| Procedure Division - Intrinsic Functions | Procedure Division - ENTER - INSPECT | |
The ACCEPT statement causes data keyed at the console or supplied by the operating system to be made available to the program in a specified data item.
Alternatively, function-name
can itself be used instead of an associated mnemonic-name.
When mnemonic-name is
associated with ARGUMENT-NUMBER in the SPECIAL-NAMES paragraph, then
identifier must be an unsigned integer.
When mnemonic-name is associated with ARGUMENT-VALUE or
ENVIRONMENT-VALUE in the SPECIAL-NAMES paragraph, then identifier must be an
alphanumeric data item.
Identifier can be a USAGE
DISPLAY-1 (DBCS) item or an external floating-point data item.
Identifier can be an internal
floating-point data item.
The EXCEPTION phrase may be specified only if FROM is specified with either
ENVIRONMENT-NAME or ARGUMENT-VALUE, or with the mnemonic-names associated
with them. Identifier can be an internal
floating-point or external floating-point item. No restrictions apply to the
class, category or usage of the identifier. However, the actual value placed
into the identifier and the validity of moving such values to the identifier
are dependent on the FROM clause. See the Format 3 General Rules for more
information. Screen-name cannot be an item
with an OCCURS clause. The LINE and COLUMN phrases can
appear in any order.
EXCEPTION and ESCAPE are
equivalent.
Identifier-4 must be a PIC 9(4)
or a PIC 9(6) data item. Identifier-8 must be an
integer. It can be signed.
Integer-7 can be signed.
An ACCEPT statement whose
operand is not a screen-name is treated as a Format 5 ACCEPT statement if it
has an AT phrase, a FROM phrase with the CRT option, a WITH phrase, a MODE
IS BLOCK phrase, or an EXCEPTION phrase; or if it has no FROM phrase but the
CONSOLE option clause is specified in the SPECIAL-NAMES paragraph. If it has
the FROM phrase with the CONSOLE option, or if it has no FROM phrase and the
CONSOLE IS CRT clause is not specified in the SPECIAL-NAMES paragraph, it is
treated as a Format 1 ACCEPT statement.
The phrases following the
identifier can be in any order.
The SPACE-FILL, ZERO-FILL,
LEFT-JUSTIFY, RIGHT-JUSTIFY, PROMPT and TRAILING-SIGN options are allowed
only if the operand is an elementary item.
Elementary data items in
identifier-1 must be of USAGE DISPLAY.
No elementary item in
identifier-1 may be longer than 8191 bytes. If the MODE IS BLOCK phrase is
used, the whole of identifier-1 must be no longer than 8191 bytes. The size of a data transfer is determined by the device and the run-time environment (see The SPECIAL-NAMES Paragraph for a list of function-names that can be used and your COBOL system documentation for details of devices and the limits of data transfer sizes).
If the device is capable of transferring data of the same size as the receiving data item, the transferred data is stored in the receiving data item. If otherwise, then:
If the FROM option is not given, it is equivalent to specifying FROM CONSOLE.
If the function-name
COMMAND-LINE, or a mnemonic-name associated with the function-name
COMMAND-LINE, is specified, the contents of a system-dependent command-line
buffer are transferred to the receiving data item.
If the mnemonic-name
associated with the function-name ARGUMENT-NUMBER is used, identifier
receives the number of arguments contained in the command line.
(This includes the program-name itself and any arguments following the program-name.)
When mnemonic-name is
associated with ARGUMENT-VALUE, the current command-line argument is placed
into identifier. The determination of which command-line argument is current
is as follows:
If the current command-line argument number is 0, it is intended that the program-name of the main program of the run unit will be returned. However, the effects of variations on program invocation can impact these results so that a utility or calling program can be returned instead.
If the current command-line argument number
is not in the range 0 to 99
inclusive or
exceeds the number of arguments on the command line when this ACCEPT statement is executed, imperative-statement-1, if specified, will be executed.
If the mnemonic-name
associated with the function-name ENVIRONMENT-VALUE is used then,
The effect of retrieving
command line arguments and the number of arguments in a program that is
called by another program is defined as the same as if they were retrieved
by the first program in the run unit.
DAY-OF-WEEK,
and TIME are conceptual data items and, therefore, are not described in the COBOL program.
DAY-OF-WEEK is composed of a
single data element whose content represents the day of the week.
DAY-OF-WEEK, when accessed by a COBOL program, behaves as if it had been
described in a COBOL program as an unsigned elementary numeric integer data
item one digit in length. In DAY-OF-WEEK, the value 1 represents Monday, 2
represents Tuesday, ... , 7 represents Sunday.
The value returned from the
ACCEPT FROM LINE NUMBER phrase is always numeric. The value is
implementation-dependent.
The FROM USER NAME option
returns a user-id number on UNIX systems, or spaces on systems where such a
concept is meaningless.
The FROM ESCAPE KEY option
returns the two digit code generated by a termination key.
The EXCEPTION STATUS item
contains a three-digit numeric value that identifies the type of exception
condition that has occurred during the execution of a CALL statement.
If EXCEPTION STATUS is to be examined this should be done immediately following the CALL statement. There must be nothing between the CALL and ACCEPT FROM EXCEPTION STATUS. File I-O operations will alter the exception status value making it unpredictable.
This format of the ACCEPT
statement accepts screen items,which are defined in the Screen Section of
the program, and allows full access to the enhanced screen handling
facilities. The order of execution of an
ACCEPT statement is always:
The AT phrase gives the
absolute address on the screen where the ACCEPT operation is to start.
If integer-3 or identifier-4 is
4 digits long, the first two digits specify the line, the second two the
column. If 6 digits long, the first three digits specify the line, while the
second three specify the column.
