Unfortunately, I believe perl is the wrong tool to use for web-scraping these days. It's fine for Web 1.0 applications, but for 2.0 apps you are much better off using the browser itself. You really need a complete Javascript/DOM environment to do it adequately.

I'd investigate Firefox plugins like GreaseMonkey or Selenium. Selenium is controllable via perl, so there is still room for perl, but all the heavy lifting is going to be done by Firefox and the Selenium plugin.