Jeff Trout <threshar@threshar.is-a-geek.com> writes:
> On Jan 24, 2007, at 12:24 PM, Tom Lane wrote:
>> Hmm. This should yield an error (SQL function not returning the type
>> it claims to), and we probably should plug the hole by invoking
>> check_sql_fn_retval every time not just at creation.
> The really curious thing is that it does't complain when restoring
> from the dump - or are those error supressed?
Yeah, pg_dump sets check_function_bodies = off. It more or less has to
because it can't understand the function contents well enough to know
what objects would need to be loaded first (and in fact there might be
no solution --- consider two mutually recursive functions ...)
regards, tom lane