Hi,
I've recently migrated from MySQL to PostgreSQL and as impressed as I am
with Postgres, I have found one seemingly missing feature to be a little
bothersome..
'mysqlimport' has the ability to skip duplicate records when doing bulk
imports from non-binary files. PostgreSQL doesn't seem to have this
feature, and it causes a problem for me as I import extremely large
amounts of data into Postgres using 'copy' and it rejects the whole file
if one record breaches the primary key.
I have managed to get around this by hacking
src/backend/access/nbtree/nbtinsert.c to call elog with NOTICE instead of
ERROR, causing it to skip the duplicate record and continue importing.
Is there a way to get around this without changing the code? If not, will
a future release of Postgres optionally implement this?
Thanks in advance,
Steve Micallef