W3C home > Mailing lists > Public > xproc-dev@w3.org > November 2008

Re: Difference between 'contained steps' and 'subpipeline'

From: Jeni Tennison <jeni@jenitennison.com>
Date: Sat, 22 Nov 2008 15:45:51 +0000
To: "David A. Lee" <dlee@calldei.com>
Message-Id: <0F7B266F-EE05-4E9D-892B-AB949E92FD62@jenitennison.com>
Cc: "XProc Dev" <xproc-dev@w3.org>


On 22 Nov 2008, at 14:39, David A. Lee wrote:
> I'm a little (ok a LOT) confused between the difference of  
> 'contained steps' and 'subpipeline'
> E.g in the spec in this part:
> 4.1
> <p:pipeline
>   name? = NCName
>   type? = QName
>   psvi-required? = boolean
>   xpath-version? = string
>   exclude-inline-prefixes? = prefix list>
>     (p:input |
>      p:output |
>      p:option |
>      p:log |
>      p:serialization)*,
>     (p:declare-step |
>      p:import)*,
>     subpipeline
> </p:pipeline>
> What's the difference between the sequence of p:declare-step and  
> subpipeline ?

It's the same as the difference between the declaration of a function  
in your programming language of choice, and the code that makes up the  
body of a function. A <p:declare-step> is a declaration of a pipeline  
that you can later invoke within the subpipeline. The subpipeline is  
the sequence of steps that is performed when the pipeline is called.

> What really confuses me is this comment : from 2.1
> ------
> Note
> User-defined pipelines (identified with pfx:user-pipeline in the  
> preceding syntax summary) are atomic. A pipeline declaration may  
> contain a subpipeline, but the invocation of that pipeline does not.
> ------
> So that last sentance ... is it saying the "subpipeline" in the  
> above p:pipeline is not executed ?
> If so what is done with it ?

This is the same as asking whether the code within a function  
definition is executed. It is executed, but only when the function  
(pipeline) is invoked. The note you quote is saying that you can't  
create user-defined pipelines that hold subpipelines *when you invoke  
them*, so you can't create user-defined pipelines that work like the  
compound steps <p:for-each>, <p:viewport>, <p:choose> and so on. But  
you can create user-defined pipelines that work like the various  
atomic steps such as <p:replace> or <p:xslt>.

Does that make any more sense?

Jeni Tennison
Received on Saturday, 22 November 2008 15:46:26 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 23:16:44 UTC