Kris Jurka <books(at)ejurka(dot)com> writes:
> Testing out the new pg_dump exclusion switches I've found that excluding a
> table means that no functions or types will be dumped. Excluding one
> table shouldn't exclude these objects.
I tend to agree ... will see if I can make it happen. (I never did get
around to reviewing that patch, anyway ...)
One issue is what to do with procedural languages and large objects,
which don't have any associated schema. If we treat them as being
outside all schemas, we'd have semantics like this: dump the PLs and
blobs unless one or more --schema switches appeared. Is that OK?
regards, tom lane