> From: Sylvain Wallez [mailto:[EMAIL PROTECTED]] 
> Piroumian Konstantin wrote:
> 
> >Hi all!
> >
> >I've almost finished the Schema for sitemap, but I'm a bit 
> stuck with the
> >pipeline content description and would like to prove some statements:
> >
> >Terms
> >     Pipeline content - content model of 'pipeline', 'match', 'act',
> >'select/when' elements (with some restrictions also of 
> 'resource' and 'view'
> >elements).
> >
> >A pipeline can contain the following elements:
> >Direct:
> >           - mount
> >           - redirect-to
> >           - read
> >           - generate
> >           - aggregate
> >           - transform
> >           - serialize
> >Indirect:
> >           - act
> >           - match
> >           - select
> >
> >Statements:
> >        1. There can be no other elements if one of the 
> 'mount' | 'redirect-to' | 'read' is present.
> >
> 
> Yes.
> 
> >        2. If first element is one of the generating types: 
> 'generate' | 'aggregate' | 'call' then the last element must 
> be one of the outputting types: 'serialize' | 'call'.
> >  
> >
> 
> This is too restrictive, as there can be nested indirect 
> components to 
> drive the pipeline construction after the generate.

Yes, didn't think about it.

> 
> >        3. Indirect elements can appear more then once in 
> any order, but should not break generation -> 
> [transformation] -> output sequences.
> >  
> >
> 
> Is it possible to write such a rule in XML Schema ?

Not sure. I am stuck with this one. But after a little more thinking I came
to conclusion that this one is not true, e.g.:
<generate />
<transform/>
<select type="browser">
  <when>
        <serialize />
  </when>
</select>

And I was wrong thinking that pipelines should be like {aggregate | generate
| call} -> [transform | call]* -> {serialize | call} without intermediate
'select', 'match' or 'act'.

> 
> >Is the above correct and did I forget anything?
> >
> >Regards,
> >  Konstantin
> >
> >P.S. I have some more questions for XML Schema gurus, if any. Steven?
> >
> >P.S.2. Should Cocoon sitemap Schema belong to Forrest catalogue too?
> >  
> >
> 
> IMO, it belongs to Cocoon, as it has to evolve with Cocoon.

But what about the public identifier and central repository for all DTDs and
Schemas? E.g. XML Spy can get the DTD by HTTP from the server. If
xml.apache.org go Forrest then it'll be the right place for the repository
or at least provide access to sub-project repositories.

> 
> What about adding it inside cocoon.jar so that the actual schema used 
> can be obtained using e.g. 
> "resource://org/apache/cocoon/sitemap/sitemap.xsd" ?

Why not the entity catalogue? The main purpose of the Schema for me were
support in sitemap writing, so I'd like it to be in the entity catalogue and
not the sources.

Konstantin

> 
> Sylvain
> 
> -- 
> Sylvain Wallez
>   Anyware Technologies                  Apache Cocoon
>   http://www.anyware-tech.com           mailto:[EMAIL PROTECTED]
> 
> 
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, email: [EMAIL PROTECTED]
> 

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]

Reply via email to