IBM Mainframe Forum

by **Eagle** » Thu Nov 04, 2010 8:40 pm

I have two input files, "File 1" and "File 2" that have the same record length and format. I want to use Syncsort to to remove the records that exist on File 1 from File 2 and write all records from File 1 and only the unmatched records from File 2 in one output file. The matched records from File 2 can be discarded.

Questions:
What tool should I use?
Can this be done in one sort step?
If so, can you provide a generic example?
If not, I can easily accomplish this with two steps.

The key is the first 33 bytes of the 500-byte LRECL. Sample data:

Select all

IN 00101ABC00000000000123456789
IN 00101DEF00000000000234567890

by **Eagle** » Thu Nov 04, 2010 10:45 pm

One additional note: Duplicates should not be removed.

by **Alissa Margulies** » Thu Nov 04, 2010 11:47 pm

Here is a sample job that may help you:

Select all

//STEP1 EXEC PGM=SORT
//SORTJNF1 DD DSN=FILE1
//SORTJNF2 DD DSN=FILE2
//SORTOUT DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//SYSIN DD *
JOINKEYS FILES=F1,FIELDS=(1,33,A)
JOINKEYS FILES=F2,FIELDS=(1,33,A)
JOIN UNPAIRED
REFORMAT FIELDS=(F1:1,500,F2:1,500)
SORT FIELDS=COPY
OUTREC IFTHEN=(WHEN=(1,1,CH,EQ,C' '),BUILD=(501,500)),
IFTHEN=(WHEN=NONE,BUILD=(1,500))

by **Eagle** » Fri Nov 05, 2010 12:08 am

Thank you, Alissa!

IBM Mainframe Forum

Syncsort Join

Syncsort Join

Re: Syncsort Join

Re: Syncsort Join

Re: Syncsort Join