Certain combinations of line
and column numbers have special meanings, as follows:
Prior
to the execution of this ACCEPT statement, a DISPLAY statement which
specifies the same screen-name or identifier-1 as is specified in this
ACCEPT statement must have been executed. There must not have been any
ACCEPT or DISPLAY statement executed since then.
If the
ON EXCEPTION phrase is specified, imperative-statement-1 is executed if
the ACCEPT operation finishes with anything other than a normal termination.
If the NOT ON EXCEPTION phrase is specified, imperative-statement-2 is
executed if the ACCEPT operation terminates normally. (See the section The
CRT STATUS Clause earlier in this chapter for possible types of
termination.) The ACCEPT operation starts at
line 1, column 1 if no AT phrase is specified.
If identifier is a group item
and no MODE IS BLOCK phrase exists, then those elementary subordinate items
which have names other than FILLER are accepted. They are positioned on the
screen in the order their descriptions appear in the Data Division and are
separated by the lengths of the FILLER items in the group. For this purpose,
the first position on a line is regarded as immediately following the last
position on the previous line. The items are accepted in the same order.
Unless otherwise specified in the CURSOR IS clause (see the section The CURSOR IS Clause earlier in this chapter), the cursor is initially positioned at the start of the first item. As the ACCEPT operation into each item is terminated, the cursor moves to the start of the next item.
If identifier-1 is a group item that has a variable-occurrence data item
subordinate to it, the ACCEPT statement acts as if the MODE IS BLOCK clause
were specified.
The MODE IS BLOCK phrase
indicates that the identifier is to be treated as an elementary item. Thus,
even if it is a group item it is displayed as one item.
If the PROMPT option is not
specified, no character is output to mark empty character-positions. PROMPT
without the CHARACTER option causes this to be done, using the character
configured for this purpose.
If the PROMPT option is not
specified, those character-positions into which the operator does not enter
data produce spaces in the data item.
The WITH phrase allows you to
specify certain options available during the operation. (See The Screen
Section earlier in this chapter for descriptions of these clauses.)
In addition to the options available as screen description clauses, the following options can be used in the WITH phrase; SPACE-FILL, ZERO-FILL, LEFT-JUSTIFY, RIGHT-JUSTIFY TRAILING-SIGN and UPDATE. ZERO-FILL appears in this list and as a screen description clause because it has two different uses. Its second use is documented later in this chapter.
A configuration option is available which allows the entry of data into numeric and numeric edited screen fields in free format mode. In COBOL, nonedited numeric data items are intended for holding data in an internal form; however, this format enables such data items to appear on the screen. See your COBOL system documentation on user interfaces for more details. If free format mode is in effect, the data will appear automatically reformatted as follows:
The SPACE-FILL, ZERO-FILL, LEFT-JUSTIFY, RIGHT-JUSTIFY and TRAILING-SIGN options amend this format.
The SPACE-FILL option causes
data in free format nonedited numeric data items to appear on the screen
with zero-suppression in all integer character positions. This option
affects only free format, nonedited numeric data items. This takes effect
when initial data in the data item is displayed and again when the ACCEPT
operation into the data item is terminated. Any leading sign is displayed in
the rightmost space.
The ZERO-FILL option causes
data in free format nonedited numeric data items to appear on the screen
with no zero-suppression. This takes effect when initial data in the data
item is displayed and again when the ACCEPT operation into the data item is
terminated. (See the section The ZERO-FILL Clause earlier in this
chapter for this option's effect when used with alphabetic or alphanumeric
data items.)
The LEFT-JUSTIFY option is
documentary only.
The RIGHT-JUSTIFY option causes
operator-keyed characters to be moved on the screen to the rightmost
character positions of the field. This option affects only free format
nonedited numeric data items. This takes effect upon display of the initial
data (the current contents displayed) in the data item and also upon
termination of the ACCEPT operation.
The TRAILING-SIGN option causes
the operational sign to appear in the rightmost character position of the
field. This takes effect upon display of initial data in the data item and
also upon termination of the ACCEPT operation. This option affects only
signed, nonedited numeric data items which are in free format mode.
The UPDATE option causes the
current contents (initial data) of the data item to be displayed before the
operator is prompted to key in new input. If the operator does not key in
any new data, the initial data is then treated as though it were
operator-keyed. If the UPDATE option is not specified, the display of
initial data is a configuration option. (See your COBOL system documentation
on user interfaces for details of configuration options.)
The WITH UPPER option forces
input into upper case.
The WITH LOWER option forces
input into lower case.
If REDEFINES is used with
identifier-1, the first description of the redefined data area is used and
subsequent descriptions are ignored. If OCCURS or nested OCCURS are used the
repeated data item is expanded into the full number of times it occurs, so
that one definition is repeated for many fields.
If identifier-8 or integer-7
has a negative value, this represents a request that no time-out "exception"
should occur no matter how long there is between/before key-strokes.
If identifier-8 or integer-7 is
zero, this indicates that a time-out should not occur if characters are
waiting. However, if no characters are waiting (when the ACCEPT is
processed), then issue a time-out immediately.
The ON EXCEPTION clause, if
present, will be executed when a time-out occurs and a TIME-OUT clause is
specified. The NOT ON EXCEPTION clause, if present, will be executed when a
TIME-OUT clause is specified but no time-out (or other exception) occurs.
If a time-out exception occurs,
then the contents of any ACCEPT resultant-field will be defined as follows:
If at run time a positive
timeout interval greater than 2,147,483,647 hundredths of a second is
detected the timeout value will be reset to 2,147,483,647 hundredths of a
second (which is approximately eight months).
