interned names [Re: Next Round]

[james anderson]

John Cowan wrote:
> 
> james anderson wrote:
> 
> > Neither the prefix nor the qualified name have the same permanence as the
> > local name, the namespace and the expanded name. One could well collect all
> > prefixes in connection with which as symbol appeared, but the values are of no
> > use as the bindings have dynamic extent.
> 
> Granted, but it is still necessary to publish the current
> prefix->URI map, as things that are semantically interpretable by a

It should well be published as an interface which makes use of the
prefix<->uri map, with the restriction that either it depends the dynamic
context - and as such is unambiguous - or falls back on the cumulative
bindings - and as such has no guaranteed result.
To publish the prefix<->uri binding as an aspect of a name is misleading.

> specific application as QNames may appear in #PCDATA content, CDATA
> attribute values, PI content, or other such places where
> an unassisted parser cannot detect and decode them.  The application
> must therefore determine the URI itself.

To depend on the prefix-URI bindings outside of the dynamic context is to
permit erroneous results. It suffices to permit the application to process
names in the parser's dynamic context. To this end it need no explicit access
to the prefix<->uri map, just to an interface which interns qualified names
given the parser's dynamic context.

Reserialization is handled by letting the parser generate unique names for
namespaces as discovered and using those names as prefixes when encoding.

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/
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)