match with exception

vit has asked for the wisdom of the Perl Monks concerning the following question:

Replies are listed 'Best First'.
Re: match with exception by davis (Vicar) on May 19, 2009 at 12:57 UTC
I'd be inclined to use the "does not match" feature of !~: `#!/usr/bin/perl use warnings; use strict; use Data::Dumper; my $string = "asdfxyzasdf"; print "Not found\n" if($string !~ /xyz/);` [download] Of course, that assumes a regular expression is the best way of (not) matching your string -- if it's just a simple string, then the perl function index() may be more suitable. Edit: Changed wording to more accurately reflect test. davis	[reply] [d/l]
Re: match with exception by zwon (Abbot) on May 19, 2009 at 13:00 UTC
maybe you searching something like that: `$str !~ /xyz/; do_something unless $str =~ /xyz/; do_something unless $str =~ /^xyz$/;` [download]	[reply] [d/l]
Re: match with exception by paulchernoch (Acolyte) on May 19, 2009 at 14:18 UTC
The other answers are correct if all you want to do is not match an entire pattern. But sometimes you want an expression that matches in one part but does not match in another. To do this you need to use look around assertions. For example, if you want a pattern that matches "abc" if it is not followed by "xyz" you would do this: `my $s = "abcxyQ"; if ($s =~ /abc(?!xyz)/) { print "Matches '$s'\n"; } else { print "Doesn't Match '$s'\n"; } my $t = "abcxyz"; if ($t =~ /abc(?!xyz)/) { print "Matches '$t'\n"; } else { print "Doesn't Match '$t'\n"; }` [download] The look-ahead assertion is inside the (?!). Read the perl documentation about this feature - it is tricky and does not always work as expected (but it does work as designed!) - Paul	[reply] [d/l]
Re: match with exception by AnomalousMonk (Archbishop) on May 19, 2009 at 15:34 UTC
This depends on just how you define 'everything except sequence xyz'. One way (note that a piece of 'xyz', the 'yz', is captured – after all, 'yz' is not 'xyz'!): `>perl -wMstrict -le "my $s = 'abcxyzdef'; my $exclude = qr{ xyz }xms; my $all_not_xyz = qr{ (?! $exclude) . }xms; my @matches = $s =~ m{ ($all_not_xyz+) }xmsg; print qq{@matches}; " abc yzdef` [download] Another way (note that the 'xyz' sequence is entirely excluded): `>perl -wMstrict -le "my $s = 'abcxyzdef'; my $exclude = qr{ xyz }xms; my @matches = split m{ $exclude }xms, $s; print qq{@matches}; " abc def` [download] These approaches generalize nicely to multiple exclusion sequences: `>perl -wMstrict -le "my $s = 'abcFOOdefBARghi'; my $exclude = qr{ FOO \| BAR }xms; my $all_but = qr{ (?! $exclude) . }xms; my @matches = $s =~ m{ ($all_but+) }xmsg; print qq{@matches}; @matches = split m{ $exclude }xms, $s; print qq{@matches}; " abc OOdef ARghi abc def ghi` [download]	[reply] [d/l] [select]
Re: match with exception by JavaFan (Canon) on May 19, 2009 at 15:44 UTC
I would write (untested, may contain typos): `/^[^x](?:x(?!yz)[^x])$/` [download] This only does a lookahead after finding an 'x', which ought to be faster than doing a lookahead for each character (which is what `/^(?:(?!xyz).)$/s` based solutions do). Of course, the performance difference may not be relevant for you. If needed, the lookahead can entirely be eliminated, but that requires a more complicated regexp - and then I would have to think.	[reply] [d/l] [select]