The TIME-OUT value specifies
the number of seconds (or tenths of a second) after the ACCEPT statement
begins processing until a time-out exception occurs. A new ADISCF
configuration option controls whether or not the timeout clock is "reset"
each time a new keyboard action is detected. An application which wants some
ACCEPT statements to be handled with a RESET and some without, can make
specific calls to the ADIS run-time interface before or after the ACCEPT
statements that need changes from the default. For example when ADIS was
configured not to do resets, if a program had:
ACCEPT INPUT-FIELD TIME-OUT AFTER +10
a timeout would occur after 10 seconds had elapsed from the beginning of the ACCEPT, even if some characters were entered after 5 seconds into the ACCEPT processing.
If, on the other hand, ADIS were configured to do resets, if a program had the same code given above, then each time a new character were entered, the TIMEOUT "clock" would be reset to zero.
If a timeout exception occurs
and no ON EXCEPTION phrase is specified, the CRT Status keys (if specified)
are updated, the application continues to the next logical phrase, and the
contents of the ACCEPT receiving field are as defined above. (The above is
true whether or not a NOT ON EXCEPTION phrase is specified.)
All references to (NOT) ON
EXCEPTION phrases also apply to (NOT) ON ESCAPE phrases. For alphanumeric data items,
the size of the field that is accepted from the screen during an ACCEPT
statement is exactly the same size as the target field. Therefore, if you
want data to appear at the right hand side of a field, you must enter it
there. The ADD statement causes two or more numeric operands to be summed and the result to be stored.
Floating-point literals and
floating-point data items can be used anywhere a numeric data item or
literal can be specified. Your COBOL system ensures that
enough places are carried so as not to lose any significant digits during
execution. The ALTER statement modifies a predetermined sequence of operations.
The ALTER statement is classed
as an obsolete element in the ANSI'85 standard and is scheduled to be deleted
from the next full revision of the ANSI Standard.
All dialects in this COBOL
implementation fully support this syntax. The FLAGSTD Compiler directive can
be used to detect all occurrences of this syntax.
Although it is a part of the
standard COBOL definition, this verb is explicitly excluded from the X/Open
COBOL language definitions and should not be used in a conforming X/Open
COBOL source program.
All other uses of the ALTER statement are valid and are performed even if procedure-name-1, procedure-name-3 is in an overlayable fixed segment.
The CALL statement causes control to be transferred from one object program to another within the run unit.
You may
CALL the target object program procedure indirectly, by address. Data,
pointers to data, and pointers to procedures can be passed as parameters to
the called program procedure, and returned by it. Parameter passing
conventions can be controlled in order to allow the called program procedure
to be written in languages other than COBOL, using different conventions. A
called COBOL program can recursively call itself when a Local-Storage Section
exists.
Syntax rules marked with an asterisk (*) are common to the CALL statement and to the CHAIN statement. In these rules, where "CALL or CHAIN statement" is specified, you should read "CALL statement" or "CHAIN statement" when applying the rules to the CALL statement or the CHAIN statement respectively.
Mnemonic-name is necessary only
if the object program that is to be invoked will expect a calling convention
other than that used as the default by your COBOL system. Typically, the
default COBOL calling convention is consistent with that used by a
significant implementation of a non-COBOL language for the run-time
environment.
Mnemonic-name must be defined
in the SPECIAL-NAMES paragraph. See the section The SPECIAL-NAMES
Paragraph earlier in this chapter for details of how to do this and
your COBOL system documentation on interfacing for details
of which calling conventions are supported in your run-time environment.
, literal-2
, and literal-3
must be nonnumeric literals and cannot be a figurative constant.
* Integer-1 may be signed.
Local-Storage Section
Communications Section or Linkage Section, and must be a level 01 data item, or a level 77 data item
, or an elementary item
, or any group item.
* Identifier-3 must be defined
as a data item in the Linkage Section
, Local-Storage Section or
Working-Storage Section,
and must be a level 01 data
item, a level 77 data item
, or other level data item.
* Neither identifier-5,
identifier-6, identifier-7, nor identifier-8 can be a function-identifier.
* GIVING and RETURNING are
equivalent.
* Identifier-8 must be 4 bytes
in size.
Identifier-9 must be defined as
a data item in the Linkage Section with a level number of 01 or 77.
If the CALL statement specifies
literal-1 (but not if it specifies identifier-1 or procedure-pointer-1), and
if a call prototype (a program with the EXTERNAL clause in its PROGRAM-ID
paragraph) has been included in the current compilation unit with a name
that matches literal-1, then the following items are validated during syntax
checking:
General rules marked with an asterisk (*) are common to the CALL statement and to the CHAIN statement. In these rules, where "CALL or CHAIN statement" are specified, you should read "CALL statement" or "CHAIN statement" when applying the rules to the CALL statement or the CHAIN statement respectively.
Mnemonic-name identifies the
calling convention that this program will use when invoking the called
program. If the calling convention assumed by the called program is
different to that implied by mnemonic-name, then the COBOL system may become
corrupt.
or called procedure.
The program in which the CALL statement appears is the calling program. If the program being called is a COBOL program, literal-1 or the content of the data item referenced by identifier-1 must contain the program-name contained in the PROGRAM-ID paragraph of the called program (see The Program-ID Paragraph earlier in this chapter, particularly in regard to case sensitivity),
or the entry-name contained in
the ENTRY statement of a called program that must have been called
previously and not canceled since. Alternatively, the name of the called
program may be a name which identifies a file that contains executable
program code for the called program. The result of specifying different
names that identify the same file in two CALL statements or a CALL and a
CANCEL statement are operating system specific. See your COBOL system
documentation for details of how your system matches the given name to
previously called programs, program filenames or programs held in library
files.
