Currently we only implement _get_html and _get_html_range and only a subset of
html tags are identified properly. This was intentional as it's easy to add
additional tags and because we're also introducing testing to Pidgin in this
commit.

Ran the unit tests under valgrind without issues.