Re: Counting records in a PL/pgsql cursor - Mailing list pgsql-general

From Magnus Hagander
Subject Re: Counting records in a PL/pgsql cursor
Date
Msg-id 6BCB9D8A16AC4241919521715F4D8BCEA3589D@algol.sollentuna.se
Whole thread Raw
In response to Re: Counting records in a PL/pgsql cursor  (Martijn van Oosterhout <kleptog@svana.org>)
Responses Re: Counting records in a PL/pgsql cursor
List pgsql-general
> > Is there any way to count how many hits I got in a cursor
> in PL/pgsql?
> >
> > I have a function that will "window" through the result of
> a (large)
> > query based on two parameters, but I also want to return
> the number of
> > hits to the client. Right now I'm looping through the entire cursor
> > and incrementing a local variable, which I later return (along with
> > the first <n> records in the resultset) to the client. But
> this seems
> > horribly inefficient... I'd just like to ask "how many rows are in
> > this cursor", is there a way to do that without looping
> through them all?
>
> You can move to the end, look at the row number, then move to
> the beginning. It will still need to materialise the entire
> resultset though.

How do I do that? remember this is a pL/pgsql cursor. From what I can
find at
http://www.postgresql.org/docs/8.1/static/plpgsql-cursors.html#PLPGSQL-C
URSOR-USING, I can only do FETCH to get the next row, or CLOSE.

I can deal with materializing the resultset, but I want to get away from
the loop-a-thousand-times-doing-plus-one...

//Magnus

pgsql-general by date:

Previous
From: Shane Ambler
Date:
Subject: Re: new and old not available in rule subselects?
Next
From: "redhog"
Date:
Subject: Re: new and old not available in rule subselects?