In message <00e301c27366$e2d95eb0$[EMAIL PROTECTED]>, "Bob Dickinson \(BSL\)" writes: >ORO has handled these expressions just fine with only one exception: the >SpamAssassin rules use the negated case insensitivity extended regular >expression, as in (?-i). Since ORO does not support this, I have to throw >out about 20 expressions.
It looks like this first appeared in Perl 5.005. Perl5Compiler/Matcher are still mostly just Perl 5.003 compatible and we seem to have settled into adding later Perl features on an as-need basis rather than all at once. I just added support for negated modifiers. Note, however, that we still do not support Perl 5.005+ group-local modifiers such as (?imsx-imsx:pattern). You can do an anonymous CVS checkout of the latest source and test it out. If it passes muster I think we should just release 2.0.8 instead of a 2.0.7-dev-2. >Are there any plans to add this functionality? If no (or not soon), then is >this something that would be straightforward for me to do, and where would I >start? This one happened to be straightforward and only required a couple changes in Perl5Compiler. Other changes for things like group-local modifiers and look-behind assertions will be significantly more involved. All told it's a matter of enough nagging and enough available time to bring the Perl stuff up to full Perl 5.6 (or are we at 6.0 yet) compatibility. However, we really need some good unit tests before making the full charge. It sounds like your Spam Assassin expressions would make a good start for both regression and performance tests. daniel -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
