There's currently no way other than crawling the web site. The HTML and PDF files are intermingled with the original files.

(I'm the author of the blog post, walkthrough, and archive.)