CFF multiple instance

[NSIT-DVP-01]

Hi all,

I want to design a job to read the EBCDIC file with records of multiple structures and save records with the same structure in one sequential file. I prepare to use CFF as input, and several sequential files as output. I have copybooks and table definitions in datastage for all the structures of records.

My question is that I have several EBCDIC files and every of them contains records of different structures, and I want to use one job allowing multiple instance to make it work for all EBCDIC files.

I wonder if there is a way to load the table definition via a schema file in CFF stage like what we can do in a sequential file stage. Then I can set the schema file name as a parameter and pass the value dynamically to the job in runtime. Like that I can make this job universal for all EBCDIC files no matter what's its structure. Can anybody give me an advice?

Thanks in advance!

[eostic]

I haven't tried anything like this, but I tend to think that it will be very difficult if not impossible, because the CFF underpinnings combine a lot of stuff. ...been awhile since I've looked under the covers, but when it was first written, the EE CFF was ultimately reaching out to the already existing various restructure operators, depending on what repeating groups, occurs, etc. that it needed to utilize. ...I wonder if the restructure stages would be more maleable for the technique you are looking for? ...might be harder to code the initial job, but perhaps they are flexible enough to take schema defs?

Ernie

[NSIT-DVP-01]

Hi Ernie,

Thanks for your reply. But I think the restructure stages will make things much more complicate. What I need is to load the table definition dynamically in runtime. Maybe via a schema file or other else. I think it is hardly possible considering the number of table definitions to load for each EBCDIC file is variable. Because for me, every EBCDIC file has several types of record structures and the number of which is not certain...

I think that even if I reach to solve the problem of loading one table definition dynamically, I can't reach to solve the problem of variable number of table definitions to load. What's more, I have to solve the problem of the variable number of output...

I post here just hoping to finding out someone having done such thing.Maybe there is... Anyway thanks for you advice.

regards