W3C home > Mailing lists > Public > www-rdf-interest@w3.org > August 2000

Re: M&S/Parser question

From: Jan Wielemaker <jan@swi.psy.uva.nl>
Date: Tue, 29 Aug 2000 17:07:34 +0200
To: Stefan Kokkelink <skokkeli@mathematik.uni-osnabrueck.de>
Cc: RDF interest group <www-rdf-interest@w3.org>
Message-Id: <00082917155906.13688@gollem>
On Tue, 29 Aug 2000, Stefan Kokkelink wrote:
>Jan Wielemaker wrote:
>
>> ...
>> Throwing it at the SWI-Prolog parser suggests what you say:
>> 
>> 'URL1'  http://purl.org/metadata/dublin_core/:Prop   each(pages)
>> 
>> (It does not expand bags into separate tripples, but rather returns
>>  a term each(BagId))
>> 
>> But looking at it, I wonder whether is correct.  Shouldn't it reify
>> the inlined description and generate a (URI, DC:Prop, <BagIdOfReified>)
>> tripple?
>> 
>> I.e. if a description is the value of an attribute, reify the
>> description and make the value the reified bag?
>
>Hm, I don't think so (although I'm not sure ;-). Without
>'aboutEach' the role of a description element is 
>quite clear: a description element represents a (single)
>resource, which can act as a target and as a source
>for arcs in the RDF graph.

Not so sure. SirPAC and the SWI-Prolog agree on the following:

<?xml version="1.0"?>
<RDF xmlns="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
     xmlns:DC="http://purl.org/metadata/dublin_core/">
<Description about="URL1">
  <DC:Prop>
    <Description about="pages">
      <DC:Creator>Ora Lassila</DC:Creator>
    </Description>
  </DC:Prop>
</Description>
</RDF>

to create (SirPAC):

triple('http://purl.org/metadata/dublin_core/Creator',
       'online#pages',
       'Ora Lassila').
triple('http://purl.org/metadata/dublin_core/Prop',
       'online#URL1',
       'online#pages').

Ok, the first sounds reasonable, but the second sounds a bit odd.
To me, the original statement is something like

``URL1 has Prop (pages has creator Ora Lasilla)''

which RDF should state using reification.  Ok, I've been on holiday
after writing the SWI-Prolog-based RDF parser and not looked in detail
at RDF syntax (only at the triples :-) afterwards, so please correct
me if I'm talking nonsense!

>When 'aboutEach' is specified,
>the resource represented by the description element is 'replaced'
>with the set of resources contained in the named container. 
>Why shouldn't the role of the resource as target 
>don't expand in the same manner?

I agree with this statement.

	Regards --- Jan
Received on Tuesday, 29 August 2000 11:16:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:44 GMT