITY, and IN2 flavor accounts, the READLIST statement has the following syntax:
READLIST dynamic.array FROM listname [SETTING variable]
{THEN statements [ELSE statements] | ELSE statements}
In these flavors the READLIST statement reads a saved select list from the &SAVEDLISTS& file without
activating a select list. In PICK and IN2 flavor accounts, READLIST lets you access a saved select list
without changing the currently active select list if there is one.
The select list saved in listname in the &SAVEDLISTS& file is put in dynamic.array. The elements of the list
are separated by field marks.
or
record.ID account.name
record.ID specifies the record ID of the list in &SAVEDLISTS&, and account.name specifies the name of
another WebSphere DataStage account in which to look for the &SAVEDLISTS& file.
The SETTING clause assigns the count of the elements in the list to variable.
If the list is retrieved successfully (the list must not be empty), the THEN statements are executed; if not,
the ELSE statements are executed. If listname evaluates to the null value, the READLIST statement fails
and the program terminates with a run-time error message.
In PICK, REALITY, and IN2 flavor accounts, use the -VAR.SELECT option of the $OPTIONS statement to
get READLIST to behave as it does in IDEAL flavor accounts.
READNEXT statement
Syntax
READNEXT dynamic.array [ ,value [ ,subvalue ] ] [FROM list]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the READNEXT statement to assign the next record ID from an active select list to dynamic.array.
list specifies the select list. If none is specified, select list 0 is used. list can be a number from 0 through 10
indicating a numbered select list, or the name of a select list variable.
The BASIC SELECT statements statement or the WebSphere DataStage GET.LIST, FORM.LIST, SELECT, or
SSELECT commands create an active select list; these commands build the list of record IDs. The
READNEXT statement reads the next record ID on the list specified in the FROM clause and assigns it to
the dynamic.array.
If list evaluates to the null value, the READNEXT statement fails and the program terminates with a
run-time error message.
A READNEXT statement with value and subvalue specified accesses an exploded select list. The record ID
is stored in dynamic.array, the value number in value, and the subvalue number in subvalue. If only
dynamic.array is specified, it is set to a multi-valued field consisting of the record ID, value number, and
subordinate value number, separated by value marks.
INFORMATION Flavor
In INFORMATION flavor accounts READNEXT returns an exploded select list. Use the RNEXT.EXPL
option of the $OPTIONS statement to return exploded select lists in other flavors.
Example
OPEN ’’,’SUN.MEMBER’ ELSE STOP "CAN’T OPEN FILE"
SELECT TO 1
10: READNEXT MEM FROM 1 THEN PRINT MEM ELSE GOTO 15:
GOTO 10:
*
15: PRINT
OPEN ’’,’SUN.SPORT’ TO FILE ELSE STOP
SELECT FILE
COUNT=0
20*
READNEXT ID ELSE
PRINT ’COUNT= ’,COUNT
STOP
END
COUNT=COUNT+1
GOTO 20
READSEQ statement
Syntax
READSEQ variable FROM file.variable [ON ERROR statements]
{THEN statements [ELSE statements] | ELSE statements}
Use the READSEQ statement to read a line of data from a file opened for sequential processing.
Sequential processing lets you process data one line at a time. WebSphere DataStage keeps a pointer at
the current position in the file. The OPENSEQ statement sets this pointer to the first byte of the file, and
it is advanced by READSEQ, READBLK, WRITESEQ, and WRITEBLK statements.
Each READSEQ statement reads data from the current position in the file up to a newline and assigns it
to variable. The pointer is then set to the position following the newline. The newline is discarded.
file.variable specifies a file previously opened for sequential processing. The FROM clause is required. If
the file is neither accessible nor open, or if file.variable evaluates to the null value, the READSEQ
statement fails and the program terminates with a run-time error message.
If data is read from the file, the THEN statements are executed, and the ELSE statements are ignored. If
the file is not readable, or the end of file is encountered, the ELSE statements are executed; any THEN
statements are ignored.
In the event of a timeout, READSEQ returns no bytes from the buffer, and the entire I/O operation must
be retried.
If NLS is enabled, the READSEQ and other BASIC statements that perform I/O operations always map
external data to the internal character set using the appropriate map for the input file if the file has a
map associated with it. For more information about maps, see WebSphere DataStage National Language
Support Guide.
The ON ERROR clause is optional in the READSEQ statement. Its syntax is the same as that of the ELSE
clause. The ON ERROR clause lets you specify an alternative for program termination when a fatal error
is encountered during processing of the READSEQ statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
READT statement
Syntax
READT [UNIT (mtu)] variable
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the READT statement to read the next tape record from a magnetic tape unit and assign its contents
to a variable.
The UNIT clause specifies the number of the tape drive unit. Tape unit 0 is used if no unit is specified.
mtu is an expression that evaluates to a code made up of three decimal digits, as shown in the following
table:
Table 51. mtu Codes
Code Available Options
m (mode) 0 = No conversion
1 = EBCDIC conversion
The mtu expression is read from right to left. Therefore, if mtu evaluates to a one-digit code, it represents
the tape unit number. If mtu evaluates to a two-digit code, the rightmost digit represents the unit number
and the digit to its left is the track number; and so on.
If either mtu or variable evaluates to the null value, the READT statement fails and the program
terminates with a run-time error message.
Each tape record is read and processed completely before the next record is read. The program waits for
the completion of data transfer from the tape before continuing.
If the next tape record exists, variable is set to the contents of the record, and the THEN statements are
executed. If no THEN statements are specified, program execution continues with the next statement.
The largest tape record that the READT statement can read is system-dependent. If a tape record is larger
than the system maximum, only the bytes up to the maximum are assigned to variable.
The STATUS function returns 1 if READT takes the ELSE clause, otherwise it returns 0.
If NLS is enabled, the READT and other BASIC statements that perform I/O operations always map
external data to the internal character set using the appropriate map for the input file if the file has a
map associated with it. For more information about maps, see WebSphere DataStage National Language
Support Guide.
PIOPEN Flavor
If you have a program that specifies the syntax UNIT ndmtu, the nd elements are ignored by the compiler
and no errors are reported.
Examples
The next example reads a record from tape drive 3, doing an EBCDIC conversion in the process:
READT UNIT(103) RECORD ELSE PRINT "COULD NOT READ"
READU statement
Use the READU statement to set an update record lock and perform the READ statement. For details, see
the statement.
READV statement
Use the READV statement to read the contents of a specified field of a record in a WebSphere DataStage
file. For details, see the statement.
READVL statement
Use the READVL statement to set a shared record lock and perform the READV statement. For details,
see the statement.
READVU statement
Use the READVU statement to set an update record lock and read the contents of a specified field of a
record in a WebSphere DataStage file. For details, see the statement.
REAL function
Syntax
REAL (number)
Use the REAL function to convert number into a floating-point number without loss of accuracy. If number
evaluates to the null value, null is returned.
RECORDLOCK statements
Syntax
RECORDLOCKL file.variable , record.ID [ ON ERROR statements ]
[ LOCKED statements ]
RECORDLOCKU file.variable , record.ID [ ON ERROR statements ]
[ LOCKED statements ]
Description
Use RECORDLOCK statements to acquire a record lock on a record without reading the record.
Use this statement...
To acquire this lock without reading the record...
RECORDLOCKL
Shared record lock
RECORDLOCKU
Update record lock
record.ID is an expression that evaluates to the record ID of the record that is to be locked.
The RECORDLOCKL statement lets other users lock the record using RECORDLOCK or any other
statement that sets a shared record lock, but cannot gain exclusive control over the record with
FILELOCK statement, or any statement that sets an update record lock.
The RECORDLOCKU statement prevents other users from accessing the record using a FILELOCK
statement or any statement that sets either a shared record lock or an update record lock. You can reread
a record after you have locked it; you are not affected by your own locks.
The ON ERROR clause is optional in RECORDLOCK statements. The ON ERROR clause lets you specify
an alternative for program termination when a fatal error is encountered while a RECORDLOCK
statement is being processed.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
The LOCKED clause handles a condition caused by a conflicting lock (set by another user) that prevents
the RECORDLOCK statement from processing. The LOCKED clause is executed if one of the following
conflicting locks exists:
Table 52. Conditions Causing Execution of LOCKED Clause
In this statement... This requested lock... Conflicts with these locks...
RECORDLOCKL Shared record lock Exclusive file lock Update record lock
RECORDLOCKU Update record lock Exclusive file lock Intent file lock
Shared file lock Update record lock
Shared record lock
If the RECORDLOCK statement does not include a LOCKED clause, and a conflicting lock exists, the
program pauses until the lock is released.
If a LOCKED clause is used, the value returned by the STATUS function is the terminal number of the
user who owns the conflicting lock.
Releasing Locks
A shared record lock can be released with a CLOSE statement, RELEASE statement, or STOP statement.
An update record lock can be released with a CLOSE,DELETE, MATWRITE statements, RELEASE, STOP,
, or statement.
Locks acquired or promoted within a transaction are not released when the previous statements are
processed.
All record locks are released when you return to the command prompt.
Example
In the following example, the file EMPLOYEES is opened. Record 23694 is locked. If the record was
already locked, the program terminates, and an appropriate message is displayed. The RECORDLOCKL
statement allows other users to read the record with READL or lock it with another RECORDLOCKL, but
prevents any other user from gaining exclusive control over the record.
OPEN ’’,’EMPLOYEES’ TO EMPLOYEES ELSE STOP ’Cannot
open file’ RECORDLOCKL EMPLOYEES,’23694’
LOCKED STOP ’Record previously locked by user
’:STATUS( )
RECORDLOCKED function
Syntax
RECORDLOCKED ( file.variable , record.ID)
record.ID is an expression that evaluates to the record ID of the record that is to be checked.
An insert file of equate names is provided to let you use mnemonics (see the following table). The insert
file is called RECORDLOCKED.INS.IBAS, and is located in the INCLUDE directory in the WebSphere
DataStage engine directory. In PIOPEN flavor accounts, the VOC file has a file pointer called SYSCOM.
SYSCOM references the INCLUDE directory in the WebSphere DataStage engine directory.
To use the insert file, specify $INCLUDE SYSCOM RECORDLOCKED.INS.IBAS when you compile the
program.
Table 53. RECORDLOCKED.INS.IBAS File Equate Names
Equate Name Value Meaning
LOCK$MY.FILELOCK 3 This user has a FILELOCK.
LOCK$MY.READL 2 This user has a shared record lock.
LOCK$MY.READU 1 This user has an update record lock.
LOCK$NO.LOCK 0 The record is not locked.
LOCK$OTHER.READL -1 Another user has a shared record
lock.
LOCK$OTHER.READU -2 Another user has an update record
lock.
LOCK$OTHER.FILELOCK -3 Another user has a FILELOCK.
If you have locked the file, the RECORDLOCKED function indicates only that you have the file lock for
that record. It does not indicate any update record or shared record lock that you also have on the record.
Possible values returned by the STATUS function, and their meanings, are as follows:
>0 A positive value is the terminal number of the owner of the lock (or the first terminal number
encountered, if more than one user has locked records in the specified file).
<0 A negative value is -1 times the terminal number of the remote user who has locked the record or
file.
Examples
The following program checks to see if there is an update record lock or FILELOCK held by the current
user on the record. If the locks are not held by the user, the ELSE clause reminds the user that an update
record lock or FILELOCK is required on the record. This example using the SYSCOM file pointer, only
works in PI/open flavor accounts.
$INCLUDE SYSCOM RECORDLOCKED.INS.IBAS
OPEN ’’,’EMPLOYEES’ TO EMPLOYEES
ELSE STOP ’CANNOT OPEN FILE’
.
.
.
IF RECORDLOCKED(EMPLOYEES,RECORD.ID) >= LOCK$MY.READU
The next program checks to see if the record lock is held by another user and prints a message where the
STATUS function gives the terminal number of the user who holds the record lock:
$INCLUDE SYSCOM RECORDLOCKED.INS.IBAS
OPEN ’’,’EMPLOYEES’ TO EMPLOYEES
ELSE STOP ’CANNOT OPEN FILE’
.
.
.
IF RECORDLOCKED(EMPLOYEES,RECORD.ID) < LOCK$NO.LOCK
THEN
PRINT ’Record locked by user’ : STATUS()
END
RELEASE statement
Syntax
RELEASE [ file.variable [ ,record.ID ] ] [ ON ERROR statements ]
Description
Use the RELEASE statement to unlock, or release, locks set by a FILELOCK, MATREADL, MATREADU,
READL, READU, READVL, READVU, or OPENSEQ statement. These statements lock designated records
to prevent concurrent updating by other users. If you do not explicitly release a lock that you have set, it
is unlocked automatically when the program terminates.
file.variable specifies an open file. If file.variable is not specified and a record ID is specified, the default file
is assumed (for more information on default files, see the OPEN statement). If the file is neither accessible
nor open, the program terminates with a run-time error message.
record.ID specifies the lock to be released. If it is not specified, all locks in the specified file (that is, either
file.variable or the default file) are released. If either file.variable or record.ID evaluates to the null value, the
RELEASE statement fails and the program terminates with a run-time error message.
When no options are specified, all locks in all files set by any FILELOCK, READL, READU, READVL,
READVU, WRITEU, WRITEVU, MATREADL, MATREADU , MATWRITEU , or OPENSEQ statement
during the current login session are released.
The ON ERROR clause is optional in the RELEASE statement. The ON ERROR clause lets you specify an
alternative for program termination when a fatal error is encountered during processing of the RELEASE
statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
Examples
The following example releases all locks set in all files by the current program:
RELEASE
The next example releases all locks set in the NAMES file:
RELEASE NAMES
The next example releases the lock set on the record QTY in the PARTS file:
RELEASE PARTS, "QTY"
REM function
Syntax
REM (dividend, divisor)
Description
Use the REM function to calculate the remainder after integer division is performed on the dividend
expression by the divisor expression.
The REM function calculates the remainder using the following formula:
REM (X, Y) = X - (INT (X / Y) * Y)
dividend and divisor can evaluate to any numeric value, except that divisor cannot be 0. If divisor is 0, a
division by 0 warning message is printed, and 0 is returned. If either dividend or divisor evaluates to the
null value, null is returned.
Example
X=85; Y=3
PRINT ’REM (X,Y)= ’,REM (X,Y)
REM statement
Syntax
REM [comment.text]
Description
Use the REM statement to insert a comment in a BASIC program. Comments explain or document
various parts of a program. They are part of the source code only and are nonexecutable. They do not
affect the size of the object code.
Any text that appears between a comment designator and the end of a physical line is treated as part of
the comment. If a comment does not fit on one physical line, it can be continued on the next physical line
only by starting the new line with a comment designator. If a comment appears at the end of a physical
line containing an executable statement, you must treat it as if it were a new statement and put a
semicolon ( ; ) after the executable statement, before the comment designator.
Example
PRINT "HI THERE"; REM This part is a comment.
REM This is also a comment and does not print.
REM
IF 5<6 THEN PRINT "YES"; REM A comment; PRINT "PRINT
ME"
REM BASIC thinks PRINT "PRINT ME" is also part
REM of the comment.
IF 5<6 THEN
PRINT "YES"; REM Now it doesn’t.
PRINT "PRINT ME"
END
REMOVE function
Syntax
REMOVE (dynamic.array, variable)
Description
Use the REMOVE function to successively extract and return dynamic array elements that are separated
by system delimiters, and to indicate which system delimiter was found. When a system delimiter is
encountered, the value of the extracted element is returned. The REMOVE function is more efficient than
the EXTRACT function for extracting successive fields, values, and so on, for multi-value list processing.
variable is set to a code corresponding to the system delimiter which terminates the extracted element.
The contents of variable indicate which system delimiter was found, as follows:
0 End of string
1 Item mark ASCII CHAR(255)
2 Field mark ASCII CHAR(254)
3 Value mark ASCII CHAR(253)
4 Subvalue mark ASCII CHAR(252)
5 Text mark ASCII CHAR(251)
6 ASCII CHAR(250) (Not available in the PIOPEN flavor)
7 ASCII CHAR(249) (Not available in the PIOPEN flavor)
The REMOVE function extracts one element each time it is executed, beginning with the first element in
dynamic.array. The operation can be repeated until all elements of dynamic.array are extracted. The
REMOVE function does not change the dynamic array.
As each successive element is extracted from dynamic.array, a pointer associated with dynamic.array is set
to the beginning of the next element to be extracted. Thus the pointer is advanced every time the
REMOVE function is executed.
The pointer is reset to the beginning of dynamic.array whenever dynamic.array is reassigned. Therefore,
dynamic.array should not be assigned a new value until all elements have been extracted (that is, until
variable is 0).
If dynamic.array evaluates to the null value, null is returned and variable is set to 0 (end of string). If an
element in dynamic.array is the null value, null is returned for that element, and variable is set to the
appropriate delimiter code.
Unlike the EXTRACT function, the REMOVE function maintains a pointer into the dynamic array. (The
EXTRACT function always starts processing at the beginning of the dynamic array, counting field marks,
value marks, and subvalue marks until it finds the correct element to extract.)
See the REMOVE statement for the statement equivalent of this function.
Examples
The first example sets the variable FIRST to the string MIKE and the variable X to 2 (field mark). The
second example executes the REMOVE function and PRINT statement until all the elements have been
extracted, at which point A = 0. Printed lines are 12, 4, 5, 7654, and 00.
Source Lines
Program Output
FM=CHAR(254) NAME=’MIKE’:FM:’JOHN’:FM X=REMOVE(NAME,FIRST) PRINT ’FIRST = ’:FIRST,
’X = ’:X
FIRST = 2 X = MIKE
VM=CHAR(253) A = 1 Z=12:VM:4:VM:5:VM:7654:VM:00 FOR X=1 TO 20 UNTIL A=0 A =
REMOVE(Z,Y) PRINT ’Y = ’:Y, ’A = ’:A NEXT X
Y = 3 A = 12
Y = 3 A = 4
Y = 3 A = 5
Y = 3 A = 7654
Y = 0 A = 0
REMOVE statement
Syntax
REMOVE element FROM dynamic.array SETTING variable
Description
Use the REMOVE statement to successively extract dynamic array elements that are separated by system
delimiters. When a system delimiter is encountered, the extracted element is assigned to element. The
REMOVE statement is more efficient than the EXTRACT function for extracting successive fields, values,
and so on, for multi-value list processing.
The REMOVE statement extracts one element each time it is executed, beginning with the first element in
dynamic.array. The operation can be repeated until all elements of dynamic.array are extracted. The
REMOVE statement does not change the dynamic array.
As each element is extracted from dynamic.array to element, a pointer associated with dynamic.array is set
to the beginning of the next element to be extracted. Thus, the pointer is advanced every time the
REMOVE statement is executed.
The pointer is reset to the beginning of dynamic.array whenever dynamic.array is reassigned. Therefore,
dynamic.array should not be assigned a new value until all elements have been extracted (that is, until
variable = 0).
If an element in dynamic.array is the null value, null is returned for that element.
Unlike the EXTRACT function, the REMOVE statement maintains a pointer into the dynamic array. (The
EXTRACT function always starts processing at the beginning of the dynamic array, counting field marks,
value marks, and subvalue marks until it finds the correct element to extract.)
See the REMOVE function for the function equivalent of this statement.
Examples
The first example sets the variable FIRST to the string MIKE and the variable X to 2 (field mark). The
second example executes the REMOVE and PRINT statements until all the elements have been extracted,
at which point A = 0. Printed lines are 12, 4, 5, 7654, and 00.
Source Lines
Program Output
FM=CHAR(254) NAME=’MIKE’:FM:’JOHN’:FM REMOVE FIRST FROM NAME SETTING X PRINT
’X= ’:X, ’FIRST= ’:FIRST
X= 2 FIRST= MIKE
VM=CHAR(253) A=1 Z=12:VM:4:VM:5:VM:7654:VM:00 FOR X=1 TO 20 UNTIL A=0 REMOVE Y
FROM Z SETTING A PRINT ’Y= ’:Y, ’A= ’:A NEXT X
Y= 12 A= 3
Y= 4 A= 3
Y= 5 A= 3
Y= 7654 A= 3
Y= 0 A= 0
REPLACE function
Syntax
REPLACE (expression, field#, value#, subvalue# { , | ; } replacement)
REPLACE (expression [ ,field# [ ,value#] ] ; replacement)
variable < field# [ ,value# [ ,subvalue#] ] >
Description
Use the REPLACE function to return a copy of a dynamic array with the specified field, value, or
subvalue replaced with new data.
The expressions field#, value#, and subvalue# specify the type and position of the element to be replaced.
These expressions are called delimiter expressions.
The value# and subvalue# are optional. However, if either subvalue# or both value# and subvalue# are
omitted, a semicolon ( ; ) must precede replacement, as shown in the second syntax.
You can use angle brackets to replace data in dynamic arrays. Angle brackets to the left of an assignment
operator change the specified data in the dynamic array according to the assignment operator. Angle
brackets to the right of an assignment operator indicate that an EXTRACT function is to be performed
(for examples, see the EXTRACT function).
The three possible results of delimiter expressions are described as case 1, case 2, and case 3.
Case 1:
Both value# and subvalue# are omitted or are specified as 0. A field is replaced by the value of
replacement.
v If field# is positive and less than or equal to the number of fields in the dynamic array, the field
specified by field# is replaced by the value of replacement.
v If field# is negative, a new field is created by appending a field mark and the value of
replacement to the last field in the dynamic array.
v If field# is positive and greater than the number of fields in the dynamic array, a new field is
created by appending the proper number of field marks, followed by the value of replacement;
thus, the value of field# is the number of the new field.
Case 2:
subvalue# is omitted or is specified as 0, and value# is nonzero. A value in the specified field is
replaced with the value of replacement.
v If value# is positive and less than or equal to the number of values in the field, the value
specified by the value# is replaced by the value of replacement.
v If value# is negative, a new value is created by appending a value mark and the value of
replacement to the last value in the field.
In IDEAL, PICK, PIOPEN, and REALITY flavor accounts, if replacement is an empty string and an attempt
is made to append the new element to the end of the dynamic array, field, or value, the dynamic array,
field, or value is left unchanged; additional delimiters are not appended. Use the EXTRA.DELIM option
of the $OPTIONS statement to make the REPLACE function append a delimiter to the dynamic array,
field, or value.
If replacement is the null value, the stored representation of null (CHAR(128)) is inserted into
dynamic.array. If dynamic.array evaluates to the null value, it remains unchanged by the replacement. If the
REPLACE statement references a subordinate element of an element whose value is the null value, the
dynamic array is unchanged.
In INFORMATION and IN2 flavor accounts, if expression is an empty string and the new element is
appended to the end of the dynamic array, the end of a field, or the end of a value, a delimiter is
appended to the dynamic array, field, or value. Use the -EXTRA.DELIM option of the $OPTIONS
statement to make the REPLACE function work as it does in IDEAL, PICK, and REALITY flavor
accounts.
Examples
In the following examples a field mark is shown by F, a value mark is shown by V, and a subvalue mark
is shown by S.
The next example replaces the first subvalue of the third value in field 2 with # and sets Q to
FAVBV#SEFDFFF:
Q=R
Q<2,3,1>="#"
The next example replaces the first value in fields 1 through 4 with # and sets Q to #F#VBVDSEF#F#FF:
The next example appends a value mark and # to the last value in field 2 and sets Q to
FAVBVDSEV#FDFFF:
Q=R
Q=REPLACE(Q,2,-1;"#")
RETURN statement
Syntax
RETURN [TO statement.label]
Description
Use the RETURN statement to terminate a subroutine and return control to the calling program or
statement.
If the TO clause is not specified, the RETURN statement exits either an internal subroutine called by a
GOSUB statement or an external subroutine called by a CALL statement. Control returns to the statement
that immediately follows the CALL or GOSUB statement.
Use a RETURN statement to terminate an internal subroutine called with a GOSUB statement to ensure
that the program proceeds in the proper sequence.
Use a RETURN statement or an END statement to terminate an external subroutine called with a CALL
statement. When you exit an external subroutine called by CALL, all files opened by the subroutine are
closed, except files that are open to common variables.
Use the TO clause to exit only an internal subroutine; control passes to the specified statement label. If
you use the TO clause and statement.label does not exist, an error message appears when the program is
compiled.
Note: Using the TO clause can make program debugging and modification extremely difficult. Be careful
when you use the RETURN TO statement, because all other GOSUBs or CALLs active at the time the
GOSUB is executed remain active, and errors can result.
If the RETURN or RETURN TO statement does not have a place to return to, control is passed to the
calling program or to the command language.
Example
In the following example, subroutine XYZ prints the message ″THIS IS THE EXTERNAL SUBROUTINE″
and returns to the main program:
20: GOSUB 80:
25: PRINT "THIS LINE WILL NOT PRINT"
30: PRINT "HI THERE"
40: CALL XYZ
60: PRINT "BACK IN MAIN PROGRAM"
70: STOP
80: PRINT "THIS IS THE INTERNAL SUBROUTINE"
90: RETURN TO 30:
END
Description
Use the RETURN (value) statement to return a value from a user-written function.
expression evaluates to the value you want the user-written function to return. If you use a RETURN
(value) statement in a user-written function and you do not specify expression, an empty string is returned
by default.
You can use the RETURN (value) statement only in user-written functions. If you use one in a program or
subroutine, an error message appears.
REUSE function
Syntax
REUSE (expression)
Description
Use the REUSE function to specify that the value of the last field, value, or subvalue be reused in a
dynamic array operation.
expression is either a dynamic array or an expression whose value is considered to be a dynamic array.
When a dynamic array operation processes two dynamic arrays in parallel, the operation is always
performed on corresponding subordinate values. This is true even for corresponding fields, each of which
contains a single value. This single value is treated as the first and only subvalue in the first and only
value in the field.
A dynamic array operation isolates the corresponding fields, values, and subordinate values in a dynamic
array. It then operates on them in the following order:
1. The subordinate values in the values
2. The values in the fields
3. The fields of each dynamic array
A dynamic array operation without the REUSE function adds zeros or empty strings to the shorter array
until the two arrays are equal. (The DIVS function is an exception. If a divisor element is absent, the
divisor array is padded with ones, so that the dividend value is returned.)
The REUSE function reuses the last value in the shorter array until all elements in the longer array are
exhausted or until the next higher delimiter is encountered.
After all subordinate values in a pair of corresponding values are processed, the dynamic array operation
isolates the next pair of corresponding values in the corresponding fields and repeats the procedure.
If expression evaluates to the null value, the null value is replicated, and null is returned for each
corresponding element.
Example
B = (1:@SM:6:@VM:10:@SM:11)
A = ADDS(REUSE(5),B)
PRINT "REUSE(5) + 1:@SM:6:@VM:10:@SM:11 = ": A
*
PRINT "REUSE(1:@SM:2) + REUSE(10:@VM:20:@SM:30) = ":
PRINT ADDS(REUSE(1:@SM:2),REUSE(10:@VM:20:@SM:30))
*
PRINT "(4:@SM:7:@SM:8:@VM:10)*REUSE(10) = ":
PRINT MULS((4:@SM:7:@SM:8:@VM:10 ),REUSE(10))
REVREMOVE statement
Syntax
REVREMOVE element FROM dynamic.array SETTING variable
Description
Use the REVREMOVE statement to successively extract dynamic array elements that are separated by
system delimiters. The elements are extracted from right to left, in the opposite order from those
extracted by the REMOVE statement. When a system delimiter is encountered, the extracted element is
assigned to element.
dynamic.array is an expression that evaluates to the dynamic array from which to extract elements.
variable is set to a code value corresponding to the system delimiter terminating the element just
extracted. The delimiter code settings assigned to variable are as follows:
0 End of string
1 Item mark ASCII CHAR(255)
2 Field mark ASCII CHAR(254)
3 Value mark ASCII CHAR(253)
4 Subvalue mark ASCII CHAR(252)
5 Text mark ASCII CHAR(251)
6 ASCII CHAR(250)
7 ASCII CHAR(249)
8 ASCII CHAR(248)
The REVREMOVE statement extracts one element each time it is executed, beginning with the last
element in dynamic.array. The operation can be repeated until all elements of dynamic.array are extracted.
