[xml-dev] namespaces v. processing instructions for processorinvocation

[Paul Schofield <atobenews@hotmail.com>]

Hi,

I am implementing a (hopefully) flexible XML transformer that uses external
context information to define how the transform will work. At the heart of
the transformer is an XSLT processor, however the style sheets are hidden
and are defined by a sort of "meta" style sheet that is context aware.

Although the engine works I am in doubt about which is the best way to
define the meta file location from an XML file.

My first preference was to use namespaces -

<myxml xmlns="http://www.client.com/data"
             xmlns:meta="http://www.ourcompany.com/context-transform"
             meta:location="http://www.client.com/context-file.meta" >

  <stuff>bla bla</stuff>

</myxml>

As my processor "knows" what to do with elements and attributes in the
http://www.ourcompany.com/context-transform namespace it can processes the 
XML accordingly. I can also create other processors in the same way
creating a processing chain.

However XSLT uses a processing instruction to define the location, using
this model my data will be -

<?context-transform href="http://www.client.com/contextfile.meta" ?>
<myxml xmlns="http://www.client.com/data" >
  <stuff>bla bla</stuff>
</myxml

Which method, if any, is considered best practice, are there better methods?

If the namespace method is best why is XSLT implemented otherwise?

Regards
Paul Schofield


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp