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

Re: RDF testcase rdfms-rdf-names-use/test-025.rdf

From: Brian McBride <bwm@hplb.hpl.hp.com>
Date: Tue, 03 Dec 2002 09:37:07 +0000
Message-Id: <5.1.0.14.0.20021203092243.01f32e50@0-mail-1.hpl.hp.com>
To: Chet Murthy <chet@watson.ibm.com>, Chet Murthy <chet@watson.ibm.com>
Cc: www-rdf-interest@w3.org, Dave.Beckett@bristol.ac.uk

At 13:33 02/12/2002 -0500, Chet Murthy wrote:


>[I'm not sure this is the right place to send a note like this -- if
>not, please advise.]
>
>I'm starting down the RDF road, writing tools, etc, and am at the
>point of writing tools which will pass the testcases in
>RDF-TESTCASES.  I was beavering down this path, when I got derailed by
>the subject testcase.

In cases like this its helpful to folks if you could include the uri of the 
test case causing problems.  It just makes it easier to locate the problem.

As it happens the rdf you quote below does not match the test you quote in 
the subject line:

   http://www.w3.org/2000/10/rdf-tests/rdfcore/rdfms-rdf-names-use/test-025.rdf

Could you please clarify.



><rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
>   <rdf:Description rdf:about="http://example.org/node1">
>     <rdf:li rdf:resource="http://example.org/node2"/>
>   </rdf:Description>
></rdf:RDF>
>
>Now, I also looked at the RDF issues list, and as far as I can tell,
>it stipulates there that the above RDFXML is in error.
>
>It seems to me that the above is equivalent to
>
>
><rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>          xmlns:eg="http://example.org/">
>   <eg:node1>
>     <rdf:li rdf:resource="http://example.org/node2"/>
>   </eg:node1>
></rdf:RDF>

Sorry no.

Best to think of this in terms of the graph which I'll write in n-triples.

The first of these is equivalent to :

   <http://example.org/node1> rdf:_1 <http://example.org/node2> .

Your version is equivalent to:

   _:something rdf:type http://example.org/node1 .
   _:something rdf:_1   http://example.org/node2 .

These are quite different.

Where you have to switched to the typed node production, e.g. where you 
introduced

   <eg:node1>

is a shorthand way of saying that a resource is of a particular type, not a 
shorthand for giving the name of the resource.

[...]


>The upshot, it seems to me, is that the rdf:li property, anywhere,
>_forces_ a per-node counter to be incremented, in a manner which
>assigns to each rdf:li property an ascending ordinal name.

Sorta.  rdf:li is a purely syntactic thing, like a keyword in a programming 
language.  This is turned into the appropriate rdf:_nnn property by the 
parser, as you say, incrementing as it goes.


>The only explanation I can see for arguing that rdf:li, as a property
>of an "rdf:Description" node, is that such a node is intrinsically

I can't parse that sentence.

>_different_ from the abbreviation in which its rdf:type is converted
>into a namespaced XML tag, (e.g., "eg:node1" above).
>
>So, in sum, I don't understand why the Ntriple output for the first
>RDFXML is
>
><http://example.org/node1> <http://www.w3.org/1999/02/22-rdf-syntax-ns#li> 
><http://example.org/node2> .

Its not, using an abbreviated form of n-triple its what I gave above.


>rather than
>
><http://example.org/node1> <http://www.w3.org/1999/02/22-rdf-syntax-ns#_1> 
><http://example.org/node2> .

Which is the same as this.

Brian
Received on Tuesday, 3 December 2002 04:35:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:57 GMT