how: reject file rows that have too many cols (parall jb)

jherr22 · Post by **jherr22** » Mon Feb 05, 2007 2:51 pm

I'm using Paralled Jobs (in DataStage 7.5.2)
I have an input Seq File Stage (pointing to a CSV file), linked to
a Transform Stage, then output to another Seq File Stage.
When I run, the output log shows 4 rows that were not processed because
they had too many columns (a text field occassionally had a coma in the text).
I cannot find a way to output those 4 rows out a Reject link (Seq File Stage).

Thanks in advance,
-- john
(jherr22)

ray.wurlod · Post by **ray.wurlod** » Mon Feb 05, 2007 3:44 pm

Welcome aboard. :D

The Sequential File stage can have a second output link, called the "rejects" link. This link will capture, as a raw string, any row that does not match the metadata (table definition) being used to read the file.

Subsequently in the job, rows on the rejects link can be reprocessed, for example using a Column Import stage if you are aware of an alternative record schema that might be apposite.