Re: [JSON] Examples for RDF JSON serializations side-by-side comparison

On Wed, 2011-03-23 at 13:55 +0000, Andy Seaborne wrote:
> 
> On 23/03/11 13:21, Thomas Steiner wrote:
> > Hi all,
> >
> > As per ACTION-20 (http://www.w3.org/2011/rdf-wg/track/actions/20) I am
> > to create a side-by-side comparison table for the proposed RDF JSON
> > serializations so far. I have come up with two examples, one basic,
> > and one advanced example, and wanted to get your feedback on the
> > samples. Are they complex enough? Do they miss an RDF feature you
> > would like to see covered? Are they too real-world-ish or too
> > theoretic? Please feel free to edit/add your feedback on the Wiki
> > section:
> >
> > =>  http://www.w3.org/2011/rdf-wg/wiki/TF-JSON#Sample_Graphs<=
> >
> > I will take whatever we have agreed on by Saturday, and try to convert
> > the two examples in a couple of JSON docs.
> >
> > Best,
> > Tom
> >
> 
> Real world is more useful in seeing what the options might look like - 
> I've tried here to write out all the features I can think of as 
> synthetic test data for coverage.
> 
> 
> 
> @prefix rdfs:    <http://www.w3.org/2000/01/rdf-schema#> .
> @prefix xsd:     <http://www.w3.org/2001/XMLSchema#> .
> @prefix : <http://example/> .
> 
> :x1 :p1 1 ;
>      :p2 1.0 ;
>      :p3 1.0e0 ;
>      :p4 () ;
>      :p5 (1 2 "three" ) ;
>      :p6  [ :q 57 ; :q 89 ] ;
>      :p7 _:a .
> 
> _:a  :p1 :z ;
>       :p2 "str" ;
>       :p3 "str"^^xsd:string .
> 
> []  :p :z .
> 
> :z rdfs:label "Swansea"@en , "Abertawe"@cy ;
>     :q1 "2011-03-23T13:40:22.489+00:00"^^xsd:dateTime ;
>     :q2 "2011-03-23Z"^^xsd:date ;
>     :q3 "2011-03-23"^^xsd:date ;
>     :q4 "2011+01:00"^^xsd:gYear ;
>     :q5 "2011"^^xsd:gYear .
> 
> ("a" "b" "c" ) :p [ :p (1 2) , (3 4) ] .
> 
> [ :p (1 2) , (3 4) ] :p ("a" "b" "c" ) .
> 
> # These are not strict Turtle where everything
> # must be "subject-predicate" form.
> # They are SPARQL.
> 
> ("a" "b" "c" ) .

This should be flagged a little more clearly as "NOT RDF".  A mapping
can reject this one and still be 100% RDF.

> [ :p (1 2) , (3 4) ] .

Oh that's really not in Turtle?   Sad, if true.

> 
> Attached:
>    D.ttl (the data above)
>    D.nt (converted to N-triples)

Brilliant!   And cruel!    I'm trying to imagine Tom sitting on the
plane sorting this out against the various specs.    He might find he'd
rather just write the code.

   -- Sandro

Received on Wednesday, 23 March 2011 14:47:04 UTC