DSXchange

An answer from a slightly different perspective. (1) Resource allocation in server jobs and job sequences is "on demand". The "file" in which the demands are recorded is not, indeed, a file but rather a database table in the Repository called RT_CONFIGnnn where nnn is the job number. Records in this...

Downstream of the Execute Command activity - on any activity (Job Activity, Routine Activity, etc.) when you are entering a job parameter value you can nominate one of the upstream activity variables - in this case the output from your Execute Command activity. Assuming you're on version 7.5, you ca...

Delimited or fixed-width source? If delimited there's nothing; int32 can have up to 10 digits. Your job will need to enforce the business rule downstream. Or you could declare it as string[3] for the import, and convert to int32 subsequently.

My preference is to use an explicit Sort stage to tell DataStage not to sort. By this means I can document what the job is doing, to aid the future maintainer, and not have to be concerned whether the environment variables are set correctly in test and production.

Data Sets work like text files, only in a parallel environment. It is the operating system that limits you, not DataStage. Therefore there is no workaround in DataStage.

If the routine exists in your repository you can examine its documentation and source code, which will tell you what (if anything) you need to provide in the Input Values field.

But the hashed file solution will be WAY easier on memory.

Is it not said that before you can define recursion you need to define recursion?

That's why I had to ask the question about "/" in the STRIPLIST. It couldn't have parsed "1/2" as a single token. I think your first solution was better.

You're right of course, a pattern override does not provide the mechanism for splitting the token, so you'll need to use PAL. Not having the books with me at the moment (and not being fluent in PAL) I'll have to postpone that answer. Meanwhile, take a scan through the PAL manual; you might spot some...

There can be more than one in the case of normal lookups.

There can only be one if it is a sparse lookup.

You've quoted the manual out of context. But you've understood the rules.

Please verify whether this is a server job (as per job type) or parallel job (as per forum in which posted). The answer will depend upon this fact.

... which is precisely what you get with a USING clause.

Override for >T in the xxADDR rule set.

">" is a "mixed, leading numeric" token

Is "/" in your STRIPLIST ? If so you only have five pattern components to parse.

How do you know the correct street type is ST ?
Use a pattern override for ^R^

Through the GUI only via a rules override.