The REVREMOVE statement does not change the dynamic array.
The pointer is reset to the beginning of dynamic.array whenever dynamic.array is reassigned. Therefore,
dynamic.array should not be assigned a new value until all elements have been extracted (that is, until
variable = 0).
If an element in dynamic.array is the null value, null is returned for that element.
Use REVREMOVE with the REMOVE statement. After a REMOVE statement, REVREMOVE returns the
same string as the preceding REMOVE, setting the pointer to the delimiter preceding the extracted
element. Thus, a subsequent REMOVE statement extracts the same element yet a third time.
If no REMOVE statement has been performed on dynamic.array or if the leftmost dynamic array element
has been returned, element is set to the empty string and variable indicates end of string (that is, 0).
Example
DYN = "THIS":@FM:"HERE":@FM:"STRING"
REMOVE VAR FROM DYN SETTING X
PRINT VAR
REVREMOVE NVAR FROM DYN SETTING X
PRINT NVAR
REMOVE CVAR FROM DYN SETTING X
PRINT CVAR
REWIND statement
Syntax
REWIND [UNIT (mtu)]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the REWIND statement to rewind a magnetic tape to the beginning-of-tape position.
The UNIT clause specifies the number of the tape drive unit. Tape unit 0 is used if no unit is specified. If
the UNIT clause is used, mtu is an expression that evaluates to a code made up of three decimal digits.
Although the mtu expression is a function of the UNIT clause, the REWIND statement uses only the third
digit (the u). Its value must be in the range of 0 through 7. If mtu evaluates to the null value, the
REWIND statement fails and the program terminates with a run-time error message.
Before a REWIND statement is executed, a tape drive unit must be attached to the user. Use the ASSIGN
command to assign a tape unit to a user. If no tape unit is attached or if the unit specification is incorrect,
the ELSE statements are executed.
The STATUS function returns 1 if REWIND takes the ELSE clause, otherwise it returns 0.
PIOPEN Flavor
If you have a program that specifies the syntax UNIT ndmtu, the nd elements are ignored by the compiler
and no errors are reported.
RIGHT function
Syntax
RIGHT (string, n)
Description
Use the RIGHT function to extract a substring comprising the last n characters of a string. It is equivalent
to the following substring extraction operation:
string [ length ]
If you use this function, you need not calculate the string length.
If string evaluates to the null value, null is returned. If n evaluates to the null value, the RIGHT function
fails and the program terminates with a run-time error message.
Example
PRINT RIGHT("ABCDEFGH",3)
RND function
Syntax
RND (expression)
Description
Use the RND function to generate any positive or negative random integer or 0.
expression evaluates to the total number of integers, including 0, from which the random number can be
selected. That is, if n is the value of expression, the random number is generated from the numbers 0
through (n - 1).
See the RANDOMIZESyntax statement for details on generating repeatable sequences of random
numbers.
Example
A=20
PRINT RND(A)
PRINT RND(A)
PRINT RND(A)
PRINT RND(A)
ROLLBACK statement
Syntax
ROLLBACK [ WORK ] [ THEN statements ] [ ELSE statements ]
Description
Use the ROLLBACK statement to cancel all file I/O changes made during a transaction. The WORK
keyword provides compatibility with SQL syntax conventions; it is ignored by the compiler.
A transaction includes all statements executed since the most recent BEGIN TRANSACTION statement.
The ROLLBACK statement rolls back all changes made to files during the active transaction. If a
subordinate transaction rolls back, none of the changes resulting from the active subordinate transaction
affect the parent transaction. If the top-level transaction rolls back, none of the changes made are
committed to disk, and the database remains unaffected by the transaction.
Use the ROLLBACK statement in a transaction without a COMMIT statement to review the results of a
possible change. Doing so does not affect the parent transaction or the database. Executing a ROLLBACK
statement ends the current transaction. After the transaction ends, execution continues with the statement
following the next END TRANSACTION statement.
If no transaction is active, the ROLLBACK statement generates a run-time warning, and the ELSE
statements are executed.
Example
This example begins a transaction that applies locks to rec1 and rec2. If errors occur (such as a failed or a
failed WRITE statements), the ROLLBACK statements ensure that no changes are written to the file.
BEGIN TRANSACTION
READU data1 FROM file1,rec1 ELSE ROLLBACK
READU data2 FROM file2,rec2 ELSE ROLLBACK
.
.
.
WRITE new.data1 ON file1,rec1 ELSE ROLLBACK
WRITE new.data2 ON file2,rec2 ELSE ROLLBACK
COMMIT WORK
END TRANSACTION
The update record lock on rec1 is not released on successful completion of the first WRITE statement.
RPC.CALL function
Syntax
RPC.CALL (connection.ID, procedure, #args, MAT arg.list, #values,
MAT return.list)
Description
Use the RPC.CALL function to make requests of a connected server. The request is packaged and sent to
the server using the C client RPC library. RPC.CALL returns the results of processing the remote request:
1 for success, 0 for failure.
arg.list is a two-dimensional array (matrix) containing the input arguments to pass to the RPC server. The
elements of this array represent ordered pairs of values. The first value is the number of the argument to
the server operation, the second value is an argument-type declarator. (Data typing generalizes the RPC
interface to work with servers that are data-type sensitive.)
return.list is a dimensioned array containing the results of the remote operation returned by RPC.CALL.
Like arg.list, the results are ordered pairs of values.
RPC.CALL builds an RPC packet from #args and arg.list. Functions in the C client RPC library transmit
the packet to the server and wait for the server to respond. When a response occurs, the RPC packet is
separated into its elements and stored in the array return.list.
Use the STATUS function after an RPC.CALL function is executed to determine the result of the
operation, as follows:
81001 Connection closed, reason unspecified.
81002 connection.ID does not correspond to a valid bound connection.
81004 Error occurred while trying to store an argument in the transmission packet.
81005 Procedure access denied because of a mismatch of RPC versions.
81008 Error occurred because of a bad parameter in arg.list.
81009 Unspecified RPC error.
81010 #args does not match expected argument count on remote machine.
81015 Timeout occurred while waiting for response from server.
Example
The following example looks for jobs owned by fred. The server connection was made using the
RPC.CONNECT function.
args (1,1) = "fred"; args (1,2) = UVRPC.STRING
IF (RPC.CALL (server.handle, "COUNT.USERS", 1, MAT
args,
return.count, MAT res)) ELSE
PRINT "COUNT.JOBS request failed, error code is: "
STATUS()
GOTO close.connection:
END
RPC.CONNECT function
Syntax
RPC.CONNECT (host, server)
Use the RPC.CONNECT function to establish a connection to a server process. Once the host and server
are identified, the local daemon tries to connect to the remote server. If the attempt succeeds,
RPC.CONNECT returns a connection ID. If it fails, RPC.CONNECT returns 0. The connection ID is a
nonzero integer used to refer to the server in subsequent calls to RPC.CALL function and
RPC.DISCONNECT function.
server is the name, as defined in the remote /etc/services file, of the RPC server class on the target host.
If host is not in the /etc/hosts file, or if server is not in the remote /etc/services file, the connection
attempt fails.
Use the STATUS function after an RPC.CONNECT function is executed to determine the result of the
operation, as follows:
81005 Connection failed because of a mismatch of RPC versions.
81007 Connection refused because the server cannot accept more clients.
81009 Unspecified RPC error.
81011 Host is not in the local /etc/hosts file.
81012 Remote dsrpcd cannot start service because it could not fork the process.
81013 Cannot open the remote dsrpcservices file.
81014 Service not found in the remote dsrpcservices file.
81015 Connection attempt timed out.
Example
RPC.DISCONNECT function
Syntax
RPC.DISCONNECT (connection.ID)
Description
Use the STATUS function after an RPC.DISCONNECT function is executed to determine the result of the
operation, as follows:
81001 The connection was closed, reason unspecified.
81002 connection.ID does not correspond to a valid bound connection.
81009 Unspecified RPC error.
Example
The following example closes the connection to a remote server called MONITOR on HOST.A:
MAT args(1,2), res(1,2)
server.handle = RPC.CONNECT ("HOST.A", "MONITOR")
IF (server.handle = 0) THEN
PRINT "Connection failed, error code is: ":
STATUS()
STOP
END
.
.
.
close.connection:
IF (RPC.DISCONNECT (server.handle)) ELSE
PRINT "Bizarre disconnect error, result code is: "
STATUS()
END
SADD function
Syntax
SADD (string.number.1, string.number.2)
Description
Use the SADD function to add two string numbers and return the result as a string number. You can use
this function in any expression where a string or string number is valid, but not necessarily where a
standard number is valid, because string numbers can exceed the range of numbers that standard
arithmetic operators can handle.
Either string number can evaluate to any valid number or string number.
If either string number contains nonnumeric data, an error message is generated, and 0 replaces the
nonnumeric data. If either string number evaluates to the null value, null is returned.
Example
A = 88888888888888888
B = 77777777777777777
X = "88888888888888888"
Y = "77777777777777777"
PRINT A + B
PRINT SADD(X,Y)
Description
Use the SCMP function to compare two string numbers and return one of the following three numbers: -1
(less than), 0 (equal), or 1 (greater than). If string.number.1 is less than string.number.2, the result is -1. If
they are equal, the result is 0. If string.number.1 is greater than string.number.2, the result is 1. You can use
this function in any expression where a string or string number is valid.
Either string number can be a valid number or string number. Computation is faster with string numbers.
If either string number contains nonnumeric data, an error message is generated and 0 is used instead of
the nonnumeric data. If either string number evaluates to the empty string, null is returned.
Example
X = "123456789"
Y = "123456789"
IF SCMP(X,Y) = 0 THEN PRINT "X is equal to Y"
ELSE PRINT "X is not equal to Y"
END
SDIV function
Syntax
SDIV (string.number.1, string.number.2 [ ,precision])
Description
Use the SDIV function to divide string.number.1 by string.number.2 and return the result as a string
number. You can use this function in any expression where a string or string number is valid, but not
necessarily where a standard number is valid, because string numbers can exceed the range of numbers
which standard arithmetic operators can handle. Either string number can be a valid number or a string
number.
precision specifies the number of places to the right of the decimal point. The default precision is 14.
If either string number contains nonnumeric data, an error message is generated and 0 is used for that
number. If either string number evaluates to the null value, null is returned.
Example
X = "1"
Y = "3"
Z = SDIV (X,Y)
ZZ = SDIV (X,Y,20)
PRINT Z
PRINT ZZ
Description
Use the SEEK statement to move the file pointer by an offset specified in bytes, relative to the current
position, the beginning of the file, or the end of the file.
offset is the number of bytes before or after the reference position. A negative offset results in the pointer
being moved before the position specified by relto. If offset is not specified, 0 is assumed.
Note: On Windows NT systems, line endings in files are denoted by the character sequence RETURN +
LINEFEED rather than the single LINEFEED used in UNIX files. The value of offset should take into
account this extra byte on each line in Windows NT file systems.
If the pointer is moved, the THEN statements are executed and the ELSE statements are ignored. If the
THEN statements are not specified, program execution continues with the next statement.
If the file cannot be accessed or does not exist, the ELSE statements are executed; any THEN statements
are ignored.
If file.variable, offset, or relto evaluates to the null value, the SEEK statement fails and the program
terminates with a run-time error message.
Note: On Windows NT systems, if you use the OPENDEV statement to open a 1/4-inch cartridge tape
(60 MB or 150 MB) for sequential processing, you can move the file pointer only to the beginning or the
end of the data. For diskette drives, you can move the file pointer only to the start of the data.
Seeking beyond the end of the file and then writing creates a gap, or hole, in the file. This hole occupies
no physical space, and reads from this part of the file return as ASCII CHAR 0 (neither the number nor
the character 0).
For more information about sequential file processing, see the OPENSEQ statement, READSEQ statement,
and WRITESEQ statements.
Example
The following example reads and prints the first line of RECORD4. Then the SEEK statement moves the
pointer five bytes from the front of the file, then reads and prints the rest of the current line.
OPENSEQ ’FILE.E’, ’RECORD4’ TO FILE ELSE ABORT
READSEQ B FROM FILE THEN PRINT B
SEEK FILE,5, 0 THEN READSEQ A FROM FILE
THEN PRINT A ELSE ABORT
Description
Use the SEEK(ARG.) statement to move the command line argument pointer to the next command line
argument from left to right, or to a command line argument specified by arg#. The command line is
delimited by blanks, and the first argument is assumed to be the first word after the program name.
When a cataloged program is invoked, the argument list starts with the second word in the command
line.
Blanks in quoted strings are not treated as delimiters. A quoted string is treated as a single argument.
arg# specifies the command line argument to move to. It must evaluate to a number. If arg# is not
specified , the pointer moves to the next command line argument. SEEK(ARG.) works similarly to
GET(ARG.) statement except that SEEK(ARG.) makes no assignments.
THEN and ELSE statements are both optional. The THEN clause is executed if the argument is found.
The ELSE clause is executed if the argument is not found. The SEEK(ARG.) statement fails if arg#
evaluates to a number greater than the number of command line arguments or if the last argument has
been assigned and a SEEK(ARG.) with no arg# is used. To move to the beginning of the argument list, set
arg# to 1.
If arg# evaluates to the null value, the SEEK(ARG.) statement fails and the program terminates with a
run-time error message.
Example
Description
Use a SELECT statement to create a numbered select list of record IDs from a WebSphere DataStage file
or a dynamic array. A subsequent READNEXT statement can access this select list, removing one record
ID at a time from the list. READNEXT instructions can begin processing the select list immediately.
variable can specify a dynamic array or a file variable. If it specifies a dynamic array, the record IDs must
be separated by field marks (ASCII 254). If variable specifies a file variable, the file variable must have
previously been opened. If variable is not specified, the default file is assumed (for more information on
default files, see the OPEN statement). If the file is neither accessible nor open, or if variable evaluates to
the null value, the SELECT statement fails and the program terminates with a run-time error message.
If the file is an SQL table, the effective user of the program must have SQL SELECT privilege to read
records in the file. For information about the effective user of a program, see the AUTHORIZATION
statement.
You must use a file lock with the SELECT statement when it is within a transaction running at isolation
level 4 (serializable). This prevents phantom reads.
The TO clause specifies the select list that is to be used. list.number is an integer from 0 through 10. If no
list.number is specified, select list 0 is used.
The record IDs of all the records in the file, in their stored order, form the list. Each record ID is one
entry in the list.
The SELECT statement does not process the entire file at once. It selects record IDs group by group. The
@SELECTED variable is set to the number of elements in the group currently being processed.
You often want a select list with the record IDs in an order different from their stored order or with a
subset of the record IDs selected by some specific criteria. To do this, use the SELECT or SSELECT
statement commands in a BASIC EXECUTE statement. Processing the list by READNEXT is the same,
regardless of how the list is created.
Use the SELECTV statement to store the select list in a named list variable instead of to a numbered
select list. list.variable is an expression that evaluates to a valid variable name. This is the default behavior
of the SELECT statement in PICK, REALITY, and IN2 flavor accounts. You can also use the VAR.SELECT
option of the $OPTIONS statement to make the SELECT statement act as it does in PICK, REALITY, and
IN2 flavor accounts.
The ON ERROR clause is optional in the SELECT statement. The ON ERROR clause lets you specify an
alternative for program termination when a fatal error is encountered during processing of the SELECT
statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
306 BASIC Reference Guide
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
You can use either the SELECT or the SELECTV statement to create a select list and store it in a named
list variable. The only useful thing you can do with a list variable is use a READNEXT statement to read
the next element of the select list.
Use the SELECTN statement to store the select list in a numbered select list. list.number is an expression
that evaluates to a number from 0 through 10. You can also use the -VAR.SELECT option of the
$OPTIONS statement to make the SELECT statement act as it does in IDEAL and INFORMATION flavor
accounts.
Example
The following example opens the file SUN.MEMBER to the file variable MEMBER.F, then creates an
active select list of record IDs. The READNEXT statement assigns the first record ID in the select list to
the variable @ID, then prints it. Next, the file SUN.SPORT is opened to the file variable SPORT.F, and a
select list of its record IDs is stored as select list 1. The READNEXT statement assigns the first record ID
in the select list to the variable A, then prints DONE.
OPEN ’’,’SUN.MEMBER’ TO MEMBER.F ELSE PRINT "NOT
OPEN"
SELECT
READNEXT @ID THEN PRINT @ID
*
OPEN ’’,’SUN.SPORT’ TO SPORT.F ELSE PRINT "NOT OPEN"
SELECT TO 1
READNEXT A FROM 1 THEN PRINT "DONE" ELSE PRINT "NOT"
SELECTE statement
Syntax
SELECTE TO list.variable
Description
Use the SELECTE statement to assign the contents of select list 0 to list.variable. list.variable is activated in
place of select list 0 and can be read with the READNEXT statement.
Description
Use the SELECTINDEX statement to create select lists from secondary indexes.
index is an expression that evaluates to the name of an indexed field in the file. index must be the name of
the field that was used in the CREATE.INDEX command when the index was built.
list.number is an expression that evaluates to the select list number. It can be a number from 0 through 10.
The default list number is 0.
alt.key is an expression that evaluates to an alternate key value contained in the index. If alt.key is
specified, a select list is created of the record IDs referenced by alt.key. If alt.key is not specified, a select
list is created of all of the index’s alternate keys.
If the field is not indexed, the select list is empty, and the value of the STATUS function is 1; otherwise
the STATUS function is 0. If index, alt.key, or file.variable evaluates to the null value, the SELECTINDEX
statement fails and the program terminates with a run-time error message.
PIOPEN Flavor
In a PIOPEN flavor account, the SELECTINDEX statement creates a select list from the secondary indexes
without duplicate keys. To implement this functionality in other flavors, use the PIOPEN.SELIDX option
with the $OPTIONS statement.
Example
In the following example, the first SELECTINDEX selects all data values to list 1. The second
SELECTINDEX selects record IDs referenced by STOREDVAL to list 2.
OPEN "", "DB" TO FV ELSE STOP "OPEN FAILED"
SELECTINDEX "F1" FROM FV TO 1
EOV = 0
LOOP
SELECTINDEX "F1" FROM FV TO 1
UNTIL EOV DO
SELECTINDEX "F1", STOREDVAL FROM FV TO 2
EOK = 0
LOOP
READNEXT KEY FROM 2 ELSE EOK=1
UNTIL EOK DO
PRINT "KEY IS ":KEY:" STOREDVAL IS ":STOREDVAL
REPEAT
REPEAT
END
SELECTINFO function
Syntax
SELECTINFO (list, key)
Use the SELECTINFO function to determine whether a select list is active, or to determine the number of
items it contains.
list is an expression evaluating to the number of the select list for which you require information. The
select list number must be in the range of 0 through 10.
key specifies the type of information you require. You can use equate names for the keys as follows:
IK$SLACTIVE
Returns 1 if the select list specified is active, and returns 0 if the select list specified is not active.
IK$SLCOUNT
Returns the number of items in the select list. 0 is returned if the select list is not active or is an
empty select list.
Equate Names
An insert file of equate names is provided for the SELECTINFO keys. To use the equate names, specify
the directive $INCLUDE SYSCOM INFO_KEYS.INS.IBAS when you compile your program.
Example
In the following example, the insert file containing the equate name is inserted by the $INCLUDE
statement. The conditional statement tests if select list 0 is active.
$INCLUDE SYSCOM INFO_KEYS.INS.IBAS
IF SELECTINFO(0,IK$SLACTIVE)
THEN PRINT ’SELECT LIST ACTIVE’
ELSE PRINT ’SELECT LIST NOT ACTIVE’
END
SEND statement
Syntax
SEND output [:] TO device
{ THEN statements [ ELSE statements ] | ELSE statements }
Description
Use the SEND statement to write a block of data to a device. The SEND statement can be used to write
data to a device that has been opened for I/O using the OPENDEV statement or OPENSEQ statement.
output is an expression evaluating to a data string that will be written to device. If the optional colon is
used after output, the terminating newline is not generated.
device is a valid file variable resulting from a successful OPENDEV or OPENSEQ statement. This is the
handle to the I/O device that supplies the data stream for the operation of the SEND statement.
The SEND syntax requires that either a THEN or an ELSE clause, or both, be specified. If data is
successfully sent, the SEND statement executes the THEN clause. If data cannot be sent, it executes the
ELSE clause.
The data block specified by output is written to the device followed by a newline. Upon successful
completion of the SEND operation, program control is passed to the THEN clause if specified. If an error
occurs during the SEND operation, program control is passed to the ELSE clause if specified.
The following code fragment shows how the SEND statement is used to write a series of messages on a
connected device:
OPENDEV "TTY10" TO TTYLINE ELSE STOP "CANNOT OPEN
TTY10"
LOOP
INPUT MESSAGE
WHILE MESSAGE # "QUIT" DO
SEND MESSAGE TO TTYLINE
ELSE
STOP "ERROR WRITING DATA TO TTY10"
END
REPEAT
SENTENCE function
Syntax
SENTENCE ( )
Description
Use the SENTENCE function to return the stored sentence that invoked the current process. Although the
SENTENCE function uses no arguments, parentheses are required to identify it as a function. The
SENTENCE function is a synonym for the @SENTENCE system variable.
A PERFORM statement in a program updates the system variable, @SENTENCE, with the command
specified in the PERFORM statement.
Example
PRINT SENTENCE()
SEQ function
Syntax
SEQ (expression)
Description
Use the SEQ function to convert an ASCII character to its numeric string equivalent.
expression evaluates to the ASCII character to be converted. If expression evaluates to the null value, null is
returned.
In NLS mode, use the UNISEQ function to return Unicode values in the range x0080 through x00F8.
Using the SEQ function to convert a character outside its range results in a run-time message, and the
return of an empty string.
For more information about these ranges, see WebSphere DataStage National Language Support Guide.
In PICK, IN2, and REALITY flavors SEQ(″ ″) is 255 instead of 0. In IDEAL and INFORMATION flavor
accounts, use the SEQ.255 option of the $OPTIONS statement to cause SEQ(″ ″) to be interpreted as 255.
Example
G="T"
A=SEQ(G)
PRINT A, A+1
PRINT SEQ("G")
SEQS function
Syntax
SEQS (dynamic.array)
CALL -SEQS (return.array, dynamic.array)
CALL !SEQS (return.array, dynamic.array)
Description
Use the SEQS function to convert a dynamic array of ASCII characters to their numeric string
equivalents.
dynamic.array specifies the ASCII characters to be converted. If dynamic.array evaluates to the null value,
null is returned. If any element of dynamic.array is the null value, null is returned for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
In NLS mode, you can use the UNISEQS function to return Unicode values in the range x0080 through
x00F8.
Using the SEQS function to convert a character outside its range results in a run-time message, and the
return of an empty string.
For more information about these ranges, see WebSphere DataStage National Language Support Guide.
Example
G="T":@VM:"G"
A=SEQS(G)
PRINT A
PRINT SEQS("G")
Use the SET TRANSACTION ISOLATION LEVEL statement to set the default transaction isolation level
you need for your program.
Note: The isolation level you set with this statement remains in effect until another such statement is
issued. This affects all activities in the session, including WebSphere DataStage commands and SQL
transactions.
The SET TRANSACTION ISOLATION LEVEL statement cannot be executed while a transaction exists.
Attempting to do so results in a run-time error message, program failure, and the rolling back of all
uncommitted transactions started in the execution environment.
Examples
The following example sets the default isolation level to 3 then starts a transaction at isolation level 4.
The isolation level is reset to 3 after the transaction finishes.
SET TRANSACTION ISOLATION LEVEL REPEATABLE.READ
PRINT "We are at isolation level 3."
BEGIN TRANSACTION ISOLATION LEVEL SERIALIZABLE
PRINT "We are at isolation level 4."
COMMIT WORK
END TRANSACTION
PRINT "We are at isolation level 3"
Description
In NLS mode, use the SETLOCALE function to enable or disable a locale for a specified category or
change its setting.
category is one of the following tokens that are defined in the UVNLSLOC.H file:
UVLC$ALL
Sets or disables all categories as specified in value. value is the name of a locale, OFF, or
DEFAULT. value can also be a dynamic array whose elements correspond to the categories.
UVLC$TIME
Sets or disables the Time category. value is the name of a locale, OFF, or DEFAULT.
UVLC$NUMERIC
Sets or disables the Numeric category. value is the name of a locale, OFF, or DEFAULT.
UVLC$MONETARY
Sets or disables the Monetary category. value is the name of a locale, OFF, or DEFAULT.
UVLC$CTYPE
Sets or disables the Ctype category. value is the name of a locale, OFF, or DEFAULT.
UVLC$COLLATE
Sets or disables the Collate category. value is the name of a locale, OFF, or DEFAULT.
UVLC$SAVE
Saves the current locale state, overwriting any previous saved locale. value is ignored.
UVLC$RESTORE
Restores the saved locale state. value is ignored.
value specifies either a dynamic array whose elements are separated by field marks or the string OFF. An
array can have one or five elements:
v If the array has one element, all categories are set or not set to that value.
v If the array has five elements, it specifies the following values in this order: TIME, NUMERIC,
MONETARY, CTYPE, and COLLATE.
The MD, MR, and ML conversions require both Numeric and Monetary categories to be set in order for
locale information to be used.
The STATUS function returns 0 if SETLOCALE is successful, or one of the following error tokens if it
fails:
LCE$NO.LOCALES
WebSphere DataStage locales are disabled.
LCE$BAD.LOCALE
value is not the name of a locale that is currently loaded, or the string OFF.
LCE$BAD.CATEGORY
You specified an invalid category.
LCE$NULL.LOCALE
value has more than one field and a category is missing.
For more information about locales, see WebSphere DataStage National Language Support Guide.
Examples
The following example sets all the categories in the locale to FR-FRENCH:
status = SETLOCALE(UVLC$ALL,"FR-FRENCH")
The next example saves the current locale. This is the equivalent of executing the SAVE.LOCALE
command.
status = SETLOCALE(UVLC$SAVE,"")
The next example disables the Monetary category. WebSphere DataStage behaves as though there were no
locales for the Monetary category only.
status = SETLOCALE(UVLC$MONETARY,"OFF")
The next example completely disables locale support for all categories:
status = SETLOCALE(UVLC$ALL,"OFF")
SETREM statement
Syntax
SETREM position ON dynamic.array
Description
Use the SETREM statement to set the remove pointer in dynamic.array to the position specified by position.
position is an expression that evaluates to the number of bytes you want to move the pointer forward. If it
is larger than the length of dynamic.array, the length of dynamic.array is used. If it is less than 0, 0 is used.
dynamic.array must be a variable that evaluates to a string. If it does not evaluate to a string, an improper
data type warning is issued.
If the pointer does not point to the first character after a system delimiter, subsequent REMOVE
statement and REVREMOVE statement act as follows:
v A REMOVE statement returns a substring, starting from the pointer and ending at the next delimiter.
v A REVREMOVE statement returns a substring, starting from the previous delimiter and ending at the
pointer.
If NLS is enabled and you use a multibyte character set, use GETREM function to ensure that position is
at the start of a character. For more information about locales, see WebSphere DataStage National Language
Support Guide.
SIN function
Syntax
SIN (expression)
Description
Use the SIN function to return the trigonometric sine of an expression. expression represents the angle
expressed in degrees. Numbers greater than 1E17 produce a warning message, and 0 is returned. If
expression evaluates to the null value, null is returned.
Example
PRINT SIN(45)
SINH function
Syntax
SINH (expression)
Description
Use the SINH function to return the hyperbolic sine of expression. expression must be numeric and
represents the angle expressed in degrees. If expression evaluates to the null value, null is returned.
Example
PRINT "SINH(2) = ":SINH(2)
SLEEP statement
Syntax
SLEEP [seconds]
Description
Use the SLEEP statement to suspend execution of a BASIC program, pausing for a specified number of
seconds.
Chapter 6. BASIC Statements and Functions 315
seconds is an expression evaluating to the number of seconds for the pause. If seconds is not specified, a
value of 1 is used. If seconds evaluates to the null value, it is ignored and 1 is used.
Example
In the following example the program pauses for three seconds before executing the statement after the
SLEEP statement. The EXECUTE statement clears the screen.
