Return-Path: <Giacomo.Pati@pwr.ch>
Mailing-List: contact cocoon-dev-help@xml.apache.org; run by ezmlm
Delivered-To: mailing list cocoon-dev@xml.apache.org
Received: (qmail 39987 invoked from network); 25 Jun 2000 21:26:27 -0000
Received: from stargazer.dataway.ch (HELO dataway.ch) (qmailr@195.216.64.144)
  by locus.apache.org with SMTP; 25 Jun 2000 21:26:27 -0000
Received: (qmail 5674 invoked by uid 0); 25 Jun 2000 21:26:25 -0000
Received: from zh2-9.dial.dataway.ch (HELO pwr.ch) (root@195.216.80.137)
  by stargazer.dataway.ch with SMTP; 25 Jun 2000 21:26:25 -0000
Received: (qmail 21701 invoked from network); 25 Jun 2000 21:12:39 -0000
Received: from donald.pwr.ch (HELO pwr.ch) (giacomo@10.20.30.103)
  by simba.pwr.ch with SMTP; 25 Jun 2000 21:12:39 -0000
Sender: giacomo
Message-ID: <3956762D.7A2D1198@pwr.ch>
Date: Sun, 25 Jun 2000 23:14:21 +0200
From: Giacomo Pati <Giacomo.Pati@pwr.ch>
Organization: PWR Organisation & Entwicklung
X-Mailer: Mozilla 4.72 [en] (X11; U; Linux 2.2.14 i686)
X-Accept-Language: de, en
MIME-Version: 1.0
To: cocoon-dev@xml.apache.org
Subject: Re: [C2] Matcher concept
References: <Pine.LNX.4.10.10006251437270.5977-100000@localhost.localdomain>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Rating: locus.apache.org 1.6.2 0/1000/N

Donald Ball wrote:
> 
> On Sun, 25 Jun 2000, Giacomo Pati wrote:
> 
> > While I'm considering how a sitemap class may be constructed I've
> > encountered the following problem concerning the java layout of a
> > pipeline.
> >
> > As we have proposed that a matcher implements the interface
> >
> >   public interface CompiledMatcher implements Component {
> >     public Map match(...);
> >   }
> >
> > and also consider the following snipped from a hypothetical sitemap
> >
> > <match type="user-agent" pattern="Mozilla */* *">
> >   <match type="uri" pattern="/my/page/*">
> >     <generator type="parser" src="/my/data/$1.xml/>
> >     ...
> >   </match>
> > </match>
> >
> > Now, what is the $1 representing. The browser match or the uri match?
> > Should the pipeline code provide the Map to the matcher so that they
> > could add stuff to it? And then, must a pipeline maintainer count all
> > wildcards from the topmost matcher in the pipeline to get the right
> > replacement, so that in my example the $1 should be written as $4?
> 
> perl would say that the $number convenience variables are reset after the
> next regexp with variables, so you'd need to explicitly save them if you
> need to use them later:
> 
>  <match type="user-agent" pattern="Mozilla */* *">
>    <variable name="agent" value="$1"/>
>    <match type="uri" pattern="/my/page/*">
>      <generator type="parser" src="/my/data/$variable.xml/>
>      ...
>    </match>
>  </match>
> 
> i think that's probably reasonable.

Yes.

> 
> perl also says, though, that wildcarding by itself doesn't save a
> variable - you use parentheses to indicate portions of the regexp to save,
> e.g.:
> 
>  <match type="user-agent" pattern="Mozilla ([^/]+)/* *">

We never said we can mix wildcards and regexp! So you example should be
written like:

  <match type="user-agent" pattern="Mozilla ([^/]+)/([^/]*) ([^/]*)">

because wildcards don't need parenthesis.

Giacomo

-- 
PWR GmbH, Organisation & Entwicklung      Tel:   +41 (0)1 856 2202
Giacomo Pati, CTO/CEO                     Fax:   +41 (0)1 856 2201
Hintereichenstrasse 7                     Mailto:Giacomo.Pati@pwr.ch
CH-8166 Niederweningen                    Web:   http://www.pwr.ch