IBM Mainframe Forum

by **ranga_subham** » Sun Dec 21, 2008 2:51 pm

Hi,

I have two files and want the matched records only in the output from both the files :!:

A temporary identification can be applied to both the files to determine the origin from where there come in the output.

INPUT-1:

Select all

----+----1----+----2----+----3----
VX2 GGGVX2 2460
BC2 MMMBC2 2160
GF2 KKKGF2 2460

INPUT-2:

Select all

----+----1----+----2----+----3----
VX2 GGGVX2 2442
VX2 GGGVX2 2442
GF2 KKKGFM 2442
GF2 KKKGFM 2442
GF2 KKKGFM 2442
GF2 KKKGFM 2442
J42 KRTJ42 2442
JK2 KRTJK2 2442

OUTPUT WITH MATCHED RECORDS:

Select all

----+----1----+----2----+----3----+----4
VX2 GGGVX2 2460 1
VX2 GGGVX2 2442 2
VX2 GGGVX2 2442 2
GF2 KKKGF2 2460 1
GF2 KKKGFM 2442 2
GF2 KKKGFM 2442 2
GF2 KKKGFM 2442 2
GF2 KKKGFM 2442 2

Note: Both the files may contain duplicates and are NOT in sorted order.

Please suggest.

Thanks.

by **Alissa Margulies** » Mon Dec 22, 2008 9:18 pm

Is the temporary file identification a requirement? Also, what fields are you matching (1-3,20-25,1-25)?

by **ranga_subham** » Tue Dec 23, 2008 11:42 am

The field to be compared starts from 16th column (10 bytes). Applying an id is not part of requirement but to know from which file the records have come because matching records are to be pulled out from both the files including duplicates.
Thanks.

by **Alissa Margulies** » Fri Jan 02, 2009 11:23 pm

If your records can be pre-processed with a sequence number in addition to the file id, then here is a sample application you can use:

Select all

//STEP1 EXEC PGM=SORT
//SORTOUT DD DISP=(NEW,PASS),DSN=TEMP.DATASET,...
//SORTJNF1 DD *
VX2 GGGVX2 2460 1 1
BC2 MMMBC2 2160 1 2
GF2 KKKGF2 2460 1 3
//SORTJNF2 DD *
VX2 GGGVX2 2442 2 1
VX2 GGGVX2 2442 2 2
GF2 KKKGFM 2442 2 3
GF2 KKKGFM 2442 2 4
GF2 KKKGFM 2442 2 5
GF2 KKKGFM 2442 2 6
J42 KRTJ42 2442 2 7
JK2 KRTJK2 2442 2 8
//SYSOUT DD SYSOUT=*
//SYSIN DD *
JOINKEYS FILE=F1,FIELDS=(1,3,A)
JOINKEYS FILE=F2,FIELDS=(1,3,A)
REFORMAT FIELDS=(F1:1,44,F2:1,44)
SORT FIELDS=COPY
OUTFIL OUTREC=(1:1,44,/,1:45,44)
//*********************************************
//STEP2 EXEC PGM=SORT
//SORTIN DD DISP=SHR,DSN=TEMP.DATASET
//SORTOUT DD DSN=FINAL.OUTPUT,...
//SYSOUT DD SYSOUT=*
//SYSIN DD *
SORT FIELDS=(20,6,CH,A,36,1,CH,A,37,8,CH,A)
DUPKEYS FIELDS=NONE
OUTREC BUILD=(1,36)
/*

NOTE:
Support for DUPKEYS was included in SyncSort for z/OS 1.3. If you are running an earlier release, you can replace DUPKEYS with SUM.

by **ranga_subham** » Sat Jan 03, 2009 4:28 pm

Thank you. We are still at v1.2 only

by **Alissa Margulies** » Mon Jan 05, 2009 9:23 am

Then as I previously mentioned, you can code SUM FIELDS instead of DUPKEYS FIELDS.

IBM Mainframe Forum

Matched records alone need in o/p from both file !

Matched records alone need in o/p from both file !

Re: Matched records alone need in o/p from both file !

Re: Matched records alone need in o/p from both file !

Re: Matched records alone need in o/p from both file !

Re: Matched records alone need in o/p from both file !

Re: Matched records alone need in o/p from both file !