xml-dev - Re: Mapping XHTML to XLink via Architectural Forms

[Jonathan Borden <jborden@mediaone.net>]

Simon St.Laurent wrote:
> At 10:14 AM 10/27/00 -0400, Jonathan Borden wrote:
> >    On the other hand if what we need to do is 'in-place' renaming of
> >attributes (or elements for that matter) so that <a href="...foo">
becomes
> ><a xlink:href="...foo"> AF is the easiest (and thus in my book best) way
to
> >accomplish this, with the strong added benefit that it is already
specified,
> >standardized and implemented.
>
> I'm still pondering your proposal to move forward, but wanted to add
> another data point.
>
> Eric van der Vlist raised some issues [1] with the recent working draft of
> 'Syntax of CSS rules in HTML's "STYLE" attribute' [2].  HTML, XHTML, SVG,
> and MathML all use this 'style' attribute.
>
> It's not entirely clear what the namespace status of that attribute is
> (since the default NS doesn't apply to attributes), but it's reasonably
> clear that it hasn't been defined as a 'global attribute' as described in
> (non-normative) A.2 of Namespaces in XML. [3]
>
    I agree that this is another wrinkle on the same problem. Curious...I
think there is no question about the namespace status of the "STYLE"
attribute under any circumstance. I'm not exactly sure what the
(non-normative!) meaning of a "global attribute" is, but I think it is quite
clear from XML Names that the attribute "STYLE" or "CLASS" has an empty
namespace-uri() and a local-name()="STYLE", or another way of saying this is
that the "STYLE" attribute is not within a namespace REGARDLESS of which
element it is declared within, or the presence of a default namespace on
that element.
    Processors are free to *interpret* the attribute in any way they wish,
for example to consider and process "html:STYLE" and "STYLE" the same if the
containing element is within the html namespace. This has caused a great
deal of confusion in RDF. The fact that a language processor can 'equate'
attributes for some purpose, does not mean that the attributes have the same
expanded name. For example:

<html:foo xmlns:html="http://example.org" STYLE="bar" html:STYLE="baz" />

is ALWAYS well-formed XML and

<html:foo xmlns:html="http://example.org" xmlns="http://example.org"
STYLE="bar" html:STYLE="baz" />

is also always well-formed XML even if either or both may not conform to a
particular specification of be valid to a particular schema.

(Hammering on this issue because it keeps on rearing its ugly head.)

Jonathan Borden
The Open Healthcare Group
http://www.openhealth.org