Shout out to Web::Scraper. It takes some time to wrap your head around it, but it's pretty good for writing robust scrapers. Relying on the DOM itself makes your scraper very brittle, especially if this is an HTML source that you do not control.