Using the Compiler

Linking

Chapter 6: Directives for Compiler

This chapter lists and describes the directives you can use to control the Compiler . The first section is intended to help you find directives for particular purposes; it lists their names, organized into categories, with a very brief description of the purpose of each. The second section contains complete descriptions of the directives, in alphabetical order.

6.1 Directives by Category

These are categories of Compiler directives. Select one to see a list of directives in that category. You can then select a directive to see its parameters and other details.

Enabling Language Features
Choosing Run-time Behavior
Compiler Control
Compiling for Debugging
Format of Your Data Files
Object Code, Size and Optimization
Reserved Directives

An asterisk (*) next to the description of a directive means you need an add-on product before you can use it. If you use it without the appropriate add-on product, the Compiler does not give an error message but the directive does not have the desired effect.

6.1.1 Enabling Language Features

These directives change which language features your Compiler accepts.

6.1.1.1 Add Features

6.1.1.2 Dialect

6.1.1.3 Mainframe Compatibility

6.1.1.4 Reserved Word Control

6.1.1.5 Speed

6.1.2 Run-time Behavior

These directives change the definition of language features and so change the logic of your program.

Some directives listed under Dialect also affect behavior.

6.1.2.1 Compatibility with Other COBOL Dialects

6.1.2.2 Compatibility with Older Micro Focus Products

6.1.2.3 Mainframe Compatibility

6.1.2.4 Speed or Size

6.1.3 Compiler Control

6.1.3.1 Compile/Link Files

6.1.3.2 Directives Control

6.1.3.3 Error & Flag Messages

6.1.3.4 Listing

6.1.3.5 Screen

6.1.4 Compiling for Debugging

These directives make the Compiler create files containing extra information about the program, for input to software used for debugging.

6.1.4.1 Add-on Products

6.1.5 Format of Your Data Files

These directives change the format in which data is stored in your data files.

6.1.6 Object Code, Size and Optimization

These directives change the object code produced without changing the logic of your program.

6.1.6.1 External Handlers

6.1.6.2 File Handling

6.1.6.3 Interprogram Communication

6.1.6.4 Object File Format

6.1.6.5 Size and Speed

6.1.7 Reserved Directives

These directives are reserved or are maintained for compatibility with earlier products and have no effect. Do not use them.

6.1.7.1 Add-on Products

6.1.7.2 Internal or Future Use Only

6.2 Key

Descriptions for all of the Compiler directives appear alphabetically. Each description contains the following entries.

A brief description of the function of DIRECTIVE-NAME.

Syntax:

Syntax in the form:

                                 +--------------™ 
                                 v              | 
 >>-.---.-.-------DIRECTIVE-NAME--"parameter(s)"-.-><  
    .-/-+ ..----.-DIRECTIVE-NAME-----------------+ 
           .-NO-+

The syntax of the Compiler directives is shown using diagrams called "railroad tracks", in which a directive and its parameters are shown joined by lines indicating the order in which they should be written. You read these diagrams from left-to-right. Each diagram starts with >> and ends with '". Sometimes the track forks to show alternatives and then joins up again. The length of a track has no significance.

The loop over the parameter in the above example indicates that it can be repeated one or more times.

Parameters for directives are shown in quotation marks (" "), although you can use parentheses instead unless otherwise stated. On UNIX, the equals sign (=) can be used as an alternative. When quotation marks are used the parameter can contain spaces, whereas no spaces are allowed in a parameter surrounded by parentheses.

On DOS, Windows and OS/2, if a parameter is specified after a comma, but the filename is omitted, the directive must be preceded by a slash (/). Otherwise, the directive is incorrectly assumed to be a filename.

On UNIX, if you use quotation marks or the equals sign, you must escape them using the backslash character (\). The slash (/) in the syntax diagrams is for use on DOS, Windows and OS/2. It should be ignored for UNIX environments and does not appear in diagrams for UNIX-specific directives.

Parameters:

Lists and describes valid parameter(s), if any, for the directive.

Properties:

Dependencies:

Shows if the setting of this directive changes the setting of any other directives, or if any other directives affect the setting of this directive.

There are a number of different cases, identified under this heading by the following keywords:

Comments:

Contains any additional information about the directive.

In the menus, an asterisk (*) next to the name of a directive means you need an add-on product before you can use it. If you use it without the appropriate add-on product, the Compiler does not give an error message but the directive does not have the desired effect.

6.3 List of Directives

 01SHUFFLE

Changes the addresses of level-01 items in the Working-Storage Section so that they do not cross 64K segment boundaries.

Syntax:

 >>-.---.-.----.--01SHUFFLE-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to NO01SHUFFLE at end by NESTCALL.

Comments:

If a data item would cross as segment boundary, specifying 01SHUFFLE causes the level-01 item to be moved in memory so that it starts at a segment boundary. The gap left is filled by any following level-01 items that fit. This results in level-01 items not being contiguous in memory when the program is loaded.

Use of this directive can cause an increase in the size of the data segment of a program. This directive is not valid for nested programs.

Although this directive does work with 32-bit COBOL systems, it is only really useful with the 16-bit COBOL system.

See also:

REF Compiler directive
SHOWSHUFFLE Compiler directive

 64KPARA

Makes the Compiler produce segment-breaking information more often than at each paragraph, for programs containing paragraphs that produce more than 64K of code.

Syntax:

 >>-.---.-.----.--64KPARA-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Must set SEGSIZE"integer" so integer is nonzero.

Comments:

Use of this directive lengthens the compilation process slightly but does not affect the code produced.

This directive is mandatory for programs which contain paragraphs that produce more than 64K of code.

See also:

64KSECT Compiler directive
SEGSIZE Compiler directive

 64KSECT

Makes the Compiler produce segment-breaking information at each paragraph instead of each section, for programs containing sections that produce more than 64K of code.

Syntax:

 >>-.---.-.----.--64KSECT-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Must set SEGSIZE"integer" so integer is nonzero.

Comments:

Use of this directive lengthens the compilation process slightly but does not affect the code produced.

This directive is mandatory for programs which contain sections that produce more than 64K of code.

See also:

64KPARA Compiler directive
SEGSIZE Compiler directive

ACCEPTREFRESH

Specifies whether the data areas associated with Screen Section data are updated from their corresponding Working-Storage Section items before an ACCEPT statement.

Syntax:

 >>-.---.-.----.--ACCEPTREFRESH-------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to ACCEPTREFRESH immediately by MS, IBM-MS or PC1.

Comments:

With ACCEPTREFRESH specified, before an ACCEPT statement that references a Screen Section data item, all data areas associated with Screen Section items subsidiary to the item being accepted are modified as follows:

• If the Screen Section item includes the USING phrase, an implicit MOVE is generated from the corresponding Working-Storage Section item

• If the Screen Section item includes the TO phrase the data area is initialized to spaces.

With NOACCEPTREFRESH specified, the Screen Section data area is left as it was following the last ACCEPT or DISPLAY.

ADDRSV

Adds one or more specific reserved words to the reserved word list so that they are recognized as reserved words in your program. The specified reserved words are added, whatever dialect directives are in force.

Syntax:

                   +----------™ 
                   v          | 
 >>--.---.--ADDRSV--"rsv-word"--------------><  
     .-/-+

Parameters:

Properties:

Comments:

The specified reserved words are added, whatever dialect directives are in force.

This directive does not appear in the list created with the SETTING directive.

See also:

SETTING Compiler directive

ADDSYN

Defines a user-defined reserved word to be synonymous with an existing reserved word.

Syntax:

On DOS, Windows and OS/2:

 >>-.---.--ADDSYN--"rsv-word" = "user-word"->< 
    .-/-+

On UNIX:

 >>--ADDSYN-"(rsv-word) == (user-word)"-----'"

Parameters:

Properties:

Comments:

The equals signs (==) must be surrounded by spaces.

This directive does not appear in the list created with the SETTING directive.

ADV

Causes a control character to be inserted at the start of each line in a print file. This is for compatibility with mainframe operation.

Syntax:

 >>-.---.-.----.--ADV-----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

ADV has no effect unless FILETYPE"11" is set.

See also:

FILETYPE Compiler directive

ALIAS

Prevents the Compiler from performing certain optimizations that are unsafe when a program uses aliasing.

Syntax:

 >>-.---.-.----.--ALIAS---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

Using the SET ADDRESS OF statement you can access the same data in two different ways; that is, in the Working-Storage Section and as an item in the Linkage Section. This technique is known as "aliasing". Any program that uses aliasing must be compiled with ALIAS specified.

ALIGN

Specifies the memory boundaries on which data items of level-01 or level-77 are aligned.

Syntax:

 >>-.---.--ALIGN--"integer"-----------------><  
    .-/-+

Parameters:

Properties:

Dependencies:

Set to ALIGN"8" immediately by IBMCOMP or NORM.
Set to ALIGN"2" immediately by RM or RM"ANSI".

Comments:

Data items of level-01 are aligned on byte boundaries that are multiples of the value given.

This can help produce more efficient execution results but does mean more memory might be used.

Setting integer to a number that is not a multiple of four could impair the efficiency of the generated code.

ALPHASTART

Sets the number from which the Compiler counts positions in the collating sequence when compiling the ALPHABET clause.

Syntax:

 >>-.---.--ALPHASTART--"integer"------------><  
    .-/-+

Parameters:

Properties:

Comments:

For ANSI conformance use ALPHASTART"1".

Example:

With ALPHASTART"1", the COBOL statement:

alphabet myalpha is 66, 67

declares an alphabet consisting of "A" and "B" since, counting from 1, these are the 66th and 67th characters in the ASCII collating sequence. With ALPHASTART"0", MYALPHA consists of "B" and "C".

ALTER

Allows ALTER statements in your program.

Syntax:

 >>-.---.-.----.--ALTER---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

If ASSUME is set, set to NOALTER at end by NOTRICKLE during the generate phase.

Comments:

If you know there are no ALTER statements in your program, specifying NOALTER enables the Compiler to produce slightly more efficient code.

AMODE

Provides compatibility with mainframe-style pointers.

Syntax:

>>--+---+---+----+--AMODE----"format"--------><
    |-/-|   |-NO-|

Parameters:

Properties:

Comments:

The AMODE directive must be used with any program module which:

• Must create mainframe-style pointer items (using SET pointer-item TO ADDRESS OF data-item syntax)

• Uses mainframe-style pointer items (using SET ADDRESS OF linkage-item TO pointer-item syntax)

• Contains a working-storage item that will be passed to a subprogram that creates a mainframe-style pointer item to point to that working-storage item.

In general, to eliminate potential problems that might occur when trying to be selective about which programs are compiled with this directive, we strongly recommended that you use the AMODE directive on all programs in the application, and that you use the same format for each subprogram.

The possible values of format are:

The AMODE directive cannot be used for programs that need to call a Micro Focus call-by-name or call-by-number routine or an external API that expects pointer items to be in the native machine format.

The directives RDEFPTR and CONVERTPTR must not be used with the AMODE directive.

ANALYZE

Makes the Compiler produce extra information so that you can use the Analyzer feature of Advanced Animator and Animator V2.

Syntax:

 >>-.---.-.----.--ANALYZE-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

ANALYZE sets ANIM immediately.

Comments:

This directive is reserved for use with products containing the Analyzer facility, such as Workbench and Object COBOL for UNIX. Do not change its setting unless you have an appropriate COBOL system.

ANIM

Makes the Compiler produce extra information so that you can debug your program using use Animator or Animator V2, or Xilerator with the 16-bit COBOL system.

Syntax:

 >>-.---.-.----.--ANIM----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to ANIM immediately by WB.
Set to ANIM at end by ANALYZE, BROWSE, CSI, GANIM or STRUCT.
Set to NOANIM at end by NOINT or RNIM.

Comments:

An intermediate code file is created, during the syntax check phase in the format required to enable the program to be animated.

In the generate phase, ANIM is treated the same as GANIM.

The intermediate code is held in a file with the extension .int (unless the OPT"0" directive is specified). This can be animated directly.

If you specify the OPT"0" directive, the intermediate code is held in a file with the extension .obj. This file has to be linked before it can be animated.

In all cases, a file with extension .idy, is also created. This contains the additional information needed to animate the program.

With the 16-bit COBOL system, the .idy file also contains information for debugging native code using Xilerator.

With the 32-bit COBOL system on Windows NT and OS/2 V2, an .obj file containing native code, and a file with extension .gdy are also created. The .gdy file contains information needed to animate native code. You must link the .obj file to create an .exe or a .dll file before you can debug native code.

The location of the .idy file is controlled by the COBIDY directive.

See also:

COBIDY Compiler directive
DEFFILE Compiler directive
GANIM Compiler directive
LINKLIB Compiler directive
OPT Compiler directive

ANS85

Specifies that words reserved in the ANSI'85 COBOL Standard are to be treated as reserved words, and changes the behavior of certain features to be compatible with that Standard.