If the program being called is
not a COBOL program, the rules for the formation of the program or
procedure name are given in the COBOL system documentation on interfacing.
The content of the data item
referenced by procedure-pointer-1 is the address of the called program that
must be already resident in memory. If the program being called is a COBOL
program, the address of the called program must originally have been derived
from a SET statement with an ENTRY phrase and no CANCEL statement for that
program must have been subsequently executed. If the program being called is
not a COBOL program then the address of the called program must originally
have been derived from a CALL to a non-COBOL program or procedure. It is an
error to attempt to CALL a program with an address that is equal to NULL.
or ON EXCEPTION
phrase is ignored and control is passed to the end of the CALL statement
or, if the NOT ON EXCEPTION
phrase is specified control is passed to imperative-statement-2. If control
is passed to imperative-statement-2, execution continues according to the
rules for each statement specified in imperative-statement-2. If a
procedure branching or conditional statement which causes explicit transfer
of control is executed, control is transferred in accordance with the rules
for that statement; otherwise, upon completion of the execution of
imperative-statement-2, control is transferred to the end of the CALL
statement.
If the program specified by the CALL is not resident in memory, is not available for dynamic loading or is available but finds insufficient memory in which to load, then the program cannot be made available at that time and one of the following two actions will occur:
or ON EXCEPTION
phrase is specified in the CALL statement, control is transferred to imperative-statement-1. Execution then continues according to the rules for each statement specified in imperative-statement-1. If a procedure branching or conditional statement which causes explicit transfer of control is executed, control is transferred in accordance with the rules for that statement; otherwise, upon completion of execution of imperative-statement-1, control is transferred to the end of the CALL statement
and the NOT ON EXCEPTION
phrase, if specified, is ignored.
If the ON OVERFLOW or ON
EXCEPTION phrase is not specified in the CALL statement, the NOT ON
EXCEPTION phrase, if specified, is ignored.
Program execution will stop,
without an implicit CLOSE statement for any file that is in open mode in
the run unit, and a run time error will be displayed.
The run unit remains unaware of
whether a called program is a COBOL program or a non-COBOL program or
procedure until the called program has been located, prior to being made
available for execution. The form of the program-name cannot be used, in
general, to determine whether the program is a COBOL program or not. It is
an error for a COBOL program and a non-COBOL program or for two non-COBOL
programs to have the same program-name.
Two or more COBOL programs in
the run unit can have the same program-name, and the reference in a CALL
statement to such a program-name is resolved by using the scope of names
conventions for program-names. See the section Conventions for
Program-names in the chapter Concepts of the COBOL Language.
For example, when two programs
in the run unit have the same name as that specified in a CALL statement:
The mechanism used in this
example is as follows:
If the called program is not COBOL, it is only in its initial state the first time it is called within a run unit. A CANCEL of such a program has no effect.
On all other entries into the called program, the state of the program and each program directly or indirectly contained within it remains unchanged from its state when last exited. Some implementations of languages other than COBOL may provide alternative semantics and, if available, these will be detailed in your COBOL system documentation on interfacing.
If the called program possesses
the intial attribute, it and each program directly or indirectly contained
within it, is placed into its initial state every time the called program is
called in a run unit. Whether a called program is in the initial state or
last used state is sensitive to the DYNAM Compiler directive.
Files associated with a called
program's internal file connectors are not in the open mode when the program
is in an initial state. See the section Intial State Of A Program
earlier in this chapter.
On all other entries into the
called program, the states and positioning of all such files is the same as
when the called program was last exited.
The process of calling a
program or exiting from a called program does not alter the status or
positioning of a file associated with any external file connector.
or ENTRY statement
of the called program, in which case the number of operands in each USING phrase must be identical.
The number of operands need
not be identical.
If the program being called is
not a COBOL program, the USING phrase is included in the CALL statement only
if one or more parameters are declared for the called program, in which case
the number of operands in the USING phrase must be identical to the number
of parameters in the called program. (Some implementations of languages
other than COBOL may allow the number of operands to be other than
identical.) The COBOL language places no restrictions on the alignment of
the addresses of data items, whereas non-COBOL languages typically do make
assumptions about addresses and will fail in some way if a misaligned data
item is referenced. Alignment can be achieved by one or more of the
following actions:
or CHAIN
statement and in the corresponding USING phrase in the called program's PROCEDURE DIVISION header
or ENTRY statement or
parameter list if the called program is not COBOL
determines the correspondence between the data names used by the calling and called programs. This correspondence is positional and not by name equivalence; the first operand in the CALL
or CHAIN
statement's USING phrase corresponds to the first operand in the called program's USING phrase or parameter list, the second to the second, etc.
If the correspondence cannot
be completed because the numbers of operands are different, then if the
remaining unmatched operands are in the CALL or CHAIN statement they are
ignored and if they are in the PROCEDURE DIVISION header or ENTRY statement
or parameter list then they must not be referenced within the called
program. The positional correspondence may vary depending on the calling
conventions implied by mnemonic-name.
In the case of index-name, no such correspondence is established. Index-names in the called and calling programs always refer to separate indices.
* The values of the parameters
referenced in the USING phrase of the CALL statement are made available to
the called program at the time the call statement is executed.
* The BY CONTENT, BY REFERENCE
and BY VALUE
phrases are transitive across
the parameters which follow them until another BY CONTENT, BY REFERENCE
or BY VALUE
phrase is encountered.
If no BY CONTENT, BY REFERENCE
or BY VALUE
is specified prior to the
first parameter, the BY REFERENCE phrase is assumed.
either specified or
implied for a parameter, the program operates as if the corresponding data item in the called program occupies the same storage area as the data item in the calling program. The description of the data item in the called program must describe the same number of character positions as described by the description of the corresponding data item in the calling program.
