- From: Jonathan Borden <jborden@mediaone.net>
- Date: Thu, 21 Sep 2000 18:37:34 -0400
- To: <www-rdf-interest@w3.org>
- Cc: <jason@injektilo.org>
In my investigations of simplfied XML syntax for RDF, and extracting RDF from arbitrary or colloqial XML, I have now come to the conclusion that the essence of TimBL's strawman, in which rdf:parseTyle='Resource' is the default, provides the best option for extracting RDF statements from XML. http://www.w3.org/DesignIssues/Syntax.html I have incorporated Jason Diamond's rdf.xsl into a new implementation of an XSLT based RDF extractor for XML. This extractor outputs rdf:Statements. Interestingly, when the result of a transform is itself transformed, the statements are reified, hence I call this rdfExtractify.xsl: http://www.openhealth.org/RDF/rdfExtractify.xsl A brief list of features: 1) Implements XLink2RDF, now for extended XLinks as well. 2) Extracts RDF statements from plain 'ole XML 3) Extracts RDF statements from RDF with defaultParseType='Resource' (this is a param) 4) Handles collections, aboutEach and bagID, the syntax for collections is loosened so that any child element can be a member (need not only be <rdf:li>) 5) Implements QNameToURI (see http://www.openhealth.org/RDF/QNameToURI.htm) which converts namespace qualified names to URIs 6) Implements nodeIdentifier which produces an XPointer fragment identifier e.g. #xpointer(/foo[1]) or ChildSeq identifier e.g. #/1/2 (see the XPointer spec), the type of XPointer produced is set in the explicitPathIndices = 'ChildSeq' (default) param. See updated: http://www.openhealth.org/RDF/rdf_Syntax_and_Names.htm Jonathan Borden The Open Healthcare Group http://www.openhealth.org
Received on Thursday, 21 September 2000 18:29:21 UTC