W3C home > Mailing lists > Public > www-rdf-interest@w3.org > March 2002

Re: XSLT transforms application/xml+rdf to RDF database API calls

From: Dan Brickley <danbri@w3.org>
Date: Sat, 16 Mar 2002 16:45:55 -0500 (EST)
To: "Eric Prud'hommeaux" <eric@w3.org>
cc: <elenz@xyzfind.com>, <www-rdf-interest@w3.org>, <w3c-semweb-ad@w3.org>
Message-ID: <Pine.LNX.4.30.0203161605330.25060-100000@tux.w3.org>


This *is* *very* cool. But it's also the **4th** (incomplete) XSLT parser
for RDF that I'm aware of. Without detracting from the hackvalue coolness and
the hard work that went into this, I fear this emphasises a couple of
concerns about the Semantic Web / RDF developer community.

(i) we tend towards a focus on 'proof of concept' with minimal follow-thru
to production grade delivery (with increasingly numerous honourable
exceptions. we need to move beyond this.

(ii) while it's understandable for non-academics to not bother with a trip
to the library for a full literature review, not doing a Google search
before coding is rather old fashioned. cf. http://www.google.com/search?q=rdf+xslt+parser

For those not keeping track, the other partial RDF XSLT parsers being
Jason Diamond's, Dan Connolly's and Jeremy Caroll's.

So... are there any completer-finisher types out there fancy doing a
compare and contrast, and making at least one RDF XSLT parser that does
the whole RDF syntax spec?

See http://www.xmlhack.com/read.php?item=757

	New XSLT RDF Parser
	14:32, 11 Sep 2000 UTC
	Dan Brickley, urged the two authors to bring one of these tools to completion:
	I'd really like to see one of both of these efforts finished to
	completion, as part of our attempt to pin down the various issues/problems with the spec. [...]
	In particular, Bag/Seq/Alt handling seems crucial if the XSLT parsers are
	ever going to be used in anger.

Raw materials:

	Transforming RDF with XSLT
	$Revision: 1.5 $ of $Date: 2001/05/25 13:18:15 $ by $Author: connolly $
	no support for collections
	@@need to report "no endmarker" bug.
	no support for quoted (reified) statements
	no support for parsetype="literal" nor parsetype="resource"

	XSLT RDF Parser
	From: Jason Diamond (jason@injektilo.org)
	Date: Mon, Sep 11 2000
	see also http://www.injektilo.org/rdf/rdft.html
	with a slight variation at
	...to support N-Triples (Jason -- did this get folded into copy on your site?)

	Snail - Excruciatingly Slow RDF Parsing, Jeremy Carroll
	(interestingly different).

	The current implementation has no parseType="Literal" or
	parseType="daml:list" or typedNode ID=foo support. We'll get to that.

Getting *at least one* full-featured RDF parser in pure XSLT (whether via
API calls or by transform into an XML or text triple syntax) would be
extremely useful. I want one for my Ruby RDF tools, for example.

So what would it take to get someone to finish this off? Beer? A W3C T-shirt
bribe? These things can be arranged... Maybe this would suit a student
project? I'd be happy to help supervise one if anyone on this list has a
student who'd be interested in production-gradifying some of this stuff.



(your RDF Interest Group chair who thinks its time we shipped Semantic Web v1.0)

Received on Saturday, 16 March 2002 16:45:56 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:07:39 UTC