Mailing-List: contact cocoon-dev-help@xml.apache.org; run by ezmlm
Precedence: bulk
Reply-To: cocoon-dev@xml.apache.org
Message-ID: <3DA6DE91.2090906@apache.org>
Date: Fri, 11 Oct 2002 16:22:09 +0200
From: Stefano Mazzocchi <stefano@apache.org>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
 rv:1.2a) Gecko/20020910
MIME-Version: 1.0
To: cocoon-dev@xml.apache.org
Subject: Re: [Announcement] sitemap variables
References: <CC92E7C2663B1D498A745BD686C45E1F3F2D18@exch2000.cti2.com>
 <1034215741.811.45.camel@defiant> <3DA4E4E8.9050301@gmx.de>
 <200210101056.46427.tcurdt@dff.st> <3DA568F7.3090709@gmx.de>
 <1034263269.3da59ae52bf97@imp.dff.local> <3DA5B214.1020705@anyware-tech.com>
 <3DA5C6DF.1070400@kaon.com> <3DA5FEF9.8010205@apache.org>
 <3DA691DA.50702@anyware-tech.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Sylvain Wallez wrote:

> This can often happen with matchers when e.g. an area of the application 
> is protected :
> 
> <map:match type="wildcard" pattern="admin/*">
>  <map:act type="auth">
>    <map:match type="wildcard" pattern="admin/adduser">
>      ...
>    </map:match>
>  </map:act>
> </map:match>

Ok, good point.

> So explicit naming as proposed by Ilya finally seems the best approach 
> to me, as it removes all possible ambiguities at the price of a little 
> added verbosity. Since this added verbosity only happens when this 
> feature is used, this actually does no harm.

Agreed.

> Let's see now how we can write that...

[skipped my tries]

> Puke, puke, puke ;-))

:-Prrr

> You're trying to find an alternative to the ':' which is used to 
> separate the prefix from the attribute name, because you make the 
> assumption that what's before the ':' *is* an inputmodule name.
> 
> But isn't it precisely this prefix (i.e. the namespace) that should be 
> used to distinguish between inputmodules and sitemap variables ? If yes 
> (and I think so), this means we have to find a naming convention for the 
> prefix, and not for the separator.
> 
> You may puke (if you still can ;-), but what about considering 
> sitemap-defined namespaces as internal to the sitemap, and thus prefix 
> them with a "#", that we all know as being used for internal links in 
> web pages ?
> 
> This would lead to (adapted from Ilya's proposal) :
> 
> <map:match pattern="*-*" name="mtch1">
>  <map:act type="my-action" name="act1">
>    <map:generate type="serverpages" src="{#mtch:1}.xsp">
>      <map:parameter name="display" value="{#mtch:2}"/>
>      <map:parameter name="param1" value="{#act1:1}"/>
>      <map:parameter name="param2" value="{#act1:2}"/>
>      <map:parameter name="param3" value="{request:1}"/>
>      <map:parameter name="param4" value="{1}"/>
>    </map:generate>
>  </map:act>
> </map:match>
> 
> Explanation :
> - {#mtch1:xxx} come from statement named "mtch1"
> - {#act1:xxx} come from statement named "act1"
> - {request:xxx} come from input-module "request"
> - {xxx} works as usual and gets its value from the enclosing statement.
> 
> I changed "param-prefix" proposed by Ilya to "name" since this is 
> actually what it is : a name given to a sitemap statement.
> 
> Thoughts ?

know what? I like it. It seems a pretty nice solution and the use of 
anchor-like syntax is simple and understandable.

So you get my +1 on this.

-- 
Stefano Mazzocchi                               <stefano@apache.org>
--------------------------------------------------------------------


---------------------------------------------------------------------
To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
For additional commands, email: cocoon-dev-help@xml.apache.org