Re: Re: MVT COBOL under MVS3.8j.
| Newsgroups | gmane.comp.emulators.hercules390.mvs |
|---|---|
| Message-ID | <CAPcd4G_7UafkT72JjBa5w5LBAMzZiV0ZvDp+kp0OqYhkss_uPw@mail.gmail.com> |
Remember Peppe, this is '68 Cobol, not '74 Cobol.
You use an occurs with the depending on in the data record.
FD BLAH BLAH BLAH
DATA RECORD IS DATA-RECORD
LABEL RECORDS ARE STANDARD.
01 DATA-RECORD.
02 FIELD-1 PIC 99.
02 FIELD-2 PIC X OCCURS 1 TO 99 TIMES
DEPENDING ON FIELD-1.
Joe
On Mon, Sep 16, 2019 at 4:19 AM Giuseppe Vitillaro giuseppe-yuD/ahkh7LvrZ44/[email protected]
[H390-MVS] <[email protected]> wrote:
>
>
> A side question about this compiler, not related
> with the XREF problem.
>
> I'm reading the IBM manuals about variable length
> datasets.
>
> I can't see a way to read a variable length record
> and retrieve the length of the record at runtime,
> like what is done in modern implementation of
> mainframe COBOL compilers with a syntax like this:
>
> FD VREDDV01-FILE
> DATA RECORD IS VREDDV01-REC
> RECORDING MODE IS V
> RECORD VARYING FROM 00001 TO 00080
> DEPENDING ON VREDDV01-LRECL,
>
> The GC28-6399-2 manual claims , page 136:
>
> Variable-length record descriptions, for
> input and output files, must not define
> space for the control bytes. Control bytes
> are automatically provided when a record is
> written and are not communicated to the
> use when a file is read.
>
> So, my question.
>
> How were variable length records handled
> at the time this compiler was in use?
>
> Peppe.
>
>