W3C home > Mailing lists > Public > xsl-editors@w3.org > January to March 2003

NCName -> QName in property functions

From: Paul Grosso <pgrosso@arbortext.com>
Date: Thu, 13 Mar 2003 05:59:46 -0600
Message-Id: <4.3.2.7.2.20030313055905.00b78fa8@172.27.10.30>
To: xsl-editors@w3.org


In 5.10.4 Property Value Functions at [0], all the functions (except 
proportional-column-width()) say they take an NCName (as defined at [1] 
in the Namespace spec) as an argument.

However, I think it makes more sense for this to be a QName (as defined
at [2] in the Namespace spec); that is, something that could have an
optional namespace prefix.  This would match what an XML attribute name 
can be which would be a logically more correct situation.  (An attribute
name is defined as "Name" [3] in the XML spec at [4], but the Namespace
spec makes it clear at [5] that all "constructs corresponding to the 
nonterminal Name" in the XML spec are really QNames as opposed to NCNames.)

The problem with what we have now can be seen when you consider that
a user might want to use an extended property in one of these functions.
An extended property has to be in another namespace, but then it has to
be prefixed.  But then you can't put it into any of these property
functions as currently restricted.

I think this is just an erratum, and we should change these function
definitions to say they take a QName.  This is an upward compatible
change and allows users to use extended properties in these property
value functions.

paul

[0] http://www.w3.org/TR/2001/REC-xsl-20011015/xslspec.html#section-N8624-Property-Value-Functions
[1] http://www.w3.org/TR/REC-xml-names#NT-NCName
[2] http://www.w3.org/TR/REC-xml-names#NT-QName
[3] http://www.w3.org/TR/REC-xml#NT-Name
[4] http://www.w3.org/TR/REC-xml#NT-Attribute
[5] http://www.w3.org/TR/REC-xml-names#dt-qname
Received on Thursday, 13 March 2003 07:16:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:59:53 GMT