[Beaver-devel] GNU regex library is trash.

[Leslie Polzer]

Hi there,

after messing around with GNU regex for about 6 hours I can definitely
say I do not wish to use it. As a subsitute I would like to suggest
TRE (http://kouli.iki.fi/~vlaurika/tre/) which we can distribute with
Beaver. I haven't tried it yet but it offers a lot, like

- multibyte support, very important for us of course;
- fuzzy matching, a really nice thing, too; I have worked with the
Levenshtein algorithm before and really liked it

...and it is under active maintenance, whereas the GNU developers did
not address the problem of 'greediness' since 1992, a thing which kills
all of my efforts. Yes, unfortunately the TRE people also haven't managed
to do this yet, but maybe I can do something about that. I am not having
much of a choice apart from that.

It is a real shame that a feature which has been very common in Perl for
many years now and which is very important for professional regexing isn't
available in a single C regex library.


Leslie