- From: Henry Story <Henry.Story@Sun.COM>
- Date: Mon, 08 Sep 2008 11:09:19 +0200
- To: Jean-Marc Vanel <jeanmarc.vanel@gmail.com>
- Cc: public-cwm-talk@w3.org
I think what you would like is cwm enhanced with POWDER rules. Then you could express that using it's vocab. http://www.w3.org/2007/powder/ There would then be a question as to whether adding shortcuts to the N3 language to make that easier to express would be a good idea. Henry On 8 Sep 2008, at 10:57, Jean-Marc Vanel wrote: > > Hi > > Would it make sense to add some kind of joker in N3 queries ? > What do I mean with "joker" ? > > Something like : > > ?X a myPrefix:* . > > where * stands for any resource name in that namespace . > > I guess the answer is no ... > Indeed, there is nothing like that in http://www.w3.org/TeamSubmission/n3/ > . > > My use case is the following . I'm using Euler Proof Mechanism > <http://eulersharp.sourceforge.net/> with its new Coherent Logic > engine to infer (or instantiate if you prefer) Java objects > represented by triples such as : > > _:sk2 a http://java.sun.com/class#myPackage_MyClass . > > and I'd like to get hold of all those triples by this query : > ?X a http://java.sun.com/class#* . > > The alternative for me would be to assert all subClassOf relations for > the Java classes I use , and query this : > ?X a http://java.sun.com/class#java_lang_Object . > > Any idea welcome. > > Notes: > - I tried to find "standard" URI's for representing Java classes but > couldn't > - I was obliged to use underscore instead of point as a separator of > Java packages because CWM ( and Euler ) refuse points in Qnames > > -- > Jean-Marc Vanel > http://jmvanel.free.fr/ >
Received on Monday, 8 September 2008 09:10:19 UTC