Load html

[gkellogg]

Load JSON-LD from HTML documents.

• Adds options parameter to documentLoader
• Uses xmldom, if loaded.
• Adds util.ParseContentTypeHeader
• Adds documentLoader implementations for xhr and node (still requires tests).
• Set default for expandAllScripts to true for flatten and toRdf.

[gkellogg]

Some tests still required for XHR and Node documentLoaders.

Some work needed to optionally append text/html (and application/xhtml+xml) to accept headers if we can parser HTML.

Perhaps some abstracted parsing support to handle other HTML parsers?

Otherwise, please see if you agree with the general direction of this work.

[gkellogg]

Perhaps the webpack build is failing because it doesn't include xmldom? Should there be a way to condition tests on this? How?

[dlongley]

Thanks @gkellogg! This is great. A few minor changes needed plus one architectural thing we need some feedback from @davidlehn on (regarding how to pull in a dom parser).

[gkellogg]

@davidlehn are we waiting for something else to merge this?

[davidlehn]

@gkellogg I was waiting on me to look into how to improve the way the xmldom lib is used. Otherwise I think patch is good. As Dave's comment mentioned, might want some sort of registered parser pattern. Need to check how the current code works with bundlers and such.