Syntax:

 >>-.---.-.-------ANS85--"SYNTAX"--.--------><  
    .-/-+ ..----.-ANS85------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

Set to ANS85 at end by VSC2 or SAA.

Comments:

This directive causes the following behavior changes:

• The I/O status values are those defined in ANSI'85.

• The ALPHABETIC class test regards lower case as alphabetic.

• Control variables in PERFORM operations are initialized in the order specified by ANSI'85.

• Move operations to a variable-length group item use the maximum length of the target if the DEPENDING ON item is in the group.

APOST

Makes the Compiler interpret the figurative constant QUOTE as the single-quote character (').

Syntax:

 >>-------APOST-----------------------------><

Parameters:

None

Properties:

Comments:

The opposite of this directive is the directive QUOTE which causes the double-quote character (") to be used.

See also:

QUOTE Compiler directive

AREACHECK

Causes the Compiler to treat any token which starts in area A in the Procedure Division as a paragraph or section label, regardless of the preceding tokens.

Syntax:

 >>-.---.-.----.--AREACHECK-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

If AREACHECK is not specified, only tokens which follow a period are treated as possible labels. This directive provides closer compatibility with mainframe error handling, where omitting a period before the label produces a less serious message. It is recommended that such erroneous source code is corrected.

ARITHMETIC

Specifies how arithmetic expressions are to be evaluated.

Syntax:

 >>-.---.--ARITHMETIC--"arith-type"---------><  
    .-/-+

Parameters:

Properties:

Comments:

The possible values of arith-type are:

With ARITHMETIC"MF" specified, expressions are calculated as accurately as possible because no truncation takes place.

ASM

On the 16-bit COBOL system, controls the presence of an assembly listing in the report-file produced by the generate-phase.

On the 32-bit COBOL system for OS/2 and Windows, controls the presence of the report-file (.grp) produced by the generate-phase.

Syntax:

On the 16-bit COBOL system:

 >>-.---.-.----.--ASM-----------------------><  
    .-/-+ .-NO-+

On 32-bit COBOL systems for OS/2 and Windows:

 >>-.---.-.-------ASM-.-"destination"..-----><  
    .-/-+ |           .-()-----------+| 
          ..----.-ASM-----------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

On the 16-bit COBOL system:

NOASM sets NOMASM at end.
Set to ASM immediately by ASMLIST.

Comments:

16-bit:
On the 16-bit COBOL system, specifying ASM causes the Compiler to insert disassembled generated code into the Generator report (.grp) file. The presence of the file is controlled by the ASMLIST directive. You can produce a .grp file which contains Generator messages without assembler code by specifying ASMLIST() NOASM (in that order).

32-bit on OS/2 and Windows NT:
On 32-bit COBOL systems for OS/2 and Windows NT, the report file always contains an assembly listing. When NOASM is specified, no listing file is produced. If you specify ASM with no filename, the assembly listing is sent to the screen. If you specify an existing file it is overwritten. ASM() causes the assembly listing to be put in the file source-name.grp where source-name is the root of the name of the program being compiled. If you specify ASM ASMLIST, the Compiler creates a listing file containing a hex dump of the generated code as well as assembler mnemonics. This is required by some Assemblers.

See also:

ASMLIST Compiler directive

ASMLIST

On the 16-bit COBOL system, controls the presence of the report-file (.grp) created during the generate-phase.

On 32-bit COBOL systems for Windows and OS/2, controls the presence of a hex dump in the report-file created during the generate-phase.

Syntax:

On the 16-bit COBOL system:

 >>-.---.-.-------ASMLIST-.-"destination"..-><  
    .-/-+ |               .-()-----------+| 
          ..----.-ASMLIST-----------------+ 
           .-NO-+

On 32-bit COBOL systems:

 >>-.---.-.----.--ASMLIST-------------------><  
    .-/-+ .-NO-+

Parameters:

Properties:

Dependencies:

On the 16-bit COBOL system:

ASMLIST sets ASM immediately.
NOASMLIST sets NOREF at end.

Comments:

16-bit:
On the 16-bit COBOL system, when NOASMLIST is specified, no listing file is produced. If you specify ASMLIST with no filename, the assembly listing is sent to the screen. If you specify an existing file it is overwritten. ASMLIST() causes the assembly listing to be put in the file source-name.grp where source-name is the root of the name of the program being compiled. Specifying ASMLIST() NOASM produces a report file containing no assmbly listing, just error messages.

On the 16-bit COBOL system, if you use ASMLIST together with SOURCEASM and you follow ASMLIST by the word NOASM, you get an assembly listing showing source code but not assembly code. This is useful because of the "BADCODE" comments you get on the source lines.

32-bit on OS/2 and Windows NT:
On 32-bit COBOL systems for OS/2 and Windows NT, the presence of the listing file is controlled by the ASM directive.

See also:

ASM Compiler directive
MASM Compiler directive
PARAS Compiler directive
SOURCEASM Compiler directive for details of "BADCODE".

ASSIGN

Specifies how to assign a filename when neither EXTERNAL nor DYNAMIC appear in the SELECT statement.

Syntax:

 >>-.---.--ASSIGN--"assign-type"------------><  
    .-/-+

Parameters:

Properties:

Comments:

For more details, see your Programmer's Guide to File Handling.

Under DOS, Windows and OS/2 (16-bit), for any indexed files or for all files if the program is compiled using the CALLFH directive, filename mapping overrides any syntax definition or use of the ASSIGN directive.

ASSIGN-PRINTER

Specifies how to assign the output from an ASSIGN TO PRINTER clause when the clause does not specify a filename.

Syntax:

 >>-.---.-.-----ASSIGN-PRINTER.-"filename"-..-><  
    .-/-+ |                   .-()---------+| 
          .-NO--ASSIGN-PRINTER--------------+

Parameters:

Properties:

Comments:

This directive has no effect if you specify a filename as part of the ASSIGN TO PRINTER clause.

On DOS, Windows and OS/2, NOASSIGN-PRINTER causes the output to be sent to the device LST:. On UNIX it causes the output to be placed in a file called LPT1.

ASSIGN-PRINTER"filename" causes the output to be directed to the filename specified. The filename can be fully specified, including a path-name, base-name, and extension.

ASSIGN-PRINTER() results in the same behavior as including the following COBOL statement:

select filename-1 assign to printer filename-1

That is, the filename used in your COBOL program is also used as the filename for your output. If the internal filename is too long for your operating system to handle, it is truncated to the maximum length the operating system allows.

By default, the filename does not include an extension, but you can specify an extension by using the PRINT-EXT directive.

See also:

PRINT-EXT Compiler directive

ASSUME

Controls the interaction between certain Generator directives.

Syntax:

 >>-.---.-.----.--ASSUME--------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

If ASSUME then the following dependencies apply:
NOTRICKLE sets NOALTER at end.
REGPARM"OUT" sets LITLINK at end.
REGPARM"IN" sets NOFASTLINK at end.

AUTOLOCK

Makes the default locking AUTOMATIC rather than EXCLUSIVE for files opened I-O or EXTEND in a multi-user environment.

Syntax:

 >>-.---.-.----.--AUTOLOCK------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This directive does not appear in the SETTING list if its state is the same as WRITELOCK. In this case, the state of the two directives is indicated by the FILESHARE directive.

This directive is included for compatibility with earlier file-sharing products. When writing new programs you should use the locking syntax rather than this directive.

BADSIGNS

Allows arithmetic with COMP-3 items with illegal sign nibbles (such as those produced by a redefinition of the item) to produce the desired result.

With 32-bit COBOL systems, this directive has been replaced by the HOSTSIGNS directive.

Syntax:

 >>-.---.-.----.--BADSIGNS------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

Specifying BADSIGNS causes COMP-3 arithmetic to be performed by a nonoptimized route, thus reducing the efficiency of your programs. For smaller, faster code you should not specify BADSIGNS.

BELL

Makes the bell sound at points such as when compilation stops, either because of an error or because it has finished.

Syntax:

 >>-.---.-.----.--BELL----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

BOUND

Specifies that the subscript or index value is to be checked to ensure it is within the limits defined by the OCCURS clause.

Syntax:

 >>-.---.-.----.--BOUND---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

BOUND sets NOTABLESEGCROSS and NOBOUNDOPT at end.

Comments:

For multi-dimensional tables, only the composite subscript is checked. If any of the individual subscripts or indices is beyond its limit, but the reference remains within the table, no error is produced.

BOUNDOPT

Optimizes the code generated for USAGE DISPLAY subscripts.

Syntax:

 >>-.---.-.----.--BOUNDOPT------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to NOBOUNDOPT at end by BOUND or TABLESEGCROSS.

Comments:

If BOUNDOPT is used, any digits in a USAGE DISPLAY subscript above the size of the table are ignored.

Can only be used when NOBOUND is specified, and only affects generated code created with OPT"1" or above. NOBOUNDOPT must be specified if a program references beyond the end of a table.

Example:

For example, for a table with 50 entries, a PIC 9(3) subscript is treated as PIC 9(2), with the most significant digit ignored.

BRIEF

Makes the Compiler produce only error numbers and no message texts.

Syntax:

 >>-.---.-.----.--BRIEF---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

BROWSE

Makes the Compiler create an .sbr file for use with the Microsoft Source Browser supplied with the Microsoft utilities. Requires the Microsoft components to be present at compile time.

Syntax:

 >>-.---.-.----.--BROWSE--------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

BROWSE sets ANIM immediately.

BWZSTAR

Determines whether the BLANK WHEN ZERO clause is allowed in the Data Division for those fields defined with the "*" PICTURE symbol.

Syntax:

 >>-.---.-.----.--BWZSTAR-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

The BWZSTAR directive is Early User Syntax support. You must set the EARLY-RELEASE directive to enable this feature. This directive might change or be removed in a later revision of this COBOL system.

If BWZSTAR is specified, the BLANK WHEN ZERO clause is allowed with fields defined using the "*" PICTURE symbol, and BLANK WHEN ZERO is effective when the field is the target of an editing operation and the result is zero. If NOBWZSTAR is specified, a BLANK WHEN ZERO clause associated with a PIC * field is rejected.

See also:

EARLY-RELEASE Compiler directive

BYTE-MODE-MOVE

Controls the granule size for moves between overlapping data items.

Syntax:

 >>-.---.-.----.--BYTE-MODE-MOVE------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

The Compiler also accepts the directive-name without hyphens; that is BYTEMODEMOVE.

This directive is provided for compatibility with IBM mainframe compilers in the handling of forward overlapping moves - that is, where the start of the target data item overlaps the end of the source data item.

If BYTE-MODE-MOVE is specified, data is moved one byte at a time from the source to the target.

If NOBYTE-MODE-MOVE is specified, the data is moved in granules of two, four or more bytes at a time (depending on environment) from the source to the target. Consequently, if the overlap is less than the size of the granule, each granule moved overwrites part of the next granule to be moved.

CALLADIS

Causes all format 1 ACCEPT, DISPLAY and EXHIBIT statements to be routed through the specified handler. The output generated by TRACE is also routed through the specified handler.

Syntax:

 >>-.---.-.-------CALLADIS--"handler-name"-.><  
    .-/-+ ..----.-CALLADIS-----------------+ 
           .-NO-+

Parameters:

Properties:

Comments:

The CALLADIS directive is Early User Syntax support. You must set the EARLY-RELEASE directive to enable this feature. This directive might change or be removed in a later revision of this product.

If handler-name is not specified, EXTADIS is assumed.

See also:

EARLY-RELEASE Compiler directive

CALLFH

Makes the Compiler generate direct calls for all file I/O operations, using the Callable File Handler interface.

Syntax:

 >>-.---.-.-------CALLFH--"handler-name"-.--><  
    .-/-+ ..----.-CALLFH-----------------+ 
           .-NO-+

Parameters:

Properties:

Comments:

handler-name can be one of the following:

If NOCALLFH is specified, file I/O is handled by the run-time system. When CALLFH is specified, all file I/O statements are converted to calls to the file handler specified by the directive.

If handler-name is not specified, EXTFH is assumed.

This directive is also used to direct all file handling calls through converter modules, such as XFH2BTR, to make use of non-COBOL file handlers.

Use this directive to make your programs call your file handler instead of the standard file handler.

CALLMAP

Causes all call operations to be converted to calls to the specified call handler program.

Syntax:

 >>-.---.-.----.--CALLMAP--"handler-name"---><  
    .-/-+ .-NO-+

Parameters:

Properties:

Comments:

The CALLMAP directive is Early User Syntax support. You must set the EARLY-RELEASE directive to enable this feature. This directive might change or be removed in a later revision of this COBOL system.

