"Karl O. Pinc" <kop@meme.com> writes:
> Yes. I'm wrong. For some reason I thought you could use DO to make
> an anonymous code block that would act as a SETOF function,
> allowing RETURN NEXT expr (et-al) to be used in the
> plpgsql code, allowing DO to return table results.
> (Or, perhaps, instead, be used in place of a table in a SELECT
> statement.) Oh well.
My key for remembering about that point is that DO is a utility command,
not a query. Now, the proposal I pushed last time we opened that very
can of worms was to have inline functions rather than anonymous code
blocks:
WITH FUNCTION foo(integer) returns bigint language SQL AS $$ SELECT $1 + 1; $$,
Not sure how much that relates to $topic, but still something that
raises in my mind with enough presence that I need to write about it so
that it stops calling for attention :)
Regards,
--
Dimitri Fontaine
http://2ndQuadrant.fr PostgreSQL : Expertise, Formation et Support