My application is adding a bytea column to a table into which data is
dumped in approximately 4k row batches, one batch approximately every
10 seconds. To this point, those dumps have used copy from stdin;
however, I'm having some difficulty getting bytea encodings to work
with it. Specifically, I can't seem to get the parser to recognize
that what I'm handing it is an escaped string that needs to be parsed
back into individual bytes rather than stored as-is. The encoding is
very straightforward for INSERT, of course, but the COPY ... FROM
STDIN CSV doesn't seem to want to work no matter what I've tried:
\\000
"\\000"
"E'\\000'"
etc.
Is there a trick to this that I just didn't see in the documentation,
or is this some limitation of CSV copy-in? If the latter, are there
suggestions for workarounds other than to fallback on the inserts?
Using 8.3.3, and this is specifically via libpq, if that makes a difference.
Thanks much.
--
- David T. Wilson
david.t.wilson@gmail.com