See also:

EARLY-RELEASE Compiler directive

CALLMCS

This directive is reserved for internal use by this COBOL system. Because it might appear in the list of settings, it is included for completeness. It is not intended for your use, and its setting should not be changed.

Causes all Message Control System operations to be routed through the specified handler.

Syntax:

 >>-.---.-.-------CALLMCS--"mcs-name"-.-----><  
    .-/-+ ..----.-CALLMCS-------------+ 
           .-NO-+

Parameters:

Properties:

Comments:

If handler-name is not specified, EXTMCS is assumed.

CALLSORT

Defines the program to be called to handle all SORT and MERGE operations.

Syntax:

>>-.---.-.-------CALLSORT--"sort-name"-.---><  
   .-/-+ ..----.-CALLSORT--------------+ 
          .-NO-+

Parameters:

Properties:

CANCELLBR

Makes the Compiler close an .lbr file used as a library in a COPY statement once that copy operation is complete.

Syntax:

 >>-.---.-.----.--CANCELLBR-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

NOCANCELLBR causes such .lbr files to be left open until the end of the compilation. Subsequent COPY statements that do not specify a library search for the copyfile in all open .lbr libraries (last one opened is searched first) before the current directory.

This directive is only relevant if COPYLBR is on.

[NO]CANCELLBR can be used on any $SET statement. Only .lbr libraries referenced in COPY statements after the $SET statement are affected. Any left open from previous COPY statements are only closed if specified again on a COPY statement following a CANCELLBR directive.

CASE

Prevents external symbols (such as Program-ID and names of called programs) being converted to upper case.

Syntax:

 >>-.---.-.----.--CASE----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This is useful if you are calling a program written in C, where names are case dependent.

CHANGE-MESSAGE

Changes the severity of errors/messages. Messages can also be supressed or returned to their original severity.

Syntax:

                               +-------------™ 
                               v             | 
 >>-.---.-.-----CHANGE-MESSAGE--"error svrty"-.->< 
    .-/-+ .-NO--CHANGE-MESSAGE----------------+

Parameters:

Properties:

Dependencies:

CHANGE-MESSAGE overrides the general settings of directives such as FLAGAS.

Comments:

CHANGE-MESSAGE can take as many parameter pairs as required. If CHANGE-MESSAGE is specified multiple times the effects are cumulative.

CHANGE-MESSAGE replaces the HIDE-MESSAGE Compiler directive. However, if you use HIDE-MESSAGE in conjunction with CHANGE-MESSAGE the effects are also cumulative.

Serious errors cannot have their severity reduced unless they were changed to serious (from a lower severity) by another CHANGEMESSAGE directive or the FLAGAS or FLAGCD directives.

Examples:

CHANGE-MESSAGE"ALL R" 
CHANGEMESSAGE"10 S 135 E 100 N"

See also:

FLAGAS Compiler directive
FLAGCD Compiler directive
HIDE-MESSAGE Compiler directive

CHARSET

Defines the character set of the environment.

Syntax:

 >>-.---.--CHARSET--"char-set"--------------><  
    .-/-+

Parameters:

Properties:

Dependencies:

CHARSET"ASCII" sets DEFAULTBYTE"32", SIGN"ASCII" and NATIVE"ASCII" immediately.
CHARSET"EBCDIC" sets DEFAULTBYTE"0" immediately and SIGN"EBCDIC" and NATIVE"EBCDIC" at end.

Comments:

DOS, WINDOWS and OS/2:
On DOS, Windows and OS/2, to use this directive you must have the full Micro Focus COBOL Workbench system.

16-bit:
With the 16-bit COBOL system, do not change its setting if your application is to be linked with the static-linked run-time system.

All literals and collating sequences are handled in the character set specified.

For current limitations see you Programmer's Guide to Writing Programs.

With CHARSET"EBCDIC" set, COBOL system library routines that receive or return alpahnumeric data in any parameters do not work. The alphanumeric data must be in ASCII.

CHECKDIV

Controls the behavior of your program if it tries to divide by zero in a statement that has no ON SIZE ERROR phrase.

Syntax:

 >>-.---.-.-------CHECKDIV--"dialect"-.-----><  
    .-/-+ ..----.-CHECKDIV------------+ 
           .-NO-+

Parameters:

Properties:

Comments:

With CHECKDIV or CHECKDIV"ANSI" specified, the program continues with an undefined result if it tries to divide by zero. With NOCHECKDIV set, the behavior is undefined. Setting NOCHECKDIV results in optimal code for divides.

Specifying CHECKDIV"OSVS", CHECKDIV"VSC2", or CHECKDIV"COBOL370" has the same effect; trying to divide by zero produces run-time error 048 (Attempt to divide by zero). This error can be disabled using the -O RTS switch.

This directive has no effect on arithmetic statements that use the ON SIZE ERROR phrase.

See also:

O RTS switch

CHECKNUM

Checks whether numeric fields contain valid data before performing operations on them.

Syntax:

 >>-.---.-.----.--CHECKNUM------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

Specifying this directive causes extra code to be generated to ensure that numeric fields contain valid data before any operations are performed on them. With CHECKNUM set, run-time error message 163 ("Illegal character in numeric field") is displayed if any numeric field is found to contain nonnumeric data. This error can be disabled using the -F RTS switch.

Specifying CHECKNUM causes extra code to be produced, reducing the efficiency of your programs. For smaller, faster code you should not specify CHECKNUM.

See also:

F RTS switch

CHECKSTACK

Adds code to all calls to check the stack and report if an incorrect number of parameters has been passed.

Syntax:

 >>-.---.-.----.--CHECKSTACK----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

Extra code is added for all calls, including calls to the system library routines (CBL_).

If an incorrect number of parameters has been used, the code tries to detect the resulting stack corruption and give run-time system error 168 ("Stack Overflow") at run-time.

The extra instructions slightly reduce run-time performance, so the directive should only be used when debugging.

CHIP

Tells the Compiler that a specific microprocessor architecture is in use.

Syntax:

 >>-.---.-.-----CHIP--"integer"-.-----------><  
    .-/-+ .-NO--CHIP------------+

Parameters:

Properties:

Comments:

The information produced is only relevant for 16-bit environments.

In the case of the Intel 80x86 family, the use of CHIP"16" indicates that the maximum segment addressability is given by 16 bits, and anything over this size is split over an addressing boundary. Knowing this, the Compiler can alter certain code structures to avoid potential crossing boundaries conditions (for example, the internal print buffer used with Report Writer syntax).

When used with the directive FLAG-CHIP, this directive causes the Compiler to indicate chip specific boundary conditions. These might indicate potential problems or performance losses on the particular chip.

CICS

Enables the use of CICS by updating BLL cells.

Syntax:

 >>-.---.-.----.--CICS----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This directive is reserved for use with add-on CICS products supplied by Micro Focus. Do not change its setting unless you have an appropriate add-on product. This is not for use with CICS OS/2.

With CICS, if CICS-CPY is also set, the Compiler inserts:

     COPY "CICS.CPY"

at the beginning of your Linkage Section. If your program has no Linkage Section, the Compiler inserts one as follows:

 LINKAGE SECTION. 
 COPY "CICS.CPY". 
 01 DFHCOMMAREA  PIC X(n).

where the parameters are:

The Linkage Section item following the DFHCOMMAREA record is assumed to contain the BLL cells, each named and with the description PIC S9(8) COMP. The BLL cells are matched with the corresponding Linkage Section records according to CICS rules.

If the Compiler inserts the DFHCOMMAREA record, it is not visible during animation or in printer listings. However, the record can be queried during animation. If you intend to animate a program that uses BLL cells, set this directive on.

CICS-CPY

Inserts the statement COPY "CICS.CPY" provided the CICS directive is specified.

Syntax:

 >>-.---.-.----.--CICS-CPY------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This directive is reserved for use with add-on CICS products supplied by Micro Focus. Do not change its setting unless you have an appropriate add-on product. This is not for use with CICS OS/2.

See also:

CICS Compiler directive

CICSOPTIMIZE

Optimizes the handling of CICS BLL cells.

Syntax:

 >>-.---.-.----.-.-CICSOPT------.-----------><  
    .-/-+ .-NO-+ .-CICSOPTIMISE-+
                 .-CICSOPTIMIZE-+

Parameters:

None

Properties:

Comments:

This directive is reserved for use with add-on CICS products supplied by Micro Focus. Do not change its setting unless you have an appropriate add-on product. This is not for use with CICS OS/2.

With NOCICSOPTIMIZE, the Compiler inserts instructions whenever a BLL-cell is updated to ensure that the addressability of Linkage Section items corresponds to the current values in the BLL cells. With CICSOPTIMIZE, the Compiler does not insert these instructions. You must use the SERVICE RELOAD instruction to ensure that the addressability is updated.

CMPR2

For compatibility with the mainframe compiler option of the same name which returns behavior of the COBOL 370, VS COBOL II version 3 and VS COBOL II version 4 compilers to that of the VS COBOL II version 2 compiler.

Syntax:

 >>-.---.-.----.--CMPR2---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Requires VSC2"3", VSC2"4" or COBOL370 flagging to be selected.

Comments:

A conflicting directives message is issued (with FLAGCD only) if one of the required flagging levels is not selected.

Setting this directive is not quite the same as setting VSC2"2". It emulates the mainframe behavior quite closely, and also causes different run-time behavior in some cases. If you use this directive on the mainframe with one of the specified compilers, use this directive instead of VSC2"2".

The FLAGMIG directive flags the items that give different run-time behavior.

See also:

FLAGMIG Compiler directive

COBFSTATCONV

Causes the Callable File Handler ExtFH to use the user-supplied module specified in the COBFSTATCONV environment variable to convert the file status codes if an I/O error is encountered on a file.

Syntax:

 >>-.---.-.----.--COBFSTATCONV--------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

COBIDY

Specifies the path where the Animator information file (.idy file) and COBOL Source Information file (.csi) are to be written.

Syntax:

 >>-.---.-.-------COBIDY--"pathname"-.-----><  
    .-/-+ ..----.-COBIDY--------------+ 
           .-NO-+

Parameters:

Properties:

Comments:

If you do not specify pathname, the pathname in the COBIDY environment variable is used; if the environment variable contains multiple path-names the first of them is used. With NOCOBIDY, the .idy and .csi files are written to the same path as the object program.

The .csi files are only produced if the CSI directive is used. This should only be used if you have Object COBOL for UNIX or Workbench.

See also:

CSI Compiler directive

COBOL370

Specifies that words reserved in IBM COBOL/370 are to be treated as reserved words, and allows features selectively for compatibility with a given level of that product.

Syntax:

 >>-.---.-.-------COBOL370--"integer"-.-----><  
    .-/-+ ..----.-COBOL370------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

COBOL370 sets DBSPACE and DBCS"3" immediately.

Comments:

Specifying COBOL370 without a parameter is the same as specifying COBOL370"1". This specifies compatibility with Version 1 Release 1 of IBM COBOL/370.

Setting integer to:

Specifying COBOL370 without a parameter is the same as specifying COBOL370"2".

COBOLDIR

Specifies whether the Compiler should process the directives in a cobol.dir file or ignore it.

Syntax:

 >>-.---.-.----.--COBOLDIR------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

COMP

Makes the Compiler produce very compact and efficient code for some statements involving COMP data items, by treating COMP items as COMP-X.

Syntax:

 >>-.---.-.----.--COMP----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This COMP code produced by the Compiler behaves in a way that is not the ANSI standard in cases of numeric overflow. You should use this directive only if you know that your program does not lead to numeric overflow, or if you want to use the defined (but nonstandard) behavior on overflow.

COMP-5

Specifies whether the sign is to be dropped when a value is stored in an unsigned COMP-5 data item.

Syntax:

 >>-.---.--COMP-5--"integer"----------------><  
    .-/-+

Parameters:

Properties:

Comments:

The possible values of integer are:

COMP-6

Specifies whether COMP-6 data is to be held in binary or packed decimal format.

Syntax:

 >>-.---.--COMP-6--"integer"----------------><  
    .-/-+

Parameters:

Properties:

Dependencies:

RM or MF must be set.
RM and RM"ANSI" set COMP-6"1".
NORM sets COMP-6"2".

Comments:

The possible values of integer are:

Even if you specify the COMP-6 directive, the reserved word COMP-6 is recognized only if it belongs to the dialect specified by RM or MF"10".

Example:

With COMP-6"2" specified:

PIC 99 COMP-6 VALUE 87 is stored in one byte as x"87"
PIC S99 COMP-6 VALUE 87 is stored in two bytes as x"087C"

COMS85

Allows your program to contain syntax introduced in the Communications Module of the ANSI'85 COBOL Standard.

