Re: Deterministic behavior in processors

John Cowan (cowan@locke.ccil.org)
Wed, 05 Aug 1998 13:53:40 -0400


David Brownell wrote:

> And maybe more. For example, it's very useful to see the
> first validation error be "no DTD provided"!! :-)

My point is that there is no such Validation Constraint.
A WF document that doesn't have a DTD will always provoke
at least one "VC:Element Valid" error, since a WF document
has to have at least one element. But having a DTD
is not *as such* a VC.

It's true that there's nothing in the spec preventing
parsers from reporting errors where there are no errors:
presumably that is a QOI issue.

> > Note that clause 1.2 says validation errors should be
> > reported at user option, whereas clause 5.1 says validation
> > errors must be reported, period.
>
> The XML editors should know about such internal inconsistencies
> in the spec, and address this in the errata or a forthcoming
> revision of the document.
>
> > I note that "fatal errors" are of only three kinds: failure to be
> > WF, an encoding declaration specifying an encoding the processor
> > cannot handle, and disallowed entity references (no unparsed entity
> > refs, no general entity refs in the DTD, no external entity refs in
> > attribute values).
>
> But (following on an earlier thread) if you don't handle
> external entities, you're not required to report all WF
> errors ... sigh.
>
> - Dave

-- 
John Cowan	http://www.ccil.org/~cowan		cowan@ccil.org
	You tollerday donsk?  N.  You tolkatiff scowegian?  Nn.
	You spigotty anglease?  Nnn.  You phonio saxo?  Nnnn.
		Clear all so!  'Tis a Jute.... (Finnegans Wake 16.5)