Ignoring parse errors with Html Agility Pack?

I'm trying to parse out a single page from YouTube... Which isn't really free of syntax errors. Html Agility Pack screams about th开发者_运维百科ese errors, and returns nothing in result.

http://codepaste.net/gh3hco

---

I haven't tried this, but based on a suggestion in their forum you can use HTML Tidy or Tidy.NET to clean the HTML first. Optionally, you could find the erroneous tags and remove them in a pre-process step.