Well, actually not so obfuscated. The classic thing just happened, because I wrote this piece of code months ago and now I've completely forgotten how or why it actually works. Oh, well. It doesn't look complex, but I just can't get a hold of the idea anymore.

# A simple script to extract HTML/XML tag options

use strict;
use warnings;

foreach my $file (@ARGV) {
    open (FILE, "< $file") or die "No such file: $file";
    my @d = <FILE>;
    close FILE;
    my $t = join(' ', @d);
    $t =~ m/\<(\?)?
            ([a-zA-Z0-9]+)(?:\s+)
                (?{ print "<$^N>\n"; })
            (
                ((?>[a-zA-Z0-9]+)=(?>(".*?")|([^ "<>]+)))
                    (?{ 
                        my ($opt, $val) = split(\/=\/, $^N); 
                        print "\t$opt\t\t$val\n"; 
                    })
            \s*?)*
            ([ ]*\/)?>
            (.*?)
            (<\/\2>)?
            \1
            /isx;
}
[download]