PRINT "STUDY THE FOLLOWING SENTENCE CLOSELY:"
PRINT
PRINT
PRINT "There are many books in the"
PRINT "the library."
SLEEP 3
EXECUTE ’CS’
PRINT "DID YOU SEE THE MISTAKE?"
SMUL function
Syntax
SMUL (string.number.1, string.number.2)
Description
Use the SMUL function to multiply two string numbers and return the result as a string number. You can
use this function in any expression where a string or string number is valid, but not necessarily where a
standard number is valid, because string numbers can exceed the range of numbers that standard
arithmetic operators can handle.
If either string number contains nonnumeric data, an error message is generated and 0 is used for that
number. If either string number evaluates to the null value, null is returned.
Example
X = "5436"
Y = "234"
Z = SMUL (X,Y)
PRINT Z
SOUNDEX function
Syntax
SOUNDEX (expression)
The SOUNDEX function evaluates expression and returns the most significant letter in the input string
followed by a phonetic code. Characters that are not alphabetic are ignored. If expression evaluates to the
null value, null is returned.
This function uses the soundex algorithm (the same as the one used by the SAID keyword in RetrieVe) to
analyze the input string. The soundex algorithm returns the first letter of the alphabetic string followed
by a one- to three-digit phonetic code.
Example
Source Lines
Program Output
DATA ″MCDONALD″, ″MACDONALD″, ″MACDOUGALL″ FOR I=1 TO 3 INPUT CUSTOMER
PHONETIC.CODE=SOUNDEX(CUSTOMER) PRINT PHONETIC.CODE NEXT
?MCDONALD
M235
?MACDONALD
M235
?MACDOUGALL
M232
SPACE function
Syntax
SPACE (expression)
Description
Use the SPACE function to return a string composed of blank spaces. expression specifies the number of
spaces in the string. If expression evaluates to the null value, the SPACE function fails and the program
terminates with a run-time error message.
Example
PRINT "HI":SPACE(20):"THERE"
*
*
VAR=SPACE(5)
PRINT "TODAY IS":VAR:OCONV(DATE(),"D")
SPACES function
Syntax
SPACES (dynamic.array)
CALL -SPACES (return.array, dynamic.array)
CALL !SPACES (return.array, dynamic.array)
Use the SPACES function to return a dynamic array with elements composed of blank spaces.
dynamic.array specifies the number of spaces in each element. If dynamic.array or any element of
dynamic.array evaluates to the null value, the SPACES function fails and the program terminates with a
run-time error message.
There is no limit to the number of blank spaces that can be generated except available memory.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
SPLICE function
Syntax
SPLICE (array1, expression, array2)
CALL -SPLICE (return.array, array1, expression, array2)
CALL !SPLICE (return.array, array1, expression, array2)
Description
Use the SPLICE function to create a dynamic array of the element-by-element concatenation of two
dynamic arrays, separating concatenated elements by the value of expression.
Each element of array1 is concatenated with expression and with the corresponding element of array2. The
result is returned in the corresponding element of a new dynamic array. If an element of one dynamic
array has no corresponding element in the other dynamic array, the element is returned properly
concatenated with expression. If either element of a corresponding pair is the null value, null is returned
for that element. If expression evaluates to the null value, null is returned for the entire dynamic array.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A="A":@VM:"B":@SM:"C"
B="D":@SM:"E":@VM:"F"
C=’-’
PRINT SPLICE(A,C,B)
SQRT function
Syntax
SQRT (expression)
Description
Use the SQRT function to return the square root of expression. expression must evaluate to a numeric value
that is greater than or equal to 0. If expression evaluates to a negative value, the result of the function is
SQRT(-n) and an error message is printed. If expression evaluates to the null value, null is returned.
Example
A=SQRT(144)
PRINT A
*
PRINT "SQRT(45) IS ":SQRT(45)
SQUOTE function
Syntax
SQUOTE (expression )
CALL !SQUOTE ( quoted.expression, expression )
Description
Use the SQUOTE function to enclose an expression in single quotation marks. If expression evaluates to
the null value, null is returned, without quotation marks.
Example
PRINT SQUOTE(12 + 5) : " IS THE ANSWER."
END
SSELECT statement
Syntax
SSELECT [variable] [TO list.number] [ON ERROR statements]
SSELECTN [variable] [TO list.number] [ON ERROR statements]
SSELECTV [variable] TO list.variable [ON ERROR statements]
Description
Use an SSELECT statement to create a numbered select list of record IDs in sorted order from a
WebSphere DataStage file or a dynamic array. You can then access this select list by a subsequent
READNEXT statement which removes one record ID at a time from the list.
variable can specify a dynamic array or a file variable. If it specifies a dynamic array, the record IDs must
be separated by field marks (ASCII 254). If variable specifies a file variable, the file variable must have
previously been opened. If variable is not specified, the default file is assumed (for more information on
default files, see the OPEN statement). If the file is neither accessible nor open, or if variable evaluates to
the null value, the SSELECT statement fails and the program terminates with a run-time error message.
If the file is an SQL table, the effective user of the program must have SQL SELECT privilege to read
records in the file. For information about the effective user of a program, see the AUTHORIZATION
statement.
You must use a file lock with the SSELECT statement when it is within a transaction running at isolation
level 4 (serializable). This prevents phantom reads.
The TO clause specifies the select list that is to be used. list.number is an integer from 0 through 10. If no
list.number is specified, select list 0 is used.
You often want a select list with the record IDs in an order different from their stored order or with a
subset of the record IDs selected by some specific criteria. To do this, use the SELECT statements or
SSELECT commands in a BASIC EXECUTE statement. Processing the list by READNEXT statement is the
same, regardless of how the list is created.
Use the SSELECTV statement to store the select list in a named list variable instead of to a numbered
select list. list.variable is an expression that evaluates to a valid variable name. This is the default behavior
of the SSELECT statement in PICK, REALITY, and IN2 flavor accounts. You can also use the
VAR.SELECT option of the $OPTIONS statement to make the SSELECT statement act as it does in PICK,
REALITY, and IN2 flavor accounts.
In NLS mode when locales are enabled, the SSELECT statements use the Collate convention of the
current locale to determine the collating order. For more information about locales, see WebSphere
DataStage National Language Support Guide.
The ON ERROR clause is optional in SSELECT statements. The ON ERROR clause lets you specify an
alternative for program termination when a fatal error is encountered during processing of a SSELECT
statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
In a PICK, REALITY, or IN2 flavor account, the SSELECT statement has the following syntax:
SSELECT[V] [variable] TO list.variable
SSELECTN [variable] TO list.number
You can use either the SSELECT or the SSELECTV statement to create a select list and store it in a named
list variable. The only useful thing you can do with a list variable is use a READNEXT statement to read
the next element of the select list.
Use the SSELECTN statement to store the select list in a numbered select list. list.number is an expression
that evaluates to a number from 0 through 10. You can also use the -VAR.SELECT option of the
$OPTIONS statement to make the SSELECT statement act as it does in IDEAL and INFORMATION
flavor accounts.
The following example opens the file SUN.MEMBER to the file variable MEMBER.F, then creates an
active sorted select list of record IDs. The READNEXT statement assigns the first record ID in the select
list to the variable @ID, then prints it. Next, the file SUN.SPORT is opened to the file variable SPORT.F,
and a sorted select list of its record IDs is stored as select list 1. The READNEXT statement assigns the
first record ID in the select list to the variable A, then prints DONE.
OPEN ’’,’SUN.MEMBER’ ELSE PRINT "NOT OPEN"
SSELECT
READNEXT @ID THEN PRINT @ID
*
OPEN ’’,’SUN.SPORT’ ELSE PRINT "NOT OPEN"
SSELECT TO 1
READNEXT A FROM 1 THEN PRINT "DONE" ELSE PRINT "NOT"
SSUB function
Syntax
SSUB (string.number.1, string.number.2)
Description
Use the SSUB function to subtract string.number.2 from string.number.1 and return the result as a string
number. You can use this function in any expression where a string or string number is valid, but not
necessarily where a standard number is valid, because string numbers can exceed the range of numbers
that standard arithmetic operators can handle.
If either string number contains nonnumeric data, an error message is generated, and 0 replaces the
nonnumeric data. If either string number evaluates to the null value, null is returned.
Example
X = "123456"
Y = "225"
Z = SSUB (X,Y)
PRINT Z
STATUS function
Syntax
STATUS ( )
Description
Use the STATUS function to determine the results of the operations performed by certain statements and
functions.
After a DELETE statement with an ON ERROR clause, the value returned is the error number.
After a successful execution of the FILEINFO function, STATUS returns 0. If the function fails to execute,
STATUS returns a nonzero value. For complete information, see the FILEINFO function.
After a FILELOCK statement with a LOCKED clause, the value returned is the terminal number of the
user who has a conflicting lock.
A 0 is returned if the statement was completed by a Return. The trap number is returned if the statement
was completed by one of the trapped keys (see the INPUT@ and KEYTRAP statements).
The file type is returned if the file is opened successfully. If the file is not opened successfully, the
following values might return
:
Value Description
-1 The filename was not found in the VOC file.
-21 The filename or file is null.
-3 An operating system access error occurs when you do not have permission to access a WebSphere
DataStage file in a directory. For example, this error might occur when trying to access a type 1
or type 30 file.
-41 An access error appears when you do not have operating system permissions or if DATA.30 is
missing for a type 30 file.
-5 The operating system detected a read error.
-6 The lock file header cannot be unlocked.
-7 Invalid file revision or wrong byte-ordering exists for the platform.
-81 Invalid part file information exists.
1
-9 Invalid type 30 file information exists in a distributed file.
-10 A problem occurred while the file was being rolled forward during warmstart recovery.
Therefore, the file is marked ″inconsistent.″
-11 The file is a view; therefore it cannot be opened by a BASIC program.
-12 No SQL privileges exist to open the table.
1
-13 An index problem exists.
-14 The NFS file cannot be opened.
If the file is a distributed file, the STATUS function returns the following:
-1 The partitioning algorithm does not evaluate to an integer.
-2 The part number is invalid.
If the statement includes the LOCKED clause, the returned value is the terminal number, as returned by
the WHO command, of the user who set the lock.
.
3 The unmappable character is in the record ID.
4 The unmappable character is in the record’s data.
If the statement takes the ELSE clause, the returned value is 1. Otherwise the returned value is 0.
The STATUS function returns 0 if SETLOCALE function is successful, or one of the following error tokens
if it fails:
LCE$NO.LOCALES
WebSphere DataStage locales are disabled.
LCE$BAD.LOCALE
The specified locale name is not currently loaded, or the string OFF.
LCE$BAD.CATEGORY
You specified an invalid category.
LCE$NULL.LOCALE
The specified locale has more than one field and a category is missing.
Example
Source Lines
Program Output
OPEN ’’,’EX.BASIC’ TO FILE ELSE STOP PRINT ’STATUS() IS ’:STATUS() Q=123456
Q=OCONV(Q,″MD2″) PRINT ’STATUS() IS ’:STATUS() Q=’ASDF’ Q=OCONV(Q,″D2/″) PRINT
’STATUS() IS ’:STATUS()
STATUS() IS 1
STATUS() IS 0
STATUS() IS 1
STATUS statement
Syntax
STATUS dynamic.array FROM file.variable{THEN statements [ELSE statements] | ELSE statements}
Description
Use the STATUS statement to determine the status of an open file. The STATUS statement returns the file
status as a dynamic array and assigns it to dynamic.array.
The following table lists the values of the dynamic array returned by the STATUS statement:
Table 55. STATUS Statement Values
Field Stored Value Description
1 Current position in the file Offset in bytes from beginning of the
file.
2 End of file reached 1 if EOF, 0 if not.
3 Error accessing file 1 if error, 0 if not.
4 Number of bytes available to read
1 write-only
2 read-only
3 read/write
4 delete-only
5 delete/write
6 delete/read
7 delete/read/write
29 1 if this is an SQL table, 0 if not. If
the file is a view, the STATUS
statement fails. (No information on a
per-column basis is returned.)
30 User name User name of the owner of the file.
file.variable specifies an open file. If file.variable evaluates to the null value, the STATUS statement fails and
the program terminates with a run-time error message.
If the STATUS array is assigned to dynamic.array, the THEN statements are executed and the ELSE
statements are ignored. If no THEN statements are present, program execution continues with the next
statement. If the attempt to assign the array fails, the ELSE statements are executed; any THEN
statements are ignored.
Example
OPENSEQ ’/etc/passwd’ TO test THEN PRINT "File
Opened" ELSE ABORT
STATUS stat FROM test THEN PRINT stat
field5 = stat<5,1,1>
field6 = stat<6,1,1>
field8 = stat<8,1,1>
PRINT "permissions:": field5
PRINT "filesize:": field6
PRINT "userid:": field8
CLOSESEQ test
STOP statement
Syntax
STOP [expression]
STOPE [expression]
STOPM [expression]
Description
Use the STOP statement to terminate program execution and return system control to the calling
environment, which can be a menu, a paragraph, another BASIC program, or the WebSphere DataStage
command processor.
When expression is specified, its value is displayed before the STOP statement is executed. If expression
evaluates to the null value, nothing is printed.
To stop all processes and return to the command level, use the ABORT statement.
Use the ERRMSG statement if you want to display a formatted error message from the ERRMSG file
when the program stops.
The STOPE statement uses the ERRMSG file for error messages instead of using text specified by
expression. The STOPM statement uses text specified by expression rather than messages in the ERRMSG
file. If expression in the STOPE statement evaluates to the null value, the default error message is printed:
Message ID is NULL: undefined error
In PICK, IN2, and REALITY flavor accounts, the STOP statement uses the ERRMSG file for error
messages instead of using text specified by expression. Use the STOP.MSG option of the $OPTIONS
statement to get this behavior in IDEAL and INFORMATION flavor accounts.
Example
PRINT "1+2=":1+2
STOP "THIS IS THE END"
STORAGE statement
Syntax
STORAGE arg1 arg2 arg3
Description
The STORAGE statement performs no function. It is provided for compatibility with other Pick systems.
328 BASIC Reference Guide
STR statement
Syntax
STR (string, repeat)
Description
Use the STR function to produce a specified number of repetitions of a particular character string.
repeat is an expression that evaluates to the number of times string is to be repeated. If repeat does not
evaluate to a value that can be truncated to a positive integer, an empty string is returned.
If string evaluates to the null value, null is returned. If repeat evaluates to the null value, the STR function
fails and the program terminates with a run-time error message.
Example
PRINT STR(’A’,10)
*
X=STR(5,2)
PRINT X
*
X="HA"
PRINT STR(X,7)
STRS function
Syntax
STRS (dynamic.array, repeat)
CALL -STRS (return.array, dynamic.array, repeat)
CALL !STRS (return.array, dynamic.array, repeat)
Description
Use the STRS function to produce a dynamic array containing the specified number of repetitions of each
element of dynamic.array.
repeat is an expression that evaluates to the number of times the elements are to be repeated. If it does
not evaluate to a value that can be truncated to a positive integer, an empty string is returned for
dynamic.array.
If dynamic.array evaluates to the null value, null is returned. If any element of dynamic.array is the null
value, null is returned for that element. If repeat evaluates to the null value, the STRS function fails and
the program terminates with a run-time error message.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
SUBR function
Syntax
SUBR (name, [argument [ ,argument ... ] ] )
Description
Use the SUBR function to return the value of an external subroutine. The SUBR function is commonly
used in I-descriptors.
name is an expression that evaluates to the name of the subroutine to be executed. This subroutine must
be cataloged in either a local catalog or the system catalog, or it must be a record in the same object file
as the calling program. If name evaluates to the null value, the SUBR function fails and the program
terminates with a run-time error message.
argument is an expression evaluating to a variable name whose value is passed to the subroutine. You can
pass up to 254 variables to the subroutine.
Subroutines called by the SUBR function must have a special syntax. The SUBROUTINE statement
defining the subroutine must specify a dummy variable as the first parameter. The value of the
subroutine is the value of the dummy variable when the subroutine finishes execution. Because the
SUBROUTINE statement has this dummy parameter, the SUBR function must specify one argument less
than the number of parameters in the SUBROUTINE statement. In other words, the SUBR function does
not pass any argument to the subroutine through the first dummy parameter. The first argument passed
by the SUBR function is referenced in the subroutine by the second parameter in the SUBROUTINE
statement, and so on.
Example
SUBROUTINE statement
Syntax
SUBROUTINE [name] [ ( [MAT] variable [ , [MAT] variable ... ] ) ]
Description
Use the SUBROUTINE statement to identify an external subroutine. The SUBROUTINE statement must
be the first line in the subroutine that is not a comment line. Each external subroutine can contain only
one SUBROUTINE statement.
An external subroutine is a separate program or set of statements that can be executed by other programs
or subroutines (called calling programs) to perform a task. The external subroutine must be compiled and
cataloged before another program can call it.
The SUBROUTINE statement can specify a subroutine name for documentation purposes; it need not be
the same as the program name or the name by which it is called. The CALL statement must reference the
subroutine by its name in the catalog, in the VOC file, or in the object file.
variables are variable names used in the subroutine to pass values between the calling programs and the
subroutine. To pass an array, you must precede the array name with the keyword MAT. When an external
subroutine is called, the CALL statement must specify the same number of variables as are specified in
the SUBROUTINE statement. See the CALL statement for more information.
Example
The following SUBROUTINE statements specify three variables, EM, GROSS, and TAX, the values of
which are passed to the subroutine by the calling program:
SUBROUTINE ALONE(EM, GROSS, TAX)
SUBROUTINE STATE(EM,GROSS,TAX)
SUBS function
Syntax
SUBS (array1, array2)
CALL -SUBS (return.array, array1, array2)
CALL !SUBS (return.array, array1, array2)
Use the SUBS function to create a dynamic array of the element-by-element subtraction of two dynamic
arrays.
Each element of array2 is subtracted from the corresponding element of array1 with the result being
returned in the corresponding element of a new dynamic array.
If an element of one dynamic array has no corresponding element in the other dynamic array, the missing
element is evaluated as 0. If either of a corresponding pair of elements is the null value, null is returned
for that element.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A=2:@VM:4:@VM:6:@SM:18
B=1:@VM:2:@VM:3:@VM:9
PRINT SUBS(A,B)
SUBSTRINGS function
Syntax
SUBSTRINGS (dynamic.array, start, length)
CALL -SUBSTRINGS (return.array, dynamic.array, start, length)
CALL !SUBSTRINGS (return.array, dynamic.array, start, length)
Description
Use the SUBSTRINGS function to create a dynamic array each of whose elements are substrings of the
corresponding elements of dynamic.array.
start indicates the position of the first character of each element to be included in the substring. If start is
0 or a negative number, the starting position is assumed to be 1. If start is greater than the number of
characters in the element, an empty string is returned.
length specifies the total length of the substring. If length is 0 or a negative number, an empty string is
returned. If the sum of start and length is larger than the element, the substring ends with the last
character of the element.
If an element of dynamic.array is the null value, null is returned for that element. If start or length
evaluates to the null value, the SUBSTRINGS function fails and the program terminates with a run-time
error message.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
Example
A="ABCDEF":@VM:"GH":@SM:"IJK"
PRINT SUBSTRINGS(A,3,2)
Description
Use the SUM function to calculate the sum of numeric data. Only elements at the lowest delimiter level
of a dynamic array are summed. The total is returned as a single element at the next highest delimiter
level.
The delimiters from highest to lowest are field, value, and subvalue.
There are seven levels of delimiters from CHAR(254) to CHAR(248): field mark, value mark, subvalue
mark, text mark, CHAR(250), CHAR(249), and CHAR(248).
The SUM function removes the lowest delimiter level from a dynamic array. In a dynamic array that
contains fields, values, and subordinate values, the SUM function sums only the subordinate values,
returning the sums as values. In a dynamic array that contains fields and values, the SUM function sums
only the values, returning the sums as fields. In a dynamic array that contains only fields, the SUM
function sums the fields, returning the sum as the only field of the array. SUM functions can be applied
repeatedly to raise multilevel data to the highest delimiter level or to a single value.
Nonnumeric values, except the null value, are treated as 0. If dynamic.array evaluates to the null value,
null is returned. Any element that is the null value is ignored, unless all elements of dynamic.array are
null, in which case null is returned.
Examples
In the following examples a field mark is shown by F, a value mark is shown by V, and a subvalue mark
is shown by S.
Source Lines
Program Output
X=20:@VM:18:@VM:9:@VM:30:@VM:80 PRINT ″SUM(X)=″,SUM(X)
SUM(X)= 157
X=17:@FM:18:@FM:15 Y=10:@FM:20 PRINT ″SUM(X)+SUM(Y)= ″,SUM(X)+SUM(Y)
SUM(X)+SUM(Y)= 80
X=3:@SM:4:@SM:10:@VM:3:@VM:20 Y=SUM(X) PRINT ″Y= ″,Y Z=SUM(Y) PRINT ″Z= ″,Z
Y= 17V3V20
Z= 40
SUMMATION function
Syntax
SUMMATION (dynamic.array)
CALL !SUMMATION ( result , dynamic.array )
Description
Use the SUMMATION function to return the sum of all the elements in dynamic.array. Nonnumeric
values, except the null value, are treated as 0.
Example
A=1:@VM:"ZERO":@SM:20:@FM:-25
PRINT "SUMMATION(A)=",SUMMATION(A)
SYSTEM function
Syntax
SYSTEM (expression)
Description
Use the SYSTEM function to check on the status of a system function. Use the SYSTEM function to test
whether NLS is on when you run a program, and to display information about NLS settings.
expression evaluates to the number of the system function you want to check. If expression evaluates to the
null value, the SYSTEM function fails and the program terminates with a run-time error message.
The following table lists the values for expression and their meanings. Values 100 through 107 (read-only)
for the SYSTEM function contain NLS information. See the include file UVNLS.H for their tokens.
Table 56. SYSTEM Function Values
Value Action
1 Checks to see if the PRINTER statement ON statement
has turned the printer on. Returns 1 if the printer is on
and 0 if it is not.
2 Returns the page width as defined by the terminal
characteristic settings.
3 Returns the page length as defined by the terminal
characteristic settings.
4 Returns the number of lines remaining on the current
page.
5 Returns the current page number.
6 Returns the current line number.
7 Returns the terminal code for the type of terminal the
system believes you are using.
8,n Checks whether the tape is attached. Returns the current
block size if it is and -1 if it is not. n is the number of the
tape unit. If it is not specified, tape unit 0 is assumed.
9 Returns the current CPU millisecond count.
10 Checks whether the DATA stack is active. Returns 1 if it
is active and 0 if it is not.
11 Checks whether select list 0 is active. Returns 1 if select
list 0 is active and 0 if it is not.
12 By default, returns the current system time in seconds
(local time). If the TIME.MILLISECOND option is set
(see $OPTIONS), returns the current system time in
milliseconds.
Examples
The first example returns the number of lines left to print on a page, with the maximum defined by the
TERM command. The second example returns the current page number.
Source Lines
Program Output
Q=4 PRINT ’SYSTEM(Q)’,SYSTEM(Q)
SYSTEM(Q) 20
PRINT ’X=’,SYSTEM(5)
X= 0
The next example sets a 30-second timeout for the network connection to the system ORION:
TIMEOUT SYSTEM(1200, "ORION"), 30
TABSTOP statement
Syntax
TABSTOP expression
Description
Use the TABSTOP statement to set the current tab stop width for PRINT statements. The initial tab stop
setting is 10.
If expression evaluates to the null value, the TABSTOP statement fails and the program terminates with a
run-time error message.
Example
A="FIRST"
B="LAST"
PRINT A,B
TABSTOP 15
PRINT A,B
TAN function
Syntax
TAN (expression)
Use the TAN function to return the trigonometric tangent of expression. expression represents an angle
expressed in degrees.
Trying to take the tangent of a right angle results in a warning message, and a return value of 0.
Numbers greater than 1E17 produce a warning message, and 0 is returned. If expression evaluates to the
null value, null is returned.
Example
PRINT TAN(45)
TANH function
Syntax
TANH (expression)
Description
Use the TANH function to return the hyperbolic tangent of expression. expression must be numeric and
represents the angle expressed in degrees. If expression evaluates to the null value, null is returned.
Example
PRINT TANH(45)
TERMINFO function
Syntax
TERMINFO (argument)
Description
Use the TERMINFO function to access the device-independent terminal handler string defined for the
current terminal type. The TERMINFO function returns a dynamic array containing the terminal
characteristics for the terminal type set by TERM or SET.TERM.TYPE.
argument can be 0 or 1, depending on whether the terminal characteristics are returned as stored, or
converted to printable form. If argument is 0, the function returns the terminal characteristics in the form
usable by BASIC applications for device-independent terminal handling with the TPARM function and
the TPRINT statement. If argument is 1, the function returns characteristics in terminfo source format.
Boolean values are returned as Y = true and N = false. The terminfo files contain many unprintable
control characters that might adversely affect your terminal.
If argument evaluates to the null value, the TERMINFO function fails and the program terminates with a
run-time error message.
The easiest way to access the terminfo characteristics is by including the BASIC file UNIVERSE.INCLUDE
TERMINFO in your program. The syntax is:
$INCLUDE UNIVERSE.INCLUDE TERMINFO
Example
$INCLUDE UNIVERSE.INCLUDE TERMINFO
PRINT AT.NEGATIVE.1
PRINT "Your terminal type is":TAB:TERMINAL.NAME
Description
Use the TIME function to return a string value expressing the internal time of day. The internal time is
the number of seconds that have passed since midnight to the nearest thousandth of a second (local
time).
The parentheses must be used with the TIME function to distinguish it from a user-named variable called
TIME. However, no arguments are required with the TIME function.
UNIX System V
If the TIME.MILLISECOND option of the $OPTIONS statement is set, the TIME function returns the
system time in whole seconds.
Example
PRINT TIME()
TIMEDATE function
Syntax
TIMEDATE ( )
Description
Use the TIMEDATE function to return the current system time and date in the following format:
hh:mm:ss dd mmm yyyy
hh Hours (based on a 24-hour clock)
mm Minutes
ss Seconds
dd Day
mmm Month
yyyy Year
If you want to increase the number of spaces between the time and the date, edit the line beginning with
TMD0001 in the msg.txt file in the UV account directory. This line can contain up to four hash signs (#).
Each # prints a space between the time and the date.
If NLS mode is enabled, the TIMEDATE function uses the convention defined in the TIMEDATE field in
the NLS.LC.TIME file for combined time and date format. Otherwise, it returns the time and date. For
Examples
PRINT TIMEDATE()
If the TMD0001 message contains four #s, the program output is:
11:19:07 18 JUN 1996
TIMEOUT statement
Syntax
TIMEOUT {file.variable | link.number}, time
Description
Use the TIMEOUT statement to terminate a READSEQ or READBLK statement if no data is read in the
specified time. You can also use the TIMEOUT statement to set a time limit for a network link. Use the
TTYGET and TTYSET statements to set a timeout value for a file open on a serial communications port.
time is an expression that evaluates to the number of seconds the program should wait before terminating
the READSEQ or READBLK statement or the network connections.
link.number is the network link. It is a positive number from 1 through 255 (or the number set in the
NET_MAXCONNECT VALUE for network connections).
TIMEOUT causes subsequent READSEQ and READBLK statement to terminate and execute their ELSE
statements if the number of seconds specified by time elapses while waiting for data. Use the STATUS
function to determine if time has elapsed. In the event of a timeout, neither READBLK nor READSEQ
returns any bytes from the buffer, and the entire I/O operation must be retried.
If either file.variable or time evaluates to the null value, the TIMEOUT statement fails and the program
terminates with a run-time error message.
