Please do not attempt to scrape the website. Most of it is written as a client-side javascript app (which isn’t easily scrapable), and the few pages that aren’t JS driven soon will be.
As Sunspots said, it’s better to parse the API (even with regexes) rather than attempt page scraping.