Syntax:

 >>-.---.-.----.--COMS85--------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

CONFIRM

Makes the Compiler echo all subsequent directives to the screen.

Syntax:

 >>-.---.-.----.--CONFIRM-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to CONFIRM immediately by VERBOSE.

CONSTANT

Declares a constant for use in the program.

Syntax:

 >>-.---.-CONSTANT-const-name-.-(numeric-lit)--.-><  
    .-/-+                     .-"alphanum-lit"-+

Parameters:

Properties:

Comments:

The effect is as if the constant const-name had been defined in the program in a level-78. Its value is that given in the parentheses or quotation marks. If parentheses are used, the constant is of category numeric; if quotation marks are used it is of category alphanumeric.

To declare several figurative constants using this feature, use the CONSTANT directive repeatedly.

CONVERTPTR

Specifies whether POINTER data items can be redefined as PIC 9(9) COMP data items.

Syntax:

 >>-.---.-.----.--CONVERTPTR----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

IBM VS COBOL II and COBOL/370 let you use the REDEFINES clause on a POINTER data item, to redefine it as a PIC 9(9) COMP data item. You can then perform arithmetic operations on this item, giving the program the ability to shift the address referred to by a pointer up or down.

If CONVERTPTR is specified, pointers are held in a form that is reversed with respect to native byte ordering. They are converted to the native form immediately before any statement of the form:

SET ADDRESS OF linkage-item TO pointer

and immediately after any statement of the form:

SET pointer TO ADDRESS OF ...

The CONVERTPTR directive has no effect on PROCEDURE-POINTER data items.

On the 16-bit COBOL system, unexpected behavior might occur if the arithmetic carried out causes the pointer to cross the boundary of a data segment.

If you specify the CONVERTPTR directive, you cannot use a format 4 SET statement in your program.

CONVERTRET

Causes COMP and COMP-X items specified in CALL ... RETURNING and EXIT PROGRAM ... RETURNING phrases to be converted to COMP-5.

Syntax:

 >>-.---.-.----.--CONVERTRET----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

CONVSPACE

Converts double-byte space characters in COBOL source files to single-byte spaces on input.

Syntax:

 >>-.---.-.----.--CONVSPACE-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

CONVSPACE only has an effect when the DBSPACE Compiler directive is off.

COPYEXT

Specifies, in part, the filename extension of the copyfile that the Compiler is to look for if a filename in a COPY statement is specified without an extension.

Syntax:

                              +-------------. 
                              v             | 
 >>-.---.--COPYEXT-"extension-.-------------.-"--><  
    .-/-+                     .-,-extension-+

Parameters:

Properties:

Comments:

Up to eight extensions can be specified, each extension being up to 10 characters long. A null extension is used to represent an extension of spaces; for example:

COPYEXT"src,,txt"

would first look for files with the extension .src, then for files with no extension, and then for files with the extension .txt.

If the filename specified in a COPY statement has no extension or trailing period then a list of possible extensions are tried in turn until a file is sucessfully found or the list is exhausted (and an error reported.) The list is derived from a combination of the values of the COPYEXT and OSEXT directives. The first extension in the list is either the value in the OSEXT directive or the first value in the COPYEXT directive depending on which of the two directives was last specified. The remainder of the list is the remainder of the values in the COPYEXT directive (values 2 through to the end). For example:

COPYEXT"cpy,cbl,txt"  OSEXT"src"

would have the same effect as:

COPYEXT"src,cbl,txt"

If you have many COPY statements that do not specify an extension, using COPYEXT can improve the compilation speed of your program. For example, if all of your copyfiles have the extension .cpy, specifying COPYEXT"CPY,CBL" would avoid unnecessary file access attempts.

See also:

OSEXT Compiler directive

COPYLBR

Makes the Compiler treat the library specified in a COPY statement as an .lbr file.

Syntax:

 >>-.---.-.----.--COPYLBR-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

With NOCOPYLBR, the Compiler assumes that the library is a path-name.

COPYLIST

Makes the Compiler list the contents of files named in COPY statements.

Syntax:

 >>-.---.-.----.--COPYLIST-.-----------.----><  
    .-/-+ .-NO-+           .-"integer"-+

Parameters:

Properties:

Comments:

integer is the number of a COBOL segment. It must be 0 or in the range 50 through 99. If it is not specified, the contents of all copyfiles are listed. If it is specified, the contents of all copyfiles in the first three divisions (that is, the Identification, Environment and Data Divisions), the root, and the given segment are listed. An integer of 0 refers to the first three divisions and all root segments.

NOCOPYLIST prevents the listing of the contents of any copyfiles. If a segment-number is specified with NOCOPYLIST, only copyfiles in that segment are listed. For example:

Whatever the state of this directive, the name of any copyfile open when a page heading is output is given in that heading.

CSI

Makes the Compiler produce extra information so that you can use CSI.

Syntax:

 >>-.---.-.----.--CSI-----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

CSI sets ANIM at end.

Comments:

This directive is reserved for use with products containing CSI, such as Workbench and Object COBOL for UNIX. Do not change its setting unless you have an appropriate add-on product.

The CSI file has extension .csi. Its location is controlled by the COBIDY directive.

See also:

COBIDY Compiler directive

CURRENCY-SIGN

Specifies the currency sign to be recognized in the PICTURE clause.

Syntax:

 >>-.---.--CURRENCY-SIGN--"integer"---------><  
    .-/-+

Parameters:

Properties:

Comments:

You cannot specify a valid PICTURE clause symbol. See your Language Reference for a list of these.

CURRENT-DATE

Specifies the format of the date stored in the CURRENT-DATE special register.

Syntax:

 >>-.---.--CURRENT-DATE--"date-format"------><  
    .-/-+

Parameters:

Properties:

Comments:

The DDMMYY parameter causes CURRENT-DATE to be stored in European format. The parameter can be specified in either upper case or lower case.

DATACOMPRESS

Specifies the type of data compression to be done on sequential and indexed files.

Syntax:

 >>-.---.-.-----DATACOMPRESS--"integer"-.---><  
    .-/-+ .-NO--DATACOMPRESS------------+

Parameters:

Properties:

Comments:

The only values that can be specified for data compression supported by this system are 1 and 3. Values in the range 128 through 255 indicate user-defined compression routines.

You need to specify data compression only when creating the file. Subsequently, the data compression is detected when the file is opened.

To get data compression on an individual file, use $SET statements in your source so that this directive is in effect only for the part of the source containing the file's SELECT statement.

Data compression is supported only by the Callable File Handler, ExtFH. All indexed files are processed by ExtFH. However, if you want to use data compression with sequential files, every program referencing those files must be compiled with the directive CALLFH"EXTFH".

See also:

CALLFH Compiler directive

DATALIT

Makes the Compiler put literals in the DATA segment of the program rather than the CODE segment.

Syntax:

 >>-.---.-.----.--DATALIT-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to NODATALIT at end by OPT"0".

Comments:

For segmented programs, specifying NODATALIT gives the advantage of being able to have literals in the CODE segment so that they can then be swapped out with the code that references them. There is generally no advantage in specifying NODATALIT for a nonsegmented program. However, if you are running generated code (.gnt files) under DOS, specifying NODATALIT lets you discard any unreferenced code segment.

DATE

Puts the date in the Date-Compiled paragraph and at the top of each page of the listing.

Syntax:

 >>-.---.-.-------DATE--"string"-.----------><  
    .-/-+ ..----.-DATE-----------+ 
           .-NO-+

Parameters:

Properties:

Comments:

The date and time, available from the operating system, are automatically inserted when you specify DATE. You can, however, enter the date yourself as the parameter. With NODATE, the paragraph is left unaltered.

With DATE, the operating system date or the string you enter appears at the top of each page of the listing. With NODATE, spaces are used instead.

DBCHECK

Makes the Compiler check that any Double-byte Character Set (DBCS) literals only contain valid 16-bit DBCS characters.

Syntax:

 >>-.---.-.----.--DBCHECK-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

If you specify DBCHECK in environments that support validation of DBCS data, any literal that does not contain only valid 16-bit DBCS characters results in syntax error 1048 (DBCS literal includes invalid data).

DBCS

Makes the Compiler accept characters of the Double-byte Character Set (DBCS) for use in ideographic languages such as Japanese, Chinese and Korean.

Syntax:

 >>-.---.-.-------DBCS--"integer"-.---------><  
    .-/-+ ..----.-DBCS------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

Set to DBCS"2" immediately by SAA, VSC2"2", VSC2"3" or MF"7".
Set to DBCS"3" immediately by COBOL370 or MF"integer" where integer > 7.

Comments:

The possible values of integer are:

DBCSSOSI

Defines the two characters used as the shift-out and shift-in delimiters in DBCS literals.

Syntax:

 >>-.---.-.----DBCSSOSI-"integer-1"-"integer-2"-.><  
    .-/-+ .-NO-DBCSSOSI-------------------------+

Parameters:

Properties:

Comments:

When shift-out and shift-in characters are specified by this directive, every DBCS literal must have the shift-out character immediately after the opening quotation mark and the shift-in character immediately before the closing quotation mark.

They act as additional delimiters to the literal, and are not part of its value. With NODBCSSOSI, no shift-out and shift-in characters are needed or recognized.

DBSPACE

Makes the Compiler interpret the figurative constant SPACE, when used as a DBCS figurative constant, as the double-byte space character supplied by the system.

Syntax:

 >>-.---.-.----.--DBSPACE-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

DBSPACE set immediately by COBOL370, NCHAR"2", SAA, VSC2"2" or VSC2"3".
DBSPACE set immediately by MF"integer" where integer > 7

Comments:

With DBSPACE, the Compiler uses the system-supplied double-byte space character. NODBSPACE provides compatibility with previous versions of this Compiler, where the double-byte space character was two ASCII space characters (x"2020").

DE-EDIT

Specifies the behavior of de-editing moves from numeric-edited items to other numeric-edited items or to numeric items.

Syntax:

 >>-.---.--DE-EDIT--"integer"---------------><  
    .-/-+

Parameters:

Properties:

Comments:

The possible values of integer are:

Example:

 01 a  pic 909V99 value "30456". 
 01 b  pic 9(5). 
        ... 
     move a to b

With DE-EDIT"1", b contains 30456. With DE-EDIT"2", b contains 00034; the 0 after the 3 is dropped because it corresponds to the insertion character 0 in the picture-string of a, and the .56 is dropped because B has no decimal places.

DEFAULTBYTE

Initializes each otherwise undefined byte of the Data Division to the character given.

Syntax:

 >>-.---.--DEFAULTBYTE--"integer"-----------><  
    .-/-+

Parameters:

Properties:

Dependencies:

Set to DEFAULTBYTE"32" immediately by CHARSET"ASCII".
Set to DEFAULTBYTE"0" immediately by CHARSET"EBCDIC", MS, IBM-MS or PC1.

Comments:

integer is a decimal value. For example, if you want to specify an EBCDIC space use DEFAULTBYTE"64"; to specify an ASCII space use DEFAULTBYTE"32".

DEFAULTCALLS

Specifies the default calling convention.

Syntax:

 >>-.---.-.-------DEFAULTCALLS--"integer"-.-><  
    .-/-+ ..----.-DEFAULTCALLS------------+ 
           .-NO-+

Parameters:

Properties:

Comments:

DEFAULTCALLS without the optional parameter specifies that the calling convention specified in the PROCEDURE DIVISION USING statement is to be used as the default calling convention.

DEFAULTCALLS"integer" specifies that the calling convention indicated by "integer" is to be used as the default calling convention.

NODEFAULTCALLS is equivalent to DEFAULTCALLS"0".

Individual CALL statements can override these defaults (see your Language Reference). See your Programmer's Guide to Writing Programs for a list of available calling conventions.

DEFFILE

Makes the Compiler produce a module definition file (.def file), for use by the Linker.

Syntax:

 >>-.---.-.-------DEFFILE--"filename"--.----><  
    .-/-+ ..----.-DEFFILE--------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

If OMF"GNT", DEFFILE sets OMF"OBJ" at end.

Comments:

This directive only affects .obj files.

The name of this file is included on the command line to the Linker.

If filename is omitted, the name of the file created is obj-name.def, where obj-name is the name of the object file without an extension. filename can be specified as *.ext to set the extension but use obj-name as the base-name.

The module definition file created is used when linking to create an executable file for Windows, Windows NT or OS/2. The exact contents of the file is determined by the settings of the DLL and DEFFILETYPE directives, as follows:

• The DLL directive indicates if the module definition file is to be used to create an .exe or a .dll file.

• The DEFFILETYPE directive indicates if it is for Windows, Windows NT or OS/2.

