Extracting email addresses from mailto URIs using HTML and URI parsers instead of regular expressions

[dorward]

I don't like using regular expressions on HTML documents, so my approach would be to use a proper HTML parser instead. This has a number of benefits, including the decoding of entities in the HTML representing the email address. This code uses LWP::UserAgent to fetch the HTML document, HTML::TokeParser to read it, and URI to parse the URIs in it.

#!/usr/bin/perl

use strict;
use warnings;
use LWP::UserAgent;
use HTML::TokeParser;
use URI;

my $ua = LWP::UserAgent->new;
$ua->timeout(10);
my $root_uri = 'http://example.com/';
my $response = $ua->get($root_uri);
if ($response->is_success) {
    my $html = $response->decoded_content;
    my $p = HTML::TokeParser->new( \$html );
    while (my $tag = $p->get_tag('a')) {
       my $href = $tag->[1]{href};
       next unless $href;
       my $uri = URI->new_abs( $href, $root_uri );
       next unless ($uri->scheme eq 'mailto');
       print $uri->to, "\n";
    }
} else {
    die $response->status_line;
}