W3C home > Mailing lists > Public > public-rif-wg@w3.org > February 2008

Re: namespace prefix definitions in presentation syntax

From: Axel Polleres <axel.polleres@deri.org>
Date: Fri, 29 Feb 2008 19:24:46 +0000
Message-ID: <47C85BFE.6000006@deri.org>
To: Michael Kifer <kifer@cs.sunysb.edu>
CC: axel@polleres.net, "Public-Rif-Wg (E-mail)" <public-rif-wg@w3.org>

Michael Kifer wrote:
> The RIF specification says that curies are NOT to be used in XML

Sure, in XML syntax I can use the "regular" namespace mechanism of XML. 
no problem, I know how to do this.

> only
> absolute URIs. In the presentation syntax they are used as an informal
> shorthand.

So, if I understand this correctly -  that there are no curies in the 
presentation syntax - then what I just suggest is just to formalize the 
informal shorthand use in RIF presentation syntax a bit, by adopting one 
of the ways other standards define it... shouldn't cause major trouble, 
or no? I am having major pain writing down examples otherwise...

best,
Axel


> 	--michael  
> 
>> Dear all,
>>
>> I just tried to write down some RIF presentation syntax... (I know I 
>> shouldn't do this) ... and realized that we don't have any way to define 
>> namespace prefixes there although we use curies allover...
> 
>> So I thought we shall reuse one of the established syntaxes here instead 
>> of reinventing the wheel.
>>
>> 3 alternatives come to my mind:
>>
>> TURTLE:
>>
>> @prefix prefix: <IRI>.
>>
>> SPARQL:
>>
>> PREFIX prefix: <IRI>
>>
>> XQuery:
>>
>> declare namespace prefix="IRI"
>>
>> for the moment, I suggest to us the latter, because
>> we never use angle bracketted IRI's  in the document.
>>
>> BTW: We might want to switch to angle brackets in the end, because
>> with curies alone, we cannot express all URIs (that is acxtually a 
>> severe problem in RDF/XML, btw, where properties can only be CURIEs)
>>
>>   e.g. assume I have a datatype "http://mydatatype.com/"
>>
>> I cannot write:
>>
>>    "12345"^^http://mydatatype.com/
>>
>> because we have only CURIEs in the datatype position (didn't see other 
>> examples so far) and the pain thing is that http://mydatatype.com/ 
>> cannot really be synbtactically be split into a QName.
>>
>> obviously, we don't want to end up in
>>
>>    "12345"^^"http://mydatatype.com/"^^rif:iri
>>
>> which BTW could then be written as:
>>
>>    "12345"^^"htttp://mydatatype.com/"^^"http://theIRIforRifIRI"^^rif:iri
>>
>> which actually could be written as:
>>  
>> "12345"^^"ttp://mydatatype.com/"^^"http://theIRIforRifIRI"^^"http://theIRIforRifIRI"^^rif:iri
>>
>> ...
>>
>> right?
>>
>> Axel
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> -- 
>> Dr. Axel Polleres
>> email: axel@polleres.net  url: http://www.polleres.net/
>>
>> rdfs:Resource owl:differentFrom xsd:anyURI .
>>
>>
> 
> 


-- 
Dr. Axel Polleres
email: axel@polleres.net  url: http://www.polleres.net/

rdfs:Resource owl:differentFrom xsd:anyURI .
Received on Friday, 29 February 2008 19:25:04 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:33:45 GMT