Re: POSIX regex performance bug in 7.3 Vs. 7.2

Hannu Krosing <hannu(at)tm(dot)ee> writes:
> Tom Lane kirjutas T, 04.02.2003 kell 21:18:
>> What advantages does it have to make it worth considering?

> Should be the same as pcre + support for wide chars.

Well, if someone wants to do the legwork to try it, that interface
should work just about comparably to Spencer's package: I found that
the easiest way to make it work is to use pg_mb2wchar to expand our
internal encoding into an array of pg_wchar's and then apply the
regex package to that form. So as long as sre can handle 4-byte
wide chars it ought to more or less drop in.

I've got a fair amount of cleanup to do on Spencer's package before
I can even think of committing it (ANSIfy function headers, fix comment
formatting so that pg_indent won't completely destroy 'em, etc).
But I'll be glad to send the modified interface file (adt/regexp.c)
to anyone who'd like to try getting it to work with sre.

regards, tom lane