Steve Dunham wrote:
> On 3/1/06, Emmanuel Pietriga <emmanuel.pietriga_at_inria.fr> wrote:
>> If your intent is to use :fallBackLens for displaying resources that are
>> not matched by any other lens, you don't need the notion of a "default"
>> lens.
>
>> :fallbackLens a fresnelLense ;
>> fresnel:classLensDomain "*"^^fresnel:fslSelector ;
>> fresnel:showProperties ( fresnel:allProperties ) ;
>> fresnel:group :group .
>
>> :fallbackLens a fresnelLense ;
>> fresnel:classLensDomain "SELECT ?x WHERE {?x ?y
>> ?z}"^^fresnel:sparqlSelector ;
>> fresnel:showProperties ( fresnel:allProperties ) ;
>> fresnel:group :group .
>
> That is my intent, unfortunately, the examples you give don't work.
> (The FSESelector gets created, but the simile fresnel code tries to
> cast it to a TypeSelector before adding it to an internal hash table.)
>
> The fresnel spec says that those selectors can only be used with
> fresnel:instanceLensDomain, not fresnel:classLensDomain.
> ( http://www.w3.org/2005/04/fresnel-info/manual/#propertyselection )
Yes, my mistake. I meant instanceLensDomain.
> I tried that, and it worked after fixing a lot of issues in the
> fresnel code.
Issues in the Fresnel code or the FSL code ?
> (It's very slow, though; it has to select every resource
> in the system and then check if the selected resource is in the list.
That's how FSL works if you don't give it a smaller subset of potential
starting points. I don't know how Ryan integrating FSL functionality
into the Simile Fresnel code, so I can't really tell you if it would be
possible to easily specify such a subset (in case you are able to
identify it). From a purely FSL perspective, our various Java
implementations should allow that through method
evaluatePath(FSLPath p, Vector startSet)
--
Emmanuel Pietriga
INRIA Futurs - Projet In Situ tel : +33 1 69 15 34 66
Bat 490, Université Paris-Sud fax : +33 1 69 15 65 86
91405 ORSAY Cedex http://www.lri.fr/~pietriga
Received on Thu Mar 02 2006 - 21:41:48 EST