IBM Mainframe Forum

Posted: **Fri Nov 12, 2010 10:07 pm**

Is there a way to use Syncsort to produce the following? All records with unique keys go to Output file 1 and all duplicate records plus the respective original record go to Output file 2.

Example:
Input

Select all

0001
0002
0002
0003
0004
0005
0005

Output file 1:

Select all

0001
0003
0004

Output file 2:

Select all

0002
0002
0005
0005

Posted: **Sat Nov 13, 2010 12:08 am**

Here is a SyncSort for z/OS 1.3 job that will produce the requested output:

Select all

//STEP1 EXEC PGM=SORT
//SORTIN DD *
0001
0002
0002
0003
0004
0005
0005
//SORTXDUP DD DSN=&&XDUP,DISP=(NEW,PASS)
//SORTOF01 DD DSN=&&OUT1,DISP=(NEW,PASS)
//SORTOF02 DD DSN=UNIQUE.RECORDS
//SYSOUT DD SYSOUT=*
//SYSIN DD *
INREC OVERLAY=(81:X'F1')
SORT FIELDS=(1,5,CH,A)
DUPKEYS SUM=(81,1,ZD),XDUP
OUTFIL FILES=01,INCLUDE=(81,1,ZD,GT,1)
OUTFIL FILES=02,INCLUDE=(81,1,ZD,EQ,1),BUILD=(1,80)
/*
//STEP2 EXEC PGM=SORT
//SORTIN01 DD DSN=&&XDUP,DISP=SHR
//SORTIN02 DD DSN=&&OUT1,DISP=SHR
//SORTOUT DD DSN=ALLDUPS
//SYSOUT DD SYSOUT=*
//SYSIN DD *
INREC BUILD=(1,80)
MERGE FIELDS=(1,5,CH,A)
/*

Posted: **Sat Nov 13, 2010 12:19 am**

Thanks, again!

Is there a manual for Synctool? Using Synctool, I found the following example on another forum:

Select all

//S1 EXEC PGM=SYNCTOOL
//TOOLMSG DD SYSOUT=*
//DFSMSG DD SYSOUT=*
//IN DD DSN=TEST.XSUM.INPUT,DISP=SHR
//OUT1 DD DSN=TEST.XSUM.OUTPUT.UNIQUE,DISP=SHR
//OUT2 DD DSN=TEST.XSUM.OUTPUT.DUPS,DISP=SHR
//TOOLIN DD *
SELECT FROM(IN) TO(OUT1) ON(1,04,CH) NODUPS
SELECT FROM(IN) TO(OUT2) ON(1,04,CH) ALLDUPS
/*

Posted: **Sat Nov 13, 2010 12:27 am**

Eagle,

Please see the email I sent you offline.

Regards,

IBM Mainframe Forum

Remove Duplicates + Original Record

Remove Duplicates + Original Record

Re: Remove Duplicates + Original Record

Re: Remove Duplicates + Original Record

Re: Remove Duplicates + Original Record