Hi Santiago,
Thanks for your response. Scott has raised the question of XML 1.1
support in XSLT and XPath 2.0 with the working groups.
Perhaps it would be reasonable to wait for XPath and XSLT 2.0 before
supporting fully the Namespaces 1.1 form of QNames and NCNames. However,
given that XSLT 1.0 already provides a means for a stylesheet to request
that the result document be output as XML 1.1 and permits a processor to
do that, perhaps I should have split this into a set of cascading
questions, based on the amount of invention required, to determine users'
requirements:
1. Should we add support for XML 1.1 output on both the MAIN and xslt20
branches or only on the xslt20 branch? This would require no
"extensions". I call this "minimal support for XML 1.1 output".
2. If we added minimal support for XML 1.1 output on the MAIN branch,
should we add an experimental extension to control things like Unicode
normalization of output (which is something XSLT 2.0 currently has, but
XSLT 1.0 does not)?
3. If we added basic support for XML 1.1 output on the MAIN branch, would we
want to add an experimental extension to permit elements and attributes
constructed by xsl:element and xsl:attribute to have names that conform to
the XML 1.1 and Namespaces 1.1 definitions? (XSLT 1.0 does not do this,
nor does XSLT 2.0, but the latter will almost certainly need to.)
I would call 2 and 3 together "full support for XML 1.1 output".
4. If we added full support for XML 1.1 output on the MAIN branch, would
we want to add an experimental extension to permit QNames and NCNames in
patterns and other XPath expressions to be able to match the full range of
names permitted by XML 1.1 and Namespaces 1.1? (Again, XSLT 1.0 does not do this, nor
does XSLT 2.0, but the latter will
almost certainly need to.)
Thanks,
Henry
------------------------------------------------------------------
Henry Zongaro Xalan development
IBM SWS Toronto Lab Tie Line 969-6044; Phone (905) 413-6044
mailto:[EMAIL PROTECTED]
"Santiago Pericas-Geertsen" <[EMAIL PROTECTED]>
03/03/2003 09:33 AM
Please respond to xalan-dev
To: <[EMAIL PROTECTED]>
cc:
Subject: Re: XML 1.1 and Namespaces in XML 1.1 support
----- Original Message -----
From: <[EMAIL PROTECTED]>
>
> Xerces-J recently added experimental support for XML 1.1 and
> Namespaces 1.1, so Xalan-J users acquire some limited support for XML
1.1
> in the form of input documents and stylesheets. However, a serialized
XML
> document is still restricted to version 1.0. Further, the set of
> permissible characters in QNames and NCNames has been expanded by XML
1.1
> and Namespaces 1.1, but a stylesheet would not be able to match or
select
> such a QName or NCName with a NameTest.
>
Shouldn't we wait for the XPath 2.0 working group to adopt XML 1.1 before
implementing it? Has anyone evaluated the consequences of implementing an
XPath 2.0 parser (or 1.0 for that matter) based on XML 1.1?
> It seems desirable to add experimental support for these candidate
> recommendations to the Xalan-J processors; I hope that's not a
> controversial statement. What I'd like to ask is: do people have a
> preference on which branch of Xalan-J they'd like to see that support -
> only the xslt20 branch, or the MAIN branch as well?
I prefer to see this in the xslt20 branch. In fact, if people think XML
1.1
is so important, and given that XSLT 2.0 and XPath 2.0 specs are not final
yet, why don't we ask the expert groups to consider it?
-- Santiago
