[Nitro] Param stuff (yes, I'm sorry ^^;)

[George Moschovitis]

Ok, thanks for this patch, I will have a look ASAP.

-g.

On 10/28/06, Jonathan Buch <john at oxyliquit.de> wrote:
> Hi,
>
> >> I'm really sorry, but could you either revert back to the original partial
> >> broken param handling or test with my testcase properly without randomly
> >> commenting valid testcases (which are not touching the two issues I raised
> >> in my last post about params)? ^^;
> >> With `def action(par = nil)` broken we can't release on the 31st. ^^;
> >
> > I havent randomly remarked your tests. I just haven worked on this
> > issue yet. It is in my immediate todo list. Btw, if you can make a
> > patch of your implementation against the current repo, it would be
> > really helpful. Just this single patch, it is too difficult to
> > understand your changes in that biiiig bundle you have sent me some
> > time ago.
> >
> > thanks in advance and sorry for the inconvienience,
>
> yes, randomly was indeed the wrong word here.
>
> Attached are 2 patches which do what I was proposing (probably).
> This time it should even work with routes..
>
> Sat Oct 28 15:21:40 CEST 2006  Jonathan Buch <john at oxyliquit.de>
>    * Rework parameter handling
>    Quite big change on how parameters are handled.
>
>    * Use new context.action_params instead of .query_params to distinguish
>      between GET params and the ones from nice urls.
>    * When route is given merge the route parameters into the GET params,
>      add route parameter values into .action_params.
>    * Change Router#decode_route to use Dictionary instead of Hash (to
>      preserve order, important for action arguments).
>    * action raises `ActionError, "Wrong parameter count"` on wrong usage.
>    * Uncommented testcases in tc_controller_params again
>    * Added testcase for routes (TODO: need more of those)
>    * Single line fix for tc_dispatcher, it still used NoActionError.
>
> Sat Oct 28 15:48:33 CEST 2006  Jonathan Buch <john at oxyliquit.de>
>    * Bugfix for params and actions with arity 0
>    Since the same params are used for rendering included parts and the
>    /error page, actions with arity 0 have to be treated differently.
>
> The current implementation is probably not as fast as it could be, but
> we can optimize this later I guess.
> I have to admit that I hate digging into this Nitro code...  I feel much
> more at home with Og.  ^^;
>
> Jonathan
>
> --
> Feel the love
> http://pinkjuice.com/pics/ruby.png
>
> _______________________________________________
> Nitro-general mailing list
> Nitro-general at rubyforge.org
> http://rubyforge.org/mailman/listinfo/nitro-general
>
>
>


-- 
http://www.gmosx.com
http://www.nitroproject.org