If the BY REFERENCE ADDRESS
OF phrase is specified or implied then the program operates as if an
additional data item had been declared with USAGE POINTER and that data
item passed BY REFERENCE with a value acquired by a SET data item TO
ADDRESS OF identifier-3 statement.
If identifier-3 is in the
Linkage Section and has a level number other than 01 or 77 or is in the
Working-Storage Section, it is equivalent to passing the item BY CONTENT
and the address of identifier-3 cannot be modified by the called
subprogram.
If the BY REFERENCE literal-2
phrase is specified or implied then the object program processes literal-2
as described for literal-3.
* If the BY CONTENT phrase is
specified or implied for a parameter, the called program cannot change the
value of this parameter as referenced in the CALL statement's USING phrase,
though the called program can change the value of the data item referenced
by the corresponding data-name in the called program's PROCEDURE DIVISION
header. The data description of each parameter in the BY CONTENT phrase of
the CALL statement must be the same, meaning no conversion or extension or
truncation, as the data description of the corresponding parameter in the
USING phrase of the PROCEDURE DIVISION header.
If the BY CONTENT phrase is
specified or implied for a parameter, the object program operates as if an
additional data item had been declared and this additional data item used
as the parameter in a BY REFERENCE phrase. If identifier-4 is specified
then both the implied data description of the additional data item and its
contents are identical to that of identifier-4. If literal-3 is specified
then the implied data description of the additional item is equivalent to
an alphanumeric data item with the same size as literal-3 and with its
contents set to the value of literal-3. If LENGTH OF identifier-5 is
specified then the data description of the additional item is equivalent to
PIC S9 (9) USAGE BINARY and its contents set to the number of bytes of
storage allocated to identifier-5.
* If the BY VALUE phrase is
specified or implied for a parameter, the called program cannot change the
value of this parameter as referenced in the CALL statement's USING phrase,
though the called program can change the value of the data item referenced
by the corresponding data-name in the called program's PROCEDURE DIVISION
header. The object program operates, conceptually, as if an additional data
item had been declared in a system area (typically "the stack"),
available to non-COBOL languages for passing parameters and that this
additional data item occupies the same storage area as the data item in the
called program. If identifier-6 is specified then both the implied data
description of the additional data item and its contents are identical to
that of identifier-6. If integer-1 is specified then the implied data
description of the additional item is equivalent to a signed numeric item
USAGE COMP-5 that occupies the number of bytes of storage specified by the
value of integer-2, if specified, or 4 bytes otherwise. If LENGTH OF
identifier-7 is specified then the data description of the additional item
is equivalent to PIC S9 (9) USAGE BINARY and its contents set to the number
of bytes of storage allocated to identifier-7.
Each parameter in the BY VALUE
phrase of the CALL statement has a conceptual additional data item and a
corresponding parameter declared in the USING phrase of the PROCEDURE
DIVISION header in the called program. The data description of each such
parameter in the called program must be the same, meaning no conversion or
extension or truncation, as the data description of the corresponding
conceptual additional data item. In addition, the implied size of the
conceptual data item must not exceed the maximum size of the system area
(typically 4 bytes) otherwise the system may become catastrophically
corrupt.
If the program being called is
a COBOL program, each parameter in the BY VALUE phrase of the CALL
statement must have a corresponding parameter in the USING phrase of the
PROCEDURE DIVISION header that also has the BY VALUE phrase specified or
implied.
If the program being called is
non-COBOL, see your COBOL system documentation on interfacing for details
of when you need to use the BY VALUE phrase.
A called program may pass data
to the calling program either by using a parameter declared in the USING
phrase of the calling program or by means of a returned value. A called
program returns a value by means of an EXIT GIVING or GOBACK GIVING
statement. As control is returned from the called program to the calling
program, the returned value is held in a system area generally available for
non-COBOL programs to return a value.
The size of the system area
(typically at least 4 bytes) is determined outside the COBOL system and
limits the maximum value the return value can take, but the return value
may not utilize the full size of the system area. The returned value is
available to the calling program after control is returned, either
implicitly in the special register RETURN-CODE or explicitly in the data
item specified in the GIVING phrase.
If the GIVING phrase is not
specified and the calling convention specifies updating the RETURN-CODE
special register (see Calling Convention in the earlier section Special-Names
Paragraph) then the object program operates as if the system area were
declared as a COBOL numeric data item with USAGE COMP-5 and the same size as
the system area and as if a MOVE statement had been executed with the system
area as the sending item and the RETURN-CODE special register as the
receiving item. (See the section Special Registers in the chapter
Concepts of the COBOL Language for details of RETURN-CODE.)
If the GIVING INTO phrase is
specified in the calling program, identifier-8 must describe the same number
of character positions as is required to hold the return value in the system
area and must be of the type and usage that is implied for the returned
value by the manner in which the value was returned from the called program.
When control returns to the calling program, identifier-8 contains the
return value.
If the GIVING ADDRESS OF phrase
is specified in the calling program, the called program must return a value
with explicit or implicit USAGE POINTER. The object program operates as if
the system area were declared as a COBOL data item with USAGE POINTER and as
if a SET statement had been executed with the system area as the first
operand and TO ADDRESS OF identifier-9 as the second operand.
It is an error to specify the
GIVING phrase for a called program that does not return a value.
if the Local-Storage Section
has not been declared,
a called program may not execute a CALL that directly or indirectly calls the calling program. If a CALL statement is executed within the range of a declarative, that CALL statement cannot directly or indirectly reference any called program to which control has been transferred and which has not completed execution.
