Hugo <hv@crypt.compulink.co.uk> wrote: |> I don't think I understand what aspects of the current implementation |> could change that would invalidate your \v/\V proposal without also |> undermining the documented definitions of what matches when, but |> consider the (?>...) definition to which he refers (from the latest |> perlre.pod): |> |> An "independent" subexpression, one which matches the substring |> that a I<standalone> C<pattern> would match if anchored at the given |> position, and it matches I<nothing other than this substring>. |> |> Can you write a definition of \v and \V that does not invoke the |> details of the current implementation? I guess the question then becomes how far away from "implementation" one has to step. I believe that my definition steps back as far as the general semantics of regular expressions applied with a nondeterministic finite automata based engine, and in particular, the description of Perl regex semantics given on pages p197-201 of PP3. If we were to step back further to include deterministic finite automata, my description wouldn't be valid, but then, neither would the concept :-) To be clear, I believe that the definitions I posted rely only on documented features of Perl, and on nothing implementation-specific at all. I don't believe that you can have a meaningful understanding of Perl regex semantics without understaning the concept of how a nondeterministic finite automata engine works. (But I don't believe it's important to actually know the phrase "nondeterministic finite automata", and probably a lot better not to.) My opinion is that it's better to talk up front about what's happening, and let understanding of the complex parts blossom from a true understanding of the basics. I realize, though, that not everyone will have the desire/patience for that kind of approach, which is why it's good that there are so many different forms of documentation. But a rose by any other name still smells as sweet (you can quote me on that :-), and however you choose to describe it, I believe that the \v & \V wedges would make meaningful additions to the language. Jeffrey ------------------------------------------------------------------------------ Jeffrey Friedl <jfriedl@yahoo-inc.com> Yahoo! Finance http://finance.yahoo.com