W3C home > Mailing lists > Public > www-rdf-interest@w3.org > May 2000

Transforming RDF with XSLT

From: Dan Connolly <connolly@w3.org>
Date: Tue, 02 May 2000 00:01:59 -0500
Message-ID: <390E6147.A9AEE547@w3.org>
To: www-rdf-interest@w3.org
Somebody suggested more people might be interested in this hack,
so I wrote a little explanation, and here you go...

=========
Transforming RDF with XSLT
http://www.w3.org/XML/2000/04rdf-parse/

I'm trying to flesh out some of the details of TimBL's semantic web
toolbox... to
build some layers on top of RDF such as horn clauses ala SHOE (see
SHOE->RDF materials which so far just xhtml-ize SHOE). 

In order to connect RDF with other tools to work on practical examples,
I needed a
parser. XSLT is the hammer I'm using for nails of all shapes and sizes
lately, and I
figured that once I understand what the output of an RDF parser needs to
look like,
XSLT is a good platform for experimenting with various input syntaxes.

One of the more quick-and-dirty hacks I've seen in a while is DanBri's
JS RDF
thingy, based on Jan's tinyprolog, whence comes the output syntax of
this parser.
There are quite a few other RDF/prolog hacks in the RDF interest group)
lately,
though I'd rather avoid the closed-world assumption ala Algernon and
access-limited logic. 

Also, a representation for proofs is an important layer (see Edinborough
Logical
Framework (larch trait))

What: Contents

     rdfp.xsl the parser itself. 

     Try it out (if you're feeling lucky; this depends on an XSLT
servlet that runs on
     an as-is, no-warranties basis):

     Address of RDF input: 

                   

     Makefile instructions for how to invoke it 
     example 1 from the RDF spec, parsed output 

Bugs and Limitations

These are the sort of dull engineering issues that have straightfoward,
if tedious,
answers...

     no support for collections 
     no support for quoted (reified) statements 
     no support for parsetype="literal" nor parsetype="resource" 
     only does about 1/3 of the job of reporting bad syntax 
     doesn't expande URI references w.r.t. the base 

Issues raised

These are issues that came up while I was working on this...

One of the reasons for developing this thing was for testing the RDF
spec, i.e. to
answer some of my own questions about how well-specified RDF is. I found
at
least one bug, and several interesting corner cases

     suffers from a lack of declared URIs for XML Schema datatypes 
     What URI to use for anonymous nodes? (noodling on skolem
functions...) 

Dan Connolly
$Revision: 1.1 $ of $Date: 2000/05/02 04:52:47 $ by $Author: connolly $ 

=========

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Tuesday, 2 May 2000 01:02:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:43 GMT