Dan Muey wrote: > > Howdy, Hello,
> perldoc perlre on 5.8.0 says that [:ascii:] should match any ascii > character and [:^ascii:] is the negate version. > > If I do =~m/[:^ascii:]/ or [:ascii:] on astring that is 'hi' it That character class matches the characters 'a', 'c', 'i', 's', ':' or '^' and 'hi' contains the character 'i' so it returns true. > returns true each way, so I believe it says the [] are part of > [::] conbstruct which I think means you have to have[[:class:]] Yes. > Now [^[:ascii:]] and [[:ascii:]] behave as you'd expect so I think > that's all correct ( Please tell me if I'm wrong) > > SO my question is: > > Which is faster/better/more portable/compliant etc.... > > [^[:ascii:] or [[:^ascii:]] ?? $ perl -le'use re qw/debug/; "xxx" =~ /[[:^ascii:]]/g' Compiling REx `[[:^ascii:]]' size 11 first at 1 1: ANYOF[-\377](11) 10: END(0) stclass `ANYOF[-\377]' minlen 1 Matching REx `[[:^ascii:]]' against `xxx' Freeing REx: `[[:^ascii:]]' $ perl -le'use re qw/debug/; "xxx" =~ /[^[:ascii:]]/g' Compiling REx `[^[:ascii:]]' size 11 first at 1 1: ANYOF[-\377](11) 10: END(0) stclass `ANYOF[-\377]' minlen 1 Matching REx `[^[:ascii:]]' against `xxx' Freeing REx: `[^[:ascii:]]' They both look the same to me. John -- use Perl; program fulfillment -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
