Sort stage writes #'s in place of null in output to seq file

jmryan2000 · Post by **jmryan2000** » Wed Aug 16, 2006 2:56 pm

I need to sort a fixed length sequential file by 2 columns and tried using the sort stage. When reviewing the output through the data viewer in DS, the data appears OK. When reviewing the output in a text editor (TextPad), any fields that contained no data were replaced with #'s (though fields that have trailing spaces retain the spaces). So a 5 char field with no data comes across as #####. The data coming into the sort stage doesn't have the #'s. Any ideas why this is occurring and how to stop it?

Thanks!

kris007 · Post by **kris007** » Wed Aug 16, 2006 3:01 pm

If you go to the Inputs tab on the (target)sequential file and then go to the Format tab what do you see in the Default padding field. Change it to what you want it to be. :wink:

jmryan2000 · Post by **jmryan2000** » Wed Aug 16, 2006 3:58 pm

That did the trick. Thanks, again!

ray.wurlod · Post by **ray.wurlod** » Thu Aug 17, 2006 8:14 am

Diagnostician Note
Don't be in too much of a hurry to attribute blame to a particular stage type.

Here the subject line suggests a problem with the Sort stage, which may mislead future searchers. The problem wasn't. DataStage did exactly what it was programmed to do - via the pad character property in the Sequential File stage.