Re: HTTP-in-RDF implementation

On 21 Nov 2008, at 17:37, Johannes Koch wrote:
> Toby A Inkster schrieb:
>> I realise it's past the deadline for feedback, but I have one  
>> thing I'd like to mention. Say I have:
>>     <http:Request>
>>       <http:methodName>GET</http:methodName>
>>       <http:absoluteURI>http://example.org/foo</http:absoluteURI>
>>       <http:response>
>>         <http:Response>
>>           <http:statusCodeNumber>200</http:statusCodeNumber>
>>         </http:Response>
>>       </http:response>
>>     </http:Request>
>> And I also have some triples pertaining to the document itself:
>>     <rdf:Description rdf:about="http://example.org/foo">
>>       <dc:title>Foo</dc:title>
>>       <dc:creator>John Citizen</dc:creator>
>>     </rdf:Description>
>> It would be nice if there were some sort of predicate for linking  
>> from the Request resource to the Description resource, or vice versa.
>
> I'd propose to add an http:body property with a cnt:Content (see  
> "Representing Content in RDF" <http://www.w3.org/TR/Content-in-RDF/ 
> >) object resource "http://example.org/foo" to the http:Response  
> subject.

I suppose that could work, but what about if http://example.org/foo  
is a foaf:Person? Are we then saying that a physical person was sent  
down the wire?

Anyway, to solve the need I had, I've created a tiny extension to the  
HTTP in RDF vocab:

	http://buzzword.org.uk/rdf/http-ext#
>


This gives two extra predicates:

1. http-ext:related-request - link from an rdfs:Resource to an  
http:Request. Indicates that the request is in some (unspecified) way  
related to the resource. For example, the request may have been an  
HTTP GET or POST request to the URI of the resource.

2. http-ext:http-equiv-headers - a link from a resource representing  
an (X)HTML document to a list or sequence of MessageHeaders which  
represent headers not actually transmitted via HTTP, but present as  
<meta http-equiv> elements in the (X)HTML document.

These, I'm sure, are too esoteric to be included in the official  
vocab, but I hope that by posting this message to the list, anyone  
who does need predicates like these can find them and won't have to  
re-invent them.

-- 
Toby A Inkster
<mailto:mail@tobyinkster.co.uk>
<http://tobyinkster.co.uk>

Received on Wednesday, 26 November 2008 21:28:03 UTC