Re: What about XQuery? (#issue-whichlangs, #issue-conformance-labels, #issue-output-formats)

On Fri, 2007-07-20 at 12:52 -0400, Jonathan Robie wrote:
> Dan Connolly wrote:
> > On Thu, 2007-07-19 at 13:17 -0400, Jonathan Robie wrote:
> > [...]
> >   
> >> I think it would be simple to change this to say "XSLT and XQuery are 
> >> specifically designed to express XML to XML transformations and has some 
> >> good safety characteristics" or add a sentence like: "XQuery, another 
> >> language designed specifically for such tasks, is not yet used in 
> >> existing GRDDL implementations."
> >>     
> >
> > OK, I salted to taste a bit:
> >
> > "While technically Javascript, C, or virtually any other programming
> > language may be used to express transformations for GRDDL, XSLT is
> > specifically designed to express XML to XML transformations and has some
> > good safety characteristics; XQuery has similar characteristics to XSLT,
> > though use of XQuery in GRDDL implementation is less widely deployed at
> > the time of this writing."
> >  -- GRDDL spec, editor's draft
> >   http://www.w3.org/2004/01/rdxh/spec#txforms
> >   1.288  2007/07/19 21:21:00
> >
> >
> >
> >   
> >> P.S.,  as you probably guess, I'm not satisfied with your answer.
> >>     
> >
> > How about now?
> 
> As far as XQuery is concerned, I think this is good now, because XQuery 
> uses the same data model as XSLT, and it's easy to infer the details 
> from here. Thanks!
> 
> I'm not at all sure how GRDDL would expect a C or Javascript 
> implementation to work, given the general framework, and I can't quite 
> tell whether your text is trying to say that GRDDL supports any 
> language, or that these other languages aren't really directly suitable 
> so you don't try to offer support for them.  Saying one of these two 
> things more clearly would be helpful.

A few paragraphs later, after giving the formal rule for the XSLT
case, we say...

"Non-XSLT transforms may indicate the RDF graph in some other,
unspecified, fashion."

Perhaps that's what you're looking for?

The actual conformance clause for a GRDDL-aware agent is later
in the document; the relevant part is:

"a GRDDL-aware agent should ...
Find each transformation ...
Selectively apply any or all discovered transformations to obtain GRDDL
results. Note selection may be guided by the agent's capabilities, local
security policies and possibly user/client intervention."

i.e. GRDDL-aware agents may support any transformation language(s).
How XSLT works is fully specified. All the test cases and
examples the WG provides are XSLT, because that's all we have experience
with, but we leave open the possibility of agents that support
other languages.

Is that satisfactory?


For reference, the relevant WG decisions and issues are:

http://www.w3.org/2001/sw/grddl-wg/issues#issue-conformance-labels
resolved 2007-02-21 to have a conformance clause for GRDDL-aware
agents as above.

http://www.w3.org/2001/sw/grddl-wg/issues#issue-whichlangs
resolved 2006-08-30 to recommend XSLT and allow other transformation
languages

http://www.w3.org/2001/sw/grddl-wg/issues#issue-output-formats
resolved 6 Dec 2006-12-06 to specify the XSLT case completely


-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Friday, 20 July 2007 17:30:09 UTC