Sorry but lxml with ETree will handle any amount of broken html you throw at it. Add in XPath and I find lxml to be a far superior, and more memory efficent, op

Sorry but lxml with ETree will handle any amount of broken html you throw at it. Add in XPath and I find lxml to be a far superior, and more memory efficent, option. Source: former professional web scraper.