DSXchange

That would be your classic "unsupported Operating System" error. And 1603 just means "Fatal error during installation".

... and here we are off in our new home, with both errant posts and their accompanying admonishments combined. Since we're talking parameters and probably Sequence jobs and / or command line stuffs, seemed like here would be as good a place as any. smahendra_s - please go back and edit your post her...

You are missing the point specific to this discussion - there is only one field. People initially made the assumption it was something like you posted, but alas 'twas much simpler and hence my question.

Nice Hitchhiker's reference, btw.

We have just one record with one field. However the one field will contain comma delimited values. OK... this has been bugging me. What is the difference between a file where each record contains "one field" and that field contains comma delimited data and a plain old regular comma delimi...

That's true, forgot about the dreaded "Double Convert".

If your target datatype is decimal you'll get all of those zeros. If you don't want them, you'll need to convert it back to a string using the appropriate mask for your desired target format.

Divide by 100.

mczern wrote:Is a dataset able to handle multiple simultaneous writes?

Nope. Multiple readers, yes. Multiple writers, no. It is still, after all, sequential media.

OK, so here we are in the correct forum now. You should be able to use OConv for this, just the "MM" mask should get you what you want.

You could also continue down James' original path as long as you know what the max occurance would be. You then always generate that max number of fields, pivot out to the maximum number of records and then filter out an unneeded / null records post pivot.

This date... is it an actual Date datatype or is it a string? For the former, there are functions to get that, for the latter you can use substring.

Also, you marked the job as Server yet posted in the PX forum, which job type is it? That will change the answer as to the function needed.

If you had said "comma delimited field" rather than "comma delimited record" right up front, I would have expanded a bit more on my answer.

Horizontal Pivot.

Which isn't quite what the OP was asking for but which would probably have the same end result in this situation. You could also build another very similar function that checks each character for a numeric digit and only passes that out as the result, that would match what the Server DIGITS transfor...

NOTE: Split from this older post: viewtopic.php?t=117804

Thripura, please go back and edit your post if I guessed wrong on your particulars - version, O/S, etc.