Maybe you remember this:
http://blog.netbsd.org/tnf/entry/efficient_wide_character_regular_expressions
Yesterday I have hooked it into the build system, so people can use
libtre as the regex engine in libc by compiling NetBSD-current
with USE_LIBTRE=yes in /etc/mk.conf.
Please test and try to hunt regressions and do benchmarks if interested!
--
NetBSD - Simplicity is prerequisite for reliability
--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-...@muc.de
Will all RE go through libtre then (e.g. regular grep(1) on simple ASCII
files), or does this only apply to specific character sets / locales?
- Hubert
> On Sat, 4 Sep 2010, Adam Hoka wrote:
> > Please test and try to hunt regressions and do benchmarks if interested!
>
> Will all RE go through libtre then (e.g. regular grep(1) on simple ASCII
> files), or does this only apply to specific character sets / locales?
>
> - Hubert
USE_LIBTRE will completely replace it in libc
--
NetBSD - Simplicity is prerequisite for reliability
--
GNU grep doesn't use the libc regex engine. The TRE code included in
libc by Adam is only includes the single byte support for now.
Joerg