COBOL : How to call nth digit of a field in occurs

riponrai · New User Joined: 15 Jun 2005 Posts: 2 Location: Somerset

hello friends,

i am confused in mapping this field from an input application to my code.
You can consider my receiving field as CONTENT X

05 HELLO OCCURS 25 TIMES.
10 SAYONARA PIC 9 VALUE 0.
10 NAMASTE PIC X(03).

How to map this ??

If NAMASTE (3:1) is 'A' in any occurence then pull corresponding SAYONARA field ??

I'm confused in declaring NAMASTE field

Thanks for any help.

Ripon

Bill Woodger · Posted: Tue Jun 21, 2011 9:02 pm

dbzTHEdinosauer · Posted: Tue Jun 21, 2011 9:12 pm

implementing CONDITION-NAMEs (Level 88)

makes the code easier to understand
removes the need for literals in PROCEDURE DIVISION.

Bill Woodger · Posted: Tue Jun 21, 2011 9:27 pm

I am certainly in agreement, dbz.

I hope no-one is going to show the TS how to subcript/index the reference modification so that they can use a literal.

1) Look it up yourself, so you know how to use it when you need to (not for occasions like this).
2) Use the 88, like dbz points out. If you are not going to reference the PIC X field, you can make it a FILLER so that no-one wonders why the field is not used.

blah, blah, blah - did you look?

dbzTHEdinosauer · Posted: Tue Jun 21, 2011 9:37 pm

yes, i did look; couldn't help myself.
the quote button makes it very easy to read posts,
as well as find out how BBCode works.

and you certainly did not think that I would pass up a chance to respond?

Anuj Dhawan · Posted: Wed Jun 22, 2011 3:26 pm

I'm, on the other hand, thinking about the "varibale names" used to demonstrate the example:

Hello - well, we all know.
SAYONARA - Japanese way of farewell remark.
Namaste - Indian way of saying Hello or good-bye.

Sorry Dick...

dbzTHEdinosauer · Posted: Wed Jun 22, 2011 3:30 pm

Marso · Posted: Wed Jun 22, 2011 7:00 pm

Maybe the TS is confused about using both OCCURS and reference modification.
I opened the COBOL Language Reference, looked for "reference modification". The third example provides a clear answer.

dbzTHEdinosauer · Posted: Wed Jun 22, 2011 7:09 pm

excellent post Marso, as usual.

Bill Woodger · Posted: Wed Jun 22, 2011 7:17 pm

Not that easy... :-)

We don't want him to use the subscripted reference modification in this case, because you can't put an 88-level on it and we don't like literals in the Procedure Division (do we, dbz?). Perfect link for him, though.

dbzTHEdinosauer · Posted: Wed Jun 22, 2011 8:16 pm

ok, Bill, you win. you can have the last post.

Marso · Posted: Wed Jun 22, 2011 8:44 pm

Last post ? hey, wait! I'm not finished!

Technically, I think I got the right answer.
But the "reference modification / 88-level / literal" issue is interesting too, and maybe to solve it there is a way around:

Bill Woodger · Posted: Wed Jun 22, 2011 9:04 pm

Bill Woodger · Posted: Wed Jun 22, 2011 10:48 pm

Have to say I do prefer dbz's solution. All the definition in the Data Division, clear what the individual field is if necessary (else FILLLER, so it is clear the field is only tested).

W-h-a-t d-o y-o-u t-h-i-n-k, dbz?

dbzTHEdinosauer · Posted: Thu Jun 23, 2011 1:01 am

I go out of my way to generate the required data layout in order
to avoid reference modification
in modules that i release.

honestly though, my personal modules (utilities, etc...)
look like assembler - only it is compiled with a cobol compiler.

so, it is one of those do as I say, not what I do