Back in the mists of time - so I can say it now :-) - Frank Yaeger and I were discussing how we might add some things to the first offering of OUTFIL in DFSORT that weren't then in Syncsort. We came up with SAVE and "Card Dealer".
SAVE has, to my mind, been most invaluable. Prior to this you had to figure out the boolean complement of the INCLUDE= clause. With more than one INCLUDE= clause in a sort it became hairier, requiring De Morgan's Law*. :-) SAVE made this a lot simpler and less error prone. "Card Dealer" not so much. I would expect both implementations - DFSORT and Syncsort - to have had these functions since the 1990s and both have gone much further. On the "tagging" that Sri Hari mentioned: 1) You needn't tag unless you want to - and there will be some occasions where you don't want to. One example might be if you've been counting with SUM FIELDS= and a binary "ONE" field. You might throw away records where the count is less than, say, 100. 2) One other thing you might want to do is to throw away, say, 90% of the records on the way in with INCLUDE/OMIT and then divide up the surviving records on the way out, using OUTFIL INCLUDE=/OMIT= and SAVE. Just some thoughts from someone who's loved playing with what DFSORT can do but has never actually been a DFSORT developer. * Augustus De Morgan founded the Maths Department at UCL, where I learnt a trick or two. :-) Cheers, Martin Martin Packer Systems Investigator & Performance Troubleshooter, IBM +44-7802-245-584 email: martin_pac...@uk.ibm.com Twitter / Facebook IDs: MartinPacker Blog: https://mainframeperformancetopics.com Mainframe, Performance, Topics Podcast Series (With Marna Walle): https://anchor.fm/marna-walle Youtube channel: https://www.youtube.com/channel/UCu_65HaYgksbF6Q8SQ4oOvA From: Sri h Kolusu <skol...@us.ibm.com> To: IBM-MAIN@LISTSERV.UA.EDU Date: 22/09/2020 19:24 Subject: [EXTERNAL] Re: SORT question: How to both remove records from a file AND write removed records to a 2nd output? Sent by: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> > In one pass of an input file is there a way to use SORT to both > remove identified records from an input file AND write those removed > records to a second output file? Peter, You don't need 2 passes of data, you can tag the records that match your identification and then use that tag to filter the records on OUTFIL. Some thing like this //STEP0100 EXEC PGM=SORT //SYSOUT DD SYSOUT=* //SORTIN DD * ABC - DROP THIS RECORD DEF - PICK THIS RECORD EFG - PICK THIS RECORD XYZ - DROP THIS RECORD ZZZ - DROP THIS RECORD //PICKREC DD SYSOUT=* //DROPREC DD SYSOUT=* //SYSIN DD * OPTION COPY INREC IFTHEN=(WHEN=(1,3,SS,EQ,C'DEF,EFG'), OVERLAY=(81:C'A')) OUTFIL FNAMES=PICKREC, INCLUDE=(81,1,CH,EQ,C'A'), BUILD=(1,80) OUTFIL FNAMES=DROPREC,SAVE, BUILD=(1,80) /* The PICKREC file will have the following records DEF - PICK THIS RECORD EFG - PICK THIS RECORD The DROPREC file will have the following records ABC - DROP THIS RECORD XYZ - DROP THIS RECORD ZZZ - DROP THIS RECORD For VB files, make sure that you have the Indicator right after the RDW so that you retain the variable length records as is Something like this //SYSIN DD * OPTION COPY INREC IFTHEN=(WHEN=INIT, BUILD=(1,4, $ RDW X, $ ID TAG 5)), $ DATA AS-IS IFTHEN=(WHEN=(6,3,SS,EQ,C'DEF,EFG'), OVERLAY=(05:C'A')) OUTFIL FNAMES=PICKREC, INCLUDE=(05,1,CH,EQ,C'A'), BUILD=(1,04, $ RDW 6) $ DATA AS-IS OUTFIL FNAMES=DROPREC,SAVE, BUILD=(1,04, $ RDW 6) $ DATA AS-IS /* Fruther if you have any questions please let me know Thanks, Kolusu DFSORT Development IBM Corporation ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