Examples
TIMEOUT SUN.MEMBER, 10
READBLK VAR1 FROM SUN.MEMBER, 15 THEN PRINT VAR1 ELSE
IF STATUS() = 2 THEN
PRINT "TIMEOUT OCCURRED"
END ELSE
PRINT "CANNOT OPEN FILE"
END
GOTO EXIT.PROG
END
The following example sets a 30-second timeout for the network connection to the system ORION:
TPARM function
Syntax
TPARM (terminfo.string, [arg1], [arg2], [arg3], [arg4], [arg5],
[arg6], [arg7], [arg8] )
Description
terminfo.string represents a string of characters to be compiled by the terminfo compiler, tic. These
terminal descriptions define the sequences of characters to send to the terminal to perform special
functions. terminfo.string evaluates to one of four types of capability: numeric, Boolean, string, or
parameterized string. If terminfo.string or any of the eight arguments evaluates to the null value, the
TPARM function fails and the program terminates with a run-time error message.
Numeric capabilities are limited to a length of five characters that must form a valid number. Only
nonnegative numbers 0 through 32,767 are allowed. If a value for a particular capability does not apply,
the field should be left blank.
Boolean capabilities are limited to a length of one character. The letter Y (in either uppercase or
lowercase) indicates that the specified capability is present. Any value other than Y indicates that the
specified capability is not present.
String capabilities are limited to a length of 44 characters. You can enter special characters as follows:
\E or \e
The ESC character (ASCII 27).
\n or \l
The LINEFEED character (ASCII 10).
\r The RETURN character (ASCII 13).
\t The TAB character (ASCII 9).
\b The BACKSPACE character (ASCII 8).
\f The formfeed character (ASCII 12).
\s A space (ASCII 32).
^x The representation for a control character (ASCII 0 through 31). The character can be either
uppercase or lowercase. A list of some control character representations follows:
Representation
Control Character
^A ^a
Parameterized string capabilities, such as cursor addressing, use special encoding to include values in the
appropriate format. The parameter mechanism is a stack with several commands to manipulate it:
Table 58. Parameter Mechanism
Code Explanation
%pn Push parameter number n onto the stack. Parameters
number 1 through 8 are allowed and are represented by
arg1 through arg8 of the TPARM function.
%’c’ The ASCII value of character c is pushed onto the stack.
%[nnn] Decimal number nnn is pushed onto the top of the stack.
%d Pop the top parameter off the stack, and output it as a
decimal number.
%nd Pop the top parameter off the stack, and output it as a
decimal number in a field n characters wide.
%0nd Like %nd, except that 0s are used to fill out the field.
%c The top of the stack is taken as a single ASCII character
and output.
%s The top of the stack is taken as a string and output.
The top two elements are popped off the stack and
%+ %- %* %/ added, subtracted, multiplied, or divided. The result is
pushed back on the stack. The fractional portion of a
quotient is discarded.
%m The second element on the stack is taken modulo of the
first element, and the result is pushed onto the stack.
%& % | %^ The top two elements are popped off the stack and a
bitwise AND, OR, or XOR operation is performed. The
result is pushed onto the stack.
%= %< %> The second element on the stack is tested for being equal
to, less then, or greater than the first element. If the
comparison is true, a 1 is pushed onto the stack,
otherwise a 0 is pushed.
A delay in milliseconds can appear anywhere in a string capability. A delay is specified by $<nnn>, where
nnn is a decimal number indicating the number of milliseconds (one thousandth of a second) of delay
desired. A proper number of delay characters will be output, depending on the current baud rate.
TPRINT statement
Syntax
TPRINT [ON print.channel] [print.list]
Description
Use the TPRINT statement to send data to the screen, a line printer, or another print file. TPRINT is
similar to the PRINT statement, except that TPRINT lets you specify time delay expressions in the print
list.
The ON clause specifies the logical print channel to use for output. print.channel is an expression that
evaluates to a number from -1 through 255. If you do not use the ON clause, logical print channel 0 is
used, which prints to the user’s terminal if PRINTER statement OFF is set (see the PRINTER statement).
If print.channel evaluates to the null value, the TPRINT statement fails and the program terminates with a
run-time error message. Logical print channel -1 prints the data on the screen, regardless of whether a
PRINTER ON statement has been executed.
You can specify HEADING statement, FOOTING statement, PAGE statement, and PRINTER statement
CLOSE statements for each logical print channel. The contents of the print files are printed in order by
logical print channel number.
print.list can contain any BASIC expression. The elements of the list can be numeric or character strings,
variables, constants, or literal strings. The list can consist of a single expression or a series of expressions
separated by commas ( , ) or colons ( : ) for output formatting. If no print.list is designated, a blank line is
printed. The null value cannot be printed.
print.list can also contain time delays of the form $<time>. time is specified in milliseconds to the tenth of
a millisecond. As the print list is processed, each time delay is executed as it is encountered.
Expressions separated by colons are concatenated. That is, the expression following the colon is printed
immediately after the expression preceding the colon. To print a list without a LINEFEED and RETURN,
end print.list with a colon ( : ).
If NLS is enabled, the TPRINT statement maps data in the same way as the PRINT statement. For more
information about maps, see WebSphere DataStage National Language Support Guide.
Example
The following example prints the string ALPHA followed by a delay of 1 second, then the letters in the
variable X. The printing of each letter is followed by a delay of one tenth of a second.
X="A$<100>B$<100>C$<100>D$<100>E"
TPRINT "ALPHA$<1000.1> ":X
TRANS function
Syntax
TRANS ( [DICT] filename, record.ID, field#, control.code)
Description
Use the TRANS function to return the contents of a field or a record in a WebSphere DataStage file.
TRANS opens the file, reads the record, and extracts the specified data.
filename is an expression that evaluates to the name of the remote file. If TRANS cannot open the file, a
run-time error occurs, and TRANS returns an empty string.
record.ID is an expression that evaluates to the ID of the record to be accessed. If record.ID is multi-valued,
the translation occurs for each record ID and the result is multi-valued (system delimiters separate data
translated from each record).
field# is an expression that evaluates to the number of the field from which the data is to be extracted. If
field# is -1, the entire record is returned, except for the record ID.
control.code is an expression that evaluates to a code specifying what action to take if data is not found or
is the null value. The possible control codes are:
X (default) Returns an empty string if the record does not exist or data cannot be found.
V Returns an empty string and produces an error message if the record does not exist or data
cannot be found.
C Returns the value of record.ID if the record does not exist or data cannot be found.
N Returns the value of record.ID if the null value is found.
The returned value is lowered. For example, value marks in the original field become subvalue marks in
the returned value. For more information, see the LOWER function.
Example
X=TRANS("VOC","EX.BASIC",1,"X")
PRINT "X= ":X
*
FIRST=TRANS("SUN.MEMBER","6100",2,"X")
LAST=TRANS("SUN.MEMBER","6100",1,"X")
PRINT "NAME IS ":FIRST:" ":LAST
TRANSACTION statements
Syntax
BEGIN TRANSACTION
[statements]
{ COMMIT [WORK] | ROLLBACK [WORK] }
[statements]
[{ COMMIT [WORK] | ROLLBACK [WORK] }
[statements]
.
.
. ]
END TRANSACTION
Syntax (PIOPEN)
TRANSACTION START
{THEN statements [ELSE statements] | ELSE statements}
TRANSACTION COMMIT
{THEN statements [ELSE statements] | ELSE statements}
TRANSACTION ABORT
Description
Use transaction statements to treat a sequence of file I/O operations as one logical operation with respect
to recovery and visibility to other users. These operations can include file I/O operations or subordinate
transactions.
Note: BASIC accepts PI/open syntax in addition to WebSphere DataStage syntax. You cannot mix both
types of syntax within a program.
Description
Use the TRANSACTION ABORT statement to cancel all file I/O changes made during a transaction.
After the transaction ends, execution continues with the statement following the TRANSACTION ABORT
statement.
Example
The following example shows the use of the TRANSACTION ABORT statement to terminate a
transaction if both the ACCOUNTS RECEIVABLE file and the INVENTORY file cannot be successfully
updated:
PROMPT ’’
OPEN ’ACC.RECV’ TO ACC.RECV ELSE STOP ’NO OPEN
ACC.RECV’
OPEN ’INVENTORY’ TO INVENTORY ELSE STOP ’NO OPEN
INVENTORY’
PRINT ’Customer Id : ’:
INPUT CUST.ID
PRINT ’Item No. : ’:
INPUT ITEM
PRINT ’Amount : ’:
INPUT AMOUNT
END
Description
Use the TRANSACTION COMMIT statement to commit all file I/O changes made during a transaction.
The TRANSACTION COMMIT statement can either succeed or fail. If the TRANSACTION COMMIT
statement succeeds, the THEN statements are executed; any ELSE statements are ignored. If the
TRANSACTION COMMIT statement fails, the ELSE statements, if present, are executed, and control is
transferred to the statement following the TRANSACTION COMMIT statement.
Description
You must have a THEN clause or an ELSE clause, or both, in a TRANSACTION START statement.
If the TRANSACTION START statement successfully begins a transaction, the statements in the THEN
clause are executed. If for some reason WebSphere DataStage is unable to start the transaction, a fatal
error occurs, and you are returned to the command prompt.
TRIM function
Syntax
TRIM (expression [ ,character [ ,option] ] )
Description
Use the TRIM function to remove unwanted characters in expression. If only expression is specified,
multiple occurrences of spaces and tabs are reduced to a single tab or space, and all leading and trailing
spaces and tabs are removed. If expression evaluates to one or more space characters, TRIM returns an
empty string.
If expression evaluates to the null value, null is returned. If option evaluates to the null value, null is
ignored and option R is assumed. If character evaluates to the null value, the TRIM function fails and the
program terminates with a run-time error message.
If NLS is enabled, you can use TRIM to remove other white space characters such as Unicode values
0x2000 through 0x200B, 0x00A0, and 0x3000, marked as TRIMMABLE in the NLS.LC.CTYPE file entry for
the specified locale. For more information about Unicode values, see WebSphere DataStage National
Language Support Guide.
Example
A=" Now is the time for all good men to"
PRINT A
PRINT TRIM(A)
TRIMB function
Syntax
TRIMB (expression)
Description
Use the TRIMB function to remove all trailing spaces and tabs from expression. All other spaces or tabs in
expression are left intact. If expression evaluates to the null value, null is returned.
If NLS is enabled, you can use TRIMB to remove white space characters such as Unicode values 0x2000
through 0x200B, 0x00A0, and 0x3000, marked as TRIMMABLE in the NLS.LC.CTYPE file entry for the
specified locale. For more information about Unicode values, see WebSphere DataStage National Language
Support Guide.
Example
A=" THIS IS A SAMPLE STRING "
PRINT "’":A:"’": " IS THE STRING"
PRINT "’":TRIMB(A):"’":" IS WHAT TRIMB DOES"
END
TRIMBS function
Syntax
TRIMBS (dynamic.array)
CALL -TRIMBS (return.array, dynamic.array)
Description
Use the TRIMBS function to remove all trailing spaces and tabs from each element of dynamic.array.
TRIMBS removes all trailing spaces and tabs from each element and reduces multiple occurrences of
spaces and tabs to a single space or tab.
If dynamic.array evaluates to the null value, null is returned. If any element of dynamic.array is null, null is
returned for that value.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
If NLS is enabled, you can use TRIMBS to remove white space characters such as Unicode values 0x2000
through 0x200B, 0x00A0, and 0x3000, marked as TRIMMABLE in the NLS.LC.CTYPE file entry for the
specified locale. For more information about Unicode values, see WebSphere DataStage National Language
Support Guide.
TRIMF function
Syntax
TRIMF (expression)
Description
Use the TRIMF function to remove all leading spaces and tabs from expression. All other spaces or tabs in
expression are left intact. If expression evaluates to the null value, null is returned.
If NLS is enabled, you can use TRIMF to remove white space characters such as Unicode values 0x2000
through 0x200B, 0x00A0, and 0x3000, marked as TRIMMABLE in the NLS.LC.CTYPE file entry for the
specified locale. For more information about Unicode values, see WebSphere DataStage National Language
Support Guide.
Example
A=" THIS IS A SAMPLE STRING "
PRINT "’":A:"’":" IS THE STRING"
PRINT "’":TRIMF(A):"’":" IS WHAT TRIMF DOES"
END
Description
Use the TRIMFS function to remove all leading spaces and tabs from each element of dynamic.array.
TRIMFS removes all leading spaces and tabs from each element and reduces multiple occurrences of
spaces and tabs to a single space or tab.
If dynamic.array evaluates to the null value, null is returned. If any element of dynamic.array is null, null is
returned for that value.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
If NLS is enabled, you can use TRIMFS to remove white space characters such as Unicode values 0x2000
through 0x200B, 0x00A0, and 0x3000, marked as TRIMMABLE in the NLS.LC.CTYPE file entry for the
specified locale. For more information about Unicode values, see WebSphere DataStage National Language
Support Guide.
TRIMS function
Syntax
TRIMS (dynamic.array)
CALL -TRIMS (return.array, dynamic.array)
Description
Use the TRIMS function to remove unwanted spaces and tabs from each element of dynamic.array.
TRIMS removes all leading and trailing spaces and tabs from each element and reduces multiple
occurrences of spaces and tabs to a single space or tab.
If dynamic.array evaluates to the null value, null is returned. If any element of dynamic.array is null, null is
returned for that value.
If you use the subroutine syntax, the resulting dynamic array is returned as return.array.
If NLS is enabled, you can use TRIMS to remove white space characters such as Unicode values 0x2000
through 0x200B, 0x00A0, and 0x3000, marked as TRIMMABLE in the NLS.LC.CTYPE file entry for the
specified locale. For more information about Unicode values, see WebSphere DataStage National Language
Support Guide.
TTYCTL statement
Syntax
TTYCTL file.variable, code#
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the TTYCTL statement to set terminal device characteristics on Berkeley terminal drivers. code#
specifies the action to take.
file.variable specifies a file previously opened for sequential access to a terminal device. If file.variable
evaluates to the null value, the TTYCTL statement fails and the program terminates with a run-time error
message.
If the action is taken, the THEN statements are executed. If no THEN statements are present, program
execution continues with the next statement.
If an error is encountered during the execution of the TTYCTL operation, or if the file variable is not
open to a terminal device, the ELSE statements are executed; any THEN statements are ignored.
Example
OPENSEQ ’FILE.E’, ’RECORD4’ TO FILE ELSE ABORT
*
TTYCTL FILE, 0
THEN PRINT ’THE FILE IS A TTY’
ELSE PRINT ’THE FILE IS NOT A TTY’
TTYGET statement
Syntax
TTYGET variable [FROM {file.variable | LPTR [n] | MTU [n] } ]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the TTYGET statement to assign the characteristics of a terminal, line printer channel, or tape unit as
a dynamic array to variable. If the FROM clause is omitted, a dynamic array of the terminal characteristics
for your terminal is assigned to variable.
file.variable is a terminal opened for sequential processing with the OPENDEV statement or OPENSEQ
statement. If file.variable is specified, the terminal characteristics for the specified terminal are retrieved.
If the terminal characteristics are retrieved, the THEN statements are executed.
If the device does not exist or cannot be opened, or if no dynamic array is returned, the ELSE statements
are executed; any THEN statements are ignored.
If either file.variable or n evaluates to the null value, the TTYGET statement fails and the program
terminates with a run-time error message.
The best way to access the information in the dynamic array is to include the BASIC code
UNIVERSE.INCLUDE TTY. The syntax for including this file is:
$INCLUDE UNIVERSE.INCLUDE TTY
This file equates each value of the dynamic array to a name, so that each value can be easily accessed in
your program. To take advantage of this code you must call variable tty$. Once this code has been
included in your program, you can use the names to access the values of the dynamic array. To set values
for a terminal line, use the TTYSET statement.
The following table lists the equate names to the values of the dynamic array, and describes each value.
The final columns indicate which values are available on different operating systems: SV indicates System
V, B indicates Berkeley UNIX, and NT indicates Windows NT.
Table 60. TTYGET Statement Values
Value Name Description Availability
SV B NT
Field 1
1 mode.type One of these 3333 3333 33
modes:
MODE$LINE or 0
= line
MODE$RAW or 1
= raw
MODE$CHAR or 2
= character
MODE$EMULATE
or 3 = emulated
2 mode.min Minimum number 3 3 3
of characters before
input.
3 mode.time Minimum time in 3 3 3
milliseconds before
input.
Field 2
1 cc.intr Interrupt character. 3 3 3
-1 undefined.
2 cc.quit Quit character. -1 3 3
undefined.
ECHOE$BS or 1 =
echo as backspace
ECHOE$BSB or 2
= echo as
backspace, space,
backspace
ECHOE$PRINTER
or 3 = echo as a
printer
3 echo.kill ECHOK$KILL or 0 3 3
= kill as kill
character
ECHOK$LF or 1 =
kill as RETURN,
LINEFEED
ECHOK$ERASE or
2 = kill as series of
erases
4 echo.ctrl Set control to echo 3 3
as ^ character
5 echo.lf When echo is off, 3 3 3
echo RETURN as
RETURN,
LINEFEED
Field 8
1 handshake.xon 1 = turns on 3 3 3
X-ON/X-OFF
protocol
0 = turns off
X-ON/X-OFF
protocol
0 = only X-ON
character acts as
X-ON
3 handshake. 1 = when input 3 3 3
tandem buffer is nearly
full, X-OFF is sent
0 = turns off
automatic X-OFF,
X-ON mode
4 handshake.dtr 1 = turns on DTR 3 3
2 = 75 10 = 1800
3 = 110 11 = 2400
4 = 134 12 = 4800
5 = 150 13 = 9600
6 = 200 14 or EXTA
= 19200
7 = 300 15 = EXTB
8 = 600
5 = 5 bits 7 = 7 bits
6 = 6 bits 8 = 8 bits
4 protocol.stop 2 = 2 stopbits 1 = 1 3 3 3
stopbit
5 protocol.output Output parity: 33 33 333
0 = no parity 1 =
even parity
2 = odd parity
6 protocol.input Input parity: 33333 33333 3333
0 = disable input
parity checking
1 = enable input
parity checking
2 = mark parity
errors 3 = mark
parity errors with a
null 4 = ignore
parity errors
7 protocol.strip 1 = strip to 7 bits 0 3 3
= 8 bits
Field 11
1 signals.enable Enable signal keys: 3 3
Interrupt, Suspend,
Quit.
2 signals.flush Flush type-ahead 3 3
buffer.
3 signals.brkkey 0 = break ignored 3 3
1 = break as
interrupt
2 = break as null
TTYSET statement
Syntax
TTYSET dynamic.array [ON {file.variable | LPTR [n] | MTU [n] } ]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the TTYSET statement to set the characteristics of a terminal, line printer channel, or tape unit. If
only dynamic.array is specified, the terminal characteristics for your terminal are set based on the contents
file.variable is a terminal opened for sequential processing with the OPENDEV or OPENSEQ statements. If
file.variable is specified, the terminal characteristics for the specified terminal are set.
n specifies a logical print channel with LPTR or a tape unit with MTU. If n is specified, the characteristics
for the print channel or tape unit are set. n is in the range of 0 through 225 for logical print channels; the
default is 0. n is in the range of 0 through 7 for tape units; the default is 0. On Windows NT you cannot
specify a tape unit.
If the terminal characteristics are set, the THEN statements are executed.
If the device does not exist or cannot be opened, or if no dynamic array is returned, the ELSE statements
are executed; any THEN statements are ignored.
If dynamic.array, file.variable, or n evaluates to the null value, the TTYSET statement fails and the program
terminates with a run-time error message.
To build dynamic.array, get the current values of the terminal line using the TTYGET statement,
manipulate the values, and reset them with the TTYSET statement. The best way to access the
information in the dynamic array is to include the BASIC code UNIVERSE.INCLUDE TTY. The syntax for
including this file is:
$INCLUDE UNIVERSE.INCLUDE TTY
This file equates each value of variable from the TTYGET statement with a name, so that each value can
be easily accessed in your program. To take advantage of this code you must call variable tty$. Once this
code is included in your program, you can use the names to access the values of the dynamic array. The
TTYGET Statement Values table lists the names equated to the values of the dynamic array and describes
the values.
Timeout Handling
You can set the MODE.MIN and MODE.TIME values to define timeouts for read operations over a
communications line. MODE.MIN specifies the minimum number of characters to be received.
MODE.TIME specifies time in tenths of a second. The two values interact to provide four cases that can
be used as follows.
Intercharacter Timer
When you set the values of both MODE.MIN and MODE.TIME to greater than 0, MODE.TIME specifies
the maximum time interval allowed between successive characters received by the communication line in
tenths of a second. Timing begins only after the first character is received.
Blocking Read
When you set the value of MODE.MIN to greater than 0 and MODE.TIME to 0, no time limit is set, so
the read operation waits until the specified number of characters have been received (or a newline in the
case of READSEQ) statement.
Read Timer
When you set the value of MODE.MIN to 0 and MODE.TIME to greater than 0, MODE.TIME specifies
how long the read operation waits for a character to arrive before timing out. If no characters are
Nonblocking Read
When you set the values of both MODE.MIN and MODE.TIME to 0, data is read as it becomes available.
The read operation returns immediately.
v If any characters are received:
– READBLK returns as many characters as specified in the blocksize argument, or all the characters
received, whichever is fewer.
– READSEQ returns characters up to the first newline, or all the characters received if no newline is
received.
v If no characters are received, READSEQ and READBLK use the ELSE clause if there is one.
UNASSIGNED function
Syntax
UNASSIGNED (variable)
Description
Use the UNASSIGNED function to determine if variable is unassigned. UNASSIGNED returns 1 (true) if
variable is unassigned. It returns 0 (false) if variable is assigned a value, including the null value.
Example
A = "15 STATE STREET"
C = 23
X = UNASSIGNED(A)
Y = UNASSIGNED(B)
Z = UNASSIGNED(C)
PRINT X,Y,Z
UNICHAR function
Syntax
UNICHAR (unicode)
Description
Use the UNICHAR function to generate a single character from a Unicode value.
unicode is a decimal number from 0 through 65535 that is the value of the character you want to generate.
If unicode is invalid, an empty string is returned. If unicode evaluates to the null value, null is returned.
The UNICHAR function operates the same way whether NLS mode is enabled or not.
Note: Use BASIC @Variables to generate WebSphere DataStage system delimiters. Do not use the
UNICHAR function.
Description
Use the UNICHARS function to generate a dynamic array of characters from a dynamic array of Unicode
values.
dynamic.array is an array of decimal Unicode values separated by system delimiters. If any element of
dynamic.array is invalid, an empty string is returned for that element. If dynamic.array evaluates to the null
value, null is returned. If any element of dynamic.array is null, null is returned for that element.
The UNICHARS function operates the same way whether NLS mode is enabled or not.
Note: Use BASIC @Variables to generate WebSphere DataStage system delimiters. Do not use the
UNICHARS function.
UNISEQ function
Syntax
UNISEQ (expression)
Description
The first character of expression is converted to its Unicode value, that is, a hexadecimal value in the
range 0x0000 through 0x1FFFF. If expression is invalid, for example, an incomplete internal string, an
empty string is returned. If expression evaluates to the null value, null is returned.
The UNISEQ function operates the same way whether NLS mode is enabled or not.
Note: UNISEQ does not map system delimiters. For example, UNISEQ(″û″) returns 251 (0x00FB), and
UNISEQ(@TM) returns 63739 (0xF8FB). The Unicode value returned is the internal representation of the
text mark character that is mapped to a unique area so that it is not confused with any other character.
Note that this behaves differently from SEQ(@TM), which returns 251.
For more information about Unicode values and tokens defined for system delimiters, see WebSphere
DataStage National Language Support Guide.
UNISEQS function
Syntax
UNISEQS (dynamic.array)
Description
Use the UNISEQS function to generate an array of Unicode values from a dynamic array of characters.
dynamic.array specifies an array of characters with the elements separated by system delimiters. The first
character of each element of dynamic.array is converted to its Unicode value, a hexadecimal value in the
The UNISEQS function operates the same way whether NLS mode is enabled or not.
Note: UNISEQS does not map system delimiters. For example, UNISEQS(″û″) returns 251 (0x00FB), and
UNISEQS(@TM) returns 63739 (0xF8FB). The Unicode value returned is the internal representation of the
text mark character that is mapped to a unique area so that it is not confused with any other character.
Note that this behaves differently from SEQ(@TM), which returns 251.
For more information about Unicode values and tokens defined for system delimiters, see WebSphere
DataStage National Language Support Guide.
UNLOCK statement
Syntax
UNLOCK [expression]
Description
Use the UNLOCK statement to release a process lock set by the LOCK statement.
expression specifies an integer from 0 through 63. If expression is not specified, all locks are released (see
the LOCK statement).
If expression evaluates to an integer outside the range of 0 through 63, an error message appears and no
action is taken.
If expression evaluates to the null value, the UNLOCK statement fails and the program terminates with a
run-time error message.
Examples
The next example unlocks all locks set during the current login session:
UNLOCK
UPCASE function
Syntax
UPCASE (expression)
Description
Use the UPCASE function to change all lowercase letters in expression to uppercase. If expression evaluates
to the null value, null is returned.
Example
A="This is an example of the UPCASE function: "
PRINT A
PRINT UPCASE(A)
UPRINT statement
Syntax
UPRINT [ ON print.channel ] [ print.list ]
Description
In NLS mode, use the UPRINT statement to print data that was mapped to an external format using
OCONV mapname. The UPRINT statement subsequently sends the mapped data to the screen, a line
printer, or another print file with no further mapping.
The ON clause specifies the logical print channel to use for output. print.channel is an expression that
evaluates to a number from -1 through 255. If you do not use the ON clause, logical print channel 0 is
used, which prints to the user’s terminal if PRINTER OFF is set (see the PRINTER statement). If
print.channel evaluates to the null value, the PRINT statement fails and the program terminates with a
run-time error message. Logical print channel -1 prints the data on the screen, regardless of whether a
PRINTER ON statement has been executed.
You can specify HEADING, FOOTING, PAGE, and PRINTER CLOSE statements for each logical print
channel. The contents of the print files are printed in order by logical print channel number.
print.list can contain any BASIC expression. The elements of the list can be numeric or character strings,
variables, constants, or literal strings; the null value, however, cannot be printed. The list can consist of a
single expression or a series of expressions separated by commas ( , ) or colons ( : ) for output formatting.
If no print.list is designated, a blank line is printed.
Expressions separated by commas are printed at preset tab positions. The default tab stop setting is 10
characters. For information about changing the default setting, see the TABSTOP statement. Use multiple
commas together for multiple tabulations between expressions.
Expressions separated by colons are concatenated. That is, the expression following the colon is printed
immediately after the expression preceding the colon. To print a list without a LINEFEED and RETURN,
end print.list with a colon ( : ).
If NLS is disabled, the UPRINT statement behaves like the PRINT statement.
For more information about maps, see WebSphere DataStage National Language Support Guide.
WEOF statement
Syntax
WEOF [UNIT (mtu) ] {THEN statements [ELSE statements] |
ELSE statements}
The UNIT clause specifies the number of the tape drive unit. Tape unit 0 is used if no unit is specified.
mtu is an expression that evaluates to a three-digit code (decimal). Although the mtu expression is a
function of the UNIT clause, the WEOF statement uses only the third digit (the u). Its value must be in
the range of 0 through 7 (see the READT statement for details on the mtu expression). If mtu evaluates to
the null value, the WEOF statement fails and the program terminates with a run-time error message.
Before a WEOF statement is executed, a tape drive unit must be attached (assigned) to the user. Use the
ASSIGN command to assign a tape unit to a user. If no tape unit is attached or if the unit specification is
incorrect, the ELSE statements are executed.
The STATUS function returns 1 if WEOF takes the ELSE clause, otherwise it returns 0.
Example
WEOF UNIT(007) ELSE PRINT "OPERATION NOT COMPLETED."
WEOFSEQ statement
Syntax
WEOFSEQ file.variable [ON ERROR statements]
Description
Use the WEOFSEQ statement to write an end-of-file (EOF) mark in a file opened for sequential access.
The end-of-file mark is written at the current position and has the effect of truncating the file at this
point. Any subsequent READSEQ statement has its ELSE statements executed.
file.variable specifies a file opened for sequential access. If file.variable evaluates to the null value, the
WEOFSEQ statement fails and the program terminates with a run-time error message.
Note: On Windows NT systems, you cannot use the WEOFSEQ statement with a diskette drive that you
opened with the OPENDEV statement. For 1/4- inch cartridge tape drives (60 MB or 150 MB) you can
use WEOFSEQ to write an end-of-file (EOF) mark at the beginning of the data or after a write.
The ON ERROR clause is optional in the WEOFSEQ statement. The ON ERROR clause lets you specify
an alternative for program termination when a fatal error is encountered during processing of the
WEOFSEQ statement.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
See the OPENSEQ statement, READSEQ statement, and WRITESEQ statements for more information
about sequential file processing.
Note: Some systems do not support the truncation of disk files. WEOFSEQ is ignored on these systems,
except that WEOFSEQ always works at the beginning of a file.
Example
The following example writes an end-of-file mark on the record RECORD in the file TYPE1:
OPENSEQ ’TYPE1’,’RECORD’ TO FILE ELSE STOP
WEOFSEQ FILE
WRITE statements
Syntax
WRITE[U] expression {ON | TO} [file.variable,] record.ID[ON ERROR statements] [LOCKED statements]
[THEN statements] [ELSE statements]
WRITEV[U] expression {ON | TO} [file.variable,] record.ID, field#[ON ERROR statements] [LOCKED statements]
[THEN statements] [ELSE statements]
Description
Use WRITE statements to write new data to a record in a WebSphere DataStage file. The value of
expression replaces any data previously stored in the record.
Table 61. Effects of WRITE Statements
Use this statement... To do this...
WRITE Write to a record.
WRITEU Write to a record, retaining an update record lock.
WRITEV Write to a field.
WRITEVU Write to a field, retaining an update record lock.
If expression evaluates to the null value, the WRITE statement fails and the program terminates with a
run-time error message.
file.variable specifies an open file. If file.variable is not specified, the default file is assumed (for more
information on default files, see the OPEN statement. If the file is neither accessible nor open, the
program terminates with a run-time error message, unless ELSE statements are specified.
The system searches the file for the record specified by record.ID. If the record is not found, WRITE
creates a new record.
If file.variable, record.ID, or field# evaluates to the null value, all WRITE statements (WRITE, WRITEU,
WRITEV, WRITEVU) fail and the program terminates with a run-time error message.
The new value is written to the record, and the THEN statements are executed. If no THEN statements
are specified, execution continues with the statement following the WRITE statement. If WRITE fails, the
ELSE statements are executed; any THEN statements are ignored.
The WRITE statement does not strip trailing field marks enclosing empty strings from expression. Use the
MATWRITE statements statement if that operation is required.
Tables
If the file is a table, the effective user of the program must have SQL INSERT and UPDATE privileges to
read records in the file. For information about the effective user of a program, see the AUTHORIZATION
statement.
If the OPENCHK configurable parameter is set to TRUE, or if the file is opened with the OPENCHECK
statement, all SQL integrity constraints are checked for every write to an SQL table. If an integrity check
fails, the WRITE statement uses the ELSE clause. Use the ICHECK function to determine what specific
integrity constraint caused the failure.
NLS Mode
WRITE and other BASIC statements that perform I/O operations map internal data to the external
character set using the appropriate map for the output file.
WebSphere DataStage substitutes the file map’s unknown character for any unmappable character. The
results of the WRITE statements depend on the following:
v The inclusion of the ON ERROR clause
v The setting of the NLSWRITEELSE parameter in the uvconfig file
v The location of the unmappable character
The values returned by the STATUS function and the results are as follows:
Table 62. Values and results
STATUS Value and Results ON ERROR and Parameter Setting Unmappable Character Location
3 The WRITE fails, no records ON ERROR Record ID
written.
Data
4 The WRITE fails, no records
written.
Program terminates with a run-time No ON ERROR, and Record ID or data
error message.
NLSWRITEELSE = 1
Program terminates with a run-time No ON ERROR, NLSWRITEELSE = 0 Record ID
error message.
Data
Record is written with unknown
characters; lost data.
For more information about unmappable characters , see WebSphere DataStage National Language Support
Guide.
Use the STATUS function after a WRITE statement is executed, to determine the result of the operation,
as follows:
0 The record was locked before the WRITE operation.
-2 The record was unlocked before the WRITE operation.
The ON ERROR clause is optional in WRITE statements. Its syntax is the same as that of the ELSE clause.
The ON ERROR clause lets you specify an alternative for program termination when a fatal error is
encountered during processing of the WRITE statement.
If a fatal error occurs, and the ON ERROR clause was not specified or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
The LOCKED clause is optional, but recommended. Its format is the same as that of the ELSE clause.
The LOCKED clause handles a condition caused by a conflicting lock (set by another user) that prevents
the WRITE statement from processing. The LOCKED clause is executed if one of the following conflicting
locks exists:
v Exclusive file lock
v Intent file lock
v Shared file lock
v Update record lock
v Shared record lock
If the WRITE statement does not include a LOCKED clause, and a conflicting lock exists, the program
pauses until the lock is released.
If a LOCKED clause is used, the value returned by the STATUS function is the terminal number of the
user who owns the conflicting lock.
Use the WRITEU statement to update a record without releasing the update record lock set by a previous
READU statement (see the “READ statements” on page 273). To release the update record lock set by a
READU statement and maintained by a WRITEU statement, you must use a RELEASE statement, WRITE
statements, MATWRITE statements, or WRITEV statement. If you do not explicitly release the lock, the
If expression evaluates to the null value, the WRITEU statement fails and the program terminates with a
run-time error message.
Use the WRITEV statement to write a new value to a specified field in a record. The WRITEV statement
requires that field# be specified. field# is the number of the field to which expression is written. It must be
greater than 0. If either the record or the field does not exist, WRITEV creates them.
If expression evaluates to the null value, null is written to field#, provided that the field allows nulls. If the
file is an SQL table, existing SQL security and integrity constraints must allow the write.
Use the WRITEVU statement to update a specified field in a record without releasing the update record
lock set by a previous READU statement (see the READ statement). The WRITEVU syntax is like that of
the WRITEV and WRITEU statements.
If expression evaluates to the null value, null is written to field#, provided that the field allows nulls. If the
file is an SQL table, existing SQL security and integrity constraints must allow the write.
Remote Files
If in a transaction you try to write to a remote file over and the write statement fails, the transaction is
rolled back, and the program terminates with a run-time error message.
Example
CLEAR
DATA "ELLEN","KRANZER","3 AMES STREET","CAMBRIDGE"
DATA "MA","02139","SAILING"
OPEN ’’,’SUN.MEMBER’ TO FILE ELSE
PRINT "COULD NOT OPEN FILE"
STOP
END
PRINT "ENTER YOUR FIRST NAME"
INPUT FNAME
PRINT "ENTER YOUR LAST NAME"
INPUT LNAME
PRINT "ENTER YOUR ADDRESS (PLEASE WAIT FOR PROMPTS)"
PRINT "STREET ADDRESS"
INPUT STREET
PRINT "ENTER CITY"
INPUT CITY
PRINT "ENTER STATE"
INPUT STATE
PRINT "ENTER ZIP CODE"
INPUT ZIP
PRINT "ENTER YOUR INTERESTS"
INPUT INTERESTS
RECORD<1>=LNAME
RECORD<2>=FNAME
RECORD<3>=STREET
RECORD<4>=CITY
RECORD<5>=STATE
RECORD<6>=ZIP
RECORD<7>=1989
1 records listed.
WRITEBLK statement
Syntax
WRITEBLK expression ON file.variable{THEN statements [ELSE statements] | ELSE statements}
Description
Use the WRITEBLK statement to write a block of data to a file opened for sequential processing. Each
WRITEBLK statement writes the value of expression starting at the current position in the file. The current
position is incremented to beyond the last byte written. WRITEBLK does not add a newline at the end of
the data.
Note: On Windows NT systems, if you use the WRITEBLK statement to write to a 1/4-inch cartridge
tape (60 MB or 150 MB) that you opened with the OPENDEV statement, you must specify the block size
as 512 bytes or a multiple of 512 bytes.
The value of expression is written to the file, and the THEN statements are executed. If no THEN
statements are specified, program execution continues with the next statement. If the file cannot be
accessed or does not exist, the ELSE statements are executed; any THEN statements are ignored.
If either expression or file.variable evaluates to the null value, the WRITEBLK statement fails and the
program terminates with a run-time error message.
If NLS is enabled, the data written is mapped using the appropriate output file map. For more
information about maps, see WebSphere DataStage National Language Support Guide.
Example
OPENSEQ ’FILE.E’,’RECORD4’ TO FILE ELSE ABORT
WEOFSEQ FILE
DATA1=’ONE’
DATA2=’TWO’
*
WRITEBLK DATA1 ON FILE ELSE ABORT
WRITEBLK DATA2 ON FILE ELSE ABORT
* These two lines write two items to RECORD4 in
FILE.E without
* inserting a newline between them.
WEOFSEQ FILE
WRITELIST statement
Syntax
WRITELIST dynamic.array ON listname
Description
Use the WRITELIST statement to save a list as a record in the &SAVEDLISTS& file.
dynamic.array is an expression that evaluates to a string made up of elements separated by field marks. It
is the list to be saved.
or
record.ID account.name
record.ID is the record ID of the select list created in the &SAVEDLISTS& file. If listname includes
account.name, the &SAVEDLISTS& file of the specified account is used instead of the one in the local
account. If record.ID exists, WRITELIST overwrites the contents of the record.
If either dynamic.array or listname evaluates to the null value, the WRITELIST statement fails and the
program terminates with a run-time error message.
WRITESEQ statement
Syntax
WRITESEQ expression {ON | TO} file.variable [ON ERROR
statements]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the WRITESEQ statement to write new lines to a file opened for sequential processing. WebSphere
DataStage keeps a pointer to the current position in the file while it is open for sequential processing. The
OPENSEQ statement sets this pointer to the first byte of the file, and it is advanced by the READSEQ
statement, READBLK statement, WRITESEQ, and WRITEBLK statements.
WRITESEQ writes the value of expression followed by a newline to the file. The data is written at the
current position in the file. The pointer is set to the position following the newline. If the pointer is not at
the end of the file, WRITESEQ overwrites any existing data byte by byte (including the newline), starting
from the current position.
If expression or file.variable evaluates to the null value, the WRITESEQ statement fails and the program
terminates with a run-time error message.
After executing a WRITESEQ statement, you can use the STATUS function to determine the result of the
operation:
0 The record was locked before the WRITESEQ operation.
-2 The record was unlocked before the WRITESEQ operation.
File Buffering
Normally WebSphere DataStage uses buffering for sequential input and output operations. If you use the
NOBUF statement after an OPENSEQ statement, buffering is turned off and writes resulting from the
WRITESEQ statement are performed right away.
You can also use the FLUSH statement after a WRITESEQ statement to cause all buffers to be written
right away.
For more information about buffering, see the FLUSH and NOBUF statements.
The ON ERROR clause is optional in the WRITESEQ statement. The ON ERROR clause lets you specify
an alternative for program termination when a fatal error is encountered while the WRITESEQ statement
is being processed.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
If NLS is enabled, WRITESEQ and other BASIC statements that perform I/O operations always map
internal data to the external character set using the appropriate map for the output file. For more
information about maps, see WebSphere DataStage National Language Support Guide.
Example
DATA ’NEW ITEM 1’, ’NEW ITEM 2’
OPENSEQ ’FILE.E’, ’RECORD1’ TO FILE ELSE ABORT
READSEQ A FROM FILE ELSE STOP
*
WRITESEQF statement
Syntax
WRITESEQF expression {ON | TO} file.variable [ON ERROR
statements]
{THEN statements [ELSE statements] | ELSE statements}
Description
Use the WRITESEQF statement to write new lines to a file opened for sequential processing, and to
ensure that data is physically written to disk (that is, not buffered) before the next statement in the
program is executed. The sequential file must be open, and the end-of-file marker must be reached before
you can write to the file. You can use the FILEINFO function to determine the number of the line about
to be written.
Normally, when you write a record using the WRITESEQ statement, the record is moved to a buffer that
is periodically written to disk. If a system failure occurs, you could lose all the updated records in the
buffer. The WRITESEQF statement forces the buffer contents to be written to disk; the program does not
execute the statement following the WRITESEQF statement until the buffer is successfully written to disk.
A WRITESEQF statement following several WRITESEQ statements ensures that all buffered records are
written to disk.
WRITESEQF is intended for logging applications and should not be used for general programming. It
increases the disk I/O of your program and therefore degrades performance.
The value of expression is written to the file as the next line, and the THEN statements are executed. If
THEN statements are not specified, program execution continues with the next statement. If the specified
file cannot be accessed or does not exist, the ELSE statements are executed; any THEN statements are
ignored.
If expression or file.variable evaluates to the null value, the WRITESEQF statement fails and the program
terminates with a run-time error message.
If NLS is enabled, WRITESEQF and other BASIC statements that perform I/O operations always map
internal data to the external character set using the appropriate map for the output file. For more
information about maps, see WebSphere DataStage National Language Support Guide.
The ON ERROR clause is optional in the WRITESEQF statement. Its syntax is the same as that of the
ELSE clause. The ON ERROR clause lets you specify an alternative for program termination when a fatal
error is encountered while the WRITESEQF statement is being processed.
If a fatal error occurs, and the ON ERROR clause was not specified, or was ignored (as in the case of an
active transaction), the following occurs:
v An error message appears.
v Any uncommitted transactions begun within the current execution environment roll back.
v The current program terminates.
v Processing continues with the next statement of the previous execution environment, or the program
returns to the command prompt.
If the ON ERROR clause is used, the value returned by the STATUS function is the error number.
Key 14 (FINFO$CURRENTLINE) of the FILEINFO function can be used to determine the number of the
line about to be written to the file.
Example
In the following example, the print statement following the WRITESEQF statement is not executed until
the record is physically written to disk:
WRITESEQF ACCOUNT.REC TO ACCOUNTS.FILE
THEN WRITTEN = TRUE
ELSE STOP "ACCOUNTS.FILE FORCE WRITE ERROR"
PRINT "Record written to disk."
WRITET statement
Syntax
WRITET [UNIT (mtu)] variable{THEN statements [ELSE statements] | ELSE statements}
Description
Use the WRITET statement to write a tape record on a magnetic tape. The value of variable becomes the
next tape record. variable is an expression that evaluates to the text to be written to tape.
The UNIT clause specifies the number of the tape drive unit. Tape unit 0 is used if no unit is specified. If
the UNIT clause is used, mtu is an expression that evaluates to a code made up of three decimal digits, as
shown in the following table:
1 = EBCDIC conversion
The mtu expression is read from right to left. If mtu evaluates to a one-digit code, it represents the tape
unit number. If mtu evaluates to a two-digit code, the rightmost digit represents the unit number and the
digit to its left is the track number.
If either mtu or variable evaluates to the null value, the WRITET statement fails and the program
terminates with a run-time error message.
Each tape record is written completely before the next record is written. The program waits for the
completion of data transfer to the tape before continuing.
Before a WRITET statement is executed, a tape drive unit must be attached (assigned) to the user. Use the
ASSIGN command to assign a tape unit to a user. If no tape drive unit is attached or if the unit
specification is incorrect, the ELSE statements are executed.
The largest record that the WRITET statement can write is system-dependent. If the actual record is
larger, bytes beyond the system byte limit are not written.
Note: WebSphere DataStage BASIC does not generate tape labels for the tape file produced with the
WRITET statement.
The STATUS function returns 1 if READT takes the ELSE clause, otherwise it returns 0.
If NLS is enabled, WRITET and other BASIC statements that perform I/O operations always map
external data to the internal character set using the appropriate map for the file. The map defines the
external character set for the file that is used to input data on a keyboard, display data on a screen, and
so on. For more information about maps, see WebSphere DataStage National Language Support Guide.
PIOPEN Flavor
If you have a program that specifies the syntax UNIT ndmtu, the nd elements are ignored by the compiler
and no errors are reported.
Examples
The next example writes the numeric constant 50 to tape drive 2, a 9-track tape with no conversion:
WRITET UNIT(002) "50" ELSE PRINT "COULD NOT WRITE"
WRITEV statement
Use the WRITEV statement to write on the contents of a specified field of a record of a WebSphere
DataStage file. For details, see the statement.
WRITEVU statement
Use the WRITEVU statement to maintain an update record lock while writing on the contents of a
specified field of a record of a WebSphere DataStage file. For details, see the WRITE statement.
XLATE function
Syntax
XLATE ( [DICT] filename, record.ID, field#, control.code)
Description
Use the XLATE function to return the contents of a field or a record in a WebSphere DataStage file.
XLATE opens the file, reads the record, and extracts the specified data.
filename is an expression that evaluates to the name of the remote file. If XLATE cannot open the file, a
run-time error occurs, and XLATE returns an empty string.
record.ID is an expression that evaluates to the ID of the record to be accessed. If record.ID is multi-valued,
the translation occurs for each record ID and the result is multi-valued (system delimiters separate data
translated from each record).
field# is an expression that evaluates to the number of the field from which the data is to be extracted. If
field# is -1, the entire record is returned, except for the record ID.
control.code is an expression that evaluates to a code specifying what action to take if data is not found or
is the null value. The possible control codes are:
X (default) Returns an empty string if the record does not exist or data cannot be found.
V Returns an empty string and produces an error message if the record does not exist or data
cannot be found.
C Returns the value of record.ID if the record does not exist or data cannot be found.
N Returns the value of record.ID if the null value is found.
The returned value is lowered. For example, value marks in the original field become subvalue marks in
the returned value. For more information, see the LOWER function.
If filename, record.ID, or field# evaluates to the null value, the XLATE function fails and the program
terminates with a run-time error message. If control.code evaluates to the null value, null is ignored and X
is used.
XTD function
Syntax
XTD (string)
Description
Use the XTD function to convert a string of hexadecimal characters to an integer. If string evaluates to the
null value, null is returned.
Example
Y = "0019"
Z = XTD (Y)
PRINT Z
Compiler Directives
* Identifies a line as a comment line. Same as the !, $*, and REM statements.
! Identifies a line as a comment line. Same as the *, $*, and REM statements.
#INCLUDE
Inserts and compiles WebSphere DataStage BASIC source code from another program into the
program being compiled. Same as the $INCLUDE and INCLUDE statements.
$* Identifies a line as a comment line. Same as the *, !, and REM statements.
$CHAIN
Inserts and compiles WebSphere DataStage BASIC source code from another program into the
program being compiled.
$COPYRIGHT
Inserts comments into the object code header. (WebSphere DataStage supports this statement for
compatibility with existing software.)
$DEFINE
Defines a compile time symbol.
$EJECT
Begins a new page in the listing record. (WebSphere DataStage supports this statement for
compatibility with existing software.) Same as the $PAGE statement.
$IFDEF
Tests for the definition of a compile time symbol.
Declarations
COMMON
Defines a storage area in memory for variables commonly used by programs and external
subroutines.
DEFFUN
Defines a user-written function.
DIMENSION
Declares the name, dimensionality, and size constraints of an array variable.
FUNCTION
Identifies a user-written function.
PROGRAM
Identifies a program.
SUBROUTINE
Identifies a series of statements as a subroutine.
Assignments
ASSIGNED( )
Determines if a variable is assigned a value.
CLEAR
Assigns a value of 0 to specified variables.
LET Assigns a value to a variable.
File I/O
AUTHORIZATION
Specifies the effective run-time UID (user identification) number of the program.
BEGIN TRANSACTION
Indicates the beginning of a set of statements that make up a single transaction.
BSCAN
Scans the leaf-nodes of a B-tree file (type 25) or a secondary index.
CLEARFILE
Erases all records from a file.
CLOSE
Writes data written to the file physically on the disk and releases any file or update locks.
COMMIT
Commits all changes made during a transaction, writing them to disk.
DELETE
Deletes a record from a WebSphere DataStage file.
DELETEU
Deletes a record from a previously opened file.
END TRANSACTION
Indicates where execution should continue after a transaction terminates.
FILELOCK
Sets a file update lock on an entire file to prevent other users from updating the file until this
program releases it.
FILEUNLOCK
Releases file locks set by the FILELOCK statement.
INDICES( )
Returns information about the secondary key indexes in a file.
MATREAD
Assigns the data stored in successive fields of a record from a WebSphere DataStage file to the
consecutive elements of an array.
MATREADL
Sets a shared read lock on a record, then assigns the data stored in successive fields of the record
to the consecutive elements of an array.
MATREADU
Sets an exclusive update lock on a record, then assigns the data stored in successive fields of the
record to the consecutive elements of an array.
MATWRITE
Assigns the data stored in consecutive elements of an array to the successive fields of a record in
a WebSphere DataStage file.
Tape I/O
READT
Assigns the contents of the next record from a magnetic tape unit to the named variable.
REWIND
Rewinds the magnetic tape to the beginning of the tape.
WEOF Writes an end-of-file mark to a magnetic tape.
WRITET
Writes the contents of a record onto magnetic tape.
String Handling
ALPHA( )
Determines whether the expression is an alphabetic or nonalphabetic string.
CATS( )
Concatenates elements of two dynamic arrays.
CHANGE( )
Substitutes an element of a string with a replacement element.
CHECKSUM( )
Returns a cyclical redundancy code (a checksum value).
COL1( )
Returns the column position immediately preceding the selected substring after a BASIC FIELD
function is executed.
COL2( )
Returns the column position immediately following the selected substring after a BASIC FIELD
function is executed.
COMPARE( )
Compares two strings for sorting.
CONVERT
Converts specified characters in a string to designated replacement characters.
NLS
$MAP Directs the compiler to specify the map for the source code.
AUXMAP
Assigns the map for the auxiliary printer to print unit 0 (i.e., the terminal).
BYTE( )
Generates a string made up of a single byte.
BYTELEN( )
Generates the number of bytes contained in the string value in an expression.
BYTETYPE( )
Determines the function of a byte in a character.
BYTEVAL( )
Retrieves the value of a byte in a string value in an expression.
FMTDP( )
Formats data for output in display positions rather than character lengths.
FMTSDP( )
Formats elements of a dynamic array for output in display positions rather than character
lengths.
Mathematical Functions
ABS( )(
Returns the absolute (positive) numeric value of an expression.
ABSS( )
Creates a dynamic array containing the absolute values of a dynamic array.
ACOS( )
Calculates the trigonometric arc-cosine of an expression.
ADDS( )
Adds elements of two dynamic arrays.
ASIN( )
Calculates the trigonometric arc-sine of an expression.
ATAN( )
Calculates the trigonometric arctangent of an expression.
BITAND( )
Performs a bitwise AND of two integers.
BITNOT( )
Performs a bitwise NOT of two integers.
Relational Functions
ANDS( )
Performs a logical AND on elements of two dynamic arrays.
EQS( )
Compares the equality of corresponding elements of two dynamic arrays.
GES( )
Indicates when elements of one dynamic array are greater than or equal to corresponding
elements of another dynamic array.
fGTS( )
Indicates when elements of one dynamic array are greater than corresponding elements of
another dynamic array.
IFS( ) Evaluates a dynamic array and creates another dynamic array on the basis of the truth or falsity
of its elements.
ISNULL( )
Indicates when a variable is the null value.
ISNULLS( )
Indicates when an element of a dynamic array is the null value.
LES( ) Indicates when elements of one dynamic array are less than or equal to corresponding elements
of another dynamic array.
LTS( ) Indicates when elements of one dynamic array are less than corresponding elements of another
dynamic array.
NES( )
Indicates when elements of one dynamic array are not equal to corresponding elements of
another dynamic array.
NOT( )
Returns the complement of the logical value of an expression.
NOTS( )
Returns the complement of the logical value of each element of a dynamic array.
ORS( )
Performs a logical OR on elements of two dynamic arrays.
System
DATE( )
Returns the internal system date.
DEBUG
Invokes RAID, the interactive WebSphere DataStage BASIC debugger.
ERRMSG
Prints a formatted error message from the ERRMSG file.
INMAT( )
Used with the MATPARSE, MATREAD, and MATREADU statements to return the number of
array elements or with the OPEN statement to return the modulo of a file.
ITYPE( )
Returns the value resulting from the evaluation of an I-descriptor.
Miscellaneous
CLEARPROMPTS
Clears the value of the in-line prompt.
EOF(ARG.)
Checks whether the command line argument pointer is past the last command line argument.
FILEINFO( )
Returns information about the specified file’s configuration.
ILPROMPT( )
Evaluates strings containing in-line prompts.
GET(ARG.)
Retrieves a command line argument.
SEEK (ARG.)
Moves the command line argument pointer.
This appendix describes the correlative and conversion codes used in dictionary entries and with the
ICONV, ICONVS, OCONV, and OCONVS functions in BASIC. Use conversion codes with the ICONV
function when converting data to internal storage format and with the OCONV function when converting
data from its internal representation to an external output format. Read this entire appendix and both the
ICONV function and OCONV function sections before attempting to perform internal or external data
conversion.
Note: If you try to convert the null value, null is returned and the STATUS function returns 1 (invalid
data).
The NLS extended syntax is supported only for Release 9.4.1 and above.
Algebraic Functions
Format
A [ ; ] expression
The A code converts A codes into F codes in order to perform mathematical operations on the field
values of a record, or to manipulate strings. The A code functions in the same way as the F code but is
easier to write and to understand.
A function
Table 67. A function
R(exp) Remainder after integer division of the first operand by
the second. For example, R(2,″5″) returns the remainder
when field 2’s value is divided by 5.
S(exp) Sum all multi-values in exp. For example, S(6) sums the
multi-values of field 6.
IN(exp) Test for the null value.
[] Extract substring. Field numbers, literal numbers, or
expressions can be used as arguments within the
brackets. For example, if the value of field 3 is 9, then
7[″2″,3] returns the second through ninth characters of
field 7. The brackets are part of the syntax and must be
typed.
An arithmetic operator
* Multiply operands.
/ Divide operands. Division always returns an integer result: for example, ″3″ / ″2″ evaluates to 1,
not to 1.5.
+ Add operands.
- Subtract operands.
: Concatenate operands.
A relational operator
= Equal to
< Less than
A conditional operator
AND Logical AND
OR Logical OR
In most cases F and A codes do not act on a data string passed to them. The code specification itself
contains all the necessary data (or at least the names of fields that contain the necessary data). So the
following A codes produce identical F codes, which in turn assign identical results to X:
X = OCONV( "123", "A;’1’ + ’2’" )
X = OCONV( "", "A;’1’ + ’2’" )
X = OCONV( @ID, "A;’1’ + ’2’" )
X = OCONV( "The quick brown fox jumped over a lazy
dog’s
back","A;’1’ + ’2’" )
The data strings passed to the A code-123, the empty string, the record ID, and ″The quick brown fox...″
string-simply do not come into play. The only possible exception occurs when the user includes the LPV
(load previous value) special operand in the A or F code. The following example adds the value 5 and
the previous value 123 to return the sum 128:
X = OCONV( "123", "A;’5’ + LPV" )
It is almost never right to call an A or F code using the vector conversion functions OCONVS and
ICONVS. In the following example, Y = 123V456V789:
X = OCONVS( Y, "A;’5’ + ’2’ )
The statement says, ″For each value of Y, call the A code to add 5 and 2.″ (V represents a value mark.)
The A code gets called three times, and each time it returns the value 7. X, predictably, gets assigned 7.
The scalar OCONV function returns the same result in much less time.
What about correlatives and conversions within an A or F code? Since any string in the A or F code can
be multi-valued, the F code calls the vector functions OCONVS or ICONVS any time it encounters a
secondary correlative or conversion. In the following example, the F code-itself called only once-calls
OCONVS to ensure that the G code gets performed on each value of @RECORD< 1 >. X is assigned the
result cccVfff:
@RECORD< 1 > = aaa*bbb*cccVddd*eee*fff
X = OCONV( "", "A;1(G2*1)"
The value mark is reserved to separate individual code specifications where multiple successive
conversions must be performed.
The following dictionary entry specifies that the substring between the first and second asterisks of the
record ID should be extracted, then the first four characters of that substring should be extracted, then
the masked decimal conversion should be applied to that substring:
001: D
002: 0
003: G1*1VT1,4VMD2
004: Foo
005: 6R
006: S
are invalid. The first returns an empty string (the original value) and a status of 2. The second returns the
record ID; if the STATUS function were accessible from dictionary entries, it would also be set to 2.
