Re: Adding test cases to the suite to cover ISSUE-131

Ivan, Gregg,

thanks for your verifications! I have added and pushed these test
cases now (0289-0292). I followed Ivan's advice and made two test
cases for @about (one for just the overriding and one to verify its
use in nested descriptions).

This should complete my ACTION-113.

I've also run the thus updated public test suite against our
implementations. (As mentioned I had to update clj-rdfa a bit.) That
one and pyRdfa pass these new tests, but RDF.rb fails on #0290: "@href
becomes subject when @property and @datatype are present". This is
about the spec fix (in step 11) to also account for an empty @datatype
when determining the object.

Best regards,
Niklas


2012/3/23 Ivan Herman <ivan@w3.org>:
>
> On Mar 22, 2012, at 22:57 , Niklas Lindström wrote:
>
>> Hi all!
>>
>> I'm catching up on my ACTION-113 [1] to add the test cases needed to
>> cover the resolution of ISSUE-131 [2].
>>
>> 1. The first part should be straightforward. I'll just add:
>>
>>    <a href="http://example.org/" property="rdf:value"
>> content="value">ignored</a>
>>
>> Which results in (as it always has):
>>
>>    <http://example.org/> rdf:value "value" .
>
> I am surprised that such simple case was not in the test suite but you are right. I have not found any at first glance:-(
>
> Ie: I concur, this should be added to the test suite with that result.
>
>>
>> 2. The other part is the spec fix for when @datatype is present. I'd
>> like to verify that I'm reading things correctly. I figure that this:
>>
>>    <a href="http://example.org/" property="rdf:value" datatype="">value</a>
>>
>> Should now also result in:
>>
>>    <http://example.org/> rdf:value "value" .
>>
>
> Yes, that is also correct. Either @content or @datatype sets back @property to its old self:-)
>
> Ie: I concur, this should be added to the test suite with that result.
>
>> (Before the fix, the result was that @href was used for both subject
>> and object.)
>>
>> 3. For completeness, I should also add a test to verify that @about
>> effectively overrides @href here. I.e.:
>>
>>    <a about="http://example.net/" href="http://example.org/"
>> property="rdf:value" content="value">ignored</a>
>>
>> Should become:
>>
>>    <http://example.net/> rdf:value "value" .
>
> Again, that is correct, and I concur with a new test
>
>>
>> I also figure that this results in <http://example.net/> being used as
>> the parent resource for nested statements, correct? Thus, I should
>> probably extend this to be:
>>
>>    <a about="http://example.net/" href="http://example.org/"
>> property="rdf:value" content="value one">
>>      <span property="rdf:value">value two</span>
>>    </a>
>>
>> Yielding:
>>
>>    <http://example.net/> rdf:value "value one", "value two" .
>>
>
> Right again:-)
>
> I would propose to keep those two tests separately rather than merge them, b.t.w.
>
> Ivan
>
>
>> If you all agree that I'm interpreting this correctly, I'll promptly
>> add these test cases and include them in the manifest.
>>
>> (To pass them I had to amend clj-rdfa. Fairly simple, but not yet
>> committed as I want this confirmed first.)
>>
>> Best regards,
>> Niklas
>>
>> [1]: http://www.w3.org/2010/02/rdfa/track/actions/113
>> [2]: http://www.w3.org/2010/02/rdfa/track/issues/131
>>
>
>
> ----
> 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
>
>
>
>
>

Received on Friday, 23 March 2012 19:22:37 UTC