Split the i/p record to 2 or 3 using cobol code

alluri12 · Posted: Tue Aug 28, 2007 9:26 pm

Hi,
I have a variable length file input 500 bytes. I need to write that to FB output file 80 bytes.
I have to copy input file based on
1st record 1 to 72 position with 72 position = spaces to the output file.
If 72nd position is not spaces then need to check 71 position if SPACES then write to output file.
Then
From position we wrote (71 or 72) to next 72 position copy to output file if 144 or 143 is spaces.
From position we wrote (143 or 144) to next 72 position copy to output file if 221 or 222 is spaces.
Follow same till end of 1st record(500 bytes). then set a flag if end of 1st record.

once end of record
follow same as above for 2nd record and so on

Bottom line i need to split each input record into different 72 or 71 bytes in the output file USING COBOL program.

Here is the small Example of input file

.) AAA 1234567890 ABCDEFGHIJKLM AAAA AAAAAA BBBBBB CCCC DDD
.) AAA 1234567890 ABCDEFGHIJKLM AAAA AAAAAA BBBBBB CCCC DDD

Output file will be

.) AAA 1234567890
ABCDEFGHIJKLM
AAAA AAAAAA
BBBBBB CCCC
DDD
.) AAA 1234567890
ABCDEFGHIJKLM
AAAA AAAAAA
BBBBBB
CCCC
DDD

Could anybody please help me on this...

Thanks
Suresh

William Thompson · Posted: Tue Aug 28, 2007 10:19 pm

alluri12 · Posted: Tue Aug 28, 2007 10:30 pm

Oh my mistake.... It should be identical. Looks like i pressed enter twice.
.) AAA 1234567890
ABCDEFGHIJKLM
AAAA AAAAAA
BBBBBB CCCC
DDD
.) AAA 1234567890
ABCDEFGHIJKLM
AAAA AAAAAA
BBBBBB CCCC
DDD

dick scherrer · Posted: Tue Aug 28, 2007 11:51 pm

Hello,

What have you coded so far?

I'm a bit confused. Your "rules" say that you want to split after 71 or 72 bytes, but your example splits after far fewer bytes.

Where does the end-of-record flag come into the picture?

If you post what you have so far, and clarify the rules, we will be better able to offer suggestions.

sandy_venkat · New User Joined: 16 May 2007 Posts: 35 Location: India

Dick, i think he is just giving us an example of how the records are split.

dick scherrer · Posted: Wed Aug 29, 2007 7:41 pm

Hello,

Could be, but if so, the "rules" should be restated to match the sample data.

The code will be basically the same, just the break-points will be at different displacements.

Also, what should happen if neither of the break-point positions contains a space? Should a value just be chopped off or should the parser "back up" to the most recent space and write the record?

socker_dad · Posted: Thu Aug 30, 2007 1:38 am

Sounds like a COBOL 101 homework assignment to me.

jayachandra · New User Joined: 08 Mar 2007 Posts: 3 Location: mumbai

Hello,

Try this piece of code
FD INPUT-FILE.
01 WS-FILE.
05 WS-FILE-READ PIC X(500).
05 WS-REDIF-FILE REDEFINES WS-FILE-READ.
10 WS-SPLIT PIC X(71) OCCURS 8 TIMES.
FD OUTPUT-FILE
01 WS-OUT-REC PIC X(80).
.
.
.
MOVE HIGH-VALUES TO WS-REDIF-FILE.
READ INPUT-FILE.

PERFORM ARR-CNT FROM 1 BY 1
UNTIL ARR-CNT > 8
OR END-OF-RECORD
MOVE ZEROES TO WS-CNT
INSPECT WS-SPLIT(ARR-CNT)
TALLYING WS-CNT FOR LEADING HIGH-VALUES
IF WS-CNT = 71
SET END-OF-RECORD TO TRUE
ELSE
MOVE SPACES TO WS-OUT-REC
WRITE WS-OUT-REC FROM WS-SPLIT(ARR-CNT)
END-IF
END-PERFORM.

HOPE THIS MEETS UR REQUIREMENT

dick scherrer · Posted: Thu Aug 30, 2007 6:17 pm

Hello jayachandra,

Please review the requirement first posted and compare it to your suggestion.

Your code does not match "the rules" posted.

star_dhruv2000 · Posted: Mon Feb 18, 2008 12:00 pm

HI,

Please let me know if following worked for you: