DSXchange

Welcome aboard.

You really need to provide more information than that. For example, what value can not be converted from NVarChar to Integer?

For instance, 9998887776 is only ten characters (digits), but is not valid as an integer.

The approach suggested by James (using Quick Find) will still work. Once you have the result of the search, hover over each job name in turn - an image of the job will appear. Use the Ctrl key to change the size of the image.

Welcome aboard.

Use a Sort stage to form the groups, and a Transformer stage to preserve the first from each group. Note that a Remove Duplicates stage is not what you want here, since you can't guarantee that the data are sorted by Column13.

No, I meant the material in Chapter 4 - the BASIC and CATALOG commands. This also means that you need to create your routine using a line editor in a "Type 19 file".

Anything developed in 8.1 will run successfully in 8.5. (The reverse is not true.) Do consider upgrading your development environment at about the same time.

Show us the code from the Build stage. Also compile and run the job in trace mode so you can determine what's coming out of the Build stage.

Has it perhaps received "3 " ("3" followed by " ") as the input?

elif was unexpected.

Verify the syntax of the if..elif..else..end construct in your particular shell on your particular UNIX.

The function name is UniSeq().

UniSeq() is to Seq() as UniChar() is to Char().

Your licence for 8.1 is good for 8.5.

No. The reason is that routines in a project are stored in a database table (DS_ROUTINES). You need to work externally to this, as I mentioned earlier.

... assuming, of course, that du even works with a file. It does not work on files (only on directories) on some UNIX variants.

Information Center for Information Server v8r1
But why? Version 8.5 is way less buggy!

What was the resolution?

Not by any published means within DataStage.

However you could create the routine at engine level (rather than at project level) and catalog it globally with the location of its source code suppressed. These techniques are described in Chapter 4 of the DataStage BASIC manual.