Mailing-List: contact dev-help@forrest.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@forrest.apache.org
Received-SPF: neutral (hermes.apache.org: local policy)
Message-ID: <42A1B905.9060308@apache.org>
Date: Sat, 04 Jun 2005 15:21:57 +0100
From: Ross Gardler <rgardler@apache.org>
User-Agent: Mozilla Thunderbird 1.0 (Windows/20041206)
MIME-Version: 1.0
To: dev@forrest.apache.org
Subject: Re: locationmap - not finishing the pipeline
References: <4998884405060406551c7ab2ee@mail.gmail.com>
In-Reply-To: <4998884405060406551c7ab2ee@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Tim Williams wrote:
> The second link in the locationmap wouldn't work because for some
> reason it's trying to use the "file src" defined in forrest-core.xconf
> and append it to the project-content-dir.  This is strange because it
> apparently finds locationmap.xml with no problem for simple link
> re-writing (the first sample).  I don't really know where to start but
> for now I made the kludge even worse by hardcoding the absolute path
> to locationmap.xml.  This gives me something to play with for now
> anyway.

Yes, I noticed that problem (which is why there is a warning on that 
section of the page saying it doesn't work yet). I'm glad you found a 
temporary solution. The longer term solution (I think) is to pass the 
path to the locationmap to the input-module from the sitemap rather than 
within xconf. I'll look into this next week.

> Anyway, so I can successfully take the /locationmap/demo.xml and
> return a real demo.xml from 127.0.0.1:8080/slide/files/demo.xml (my
> slide repo).  Everythings pretty much hard coded for now mind you. 

Fantastic. That was to be my next test (with Daisy rather than Slide, 
but same principle), glad to hear it works, at least to an extent

> Now the problems:  1) it strips the contents from demo.xml and is
> essentially just a shell xml file.  

Well that isn't very helpful, I suppose we better fix that ;-)

I have no idea how this can be happening. Are you able to confirm that 
the file is being read from slide correctly?

> 2) it doesn't apply any skinning
> to it -- it just returns the xml.  

Is your request to Forrest for demo.html or demo.xml?

>any pointers on where to start with these?

Lets start with a simple case. Lets ignore slide for now and retrieve 
something via http. A good candidate would be something from our own 
SVN, for example

http://svn.apache.org/viewcvs.cgi/*checkout*/forrest/trunk/main/fresh-site/src/documentation/content/xdocs/index.xml

(the home page of fresh-site)

Could you modify your code to retrieve that page rather than a slide one 
and put a patch against http://issues.apache.org/jira/browse/FOR-200

We can then work against the same code that way. Don't worry that we are 
using hacks like hard-coded paths. As long as you put a nice clear FIXME 
note against the changes we can tidy up later. The branch we are working 
in will not be merged with trunk until we iron all these things out. For 
now we should free to experiment.

Ross