It would be good if you can get a copy of the Nuance Grammer Developer Guide. (From Nuance)
There is a lot within the guide such as prefix to use and not to use that will help you build the Grammer you require.
______________________________
You can refer directly to compiled dynamic grammars in a dynamic grammar
database by specifying the keyword dgdb: followed by the database key and
database descriptor. The syntax for the dgdb: URI is:
dgdb:?key=val1&dbdesc=val2
where val1 is the database key and val2 is the database descriptor for the
grammar. Both val1 and val2 must be URI-encoded, which means that no
illegal characters are used. The list of characters that cannot be used include all
characters other than:
¡ a-z
¡ A-Z
¡ 0-9
¡ ; / ? : @ & = + $ , - _ . ! ~ * ' ( )
The illegal characters must be converted to hexadecimal and written as %xx. For
example, the space character, ASCII 32, becomes "%20". For more information
about URI encoding, see RFC 2396 (www.ietf.org/rfc/rfc2396.txt, section 2.4)
For example, in the following reference, the portion in bold must be
URI-encoded:
dgdb:?key=PeterGrammar&dbdesc=provider=fs,root=/usr/dbs,name=db1,
class=dgdb
To convert a DBDescriptor object to the URI-encoded string required by the
dgdb: URI scheme, use the Nuance DBDescriptor function
DBDescriptorToQueryString(). See the Nuance API Reference for more
information about this function.
