Re: Comments on RELAX NG 2001-06-08

[Murata Makoto <mura034@attglobal.net>]

Rick Jelliffe wrote:

> > In an early version of RELAX Core, I did the same thing.  But this makes
> > type assignment for simple types ambiguous, since two <attribute> for the
> > same attribute name can have different datatypes.  In particular, Relaxer
> > had a big problem about such duplicated declarations.
> 
> So this is an implementation issue, in large part?

No.

> If foo is present, the interpretation I suggest is that next must come
> (bar11, bar21).   I am not suggesting (bar12,21) | (bar12, bar22).
...
> Why isnt the appropriate derivative the following?  (eliding
> the choice)  Isn't this what would happen if foo was an element?
> 
> <group>
>    <ref name="bar11"/>
>   <ref name="bar21"/>
>  </group>

Your definition of <choice> looks very unnatural to me.  <choice> means 
that one of the two possiblities is satisfied.   Existence of an 
attribute matching <attribute name="foo"/> does not elminate the second 
possibility.  

Rick Jelliffe wrote:

> Sorry being thick, but I don't understand this answer at all.

I don't understand your proposal either.  You might want to propose 
a rewrite of the inference rule for <choice> (see the formal spec).

Cheers,

Makoto