From: | "Robert Haas" <robertmhaas(at)gmail(dot)com> |
---|---|
To: | "Pavel Stehule" <pavel(dot)stehule(at)gmail(dot)com> |
Cc: | "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "Dimitri Fontaine" <dfontaine(at)hi-media(dot)com>, pgsql-hackers(at)postgresql(dot)org, "Gregory Stark" <stark(at)enterprisedb(dot)com>, "Martijn van Oosterhout" <kleptog(at)svana(dot)org> |
Subject: | Re: WIP patch: convert SQL-language functions to return tuplestores |
Date: | 2008-10-30 17:11:42 |
Message-ID: | 603c8f070810301011y2a8dfac8m94195eaae7909a76@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
> With session variables we could implement srf function in plpgsql like
> current C srf function. Like
>
> create or replace function foo(....)
> returns record as $$
> #option with_srf_context(datatype of srf context)
> begin
> return row(...);
> end;
> $$ language plpgsql;
Oh, sure - but what you can do with this will be somewhat limited
compared to a Perl hash reference off which you can chain any
arbitrary data structure with ease. I'd want to see an actual use
case for this before anyone bothered implementing it. I was actually
thinking one way to do it would be to extend the variable declaration
syntax so that you could declare n>=0 variables as SRF context
variables, which I think is nicer, but even that I think is of limited
usefulness. I think the biggest value of PL/plgsql is the ability to
RETURN QUERY, and I think the ability to push a lazy execution model
down into that subordinate query is where the win is. That case won't
be helped at all by this sort of alternate calling convention - in
fact it'll be nearly impossible to even do that at all with this type
of execution model.
...Robert
From | Date | Subject | |
---|---|---|---|
Next Message | Joshua D. Drake | 2008-10-30 17:13:16 | Re: Decreasing WAL size effects |
Previous Message | Zdenek Kotala | 2008-10-30 16:24:46 | [PATCH] HeapTuple version extension + code cleanup |