Test & implement: Component for extracting IDL fragments from HTMLFileContents

[mdittmer]

FOAM's HTMLLexer provides a stream of openTag/closeTag/textFragments. With some basic tag matching on <pre> open/close tags, we can extract the contents of <pre class="idl"> instances. We also need to filter out instances that are inside a <* class="example"> or <* class="note"> tag (as done here in the proof-of-concept code).

[mdittmer]

This work depends on first closing foam-framework/foam2#325

[mdittmer]

This work is towards #6

[mdittmer]

The proof-of-concept code that finds these <pre>-related patterns is here: https://github.com/mdittmer/web-apis/blob/master/lib/idl/memos.es6.js#L84