- From: Dave Beckett <dave.beckett@bristol.ac.uk>
- Date: Tue, 22 Jan 2002 20:32:32 +0000
- To: Sjoerd Visscher <sjoerd@w3future.com>
- cc: www-rdf-comments@w3.org
>>>Sjoerd Visscher said: > > Other specifications such as the > > > > XQuery 1.0 and XPath 2.0 Data Model (Working Draft 20 December 2001) > > http://www.w3.org/TR/2001/WD-query-datamodel-20011220/ > > > > define their Infoset filtering operation simply: > > > > "Other information items and properties made available by the > > Infoset processor are ignored." > > -- Appendix A: XML Information Set Conformance > > http://www.w3.org/TR/2001/WD-query-datamodel-20011220/#d2e4846 > > > This is not true. They very precisely define the transformation. The above quote is true, however I wasn't going to quote the whole formal part that explains the detail that you include: > > For example: > http://www.w3.org/TR/2001/WD-query-datamodel-20011220/#infoitem-to-node > clearly defines what is a node and what is not: > > else > /* infoitem-kind(i) = "doctype" | "notation" | "unparsed-entity" */ > empty-sequence() > > And in > http://www.w3.org/TR/2001/WD-query-datamodel-20011220/#infoitem-to-element-node > > kids := collapse-text-nodes(sequence-map(infoitem-to-node, > infoset-element-children(e))), > > I think it's really important that the spec provides exact > transformations from the XML semantics to the RDF semantics. > Ofcourse you could use the RDF/XML syntax to represent both the xml > infoset and the rdf triples, and provide an XSLT tranformation, but that > wouldn't be very helpful I guess. I previously mailed that I would update the RDF/XML syntax draft to define the transformation for all Infoset items - what items are used and what are not. For example, the text node collapsing part up there is already written in words in the existing draft: http://www.w3.org/TR/2001/WD-rdf-syntax-grammar-20011218/#section-text-node based on the words in XPath 1.0: http://www.w3.org/TR/1999/REC-xpath-19991116#section-Text-Nodes However it may be too terse in the current draft and I already noted I'll look to see what is needed to be said is omitted / ignored from the Infoitems. The XPath 1.0 form (words) has proved sufficient and precise enough to implement XPath, XSLT etc. from. It isn't clear that it is really necessary to require the use of another formal mapping language in this document (there are already too many), or require the use of the above draft XQuery 1.0 / XPath 2.0 Data Model in order to implement RDF/XML. Requiring the use of the latter would in particular probably require waiting for it to be a finished RECommendation. I'll get back to you when I complete and expand the mapping from Infoitems to address the points you bring up, although this might be until the next published working draft is ready. There is some work that can't be done just yet since the RDF Core WG is still working on how 'rich' XML literals maps to the RDF model, but when we do, that will be added to the draft. By rich I mean XML content that have be mixed, may include xml:lang attributes, namespace declarations, xml:base, XML Schema datatypes, ... and so on. Dave
Received on Tuesday, 22 January 2002 15:32:35 UTC