Fixed Length CSV File Column Field names

AETLP v6.3.7.23 x32

Hi, I suspect you may have come across this and may have a neat workaround that may be more affective that the one I am considering.

One of our trading partners supplies what is supposed to be a CSV file. However, it comes to us with each row padded to a 2000 byte length! There are over 60 fields in each row. In general AETL handles this. However, when it comes to the last field in the definition it has difficulties establishing the final field names and I suspect there may be some physical size limits being hit.

I should also point out that the content of the file is in fact comma separated. it is just the last column in each row gets PADDED and then the End-Of-Line marker is put in.

The screenshots attached will explain better what we receive. What would help in this case is a RTRIM() of the row we nominate contains the column names. That way the final column name will be derived correctly.

I am considering parsing the file with AWK first and to RTRIM each row before it gets into this transformation/ but obviously this introduces another moving part in the package that could fail.

I have tried to manually manipulate the final column names in the AETL Reader File Definition without much luck.

Ideas? Thanks

We will trim it by default

Working with fixed-width files can be a real pain especially if it is very wide.
What I do is export the definition into excel edit it or even better, paste the definition from spec if I have one and paste it back.
Saves a lot of time