The END-CALL phrase delimits
the scope of the CALL statement.
Any values with implicit or
explicit USAGE POINTER
or USAGE PROCEDURE-POINTER
that were originally derived
from the address of a data item
or procedure
that is contained within a
called program become invalid and must not be used after a CANCEL statement
that references that called program is executed.
The OMITTED phrase is used to
pass unused parameters to a subprogram without having to declare a dummy
data item.
The OMITTED parameter is a BY
REFERENCE parameter, equivalent to passing a null address to the calling
subprogram. It is identical to BY VALUE 0 SIZE 4 but is subject to the type
checking of call prototypes as a BY REFERENCE parameter.
If BY REFERENCE does not
immediately precede OMITTED, subsequent parameters are passed by the
convention previously in force and the omitted parameter is treated as if
the program had specified BY VALUE 0 SIZE 4. The CANCEL statement ensures that the next time the referenced program is called it will be in its initial state.
If the name contains the entry-name contained in an ENTRY statement, the results are undefined.
When an explicit or implicit
CANCEL statement is executed, all programs contained in the program
referenced by the CANCEL statement are also canceled. The result is the same
as if a valid CANCEL statement were executed for each contained program in
the reverse order in which the programs appear in the separate program.
or GOBACK
statement.
or by execution of an EXIT
PROGRAM statement in a called program that possesses the initial attribute.
The contents of data items in
external data records described by a program are not changed when that
program is canceled.
During execution of an explicit
or implicit CANCEL statement, an implicit CLOSE statement without any
optional phrases is executed for each file in the open mode that is
associated with an internal file connector in the program named in the
CANCEL statement. Any USE procedures associated with any of these files are
not executed.
The CHAIN statement transfers control from one object program to another in the run unit with no subsequent return of control. The effect is as though the chained program is the main program in the new run unit.
The chained program must be a COBOL program that is not a nested program. The name of the chained program must contain the name contained in the PROGRAM-ID paragraph of the chained program (see The Program-ID Paragraph earlier in this chapter particularly in regard to case sensitivity). Alternatively, the name of the chained program may contain a name which identifies a file that contains executable program code for the chained program. See your COBOL system documentation for details of how your system matches the given name to previously chained programs, program filenames or programs held in library files.
The CLOSE statement terminates the processing of reels/units and files, with optional rewind and/or lock or removal where applicable.
Line Sequential,
The
REEL and
UNIT phrases should be specified only for files which have had
MULTIPLE REEL or
MULTIPLE UNIT specified in their SELECT clause. If your run-time system
does not recognize multi-unit files, the statements CLOSE REEL and CLOSE
UNIT are "null" statements; that is, they are documentary only. It
is important that no other files are open on the device(s) closed by a CLOSE
REEL or CLOSE UNIT statement.
The statements
CLOSE REEL WITH LOCK and
CLOSE UNIT WITH LOCK are treated as equivalent to
CLOSE REEL FOR REMOVAL.
The
DISP option is applicable only to tape files.
It is documentary only.
If a file is in the open mode
when a STOP RUN statement is executed, the file will be closed. If a file
has been opened in a called program and not closed in that program prior to
the execution of a CANCEL statement for the program, that file will be
closed.
Following the successful
execution of a CLOSE statement, all
record or
file locks held by the run unit on the closed file are released.
Table 12-1: Relationship of Categories of Files and the Formats of the CLOSE Statement
| CLOSE Statement Format | File Category | |||
|---|---|---|---|---|
| Non-Reel/Unit | Record Sequential Single-Reel/ Unit | Record Sequential Multi-Reel/ Unit | Non-record Sequential Single/ Multi-Reel/ Unit | |
| CLOSE | C | C, G | A, C, G | C |
| CLOSE WITH LOCK | C, E | C, E, G | A, C, E, G | C, E |
| CLOSE WITH NO REWIND record sequential only |
C, H |
B, C | A, B, C | X |
| CLOSE REEL/UNIT record sequential only |
F |
F, G |
F, G | X |
| CLOSE REEL/UNIT FOR REMOVAL record sequential only |
F |
D, F, G |
D, F , G | X |
| CLOSE REEL/UNIT WITH NO REWIND record sequential only | X | X | F, B | X |
The definitions of the symbols in the table are given below. Where the definition depends on whether the file is an input, output or input- output file, alternate definitions are given; otherwise, a definition applies to input, output, and input-output files.
Input Files and Input-Output Files:
All reels/units in the file prior to the current reel/unit are processed except those reels/units controlled by a prior CLOSE REEL/UNIT statement. If the current reel/unit is not the last in the file, the reels/units in the file following the current one are not processed.
All reels/units in the file prior to the current reel/unit are processed except those reels/units controlled by a prior CLOSE REEL/UNIT statement.
The current reel/unit is left in its current position.
Input Files and Input-Output Files (Sequential Access Mode):
If the file is at the end and label records are specified for the file, the labels are processed according to the operating system label convention. The behavior of the CLOSE statement when label records are specified but not present, or when label records are not specified but are present, is undefined. If the file is at the end and label records are not specified for the file, label processing does not take place, but other closing operations dependent on the run-time system are executed. If the file is not at the end, the closing operations dependent on the RTS are executed, but there is no ending label processing.
Input Files and Input-output Files (Random or Dynamic Access Mode); Output Files (Random, Dynamic or Sequential Access Mode):
If label records are specified for the file, the labels are processed according to the operating system standard label convention. The behavior of the CLOSE statement when label records are specified but not present, or when label records are not specified but are present, is undefined. If label records are not specified for the file, label processing does not take place, but other closing operations dependent on the run-time system are executed.
