Josh Berkus <[email protected]> writes:
> Why is a function which returns void returning a row?
Returning a scalar result that happens to be of type VOID is an entirely
different thing from returning a set result that contains no rows.
> Why is that row
> NULL if it's a SQL function and empty if it's a PLPGSQL function?
I'd say that the SQL function is probably doing the right thing. It
appears that plpgsql has hacked this specially "for backward
compatibility":
/*
* Special hack for function returning VOID: instead of NULL, return a
* non-null VOID value. This is of dubious importance but is kept for
* backwards compatibility. Note that the only other way to get here is
* to have written "RETURN NULL" in a function returning tuple.
*/
if (estate->fn_rettype == VOIDOID)
{
estate->retval = (Datum) 0;
estate->retisnull = false;
estate->rettype = VOIDOID;
}
I haven't tested, but I think that diking out this section would make
the result be a null (still of type void).
regards, tom lane
---------------------------(end of broadcast)---------------------------
TIP 6: explain analyze is your friend