beehive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Andrew McCulloch" <amccu...@gmail.com>
Subject Re: Proposal for class-level interceptor support (BEEHIVE-1076)
Date Tue, 27 Mar 2007 02:28:40 GMT
How does this ordering compare to section 2.1 of JSR 250.  I am not
suggesting any changes, I was just curious if that specification is
similiar.

On 3/26/07, Ken Tam <kentaminator@gmail.com> wrote:
>
> On 3/8/07, Chad Schoettger <chad.schoettger@gmail.com> wrote:
> > I would like to propose a solution for BEEHIVE-1076:
> >
> >  "interceptors for annotations don't get wired into generated control
> > beans if the annotation is at the class level"
> >
> >
> > Just to be clear this is not a bug, it is an enhancement to existing
> > functionality.  Since this work introduces new functionality it
> > becomes necessary to clearly define the behavior of class-level
> > interceptors in Beehive.
> >
> > I prefer a simple design which is intuitive for Beehive developers, so
> > here's what I am proposing:
> >
> > 1) Use the existing Beehive annotation @InterceptorAnnotation to
> meta-annotate
> >    class-level interceptor annotations.  The only difference between a
> >    class-level and method-level interceptor annotation would be the type
> it
> >    annotates.  This would allow a developer to create a single
> interceptor
> >    annotation which could be used at the class or method level.
> >
> > 2) A class-level interceptor gets applied to each method in the class
> >    (including inherited methods).
> >
> > 3) If a method within an class is annotated with the same interceptor
> >    annotation as the class only the method's interceptor is fired.
> >
> > 4) Interceptor prioritization does not change, method-level and
> class-level
> >    interceptor priorities could still be defined in the same manner by
> >    controls-interceptors.config file.
> >
> >
> > Questions/Comments?  Any input is greatly appreciated.
>
> +1, seems straightforward.  Just to clarify #3 -- I'm reading your
> proposal as "if a given method is somehow parameterized by the same
> interceptor annotation multiple times (via some combination of
> annotated inheritance and/or annotation at both type and method
> level), then the most specific (most inherited, method level) instance
> of the annotation wins".  Sound right?
>
> >
> >  - Thanks, Chad
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message