Re: Slow tab completion w/ lots of tables

Stephen Frost <sfrost(at)snowman(dot)net> writes:
> That's the kind of concern that I was expecting, to be honest. :) As
> Kevin's pointed out, it's not likely to be needed anyway.. There's a
> bit of an open question still regarding case-insensitive searching, but
> perhaps we let that be slow and only done if we don't get any answers
> back from a case-sensetive search?

Um, I don't believe we do any case-insensitive search now, do we?

> We'd essentially do: LIKE 'xx%', and then run quote_ident() on the
> result (I assume we can replace the whole word, right?). I'd also
> strip off any ", for the purposes of searching with tab-completion.

I think you might be saying the same thing I said in my prior message,
but not quite sure.

> I'm
> not sure how easy it'd be to have a fall-back setup. I do wonder if we
> should do what I often recommend my dev do though, which is to have a
> SQL or pl/pgsql function defined on the database-side, rather than
> sending large/complex queries to the database from the application..

The nice thing about keeping this knowledge on the psql side is it would
still work with older servers.

regards, tom lane