Concatenation
Format
C [ ; ] expression1 c expression2 [ c expression3 ] ...
c is the character to be inserted between the fields. Any nonnumeric character (except system delimiters)
is valid, including a blank. A semicolon ( ; ) is a reserved character that means no separation character is
to be used. Two separators cannot follow in succession, with the exceptions of semicolons and blanks.
expression is a field number and requests the contents of that field; or any string enclosed in single
quotation marks ( ’ ), double quotation marks ( ″ ), or backslashes ( \ ); or an asterisk ( * ), which
specifies the data value being converted.
Note: When the C conversion is used in a field descriptor in a file dictionary, the field number in the
LOC or A/AMC field of the descriptor should be 0. If it is any other number and the specified field
contains an empty string, the concatenation is not performed.
Examples
The assumption that anything following a nonseparator asterisk is a separator seems egregious, so the C
code implements:
PRINT ICONV (x, "C*1*2*3" )
xone*two*threeVfour
Date Conversion
Format
D [ n ] [ *m ] [ s ] [ fmt [ [ f1, f2, f3, f4, f5 ] ] ] [ E ] [ L ]
The D code converts input dates from conventional formats to an internal format for storage. It also
converts internal dates back to conventional formats for output. When converting an input date to
internal format, date conversion specifies the format you use to enter the date. When converting internal
dates to external format, date conversion defines the external format for the date.
If the D code does not specify a year, the current year is assumed. If the code specifies the year in
two-digit form, the years from 0 through 29 mean 2000 through 2029, and the years from 30 through 99
mean 1930 through 1999.
You can set the default date format with the DATE.FORMAT command. A system-wide default date
format can be set in the msg.text file of the UV account directory. Date conversions specified in file
dictionaries or in ICONV or OCONV functions use the default date format except where they specifically
override it. When NLS locales are enabled, the locale overrides any value set in the msg.text file.
Table 68. Override values
Value Meaning Special format options
n
Single digit (normally 1 through 4)
that specifies the number of digits of
the year to output. The default is 4.
*
Any single nonnumeric character that
separates the fields in the case where
the conversion must first do a group
extraction to obtain the internal date.
* cannot be a system delimiter.
m
Single digit that must accompany any
use of an asterisk. It denotes the
number of asterisk-delimited fields to
skip in order to extract the date.
The following table shows the format options you can use together:
Format Option
Use with These Options
Y M, MA, D, J, [f1, f2, f3, f4, f5 ]
YA M, MA, D, [f1, f , f3, f4, f5 ]
M Y, YA, D, [f1, f2, f3, f4, f5 ]
MA Y, YA, D, [f1, f2, f3, f4, f5 ]
MB Y, YA, D, [f1, f2, f3, f4, f5 ]
D Y, M, [f1, f2, f3, f4, f5 ]
N Y, M, MA, MB, D, WA [f1, f2, f3, f4, f5 ]
NA Y, M, MA, MB, D, WA [f1, f2, f3, f4, f5 ]
W Y, YA, M, MA, D
WA Y, YA, M, MA, D
WB Y, YA, M, MA, D
Q [f1 ]
J Y, [f1, f2, f3, f4, f5 ]
Z [f1 ]
Each format modifier must correspond to a format option. The following table shows which modifiers
can modify which options:
Table 69. Modifiers and options
Format Modifier Format Option
D M Y J W
A no yes yes no yes
n yes yes yes yes yes
Z yes yes yes yes no
`text’ yes yes yes yes yes
The ERA_NAMES field can contain a value followed by the format code %n[string].
Table 70. Era definitions
Format code Value Explanation
%
The value preceding the % specifies
the era name (the fmt option NA for
the D code). If there is no %
following the value, %1 is the
default. The rules are as follows:
%0-%9 The number following the % (usually
1 or 2) specifies the first year in this
era. Often the first year of an era (or
part-year of an era) precedes some
qualifying characters; therefore, the
era is divided into two values, each
with the same era name. One value
ends in %1string, and the other in
%2.
%+ The era numbers count backwards
relative to year numbers (that is, if
era year 1 corresponds to Julian year
Y, year 2 corresponds to Y-1, year 3 to
Y-2, and so forth.
%- The same as %+, but uses negative
era year numbers (for example, first
year Y is -1,Y-1 is -2, y-2 is -3, and so
forth.
%Y Uses the Julian year numbers for the
era year numbers.
n
Single digit (0 through 9) or the
characters +, - or Y. These characters
specify how to calculate the era
number.
string
Any string following the % is
appended to the year number
returned by the fmt option N.
In PICK, REALITY, and IN2 flavor accounts, the input conversion of an improper date always returns an
empty string and a status of 1.
If the data to be converted is the null value, a STATUS( ) value of 3 is set and no conversion occurs.
Example
Z modifies the day format option (D) by suppressing leading zeros (05 becomes 5). A3 modifies the
month format option (M) so that the month is represented by the first three alphabetic characters (APRIL
becomes APR). Z2 modifies the year format option (Y) by suppressing leading zeros and displaying two
digits. This conversion converts April 5, 1993 to 5 APR 93.
The international date conversion lets you convert dates in internal format to the default local convention
format and vice versa. If NLS locales are not enabled, the DI conversion defaults to D. If NLS locales are
enabled, DI uses the date conversion in the DI_FMT field. The DI_FMT field can contain any valid D
code.
The ECS code resolves clashes between the WebSphere DataStage system delimiters and the ASCII
characters CHAR(251) through CHAR(255). It converts the system delimiters and ASCII characters to
alternative characters using an appropriate localization procedure. If no localization library is in use, the
input string is returned without character conversion. This code is used with an ICONV or OCONV
function.
Fcode
Format
F [ ; ] element [ ; element ...]
The F code performs mathematical operations on the data values of a record, or manipulates strings. It
comprises any number of operands or operators in reverse Polish format (Lukasiewicz) separated by
semicolons.
The program parses the F code from left to right, building a stack of operands. Whenever it encounters
an operator, it performs the requested operation, puts the result on the top of the stack, and pops the
lower stack elements as necessary.
An operator
Operators specify an operation to be performed on top stack entries. stack1 refers to the value on the top
of the stack, stack2 refers to the value just below it, stack3 refers to the value below stack2, and so on.
*[n] Multiply stack1 by stack2. The optional n is the descaling factor (that is, the result is divided by 10
raised to the nth power).
/ Divide stack1 into stack2, result to stack1.
R Same as /, but instead of the quotient, the remainder is returned to the top of the stack.
+ Add stack1 to stack2.
- Subtract stack1 from stack2, result to stack1 (except for REALITY flavor, which subtracts stack2
from stack1).
: Concatenate stack1 string onto the end of stack2 string.
[] Extract substring. stack3 string is extracted, starting at the character specified by stack2 and
continuing for the number of characters specified in stack1. This is equivalent to the BASIC [m,n]
operator, where m is in stack2 and n is in stack1.
S Sum of multivalues in stack1 is placed at the top of the stack.
_ Exchange stack1 and stack2 values.
P or \ Push stack1 back onto the stack (that is, duplicate stack1).
^ Pop the stack1 value off the stack.
(conv) Standard conversion operator converts data in stack1, putting the result into stack1.
Logical operators compare stack1 to stack2. Each returns 1 for true and 0 for false:
= Equal to.
< Less than.
> Greater than.
# or <>
Not equal to.
[ Less than or equal to.
] Greater than or equal to.
& Logical AND.
! Logical OR.
\n\ A label is defined by a positive integer enclosed by two backslashes (\\).
#n Connection to label n if stack1 differs from stack2.
>n Connection to label n if stack1 is greater than stack2.
<n Connection to label n if stack1 is less than stack2.
=n Connection to label n if stack1 equals stack2.
}n Connection to label n if stack1 is greater than or equal to stack2.
{n Connection to label n if stack1 is less than or equal to stack2.
IN Tests stack1 to see if it is the null value.
Fnnnn If stack1 evaluates to false, branch forward nnnn characters in the F code, and continue
processing.
Bnnnn Branch forward unconditionally nnnn characters in the F code, and continue processing.
Gnnnn
Go to label nnnn. The label must be a string delimited by backslashes ( \ ).
G* Go to the label defined in stack1. The label must be a string delimited by backslashes ( \ ).
Group Extraction
Format
G [ skip ] delim #fields
The G code extracts one or more values, separated by the specified delimiter, from a field.
skip specifies the number of fields to skip; if it is not specified, 0 is assumed and no fields are skipped.
delim is any single nonnumeric character (except IM, FM, VM, SM, and TM) used as the field separator.
If Ln is specified, selection is met if the value’s length is less than or equal to n characters; otherwise an
empty string is returned.
If Ln,m is specified, selection is met if the value’s length is greater than or equal to n characters, and less
than or equal to m characters; otherwise an empty string is returned.
The MC codes let you change a field’s data to upper- or lowercase, to extract certain classes of characters,
to capitalize words in the field, and to change unprintable characters to periods.
MCA Extracts all alphabetic characters in the field, both upper- and lowercase. Nonalphabetic
characters are not printed. In NLS mode, uses the ALPHABETICS field in the NLS.LC.CTYPE file.
MC/A Extracts all nonalphabetic characters in the field. Alphabetic characters are not printed. In NLS
mode, uses the NON-ALPHABETICS field in the NLS.LC.CTYPE file.
MCD[X]
Converts decimal to hexadecimal equivalents.
MCL Converts all uppercase letters to lowercase. Does not affect lowercase letters or nonalphabetic
characters. In NLS mode, uses the UPPERCASE and DOWNCASED fields in the NLS.LC.CTYPE
file.
MCM Use only if NLS is enabled. Extracts all NLS multibyte characters in the field. Multibyte
characters are all those outside the Unicode range (x0000-x007F), the WebSphere DataStage
system delimiters, and the null value. As long as NLS is enabled, the conversion still works if
locales are off. If NLS mode is disabled, the code returns a STATUS of 2, that is, an invalid
conversion code.
MC/M Use only if NLS is enabled. Extracts all NLS single-byte characters in the field. Single-byte
characters are all those in the Unicode range x0000-x007F. As long as NLS is enabled, the
conversion still works if locales are off. If NLS mode is disabled, the code returns a STATUS of 2,
that is, an invalid conversion code.
MCN Extracts all numeric characters in the field. Alphabetic characters are not printed. In NLS mode,
uses the NUMERICS field in the NLS.LC.CTYPE file.
The MD code converts numeric input data to a format appropriate for internal storage. If the code
includes the $, F, I, or Y option, the conversion is monetary, otherwise it is numeric. The MD code must
appear in either an ICONV or an OCONV expression. When converting internal representation of data to
external output format, masked decimal conversion inserts the decimal point and other appropriate
formats into the data.
Note: If NLS is enabled and either the Numeric or Monetary categories are set to OFF, the MD code
behaves as if NLS locales were turned off.
If the value of n is 0, the decimal point does not appear in the output.
The optional m specifies the power of 10 used to scale the input or output data. On input, the decimal
point is moved m places to the right before storing. On output, the decimal point is moved m places to
the left. For example, if m is 2 in an input conversion and the input data is 123, it would be stored as
12300. If m is 2 in an output conversion and the stored data is 123, it would be output as 1.23. If m is not
specified, it is assumed to be the same as n. In both cases, the last required decimal place is rounded off
before excess digits are truncated. Zeros are added if not enough decimal places exist in the original
expression.
If NLS is enabled and the conversion is monetary, the thousands separator comes from the THOU_SEP
field of the Monetary category of the current locale, and the decimal separator comes from the DEC_SEP
field. If the conversion is numeric, the thousands separator comes from the THOU_SEP field of the
Numeric category, and the decimal separator comes from the DEC_SEP field.
Table 71. Masked decimal conversion
Option Explanation NoteS
,
Specifies that thousands separators be
inserted every three digits to the left
of the decimal point on output.
When converting data to internal format, the fx option has the following effect. If the input data has been
overlaid on a background field of characters (for example, $###987.65), the fx option is used with ICONV
to indicate that the background characters should be ignored during conversion. The f is a one- or
two-digit number indicating the maximum number of background characters to be ignored. The x
specifies the background character to be ignored. If background characters exist in the input data and you
do not use the fx option, the data is considered bad and an empty string results.
When converting data from internal representation to external output format, the fx option causes the
external output of the data to overlay a field of background characters. The f is a one- or two-digit
number indicating the number of times the background character is to be repeated. The x specifies the
character to be used as a background character. If the $ option is used with the fx option, the $ precedes
the background characters when the data is output.
Formatting Numbers
Format
ML [ n [ m ] ] [ Z ] [ , ] [ C | D | M | E | N ] [ $ ] [ F ] [ intl ] [ ( fx ) ]
MR [ n [ m ] ] [ Z ] [ , ] [ C | D | M | E | N ] [ $ ] [ F ] [ intl ] [ ( fx ) ]
The ML and MR codes allow special processing and formatting of numbers and monetary amounts. If the
code includes the F or I option, the conversion is monetary, otherwise it is numeric. ML specifies left
justification; MR specifies right justification.
Note: If NLS is enabled and either the Numeric or Monetary categories are set to OFF, the ML and MR
codes behave as if locales were turned off.
n Number of digits to be printed to the right of the decimal point. If n is omitted or 0, no decimal
point is printed.
m Descales (divides) the number by 10 to the mth power. If not specified, m = n is assumed. On
input, the decimal point is moved m places to the right before storing. On output, the decimal
point is moved m places to the left. For example, if m is 2 in an input conversion specification
and the input data is 123, it would be stored as 12300. If m is 2 in an output conversion
specification and the stored data is 123, it would be output as 1.23. If the m is not specified, it is
assumed to be the same as the n value. In both cases, the last required decimal place is rounded
off before excess digits are truncated. Zeros are added if not enough decimal places exist in the
original expression.
If NLS is enabled and the conversion is monetary, the thousands separator comes from the THOU_SEP
field of the Monetary category of the current locale, and the decimal separator comes from the DEC_SEP
field. If the conversion is numeric, the thousands separator comes from the THOU_SEP field of the
Numeric category, and the decimal separator comes from the DEC_SEP field.
When NLS locales are enabled, the <, -, C, and D options define numbers intended for monetary use.
These options override any specified monetary formatting. If the conversion is monetary and no
monetary formatting is specified, it uses the POS_FMT, NEG_FMT, POS_SIGN, and NEG_SIGN fields
from the Monetary category of the current locale.
They are unaffected by the Numeric or Monetary categories. If no options are set, the value is returned
unchanged.
The format codes precede x, the number that specifies the size of the field.
You can also enclose literal strings in the parentheses. The text is printed as specified, with the number
being processed right- or left-justified.
Monetary Conversion
Format
MM [ n ] [ I [ L ] ]
Note: If NLS is enabled and either the Numeric or Monetary categories are set to OFF, the MM code
behaves as if locales were turned off.
If NLS is enabled and the Monetary category is turned on, the MM code uses the local monetary
conventions for decimal and thousands separators. The format options are as follows:.
n Specifies the number of decimal places (0 through 9) to be maintained or output. If n is omitted,
the DEC_PLACES field from the Monetary category is used; if the I option is also specified, the
INTL_DEC_PLACES field is used. If NLS is disabled or the Monetary category is turned off, and
n is omitted, n defaults to 2.
I Substitutes the INTL_CURR_SYMBOL for the CURR_SYMBOL in the Monetary category of the
current locale. If NLS locales are off, the default international currency symbol is USD.
L Used with the I option to specify that decimal and thousands separators are required instead of
the defaults ( . and , ). The DEC_SEP and THOU_SEP fields from the Monetary category are
used.
If NLS is disabled or the category is turned off, the default decimal and thousands separators are the
period and the comma.
The MP code allows decimal numbers to be packed two-to-the-byte for storage. Packed decimal numbers
occupy approximately half the disk storage space required by unpacked decimal numbers.
Leading + signs are ignored. Leading - signs cause a hexadecimal D to be stored in the lower half of the
last internal digit. If there is an odd number of packed halves, four leading bits of 0 are added. The range
of the data bytes in internal format expressed in hexadecimal is 00 through 99 and 0D through 9D. Only
valid decimal digits (0-9) and signs ( +, - ) should be input. Other characters cause no conversion to take
place.
Packed decimal numbers should always be unpacked for output, since packed values that are output
unconverted are not displayed on terminals in a recognizable format.
Time Conversion
Format
MT [ H ] [ P ] [ Z ] [ S ] [ c ] [ [ f1, f2, f3 ] ]
The MT code converts times from conventional formats to an internal format for storage. It also converts
internal times back to conventional formats for output. When converting input data to internal storage
format, time conversion specifies the format that is to be used to enter the time. When converting internal
representation of data to external output format, time conversion defines the external output format for
the time.
MT is required when you specify time in either the ICONV or the OCONV function. The remaining
specifiers are meaningful only in the OCONV function; they are ignored when used in the ICONV
function.
The internal representation of time is the numeric value of the number of seconds since midnight.
If used with ICONV in an IDEAL, INFORMATION, or PIOPEN flavor account, the value of midnight is
0. In all other account flavors, the value of midnight is 86400.
If NLS is enabled and the Time category is active, the locale specifies the AM and PM strings, and the
separator comes from the T_FMT or TI_FMT fields in the Time category.
Table 73. Time conversion
Code Explanation Note
H
Specifies to use a 12-hour format
with the suffixes AM or PM. The
24-hour format is the default. If NLS
is enabled, the AM and PM strings
come from the AM_STR and PM_STR
fields in the Time category.
P
Same as H, but the AM and PM
strings are prefixed, not suffixed.
Z
Specifies to zero-suppress hours in
the output.
S
Specifies to use seconds in the
output. The default omits seconds.
c
Specifies the character used to
separate the hours, minutes, and
seconds in the output. The colon ( : )
is the default. If NLS is enabled and
you do not specify c, and if the Time
category is active, c uses the
DEFAULT_TIME_SEP field.
[f1, f2, f3]
Specify format modifiers. You must
include the brackets, as they are part
of the syntax. You can specify from 1
through 3 modifiers, which
correspond to the hours, minutes,
and seconds, in that order. The
format modifiers are positional
parameters: if you want to specify f3
only, you must include two commas
as placeholders. Each format modifier
must correspond to a format option.
Use the following value for the
format modifiers:
`text’ Any text you enclose in single or
double quotation marks is output
without the quotation marks and
placed after the appropriate number
for the hours, minutes, or seconds.
The MX, MO, and MB codes convert data from hexadecimal, octal, and binary format to decimal (base
10) format and vice versa.
With ICONV
The decimal, or ASCII, format is the internal format for data representation. When used with the ICONV
function, MX, MO, and MB without the 0C extension convert hexadecimal, octal, or binary data values
(respectively) to their equivalent decimal values. MX, MO, and MB with the 0C extension convert
hexadecimal, octal, or binary data values to the equivalent ASCII characters rather than to decimal
values.
Use the MU0C code only if NLS is enabled. When used with ICONV, MU0C converts data in Unicode
hexadecimal format to its equivalent in the internal character set.
Characters outside of the range for each of the bases produce conversion errors. The ranges are as
follows:
MX (hexadecimal)
0 through 9, A through F, a through f
MO (octal)
0 through 7
MB (binary)
0, 1
MU0C (Unicode)
No characters outside range
With OCONV
When used with the OCONV function, MX, MO, and MB without the 0C extension convert decimal
values to their equivalent hexadecimal, octal, or binary equivalents for output, respectively. Nonnumeric
data produces a conversion error if the 0C extension is not used.
MX, MO, and MB with the 0C extension convert an ASCII character or character string to hexadecimal,
octal, or binary output format. Each character in the string is converted to the hexadecimal, octal, or
binary equivalent of its ASCII character code.
Use the MU0C code only if NLS is enabled. When used with OCONV, MU0C converts characters from
their internal representation to their Unicode hexadecimal equivalents for output. The data to convert
must be a character or character string in the internal character set; each character in the string is
converted to its 4-digit Unicode hexadecimal equivalent. Data is converted from left to right, one
character at a time, until all data is exhausted.
ASCII Conversion
Format
MY
Characters outside of the range for each of the bases produce conversion errors. The ranges are as
follows:
MY (hexadecimal)
0 through 9, A through F, a through f
Map conversion
Format
NL
The NL code allows conversion from a locale-dependent set of alternative characters (representing digits
in the local language) to Arabic numerals. The alternative characters are the external set, the Arabic
characters are the internal set.
If NLS is not enabled, characters are checked to ensure only that they are valid ASCII digits 0 through 9,
but no characters are changed.
The NLSmapname code converts data from internal format to external format and vice versa using the
specified map. mapname is either a valid map name or one of the following: LPTR, CRT, AUX, or OS.
NR Code
Format
NR
The NR code converts Roman numerals into Arabic numerals when used with the ICONV function. The
decimal, or ASCII, format is the internal format for representation.
When used with the OCONV function, the NR code converts Arabic numerals into Roman numerals.
P code
Format
P(pattern) [ { ; | / } (pattern) ] ...
The P code extracts data whose values match one or more patterns. If the data does not match any of the
patterns, an empty string is returned.
If n is 0, any number of numeric, alphabetic, or alphanumeric characters matches. If either the data or the
match pattern is the null value, null is returned.
Parentheses must enclose each pattern to be matched. For example, if the user wanted only Social
Security numbers returned, P(3N-2N-4N) would test for strings of exactly three numbers, then a hyphen,
then exactly two numbers, then a hyphen, then exactly four numbers.
Exponential Notation
Format
QR [ n { E | . } m ] [ edit ] [ mask ]
QL [ n { E | . } m ] [ edit ] [ mask ]
QX
Q alone and QR both specify right justification. QL specifies left justification. QX specifies right
justification. QX is synonymous with QR0E0 as input and MR as output.
n specifies the number of fractional digits to the right of the decimal point. It can be a number from 0
through 9.
m specifies the exponent. It can be a number from 0 through 9. When used with E, m can also be a
negative number from -1 through -9.
Separate n and m with either the letter E or a period ( . ). Use E if you want to specify a negative
exponent.
mask allows literals to be intermixed with numerics in the formatted output field. The mask can include
any combination of literals and the following three special format mask characters:
#n Data is displayed in a field of n fill characters. A blank is the default fill character. It is used if the
format string does not specify a fill character after the width parameter.
%n Data is displayed in a field of n zeros.
*n Data is displayed in a field of n asterisks.
If NLS is enabled, the Q code formats numeric and monetary values as the ML and MR codes do, except
that the intl format cannot be specified. See the ML and MR codes for more information.
See the FMT function for more information about formatting numbers.
Range Function
Format
Rn,m [ { ; | / } n,m ] ...
S code
Format
S
The S code with no arguments specifies a soundex conversion. Soundex is a phonetic converter that
converts ordinary English words into a four-character abbreviation comprising one alphabetic character
followed by three digits. Soundex conversions are frequently used to build indexes for name lookups.
Substitution code
Format
S ; nonzero.substitute ; zero.substitute ; null.substitute
The S code substitutes one of three values depending on whether the data to convert evaluates to zero or
an empty string, to the null value, or to something else.
If the data to convert evaluates to zero or an empty string, zero.substitute is returned. If the data is
nonzero, nonempty, and nonnull, nonzero.substitute is returned. If the data is the null value, null.substitute
is returned. If null.substitute is omitted, null values are not replaced.
If it is an asterisk and the data evaluates to something other than zero, the empty string, or the null
value, the data value itself is returned.
Example
If you specify length only, the extraction is either from the left or from the right depending on the
justification specified in line 5 of the dictionary definition item. In a BASIC program if you specify length
only, the extraction is from the right. In this case the starting position is calculated according to the
following formula:
string.length - substring.length + 1
This lets you extract the last n characters of a string without having to calculate the string length.
File Translation
Format
T[DICT] filename ; c [vloc] ; [iloc] ; [oloc] [ ;bloc]
T[DICT] filename ; c ; [iloc] ; [oloc] [ ;bloc] [ ,vloc | [vloc] ]
The Tfile code converts values from one file to another by translating through a file. It uses data values in
the source file as IDs for records in a lookup file. The source file can then reference values in the lookup
file.
To access the lookup file, its record IDs (field 0) must be referenced. If no reference is made to the record
IDs of the lookup file, the file cannot be opened and the conversion cannot be performed. The data value
being converted must be a record ID in the lookup file.
Table 74. File translation
Argument Explanation Notes
DICT
Specifies the lookup file’s dictionary.
(In REALITY flavor accounts, you
can use an asterisk ( * ) to specify the
dictionary: for instance, T*filename ...
.)
filename
Name of the lookup file.
c
Translation subcode, which must be
one of the following:
V Conversion item must exist on file,
and the specified field must have a
value, otherwise an error message is
returned.
C If conversion is impossible, return the
original value-to-be-translated.
I Input verify only. Functions like V for
input and like C for output.
The international time conversion lets you convert times in internal format to the default local convention
format and vice versa. If NLS locales are not enabled, the TI conversion defaults to MT. If NLS locales are
enabled, TI uses the date conversion in the TI_FMT field of the Time category. The TI_FMT field can
contain any valid MT code.
ABORT
ABORTE
ABORTM
ABS
ABSS
ACOS
ADDS
ALL
ALPHA
AND
ANDS
ARG.
ASCII
ASIN
ASSIGN
ASSIGNED
ATAN
AUTHORIZATION
BCONVERT
BEFORE
BEGIN
BITAND
BITNOT
BITOR
BITSET
BITTEST
BITXOR
BREAK
BSCAN
BY
CALL
CALLING
CAPTURING
CASE
CAT
CATS
CHAIN
CHANGE
CHAR
CHARS
CHECKSUM
CLEAR
CLEARCOMMON
CLEARDATA
CLEARFILE
CLEARINPUT
CLEARPROMPTS
CLEARSELECT
CLOSE
CLOSESEQ
COL1
COM
COMMIT
COMMON
COMPARE
CONTINUE
CONVERT
COS
COSH
COUNT
COUNTS
CREATE
CRC32
CRT
DATA
DATE
DCOUNT
DEBUG
DECLARE
DEFFUN
DEL
DELETE
DELETELIST
DELETEU
DIAGNOSTICS
DIM
DIMENSION
DISPLAY
DIVS
DO
DOWNCASE
DQUOTE
DTX
EBCDIC
ECHO
ELSE
END
ENTER
EOF
EQ
EQS
EQU
EQUATE
EREPLACE
ERRMSG
ERROR
EXCHANGE
EXEC
EXECUTE
EXIT
EXP
EXTRACT
FADD
FDIV
FFIX
FIELD
FIELDS
FIELDSTORE
FILEINFO
FILELOCK
FILEUNLOCK
FIND
FINDSTR
FIX
FLUSH
FMT
FMTS
FMUL
FOLD
FOOTING
FOR
FORMLIST
FROM
FSUB
FUNCTION
GARBAGECOLLECT
GCI
GE
GES
GET
GETLIST
GETREM
GO
GOSUB
GOTO
GROUP
GROUPSTORE
GT
GTS
HEADING
HEADINGE
HEADINGN
HUSH
ICHECK
ICONV
ICONVS
IF
IFS
ILPROMPT
IN
INCLUDE
INDEX
INDEXS
INDICES
INMAT
INPUT
INPUTCLEAR
INPUTDISP
INPUTERR
INPUTNULL
INPUTTRAP
INS
INSERT
INT
ISNULL
ISNULLS
ISOLATION
ITYPE
KEY
KEYEDIT
KEYEXIT
KEYIN
KEYTRAP
LE
LEFT
LEN
LENS
LES
LET
LEVEL
LIT
LITERALLY
LN
LOCATE
LOCK
LOCKED
LOWER
LPTR
LT
LTS
MAT
MATBUILD
MATCH
MATCHES
MATCHFIELD
MATPARSE
MATREAD
MATREADL
MATREADU
MATWRITE
MATWRITEU
MAXIMUM
MESSAGE
MINIMUM
MOD
MODS
MTU
MULS
NAP
NE
NEG
NEGS
NES
NOBUF
NO.ISOLATION
NOT
NOTS
NULL
NUM
NUMS
OCONV
OCONVS
OFF
ON
OPEN
OPENCHECK
OPENDEV
OPENPATH
OPENSEQ
OR
ORS
OUT
PAGE
PASSLIST
PCDRIVER
PERFORM
PRECISION
PRINTER
PRINTERIO
PROCREAD
PROCWRITE
PROG
PROGRAM
PROMPT
PWR
QUOTE
RAISE
RANDOMIZE
READ
READ.COMMITTED
READ.UNCOMMITTED
READBLK
READL
READLIST
READNEXT
READSEQ
READT
READU
READV
READVL
READVU
REAL
RECIO
RECORDLOCKED
RECORDLOCKL
RECORDLOCKU
REM
REMOVE
REPEAT
REPEATABLE.READ
REPLACE
RESET
RETURN
RETURNING
REUSE
REVREMOVE
REWIND
RIGHT
RND
ROLLBACK
RPC.CALL
RPC.CONNECT
RPC.DISCONNECT
RQM
RTNLIST
SADD
SCMP
SDIV
SEEK
SELECT
SELECTE
SELECTINDEX
SELECTN
SEND
SENTENCE
SEQ
SEQS
SEQSUM
SERIALIZABLE
SET
SETREM
SETTING
SIN
SINH
SLEEP
SMUL
SOUNDEX
SPACE
SPACES
SPLICE
SQLALLOCONNECT
SQLALLOCENV
SQLALLOCSTMT
SQLBINDCOL
SQLCANCEL
SQLCOLATTRI- BUTES
SQLCONNECT
SQLDESCRIBECOL
SQLDISCONNECT
SQLERROR
SQLEXECUTE
SQLFETCH
SQLFREECONNECT
SQLFREEENV
SQLFREESTMT
SQLGETCURSORNAME
SQLNUMRESULTCOLS
SQLPREPARE
SQLROWCOUNT
SQLSETCONNECT-OPTION
SQLSETCURSORNAME
SQLSETPARAM
SQRT
SQUOTE
SSELECT
SSELECTN
SSELECTV
SSUB
START
STATUS
STEP
STOP
STOPE
STOPM
STORAGE
STR
STRS
SUBR
SUBROUTINE
SUBS
SUBSTRINGS
SUM
SUMMATION
SYSTEM
TABSTOP
TAN
TANH
TERMINFO
THEN
TIME
TIMEDATE
TIMEOUT
TO
TPARM
TPRINT
TRANS
TRANSACTION
TRIM
TRIMB
TRIMBS
TRIMF
TRIMFS
TRIMS
TTYCTL
TTYSET
UNASSIGNED
UNIT
UNLOCK
UNTIL
UPCASE
USING
WEOF
WEOFSEQ
WEOFSEQF
WHILE
WORDSIZE
WORKWRITE
WRITEBLK
WRITELIST
WRITESEQ
WRITESEQF
WRITET
WRITEU
WRITEV
WRITEVU
XLATE
XTD
The EXECUTE statement initializes the values of stacked @variables either to 0 or to values reflecting the
new environment. These values are not passed back to the calling environment. The values of nonstacked
@variables are shared between the EXECUTE and calling environments. All @variables listed here are
stacked unless otherwise indicated.
