[dsdl-discuss] Syntactical sugar for Namespace Routing Language

"Attach" feels funny: not sure if it is the idea or the syntax. And the current text
for "attach" is not as luminous as James' usual prose. Perhaps
might be clearer to allow nested modes and so remove one level of indirection.

In this example of a possible improved syntax, I have
  - replaced the namespace element with a <section> element
  - allowed multiple <namespace> elements as children under <section>
  - made an element for attaching elements, called <also> to replace
    the deep <attach> which seems in the wrong place
  - thereby allowed nested <mode>s

 <rules startMode="root"
       xmlns="http://www.thaiopensource.com/validate/nrl">
  <mode name="root">
    <section>
      <namespace>http://www.w3.org/1999/xhtml</namespace>
      <validate schema="xhtml.rng">
        <also>
            <context path="head">
               <mode name="rdf+links">
                <section>
                  <namespace>http://www.w3.org/1999/02/22-rdf-syntax-ns#</namespace>
                  <namespace>http://www.w3.org/???/????/xlink</namespace>
                  <validate schema="rdfxml.rng" />
                </section>
              </mode>
            </context>
        </also>
      </validate>
    </section>
  </mode>
</rules>

The idea would be that a context or mode under a <validate> would create
a separate section, while an <also> makes the sections validate in parallel.
Something like that.

It seems there are two different problems being solved with NRL's attach:
one is that we may have many namespaces being validate by the same
or many schemas, the second is that we may have subelements
of a different namespace being validated by the same schema.

I think the first would be better handled by allowing multiple namespaces
at the same level as multiple validations. And the second would be
(from my POV) more pleasant if we had sugar to avoid references.

Cheers
Rick Jelliffe

--
DSDL members discussion list
To unsubscribe, please send a message with the
command  "unsubscribe" to dsdl-discuss-request@dsdl.org
(mailto:dsdl-discuss-request@dsdl.org?Subject=unsubscribe)