- From: Emmanuel Pietriga <epietriga@yahoo.fr>
- Date: Thu, 23 Oct 2003 10:38:50 +0200
- To: Dominique Hazaël-Massieux <dom@w3.org>, www-rdf-interest@w3.org
- Cc: Reto Bachmann-Gmuer <reto@gmuer.ch>
Dominique Hazaël-Massieux wrote: > Le lun 20/10/2003 à 14:17, Reto Bachmann-Gmuer a écrit : > >>I'm trying to develop a tool for rendering RDF-Resource to text based >>formats such as HTML or XML. >> >>The first alpha-implementation written in java is available for download at: >>http://prdownloads.sourceforge.net/knobot/wymiwyg-rdfstyles.zip?download >> >>The rdf-schema can be found at: http://wymiwyg.org/ontologies/rdfstyles.rdf > > > Unless I misunderstood, this is very close to GSS, an RDF style sheet > language used by Isaviz: > http://www.w3.org/2001/11/IsaViz/gss/gssmanual.html > > It would be interesting to see the two projects work together on this. The current version of GSS (which is an RDF-based stylesheet language for styling RDF) is aimed at styling node-link representations of RDF graphs using CSS and SVG properties. So it's not addressing exactly the same issue, but there is clearly an intersection between the two. Some precisions about GSS: GSS also has a cascading mechanism. The selector language defined by GSS is somewhat inspired from the RDF reification mechanism, so it tends to be verbose, and I'm not really happy with it (hopefully, the graphical front end for creating GSS stylesheets should hide the complexity of selectors). But GSS would clearly benefit from an XPath-like RDFPath language. As far as the execution model is concerned, we can draw a (loose) parallel with XSLT: the GSS engine walks the entire graph (nodes and arcs) and tries to match them with the selectors defined in the stylesheet. Styling instructions associated with matching selectors are executed on the current node (or arc). Note that there is no mechanism similar to apply-templates yet as it is less obvious in the context of a graph structure. If several selectors are match the current node, there is an XSLT/XPath-inspired algorithm for computing the "longest match", meaning that the most specific selector is applied (*). (*) actually, it is somewhat more complex than that, as the less specific selectors might still have some effect if they set a styling property that has not been defined by more specific selectors. I still have to catch on this RDFStyles thread, but I would be glad to look further if we can collaborate on this. Note: Ryan Lee has written an OWL Schema for GSS, available at [1] (informal Notation 3 version at [2]), and as Dominique pointed out, there is more information about IsaViz and GSS at [3] and [4]. [1] http://www.w3.org/2001/11/IsaViz/graphstylesheets [2] http://www.w3.org/2001/11/IsaViz/graphstylesheets-informal.n3 [2] http://www.w3.org/2001/11/IsaViz/ [3] http://www.w3.org/2001/11/IsaViz/gss/gssmanual.html Emmanuel -- Emmanuel Pietriga (epietriga@nuxeo.com) tel (mobile): +33 6 88 51 94 98 http://claribole.net
Received on Thursday, 23 October 2003 04:38:05 UTC