[
Lists Home |
Date Index |
Thread Index
]
David Carlisle wrote:
> People *are* indeed always using the same prefixes (xsl, xsd, xsi...)
> and declaring them in the root...
>
>some people are but much XML isn't written by people.
>If you generate a mixed namespace document with XSLT (and probably many
>other systems) the natural place that the namespace declarations appear
>is on each use not at the top of the document, as XSLT is designed to
>allow the result document to be linearised as it is generated, so there
>is no way of telling it to back up and add a namespace declaration at
>the top if you discover half way through generating your XHTML document
>that you need to add some svg or mathml or something. It was of course
>concerns such as this that led to the namespace rec allowing
>declarations mid document in the first place.
>
>
>
Sure, but the namespace rec never expressed itself about qnames in content.
If this requirement had been foreseen, I am sure the rec would have
included expanded names {iri}:localName, thereby *forcing* implementors
that want to be conformant with the spec to add a preprocessing run that
expands all prefixes.
> The annoying exceptions can very clearly be marked by any implementation
> as not supported, and it is quite easy to turn an unsupported XML
> document into a supported one (changing, moving prefixes).
>
>Easy with a text editor, but is it easy with "XML programming languages"
>such as Xquery or XSLT?
>
>
>
In that case, very easy to specify a postprocessing run that makes the
XML sane again.
Surely performance suffers, but correctness wins.
cheers,
Burak
--
Burak Emir
http://lamp.epfl.ch/~buraq
|
