IBM Mainframe Forum

by **Praveen433** » Tue Jan 18, 2011 8:43 pm

Hi ,
Please help me in achieving the following criteria
There are two files with same record layout. The first two fields will act as a key(these two fields are common in both files), and the requirement is to compare the remaining fields of those two similar records. If any field is different then the output file should be written with the key fields followed by the mismatch field Number. The output file should contain records which have mismatch fields.

It looks like this

Select all

File 1:
1 2 3 4 5
Record 1: ABC XYZ SSS AAA BBB
Record 2: PQR CBA EEE OOO CCC
Record 3: EPF ETR UUU JJJ GGG

Select all

File 2:
1 2 3 4 5
Record 1: ABC XYZ SSS PPP BBB
Record 2: PQR CBA EEE OOO CCC
Record 3: EPF ETR MMM JJJ QQQ

Select all

Output File:
Record 1: ABC XYZ 4
Record 2: EPF ETR 3 5

Any change in achieving the requirement is always welcome.
My main requirement is to get the field names which are not matching.

by **NicC** » Wed Jan 19, 2011 2:10 am

Q. Can either file have duplicates? If they can, then what is the criteria?

I sus pect that a sort product may have the ability to do what you want. Certainly you could use a programming language to do it. You CANNOT do it with JCL (although JCL could be used to present your solution to the machine for processing).

by **dick scherrer** » Wed Jan 19, 2011 2:16 am

Hello,

Which sort product is used on your system?

As NicC mentioned, this could easily be done with a bit of your own code.

IBM Mainframe Forum

Comparing two files with unmatching records written to file

Comparing two files with unmatching records written to file

Re: Comparing two files with unmatching records written to f

Re: Comparing two files with unmatching records written to f