Re: ANN: SAX Filters for Namespace Processing

[Paul Tchistopolskii <pault12@pacbell.net>]

----- Original Message ----- 
From: John Cowan <cowan@mercury.ccil.org>
To: Ronald Bourret <rpbourret@rpbourret.com>


> Ronald Bourret scripsit:
> 
> > 1) Is there any well-formed XML that a DTD can't model? I can't think of
> > any (namespace issues aside).

> I don't understand the point of this question.  

Neither do I. One may write some DTD for XSL FO, but it is impossible 
to express the actual constraints of XSL FO WD using the DTD syntax.

> You can obviously look at
> a WF document and construct a DTD that would make it valid, thus:
> 
> 1) Declare every element type, giving it a content model of the form
> ((A, B, C, ...) | (D, E, F ...) | ..., in other words, a choice between
> sequences, where each sequence represents the actual children that
> every element of that type actually has.  Alternatively, give them all
> a content model of ANY.
> 
> 2) Declare every attribute, giving it a type of CDATA and a value of #REQUIRED
> (or #IMPLIED, or #FIXED "actual-value", or "actual-value", as you will).

Declare every element as 'ANY' and every attribute as CDATA.
You're joking, right ? 

I mean ... such a DTD is useless, right?

Rgds.Paul.