Re: The processing rules for property/rel

On Oct 28, 2011, at 16:53 , Gregg Kellogg wrote:


> On Oct 28, 2011, at 6:23 AM, Ivan Herman wrote:
> 
>> Gregg,
>> 
>> I spent some time yesterday (thanks to the train ride home) on the possible implementation of the the property/rel stuff. I am afraid the processing rules on 
>> 
>> http://www.w3.org/2010/02/rdfa/wiki/RDFaLiteWithProperty
> 
> Thanks Ivan, I sketched it out without looking as deeply as you. It's on my list to implement, but so are Microdata to RDF spec changes and implementation (I'm getting seriously behind!).

:-)

> 
>> are not complete. At least for what I thought we would like to do. I take your example:
>> 
>> <div vocab="http://schema.org/" typeof="Event">
>> <a property="url" href="nba-miami-philidelphia-game3.html">
>>   <span property="description">
>>     NBA Eastern Conference First Round Playoff Tickets:
>>     Miami Heat at Philadelphia 76ers - Game 3 (Home Game 1)
>>   </span>
>> </a>
>> </div>
>> 
>> and what we want is
>> 
>> [ a schema:Event;
>> schema:url <nba-miami-philidelphia-game3.html>;
>> schema:description """
>>     NBA Eastern Conference First Round Playoff Tickets:
>>     Miami Heat at Philadelphia 76ers - Game 3 (Home Game 1)
>>   """ ] .
>> 
>> 
>> However, in this case, step #5 will also kick in before #11 that you modified. That will set the new subject to <nba-miami-philidelphia-game3.html>; in step #13, this will modify the value of parent subject as used when getting into the recursive step, which means that this will be the subject of everything down the line. Ie, there is a chaining no matter what (in the current model). 
>> 
>> If we want to define the processing as you indicate, that means modifying #5. Essentially, if there is a @property and no @rel/@rev then, I believe, what should happen is:
>> 
>> - @about or @typeof should set a new subject (and that should be valid for the @property as well) otherwise the subject is the inherited one
>> - @href/@src/@data should _not_ set the new subject (but used as an object for @property) later, in your modified version of #11.
>> 
>> Editorially, I would probably prefer to cut processing step #5 into two, b.t.w.
>> 
>> Do I miss something?
> 
> This looks right, but I wonder if there are some remaining issues this wouldn't detect. E.g.:
> 
> <div rel="foo"><a href="bar"><span property="baz"></span></a></div>

Yes. But what I described as a modification only applies in the case when there _is_ a @property on the same element. Ie, in this case, the old rule would apply.

> 
> In this case, the href _should_ create a new subject in 5, I think. This might need attention in step 12. Did you run the test suite after making your proposed change?

I have not implemented the whole thing yet. Actually, I did and then I found a different bug somewhere else in my code, which I had to take care of before coming back to this again... I hope to do that next week.

Ivan


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


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