RE: [picocontainer-dev] Trying an experiment to kill off ContainerRecorder (Nano)

--- Michael Rimov <[email protected]> wrote:

> > Instance components for session level is not
> something I thought of
> > with the ThreadLocal/Storing thing.
> > 
> > If application level, then regular caching is the
> strategy
> > 
> > Request level, in-theory, a no-caching and
> no-storing CA would still
> > cache instance registrations.
> 
> Right, and it was forcing me to completely rebuild
> the container with each
> servlet request.  (I don't use Session level
> Container)
> 
> > 
> > Where do you use instance registrations dude ?
> 
> //Example...
> public class Action {
> 
> 	public Action(HttpServletRequest request,
> HttpServletResponse response)
> {
> 		//.....
> 	}
> }
> 
> pico.addComponent(Action.class);
> pico.addComponent(HttpServletRequest.class,
> request);
> pico.addComponent(HttpServletResponse.class,
> response);
> pico.addComponent(HttpSession.class,
> request.getSession(true));
> Action action = pico.getComponent(Action.class);

Action depending on servlet stuff is pretty uncool by
itself.  What framework are you using? 

I know that WW2/XW2 could  be slight overkill,  but 
at least you will get request parsing and parameter
passing for free. 

regards,

----[ Konstantin Pribluda http://www.pribluda.de ]----------------
JTec quality components: http://www.pribluda.de/projects/


      

---------------------------------------------------------------------
To unsubscribe from this list, please visit:

    http://xircles.codehaus.org/manage_email