- From: Dan Connolly <connolly@w3.org>
- Date: Thu, 31 May 2001 11:15:37 -0500
- To: Brian McBride <bwm@hplb.hpl.hp.com>
- CC: Aaron Swartz <aswartz@upclink.com>, Jan Grant <Jan.Grant@bristol.ac.uk>, timbl@w3.org, RDFCore Working Group <w3c-rdfcore-wg@w3.org>
Brian McBride wrote: > > Dan Connolly wrote: > [...] > > Actually, that smiley-point is well-made: this testing format > > shouldn't depend on all the RDF/n3 specs, code, and > > tutorials, which are in flux... > > Yes > > [...] > > > > We've got terms of the form > > _:name for "anonymous" terms > > <absURIref> for URIs > > "lskdjf" for string literals. > > How would we handle relative URI's, e.g.: > > <rdf:Description rdf:ID='foo'/> I'm proposing that they get absolutized in the expected results; in this case: http://example/whatever-the-base-is#foo The input to the test is an XML document, and one of the properties of an XML document is its base URI. (I suppose some tests might have syntax errors at the XML level; but in that case, there are no expected results anyway.) This might be somewhat tedious: if/when we move the tests, we have to updated the expected result; copying them to local disk has to be done in such a way that the test harness remembers where it came from; etc. But I think the alternatives are all worse. -- Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Thursday, 31 May 2001 12:17:06 UTC