Peter Eisentraut <peter_e@gmx.net> writes:
> On 4/4/14 6:21 PM, Andres Freund wrote:
> + /* dependency on transform used by return type, if any */
> + if ((trfid = get_transform_oid(returnType, languageObjectId, true)))
>> Should be compared to InvalidOid imo, rather than implicitly assuming
>> that InvalidOid evaluates to false.
> I think it's widely assumed that InvalidOid is false.
That's not the point; the point is that some nonzero number of compilers
(and probably Coverity too) will warn about this construct, on the not
unreasonable grounds that = might be a typo for ==. (Those extra parens
might satisfy gcc, but not other tools.) Please put in an explicit
comparison of the assignment result, as is done in approximately 100% of
the other places where this idiom appears in Postgres.
regards, tom lane