Example:

DEFFILE"*.dfa" DEFFILETYPE"OS2" DLL

creates a file called obj-name.dfa containing the lines:

LIBRARY INITINSTANCE 
PROTMODE 
DATA NONSHARED 
EXPORTS source-name @1

See also:

DEFFILETYPE Compiler directive
DLL Compiler directive

DEFFILETYPE

Specifies the type of module definition file (.def file) to be produced when the DEFFILE directive is specified.

Syntax:

 >>-.---.--DEFFILETYPE--"type"--------------><  
    .-/-+

Parameters:

Properties:

Comments:

Produces a .def file suitable for linking to create an .exe file or a .dll file for use on the environment specified by type:

On the 16-bit COBOL system, the DLL directive indicates if the module definition file is to be used to create an .exe or a .dll file.

This directive affects only .obj files.

See also:

DEFFILE Compiler directive
DLL Compiler directive

DETECT-LOCK

Makes READ statements detect when a record is locked by another program.

Syntax:

 >>-.---.-.----.--DETECT-LOCK---------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

The Compiler also accepts the name without a hyphen; that is DETECTLOCK

With DETECT-LOCK, if a READ statement reads a record locked by another program, it returns an I/O status of 9/068. With NODETECT-LOCK, it returns 0/000. In both cases it reads the record successfully.

If a READ statement attempts to lock a record (that is, if the statement LOCK MODE IS AUTOMATIC has been specified, or a WITH LOCK phrase is used on the READ statement), and the record is locked by another program, an I/O status of 9/068 is returned. If NODETECT-LOCK is specified, and the READ statement is not attempting to lock a record an I/O status of 0/000 is returned.

DG

Changes the behavior of certain features to be compatible with Data General Interactive COBOL rev 1.30.

Syntax:

 >>-.---.-.----.--DG------------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

See your Language Reference for details of syntax support for Data General Interactive COBOL rev 1.30.

DIR

Makes the Compiler read directives from a file.

Syntax:

 >>-.---.-.-DIRECTIVES-.-"filename"---------><  
    .-/-+ .-DIR--------+

Parameters:

Properties:

Comments:

A directives file is a text file containing directives. Directives are separated by a space or the end of line. A directive cannot be broken across two lines.

You can include a comment line with an ampersand (&) in column 1, followed by a space. If you omit the space the comment is treated as a directive, and the syntax check on that file fails.

The directives are read from the file until the end of file is reached or another DIRECTIVES directive is encountered. The maximum length of a line is 128 characters.

You can specify more than one directives file in a program by specifying several DIRECTIVES"filename". If you specify the DIRECTIVES directive in a directives file, the Compiler switches to the new directives file, reads all the directives in it, returns to the original directives file, and continues to read the directives specified after the DIRECTIVES directive. You can nest directives files to any depth.

The directives file is searched for in the current and COBOL system directories. If no extension is specified, a file extension of .dir is assumed. If no file is found, the search is repeated with no extension.

See also:

USE Compiler directive

DIRECTIVES-IN-COMMENTS

Makes the Compiler process $SET statements held in comment lines.

Syntax:

 >>-.---.-.----.--DIRECTIVES-IN-COMMENTS----><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This directive enables sources to contain directives used when compiling with the Micro Focus Compiler, but ignored (as comment lines) by any other compiler, such as a mainframe compiler.

$SET can appear anywhere on a line, provided it is the first thing on the line. However, if this directive is set, the $SET statement is processed even if the line is a comment line (asterisk in column 7).

DLL

Makes the Compiler produce a module definition file suitable for creating a .dll file instead of an .exe file when you specify the DEFFILE directive.

Syntax:

 >>-.---.-.----.--DLL-----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This directive only affects .obj files.

Example:

DEFFILE DEFFILETYPE"WIN" NODLL

causes the Compiler to produce a module definition file of name source-name.def suitable for creating an .exe file for Windows.

See also:

DEFFILE Compiler directive
DEFFILETYPE Compiler directive

DOSVS

Specifies that words reserved in IBM DOS/VS COBOL are to be treated as reserved words.

Syntax:

 >>-.---.-.----.--DOSVS---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to NODOSVS immediately by OSVS.

See also:

OSVS Compiler directive

DYNAM

Specifies that CANCEL statements are not to be ignored.

Syntax:

 >>-.---.-.----.--DYNAM---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

With NODYNAM, CANCEL statements in the program are ignored.

EANIM

Makes the Compiler put information about line numbers and symbols into the .obj file for use by a run-time debugger such as Microsoft CodeView.

Syntax:

 >>-.---.-.-------EANIM--"integer"----.-----><  
    .-/-+ ..----.-EANIM---------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

If OMF"GNT", EANIM sets OMF"OBJ" at end.
If OPT"0", EANIM sets OPT"1" at end.
EANIM sets ANIM immediately.

Comments:

integer can be one of two values:

Specifying EANIM is the same as specifying EANIM"2".

For CodeView V3.00 and V3.1x, you can use either EANIM or EANIM"1". For CodeView V4.xx you must use EANIM"1", and you must use the Linker that came with your copy of CodeView.

This directive affects only .obj files.

If you subsequently link the program using the options /CO, the debugging information is copied into the .exe file.

EANIM causes FILLER data items to be placed in the dictionary, as is required by CodeView V3.50.

EARLY-RELEASE

Enables support for Early User Syntax.

Syntax:

 >>-.---.-.----.--EARLY-RELEASE-------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

If NOEARLY-RELEASE and MF are set, MF defaults to MF"10".
If EARLY-RELEASE and MF are set, MF defaults to MF"11".

Comments:

You must specify this directive if you want to use any of the early-release directives.

ECHO

Makes the Compiler display error lines and error messages on the screen.

Syntax:

 >>-.---.-.----.--ECHO----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to NOECHO immediately by ECHOALL.

Comments:

For each error, the source line is displayed together with an error number and (unless BRIEF is set) an explanatory message.

See also:

BRIEF Compiler directive

ECHOALL

Sends a full listing to the screen as well as to a printer or other device specified with the LIST or PRINT directive.

Syntax:

 >>-.---.-.----.--ECHOALL-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

ECHOALL sets NOECHO immediately.

EDITOR

Makes the Compiler send error messages to a file in a format compatible with a specified editor.

Syntax:

 >>-.---.-.-----EDITOR--"editor-id"-.-------><  
    .-/-+ .-NO--EDITOR--------------+

Parameters:

Properties:

Dependencies:

EDITOR"MS" sets NOENSUITE immediately.
EDITOR"MF" sets ENSUITE"1" immediately.

Comments:

The possible values of editor-id are:

You are recommended to use the NOECHO and NOQUERY directives in addition to the EDITOR directive.

See also:

ECHO Compiler directive
FLAGSINEDIT Compiler directive
QUERY Compiler directive

ENSUITE

Reserved for internal use only.

ERRLIST

Specifies that the listing is to contain no source lines except those that have errors or flags.

Syntax:

 >>--.---.--.----.---ERRLIST---"option"--------><
     .-/-+  .-NO-+ 

Parameters:

Properties:

Dependencies:

ERRLIST"END"is switched to ERRLIST"VERBOSE" if FLAGQ or ERRQ are set.

Comments:

The possible values of option are:

ERRQ

Makes the Compiler ask, each time it gives an error message, whether you want to stop compiling.

Syntax:

 >>-.---.-.----.--ERRQ----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

EXPANDDATA

Makes the Compiler store the program data in uncompressed form.

Syntax:

 >>-.---.-.----.--EXPANDDATA----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This directive only affects .obj files.

Normally the COBOL system stores the program data in compressed form and expands it when the program is loaded. Specifying EXPANDDATA when compiling to an .obj file causes the data to be stored in its expanded form.

This saves a small amount of time at initial program entry, and on DOS it gives a smaller .exe file because information to guide the expansion can be omitted; DOS always allocates the space for the expanded form in any case. On OS/2 it gives a slightly bigger .exe file.

When data is stored in this way, it is only initialized once, when the program is physically loaded. Hence the IS INITIAL phrase in the Program-Id paragraph does not work. Also, a program which is canceled and called again is only reinitialized if it is a separate module, such that this cancel/call cycle causes a physical reload from disk. The directive is ignored when compiling to a.gnt file.

EXTINDEX

Specifies that index-names defined in a table in an EXTERNAL record are to be treated as EXTERNAL.

This directive is provided for compatibility with earlier releases of this COBOL system. It should not be used with new programs, and existing programs should be recoded to make the use of this directive unnecessary. It will be removed in a future release.

Syntax:

 >>-.---.-.----.--EXTINDEX------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

With EXTINDEX, behavior is as in earlier versions of this Compiler. With NOEXTINDEX, such items are not regarded as EXTERNAL, thus conforming to the ANSI'85 standard.

FASTLINK

Tells the Compiler that the parameters in the USING clauses of the Procedure Division statement and each ENTRY statement conform to certain restrictions. This enables it to produce faster code.

Syntax:

 >>-.---.-.----.--FASTLINK------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

If OMF"GNT", FASTLINK sets OMF"OBJ" at end.
If ASSUME is set, set to NOFASTLINK at end by REGPARM"IN".

Comments:

This directive only affects .obj files.

The restrictions are:

• The first parameter in every USING clause is the first item in the Linkage Section.

• The parameters must appear in the same order they appear in the Linkage Section.

• Any Linkage Section items that do not appear in the clause are defined after those that do.

• No BY VALUE parameters are used.

• Only allowed with OMF"OBJ".

• The program must not have any EXTERNAL data items.

FASTSORT

This directive is reserved for internal use by this COBOL system. Because it might appear in the list of settings, it is included for completeness. It is not intended for your use, and its setting should not be changed.

Syntax:

 >>-.---.-.----.--FASTSORT------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

FCDREG

Makes the Compiler define special-registers giving access to File Control Descriptions (FCD) and Key Definition Blocks.

Syntax:

 >>-.---.-.----.--FCDREG--------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

FCDREG causes a special register, FH--FCD, to be created for each File Definition (FD) in the program. This register points to the File Control Description (FCD) for the file. Thus the program can read or change information in the FCD.

For each indexed file, an additional special register, FH--KEYDEF, is created. This register points to the Key Definitions Block for the file.

The layout of the FCD and the Key Definitions Block is given in your Programmer's Guide to File Handling.

FDCLEAR

Specifies that the record buffer for a file should be cleared after every write operation. The buffer is cleared to the value specified by the DEFAULTBYTE directive.

Syntax:

 >>-.---.-.----.--FDCLEAR-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

The FDCLEAR directive is effective while the SELECT clauses are being processed. Use $SET statements before and after each SELECT clause to apply this feature to selected files. Alternatively, use it once before processing any SELECT statements to apply it to all files in a program.

The FDCLEAR directive is ignored for files subject to a SAME RECORD AREA clause in the I-O-Control paragraph.

Example:

In the following code, FDCLEAR is applied only to file-2:

 file-control. 
     select file-1 ... 
$set fdclear 
     select file-2 ... 
$set nofdclear 
     select file-3 ...

FILESHARE

Causes the default locking to become AUTOMATIC rather than EXCLUSIVE for files in a multi-user environment, and automatically locks records on a WRITE or REWRITE statement when the program is locking multiple records.

This directive is not related in any way to Fileshare Version 2.

You must not use this directive when using Fileshare V2.

Syntax:

 >>-.---.-.----.--FILESHARE-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This directive is included for compatibility with earlier file-sharing products. When writing new programs you should use the locking syntax rather than this directive.

FILESHARE is equivalent to WRITELOCK and AUTOLOCK together. NOFILESHARE is equivalent to NOAUTOLOCK and NOWRITELOCK together.

FILETYPE

Specifies the file format to use when creating files.

Syntax:

 >>-.---.--FILETYPE--"integer"--------------><  
    .-/-+

Parameters:

Properties:

Dependencies:

FILETYPE"integer" sets IDXFORMAT"integer" immediately.

Comments:

The possible values of integer are:

Type 14 is available for all file organizations only in the 16-bit COBOL system, and for indexed files in the 32-bit COBOL system for OS/2 V2.

This directive only works on files processed by the Callable File Handler. Use the CALLFH directive if you are processing files other than indexed files.

To produce print files in the style of an IBM mainframe using FILETYPE"11", you must also:

• Set the directives OSVS or VCS2

• Use WRITE AFTER ADVANCING syntax

• Set the CALLFH directive (so your program uses the Callable File Handler)

• Use a SELECT statement in the form:

     select printfile assign "print" organization is 
    sequential.

(Print files in the style of an IBM mainframe have print control character in the first column.)

See also:

ADV Compiler directive
CALLFH Compiler directive

FIXING

Tells the Compiler how many times to perform its final pass which shortens jump instructions.

Syntax:

 >>-.---.--FIXING--"integer"----------------><  
    .-/-+

Parameters:

Properties:

Comments:

The Compiler tries to convert 80x86 jump instructions into their shorter form if possible; this gives smaller object code. Shortening some jump instructions might result in other jumps becoming eligible to be shortened and so further fixing passes might result in further size reduction.

