Something like that was discussed on commons ( I think ). The problem seems to be that different regex packages use different rules ( Perl-like, etc ), there is not standard syntax for regexp.
That means the behavior would be dependent on which regexp engine is used. I would rather pick one and use it, and deal with the build problems. Probably we should just use what apache uses, it seems like a reusable piece of code. Other opinions ? Costin Mladen Turk wrote: > Hi, > > I've posted a uriMap code that matches uri's using regular expressions > couple of weeks ago. > There was a consensus that we need to wait until 2.1 and eventually that > we use the regex from APR (if it ever will be there). > > Ok, so I'm proposing that we make our own regex code (?). > Why? > > 1. Apache2 has pcre.lib and pcreposix.lib as a part of standard > distribution > 2. Apache1.3.x has a regex.lib > 3. For IIS we may use the standard pcre-3.9 or the one from Apache2 > (they are the same) > 4. Most unixes has pcre or regex as part of standard distribution. > 5. If ever pcre gets to the APR we may use that one like any other. > > So, we will need single file with jk2_regex_compile, jk2_regex_match and > will be able to use any of the upper combinations? > > Thoughts? > > MT. -- Costin -- To unsubscribe, e-mail: <mailto:tomcat-dev-unsubscribe@;jakarta.apache.org> For additional commands, e-mail: <mailto:tomcat-dev-help@;jakarta.apache.org>
