XML 1.0 spec appears to violate itself

[David Brownell]

Takuki Kamiya wrote:
> 
>   <!ENTITY lt     "<">

Yes, that's broken.  It's because of the WF-ness constraint that
you quoted from 4.3.2, not any other reason.

It's the case of a "broken document" ... and in the mode where a
conformant processor is not required to report the error.  Since
that entity value will not be expanded, its WF-ness isn't tested.
 

> Even if we assume "lt" could be redeclared in the internal DTD subset for now
> (I guess we can even extrapolate this is the case since the spec itself
> redeclares it) ...

It's clear that it _is_ permitted to redeclare entities (including the
predeclared ones) in the internal subset.  See section 4.6 where it
talks about various "may" (may redeclare) and "must" (must still be WF,
and not change the standard effect) cases, none of which could work
if such redeclaration were disallowed.


> I think that XML processor shall find the entity as "not well-formed" when it
> tries to expand references to "lt" (i.e. &lt; ) if it have been redeclared
> as "<", because "<" does not match the production rule for content.
> 
> Am I missing something?

Only that, as Richard implied, most XML processors treat those entities
as "predefined" and so any additional declarations are often ignored.
That's permitted; the spec says a warning "may" be issued, but they're
not required.

- Dave

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)