The Compiler does this very quickly, so there should be no need to change this directive. It is maintained only for compatibility with earlier versions of the Compiler.

FIXOPT

Changes the way in which generated code accesses the Data Division in object (.obj) code.

Syntax:

 >>--.---.--.----.-------FIXOPT--------------><
     .-/-+  .-NO-+

Parameters:

None

Properties:

Comments:

If you receive code generation error 78 (Too many code relocations (limit=limit, actual=actual-relocations)) when creating object code, you must recompile specifying FIXOPT.

FLAG

Makes the Compiler produce language-level certification flags when it finds syntax that is not part of a specified dialect of COBOL.

Syntax:

 >>-.---.-.-----FLAG--"dialect"-.-----------><  
    .-/-+ .-NO--FLAG------------+

Parameters:

Properties:

Comments:

The possible values of dialect are:

You cannot use DG, RM or MS as a dialect.

When creating a program that is to be fully ANSI'85 conforming, use:

ANS85 FLAG"ANS85"

and correct anything that causes a flagging message to be produced.

For VSC2, the flagging given depends on the version of VS COBOL II selected by the VSC2 directive. For SAA, the flagging given depends on the level of SAA selected by the SAA directive. For COBOL370, the flagging depends on the level of COBOL/370 selected by the COBOL370 directive.

FLAG-CHIP

Makes the Compiler produce informational messages indicating potential problem areas on the chip architecture in use.

Syntax:

 >>-.---.-.----.--FLAG-CHIP-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

You must set the CHIP directive and ensure that WARNING"3" is set.

Causes the Compiler to generate special informational messages for potential problem areas for the chip architecture given by the CHIP directive. It has no effect if NOCHIP is set. To see the messages, you also need to set WARNING"3". See the chapter Writing Programs in your Programmer's Guide to Writing Programs for more details.

FLAGAS

Makes the Compiler output flagging messages as error messages, warning messages or informational messages.

Syntax:

 >>-.---.-.-----FLAGAS--"severity"-.--------><  
    .-/-+ .-NO--FLAGAS-------------+

Parameters:

Properties:

Comments:

The possible values of severity are:

FLAGCD

When used with the FLAG directive, FLAGCD flags any directive settings that cause incompatibility with the specified dialect of COBOL.

Syntax:

 >>-.---.-.--------FLAGCD--"severity"-.-----><  
    .-/-+ ..----.--FLAGCD-------------+ 
           .-NO-+

Parameters:

Properties:

Comments:

The possible values of severity are:

If FLAGCD is specified with no parameters, the messages are assigned the severity given by the FLAGAS directive, if specified; if FLAGAS is not specified, they are produced as flagging messages.

FLAGMIG

Causes the Compiler to flag any syntax that behaves differently at run time depending on the setting of CMPR2.

Syntax:

 >>-.---.-.----.--FLAGMIG-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Remarks

Such syntax receives W-level messages, with the additional text "MIGR" in the message line to show the warning results from specifying the FLAGMIG directive.

See also:

CMPR2 Compiler directive

FLAGQ

Makes the Compiler ask, each time it gives a flagging message, whether you want to stop compiling.

Syntax:

 >>-.---.-.----.--FLAGQ---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

FLAGSINEDIT

Specifies whether flagging messages are to be included in an error file.

Syntax:

 >>-.---.-.----.--FLAGSINEDIT---------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

This directive has no effect if NOEDITOR is specified. If EDITOR and FLAGSINEDIT are specified, the error file produced contains all flagging messages produced by the Compiler.

See also:

EDITOR Compiler directive

FLAGSTD

Makes the Compiler produce language-level certification flags when it finds syntax that is not part of a specified level of the ANSI'85 Standard.

Syntax:

 >>-.---.-.-----FLAGSTD--"string"-.---------><  
    .-/-+ .-NO--FLAGSTD-----------+

Parameters:

Properties:

Comments:

string contains a list of parameters, each defining a language level or optional module of the ANSI Standard. A feature is not flagged if it is in one of the specified levels or below, or in one of the specified optional modules. A feature is flagged if it is an extension, in a higher level of the Standard, or in an optional module not specified. In addition, it is possible to flag elements identified as Obsolete by the ANSI Standard.

string consists of a combination of the following parameters:

They can appear in any order but must be separated by at least one space. The following combination of flags can be specified:

• You must choose one of the following: M, I, H

• You can choose as many of the following as desired: C1 or C2, D1 or D2, S1 or S2, R

• You can also choose the following flag: O

FLAG and FLAGSTD provide similar functionality and thus only one can be used.

The ANS85 directive must be on.

FOLD-CALL-NAME

Folds the identifier/literal associated with CALL, CANCEL, ENTRY, and CHAIN statements and the program-name in the PROGRAM-ID paragraph to upper or lower case.

Syntax:

 >>-.---.-.-----FOLD-CALL-NAME--"case"-.----><  
    .-/-+ .-NO--FOLD-CALL-NAME---------+

Parameters:

Properties:

Comments:

The Compiler also accepts the directive-name without hyphens; that is FOLDCALLNAME.

In a CALL identifier statement, the name called is given the case specified by this directive, but the actual contents of identifier are not changed.

The FOLD-CALL-NAME directive can help in the transfer of COBOL code from an environment with filenames which are not case specific to one which is. For example, from OS/2 to UNIX.

It also makes the Compiler conform with the ANSI'85 standard interprogram communications module, in which the case of a program-name in a CALL statement does not matter.

With NOFOLD-CALL-NAME, by normal COBOL convention the program-name is folded to upper case unless enclosed in quotes. This also applies to the root program of a nested program. All the subprogram-names are folded to upper case even when in quotes. (Enclosing program-names in quotes is a Micro Focus extension. See the section The Program-Id Paragraph in your Language Reference.)

The program-name is also accessible as an entry point in a fully linked executable.

FOLD-COPY-NAME

Determines whether copyfile-names should be converted to upper case or lower case.

Syntax:

 >>-.---.-.-----FOLD-COPY-NAME--"case"-.----><  
    .-/-+ .-NO--FOLD-COPY-NAME---------+

Parameters:

Properties:

Comments:

The Compiler also accepts the directive-name without hyphens; that is FOLDCOPYNAME.

The FOLD-COPY-NAME directive affects both text and library-names regardless of whether they are specified as user-defined words or literals.

Example:

Specifying FOLD-COPY-NAME"UPPER" converts copyfile names to upper case (also known as folding to upper case).

FORM

Specifies the number of lines on each page of the listing.

Syntax:

 >>-.---.-.-----FORM--"integer"-.-----------><  
    .-/-+ .-NO--FORM------------+

Parameters:

Properties:

Dependencies:

Set to NOFORM at end by NOLIST or MASM.
Set to FORM"60" immediately by LIST.

Comments:

With FORM"integer", a form-feed character is always produced at the head of the listing file. With NOFORM, no form-feed characters or page headings are produced anywhere in the listing.

FP-ROUNDING

Determines whether one floating-point receiving item can affect the results of other, nonfloating-point receiving items.

Syntax:

 >>-.---.-.-----FP-ROUNDING--"dialect"---.--><  
    .-/-+ .-NO--FP-ROUNDING--------------+

Parameters:

Properties:

Comments:

In OS/VS COBOL and VS COBOL II, if any of the sending data items of an arithmetic statement are floating-point, all receiving fields are rounded, regardless of the presence or absence of the ROUNDED phrase and regardless of whether or not the receiving fields are floating-point.

In VS COBOL II, if any receiving data item is defined as floating-point, rounding takes place for all receiving data items regardless of the presence or absence of the ROUNDED phrase, of whether or not the receiving field is floating-point, and of whether or not there are any sending items which are floating-point.

With NOFP-ROUNDING, no rounding takes place for fixed-point targets in calculations involving floating-point operands unless the ROUNDED phrase is explicitly used in the source.

Example:

If two data items are defined as follows:

     05 numeric-field    pic 999. 
     05 floating-field   comp-1.

and the following statement is executed:

     compute numeric-field floating-field = 7.7 + 1

specifying FP-ROUNDING"VSC2" results in numeric-field containing 9 (rounded), whereas specifying NOFP-ROUNDING or FP-ROUNDING"OSVS" would result in numeric-field containing 8 (truncated).

If, using the same two data items, the following statement is executed:

compute numeric-field = +7.6E0

specifying FP-ROUNDING"VSC2" or FP-ROUNDING"OSVS" results in numeric-field containing 8 (rounded), whereas specifying NOFP-ROUNDING results in numeric-field containing 7 (truncated).

GANIM

Makes the Compiler produce extra information so that you can use Xilerator (16-bit COBOL system only) or Animator V2 (32-bit COBOL system for Windows NT and OS/2 V2) to debug generated code.

This directive is retained for compatibility with earlier versions of this COBOL system, and will be removed in a future version. With this version, use the ANIM directive instead.

Syntax:

 >>-.---.-.----.--GANIM---------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

GANIM sets ANIM and NORNIM at end.
If OPT"0", GANIM sets OPT"1" at end.

Comments:

On the 16-bit COBOL system, the Compiler produces an .idy file containing information required for debugging generated code using Xilerator. The .idy file created can also be used to animate the intermediate code with Animator or Animator V2.

With the 32-bit COBOL systems on Windows NT and OS/2 V2, the Compiler produces a .gdy file for animating native code using Animator V2.

GNT

Specifies the name of the object code file.

Syntax:

 >>-.---.--.---GNT---.--"path-name\filename"--.---.---><
    .-/-+  .         .--"path-name/filename"--.   .
           .         .--"path-name\"----------.   .
           .         .--"path-name/"----------.   .
           .         .--"filename"------------.   .
           .         .--()--------------------+   .
           .-NO--GNT------------------------------+

Parameters:

Properties:

Dependencies:

In the 16-bit COBOL system, OBJ and GNT are synonymous. Setting NOGNT also sets NOOBJ.

Comments:

With NOGNT, no object code file is produced.

In the 16-bit COBOL system, setting this directive does not imply OMF"GNT". To ensure that your .gnt file contains .gnt format object code you must make sure that OMF"GNT" is also specified.

On 32-bit COBOL systems, use this directive in preference to the OMF directive.

On 32-bit COBOL systems, this directive creates a file in the format specified, with the name specified. For example, if you specify GNT"myprog.gnt", then a file myprog.gnt is produced that contains generated code.

GNTANLZ

This directive is reserved for use with add-on products. Do not change its setting unless you have the appropriate add-on product.

Specifying GNTANLZ enables the GNT Analzyer, which counts the number of times each COBOL statement has been executed.

Syntax:

 >>-.---.-.-----GNTANLZ--"count-type"-.-----><  
    .-/-+ .-NO--GNTANLZ---------------+

Parameters:

Properties:

Comments:

Possible values of count-type are:

For more details, see the chapter Analyzing Your Applications in your Workbench User Guide.

GNTLEVEL

Enables faster interprogram calls.

Syntax:

 >>-.---.--GNTLEVEL--"level"----------------><  
    .-/-+

Parameters:

Properties:

Comments:

Possible values of level are:

HIDE-MESSAGE

This directive has been replaced by the directive CHANGE-MESSAGE. It remains for compatibility only, and will be removed in a future release.

Registers the number of a syntax check error message to "hide" so that if the error is encountered it is ignored.

Syntax:

 >>-.---.-.-----HIDE-MESSAGE--"integer"-----.----><  
    .-/-+ |                                 | 
          .-NO--HIDE-MESSAGE--.-----------.-+ 
                              .-"integer"-+

Parameters:

Properties:

Comments:

The Compiler also accepts the directive-name without a hyphen; that is HIDEMESSAGE.

HIDE-MESSAGE"integer" adds integer to a list of syntax check error message numbers. To hide several error message numbers using this feature, you must use the HIDE-MESSAGE directive repeatedly.

When the program is being syntax checked, any message with severity E, W, I, or flag with its number in the list is not shown in any listing. It is not included in the error summary at the end of the compilation.

Messages with severity S can be hidden only if they were made severity S by the CHANGE-MESSAGE, FLAGAS or FLAGCD directives. Normally, they are always shown, even if their number appears in this list.

NOHIDE-MESSAGE clears the list of numbers so no messages are hidden.

NOHIDE-MESSAGE"integer" removes only the specified message number from the list.

See also:

CHANGE-MESSAGE Compiler directive
FLAGAS Compiler directive
FLAGCD Compiler directive

HOST-NUMCOMPARE

Controls the operation of comparisons between integer numeric data items of USAGE DISPLAY and alphanumeric literals or figurative constants.

