MVSFORUMS.com

[krk123]

Hi,
Could someone please advise on this. I am getting input feeds which are comma sepeated and I am formatting in the program to standard layout. Each file will have a 5 byte unique identifier which is passed in the header. This program converts the comma seperated files from different sources (i.e. different 5 byte unique client Id's). After the program converts the data, we want to store the converted data in a GDG and we want the GDG to have the 5 byte unique ID as one of the qualifier. So the GDG will look something like this ..
SAMS.&uniqueid..file1.data(+)..
We wont know the unique Id until the program reads this from the header. Is there a way to pass this parameter from the program and include in the succeding steps in the job as a symbolic..
thanks a lot.
Krk123.

[dtf]

I have not really thought this through, but you might be able to do this with a combination of dynamic allocation (to create your individual file), and then building JCL and submitting it through the internal reader in order to do your processing.

That is the only thing that comes to mind........
________
roll blunts

[kolusu]

[krk123]

Hi Kolusu,
The GDG base will already exist. These files will be coming in daily and each file after conversion need to be stored in the +1 generation of the GDG.

[superk]

It sounds as if the program will have to:

A. Check for the existence of the GDG base, and, it not found...

B. Dynamically create a new GDG base entry.

C. Then, determine all of the existing generations for the base (if any), calculate the next generation, and dynamically allocate the appropriate dataset, ultimately copying the converted data into it.

[gharisankar]

Krk123,

I think, if the above sceneria is possible, we can make this as a FLAT file. We cannot change the GDG base dynamically.
_________________
Regards
Hari

[krk123]

Hi Superk,
We dont have to check for the existance of the GDG base,The GDG base will exist. Before we setup the channel to send the Comma seperated files, we will make sure that the GDG base is setup. We just need to figure out on how to include the unique ID as one of the qualifier. Currently we are thinking of building the JCL in the job and then submit thru internal reader. Not sure if these is any other soution.

Thanks a lot .

[kolusu]

krk123,

I am totally confused now. If you already had the GDGbase with unique identifier , then why do you need to tag to the new generation?

for ex you have a base with the following name.

[krk123]

I built a proc which just repro's data. So the proc is lookign something like this..
Please ignore the syntax..
step010 exec IDCAMS.
IN1 dd dummy
OUT 1 DD DUMMY.
SYSIN DD *
REPRO IN1 OUT1 .

I have created a job which looks something like this..

job card info..
step001 exec above proc..
step010.in1 dd dsn=file output from the program.

In the program, I read the above job ASIS write to a file and I append the folloing statement at the end of the file.
step010.out1 dd dsn=HLq.uniqueid.file(+1).
I am submitting the job thru internal reader..
Let me know if this not clear..I can elaborate..thanks for all the tips..

[SureshKumar]

krk123,
Read the following posts on dynamic allocation. Do your DD and decide what works best with your design. Requirements like this always has limitations. Look for the pros/cons of using PUTENV.

Try PUTENV. IKJTSOEV/IKJEFTSR, BYXBPWDN etc. You can have generic perform paragarphs, which can be invoked for every change in criteria. Then again, we are assuming that the GDG Base exists. Hope it helps.

http://www.mvsforums.com/helpboards/viewtopic.php?t=949&highlight=putenv
http://www.mvsforums.com/helpboards/viewtopic.php?t=1905&highlight=putenv