Prepare ExtractorEngine to do the structured annotation extraction as well

Like for PkPass, we'll just add a dummy extractor that the extractor
engine will ignore and proceed with structured data extraction based
on HtmlDocument, rather than our regex/XML stream parser hack.