Syntax:

 
>>--.---.--.------HOST-NUMCOMPARE--"integer"--.--><
    .-/-+  .-NO---HOST-NUMCOMPARE-------------+

Parameters:

Properties:

Comments:

The HOST-NUMCOMPARE directive is Early User Syntax support. You must use the EARLY-RELEASE directive to enable this feature. This directive might change or be removed in a later revision of this COBOL system. integer can be:

The HOST-NUMCOMPARE directive only affects comparisons where the numeric data item contains nonnumeric data at the time of the comparison.

You cannot use the HOST-NUMCOMPARE(2) in conjunction with the HOST-NUMCOMPARE(1).

See also:

ZWB Compiler directive

HOST-NUMMOVE

Ensures that run-time error 163 (illegal characters in numeric fields) does not occur when certain MOVE statements are executed on numeric display data items.

Syntax:

 >>--.---.--.------HOST-NUMMOVE--"integer"--.--><
     .-/-+  .-NO---HOST-NUMMOVE-------------+

Parameters:

None

Properties:

Comments:

Possible values of integer are:

HOST-NUMMOVE"2" does not provide 100% compatibility with the IBM mainframe behavior, but does provide emulation for the most common cases.

You cannot use the HOST-NUMMOVE(2) directive in conjunction with the HOST-NUMMOVE directive.

See also:

EARLY-RELEASE Compiler directive
RTS error 163

HOSTSIGNS

Allows arithmetic and moves with COMP-3 items with illegal sign nibbles (such as those produced by a redefinition of the item) to produce the desired result.

This directive is synonymous with the BADSIGNS directive in the 16-bit COBOL system.

Syntax:

On the 16-bit COBOL system:

 >>-.---.-.--------HOSTSIGNS--"integer"-.---><  
    .-/-+ ..----.--HOSTSIGNS------------+ 
           .-NO-+

On the 32-bit COBOL system:

 >>-.---.-.----.---HOSTSIGNS----------------><  
    .-/-+ .-NO-+

Parameters:

Properties:

Comments:

Specifying HOSTSIGNS or HOSTSIGNS"1" causes COMP-3 arithmetic to be performed by a nonoptimized route, reducing the efficiency of your programs.

Specifying HOSTSIGNS"2" applies this to arithmetic and moves of COMP-3 data items.

For smaller, faster code you should not specify HOSTSIGNS.

IBM-MS

Specifies that words reserved in IBM COBOL V1.00 are to be regarded as reserved words, and changes the behavior of certain features to be compatible with that COBOL system.

Syntax:

 >>-.---.-.----.--IBM-MS--------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

IBM-MS sets DEFAULTBYTE"0" and ACCEPTREFRESH immediately.

Comments:

This directive is synonymous with the PC1 and MS"1" directives.

IBMCOMP

Turns on word-storage mode.

Syntax:

 >>-.---.-.----.--IBMCOMP-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

IBMCOMP sets ALIGN"8" immediately.

Comments:

In word-storage mode every data item of USAGE COMP or COMP-5 occupies either two bytes or a multiple of four bytes. If you specify IBMCOMP and you use the SYNCHRONIZED clause on any items defined as USAGE COMP or COMP-5 in your program, do not specify ALIGN"1".

If you are calling any Micro Focus systems routines directly, use IBMCOMP with great care. It might cause the parameters you specify to these routines to be incorrectly aligned and sized, causing the routines to give incorrect results. This includes parameters for the CBL_ library routines, calls to Adis or direct calls to the Callable File Handler, Extfh.

IDXFORMAT

Specifies the format to use when creating indexed files.

This directive has been replaced by the FILETYPE directive and will be removed in a future release.

Syntax:

 >>-.---.--IDXFORMAT--"integer"-------------><  
    .-/-+

Parameters:

Properties:

Dependencies:

Set to IDXFORMAT"integer" immediately by FILETYPE"integer".

Comments:

The possible values of integer are:

Existing files in any of the given formats are processed correctly without the need for this directive. This directive controls the format used when creating new files.

Specifying 3 always causes the format used by this COBOL system to be created; if you specify 0, and you are using your program with a file handler from a different system, the default for that system is created.

Specifying 4 might make the files larger than their IDXFORMAT"3" equivalents.

Micro Focus Level II format files are compatible with Micro Focus products such as Level II COBOL, Professional COBOL V1.2, and VS COBOL Workbench versions up to and including V1.3. (See your Programmer's Guide to File Handling for further information.)

You must not use the ANS85 directive to enable ANSI'85 behavior when using IDXFORMAT"2". However, you can use ANS85"SYNTAX" to enable ANSI'85 syntax.

INCLUDE-FILLER

Causes information regarding FILLER data items to be stored in the .idy file.

Syntax:

 >>-.---.-.----.--INCLUDE-FILLER------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

INCLUDE-FILLER makes FILLER data items visible to tools in Object COBOL for UNIX and Workbench.

Including these data items increases the size of the .idy file.

INDD

Causes ACCEPT statements to be read from a specified file.

Syntax:

 >>-.---.-.-------INDD--"fname rsize rtype ctype"-.>< 
    .-/-+ .-------INDD--"fname rsize rtype"-------© 
          .-------INDD--"fname rsize"-------------© 
          .-------INDD--"fname"-------------------© 
          ..----.-INDD----------------------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

INDD sets NOSYSIN immediately
Set to NOINDD immediately by SYSIN

Comments:

When INDD is specified, all format 1 ACCEPT statements which either have no FROM option or specify FROM SYSIN (or the mnemonic-name associated with SYSIN) are transformed into READ statements, reading from a file with the specified external filename.

The filename can be mapped onto physical filenames in the same way as other files with external filenames; that is, by using environment variables or the External File Mapper, MFExtmap.

The default settings for this directive are the same as those that would be used for this type of ACCEPT statement when the SYSIN directive is used.

See also:

ACCEPT statement
OUTDD Compiler directive
SYSIN Compiler directive

INFORETURN

Specifies the return-code value returned by the Compiler when it produces only informational messages.

Syntax:

 >>-.---.--INFORETURN--"integer"------------><  
    .-/-+

Parameters:

Properties:

Comments:

When the Compiler terminates it returns a value that can be tested by an operating system command to determine the success or otherwise of the compilation. The values for termination are described in your Object COBOL User Guide. This directive enables you to set the value to be returned if the Compiler only outputs informational messages.

INITCALL

Specifies modules to be called immediately before the first statement of the program is executed.

Syntax:

 >>-.---..-----INITCALL--"module"-.------------..><  
    .-/-+|                        .-"priority"-+| 
         .-NO--INITCALL-------------------------+

Parameters:

Properties:

Comments:

The possible values of priority are:

Specifying INITCALL causes the Compiler to insert a call to the named module. At run time, the module specified is called before any procedural code is executed. You cannot pass parameters to the called module.

To call several modules using this feature, you must use the INITCALL directive repeatedly. NOINITCALL clears the list of modules to be called.

Calls specified with a high priority are placed before all calls with a low priority, as well as other calls made by the Compiler. Low priority calls are placed after other calls made by the Compiler. Calls with the same priority are executed in the order they are specified.

INT

Specifies the name of the intermediate code file.

Syntax:

 >>-.---.-.-----INT-.-"pathname\filename"--.-.--><  
    .-/-+ |         .-"pathname/filename"--© | 
          |         .-"pathname\"----------© | 
          |         .-"pathname/"----------© | 
          |         .-"filename"------------© | 
          |         .-()--------------------+ | 
          .-NO--INT---------------------------+

Parameters:

Properties:

Dependencies:

NOINT sets NOANIM and NOSTRUCT immediately.

Comments:

When no .obj file is required from the Compiler, or when the ANIM directive is used, an intermediate code file is produced. This directive specifies the name for that file.

Specifying NOINT prevents the intermediate code file being produced.

If you specify an existing file, it is overwritten.

If filename is not specified, the Compiler uses the source filename (source-name) with the extension .int attached. If you specify pathname but no filename, the Compiler uses the pathname, with the filename source-name.int attached.

INT() causes intermediate code to be put in the standard file source-name.int. With this format you must use parentheses not quotation marks. So, INT"" does not give this result.

Use the INT directive with caution. Incorrect use can abort the compilation process.

INTDATE

Selects the starting date for integer format dates used with date intrinsic functions.

Syntax:

>>-.---.------INTDATE---"type"-----><
   .-/-+

Parameters:

Properties:

Comments:

The INTDATE directive is Early User Syntax support. You must set the EARLY-RELEASE directive to enable this feature. This directive might change or be removed in a later revision of this COBOL system.

This directive selects the starting date for integer format dates used with date intrinsic functions (that is, DATE-OF-INTEGER,DAY-OF-INTEGER, INTEGER-OF-DATE, INTEGER-OF-DAY).

type can be:

For more details on the date intrinsic functions, see your Language Reference.

INTLEVEL

When intermediate code is being created by the Compiler, this directive controls the level of portability of the code created to different versions of Micro Focus COBOL systems in other environments.

Syntax:

 >>-.---.-.-----INTLEVEL--"integer"-.-------><  
    .-/-+ .-NO--INTLEVEL------------+

Parameters:

Properties:

Comments:

Full details of intermediate code portability are included with the relevant COBOL systems.

NOINTLEVEL causes intermediate code to be created that is suitable for execution only in this environment.

INTLEVEL"integer" creates intermediate code that can be executed by some versions of Micro Focus products in other environments. For portability between environments, the value of integer used for compilation must be supported by the Micro Focus COBOL system on each environment on which you want to execute the intermediate code.

INTLEVEL"integer" can limit the syntax that can be used in your program.

IOCONV

Selects between a straight group move and conversion of elementary record descriptions (ANSI behavior) when processing READ ... INTO and WRITE ... FROM statements.

Syntax:

 >>-.---.-.----.--IOCONV--------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

NOIOCONV causes a straight group move for READ ... INTO and WRITE ... FROM operations. This is for backward compatibility.

IOCONV causes conversion of elementary record descriptions when required. This is normal ANSI behavior, and the default.

JAPANESE

Enables the use of Micro Focus Japanese Language Extension (PIC N, Japanese data-names and Japanese procedure-names).

Syntax:

 >>-.---.-.-------JAPANESE--"integer"-.-----><  
    .-/-+ ..----.-JAPANESE------------+ 
           .-NO-+

Parameters:

Properties:

Comments:

The possible values of integer are:

Specifying JAPANESE with no parameter is the same as specifying JAPANESE"2".

This directive is provided for compatibility purposes only. It has been replaced by the NCHAR directive. JAPANESE and NCHAR are synonymous. The JAPANESE and DBCS directives are mutually exclusive.

See also:

DBCS Compiler directive
DBSPACE Compiler directive
NCHAR Compiler directive

KEEP-INT

Specifies whether the Compiler should produce intermediate code for an unsuccessful compilation; that is a compilation that produces serious or unrecoverable errors.

Syntax:

 >>-.---.-.----.--KEEP-INT------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

The Compiler also accepts the directive-name without a hyphen; that is KEEPINT.

In the case of segmented programs, should any segment have an unsuccessful compilation with NOKEEP-INT selected then intermediate code is not produced for any of the segments.

KEYCHECK

Specifies whether the number of keys should be checked when a file is opened.

Syntax:

 >>---.---.--.----.-----KEYCHECK----------><
      .-/-+  .-NO-+

Properties:

Comments:

By specifying NOKEYCHECK the file handler will not check the number of keys specified in the SELECT statement against the number of keys held in the files information area.

KEYCOMPRESS

Specifies the type of key compression to be done on indexed files.

Syntax:

 >>-.---.-.-----KEYCOMPRESS--"integer"-.----><  
    .-/-+ .-NO--KEYCOMPRESS------------+

Parameters:

Properties:

Comments:

integer can contain any combination of the following values:

You can specify any combination of these numbers by adding these values together.

KEYCOMPRESS"0" is equivalent to NOKEYCOMPRESS.

You need specify key compression only when creating the file. Subsequently, the key compression is detected when the file is opened.

To get key compression on an individual file, use $SET statements in your source so that this directive is in effect only for the part of the source containing the relevant KEY clause in the file's SELECT statement.

LIBRARIAN

Allows -INC statements in your program.

Syntax:

 >>-.---.-.----.--LIBRARIAN-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

The -INC statement specifies a file for inclusion in the source program. The string -INC must be written as a contiguous sequence of upper case characters starting in column 1, followed by one or more spaces, and then, on the same line, by the name of a file containing COBOL source. This file is included in the program at the point where the -INC statement appears.

LINE-COUNT

Controls the detail of information displayed at the end of a source listing.

Syntax:

 >>-.---.-.-------LINE-COUNT--"integer"-.----><  
    .-/-+ ..----.-LINE-COUNT------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

LINE-COUNT"2" sets CSI immediately.

