bsoup needs lots of improvements

[dustmop]

Bsoup was written a long time ago, based mainly on the reference python implementation, without too much regard for how easy it would be to use by developers. It was also built when starlark was much younger, and was missing key features such as arbitrary attribute support.

Some things that should be fixed:

• printing nodes should work. Perhaps they could display as an expandable tree
• contents() returns weird results
• get_text is not recursive
• no method to get tag name
• parent.div should work, returning a div child node of parent. child() would be unnecessary then
• parseHtml -> bsoup() rename

Also the docs need lots of work.

[GeoffBarrett]

This issue is old, but I am also seeing issues with get_text(). I am calling it on the entire page contents and it returns an empty string, I am assuming due to the lack of recursion described in this issue.