On Sun, 7 Jun 2026 21:54:42 +0000, Robert Prins wrote:
>I need to merge several, in casu, three,121 byte FB records into a new
>longer record, but the first of the three is pseudo-variable length, it
>contains a field in UTF8 format followed by a 0x00 byte, and the data up to
>this 0x00 byte needs to be used as the first part of the newly spliced
>record.
...
Are all your data in all your data sets UTF8 or is there a mixture of CCSIDs?
Do any of your pseudo-variable length records span FB records?
Is this a job for a SORTIN exit?
Can DFSORT take all its input from an exit with no actual data set?
Can DFSORT use an (allocated) POSIX pipe as SORtIN
--
gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
