Heikki Linnakangas Wrote: > Tom Lane wrote: > > Heikki Linnakangas <[EMAIL PROTECTED]> writes: > >> Also, it would be nice to use B-M(-H) for LIKE as well. > > > > Right offhand, that seems impossible, at least in patterns with %. > > Or were you thinking of trying to separate out the fixed substrings > > of a pattern and search for them with BMH?
> Yep, something like that. Even if it only handled the special case of > '%foobar%', that would be nice, because that's a pretty common special > case. It would be a quick test to check for % at either end. But we'd also need to ensure there were none in the middle. That might slow it down. I guess while looping over the inner chars checking for more %'s we could be building a skiptable. We'd have to abort it if we found any %'s of course I think with out giving it much thought _'s could be handled by BMH, these could just be given a skip distance of 2. Only having a lossy skip table throws that out the window without having a special check for _ David. -- Sent via pgsql-patches mailing list (pgsql-patches@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-patches
