- From: Stéphane Corlosquet <scorlosquet@gmail.com>
- Date: Tue, 22 Jul 2014 00:20:46 -0400
- To: Reece Dunn <msclrhd@googlemail.com>
- Cc: Gregg Kellogg <gregg@kellogg-assoc.com>, Alex Miłowski <alex@milowski.com>, W3C RDFWA WG <public-rdfa-wg@w3.org>
- Message-ID: <CAGR+nnFv1r8eYGYQcB+hMvNT1A5tt4jehgrWQK+t_hwE2N5AJQ@mail.gmail.com>
On Thu, Jul 3, 2014 at 6:11 AM, Reece Dunn <msclrhd@googlemail.com> wrote: > On 3 July 2014 06:19, Gregg Kellogg <gregg@kellogg-assoc.com> wrote: > > > > > > On Wednesday, July 2, 2014, Alex Miłowski <alex@milowski.com> wrote: > >> > >> I don't see any problem with XML has a host language. There are > >> plenty of XML vocabularies that will benefit from RDFa. In fact, RDFa > >> is being added to DocBook and will be valid DocBook for version 5.1. > > > > > > It's not the XML is a bad host language, but that this test isn't setup > to > > run in XML mode. It could be if @lang were changed to &xml:lang, but that > > may not be the point if the test. Easiest thing would be to just remove > XML > > from the set if host languages for this particular test in the test > > manifest. > > > >> > >> Meanwhile, the test seems just incorrect. The only language attribute > >> available that is universally recognized is xml:lang. > >> > >> The simple solution is to correct the input document in the test case. > > > > Either way, perhaps the test author can chime in with specifically what > the > > purpose of the test is. > > I did not write that particular test (scor did), but it is related to > tests I did write (0330 and 0331). The original discussion was at > http://lists.w3.org/Archives/Public/public-rdfa-wg/2013Apr/0012.html. > > These were created for the web page > http://reecedunn.co.uk/espeak-for-android (HTML5 page) for which the > rdf/rdfa ruby module incorrectly extracted the: > > <li content='af' datatype='dct:RFC5646' > property='s:countriesSupported'>Afrikaans</li> > > metadata as the page had <html lang="en"> declared at the top (i.e. it > used the lang property, not the datatype property as other tools did). > > I am happy for: > > 1. these tests to be restricted to the HTML (and possibly the XHTML > tests); > 2. a new set of tests based on 0330-0332 using xml:lang instead. > Would having the polyglot notation with both lang="en" and xml:lang="en" solve the problems here? or would lang="en" still make XML unhappy? > > NOTE: The 0332 test references 7.5 step 11 in the specification. > > Thanks, > - Reece > > -- Steph.
Received on Tuesday, 22 July 2014 04:21:13 UTC