cxf-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Tully, Gary" <Gary.Tu...@iona.com>
Subject RE: http-(conduit|destination) cfg
Date Thu, 03 May 2007 15:32:50 GMT
Hi Polar,

Good point, you don't! The rule is that there should only be one
wildcard per class, so having more than one wildcard per class is an
error. Possibly we should take a fail fast approach to this sort of
mis-configuration. My initial thinking was that this is not fatal, just
log a warning, but depending on the nature of the config, it could
indeed be fatal.

A fail fast approach will necessitate aborting bus initialisation,
spring will do that if there are duplicate bean ids so I guess it is the
correct approach.  What do you think?

I was considering allowing multiple wildcards per class, it would have
it's uses but it adds unnecessary complexity at the moment and would
have an a priori ordering issue when multiple config sources are
involved. Complexity is best added as needed imo.

Thanks,
Gary.

> -----Original Message-----
> From: Polar Humenn [mailto:phumenn@iona.com] 
> Sent: 03 May 2007 14:31
> To: cxf-dev@incubator.apache.org
> Subject: Re: http-(conduit|destination) cfg
> 
> My big question is on the issue that you say "first [match] wins"; So,
> 
> How do you *know* definitively and a priori, which is going 
> to be first?
> 
> Cheers,
> -Polar
> 
> Tully, Gary wrote:
> > Hi,
> > