commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Howard M. Lewis Ship" <hlshipli...@comcast.net>
Subject RE: [HiveMind] Interceptors - CGLIB / Javassist comparison
Date Fri, 05 Mar 2004 21:29:41 GMT
> Howard - if I remember right - last time on interceptor 
> ordering you said 
> something about bundles of interceptors. It did not realy 
> understand that.

The issue is that if you contribute multiple interceptors to a service extension point, you
have to
be concerned with order. This is an offshoot of the fact that any module may contribute interceptors
to any service extension point.  This extreme case complicates things for the simple case.

But what if you contribute a group of interceptors as a single contribution? i.e.

<implementation service-id="...">
  <interceptor-group>
    <interceptor service-id="A"/>
    <interceptor service-id="B"/>
  </>
</>

This clearly orders A before B and there's no guesswork, since the group is contributed as
a whole.

Another option to explore is defining "sets" of interceptors to be contributed (possibly as
a group
with known ordering).  So:

<implementation service-id="...">
   <interceptor-set set-id="Foo"/>
</>

<interceptor-set id="Foo">
  <interceptor service-id="A"/>
  <interceptor service-id="B"/>
</>

Perhaps the entire idea of contributing just an individual interceptor is flawed; perhaps
they
should always be applied as a set or group, with the restriction that only a single set/group
may be
applied.  Then we are no longer concerned with order.

--
Howard M. Lewis Ship
Independent J2EE / Open-Source Java Consultant
Creator, Tapestry: Java Web Components 
http://howardlewisship.com

    


---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org


Mime
View raw message