xml-dev - Re: SAX2: Namespace Processing and NSUtils helper class

[james anderson <James.Anderson@mecomnet.de>]

hmm..., the thread is becoming rather tenuous, but, if you concur, that
{"","a",""}, which is descibes <... xmlns="" ... > is not the same as {
undeclared, "a", valid-prefix }, then how does the notion of nonconformance
enter into the question re

> > > {"", "a", ""} == {"", "a", "html"}

?

David Brownell wrote:
> 
> james anderson wrote:
> >
> > i had understood
> >   {"","a",""},
> > from the example, to denote a name in a namespace for which there would have
> > been a declaration, namely a binding such as
> >
> >   <... xmlns="" ... >
> >
> > for the ignominious "null" namespace. how does this come to correspond to
> >
> >   { undeclared, "a", valid-prefix } ?
> 
> It doesn't.  Namespace URIs clearly have three kinds of values, not that
> it's specified in the namespace spec very clearly (and I suspect DOM L2
> may need tweaking to get this right):
> 
>         - Undeclared (null?)
>         - Declared as no-value (empty string, "")
>         - Some URI (the case folk focus on).

don't they have four?
          - Undeclared (not "null"; that term already has a different meaning
in the REC)
          - Declared as "no namespace":      xmlns=""
          - Declares as the empty string:    xmlns:html=""
          - Some URI (the case folk focus on).

> 
> The distinction between "undefined" and "empty" will be familiar to folk
> in the database world (maybe with different terms).  SAX2 will have to be
> clear on how it exposes this.
> 
> One more edge case for applications to trip over!
> 
> - Dave
> 
> > i agree that an "==" implementation needs to account for this, but didn't
> > think it was nonconformant.
> >
> > David Brownell wrote:
> > >
> > > David Megginson wrote:
> > > >
> > > > Clark C. Evans writes:
> > > >
> > > >  > On Sun, 26 Dec 1999 uche.ogbuji@fourthought.com wrote:
> > > >  > >
> > > >  > > {"http://www.w3.org/1999/xhtml", "a", ""} ==
> > > >  > > {"http://www.w3.org/1999/xhtml", "a", "html"}.
> > > >  >
> > > >  >          {"","a",""} != {"","a","html"}
> > > >
> > > > As I understood it, the suggestion was that
> > > >
> > > >   {"", "a", ""} == {"", "a", "html"}
> > >
> > > Actually I think the namespace spec defines
> > >
> > >     { undeclared, "a", valid-prefix }
> > >
> > > as "namespace-nonconformant", AKA some kind of error which they
> > > neglected to describe well enough to support portable APIs.
> > >
> > > Meaning that any application choosing to use such a "Name"
> > > class (which IMHO is a fine notion) needs to treat this value
> > > with care ... IEEE floating point math has a variety of sorts
> > > of "NaN" values, perhaps useful models can be found there.
> > > I'd suggest that most "==" tests against such a Name ought to
> > > fail, and ordering tests should cause exceptions.
> > >
> >


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