If you're crawling the web looking for email addresses you're probably not bothering to parse the HTML. You don't need to: you can just grab the email from the raw response from the web server, along with any new links to follow.