SUMMARY: XML Validation Issues (was: several threads)

[Rick Jelliffe]

From: Tim Bray <tbray at textuality.com>

> ..but I think there are 2 separate questions here:
>
> - should a document be able to ask for validation
> - should a document be able to ask for guaranteed reading of all
>   external entities

I think these questions in turn boil down to that the XML spec's section
2.9 "Standalone Document Declarations" says that it is a "validity
constraint" if standalone="no" and the markup declarations contain
certain kinds of data, but instead it should be a kind of
"well-formedness constraint"!

This in practise would create three classes of processors:

1) XML parsers which cannot accept documents which have
standalone="yes" and which have markup declarations in the internal
prolog (the URI on the DOCTYPE declaration reliably names the document
type);

2) XML parsers which only accept documents which are standalone="yes",
whether it then validates content models or not;

3) XML parsers which accept standalone="yes" or "no", whether it then
validates or not.

In other words, if standalone="no", then a parser must read all the
external entities and handle the particular markup declarations; if it
cannot, it should spit the Draconian dummy. This does not mean it has to
validate against content models, however.

If standalone="yes", then a parser should read the internal markup
declarations and use them (for the particular cases mentioned: to get
default attribute values, to get entity values).

Tim's comment in his annotations presages this approach. If people are
finding this a big problem, can XML be tightened up ASAP to correct it?

Rick Jelliffe


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)