W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > October 2001

RE: SYNTAX: RDF Syntax Telecon Friday

From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
Date: Thu, 4 Oct 2001 16:43:06 +0100
To: "w3c-rdfcore-wg" <w3c-rdfcore-wg@w3.org>
Message-ID: <JAEBJCLMIFLKLOJGMELDEEEJCCAA.jjc@hplb.hpl.hp.com>

DanC:
> Regardless, I just revisited my old (2000/04) RDF parser in
> XSLT, and it took me about 10 minutes to add n-triples support
> for it. I've only tested it lightly, but you're
> welcome to take a look:
>
>   http://www.w3.org/XML/2000/04rdf-parse/rdfp.xsl
>   $Id: rdfp.xsl,v 1.7 2001/10/02 17:36:49 connolly Exp $
>
>   (nodes/doc: http://www.w3.org/XML/2000/04rdf-parse/)
>
>

DanC:
>
> And, at the end of the day, XSLT is turing-complete. So
> an RDF/XML->RDF/n-triples mapping in XSLT is just another
> program. If we were particularly confident, we could go
> so far as to call it a reference implementation. I'm not
> sure we'll get that far.

I wish to avoid a reference implementation. We need to have clearly
articulated rules. An implementation that is a direct implementation of such
rules would then have a higher confidence built into it.

Jan:
> More strongly, the most recent anon node encountered (in a top-down
> reading of the RDF/XML) is normally the only anon node referred to, and
> only as a subject; the only exception seems to be when it is the object
> of a statement involving the previous anon node. That sounds odd, but
> see below...

Jan's stack is fairly reminiscient of the contextual stack of RDFFilter, or
the actual stack in ARP, which both store the current Resource (in ARP as a
side effect of the current Resource being an argument to the fucntion - it's
top down recursive descent).

Both Jan's and Dan's approaches come down to specifying as to how you
generate n-triples rather than specifying what n-triples are generated.

The simplifying abstract syntax approach that Dave put in his message is
more about saying that such-and-such an RDF/XML file corresponds to the same
graph as such-and-such other (simpler) RDF/XML file. By following a chain of
such simplifications we get to a very simple RDF/XML file whose graph can be
read off. This can be done in a way that doesn't prioritise any particular
path of simplification.



Snail update:
Snail is progressing slowly ... (how else :)?).
I've revised one of the design decisions: the snail compiler compiler, that
I was planning to write in XSLT exceeded my XSLT competence, so I am writing
it in Java using the DOM.

The current status is that I have:
+ a slightly revised version of the rules I have previously posted
+ some of the snail compiler compiler, which compiles the LHS of the rules
into XPath expressions,  these are in a skeleton of an XSLT file.

I would hope to have enough for review at some point next week.

At the end of the day, the snail compiler compiler will also produce an HTML
version of the rules that will roughly correspond with my earlier posting

http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Sep/att-0224/01-Trans
forming_RDF.html

Jeremy
Received on Thursday, 4 October 2001 11:43:23 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:40:56 EDT