The reel or unit can be accessed again, in the proper order of reels or units in the file, if a CLOSE statement without the REEL or UNIT phrase has previously been executed for this file followed by the execution of an OPEN statement for the file.
This file cannot be opened again during this execution of this run unit.
Input Files and Input-Output Files (Reel/Unit Media):
If the current reel/unit is
the last or only reel/unit for the file, no reel/unit swap takes place;
the current volume pointer remains unchanged, and the file position
indicator is set to indicate that no next reel/unit exists.
Output Files (Reel/Unit Media):
The following operations take place;
Input Files, Input-Output Files, and Output Files (Non-Reel/Unit Media):
Execution of this statement
is considered successful. The file remains in the open mode, the file
position indicator remains unchanged, and no action takes place except as
specified in General Rule 4.
The current reel or analogous device is positioned at its physical beginning.
The CLOSE statement is
executed as if none of the optional phrases is present.
This is an illegal combination of a CLOSE option and a file category. The results at object time are undefined.
Except where otherwise stated in the general rules below, the terms "reel" and "unit" are equivalent and completely interchangeable in the CLOSE statement. Treatment of sequential mass storage files is logically equivalent to the treatment of a file on tape or analogous sequential media.
(This option has no connection
with the record or file locking used when sharing files.)
other than a delete file
can be executed that references that file, either explicitly or implicitly, unless an intervening OPEN statement for that file is executed.
The COMMIT statement releases all record locks in all files held by this run unit. For COBOL systems that support the WITH...ROLLBACK clause of the SELECT statement as other than documentary, COMMIT indicates the end of the current transaction and makes the effects of that transaction permanent.
The COMPUTE statement assigns to one or more data items the value of an arithmetic expression.
or a floating-point item.
"Equal" and "="
are synonymous.
The intermediate results
obtained during the expression evaluation are truncated as if they were
moved to data items whose PICTURE is determined by the COBOL system. This
behavior is selected using the ARITHMETIC Compiler directive.
The CONTINUE statement is a no operation statement, indicating that no executable statement is present.
The DEL ETE statement logically removes a record from a mass storage file. It can be specified only for files with relative or indexed organization.
This rule is not enforced.
and NOT INVALID KEY phrase
in the DELETE statement. (See the section The INVALID KEY Condition in this chapter.)
The END-DELETE phrase delimits
the scope of the DELETE statement. (See the section Explicit And
Implicit Scope Terminators in the chapter Concepts of the COBOL
Language.
When using DELETE, the
record to be deleted must not be locked by another run unit.
Following the successful
execution of a DELETE statement, any record lock held by the run unit on the
deleted record is released.
The DELETE FILE statement physically removes the specified files from the physical devices on which they reside.
The DISPLAY statement causes data to be transferred from specified data items to the appropriate hardware device, such as the CRT screen.
The DISPLAY statement also
transfers data from your program to the CRT or video terminal screen:
non-scrolling forms, into which you can enter data, are displayed.
Alternatively, function-name
can itself be used instead of an associated mnemonic-name.
This restriction can be
ignored.
When mnemonic-name is
associated with ARGUMENT-NUMBER, then identifier-1 or literal-1 must each,
if used, reference a data item defined as an unsigned integer or an unsigned
integer literal respectively. Only a single identifier-1 or literal-1 can be
used and the WITH NO ADVANCING phrase can not be used.
When mnemonic-name is associated with ENVIRONMENT-NAME or
ENVIRONMENT-VALUE, then identifier-1 or literal-1 must each, if used,
reference an alphanumeric data item or a nonnumeric literal respectively.
Only a single identifier-1 or literal-1 can be used and the WITH NO
ADVANCING phrase can not be used.
Screen-name cannot be an item
with an OCCURS clause. The LINE and COLUMN phrases can
appear in any order.
Integer-3 and identifier-4 must
be 4 or 6 digits long. A DISPLAY statement with an
operand that is not a screen-name is treated as a Format 3 DISPLAY statement
if it has an AT phrase, an UPON phrase with the CRT or CRT-UNDER option, a
WITH phrase, or a MODE IS BLOCK phrase; or if it has no UPON phrase but the
CONSOLE IS CRT clause is specified in the SPECIAL-NAMES paragraph. If it has
the UPON phrase with the CONSOLE option, or if it has no UPON phrase and the
CONSOLE IS CRT clause is not specified in the SPECIAL-NAMES paragraph, it is
treated as a Format 1 DISPLAY statement.
The phrases following the
identifier can be in any order.
Elementary data items in
identifier-1 must be of USAGE DISPLAY.
No elementary item in
identifier-1 may be longer than 8191 bytes. If the MODE IS BLOCK phrase is
used, the whole of identifier-1 must be no longer than 8191 bytes.
If a DISPLAY verb is followed
by more than one identifier-1, any WITH clause applies only to the operand
immediately preceding it. If the function-name
COMMAND-LINE, or a mnemonic-name associated with the function-name
COMMAND LINE, is specified, data overwrites the contents of a
system-dependent command-line buffer from where it can be retrieved by
subsequent use of ACCEPT FROM COMMAND-LINE. Only one operand is allowed in
this case.
All permissible function-names
other than those specifically mentioned in the general rules for Format 1
of the DISPLAY statement are treated as equivalent to CONSOLE and each
operand is transferred to the console device in the order listed. The total
size of data displayed is equal to the sum of the sizes of each operand.
The display starts from the current cursor position, overflowing onto the
following line(s) if necessary.
On earlier releases, trailing
spaces after the last operand were not displayed.
If the NO ADVANCING phrase is
specified, the cursor is then left at the space following the last character
displayed; otherwise it is positioned at the start of the next line.
