[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Types and Context
- From: Jeni Tennison <email@example.com>
- To: Jonathan Borden <firstname.lastname@example.org>
- Date: Tue, 22 May 2001 15:25:21 +0100
> actually no. the entities can be defined in an external DTD subset.
> only this would need to be edited to add an element to the
> 'this.person' pattern, but this is only an example to demonstate the
> ability to do this with XML 1.0.
I understand that this is only an example. I was talking about using
XML Schema as the schema syntax, and the point that I was making is
that having a label for the type allows you to have one edit (that
affects the schema) rather than two (whether the second edit is in the
local or remote DTD for the stylesheet). Or perhaps there's some way
of mechanically link the name of a type in the schema to the names of
the entities that you use in the stylesheet?
If you were using Schematron, you could share entity definitions of
this kind between Schematron and the stylesheet, which would be quite
a nice way of adding more abstract/general information to Schematron.
> To be clear about this, what I have suggested can be done _today_
> with todays tools. There is no need for XSLT/XPath 2.0. True, an
> extension XSLT function would be very helpful, but that hardly ought
> be a complete version increment on XPath itself.
I'm not sure whether you're arguing against there being an XPath 2.0
or just asserting that you can do today all that you need with
XSLT/XPath. I certainly won't argue against the latter, if that's what
you mean, but surely the point of the next version of XSLT/XPath is to
make it *easier* to do the difficult things (including accessing
schema information, although it's only a 'should' in the requirements,
not a 'must').