Re: Help with test case 105 and 106

Micah Dubinko wrote:
> My parser is picking up the blank node as expected, but also separately 
> completing the triple based on @href. Why is this wrong?

What is the second triple you are getting, exactly?

> Could one generalize to say that non-CURIE (and also non-html-reserved) 
> rel values can be treated as if they weren’t there? If not, what is the 
> right generalization?

It looks like the two test cases are very useful :)

The right interpretation is that the *presence* and *value* of @rel play 
different roles. Specifically, the *presence* of @rel is what completes 
the hanging triple, while its *value* determines the next triple. So 
rel="foobar" will still complete a parent triple, but since its value is 
not a CURIE, it doesn't create a second one.

Why, you might ask?

Because in a future version of RDFa, we might want to actually support 
more reserved values for @rel. But that should only decide the presence 
or absence of that one triple whose predicate is defined by @rel, not 
the structure of the other triples around it.

-Ben

Received on Monday, 4 August 2008 04:40:19 UTC