Re: [GENERAL] indexed regex select optimisation missing?

> Stuart -
> I'm forwarding a version of your last message to the hackers list, and
> to Lamar Owen, who's the keeper of the RPMs. The short verson, for those
> who haven't followed this thread over on GENERAL, is that Stuart is being
> bitten by the USE_LOCALE affect on the makeIndexable() function in the
> parser: anchored regex searches on a large table (a glossary, I believe)
> take a long time, proportional to sort position of the anchoring text:
> i.e. searching for '^zoo' is quick, '^apple' is very slow.
>
> I seems to recall the packagers here (Lamar and Oliver) asking if defining
> USE_LOCALE for the general RPM or deb would cause any problems for other
> users, who don't need locale info. Here's a real world example.
>
> The discussion about this was last June, and shifted focus into the
> multi-byte problem, as far as I can tell. Bruce, some version of this
> is on the TODO list, right?

I have beefed up the FAQ with a mention that locale disables regex
indexing, and have added to TODO:

* Allow LOCALE to use indexes in regular expression searches

--
Bruce Momjian | http://www.op.net/~candle
maillist(at)candle(dot)pha(dot)pa(dot)us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026