How to compare two datasets by sending tso commands

naveen_amudalapelly · New User Joined: 15 Oct 2004 Posts: 16

Hi

Could any one please let meknow how to compare two data sets

and Could any one also help me by sending tso commands document

Thanks in advacce

Naveen

superk · Posted: Tue Nov 16, 2004 8:29 pm

Use IBM's SuperC program, which is available under ISPF in the Utilities menu (option 3), option 12 (SuperC Compare data sets).

There are also third-party products available, such as Comparex.

jz1b0c · Posted: Tue Nov 16, 2004 8:35 pm

Hi Naveen,

you can use option 3.13 (SuperCE) to compare two data sets

There you have compare type, select apropriate one,
-------------------------------------------------------------

Compare Type
2 1. File
2. Line
3. Word
4. Byte

if you want mention some conditions for compare type E on the command prompt and enter your crieteria

Let me know if you find any difficulty

naveen_amudalapelly · New User Joined: 15 Oct 2004 Posts: 16

Hi jz1b0c,

This information has given good result for me.

and also could you please tell me how to write the records from first dataset into second data set which is not having the remaining records when we comapare the two datesets.

Thanks in advance

Naveen

honeypria567 · New User Joined: 20 Nov 2006 Posts: 14 Location: chennai

Hi All,

we can compare two dataset by using JCL Utility IEBCOMPR

regards,
Honey

mikayag · New User Joined: 10 Mar 2005 Posts: 16

Hi All,

This compare discussion is great.

I have a list of sequential files that needs to be compared against one file.

e.g.:

XXXXX.YYYYY.ABCD001
to be compared against
XXXXX.YYYYY.ABCD001.BAK

and I need to produce an output file containing records which do not exist
in the first file.

However, thsi can be tedious because I have several sets of files that needs to be compared. Take this set of example below:

XXXXX.YYYYY.ABCD001 / XXXXX.YYYYY.ABCD001.BAK
XXXXX.YYYYY.ABCD002 / XXXXX.YYYYY.ABCD002.BAK

:
:
XXXXX.YYYYY.ABCD999 / XXXXX.YYYYY.ABCD999.BAK

How can I automate this process? I was thinking of creating a program that will process these files in a loop, each time incrementing by 1. However, the JCl needs a definite name for the files we specify.

Any workaround?

dick scherrer · Posted: Thu Mar 01, 2007 2:11 am

Hello,

You can set up a small PROC that compares the common file against a file what has the 001-999 specified at run time. Use a symbolic parameter for the number part of the dataset name.

When you run this, you will still need an EXEC statement for each, but with a little bit of editing, you will have a "whole" set very quickly - after you create the "whole" set, you could remove any that you wanted to skip or did not exist.

The basic EXEC would be:

Pravin Raj · New User Joined: 16 Mar 2007 Posts: 2 Location: chennai

By honeypriya,
we can compare two dataset by using JCL Utility IEBCOMPR

She is right but i jst want to add somemorethings in that:

IEBCOMPR: mainly used to check whether it is PS or PDS.

Syntax in the JCL: COMPARE TYPORG = PS / PDS -> Return code

Note: If Return Code = 0,files are of same type(Either they r Ps or PDS)

If Return Code = 8,files are of different type(May b one Ps and other is PDS)

raak · Posted: Tue Mar 20, 2007 7:09 pm

hey,,

u can use this TSo command also.

Go inside the file(File1) u want to compare in View or Edit mode.

Then give COMP X 'FILE2'

This will show where all the difference lies between file1 and file2.

ofer71 · Posted: Tue Mar 20, 2007 8:00 pm

COMP (or COMPARE) is not a TSO command; it is an ISPF Editor Primary Command.

O.