Comments:

integer controls the detail displayed, as follows:

Specifying LINE-COUNT with no integer is the same as specifying LINE-COUNT"2".

See also:

CSI Compiler directive
LIST Compiler directive

LINKCHECK

Specifies that each time a Linkage Section item is referenced a check is to be made that it exists.

Syntax:

 >>-.---.-.----.--LINKCHECK-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

16-bit only:

On the 16-bit COBOL system, the PARAMCOUNTCHECK directive must also be specified if you intend to use the LINKCHECK directive.

If reference is made to a Linkage Section item that does not exist, a run-time error 203 (CALL parameter not supplied) is generated.

See also:

PARAMCOUNTCHECK Compiler directive

LINKCOUNT

Specifies the maximum number of Linkage Section items, external data items, and external files (as defined for interprogram communication) allowed.

Syntax:

 >>-.---.-.----.--LINKCOUNT--"integer"------><  
    .-/-+ .-NO-+

Parameters:

Properties:

Comments:

The LINKCOUNT directive is only required if your program contains file or data records (in the Data Division) that are marked as EXTERNAL. If you have none of these, you can allocate as many Linkage Section records as you like.

If you do have external items in your Data Division, when it encounters the first such item, the Compiler has to create a fixed length area to contain the references to all external areas and Linkage Section items. In this case, you use the LINKCOUNT directive to specify how many references this area needs to hold. It must be enough for all the external data items and Linkage Section items in your program.

integer must not be less than the total number of Linkage Section data items, external data items, and external files. It represents a count of every separately-addressable data item in these structures.

integer has a maximum value of 65535. However, the maximum size will depend on how many other data items are defined. If it is set too high a 781-S 'Maximum number of data items exceeded' error might occur. In this case the value in LINKCOUNT should be lowered.

LINKLIB

Specifies the link library or libraries where the Linker is to look for the run-time support routines.

Syntax:

 >>-.---.-.-----LINKLIB--"library-names"-.--><  
    .-/-+ .-NO--LINKLIB------------------+

Parameters:

Properties:

Comments:

Specifying library-names as NUL is the same as NOLINKLIB. The choice of libraries can be subsequently overridden by a directive to the Linker.

The names must be separated by a plus sign (+). If a name has no extension, .lib is assumed.

LIST

Specifies the destination of the source listing file.

Syntax:

 >>-.---.-.-------LIST-.-"destination"-..---><  
    .-/-+ |            .-()------------+| 
          ..----.-LIST------------------+ 
           .-NO-+

Parameters:

Properties:

Dependencies:

NOLIST sets NOFORM, NOREF, NOSETTING, NOSOURCEASM and NOXREF at end.
LIST sets FORM"60" immediately.
Set to LIST by ASM immediately.
Set to NOLIST by RNIM at end.

Comments:

If you specify an existing file, it is overwritten.

When NOLIST is specified, no source listing is produced. If you specify LIST with no destination, the source listing is sent to the screen. If you specify either destination or (), you cannot use this directive in a $SET statement.

destination can be the name of any suitable device. On DOS, Windows and OS/2, use CON: for the screen. On UNIX, the device selected must not be under the control of a system spooler.

NOLIST and LIST with no parameter can be used in $SET statements in a program to list selected parts of the program. The destination of the listing cannot be changed in this way.

LIST() causes the source listing to be put in the file source-name.lst, where source-name is the root of the name of the program being compiled. When used with the 16-bit Generator this creates the file source-name.grp.

Example:

If you want to list the source to a file for every compilation you do, place the LIST() directive in the cobol.dir file. This overrides the default NOLIST.

Alternatively, if you already have a LIST directive in your cobol.dir, making every listing go to the screen, you can override it by using LIST() on the command line.

LISTPATH

Specifies a path on which the list file is to be written. The name of the list file is source-name.lst.

Syntax:

 >>-.---.-.----.--LISTPATH--"list-path"-----><  
    .-/-+ .-NO-+

Parameters:

Properties:

LISTWIDTH

Sets the width of the listing.

Syntax:

 >>-.---.--.-LISTWIDTH-.-"width"------------><  
    .-/-+  .-LW--------+

Parameters:

Properties:

Dependencies:

If REF is set, and width is less than 90, sets LISTWIDTH"90" immediately. If width is less than 90, sets NOREF at end.

Comments:

LISTWIDTH"132" causes additional information to be displayed for each line listed. This includes the first eight characters of the current copyfile-name (spaces for the main file) together with the number of the line relative to the start of that file.

See also:

REF Compiler directive

LITLINK

Makes the Compiler declare the literals in CALL literal statements as public symbols, so they are resolved at link time rather than run time. (Calls generated in this way are referred to as litlinked.)

Syntax:

With the 16-bit Compiler:

 >>-.---.-.----.-LITLINK--------------------><  
    .-/-+ .-NO-+

With a 32-bit Compiler:

 >>-.---.-.-----LITLINK-.-----------.-.-----><  
    .-/-+ |             .-"integer"-+ | 
          .-NO--LITLINK---------------+

Parameters:

Properties:

Dependencies:

If OMF"GNT", LITLINK sets OMF"OBJ" at end.
Set to LITLINK at end by MODEL"SMALL", MODEL"COMPACT" or OBJLITE.
If ASSUME is set, set to LITLINK at end by REGPARM"OUT".

Comments:

This directive only affects .obj files.

With NOLITLINK on 16-bit and LITLINK"2" on 32-bit, if the name in a CALL literal statement starts with a single underscore - "_name" - a litlinked call to entry-name _name is generated. If it starts with a double underscore - "__name", a litlinked call to entry-name name is generated. (That is the entry-names are declared as public symbols.)

With LITLINK on all environments or LITLINK"1" on 32-bit, all CALL literal statements generate litlinked calls, regardless of the inclusion of underscores at the start of the literal.

With NOLITLINK on 32-bit, no CALL literal statements generate litlinked calls, regardless of the inclusion of underscores at the start of the literal.

LITLINK"2" is only intended to provide backward compatibility with code written for the 16-bit COBOL system. You should use call-convention 8 for new programs.

LITVAL-SIZE

Specifies the number of bytes to pass if the SIZE clause is omitted when passing numeric literals BY VALUE.

Syntax:

 >>-.---.--LITVAL-SIZE--"integer"-----------><  
    .-/-+

Parameters:

Properties:

LNKALIGN

Indicates that any linkage records specified in a USING statement are 01 or level-77 items, aligned according to the ALIGN directive.

Syntax:

 >>-.---.-.----.--LNKALIGN------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

Using LNKALIGN can reduce the time needed to access a linkage item at run time. However, take care when using it. You must use the correct ALIGN directive setting when compiling your program. No checks are made to ensure that the items are aligned.

The results of using this directive are machine-dependent, and you should not use it if you are in any doubt as to its effect. Unpredictable results might occur if you pass linkage items that are not aligned.

See also:

ALIGN Compiler directive

LOCALCOUNT

This directive is reserved for internal use by this COBOL system. Because it might appear in the list of settings, it is included for completeness. It is not intended for your use , and its setting should not be changed.

Syntax:

 >>-.---.--LOCALCOUNT--"integer"------------><  
    .-/-+

Parameters:

Properties:

LOCKTYPE

Specifies the type of record locking.

Syntax:

 >>-.---.--LOCKTYPE--"integer"--------------><  
    .-/-+

Parameters:

Properties:

Comments:

The possible values of integer are:

This directive only has an effect if the CALLFH directive is used.

See also:

CALLFH Compiler directive

LOGICOPT

Optimizes calls to the CBL_ logical routines so that they do not affect RETURN-CODE.

Syntax:

 >>-.---.-.----.--LOGICOPT------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

Specifying LOGICOPT optimizes COBOL logical calls, such as CBL_AND, so that they do not affect RETURN-CODE.

MAKESYN

Makes one reserved word synonymous with another.

Syntax:

On DOS, Windows and OS/2:

 >>-.---.-MAKESYN-"rsv-word-1" = "rsv-word-2"->< 
    .-/-+

On UNIX:

 >>--MAKESYN-"(rsv-word-1) == (rsv-word-2)"--->< 

Parameters:

Properties:

Comments:

The equals sign must be surrounded by spaces.

This directive does not appear in the list created with the SETTING directive.

MAPNAME

Makes the Compiler alter program-names and entry-points to make them compatible with OS/VS COBOL, DOS/VS COBOL, VS COBOL II and COBOL/370.

Syntax:

 >>-.---.-.----.--MAPNAME-------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Comments:

Specifying MAPNAME affects program-names and entry-points defined in the compilation and those referenced by the program as follows:

• ENTRY statements only in the outermost program

• CALL literal, but not CALL identifier, statements

• PROGRAM-ID entries in the outermost program, but not in nested programs.

When a name is found to be incompatible a warning message is given and a modified name is used in the object program. The rules enforced and the modifications made are:

• The first eight characters must be unique among program-names. The name is truncated to the first eight characters.

• The first character must be alphabetic. If it is numeric it is changed as follows:

           0       becomes    J 
           1 - 9   become     A - I

• Hyphen (-) is not allowed. If one appears it is changed to zero (0).

MASM

Makes the Compiler produce a listing file that can be submitted to Microsoft MASM to produce an .obj file.

Syntax:

 >>-.---.-.----.--MASM----------------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

MASM sets NOFORM at end.
Set to NOMASM by NOASM or OMF"GNT".

Comments:

This directive only affects .obj files.

You could use this directive to investigate the performance of small sections of code, as it enables you to edit the generated output. You should only use this directive on programs with one code segment, or the code produced does not work correctly.

You must set the ASMLIST directive to specify the destination of the MASM-compatible listing. You can still use SOURCEASM to enable you to include source lines as comments. An .obj file is still generated, although it might differ slightly from the .obj file produced by MASM.

MAX-ERROR

Causes the Compiler to abort when the specified number of errors have been produced.

Syntax:

 >>-.---.-.-----MAX-ERROR--"err-cnt-.-------.-"-.-><  
    .-/-+ |                         .-svrty-+   | 
          .-NO--MAX-ERROR-----------------------+ 
 

Parameters:

Properties:

Comments:

The Compiler also accepts the directive-name without a hyphen; that is MAXERROR.

If the maximum number of messages of at least the specified severity is reached then the compilation stops. The messages are only counted if produced and after any modification to the severity of the message.

Example:

Specifying the directive:

MAX-ERROR"100 E"

causes the compilation to terminate when 101 messages of severity E or S have been produced.

See also:

CHANGEMESSAGE Compiler directive
FLAG Compiler directive
FLAGAS Compiler directive
FLAGSTD Compiler directive
HIDEMESSAGE Compiler directive
WARNING Compiler directive

MF

Facilitates forward compatibility with Micro Focus COBOL systems by selectively enabling Micro Focus-specific reserved words and changing the behavior of certain features to be compatible with particular versions.

Syntax:

 >>-.---.-.-------.-MF------.-"integer"-.---><  
    .-/-+ |       .-MFLEVEL-+           | 
          ..----.-.-MF------.-----------+ 
           .-NO-+ .-MFLEVEL-+

Parameters:

Properties:

Dependencies:

MF"7" sets DBCS"2" immediately.
If integer > 7, sets DBCS"3" and DBSPACE immediately.
If NOEARLY-RELEASE and MF are set, MF defaults to MF"10".
If EARLY-RELEASE and MF are set, MF defaults to MF"11".

Comments:

The possible values of integer are:

Behavior of the BLANK LINE phrase in previous versions of this COBOL system can be implemented using the OLDBLANKLINE directive. Behavior of the NEXT SENTENCE phrase in previous versions of this COBOL system can be implemented using the OLDNEXTSENTENCE directive.

MF"11" contains all reserved words in MF"10" together with those that are part of the Early Release syntax. Specifying MF without a parameter sets MF"11" if the EARLY-RELEASE directive has already been set. Otherwise, it sets MF"10". It is not possible to set MF"11" unless EARLY-RELEASE has already been set.

See also:

EARLY-RELEASE Compiler directive
OLDBLANKLINE Compiler directive
OLDNEXTSENTENCE Compiler directive

MFCOMMENT

Treats lines with an asterisk (*) in column 1 the same as comment lines but does not show them in the source listing.

Syntax:

 >>-.---.-.----.--MFCOMMENT-----------------><  
    .-/-+ .-NO-+

Parameters:

None

Properties:

Dependencies:

Set to NOMFCOMMENT at end by SOURCEFORMAT"FREE".

Comments:

With MFCOMMENT, lines with an asterisk in column 1 are ignored by the Compiler and do not appear in the source listing (though they do appear in Animator). With NOMFCOMMENT, the asterisk forms part of the sequence number and has no special significance.

MF-OO

Enables the object orientation (OO) facilities in the Compiler.

Syntax: