Mailing-List: contact users-help@jackrabbit.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@jackrabbit.apache.org
Received-SPF: pass (herse.apache.org: local policy)
Message-ID: <456C7216.3040507@decoursey.net>
Date: Tue, 28 Nov 2006 11:29:58 -0600
From: Paul J DeCoursey <paul@decoursey.net>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: users@jackrabbit.apache.org
Subject: Re: Storing multi-part documents in repository
References: <5b3747160611280919y27f79del30da708fe87d7d4d@mail.gmail.com>
In-Reply-To: <5b3747160611280919y27f79del30da708fe87d7d4d@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

David Moss wrote:
> Hi,
>
> This isn't strictly a Jackrabbit issue, but is related to the way I 
> use it
> and I hope will be familiar to anyone who's used jackrabbit for a CMS or
> similar.
>
> I'm looking to store both single, and multi-part documents (e.g. html 
> page,
> with referenced images) within the repository and then serve these out 
> from
> the repository as part of a web application.  My first thoughts are to 
> store
> the document dependencies as child nodes of the main document node.
> However, I don't think storing the data is a problem.  The difficulty is
> with how best to retrieve it.
>
> If, for example, I simply pull an HTML document from the repository and
> stream it to a user's browser in response to a click, the links within 
> that
> document to its dependent images etc are invalid.  How can I retrieve 
> these
> as well?
>
> Does anyone have any thoughts on the best approach to this problem?
>
> I reckon I could either retrieve the files from the repository into a
> temporary directory, and serve them back to the client from there,
> or write a filter to attempt to retrieve any unrecognised url / url that
> matches a mask from the repository returning the document if found, or 
> 404
> if not.
> Neither of these seems like a neat solution.
>
> If it's useful, I'm using JSF for the user interface etc.
>
> Thanks
>
> Dave.
>
I think the solution needs to be rewriting the HTML on storage.  I am 
assuming that at some point you parse the html to get the list of linked 
images, at that point you will want to rewrite the references in the html.

Paul