I use Regexp::Common::URI for finding URLs, and it can also validate URLs.

It comes with two caveats though: It doesn't understand https (only http), so I do something like this in my code:

use Regexp::Common qw /URI/;
my $re = $RE{URI}{HTTP};
$re =~ s/http/https?/g;
[download]

The second caveat is that it doesn't understand anchors (ie the part after a # pound sign).