CPYFRMIMPF SQ99999 code 4

We had a vendor change their data exchange format from XML to CSV without notice. With the XML data, I plucked out the 14 columns I cared about and ignored the rest. The CSV data has 156 columns. I've created a new table with the intention of copying the CSV data into that table, and they doing a copy *MAP into my existing table.

The CPYFRMIMPF is failing with a bunch of:
Message ID . . . . . . : SQ99999 Severity . . . . . . . : 30
Message . . . . : Error occurred in SQL Call Level Interface
Cause . . . . . : A procedure call encountered an error. The error code is
4. Error codes are:
1 -- Memory allocation failure.
2 -- Column number not valid.
3 -- Program type out of range.
4 -- SQL data type out of range.


With a single one of these at the end:
Message ID . . . . . . : CPF2845 Severity . . . . . . . : 40
Message . . . . : The copy did not complete for reason code 13.
Cause . . . . . : The Copy From Import File (CPYFRMIMPF) command, or Copy To
Import File (CPYTOIMPF) command could not be completed for reason 13. The
reason codes are:
13 - The FROMSTMF contains a record whose length prevents it from being
properly imported. This situation occurs when a valid record delimiter is
not found before a buffer overflow situation would occur.


The ironic thing is that I was able to import the CSV using the DB Dev perspective in RDi (table->Data->Load). I don't often use RDi for DB stuff, so I can't really see how it accomplished the import. This is part of a daily process, so me manually importing the data with RDi isn't a practical option.

Does anyone have any ideas or suggestions?

TIA