DSXchange

Use transformer looping for complex problems, this is your basic boring pivot so use the Pivot stage.

Interesting idea. Thought about something similar back in the day but never had any time to do anything of the sort. Could be fun.

Based on the example posted, it looks to me like you've already got your "AId" field in the "Id" field. Can we get a more realistic example, please? One that shows why you would need an additional field.

I actually meant that you would be able to take the written requirements and turn that into DataStage PX functions. Not for you to do that so others would do it for you. However... As I said, Convert() is a powerful ally here. It can handle the first example all by itself, no need for Trim. Check th...

Well... those are Oracle regular expression functions which you could certainly continue to use. As for DataStage, since there is no direct support for "regexp", first thing in my mind would be to break down what exactly that expression is doing and put it into words. I've found that works...

If things are working as you are saying, seems to me you should contact your official support provider and ask them the why of that.

I don't know why you think you need a Sequence job to run a script, if you had one it could be run via the same ExecSH after job command you are using right now. That being said, seems to me all you need to do is add the one line to the beginning of your sql file that Chetan noted: WHENEVER SQLERROR...

Moved. I made some assumptions about your Version & O/S, please let us know if they are incorrect.

Seems like every post in the FAQ Discussion forum really belongs elsewhere. [sigh] Sad little forum.

What are the data types for your source and target fields? Strings?

Popular question, it would seem.

Someone else has the same dilemma here.

Show us your syntax.

Or "test -s" and make sure it fails.

Note that these assume you have the MKS Toolkit installed and can thus run UNIX commands. DOS commands to check filesize are... trickier.

Typically one would do something like that on the target not the source tables. As noted, if your goal is to improve the runtime that is not the way to do it.

Show us your entries.

Agreed. Purging by runs was problematic early on and may still not be the world's greatest idea. See if switching to age-based auto purging helps.