- From: Dave Beckett <dave@dajobe.org>
- Date: Fri, 10 Mar 2006 14:20:57 -0800
- To: Ian Davis <iand@internetalchemy.org>
- CC: Harry Halpin <hhalpin@ibiblio.org>, 'Semantic Web' <semantic-web@w3.org>
Ian Davis wrote:
> 
> On 10/03/2006 20:38, Harry Halpin wrote:
> 
>> I was wondering, perhaps it would be useful for the W3C or some
>> standards body to "endorse" one of the simplified XML syntax choices for
>> RDF and a compact notation ala Turtle ... 
:)
> I agree. And we're close with the syntax for expressing query patterns
> in Sparql. It wouldn't be too much work to extract the RDF bits from
> that spec and package it up as an official non-xml syntax for RDF. Of
> course it would end up looking the same as Turtle given that's where it
> came from.
That's why the SPARQL triple pattern syntax changed half way through
development from the RDQL/BRQL style, and the DAWG and myself (as Turtle
editor) have spent a lot of time making these align at various points
and in both directions.  There are still a few more Turtle updates
pending (not related to SPARQL).
Just in case people weren't aware.  Here's SPARQL-the-RDF-syntax:
PREFIX foo: <http://example.org/ns>
CONSTRUCT {
  foo:a foo:b "lit"@en ;
        foo:d <http://www.example.org> .
}
if the syntax allowed omitting the actual *query* :)
The aim was roughly this; you could put any Turtle inside CONSTRUCT { }
and it would be legal in SPARQL, and put any Turtle inside WHERE { }
and it would be a legal SPARQL triple pattern.  Then add variables, and
you have a query.  [The only bit you can't put in there is the @prefix /
PREFIX].
As to making new RDF syntaxes, the RIF WG has that in their charter, at
least for one targeted at rules.
Dave
Received on Friday, 10 March 2006 22:44:32 UTC