LAE

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: Jensen

I'm trying to read in files using the CSV node.

Due to nature of the data, there might be zero records in the file.
When this happens, the filter node, which is next in line, to change the fields into native data type will fail.

I tried creating a Cat and Dummy input node to work around this but it doesn't work.

Basically I hope to be able to still run the entire graph smoothly regardless whether there are any records from these files.

Thanks in advance!

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: stonysmith

On your CAT node - did you select ConcatenationMode=Union?
On the DUMMY node - did you list all the fields that the filter node is expecting to see?

The filter node will work fine with zero records, as long as the proper column headers are found.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: Jensen

The CAT node is on default concatenation mode because the dummy node has all the fields listed already.

I have resolved the issue by creating a dummy file which filename is read by the CSV node.
The issue I encountered is that if there is zero filename passed to the CSV node, it will generate zero records with no headers, resulting in the premature EOF error.