- From: Alessandro Vernet <avernet@orbeon.com>
- Date: Thu, 18 May 2006 17:52:57 -0700
- To: public-xml-processing-model-wg <public-xml-processing-model-wg@w3.org>
On 5/18/06, Norman Walsh <Norman.Walsh@sun.com> wrote:
> Is this what's intended?
>
> <p:pipeline>
> ...
>
> <p:pipeline name="my:subPipe">
> <p:param name="foo"/>
> <p:input name="document" label="doc"/>
> <p:output name="result" label="out"/>
>
> <p:step name="someStepName">
> <p:param name="bar" select="$foo"/>
> <p:input ref="doc"/>
> <p:output ref="out"/>
> </p:step>
> </p:pipeline>
>
> <p:step name="load">
> <p:param name="uri" select="someURI"/>
> <p:output label="xiout"/>
> </p:step>
>
> <p:step name="my:subPipe">
> <p:with-param name="foo" select="'someValue'"/>
> <p:input name="document" ref="xiout"/>
> <p:output name="result" label="pipeOut"/>
> </p:step>
>
> ...
> </p:pipeline>
>
> How is putting p:param in the load step but p:with-param in the
> my:subPipe step easier to understand? If it's with-param, why isn't it
> also with-input and with-output?
Norm,
I apologize if I am the source of a confusion here. My understanding
of Jeni's proposal is that we would use <p:with-param> to pass
parameters:
<p:step name="my:subPipe">
<p:with-param name="foo" select="'someValue'"/>
<p:input name="document" ref="xiout"/>
<p:output name="result" label="pipeOut"/>
</p:step>
And we would use <p:param> to declare a parameter that is passed to us:
<p:pipeline>
<p:param name="foo"/>
Alex
--
Blog (XML, Web apps, Open Source):
http://www.orbeon.com/blog/
Received on Friday, 19 May 2006 00:53:06 UTC