Getting a list of captures in Perl 5.10

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

I am trying to parse inter-language links in Wikipedia. I need to know to what languages a given page has links. Let's say that $string is the page's text:

$string = "bla bla bla [[en:English]][[de:German]][[ga:Irish]] bla bla
+ bla";
[download]

And i need the list:

qw(en de ga);
[download]

I can do this:

@matches = ( $string = m/\[\[(en|de|ga):.+?]\]/g );
[download]

And then i get qw(en de ga) in @matches, but that's because i have only one pair of capturing brackets, which is a limitation. If i do, for example:

@matches = ( $string = m/\[\[(en|de|ga):(.+?)]\]/g );
[download]

Then i'll get qw(en English de German ga Irish). Is there a clever way to get a list of all the results from one pair of capturing brackets? I tried using Perl 5.10's named captures and %-. Either it can't be done this way or i am doing incorrectly. I tried this:

my $string = 'bla bla [[en:English]][[de:German]][[ga:Irish]] bla bla'
+;

if (my @matches = ($string =~ m/\[\[(?<lang>en|de|ga):(.+?)\]\]/g)) {
    say 'matches';
    say 'matches: ', Dumper(\@matches);
    say 'minus  : ', Dumper(\%-);
    say 'plus   : ', Dumper(\%+);
}
[download]

i get this output:

matches
matches: $VAR1 = [
          'en',
          'English',
          'de',
          'German',
          'ga',
          'Irish'
        ];

minus  : $VAR1 = {
          'lang' => [
                      'ga'
                    ]
        };

plus   : $VAR1 = {
          'lang' => 'ga'
        };
[download]

You see - only ('ga'), but is there some way to get:

$VAR1 = {
          'lang' => [
                      'en',
                      'de',
                      'ga'
                    ]
        };
[download]

Thanks in advance for any help.

Comment on Getting a list of captures in Perl 5.10 Select or Download Code

Replies are listed 'Best First'.
Re: Getting a list of captures in Perl 5.10 by BrowserUk (Patriarch) on May 24, 2008 at 15:52 UTC
Without using any of the 5.10 features you can do: `$string = 'bla bla [[en:English]][[de:German]][[ga:Irish]] bla bla'; %matches = $string =~ m/\[ \[ ( en\|de\|ga ) : (.+?) \] \]/gx; $var = { lang => [ keys %matches ] }; pp $var; { lang => ["en", "ga", "de"] }` [download] But that makes me wonder why you are capturing the longnames just to throw them away? If you remove the capture for those you can do: `my $string = 'bla bla [[en:English]][[de:German]][[ga:Irish]] bla bla' +; my $var = { lang => [ $string =~ m/\[ \[ ( en\|de\|ga ) : .+? \] \]/gx ] + }; pp $var; { lang => ["en", "de", "ga"] }` [download] Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice. "Too many [] have been sedated by an oppressive environment of political correctness and risk aversion."	[reply] [d/l] [select]
Re^2: Getting a list of captures in Perl 5.10 (pp?) by carol (Beadle) on May 24, 2008 at 23:07 UTC
`pp $var;` [download] Somewhat off-topic, but what is this `pp`?	[reply] [d/l] [select]
Re^3: Getting a list of captures in Perl 5.10 (pp?) by BrowserUk (Patriarch) on May 25, 2008 at 05:23 UTC
See Data::Dump. Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice. "Too many [] have been sedated by an oppressive environment of political correctness and risk aversion."	[reply]
Re^2: Getting a list of captures in Perl 5.10 by amir_e_a (Hermit) on May 24, 2008 at 16:47 UTC
Thanks for the replies. The example with the hash only works with two sets of capturing parens. It's a nice hack, but i was to hoping to find something generic which will work with any number of captures.	[reply]
Re: Getting a list of captures in Perl 5.10 by moritz (Cardinal) on May 24, 2008 at 15:17 UTC
I don't think you can get exactly what you want out of 5.10's regex engine. (In Perl 6 you can get it, and PGE already implements that btw). Apart from the stuff you tried there is also another hack which uses the experimental `(?{...})` code assertions: `our @matches; m/(?>(something)(?{ push @matches, $^N }))/;` [download]	[reply] [d/l] [select]
Re: Getting a list of captures in Perl 5.10 by duelafn (Parson) on May 25, 2008 at 12:49 UTC
splice! `use Data::Dumper; my $string = 'bla bla [[en:English]][[de:German]][[ga:Irish]] bla bla' +; my $groups = 2; my @tmp = ($string =~ m/\[\[(en\|de\|ga):(.+?)\]\]/g); my @matches; push @matches, [splice @tmp, 0, $groups] while @tmp; print Dumper \@matches; $groups = 3; @tmp = ($string =~ m/\[\[(en\|de\|ga):(.)(.+?)\]\]/g); @matches = (); push @matches, [splice @tmp, 0, $groups] while @tmp; print Dumper \@matches;` [download] Good Day, Dean	[reply] [d/l]