Page 1 of 1

CFF Output Files with Occurs

Posted: Thu Jun 21, 2018 3:45 pm
by Griffin07
Hi, we have a requirement to create a CFF files that needs to go up to the MF, and in the file there needs to be an ODO. We are able to read MF files with ODO's on the inputs but now we need to create them on the output. I had read that this is not supported in the CFF stage. Does anyone know if this can be done either thru the CFF stage or another way?

thanks

Posted: Thu Jun 21, 2018 6:24 pm
by ray.wurlod
Parallel job (as marked) or server job (as posted)?

Posted: Thu Jun 21, 2018 6:39 pm
by chulett
Based on previous postings, let's go with Parallel. :wink:

Moved to the PX forum, easy enough to move back if needed.

Posted: Fri Jun 22, 2018 2:49 am
by Griffin07
yes, parallel jobs

Posted: Fri Jun 22, 2018 5:52 am
by ArndW
As you've found, writing to an occurs-depending-on using CFF isn't supported. While you could use a transform stage to "build" your variable-length Output record or perhaps use the make-vector stage, I think it might be easiest to define a fixed-occurrence OCCURS and fill the unused elements with a character value that cannot appear the file, then post-process the line (using sed or perl or awk or tr) and removing all occurrences of that character. That will effectively give you the same Thing as you would have gotten with a real occurs-depending-on.

Posted: Fri Jun 22, 2018 6:07 am
by Griffin07
Hi ArndW, the sed/pearl option sounds promising, we would need to run it on the file after it comes out of the CFF stage and has been converted to EBCDIC?