"Martin Gregorie" <martin@gregorie.org> writes:
> The errors were captured in the log file, which is safe from being
> overwritten. You're welcome to either the whole thing or to edited
> highlights from it.
That's probably fairly useless, but if you have the dump file it would
be interesting to look at that.
Are you sure that the errors aren't all attributable to a CREATE TABLE
statement failing and then the subsequent COPY command failing even to
start? Usually you need to look first at the first error, since a lot
of the subsequent ones could be cascading failures.
> 4) I was running psql with a -1 option which, rather
> to my surprise, didn't abort the run after the
> first error in the file.
psql doesn't have a -1 option, that's only for pg_restore. (I believe
pg_restore is a bit smarter about the failed-COPY case, as well.)
regards, tom lane