W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > October 2002

Re: Typed literals: current status

From: Frank Manola <fmanola@mitre.org>
Date: Mon, 21 Oct 2002 11:09:55 -0400
Message-ID: <3DB418C3.36BF93AC@mitre.org>
To: Dave Beckett <dave.beckett@bristol.ac.uk>
CC: Jeremy Carroll <jjc@hpl.hp.com>, w3c-rdfcore-wg@w3.org

Hey guys.  I'm in the process of adding datatypes to the Primer, and
this (plus the message I'm responding to and the Ntriples thread) causes
me to ask for some guidance on:

a.  how to represent typed literals in Ntriples
b.  a new abbreviated form for Ntriples

To start with (b), we'd discussed abbreviating writing out full Ntriples
in examples, and I thought we'd decided that instead of writing out full
URIrefs in angle brackets, we'd write QNames without angle brackets, as
in

ex:index.html dc:creator exstaff:85740 .
ex:index.html exterms:creation-date "August 16, 1999" .

Edging a bit into (a), the obvious extension to this abbreviated form to
handle typed literals would be something like the literal immediately
preceded (or followed?) by the QName for the datatype, as in:

exstaff:85740 exterms:age xsd:integer"27"

However, various threads seem to be discussing these issues, with
various suggestions, but it's not clear there's any consensus.  So:

1.  have we decided on the syntax for typed literals in Ntriples?  If
so, what is it?
2.  given the decided syntax for typed literals in Ntriples, do you have
any suggestions for an abbreviated
    syntax (analogous the one described above) that uses Qnames? 

--Frank


Dave Beckett wrote:
> 
> >>>Jeremy Carroll said:
> >
> > Jeremy:
> > > But
> > >if we had two new types rdf:ClassicLiteral, rdf:ClassicXMLLiteral then we
> > >could move all the complexity of XML Literals into a datatype definition.e
> 
> These should be in rdfs: given in particular since rdfs:Literal already
> exists - http://www.w3.org/TR/rdf-schema/#ch_literal
> 
> > Brian:
> > > There have been several suggestions along these lines.  I would expect
> > such
> > > a proposal to get support if it can be done quickly.
> >
> > See also Patrick's table in:
> > http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Oct/0232.html
> >
> > Unless I hear arguments to the contrary (or even simply some opposition) I
> > will try writing some text along these lines tomorrow.
> >
> > Possible problems I can identify (mainly for DaveB):
> >
> > + N-triples syntax - the following two triples are identifcal:
> >
> > <a:a> <a:p> <rdf:ClassicLiteral>"foo" .
> > <a:a> <a:p> "foo" .
> >
> > Since it is work to drop the latter, I would suggest that the former (to be
> > more precise with two reserved datatype URIs) is specifically prohibited
> > (somewhat ugly).
> 
> can we stop using <qname> please?
> 
> you mean
> 
> forbid:
>  <uri1> <uri2> <http://www.w3.org/2000/01/rdf-schema#ClassicLiteral>"foo" .
> allow (prefer):
>  <uri1> <uri2> "foo" .
> 
> forbid:
>  <uri1> <uri2> <http://www.w3.org/2000/01/rdf-schema#ClassicXMLLiteral>"blah" .
> allow (prefer):
>  <uri1> <uri2> XML"blah" .
> 
> which is ugly but I'd prefer to add these rules, if we are going down
> this route.
> 
snip


-- 
Frank Manola                   The MITRE Corporation
202 Burlington Road, MS A345   Bedford, MA 01730-1420
mailto:fmanola@mitre.org       voice: 781-271-8147   FAX: 781-271-8752
Received on Monday, 21 October 2002 11:10:17 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:52:28 EDT