* Output any image candidate markers generated by PageImages
wikimedia/mediawiki-extensions-PageImages#768464d changed how lead images are identified within PageImages, and since then an HTML comment in the form of `MW-PAGEIMAGES-CANDIDATE-$ID` is output by`doParserModifyImageHTML`, where the ID is an index in the extension output data's list of image metadata. These comments should find a way into the ParserOutput as the extension looks for those comments in a chosen document fragment when identifying qualifying candidates out of all images gathered through `onParserModifyImageHTML`.
This change does not affect compatibility with older PageImages (only the "new" hook's being called already).
* Default htmlAfter to null and update tests to expect it
Follow-up to my previous PageImages commit.