Scrolling can take place whenever the cursor is moved to a new line.
If the mnemonic-name
associated with the function-name ARGUMENT-NUMBER is used, then the position
for any subsequent ACCEPT with a mnemonic associated with ARGUMENT-VALUE is
set to retrieve the designated command-line argument. If the value
associated with identifier-1 or literal-1 is less than 0, greater than 99,
or greater than the total number of arguments on the command-line, then
results are explicitly undefined.
If the mnemonic-name
associated with the function-name ENVIRONMENT-NAME is used then the variable
to be read or set in any subsequent ACCEPT or DISPLAY with a mnemonic
associated with ENVIRONMENT-VALUE is set to the designated variable-name
specified in literal-1 or as the contents of identifier-1. The ON EXCEPTION
phrase is ignored. If any problem occurs while trying to set an
ENVIRONMENT-NAME for a subsequent ACCEPT or DISPLAY associated with
ENVIRONMENT-VALUE, then it is only the subsequent ACCEPT or DISPLAY that
must detect the ON EXCEPTION condition.
If the mnemonic-name
associated with the function-name ENVIRONMENT-VALUE is used then
If the mnemonic-name
associated with the function-name SYSERR is used, then a DISPLAY is
performed as if to the CONSOLE except that all output is directed to the
system's standard Error Device.
When using DISPLAY UPON ENVIRONMENT-VALUE, trailing spaces are included in
the value. You should use INSPECT REPLACING ALL SPACE BY LOW-VALUES to avoid
this.
Within identifier-1, internal
floating-point numbers are converted to external floating-point numbers for
display, so that:
Literal-1 may be a
floating-point literal. This format of the DISPLAY
statement displays screen items, which are defined in the screen section of
the program, and allows full access to the enhanced screen handling
facilities. The AT phrase gives the
absolute address on the screen where the DISPLAY operation is to start.
The order of execution of a
DISPLAY statement is always:
If integer-3 or identifier-4 is
4 digits long, the first two digits specify the line, the second two the
column. If 6 digits long, the first three digits specify the line, while the
second three specify the column.
Certain combinations of line
and column numbers have special meanings, as follows:
Part of this statement can be
repeated to allow the display of several data items. If the first item has
no AT phrase, it begins at line 1, column 1. Each subsequent item begins at
the current cursor position as left after the previous item.
If identifier-1 is a group item
and no MODE IS BLOCK phrase exists, those elementary subordinate items which
have names other than FILLER are displayed. They are displayed
simultaneously, positioned on the screen in the order their descriptions
appear in the Data Division, and separated by the lengths of the FILLER
items in the group. For this purpose, the first position on a line is
regarded as immediately following the last position on the previous line.
The MODE IS BLOCK phrase
indicates that the identifier is to be treated as an elementary item. Thus,
even if it is a group item it is displayed as one item.
The WITH phrase allows you to
specify certain options available during the operation. (See the section
The Screen Section earlier in this chapter for descriptions of these
clauses.)
In addition to the options
available as screen description clauses, the following options can be used
in the WITH phrase; SPACE-FILL, ZERO-FILL, LEFT-JUSTIFY, RIGHT-JUSTIFY
TRAILING-SIGN and UPDATE. ZERO-FILL appears in this list and as a screen
description clause because it has two different uses. Its second use is
documented later in this chapter.
A configuration option is
available which allows the entry of data into numeric and numeric edited
screen fields in free format mode. In COBOL, nonedited numeric data items
are intended for holding data in an internal form; however, this format
enables such data items to appear on the screen. See your COBOL system
documentation on user interfaces for more details. If free format mode is
in effect, the data will appear automatically reformatted as follows:
The SPACE-FILL, ZERO-FILL,
LEFT-JUSTIFY, RIGHT-JUSTIFY and TRAILING-SIGN options amend this format.
If the literal-1 is one of the
following figurative constants it has a special effect as follows: SPACE
clears from the specified cursor position to the end of the screen;
LOW-VALUE moves the cursor to the specified position; ALL X"01"
clears from the specified cursor position to the end of the line; ALL X"02"
clears the whole screen; ALL X"07" sounds the bell. If the literal
is a figurative constant that is not listed above and the SIZE option is not
specified, one occurrence of its value is displayed.
If the SIZE option is specified
for a figurative constant that has no special effect, then that figurative
constant is displayed as many times as necessary to reach the length
specified in the SIZE option. However, if the display wraps around to a new
line, it starts again at the beginning of the figurative constant.
If the FOREGROUND-COLOR option
is specified, then the specified color becomes the default foreground color
if and only if the entire screen is cleared by this display
.
The entire screen is cleared either when the BLANK SCREEN option is specified or when literal-1 is SPACES and the DISPLAY starts at line 1 column 1.
If the BACKGROUND-COLOR option
is specified, then the specified color becomes the default background color
if and only if the entire screen is cleared by this display.
If REDEFINES is used within
identifier-1, the first description of the redefined data area is used and
subsequent descriptions are ignored. If OCCURS or nested OCCURS are used the
repeated data item is expanded into the full number of times it occurs, so
that one definition is repeated for many fields. The DIVIDE statement divides one numeric data item into others and sets the values of data items equal to the qu otient and re mainder.
When Format 1, 2, or 3 is
used, floating-point data items and literals can be used anywhere a numeric
data item or literal can be specified. When Format 4 or 5 is used, no
floating-point data items or literals may be specified.
Copyright © 1999 MERANT International Limited. All rights reserved.
This document and the proprietary marks and names
used herein are protected by international law.
| Procedure Division - Intrinsic Functions | Procedure Division - ENTER - INSPECT | |