RE: Special cases for PrefixRewrite in C14N 2.0

> Yes this is a new requirement.
> 
> However it is problem for non disjoint subtrees too.

Not the problem of remembering the mappings. If you're within a subtree in which the prefix is declared, then you have a stack of the declarations in effect.

But with disjoint trees, and a rule that says any binding of a URI has to be to the same rewritten prefix, you must remember all those bindings for the life of the algorithm in case you happen across the URI again. I'm just asking if that new requirement is worth making the rewritten prefixes more sensible.
 
-- Scott

Received on Thursday, 28 April 2011 16:30:43 UTC