Closed chaals closed 6 years ago
I am fine with the merge of the PR.
A somewhat tangential comment. The document adds a reference to the microdata-rdf note. However, that note may need an update, in particular regarding the experimental features that are part of that note that have never materialized. A back link to this microdata spec may also have to be added and updated.
The state of the Semantic Web IG is very much in flux these days, so I am not sure re-publishing that note there is a good idea (if at all possible). Would it be a reasonable idea to re-publish an updated version of that note as a Working Group Note, alongside the publication of the microdata standard, by the Web Platform WG?
Cc: @danbri @gkellogg
The RDFa conversion needs a couple of tweaks, but, when the result is processed by a confirming HTML5+RDFa processor, including post-processing, it reproduces the native transformation results, with the addition of the rdfa:usesVocabulary triples. The JSON-LD conversion algorithm can’t do this without using the separate property value extraction defined in Microdata to RDF.
We could essentially move over the test suite from Microdata to RDF and obsolete the Note, or reduce it to just the property value extraction.
There are a couple of other uses of the direct alogorithm, but it is of marginal use now.
@gkellogg isn't it correct that the microdata-rdf note includes a number of additional features to set the datatype of the value depending on the element that is used, and that is not part of the document of @chaals? Or do you want to propose to carry that over?
I would be fine carrying those over, but I do not think that was @chaals' idea. In which case obsoleting the current Note does not look like a good idea...
The Microdata to RDF note includes the same features for extracting data type from elements and attributes that HTML5+RDFa (other than some minor differences in numeric value detection), so they amount to the same thing.
The only real addition I can think of is for the object element, for which RDFa does not examine the data attribute, bull I’ll do a PR for that.
Oops, I forgot about that, @gkellogg, you are right...
Clarify that this is a minimum required conversion, allowing for richer ones.
fix #79