|
View previous topic :: View next topic
|
| Author |
Message |
Roland_Achtsnit
New User
Joined: 30 Jul 2006
Posts: 14
|
|
|
|
Hello
I have the following input file (FB/80)
| Code: |
=COLS> ----+----1----+----2----+----3----+----4----+----5----+----6----+----7--
****** ***************************** Top of Data ******************************
000001 $AA.AA.CLIST
000002 $BB.B.CLIST
000003 $CC.C.CNTL
000004 ADD.D.DD.DDD.DDDD
000005 AEE.E.EE.EEEEEEE
000006 AFF.F.FF.FFFFFFF.FFFF
000007 PGG.G.GGGG.GGGG.G0001V00
000008 PHH.H.HH.HHHH.HHHH
000009 PII.I.II.IIII.IIII.IIIIIII.IIII.G0016V00
000010 PII.I.II.IIII.IIII.IIIIIII.IIII.G0017V00
000011 PII.I.II.IIII.IIII.IIIIIII.IIII.G0018V00
000012 PJJ.JJ.J.JJJJ.JJJJ.JJJJ.G0055V00
000013 PKK.K.KK.KKKK.KKK
000014 PLL.L.LL.LLLLLL.LLLLL.LLL.G0231V00
000015 PMM.M.MM.MMMMMM.MMM.MMMM.G0007V00
000016 PNN.N.NNN.NNNN.NNN.NNN.NNNNNNNN
000017 POO.O.OOO.OOOO.OOO.OOOOOOO.OOO
000018 PPP.P.PPP.PPPP.PPP.PPPPPPP.PPP.PPPP.G0001V00
000019 PQQ.Q.QQQ.QQQ.QQ.QQQQ.G4561V00
000020 PQQ.Q.QQQ.QQQ.QQQ.QQQQ.G7659V00
000021 PRR.R.RRR.RRR.RR.RRRR.G0023V00
000022 PSS.S.SSS.SSS.SSS.SSSS.G0456V00 |
First, I want only the records whose ending is V00
| Code: |
=COLS> ----+----1----+----2----+----3----+----4----+----5----+----6----+----7--
****** ***************************** Top of Data ******************************
000001 PGG.G.GGGG.GGGG.G0001V00
000002 PII.I.II.IIII.IIII.IIIIIII.IIII.G0016V00
000003 PII.I.II.IIII.IIII.IIIIIII.IIII.G0017V00
000004 PII.I.II.IIII.IIII.IIIIIII.IIII.G0018V00
000005 PJJ.JJ.J.JJJJ.JJJJ.JJJJ.G0055V00
000006 PLL.L.LL.LLLLLL.LLLLL.LLL.G0231V00
000007 PMM.M.MM.MMMMMM.MMM.MMMM.G0007V00
000008 PPP.P.PPP.PPPP.PPP.PPPPPPP.PPP.PPPP.G0001V00
000009 PQQ.Q.QQQ.QQQ.QQ.QQQQ.G4561V00
000010 PQQ.Q.QQQ.QQQ.QQQ.QQQQ.G7659V00
000011 PRR.R.RRR.RRR.RR.RRRR.G0023V00
000012 PSS.S.SSS.SSS.SSS.SSSS.G0456V00 |
Then I want to delete the last column such as ". G%%%% V00. 'The output file should look like this.
| Code: |
=COLS> ----+----1----+----2----+----3----+----4----+----5----+----6----+----7--
****** ***************************** Top of Data ******************************
000001 PGG.G.GGGG.GGGG
000002 PII.I.II.IIII.IIII.IIIIIII.IIII
000003 PII.I.II.IIII.IIII.IIIIIII.IIII
000004 PII.I.II.IIII.IIII.IIIIIII.IIII
000005 PJJ.JJ.J.JJJJ.JJJJ.JJJJ
000006 PLL.L.LL.LLLLLL.LLLLL.LLL
000007 PMM.M.MM.MMMMMM.MMM.MMMM
000008 PPP.P.PPP.PPPP.PPP.PPPPPPP.PPP.PPPP
000009 PQQ.Q.QQQ.QQQ.QQ.QQQQ
000010 PQQ.Q.QQQ.QQQ.QQQ.QQQQ
000011 PRR.R.RRR.RRR.RR.RRRR
000012 PSS.S.SSS.SSS.SSS.SSSS |
Could I do this with DFSORT or Icetool? I hope you can help me with my problem.
regards
Roland |
|
| Back to top |
|
 |
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
You should be able to do a Substring (SS) for "V00 " on an INCLUDE to get only those you want.
Then you'll need PARSE.
You'll only need the maximum that is valid for a DSN...
Then you need to find the first PARSEd item that is blank, and set the previous PARSEd item to space. |
|
| Back to top |
|
|
Roland_Achtsnit
New User
Joined: 30 Jul 2006
Posts: 14
|
|
|
|
Hello
Thank you for your reply. Now I have written the following job and it work's fine. I know the IF=THEN must be built from. G0 to. G9.
| Code: |
//SELECT EXEC PGM=ICEMAN
//SYSOUT DD SYSOUT=*
//SORTIN DD -------- FB/80
//SORTOUT DD -------- FB/80
//SYSIN DD *
INCLUDE COND=(2,44,SS,EQ,C'V00')
INREC IFOUTLEN=80,
IFTHEN=(WHEN=(2,44,SS,EQ,C'.G0'),
PARSE=(%00=(STARTAT=NONBLANK,ENDBEFR=C'.G0',FIXLEN=44)),
BUILD=(2:%00)),
IFTHEN=(WHEN=(2,44,SS,EQ,C'.G1'),
PARSE=(%01=(STARTAT=NONBLANK,ENDBEFR=C'.G1',FIXLEN=44)),
BUILD=(2:%01))
SUM FIELDS=NONE
SORT FIELDS=(2,44,CH,A)
/*
|
Maybe someone had a better idea? |
|
| Back to top |
|
|
enrico-sorichetti
Superior Member
Joined: 14 Mar 2007
Posts: 10873
Location: italy
|
|
| Back to top |
|
|
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
I would make the INCLUDE for "V00 " (one trailing space).
This would allow an "ordinary" DSN which just happened to end with V00 through, but I don't think one of those should be likely, and it is possible to make it more secure.
I still think the PARSE route is better, as you may have a GDG which has a qualifier which starts G<numeric> for some reason (unlikely). I have a thought, but probably can't get at it until this evening.
EDIT: Ah, whilst typing enrico has a nice solution. |
|
| Back to top |
|
|
Roland_Achtsnit
New User
Joined: 30 Jul 2006
Posts: 14
|
|
|
|
Thanks to everyone for their help. This is a very good solution to move the issue of the DSN to the right.
regards
Roland |
|
| Back to top |
|
|
|
|
