This implementations is probably completely wrong.
It serves as a half baked/but half working idea.

What is outstanding:
1) Actual functionality ie: being able to modify schema and XML through trees.
2) Breaks on some encodings.
3) A proper framework for displaying and interacting with the XML.
4) HTML WSDL SOAP etc
5) Rename plugin to mldocument since it will be doing html too.
6) Reorganise the pics folder and move out of code.