Table 75. BASIC @Variables
Variable Read-Only Value
@ABORT.CODE * A numeric value indicating the type
of condition that caused the
ON.ABORT paragraph to execute.
The values are: 1 - An ABORT
statement was executed.
!ASYNC (!AMLC)
!EDIT.INPUT
!ERRNO
!FCMP
!GET.KEY
!GET.PARTNUM
!GET.PATHNAME
!GET.USER.COUNTS
!GETPU
!INLINE.PROMPTS
!INTS
!MAKE.PATHNAME
!MATCHES
!MESSAGE
!PACK.FNKEYS
!REPORT.ERROR
!SET.PTR
!SETPU
!TIMDAT
!USER.TYPE
!VOC.PATHNAME
In addition, the subroutines listed in Table F-1 have been added to existing functions for PI/open
compatibility.
Description
Use the !ASYNC subroutine (or its synonym !AMLC) to send data to, and receive data from an
asynchronous device.
key defines the action to be taken (1 through 5). The values for key are defined in the following list:
line is the number portion from the &DEVICE& entry TTY##, where ## represents a decimal number.
carrier is an output variable that returns a value dependent on the value of key. If key is 1, 2, or 3, carrier
returns the variable specified by the user. If key has a value of 4 or 5, carrier returns 1.
You must first assign an asynchronous device using the ASSIGN command. A entry must be in the
&DEVICE& file for the device to be assigned with the record ID format of TTY##, where ## represents a
decimal number. The actions associated with each key value are as follows:
key Action
1 Inputs the number of characters indicated by the value of count.
2 Inputs the number of characters indicated by the value of count or until a linefeed character is
encountered.
3 Outputs the number of characters indicated by the value of count.
4 Returns the number of characters in the input buffer to count. On operating systems where the
FIONREAD key is not supported, 0 is returned in count. When the value of key is 4, 1 is always
returned to carrier.
5 Returns 0 in count if there is insufficient space in the output buffer. On operating systems where
the TIOCOUTQ key is not supported, 0 is returned in count. When the value of key is 5, 1 is
always returned to carrier.
Example
The !ASYNC subroutine returns the first 80 characters from the device defined by ASYNC10 in the
&DEVICE& file to the variable data.
data=
count= 80
carrier= 0
call !ASYNC(1,10,data,count,carrier)
!EDIT.INPUT subroutine
Syntax
CALL !EDIT.INPUT (keys, wcol, wrow, wwidth, buffer, startpos,
bwidth, ftable, code)
Description
Use the !EDIT.INPUT subroutine to request editable terminal input within a single-line window on the
terminal. Editing keys are defined in the terminfo files and can be set up using the KEYEDIT, KEYTRAP
and KEYEXIT statements. To ease the implementation, the UNIVERSE.INCLUDE file GTI.FNKEYS.IH can
be included to automatically define the editing keys from the current terminfo definition. We recommend
that you use the INCLUDE file.
All input occurs within a single-line window of the terminal screen, defined by the parameters wrow,
wcol, and wwidth. If the underlying buffer length bwidth is greater than wwidth and the user performs a
function that moves the cursor out of the window horizontally, the contents of buffer are scrolled so as to
keep the cursor always in the window.
If the specified starting cursor position would take the cursor out of the window, the buffer’s contents are
scrolled immediately so as to keep the cursor visible. !EDIT.INPUT does not let the user enter more than
bwidth characters into the buffer, regardless of the value of wwidth.
The specific keys to perform each function can be automatically initialized by including the $INCLUDE
UNIVERSE.INCLUDE GTI.FNKEYS.IH statement in the application program.
If any of the values appear in the trap list, its functionality is disabled and the program immediate exits
the !EDIT.INPUT subroutine when the key associated with that function is pressed.
Unsupported Functions
This implementation does not support a number of functions originally available in the Prime
INFORMATION version. Because of this, sequences can be generated that inadvertently cause the
!EDIT.INPUT function to terminate. For this reason, you can create a user-defined terminal keystroke
definition file so that !EDIT.INPUT recognizes the unsupported sequences. Unsupported sequences cause
the !EDIT.INPUT subroutine to ring the terminal bell, indicating the recognition of an invalid sequence.
The file CUSTOM.GTI.DEFS defines a series of keystroke sequences for this purpose. You can create the
file in each account or in a central location, with VOC entries in satellite accounts referencing the remote
file. There is no restriction on how the file can be created. For instance, you can use the command:
>CREATE.FILE CUSTOM.GTI.DEFS 2 17 1 /* Information
style */
or:
>CREATE-FILE CUSTOM.GTI.DEFS (1,1,3 17,1,2) /* Pick
style */
to create the definition file. A terminal keystroke definition record assumes the name of the terminal
which the definitions are associated with, i.e., for vt100 terminals the CUSTOM.GTI.DEFS file record ID
would be vt100 (case-sensitive). Each terminal keystroke definition record contains a maximum of 82
fields (attributes) which directly correspond to the keystroke code listed in the GTI.FNKEYS.IH include
file.
The complete listing of the fields defined within the GTI.FNKEYS.IH include file is shown below:
Table 79. Fields in include file
Key Name Field Description
FK$FIN 1 Finish
FK$HELP 2 Help
1
Indicates supported functionality.
Example
The following BASIC program sets up three trap keys (using the !PACK.FNKEYS subroutine), waits for
the user to enter input, then reports how the input was terminated:
$INCLUDE UNIVERSE.INCLUDE GTI.FNKEYS.IH
* Set up trap keys of FINISH, UPCURSOR and DOWNCURSOR
TRAP.LIST = FK$FIN:@FM:FK$UP:@FM:FK$DOWN
CALL !PACK.FNKEYS(TRAP.LIST, Ftable)
* Start editing in INPUT mode, displaying contents in window
KEYS = IK$INS + IK$DIS
* Window edit is at x=20, y=2, of length 10 characters;
* the user can enter up to 30 characters of input into
TextBuffer,
* and the cursor is initially placed on the first character of
the
* window.
TextBuffer=""
CursorPos = 1
CALL !EDIT.INPUT(KEYS, 20, 2, 10, TextBuffer, CursorPos, 30,
Ftable,
ReturnCode)
* On exit, the user’s input is within TextBuffer,
* CursorPos indicates the location of the cursor upon exiting,
* and ReturnCode contains the reason for exiting.
BEGIN CASE
CASE CODE = 0 * User pressed RETURN key
CASE CODE = FK$FIN * User pressed the defined FINISH
key
CASE CODE = FK$UP * User pressed the defined UPCURSOR
key
CASE CODE = FK$DOWN * User pressed the defined
DOWNCURSOR key
CASE 1 * Should never happen
END CASE
!ERRNO subroutine
Syntax
CALL !ERRNO (variable)
Description
Use the !ERRNO subroutine to return the current value of the operating system errno variable.
The !ERRNO subroutine returns the value of the system errno variable after the last call to a GCI
subroutine in variable. If you call a system routine with the GCI, and the system call fails, you can use
!ERRNO to determine what caused the failure. If no GCI routine was called prior to its execution,
!ERRNO returns 0. The values of errno that apply to your system are listed in the system include file
errno.h.
Description
Use the !FCMP subroutine to compare the equality of two floating-point numeric values as follows:
!GET.KEY subroutine
Syntax
CALL !GET.KEY (string, code)
Qualifiers
Table 80. GET.KEY subroutine
Argument Code String Value
string
Returns the character sequence of the
next key pressed at the keyboard.
code
Returns the string interpretation
value:
0 A single character that is not part of
any function key sequence. For
example, if A is pressed, code = 0 and
string = CHAR(65).
>0 The character sequence associated
with the function key defined by that
number in the GTI.FNKEYS.IH
include file. For example, on a VT100
terminal, pressing the key labelled -->
(right cursor move) returns code = 5
and string =
CHAR(27):CHAR(79):CHAR(67).
<0 A character sequence starting with an
escape or control character that does
not match any sequence in either the
terminfo entry or the
CUSTOM.GCI.DEFS file.
Description
Use the !GET.KEY subroutine to return the next key pressed at the keyboard. This can be either a printing
character, the Return key, a function key as defined by the current terminal type, or a character sequence
that begins with an escape or control character not defined as a function key.
Example
The following BASIC program waits for the user to enter input, then reports the type of input entered:
$INCLUDE GTI.FNKEYS.IH
STRING = ’ ’ ; * initial states of call variables
CODE = -999
* Now ask for input until user hits a "Q"
LOOP
UNTIL STRING[1,1] = "q" OR STRING[1,1] = "Q"
PRINT ’Type a character or press a function key (q to
quit):’:
CALL !GET.KEY(STRING, CODE)
* Display meaning of CODE
PRINT
PRINT "CODE = ":CODE:
BEGIN CASE
CASE CODE = 0
PRINT " (Normal character)"
CASE CODE > 0
PRINT " (Function key number)"
CASE 1; * otherwise
PRINT " (Unrecognised function key)"
END CASE
* Print whatever is in STRING, as decimal numbers:
PRINT "STRING = ":
FOR I = 1 TO LEN(STRING)
PRINT "CHAR(":SEQ(STRING[I,1]):") ":
NEXT I
PRINT
REPEAT
PRINT "End of run."
RETURN
END
!GET.PARTNUM subroutine
Syntax
CALL !GET.PARTNUM (file, record.ID, partnum, status)
Description
Use the !GET.PARTNUM subroutine with distributed files to determine the number of the part file to
which a given record ID belongs.
partnum (output) is the part number of the part file of the distributed file to which the given record ID
maps.
status (output) is 0 for a valid part number or an error number for an invalid part number. An insert file
of equate tokens for the error numbers is available.
An insert file of equate names is provided to allow you to use mnemonics for the error numbers. The
insert file is called INFO_ERRORS.INS.IBAS, and is located in the INCLUDE subdirectory. To use the
insert file, specify $INCLUDE SYSCOM INFO_ERRORS.INS.IBAS when you compile the program. 1
Use the !GET.PARTNUM subroutine to call the partitioning algorithm associated with a distributed file. If
the part number returned by the partitioning algorithm is not valid, that is, not an integer greater than
zero, !GET.PARTNUM returns a nonzero status code. If the part number returned by the partitioning
algorithm is valid, !GET.PARTNUM returns a zero status code.
Note: !GET.PARTNUM does not check that the returned part number corresponds to one of the available
part files of the currently opened file.
Example
In the following example, a distributed file SYS has been defined with parts and part numbers S1, 5, S2,
7, and S3, 3, respectively. The file uses the default SYSTEM partitioning algorithm.
PROMPT ’’
GET.PARTNUM = ’!GET.PARTNUM’
STATUS = 0
PART.NUM = 0
OPEN ’’, ’SYS’ TO FVAR ELSE STOP ’NO OPEN SYS’
PATHNAME.LIST = FILEINFO(FVAR, FINFO$PATHNAME)
PARTNUM.LIST = FILEINFO(FVAR, FINFO$PARTNUM)
LOOP
PRINT ’ENTER Record ID : ’:
INPUT RECORD.ID
WHILE RECORD.ID
CALL @GET.PARTNUM(FVAR, RECORD.ID, PART.NUM, STATUS)
LOCATE PART.NUM IN PARTNUM.LIST<1> SETTING PART.INDEX THEN
PATHNAME = PATHNAME.LIST <PART.INDEX>
END ELSE
PATHNAME = ’’
END
PRINT ’PART.NUM = ’:PART.NUM:’ STATUS = ’:STATUS :’
PATHNAME = ’: PATHNAME
REPEAT
END
!GET.PARTNUM returns part number 5 for input record ID 5-1, with status code 0, and part number 7 for
input record ID 7-1, with status code 0, and part number 3 for input record ID 3-1, with status code 0.
These part numbers are valid and correspond to available part files of file SYS.
!GET.PARTNUM returns part number 1200 for input record ID 1200-1, with status code 0. This part
number is valid but does not correspond to an available part file of file SYS.
!GET.PARTNUM returns part number 0 for input record ID 5-1, with status code
IE$NO.MAP.TO.PARTNUM, and part number 0 for input record ID A-1, with status code
IE$NO.MAP.TO.PARTNUM, and part number 0 for input record ID 12-4, with status code
IE$NO.MAP.TO.PARTNUM. These part numbers are not valid and do not correspond to available part
files of the file SYS.
Description
Use the !GET.PATHNAME subroutine to return the directory name and filename parts of a path name.
pathname (input) is the path name from which the details are required.
directoryname (output) is the directory name portion of the path name, that is, the path name with the last
entry name stripped off.
status (output) is the returned status of the operation. A 0 indicates success, another number is an error
code indicating that the supplied path name was not valid.
Example
!GET.PU subroutine
Syntax
CALL !GETPU (key, print.channel, set.value, return.code)
Description
Use the !GETPU subroutine to read individual parameters of any logical print channel.
The !GETPU subroutine allows you to read individual parameters of logical print channels as designated
by print.channel. Print channel 0 is the terminal unless a PRINTER ON statement has been executed to
send output to the default printer. If you specify print channel -1, the output is directed to the terminal,
regardless of the status of PRINTER ON or OFF. See the description of the !SETPU subroutine later in
this appendix for a means of setting individual print.channel parameters.
An insert file of equate names is provided to allow you to use mnemonics rather than key numbers. The
name of the insert file is ERRD.INS.IBAS. Use the $INCLUDE statement to insert this file if you want to
use equate names. The following list shows the codes returned in the argument return.code:
Code Meaning
0 No error
Examples
In this example, the file containing the parameter key equate names is inserted with the $INCLUDE
compiler directive. Later the top margin parameter for logical print channel 0 is interrogated. Print
channel 0 is the terminal unless a prior PRINTER ON statement has been executed to direct output to the
default printer. The top margin setting is returned in the argument TM.SETTING. Return codes are
returned in the argument RETURN.CODE.
$INCLUDE SYSCOM GETPU.INS.IBAS
CALL !GETPU(PU$TOPMARGIN,0,TM.SETTING,RETURN.CODE)
The next example does the same as the previous example but uses the key 4 instead of the equate name
PU$TOPMARGIN. Because the key number is used, it is not necessary for the insert file GETPU.INS.IBAS
to be included.
CALL !GETPU(4,0,TM.SETTING,RETURN.CODE)
The next example returns the current deferred time on print channel 0 in the variable TIME.RET:
CALL !GETPU(PU$DEFERTIME,0,TIME.RET,RETURN.CODE)
!GET.USER.COUNTS subroutine
Syntax
CALL !GET.USER.COUNTS (uv.users, max.uv.users, os.users)
Description
Use the !GET.USER.COUNTS subroutine to return a count of WebSphere DataStage and system users. If
any value cannot be retrieved, a value of -1 is returned.
max.uv.users (output) is the maximum number of licensed WebSphere DataStage users allowed on your
system.
!INLINE.PROMPTS subroutine
Syntax
CALL !INLINE.PROMPTS ( result , string )
Description
Use the !INLINE.PROMPTS subroutine to evaluate a string that contains in-line prompts. In-line prompts
have the following syntax:
result (output) is the variable that contains the result of the evaluation.
control specifies the characteristics of the prompt, and can be one of the following:
Table 83. INLINE prompts
Control Explanation
@(CLR) Clears the terminal screen.
@(BELL) Rings the terminal bell.
@(TOF) Issues a formfeed character: in most circumstances this
results in the cursor moving to the top left of the screen.
@ ( col , row ) Prompts at the specified column and row number on the
terminal.
A Always prompts when the in-line prompt containing the
control option is evaluated. If you do not specify this
option, the input value from a previous execution of the
prompt is used.
Cn Specifies that the nth word on the command line is used
as the input value. (Word 1 is the verb in the sentence.)
F ( filename , record . id [ , fm [ , vm [ ,sm ] ] ] ) Takes the input value from the specified record in the
specified file, and optionally, extracts a value (@VM), or
subordinate value (@SM), from the field (@FM). This
option cannot be used with the file dictionary.
In Takes the nth word from the command line, but prompts
if the word is not entered.
R ( string ) Repeats the prompt until an empty string is entered. If
string is specified, each response to the prompt is
appended by string between each entry. If string is not
specified, a space is used to separate the responses.
P Saves the input from an in-line prompt. The input is then
used for all in-line prompts with the same prompt text.
This is done until the saved input is overwritten by a
prompt with the same prompt text and with a control
option of A, C, I, or S, or until control returns to the
command prompt. The P option saves the input from an
in-line prompt in the current paragraph, or in other
paragraphs.
Sn Takes the nth word from the command (as in the In
control option), but uses the most recent command
entered at the command level to execute the paragraph,
rather than an argument in the paragraph. This is useful
where paragraphs are nested.
text The prompt to be displayed.
option A valid conversion code or pattern match. A valid
conversion code is one that can be used with the ICONV
function. Conversion codes must be enclosed in
parentheses. A valid pattern match is one that can be
used with the MATCHING keyword.
Note: Once a value has been entered for a particular prompt, the prompt continues to have that value
until a !CLEAR.PROMPTS subroutine is called, or control option A is specified. A !CLEAR.PROMPTS
subroutine clears all the values that have been entered for in-line prompts.
Example
A = ""
CALL !INLINE.PROMPTS(A,"You have requested the <<Filename>>
file")
PRINT "A"
!INTS subroutine
Syntax
CALL !INTS (result, dynamic.array)
Description
Use the !INTS subroutine to retrieve the integer portion of elements in a dynamic array.
result (output) contains a dynamic array that comprises the integer portions of the elements of
dynamic.array.
The !INTS subroutine returns a dynamic array, each element of which contains the integer portion of the
numeric value in the corresponding element of the input dynamic.array.
Example
A=33.0009:@VM:999.999:@FM:-4.66:@FM:88.3874
CALL !INTS(RESULT,A)
!MAKE.PATHNAME subroutine
Syntax
CALL !MAKE.PATHNAME (path1, path2, result, status)
Description
Use the !MAKE.PATHNAME subroutine to construct the full path name of a file. The
!MAKE.PATHNAME subroutine can be used to:
v Concatenate two strings to form a path name. The second string must be a relative path.
path1 (input) is a filename or partial path name. If path1 is an empty string, the current working directory
is used.
path2 (input) is a relative path name. If path2 is an empty string, the current working directory is used.
status (output) is the returned status of the operation. 0 indicates success. Any other number indicates
either of the following errors:
IE$NOTRELATIVE
path2 was not a relative path name.
IE$PATHNOTFOUND
The path name could not be found when !MAKE.PATHNAME tried to qualify it fully.
Example
!MATCHES subroutine
Syntax
CALL !MATCHES ( result , dynamic. array , match.pattern )
Description
Use the !MATCHES subroutine to test whether each element of one dynamic array matches the patterns
specified in the elements of the second dynamic array. Each element of dynamic.array is compared with
the corresponding element of match.pattern. If the element in dynamic.array matches the pattern specified
in match.pattern, 1 is returned in the corresponding element of result. If the element from dynamic.array is
not matched by the specified pattern, 0 is returned.
result (output) is a dynamic array containing the result of the comparison on each element in dynamic
array1.
When dynamic.array and match.pattern do not contain the same number of elements, the behavior of
!MATCHES is as follows:
v result always contains the same number of elements as the longer of dynamic.array or match.pattern.
Examples
The following example returns the value of the dynamic array as 1VM1VM1:
A=’AAA4A4’:@VM:2398:@VM:’TRAIN’
B=’6X’:@VM:’4N’:@VM:’5A’
CALL !MATCHES(RESULT,A,B)
In the next example, there are missing elements in match.pattern that are treated as though they contain a
pattern that matches an empty string. The result is 0VM0SM0FM1FM1.
R=’AAA’:@VM:222:@SM:’CCCC’:@FM:33:@FM:’DDDDDD’
S=’4A’:@FM:’2N’:@FM:’6X’
CALL !MATCHES(RESULT,R,S)
In the next example, the missing element in match.pattern is used as a test for an empty string in
dynamic.array, and the result is 1VM1FM1:
X=’AAA’:@VM:@FM:’’
Y=’3A’:@FM:’3A’
CALL !MATCHES(RESULT,X,Y)
!MESSAGE subroutine
Syntax
CALL !MESSAGE (key, username, usernum, message, status)
Description
Use the !MESSAGE subroutine to send a message to another user on the system. !MESSAGE lets you
change and report on the current user’s message status.
key (input) specifies the operation to be performed. You specify the option you require with the key
argument, as follows:
IK$MSGACCEPT
Sets message status to accept.
IK$MSGREJECT
Sets message status to reject.
IK$MSGSEND
Sends message to user.
IK$MSGSENDNOW
Sends message to user now.
IK$MSGSTATUS
Displays message status of user.
username (input) is the name of the user, or the TTY name, for send or status operations.
Note: The value of message is ignored when key is set to IK$MSGACCEPT, IK$MSGREJECT, or
IK$MSGSTATUS.
Example
CALL !MESSAGE (KEY,USERNAME,USERNUMBER,MESSAGE,CODE)
IF CODE # 0
THEN CALL !REPORT.ERROR
(’MY.COMMAND’,’!MESSAGE’,CODE)
!PACK.FNKEYS
Syntax
CALL !PACK.FNKEYS (trap.list, ftable)
Qualifiers
trap.list
A list of function numbers delimited by field marks (CHAR(254)), defining the specific keys that
are to be used as trap keys by the !EDIT.INPUT subroutine.
ftable A bit-significant string of trap keys used in the ftable parameter of the !EDIT.INPUT subroutine.
This string should not be changed in any way before calling the !EDIT.INPUT subroutine.
Description
The !PACK.FNKEYS subroutine converts a list of function key numbers into a bit string suitable for use
with the !EDIT.INPUT subroutine. This bit string defines the keys which cause !EDIT.INPUT to exit,
enabling the program to handle the specific keys itself.
trap.list can be a list of function key numbers delimited by field marks (CHAR(254)). Alternatively, the
mnemonic key name, listed below and in the UNIVERSE.INCLUDE file GTI.FNKEYS.IH, can be used:
Table 84. Mnemonic key names
Key Name Field Description
FK$FIN 1 Finish
FK$HELP 2 Help
FK$BSP 3 Backspace1
If ftable is returned as an empty string, an error in the trap.list array is detected, such as an invalid
function number. Otherwise ftable is a bit-significant string which should not be changed in any way
before its use with the !EDIT.INPUT subroutine.
Example
The following program sets up three trap keys using the !PACK.FNKEYS function, then uses the bit
string within the !EDIT.INPUT subroutine:
$INCLUDE UNIVERSE.INCLUDE GTI.FNKEYS.IH
* Set up trap keys of FINISH, UPCURSOR and DOWNCURSOR
TRAP.LIST = FK$FIN:@FM:FK$UP:@FM:FK$DOWN
CALL !PACK.FNKEYS(TRAP.LIST, Ftable)
* Start editing in INPUT mode, displaying contents in window
KEYS = IK$INS + IK$DIS
* Window edit is at x=20, y=2, of length 10 characters;
* the user can enter up to 30 characters of input into
TextBuffer,
* and the cursor is initially placed on the first character of
the
* window.
TextBuffer=""
CursorPos = 1
CALL
!EDIT.INPUT(KEYS,20,2,10,TextBuffer,CursorPos,30,Ftable,ReturnC
ode)
* On exit, the user’s input is within TextBuffer,
* CursorPos indicates the location of the cursor upon exiting,
* and ReturnCode contains the reason for exiting.
BEGIN CASE
CASE CODE = 0
* User pressed RETURN key
CASE CODE = FK$FIN
* User pressed the defined FINISH key
CASE CODE = FK$UP
* User pressed the defined UPCURSOR key
CASE CODE = FK$DOWN
* User pressed the defined DOWNCURSOR
key
CASE 1 * Should never happen
END CASE
!REPORT.ERROR subroutine
Syntax
CALL !REPORT.ERROR (command, subroutine, code)
Description
Use the !REPORT.ERROR subroutine to print explanatory text for a WebSphere DataStage or operating
system error code.
command is the name of the command that used the subroutine in which an error was reported.
subroutine is the name of the subroutine that returned the error code.
Text for values of code in the range 0 through 9999 is retrieved from the operating system. Text for values
of code over 10,000 is retrieved from the SYS.MESSAGES file. If the code has no associated text, a message
to that effect is displayed. Some WebSphere DataStage error messages allow text to be inserted in them.
In this case, code can be a dynamic array of the error number, followed by one or more parameters to be
inserted into the message text.
Examples
CALL !MESSAGE (KEY,USERNAME,USERNUMBER,MESSAGE,CODE)
IF CODE # 0
THEN CALL !REPORT.ERROR
(’MY.COMMAND’,’!MESSAGE’,CODE)
If code was IE$UNKNOWN.USER, and the user ID was joanna, !REPORT.ERROR would display the
following:
Error calling "!MESSAGE" from "MY.COMMAND" UniVerse
error 1757: joanna is not logged on
!SET.PTR subroutine
Syntax
CALL !SET.PTR (print.channel, width, length, top.margin,
bottom.margin,
mode, options)
Description
Use the !SET.PTR subroutine to set options for a logical print channel. This subroutine provides the same
functionality as the SETPTR command.
top.margin is the number of lines left at the top of the page. The default is 3.
bottom.margin is the number of lines left at the bottom of the page. The default is 3.
options are any of the printer options that are valid for the SETPTR command. These must be separated
by commas and enclosed by valid quotation marks.
