Re: question on RDFa subject resolution

This is a bug in the rdflib implementation: it should only look for the
ABOUT attribute in the ancestor hierarchy.

Michael: we should have a test case for this, to make sure random IDs
are not being picked up by mistake.

-Ben

Bob DuCharme wrote:
> The following seems oversimplified, but it does seem to reflect the
> behavior I see in the RDFlib implementation of RDFa: when an RDFa
> processor is looking for a subject to go with a predicate and an object,
> it searches up the ancestor axis until it finds either an about attribute
> or an id attribute. The first of either that it finds becomes the subject
> for the triple.
> 
> If this is an oversimplification, what am I missing? The syntax document
> is a bit vague on the potential role of id attributes as subjects as
> compared with about attributes.
> 
> thanks,
> 
> Bob
> 
>