Under zOS REXX does have a STEM Sort (via BPXWUNIX). If you have an OMVS Segment on your USERID (and as of zOS 2.1+ you should), try this:
/* Do a Binary Sort (use -tn for Text) Using 2nd Word in each StemIn. Record as primary sort key Using 1st Word in each StemIn. Record as secondary sort key More doc on Unix Sort at http://publibz.boulder.ibm.com/epubs/pdf/bpx1cd10.pdf */ x = BPXWUNIX("/bin/sort -bn -k2,2 -k1,1","StemIn.","StemOut.") Also available via the USS side of zOS is Regex support: /* Get the JESMSGLG output from a JOB, e.g. via SDSF or (E)JES and put in the STEM JesMsgLg. Use the USS grep in, Extended mode, to search for all STEPS with an RC=12, 16, or 20 Return only those steps to the BadSteps. STEM variable ... Add a step using this at the very END of a JOB to determine if something did not go as expected, and then possibly correct it! */ x = BPXWUNIX("/bin/grep -E -e'¬.{30}(.{10} (12|16|20)'", "JesMsgLg.","BadSteps.") ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN