RE: Suggestion for modifying SRGS: ABNF Token

thank you for your comment on SRGS. The time for comments on SRGS 1.0 is
well past, but we can certainly note this point for a future version of
the specification.
 
scott McGlashan
Co-chair, W3C VBWG
 
 
 -----Original Message-----
From: www-voice-request@w3.org [mailto:www-voice-request@w3.org] On
Behalf Of Dean Sturtevant
Sent: Sunday, December 14, 2003 18:32
To: www-voice@w3.org
Subject: Suggestion for modifying SRGS: ABNF Token



Suggests for modification to SRGS1.0:
 
In Appendix D: Formal Syntax for Augmented BNF Form Grammars, it says
that
a Token is defined to be a Nmtoken | DoubleQuotedCharacters. From a
speech-recognition perspective, this doesn't really make sense. The XML
specifiction for Nmtoken was not designed for this purpose, and does not
fit it particularly well. In particular, if you want to use a
contraction
or possessive as a speech token, you would need to quote it according to
this specification. I propose that the allowable unquoted tokens be any
sequence of non-white-space characters that might not otherwise have
syntactic significance, to whit:
 
Any sequence of non-white-space characters not from the following set:
 
$<>()|/[]{}!*+?~:;\
 
- Dean Sturtevant