Re: Proposal for \v and \V, the small- and large- cut regex opera tors.

On Thu, Aug 10, 2000 at 07:04:50AM -0400, Ilya Zakharevich wrote: > $quoted = qr/(['"]).*?\2/; > @a = $str =~ /($quoted)/gp; > > Here //p is the "postponed" flag. Put (?p{$quoted}) instead of > $quoted to get this semantic now (or some other char). $quoted = qr/(['"]).*?\1/; @a = $str =~ /(

Re: Proposal for \v and \V, the small- and large- cut regex opera tors.

On Thu, Aug 10, 2000 at 11:12:28AM +0200, Johan Vromans wrote: > Of course, we need group names (trivial), and group temporaries. > I needed the latter to define a generic pattern to match quoted strings: > you need to store the starting quote somewhere to find the ending quote, > but I didn't wan

Re: Proposal for \v and \V, the small- and large- cut regex opera tors.

On Thu, 10 Aug 2000 04:42:56 -0400, Ilya Zakharevich wrote: >> Variable interpolation can be handled using Damian's curried expressions. > >On XRay: > >Summary for query "curried;Damian": > > found 0 matches in 0 files. Look up RFC 23 on : "Higher order functions". -

Re: Proposal for \v and \V, the small- and large- cut regex opera tors.

On Thu, Aug 10, 2000 at 10:16:12AM +0200, Johan Vromans wrote: > > "(" (.) group(1, [[ 'a' .. 'z' ]] (*) [3,5] ) (.) ")" > > > > here (.) is the ASCII substitution for the UNICODE "REx concatenate" > > char, similarly for [[. ]] and (*). (The chars may be borrowed from > > the math repertoire.

Re: Proposal for \v and \V, the small- and large- cut regex opera tors.

On Thu, Aug 10, 2000 at 04:42:56AM -0400, Ilya Zakharevich wrote: > These are just user-defined ops. You should be able to overwrite the > normal ops, as in: > > sub match_all { > use re_ops 'overload_usual_ops'; > "(" . group(1, [ 'a' .. 'z' ] * [3,5] ) . ")" > } > > Will this go?

Re: Proposal for \v and \V, the small- and large- cut regex opera tors.

On Wed, Aug 09, 2000 at 04:20:32PM -0400, Ilya Zakharevich wrote: > It is not clear though how to design concise-but-no-line-noise > notation for \w etc. But qr/ \( ( [a-z]{3,5} ) \) / may become > > "(" (.) group(1, [[ 'a' .. 'z' ]] (*) [3,5] ) (.) ")" > > here (.) is the ASCII substitution