Re: Matching a choice grammar

Raxit,

I agree with your interpretation.

Conclusion 1:
The behavior should be as I described in my first reaction in this thread - 
see http://lists.w3.org/Archives/Public/www-voice/2009AprJun/0003.html

Conclusion 2:
The phrase
"else if ( the utterance matched a grammar belonging to a <choice> )"
in Appendix C could be perhaps reworded so that it states explicitly that 
this branch applies only in cases when the interpreter is in the <menu> to 
which the mentioned <choice> belongs.

Pavel


raxitsheth2000@gmail.com wrote:
>  >The problem is that there is no definition of the "menu grammar" term. 
> The
> 
> the term "Menu grammar" is in context of ***Grammar scope*** and not in 
> context of XML Schema/XSD/parent/child etc.
> It may not be direct children, it may be indirect children and hence may 
> fall in "Scope" for Menu grammar
> 
> 
>  >text you cited could suggest that there can be some grammars placed as
>  >direct children of the <menu> tag. These could be called "menu grammars".
>  >However, this is not true - see the following snipped of the VoiceXML 2.0
>  >schema (available at
> 
> it may be ....
> <menu>
> <choice>
> <grammar some stuff. not exact syntax !!! >
> ...
> ...
> </grammar>
> </choice>
> </menu>
> 
> 
> Anyway as i describe above, one should not check this stuff with xml 
> schmea/dtd etc as this is in context of "Grammar scope".
> 
> 
> 
>  >Then the "Grammars contained in menu choices cannot specify a scope; if
>  >they do, an error.badfetch is thrown." sentence implies that they cannot
>  >specify a scope because the scope for them is specified globally in the
>  >respective <menu> tag.
> 
> I think above para will be out of context (or there is no ambiguity in 
> that part of the spec.)
> 
> 
> -Raxit Sheth

Received on Wednesday, 22 April 2009 10:01:15 UTC