IBM Mainframe Forum

Hi,

U can take an empty file and compare this empty file with your input file in the first step of a JOB, If your Input file is empty, Then the RC wub be 28, if so, u can abend the JOB. if not, go to the next step in which your prg is executed to read the non-empty files.

P010 EXEC PGM=ISRSUPC,PARM=('LINECMP,OVSUML,UPDCNTL')
NEWDD DD DSN=InputFile,DISP=SHR
OLDDD DD DSN=An Empty File,Disp=shr
OUTDD DD SYSOUT=*
DELDD DD DUMMY
//SYSIN DD *
CMPCOLM 1:80
/*
***********************************************************
* RETURN CODE CHECK
***********************************************************
IF P010.RC not = 28 THEN
....

Hello,

I found solution for my qn,,

The posted process is Not a solution - there are several ways it can fail, making it unacceptable to most organizations.

U can take an empty file and compare this empty file with your input file in the first step of a JOB,

No. Did you not read the "requirement"? The file is not to be opened. . .

Personally, i believe the entire approach is a design flaw and some other approach should be investigated.

Hi Biily,

ya thats true i could have been more clear in defining my qn earlier.. so i briefed once i found the solution..

And to add to your prev comment, if u chk my code in prev comment after the first read, first dataset is read and wtever is the record is copied to ws-1st4data and similarly after second read ws-2nd4data will hv 2nd dataset records. I have tested and its working..

Since i'm working on enhancement proj , i couldn't change the concatenated DD step in JCL.

Thanks for all the support from all.

So how do you know for sure that the second read reads from the second file and not the first? If, by accident, the first file has 2 records in it then the second file does not get read - at least, not at this stage.

If it is an enhancement project then you can enhance the JCL by removing the concatenation.

And the words are "yes", "question", "previous", "you", "have", "project" - you seem to have no idea how annoying it is to have these non-standard abbreviations in a post. You have a full keyboard - use it, please.

I'm late to the party so I might miss the actual requirement, however, with what I've understood - may be you can use two dummy files in the existing concatnation like this:
where dummy.dataset.one and ... .two might have some indicator to show the end of respective files. E.g.: dummy.dataset.one might have @@@@ (or EOF1), in program when the first-four-charcters are these you know the program is at the end of the file.