Re: [xml] Correct behaviour of RelaxNG interleave

[Daniel Veillard <veillard redhat com>]

On Wed, Jan 24, 2007 at 03:30:07PM +0100, Bjoern Hoehrmann wrote:
> * Bjoern Hoehrmann wrote:
> > Your test case somewhat simplified:
> >
> >  a = element a { empty }
> >  b = element b { empty }
> >  c = element c { empty }
> >  d = element d { empty }
> >  
> >  e = a & b
> >  f = c & d
> >  
> >  start = element root { e & f }
>
> As Daniel asked, this would be
>
>   <grammar xmlns="http://relaxng.org/ns/structure/1.0";>
>     <define name="a">
>       <element name="a">
>         <empty/>
>       </element>
>     </define>
>     <define name="b">
>       <element name="b">
>         <empty/>
>       </element>
>     </define>
>     <define name="c">
>       <element name="c">
>         <empty/>
>       </element>
>     </define>
>     <define name="d">
>       <element name="d">
>         <empty/>
>       </element>
>     </define>
>     <define name="e">
>       <interleave>
>         <ref name="a"/>
>         <ref name="b"/>
>       </interleave>
>     </define>
>     <define name="f">
>       <interleave>
>         <ref name="c"/>
>         <ref name="d"/>
>       </interleave>
>     </define>
>     <start>
>       <element name="root">
>         <interleave>
>           <ref name="e"/>
>           <ref name="f"/>
>         </interleave>
>       </element>
>     </start>
>   </grammar>


  It doesn't look obvious to me that
     <root><a/><c/><b/><d/></root>
should be accepted. This is equivalent to

<element name="root">
  <interleave>
     <interleave>
       <element name="a">
         <empty/>
       </element>
       <element name="b">
         <empty/>
       </element>
     </interleave>
     <interleave>
       <element name="c">
         <empty/>
       </element>
       <element name="d">
         <empty/>
       </element>
     </interleave>
  </interleave>
</element>

I don't see the reduction rule which would be equivalent to flattening out
to a single interleave. The interleaves are at that point in a binary form,
explain because I don't see how your draw that conclusion.

Daniel

-- 
Red Hat Virtualization group http://redhat.com/virtualization/
Daniel Veillard      | virtualization library  http://libvirt.org/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine  http://rpmfind.net/