Re: RetrievalMethod with #frag

On Friday 15 August 2003 07:12, Kevin Burton wrote:
> I am a bit confused on how a RetrievalMethod with a URI of say
> '#myid' should be handled, could someone point out my error.

Hi Kevin, I'm a bit rusty on the spec, and as you noted I don't think 
RetrievalMethod got lots of exercise in examples ...

> The spec says the URI is dereferenced as per a Reference URI.
> Section 4.3.3.2 says the result is the identified element + all
> descendants excluding comments. So, if no transforms are
> specified we have a nodeset of the identified element and its
> descendants.

Right.

> Section 4.4.3 then says the result for KeyInfo's that have an XML
> structure is an XML element. Clearly this is not the case for
> this example. 

Is this solved if it were to say "a node corresponding to an element or a 
document." ?

> Does this imply only certain forms of
> RetrievalMethod can be successfully processed, if so how do you
> determine what is valid and what is not. Perhaps the intent is
> that you use Type attribute to perform some form of validation
> on the result to see if matches your expectation for such a key?

Yes, I think the expectation is that you do this, you've asked for something 
with an identifier, and the identifiers definition should specify what you 
should expect back.

Received on Friday, 22 August 2003 21:34:47 UTC