HTML parser improvement project

Background information: Servo uses a HTML parser implementation called html5ever. It works great, but it has several missing features that prevent correct implementation of related features in Servo. The goal of this project is to correct that!

Goals:

• fix https://github.com/servo/servo/issues/9604
  • need to modify tokenizer to track line numbers based on \n
  • need public api to retrieve current line number
  • figure out how to write tests for this
  • use new functionality for <script> in servo
• create infrastructure for document.write (see proof-of-concept)
  • need the ability to set and query the insertion point
  • need to modify the tokenizer to always use the insertion point when appending new strings
  • need to modify BufferQueue to support insertion point (maybe split current buffer at current point and add a new buffer in between?)