General

• Home
• Basic XPath demo
• Object Oriented XPath
• XML Schema for metadata interoperability
• An Application to a Rights Expression Language
• Implicit Query Transcoding
• Behind the Scenes

 

Behind the Scenes. XML/RDF Syntax

The Contorsion XPath processor works over an RDF-OWL inference engine (borrowed from the Jena API). We translate XML and XSD files to RDF and OWL files respectively, but not under a structure-mapping approach (e.g. translating each 'movie' element into a 'movie' RDF resource) but under a model-mapping approach. We have designed an ontology to represent the XML document model (the XQuery 1.0 and XPath 2.0 Data Model). Instances of this ontology are RDF representations of XML document trees, so the ontology defines some kind of XML/RDF Syntax (reversing the W3C terminology for RDF/XML Syntax).

The first step of the processor is to translate the XML instance to RDF. We use an XSLT stylesheet to perform this task.

Select an example none Simple XML example

XML instance	XML/RDF Syntax

To allow the processor to take care of schema constructs we must translate the XSD instance to OWL.

Select an example none Simple XSD example

XML Schema instance	XML/RDF Syntax

Finally is time to translate the XPath query into some RDF query language. We have choosed RDQL, and we have designed an algorithm to map the different axis, nodetests and predicates of XPath 2.0 to this language. Fortunately, the most part of of the XPath constructs have equivalent properties in our XML/RDF ontology, so the translation algorithm results very simple, and the hard part of the computation is performed by the inference engine.

Select an example none XPath example

XML Schema instance XPath query	RDQL Translation