Interesting question Jackson.
Our stuff here has all been standalone - so we've really only cared
about the servlet class/instance and the alias it was mounted on.
I'd guess there are some standard web container semantics for how
getServletName should derive it's name - don't recall any specifics in
the OSGi HttpService spec, otherwise I think we'd have implemented them
already. Shouldn't be hard to do though if it is based on a well known
Dictionary property mapping. The only thing would be how to keep this
approach "standard" for other OSGi implementations of the service. I'd
guess that would need amendment to the standard so everyone followed the
same model
- Rob
On 21/04/2010 4:10 PM, Jackson, Bruce wrote:
There are some third party frameworks (in my case, I'm using Echo 3) that use
this value for tracking of application instances. If it can't be set, all
Echo-based apps will interfere with one another.
From the spec, it looks like the Dictionary passed into registerServlet()
should be used to create the ServletContext object, which in turn has a
getServletName method. Thus, I assumed there was some item in the Dictionary
that was used for this purpose.
----- Reply message -----
From: "Sahoo"<sa...@sun.com>
Date: Tue, Apr 20, 2010 23:31
Subject: Http service and servlet name
To: "dev@felix.apache.org"<dev@felix.apache.org>
I don't think there is any such mapping defined. It's an implementation
detail. Where do you see the need for this mapping from a user's point
of view?
Thanks,
Sahoo
Jackson, Bruce wrote:
Does anyone happen to know: what is the equivalent to the<servlet-name>
tags in the web.xml for programmatically using the OSGi http service?
Do I put these into the Dictionary object passed into the registerServlet()
method, and if so, what is the key entry for the Dictionary?
Thanks
Bruce
--
Ascert - Taking systems to the Edge
r...@ascert.com
+44 (0)20 7488 3470
www.ascert.com
