I don't see a use for BeautifulSoup nowadays. I use lxml.etree for everything (with the HTML parser when needed), and do 99% of queries using XPath. It's the best way to do scraping with Python, in my experience.