If you want to leave a characteristic unchanged, supply an empty string argument and specify the option
NODEFAULT. If you want the default to be selected, supply an empty string argument without
specifying the NODEFAULT option.
Example
The following example sets the options so that printing is deferred until 12:00, and the job is retained in
the queue:
CALL !SET.PTR (0,80,60,3,3,1,’DEFER 12:00,RETAIN’)
!SET.PU subroutine
Syntax
CALL !SETPU (key, print.channel, new.value, return.code)
Description
Use the !SETPU subroutine to set individual parameters of any logical print channel.
Unlike !SET.PTR, you can specify only individual parameters to change; you need not specify parameters
you do not want to change. See the description of the !GETPU subroutine for a way to read individual
print.channel parameters.
key is a number indicating the parameter to be set (see ″Equate Names for Keys″).
return.code is the returned error code (see ″Equate Names for Return Code″).
The !SETPU subroutine lets you change individual parameters of logical print channels as designated by
print.channel. Print channel 0 is the terminal unless a PRINTER ON statement has been executed to send
In the following example, the file containing the parameter key equate names is inserted with the
$INCLUDE compiler directive. Later, the top margin parameter for logical print channel 0 is set to 10
lines. Return codes are returned in the argument RETURN.CODE.
$INCLUDE SYSCOM GETPU.INS.IBAS
CALL !SETPU(PU$TOPMARGIN,0,10,RETURN.CODE)
The next example does the same as the previous example, but uses the key 4 instead of the equate name
PU$TOPMARGIN. Because the key is used, it is not necessary for the insert file GETPU.INS.IBAS to be
included.
CALL !SETPU(4,0,10,RETURN.CODE)
!TIMDAT subroutine
Syntax
CALL !TIMDAT (variable)
Description
Use the !TIMDAT subroutine to return a dynamic array containing the time, date, and other related
information. The !TIMDAT subroutine returns a 13-element dynamic array containing information shown
in the following list.
variable is the name of the variable to which the dynamic array is to be assigned.
Field Description
1 Month (two digits).
2 Day of month (two digits).
3 Year (two digits).
4 Minutes since midnight (integer).
5 Seconds into the minute (integer).
6 Ticks1 of last second since midnight (integer). Always returns 0.
7 CPU seconds used since entering WebSphere DataStage.
8 Ticks of last second used since login (integer).
9 Disk I/O seconds used since entering WebSphere DataStage. Always returns -1.
10 Ticks of last disk I/O second used since login (integer). Always returns -1.
11 Number of ticks per second.
12 User number.
13 Login ID (user ID).
1
Tick refers to the unit of time your system uses to measure real time.
Use the following functions for alternative ways of obtaining time and date information:
Use this function...
To obtain this data...
DATE ( )
Data in fields 1, 2, and 3 of the dynamic array returned by the !TIMDAT subroutine
Example
CALL !TIMDAT(DYNARRAY)
FOR X = 1 TO 13
PRINT ’ELEMENT ’:X:’, DYNARRAY
NEXT X
!USER.TYPE subroutine
Syntax
CALL !USER.TYPE (type, admin)
Description
Use the !USER.TYPE subroutine to return the user type of the current process and a flag to indicate if the
user is an Administrator.
type is a value that indicates the type of process making the subroutine call. type can be either of the
following:
Table 87. Process types
Equate Name Value Meaning
U$NORM 1 Normal user
U$PH 65 Phantom
admin is a value that indicates if the user making the call is an Administrator. Possible values of admin are
1, if the user is an Administrator, and 0, if the user is not an Administrator.
An insert file of equate names is provided for the !USER.TYPE values. To use the equate names, specify
the directive $INCLUDE SYSCOM USER_TYPES.H when you compile your program. (For PI/open
compatibility you can specify $INCLUDE SYSCOM USER_TYPES.INS.IBAS.)
Example
In this example, the !USER.TYPE subroutine is called to determine the type of user. If the user is a
phantom, the program stops. If the user is not a phantom, the program sends a message to the terminal
and continues processing.
ERROR.ACCOUNTS.FILE: CALL !USER.TYPE(TYPE, ADMIN)
IF TYPE = U&PH THEN STOP
ELSE PRINT ’Error on opening ACCOUNTS file’
!VOC.PATHNAME subroutine
Syntax
CALL !VOC.PATHNAME (data/dict, voc.entry, result, status)
Use the !VOC.PATHNAME subroutine to extract the path names for the data file or the file dictionary of
a specified VOC entry.
An insert file of equate names is provided for the data/dict values. To use the equate names, specify the
directive $INCLUDE SYSCOM INFO_KEYS.H when you compile your program. (For PI/open
compatibility you can specify $INCLUDE SYSCOM INFO_KEYS.INS.IBAS.)
The result of the operation is returned in the status argument, and has one of the following values:
Value Result
0 The operation executed successfully.
IE$PAR
A bad parameter was used in data/dict or voc.entry.
IE$RNF
The VOC entry record cannot be found.
Example
CALL !VOC.PATHNAME (IK$DATA,"VOC",VOC.PATH,STATUS)
IF STATUS = 0
THEN PRINT "VOC PATHNAME = ":VOC.PATH
The information center is installed as a common service with IBM® Information Server. The information
center contains help for most of the product interfaces, as well as complete documentation for all product
modules in the suite.
A subset of the product documentation is also available online from the product documentation library at
publib.boulder.ibm.com/infocenter/iisinfsv/v8r1/index.jsp.
PDF file books are available through the IBM Information Server software installer and the distribution
media. A subset of the information center is also available online and periodically refreshed at
www.ibm.com/support/docview.wss?rs=14&uid=swg27008803.
You can also order IBM publications in hardcopy format online or through your local IBM representative.
You can send your comments about documentation in the following ways:
v Online reader comment form: www.ibm.com/software/data/rcf/
v E-mail: [email protected]
Contacting IBM
You can contact IBM for customer support, software services, product information, and general
information. You can also provide feedback on products and documentation.
Customer support
For customer support for IBM products and for product download information, go to the support and
downloads site at www.ibm.com/support/us/.
You can open a support request by going to the software support service request site at
www.ibm.com/software/support/probsub.html.
My IBM
You can manage links to IBM Web sites and information that meet your specific technical support needs
by creating an account on the My IBM site at www.ibm.com/account/us/.
Software services
For information about software, IT, and business consulting services, go to the solutions site at
www.ibm.com/businesssolutions/us/en.
Product feedback
You can provide general product feedback through the Consumability Survey at www.ibm.com/
software/data/info/consumability-survey.
Documentation feedback
You can click the feedback link in any topic in the information center to comment on the information
center.
You can also send your comments about PDF file books, the information center, or any other
documentation in the following ways:
v Online reader comment form: www.ibm.com/software/data/rcf/
v E-mail: [email protected]
required_item
required_item
optional_item
If an optional item appears above the main path, that item has no effect on the execution of the syntax
element and is used only for readability.
optional_item
required_item
v If you can choose from two or more items, they appear vertically, in a stack.
If you must choose one of the items, one item of the stack appears on the main path.
required_item required_choice1
required_choice2
If choosing one of the items is optional, the entire stack appears below the main path.
required_item
optional_choice1
optional_choice2
If one of the items is the default, it appears above the main path, and the remaining choices are shown
below.
default_choice
required_item
optional_choice1
optional_choice2
v An arrow returning to the left, above the main line, indicates an item that can be repeated.
required_item repeatable_item
If the repeat arrow contains a comma, you must separate repeated items with a comma.
required_item repeatable_item
A repeat arrow above a stack indicates that you can repeat the items in the stack.
v Sometimes a diagram must be split into fragments. The syntax fragment is shown separately from the
main syntax diagram, but the contents of the fragment should be read as if they are on the main path
of the diagram.
required_item fragment-name
Fragment-name:
required_item
optional_item
v Keywords, and their minimum abbreviations if applicable, appear in uppercase. They must be spelled
exactly as shown.
v Variables appear in all lowercase italic letters (for example, column-name). They represent
user-supplied names or values.
v Separate keywords and parameters by at least one space if no intervening punctuation is shown in the
diagram.
v Enter punctuation marks, parentheses, arithmetic operators, and other symbols, exactly as shown in the
diagram.
v Footnotes are shown by a number in parentheses, for example (1).
The IBM Information Server product modules and user interfaces are not fully accessible. The installation
program installs the following product modules and components:
v IBM Information Server Business Glossary Anywhere
v IBM Information Server FastTrack
v IBM Metadata Workbench
v IBM WebSphere® Business Glossary
v IBM WebSphere DataStage® and QualityStage™
v IBM WebSphere Information Analyzer
v IBM WebSphere Information Services Director
Accessible documentation
Accessible documentation for IBM Information Server products is provided in an information center. The
information center presents the documentation in XHTML 1.0 format, which is viewable in most Web
browsers. XHTML allows you to set display preferences in your browser. It also allows you to use screen
readers and other assistive technologies to access the documentation.
IBM may not offer the products, services, or features discussed in this document in other countries.
Consult your local IBM representative for information on the products and services currently available in
your area. Any reference to an IBM product, program, or service is not intended to state or imply that
only that IBM product, program, or service may be used. Any functionally equivalent product, program,
or service that does not infringe any IBM intellectual property right may be used instead. However, it is
the user’s responsibility to evaluate and verify the operation of any non-IBM product, program, or
service.
IBM may have patents or pending patent applications covering subject matter described in this
document. The furnishing of this document does not grant you any license to these patents. You can send
license inquiries, in writing, to:
For license inquiries regarding double-byte character set (DBCS) information, contact the IBM Intellectual
Property Department in your country or send inquiries, in writing, to:
The following paragraph does not apply to the United Kingdom or any other country where such
provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION
PROVIDES THIS PUBLICATION ″AS IS″ WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR
IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
NON-INFRINGEMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some
states do not allow disclaimer of express or implied warranties in certain transactions, therefore, this
statement may not apply to you.
This information could include technical inaccuracies or typographical errors. Changes are periodically
made to the information herein; these changes will be incorporated in new editions of the publication.
IBM may make improvements and/or changes in the product(s) and/or the program(s) described in this
publication at any time without notice.
Any references in this information to non-IBM Web sites are provided for convenience only and do not in
any manner serve as an endorsement of those Web sites. The materials at those Web sites are not part of
the materials for this IBM product and use of those Web sites is at your own risk.
IBM may use or distribute any of the information you supply in any way it believes appropriate without
incurring any obligation to you.
Licensees of this program who wish to have information about it for the purpose of enabling: (i) the
exchange of information between independently created programs and other programs (including this
one) and (ii) the mutual use of the information which has been exchanged, should contact:
Such information may be available, subject to appropriate terms and conditions, including in some cases,
payment of a fee.
The licensed program described in this document and all licensed material available for it are provided
by IBM under terms of the IBM Customer Agreement, IBM International Program License Agreement or
any equivalent agreement between us.
Any performance data contained herein was determined in a controlled environment. Therefore, the
results obtained in other operating environments may vary significantly. Some measurements may have
been made on development-level systems and there is no guarantee that these measurements will be the
same on generally available systems. Furthermore, some measurements may have been estimated through
extrapolation. Actual results may vary. Users of this document should verify the applicable data for their
specific environment.
Information concerning non-IBM products was obtained from the suppliers of those products, their
published announcements or other publicly available sources. IBM has not tested those products and
cannot confirm the accuracy of performance, compatibility or any other claims related to non-IBM
products. Questions on the capabilities of non-IBM products should be addressed to the suppliers of
those products.
All statements regarding IBM’s future direction or intent are subject to change or withdrawal without
notice, and represent goals and objectives only.
This information is for planning purposes only. The information herein is subject to change before the
products described become available.
This information contains examples of data and reports used in daily business operations. To illustrate
them as completely as possible, the examples include the names of individuals, companies, brands, and
products. All of these names are fictitious and any similarity to the names and addresses used by an
actual business enterprise is entirely coincidental.
COPYRIGHT LICENSE:
This information contains sample application programs in source language, which illustrate programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs
in any form without payment to IBM, for the purposes of developing, using, marketing or distributing
application programs conforming to the application programming interface for the operating platform for
which the sample programs are written. These examples have not been thoroughly tested under all
conditions. IBM, therefore, cannot guarantee or imply reliability, serviceability, or function of these
programs.
Each copy or any portion of these sample programs or any derivative work, must include a copyright
notice as follows:
© (your company name) (year). Portions of this code are derived from IBM Corp. Sample Programs. ©
Copyright IBM Corp. _enter the year or years_. All rights reserved.
If you are viewing this information softcopy, the photographs and color illustrations may not appear.
IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business
Machines Corporation in the United States, other countries, or both. If these and other IBM trademarked
terms are marked on their first occurrence in this information with a trademark symbol (® or ™), these
symbols indicate U.S. registered or common law trademarks owned by IBM at the time this information
was published. Such trademarks may also be registered or common law trademarks in other countries. A
current list of IBM trademarks is available on the Web at ″Copyright and trademark information″ at
www.ibm.com/legal/copytrade.shtml.
Adobe, the Adobe logo, PostScript, and the PostScript logo are either registered trademarks or trademarks
of Adobe Systems Incorporated in the United States, and/or other countries.
Intel, Intel logo, Intel Inside, Intel Inside logo, Intel Centrino, Intel Centrino logo, Celeron, Intel Xeon,
Intel SpeedStep, Itanium, and Pentium are trademarks or registered trademarks of Intel Corporation or its
subsidiaries in the United States and other countries.
Linux is a registered trademark of Linus Torvalds in the United States, other countries, or both.
Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in the
United States, other countries, or both.
ITIL is a registered trademark and a registered community trademark of the Office of Government
Commerce, and is registered in the U.S. Patent and Trademark Office.
UNIX is a registered trademark of The Open Group in the United States and other countries.
Cell Broadband Engine is a trademark of Sony Computer Entertainment, Inc. in the United States, other
countries, or both and is used under license therefrom.
Java and all Java-based trademarks are trademarks of Sun Microsystems, Inc. in the United States, other
countries, or both.
The United States Postal Service owns the following trademarks: CASS, CASS Certified, DPV, LACSLink,
ZIP, ZIP + 4, ZIP Code, Post Office, Postal Service, USPS and United States Postal Service. IBM
Corporation is a non-exclusive DPV and LACSLink licensee of the United States Postal Service.
Other company, product, or service names may be trademarks or service marks of others.
Notices 505
506 BASIC Reference Guide
Index
Special characters @COMMAND variable 459
@COMMAND.STACK variable 459
@SYS.BELL variable 462
@SYSTEM.RETURN.CODE variable 462
- operator [-] 14 @CONV variable 459 @SYSTEM.SET variable 462
-= operator 19 @CRTHIGH variable 459 @TERM.TYPE variable 462
: operator 15 @CRTWIDE variable 459 @TIME variable 462
:= operator 19 @DATA.PENDING variable 459 @TM variable 462
! statement 3, 57 @DATE variable 459 @TRANSACTION variable 41, 462
!ASYNC subroutine 467 @DAY variable 459 @TRANSACTION.ID variable 41, 462
!EDIT.INPUT subroutine 467 @DICT variable 459 @TRANSACTION.LEVEL variable 41,
!ERRNO subroutine 473 @FILE.NAME variable 459 462
!FCMP subroutine 474 @FILENAME variable 460 @TRUE variable 462
!GET.KEY subroutine 474 @FM variable 460 @TTY variable 462
!GET.PARTNUM subroutine 475 @FORMAT variable 460 @USER.NO variable 462
!GET.PATHNAME subroutine 477 @HDBC variable 460 @USER.RETURN.CODE variable 462
!GET.USER.COUNTS subroutine 480 @HEADER variable 460 @USER0 variable 462
!GETPU subroutine 477 @HENV variable 460 @USER1 variable 462
!INLINE.PROMPTS subroutine 480 @HSTMT variable 460 @USER2 variable 462
!INTS subroutine 482 @ID variable 460 @USER3 variable 462
!MAKE.PATHNAME subroutine 482 @IM variable 460 @USER4 variable 462
!MATCHES subroutine 483 @ISOLATION variable 41, 460 @USERNO variable 462
!MESSAGE subroutine 484 @LEVEL variable 460 @variables 41, 459, 463
!PACK.FNKEYS subroutine 485 @LOGNAME variable 460 @VM variable 462
!REPORT.ERROR subroutine 489 @LPTRHIGH variable 460 @WHO variable 463
!SET.PTR subroutine 489 @LPTRWIDE variable 460 @YEAR variable 463
!SETPU subroutine 490 @MONTH variable 460 &PH& file 25
!TIMDAT subroutine 493 @MV variable 460 # operator 17
!USER.TYPE subroutine 494 @NB variable 460 #INCLUDE statement 28, 58
!VOC.PATHNAME subroutine 495 @ND variable 460 + operator 14
/ operator 14 @NEW variable 460 += operator 19
$* statement 58 @NI variable 460 = operator 17, 19
$CHAIN statement 28, 59 @NS variable 461 ^ operator 14
$COPYRIGHT statement 60 @NULL variable 8, 16, 92, 461 ~ (tilde) 234
$DEFINE statement 28, 60 @NULL.STR variable 9, 16, 92, 108, 109,
$EJECT statement 61 461
$IFDEF statement 30, 62
$IFNDEF statement 30, 62
@NV variable 461
@OLD variable 461
A
$INCLUDE statement 28, 63 A conversion 414
@OPTION variable 461
$INSERT statement 26, 28, 63 ABORT statement 87
@PARASENTENCE variable 461
$MAP statement 64 ABORTE statement 69, 88
@PATH variable 461
$OPTIONS statement 29, 65 ABORTM statement 69, 88
@RECCOUNT variable 461
options 65, 69 ABS function 88
@RECORD variable 461
STATIC.DIM option 10 ABSS function 88
and ITYPE function 213
VEC.MATH option 15, 20 accessibility 497
@RECUR0 variable 461
$PAGE statement 71 ACID properties 39
@RECUR1 variable 461
$UNDEFINE statement 28, 71 ACOS function 89
@RECUR2 variable 461
* operator 14 ADDS function 89
@RECUR3 variable 461
* statement 3, 72 algebraic functions 414
@RECUR4 variable 461
** operator 14 ALPHA function 90
@SCHEMA variable 461
\[ \] operator 86 alphabetic characters 4
@SELECTED variable 461
\> operator 17 AND operator 18, 90, 95
@SENTENCE variable 310, 461
\>\< operator 17 ANDS function 90
@SM variable 461
\>= operator 17 angle brackets (\< \>) 72, 148, 292
@SQL.CODE variable 461
\< \> operator 17, 72, 148, 292 Arabic numeral conversion 437
@SQL.DATE variable 461
\< operator 17 arguments, passing to subroutines 2
@SQL.ERROR variable 461
\<= operator 17 arithmetic operators 14, 15
@SQL.STATE variable 461
@ expressions in INPUT statements 203 and dynamic arrays 20
@SQL.TIME variable 461
@ function 73 and multivalued data 15
@SQL.WARNING variable 462
@ABORT.CODE variable 459 and the null value 14
@SQLPROC.NAME variable 462
@ACCOUNT variable 459 array variables 10, 12
@SQLPROC.TX.LEVEL variable 462
@AM variable 459 arrays
@STDFIL variable 254, 259, 462
@ANS variable 459 assigning values to 232
@SVM variable 462
@AUTHORIZATION variable 94, 459 dimensioned 10, 134
Index 509
G INMAT function 202
after MATPARSE statement 236
levels, see isolation levels 95
line number table, suppressing 25, 27
G conversion 426 after MATREAD statement 238 list variables, see select list variables 13
GE operator 17 after OPEN statement 255 listing object code 55
GES function 176 INPUT @ statement 203 LN function 224
GET statement 176 INPUT statements 203 local cataloging 32
GET(ARG.) statement 179 and DATA statement 125 local variables 107
GETLIST statement 180 INPUT.ELSE option 67, 204 LOCALEINFO function 224
GETLOCALE function 181, 224 INPUTAT option 67 LOCATE statement 225
GETREM function 181 INPUTCLEAR statement 205 LOCATE.R83 option 67
GETX statement 176, 179 INPUTDISP statement 206 lock escalation, example 45
global cataloging 32 INPUTDP statement 206 LOCK statement 228
GOSUB statement 2, 182 INPUTERR statement 207 locks
and RETURN statement 294 INPUTIF statement 203 and MATREADL statement 239
with ON statement 253 INPUTNULL statement 207 and MATREADU statement 239
GOTO statement 183 INPUTTRAP statement 208 and READL statement 276
with ON statement 253 INS statement 208 and READU statement 242, 276, 381
granularity 35 INSERT function 210 and READVL statement 276
group extraction 426 and LOCATE statement 226 compatibility 35
GROUP function 183 INT function 212 deadlocks 38
GROUPSTORE statement 184 INT.PRECISION option 67 exclusive file lock 38
GT operator 17 intent file locks 38 file lock 159
GTS function 185 internal system date 126 granularity 35
international date conversion 424 intent file lock 38
international time conversion 442 process lock 376
H intrinsic functions, see functions 2 releasing 242, 287, 376, 381
HEADER.BRK option 66 ISNULL function 17, 212 semaphore lock 228
HEADER.DATE option 66, 171, 188 with CASE statement 105 shared file lock 37
HEADER.EJECT option 66, 187 with IF statement 194 shared record lock 35
HEADING statement 186 ISNULLS function 17, 212 transactions and 40
HEADINGE statement 69, 187 isolation levels 44, 95 types 35, 157
HEADINGN statement 69, 187 minimum locks for 44 update record lock 36
hexadecimal conversion 436 types 42 well-formed writes and 44
hexadecimal equivalents 412 isolation property 40 logical operators 18
host name and SYSTEM function 336 ISOMODE parameter 43 AND 18
HUSH statement 189 ITYPE function 213 and the null value 18
NOT 18
OR 18
I K LOOP statement 229
loops
IBM support 497 KEEP.COMMON keyword 107
FOR...NEXT 172, 246
ICHECK function 190 keyboard keys
LOOP...REPEAT 229
ICONV function 192, 413, 436 control 215, 219, 220
lost updates 43
ICONVS function 193, 416 escape 215, 219, 220
LOWER function 231
identifiers function 216, 219, 220, 485
LT operator 17
removing 28 unsupported 217
LTS function 232
replacing 28 KEYEDIT statement 214
IF KEYEXIT statement 218
operator 18 KEYIN function 219
statement 194 KEYTRAP statement 220 M
IFS function 195 keywords 1, 443 masked character conversions 427
ILPROMPT function 195 definition 2 MAT statement 232
in-line prompts, clearing 112 MATBUILD statement 233
IN2.SUBSTR option 66 MATCH operator 18, 233
include files, UVLOCALE.H 181, 224 L MATCHFIELD function 235
mathematical functions 414, 424
INCLUDE statement 28, 197 L conversion 427
INDEX function 198 MATPARSE statement 236
labels, statement 4
INDEXS function 199 and INMAT function 202
LE operator 17
INDICES function 199 MATREAD statement 237
LEFT function 221
and secondary indexes 199 and INMAT function 202
legal notices 503
in transactions 201 MATREADL statement 239
LEN function 221
INFO.ABORT option 67 and INMAT function 202
LENDP function 222
INFO.CONVERT option 67 matrices 10
length function 427
INFO.ENTER option 67 zero element 10
LENS function 222
INFO.INCLUDE option 67 MATWRITE statement 240
LENSDP function 223
INFO.LOCATE option 67, 227 MATWRITEU statement 242
LES function 223
INFO.MARKS option 67 MAXIMUM function 243
LET statement 224
INFO.MOD option 67 MAXRLOCK parameter 45
Index 511
REPEAT statement 229 SINH function 315 STATUS function (continued)
REPLACE function 73, 148, 292 SLEEP statement 315 READBLK statement 324
reserved words in BASIC 443 SMUL function 316 READL statement 324
RETURN (value) statement 175 software services 497 READSEQ statement 324
and DEFFUN statement 129 soundex conversion 440 READT statement 324
RETURN statement 2, 104, 294 SOUNDEX function 317 READU statement 324
RETURN TO statement 182 source code 1 READVL statement 324
REUSE function 23, 295 comments 3 READVU statement 324
REVREMOVE statement 296 spaces in 4 REWIND statement 324
REWIND statement 297 syntax 3 RPC.CALL function 324
RIGHT function 298 tabs in 4 RPC.CONNECT function 324
RND function 298 SPACE function 317 RPC.DISCONNECT function 324
and RANDOMIZE statement 273 spaces WEOF statement 324
RNEXT.EXPL option 68, 280 in numeric constants 8 WRITE statements 323
ROLLBACK statement 299 in source code 4 WRITET statement 324
Roman numeral conversion 437 removing 362, 363, 364, 365, 375 STATUS statement 325
RPC.CALL function 299 SPACES function 318 values 326
RPC.CONNECT function 301 special characters 4 STOP statement 328
RPC.DISCONNECT function 301 SPLICE function 318 STOP.MSG option 68, 88, 328
RUN command 31 SQRT function 318 STOPE statement 69, 328
options 31 square brackets (\) 86 STOPM statement 69, 328
SQUOTE function 319 STORAGE statement 328
SSELECT statement 319 STR function 329
S SSELECTN statement 320
SSELECTV statement 320
string functions 2, 424
string operators 15
S (soundex) conversion 440
SSUB function 321 strings
S (substitution) conversion 440
standard arrays 10 comparing 17
SADD function 302
matrices 10 determining length 18
saving variables in unnamed
vectors 10 see also character strings 17
common 107
statement labels STRS function 329
scientific notation 8
cross-reference table of 25, 26 SUBR function 330
SCMP function 303
definition 4 SUBROUTINE statement 2, 103, 331
screen readers 497
statements 1 subroutines
SDIV function 303
assignment 3 calling 103
secondary indexes and BASIC INDICES
control 3 definition 2
function 199
types 3 entering external 141
SEEK statement 304
STATIC.DIM option 10, 68, 135 list of 465
SEEK(ARG.) statement 305
STATUS function 321 passing arguments to 2
select lists
after HUSH statement 189 passing arrays to 104, 135
clearing 113
after ICONV function 192 returning from 294
variables 13, 113, 307, 320
after ICONVS function 193 returning values from 330
SELECT statement 306
after OCONV function 250 vector functions 20
SELECTE statement 307
after OCONVS function 252 SUBS function 332
SELECTINDEX statement 308
after OPENPATH statement 259 substitution 440
SELECTINFO function 309
after READSEQ statement 281 substring operator 16
SELECTN statement 69, 307
after READT statement 283, 297, 378, substrings 16, 17
SELECTV statement 69, 306
388 and assignment operators 16
semaphore locks 228
after RECORDLOCKED definition 16
SEND statement 309
statement 286 extracting 16, 86
SENTENCE function 310
after RPC.CALL function 300 SUBSTRINGS function 332
SEQ function 310
after RPC.CONNECT function 301 subtransactions 39
SEQ.255 option 68, 311
after RPC.DISCONNECT subvalue marks 11
SEQS function 311
function 302 SUM function 333
sequential I/O 4
after SELECTINDEX statement 308 SUMMATION function 333
sequential processing 114, 123, 162, 247,
after WRITESEQ statement 385 SUPP.DATA.ECHO option 68, 204
257, 260, 277, 281, 304, 355, 384, 386
BSCAN statement 100, 322 support, customer 497
serializability
DELETE statement 322 symbol table, suppressing 25, 27
lost updates and 43
FILELOCK statement 322 syntax, source code 3
property 40
FMT function 322 system date 126
SET TRANSACTION ISOLATION LEVEL
GET statement 322 system delimiters 11
statement 40, 312
GETX statement 322 SYSTEM function 334
SETLOCALE function 313
ICONV function 322 host name 336
SETPTR command 489
INPUT @ statement 323 values 335
SETREM statement 314
MATWRITE statement 323
shared file locks 37
OCONV function 322
shared memory 34
shared record locks 35
OPEN statements 323
READ statement 324
T
SIN function 315 T conversion 441
Index 513
514 BASIC Reference Guide
Printed in USA
LC18-9897-01
Spine information: