test case for transformation identification vs. contained data

My action from a couple of weeks ago was to put together this test
case and pass it to bwm who would check it, then check it into the
CVS. I'm not sure I've remembered the case properly, so it'll also
need sanity checking, and the files involved will need a little URI
adjustment to reflect their host.

In essence I believe the test would demonstrate the GRDDL-aware agent
only lifting the transformation statement from an (indirect) profile
doc, ignoring any other data there.

Attached is a HTML document which can be used as a GRDDL profile doc,
although it contains other data available via GRDDL.

If I remember correctly (not altogether likely) the test case input
will be an instance such as:

   http://dannyayers.com:88/xmlns/hdoap/samples/redland-doap.html

which refers to the profile (there's a redirect on the profile URI
which should put it in the right place).

The output will be the GRDDL result, *not including* any other RDF
contained in the profile doc, hence:

   http://dannyayers.com:88/xmlns/hdoap/samples/redland-doap.rdf

The profile doc is done using XMDP but contains some of the DOAP RDFS,
embedded as eRDF and should be GRDDLable, see:

   http://research.talis.com/2005/erdf/extract

(ouch, the redirect messes up the URIs, ah well)

btw, there's a bug in:

http://www.w3.org/2003/12/rdf-in-xhtml-xslts/grokXMDP.xsl

s/rdfs:Property/rdf:Property

- I believe that would otherwise produce much the same result from the
profile doc as the eRDF XSLT.

Cheers,
Danny.

-- 

http://dannyayers.com