[sc34wg3] CTM should be context free

[Lars Heuer]

> [...]
>> As implementor I would not so much worry about the
>> context-sensitiveness.  That can be handled efficiently. But the
>> conceptual picture a user should be as context-free as possible.

> Right. But it is a bad thing if mailto:donald at duck.org is sometimes an
> IRI and sometimes a QName. I know, you'll agree.

[Bad example here]

Better example:

       = undeclared-prefix:bla {   !! "undeclared-prefix:bla" is treated
                                   !! as IRI
       }

       
       %prefix undeclared-prefix = http://example.org/

       = undeclared-prefix:bla {   !! "undeclared-prefix:bla" is a
                                   !! QName!
       }


A parser would never detect an undeclared prefix, but would silently
assume that it is an IRI.

Best regards,
Lars
-- 
http://www.semagia.com