in reply to Find out which subpattern matched in regex

This requires a nested loop: 

use strict;
use warnings;

my %reglist = ( 
        qr/food?/      => sub { print "\t1\n" }, 
        qr/b[a4]rd?/   => sub { print "\t2\n" }, 
        qr/baz(o+ka)?/ => sub { print "\t3\n" } );

foreach (qw(foobarbaz football barcode none bazoooooka)) {
        print "$_\n";
        for my $regex (keys %reglist) {
                $reglist{$regex}() if /$regex/;
        }
}

[download]

UPDATE: If you are happy NOT to use captures in your regexes, you can just see which one has matched from the position in the matches. Due to this restriction I have changed your third regex to non-capturing parantheses. 

use warnings;

my @reglist = ( qr/food?/, qr/b[a4]rd?/, qr/baz(?:o+ka)?/ );
my $giant = join( ")|(", @reglist );

foreach (qw(foobarbaz football barcode none bazoooooka)) {
  my @hits = /($giant)/;
  my @which = map { defined $hits[$_] ? $_ : "" } 0..@hits-1;
  print "$_: @which\n";
}

[download]

Replies are listed 'Best First'.
Re^2: Find out which subpattern matched in regex
by Dallaylaen (Chaplain) on Jun 21, 2013 at 08:09 UTC

    Thanks for you comment.

    I did a quick benchmark and it turns out that nested loop solution is the fastest in most cases. In all other cases - I believe it's when regular expressions converge into a nice tree-like structure - both named captures and unnamed captures are about the same.

    Here's my code sample: 

    #!/usr/bin/perl -w

use 5.010;
use strict;
use warnings;
use Test::More tests => 3;

use Benchmark qw(cmpthese);

my @reglist = ( qr/food?/, qr/b[a4]rd?/, qr/baz(?:o+ka)?/, 100..999);
my @lines = (qw(foobarbaz b4rd perl bazooooka football));

my @expect = ("r0", "r1", "", "r2", "r0");

is_deeply(which_reg_loop(\@reglist, \@lines), \@expect, "which_reg_loo
+p")
    and
is_deeply(which_reg_capt(\@reglist, \@lines), \@expect, "which_reg_cap
+t")
    and
is_deeply(which_reg_named(\@reglist, \@lines), \@expect, "which_reg_na
+med")
    or die "Results differ, no bench";

@lines = @lines x 1000;

cmpthese ( -1, {
    loop => sub {
        which_reg_loop(\@reglist, \@lines);
    },
    capt => sub {
        which_reg_capt(\@reglist, \@lines);
    },
    named => sub {
        which_reg_named(\@reglist, \@lines);
    },
});

sub which_reg_loop {
    my ($reglist, $lines) = @_;

    my @ret;
    LINE: foreach my $str (@$lines) {
        for (my $i = 0; $i < @$reglist; $i++) {
            $str =~ $reglist->[$i] or next;
            push @ret, "r$i";
            next LINE;
        };
        push @ret, '';
    };
    return \@ret;
};

sub which_reg_capt {
    my ($reglist, $lines) = @_;

    my $giant = join "|", map { "($_)" } @$reglist;
    $giant = qr($giant);

    my @ret;
    LINE: foreach (@$lines) {
        my @hits = $_ =~ $giant;
        for (my $i = 0; $i < @hits; $i++) {
            $hits[$i] or next;
            push @ret, "r$i";
            next LINE;
        };
        push @ret, '';
    };
    return \@ret;
};

sub which_reg_named {
    my ($reglist, $lines) = @_;
    my $giant = join "|", map { "(?<r$_>$reglist->[$_])" } 0..$#$regli
+st;
    $giant = qr($giant);

    my @ret = map { $_ =~ $giant ? (keys %+) : '' } @$lines;
    return \@ret;
};

    [download]
[reply]
[d/l]

In Section Seekers of Perl Wisdom