W3C home > Mailing lists > Public > public-xml-processing-model-comments@w3.org > January 2008

Re: Comments on Editor's Draft 9 January 2008

From: Norman Walsh <ndw@nwalsh.com>
Date: Fri, 25 Jan 2008 12:41:09 -0500
To: public-xml-processing-model-comments@w3.org
Message-ID: <m2abmterbe.fsf@nwalsh.com>
/ Toman_Vojtech@emc.com was heard to say:
| So, just to make sure I understand the implicit input/output ports on
| p:pipeline (or that I am totally confused, depending on how you look at
| that :), are the following the following statements true?
| 1. Inside p:pipeline, explicit bindings cannot be provided for implicit
| input/output ports. In other words, the following is not allowed:
| <p:pipeline>
|   <p:output port="result">
|     <!-- "not-matched" is a non-primary output port of p:split-sequence
| -->
|     <p:pipe step="split" port="not-matched"/>
|   </p:output>
|   <p:split name="split" test="..."/>
| </p:pipeline>
| and should be rewritten as follows:
| <p:pipeline>
|   <p:split name="split" test="..."/>
|   <p:identity>
|     <p:input port="source">
|       <p:pipe step="split" port="not-matched"/>
|     </p:input>
|   </p:identity>
| </p:pipeline>

Or rewritten to use p:declare-step instead of p:pipeline. The
p:pipeline is really just for the very simple pipelines.

| 2. Implicit input/output on p:pipeline accept sequences of documents.


| 3. If the implicit input/output port of a pipeline is not bound to any
| document, it contains an empty sequence.

It is always an error to have an unbound input, so the pipeline
processor must provide something for the 'source' input. The output
'result' is bound to the primary output port of the last step. It's an
error if the last step has no such port.

                                        Be seeing you,

Norman Walsh <ndw@nwalsh.com> | Some people will never learn anything,
http://nwalsh.com/            | for this reason, because they
                              | understand everything too soon.-- Pope

Received on Friday, 25 January 2008 17:41:40 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:28:25 UTC