Since you are extracting links from html, you should probably use HTML::LinkExtor as recommended, but it might also be useful to point out URI::Find, which is good for extracting links from arbitrary text. It doesn't try to systematically parse the text, it just looks for something resembling a URI. Pretty handy.