On test #218: is this what we want?

The kind of corner case that tests reveals and LC is good for...

Test 0218 says:

<body>
  <div about ="">
    <p rel="rdf:value" inlist=""/>
  </div>
</body>

and the SPARQL correspondent is:

ASK WHERE {
  ?s <http://www.w3.org/1999/02/22-rdf-syntax-ns#value> <http://www.w3.org/1999/02/22-rdf-syntax-ns#nil> .
}

I must admit I was surprised seeing this. I then looked at the processing steps and I realized that, per processing steps, this is indeed right: the very last processing step seems to work that way. 

However... is this what we really want? Somehow, intuitively, I would have expected to generate no triples at all. After all, in my intuition, @inlist 'collects' the various generated triples to be put into a list. If there are no triples then...

So: we may have to run this through the WG. Modifying the processing steps is fairly easy: the whole last step should be put aside in case [local list mapping] is empty.

As for my own bias: I am more in favour of not to generate any triples at all.

Ivan


----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
FOAF: http://www.ivan-herman.net/foaf.rdf

Attachments

Received on Tuesday, 7 February 2012 17:19:45 UTC