Re: [ISSUE-131] update to NIF mapping section in spec re comments from RDF WG

Hi Dave,

Am 05.09.2013 13:19, schrieb Dave Lewis:
> Following decision on the 4th December call to opt for a query style 
> URL for the NIF string in RDF (which will also be supported in NIF) 
> when defining the mapping the following need to be changed in the spec:
>
> 1) all occurrences of RDF URLs with #char or #xpath fragments to be 
> changed to a query style as suggested by the RDF group and expanded on 
> by Felix in 
> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Sep/0000.html 
>
> i.e. all fragment identifiers for NIF strings in annex F and G should 
> be changed from, e.g.:
>
> one other associated question is, as we are using the query type to get around the limitations ofrfc 5147
> char fragment in working with XML and HTML, is it still appropriate after the above change to type the NIF string in the example with
> the subclass nif:RFC5147String? Sebastien? e.g.
>
> http://example.com/myitsservice?input=http://example.com/exampldoc.html&  <http://example.com/myitsservice?input=http://example.com/exampldoc.html&char=0,29>char=0,11
>   rdf:type nif:RFC5147String;

I am currently working on a formal ABNF definition for this, but you can 
consider it to be like this in Java:

String prefix = 
"http://example.com/myitsservice?informat=html&intype=url&input=http://example.com/exampldoc.html&" 
;
String identifier = "char=0,11" ;
String uri = prefix+identifier ;

// only identifier has to have the syntax given by the rdf:type
validate ("nif:RFC5147String", identifier) ;

So the syntax is only relevant for the identifier part. These would be 
alternative prefixes as well:
String prefixOption1 = 
"http://example.com/myitsservice/informat/html/intype/url/input/http://example.com/exampldoc.html/" 
;
String  prefixOption2 = 
"http://example.com/myitsservice?informat=html&intype=url&input=http://example.com/exampldoc.html#" 
;
String  prefixOption3 = 
"http://example.com/myitsservice?informat=html&intype=url&input=http://example.com/exampldoc.html&" 
;

It really doesn't matter and all three are valid RDF (The first one is a 
bit awkward, of course)


> 2) Once this is fixed we need to update the NIF part of the test suite and tests rerun by Felix, Leroy and Phil

As written above, this is not strictly necessary, but it is nice to be 
consistent.

>
> 3) Add the following suggested note wording to the end of Annex
> "Note: NIF allows URL for a String resource to be referenced as URIs that are fragments of the original document in the form:
> http://example.com/exampledoc.html#char=0,11
> or
> http://example.com/exampledoc.html#xpath(/html/body[1]/h2[1]/text()[1])
>
> Though this offers a potentially convenient mechanism for linking NIF resources in RDF back to the original document, the char
> fragment is defined currently only for text/plain while the xpath fragment is not defined for HTML. Therefore this URL
> recipe does fulfil the ITS requirements to support both XML and HTML and the aim of this mapping to produce resources adhering
> to the Linked Data principle of dereferenceablility. The future definition and registration of these fragment types, while a potentially
> attractive feature, is beyond the scope of this specification."


Let's change this like this:
http://example.com/doc.html#xpath(/html/body[1]/h2[1]/text()[1])
maps to
http://example.com/myitsservice?informat=html&intype=url&input=http://example.com/doc.html#char=0,11

Note that RDF is ok, with all Fragment Ids:
http://www.w3.org/TR/rdf-concepts/#section-fragID

RFC 3986 as well: http://tools.ietf.org/html/rfc3986#page-24


> The semantics of a fragment identifier are defined by the set of
>    representations that might result from a retrieval action on the
>    primary resource.  The fragment's format and resolution is therefore
>    dependent on the media type [RFC2046] of a potentially retrieved
>    representation, even though such a retrieval is only performed if the
>    URI is dereferenced.  If no such representation exists, then the
>    semantics of the fragment are considered unknown and are effectively
>    unconstrained.



The text could be like this:
"Note: NIF allows URL for a String resource to be referenced as URIs 
that are fragments of the original document in the form:
http://example.com/myitsservice?informat=html&intype=url&input=http://example.com/doc.html#char=0,11
or
http://example.com/doc.html#xpath(/html/body[1]/h2[1]/text()[1])

This offers a convenient mechanism for linking NIF resources in RDF back 
to the original document. RDF treats URIs as opaque and does not impose 
any semantic constraints on the used fragment identifiers, thus enabling 
their usage in RDF in a consistent manner. However, fragment identifiers 
get interpreted according to the retrieved mime type, if a retrieval 
action occurs as is the case in Linked Data. The char fragment is 
defined currently only for text/plain while the xpath fragment is not 
defined for HTML. Therefore this URL recipe does fulfil the ITS 
requirements to support both XML and HTML and the aim of this mapping to 
produce resources adhering to the Linked Data principle of 
dereferenceablility. The future definition and registration of these 
fragment types, while a potentially  attractive feature, is beyond the 
scope of this specification."

I will try to update the example in the spec as well.

All the best,
Sebastian

>
> cheers,
> Dave
>
>   
>
>   <http://example.com/myitsservice?input=http://example.com/exampldoc.html&char=0,29>
>
>
>
>
> On 03/09/2013 09:14, Felix Sasaki wrote:
>> 1) last call item "RDF - NIF conversion". See
>> https://www.w3.org/International/multilingualweb/lt/track/issues/131
>> and these mails
>> Phil 
>> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Sep/0001.html 
>>
>> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Aug/0066.html 
>>
>>
>> Dave
>> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Aug/0067.html 
>>
>>
>> Felix
>> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Aug/0068.html 
>>
>> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Sep/0000.html 
>>
>>
>> Goal: decide about the option 1) or 2) or something else (see a 
>> variation of option 2) in 
>> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Sep/0000.html 
>>
>> IMPORTANT: even if you are not implementing ITS <> NIF, please state 
>> your opinion since tomorrow want want to form a working group 
>> opinion, to be able to move forward.
>


-- 
Dipl. Inf. Sebastian Hellmann
Department of Computer Science, University of Leipzig
Events:
* NLP & DBpedia 2013 (http://nlp-dbpedia2013.blogs.aksw.org, Extended 
Deadline: *July 18th*)
* LSWT 23/24 Sept, 2013 in Leipzig (http://aksw.org/lswt)
Venha para a Alemanha como PhD: http://bis.informatik.uni-leipzig.de/csf
Projects: http://nlp2rdf.org , http://linguistics.okfn.org , 
http://dbpedia.org/Wiktionary , http://dbpedia.org
Homepage: http://bis.informatik.uni-leipzig.de/SebastianHellmann
Research Group: http://aksw.org