> From: Erwin [mailto:[EMAIL PROTECTED]]
> 
> Hello,
> 
...
> 
> I downloaded the latest release (2.0.2) and have studied the sitemap
dtd
> and the sitemap schema
> dtd: src/documentation/xdocs/dtd/sitemap-v02.dtd
> schema: /src/documentation/xdocs/drafts/sitemap-working-draft.xsd
> but I havent found an answer to my questions.
> 
> 
> Question 1
> 
> In the dtd I found the following:
> <!ELEMENT map:match (map:mount | map:redirect-to | map:call |
map:generate
> | map:transform | map:serialize | map:read | map:aggregate | map:act |
> map:match)*>
> 
> Do I misinterpret the dtd if I read it as follows (without the
> map-namespace): a match element can have multiple generate child
elements
> as well as multiple serialize elements?

Only one generate and serialize child directly nested into match is
allowed.


> (I also looked it up in the schema, but imho the schema is not yet
> complete)
> If I read it correctly, then could someone give me an example of how
you
> could have multiple generate elements as direct children of a match
> element. (or this it a somewhat unlucky formulation due to the
limitations
> of dtd?).

I would say that this is dtd limitations plus the fact that dtd defines
just some of the sitemap rules, and plus the fact that this dtd is not
up to date.

 
> Question 2
> 
> Also from the dtd:
> <!ELEMENT map:pipelines (map:pipeline*)>
> <!ELEMENT map:pipeline (map:match*, map:handle-errors?)>

This one is also not correct. It should be something like:

<!ELEMENT map:pipeline ((map:mount | map:redirect-to | map:call |
map:generate | map:transform | map:serialize | map:read | map:aggregate
| map:act | map:match)*, map:handle-errors?)>


> (the schema formulation is equivalent - not quite 100% the same, but
close
> enough for my question)
> 
> As far as I have understood everything, the actual pipeline is defined
> within the match-elements (although the element-names suggest
otherwise).

See above. 


> So my question is: why do the match-elements have to be children of
the
> pipeline element and can't they be children of the pipelines-element?
I
> see no use for such an architecture, aside from having different
> error-handlers per pipeline (but that is the only use is see for
having
> this construction). Or is this defined in this way, because of these
error
> handlers?

Yes. Also, map:pipeline has internal-only attribute.

 
> Also, if you have two pipeline elements as children of the
> pipelines element and inside each pipeline element you have a match
for
> e.g. index.html, then what choice does C2 make? Does it choes the
first
> one defined or what is its behaviour?

In the current implementation, it will execute sitemap linearly till
complete pipeline is being assembled. If first match does not have
serializer (pipeline with generator must always be completed with
serializer), it will proceed and execute second match also.

Vadim
 
> That are my questions for now.
> I would also like to thank everybody in advance.


---------------------------------------------------------------------
Please check that your question has not already been answered in the
FAQ before posting. <http://xml.apache.org/cocoon/faqs.html>

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

Reply via email to