> My point was that the NMTOKENS type *is* defined explicitly 
> as non-zero-length in the schema doc while the NMTOKEN type 
> is not. As you say, it is only defined as non-zero-length in 
> the grammar rules in the XML 1.0 spec and not in the schema 
> docs which is my issue. 
> 
> "Any property identified as a having a set, subset or *list* 
> value may have an empty value unless this is explicitly ruled 
> out: this is not the same as absent." Says the schema doc. I 
> imply from that it must be ruled out in the schema doc.

Well, firstly, NMTOKEN isn't a set, subset, or list, it's a string.

Secondly, a normative reference from the Schema specification to the XML
specification has the same force as spelling out the rules in the schema
specification.

Thirdly, it is in fact spelt out in the schema specification: the definition
of the xs:NMTOKEN data type in the schema for schemas has <xs:pattern
value="\c+" id="NMTOKEN.pattern">. (Search in XML Schema Part 2 for
'simpleType name="NMTOKEN"' to find it).
> 
> > for minor comments like these, the best approach is to enter a bug
> report 
> > - one per problem - in the bugzilla database.
> 
> Where is the bugzilla database to do this?
> 
http://www.w3.org/Bugs/Public/query.cgi

Michael Kay
http://www.saxonica.com/


From cmsmcq@a... Mon Jun 11 16:09:26 2007
Received: from aji.w3.org ([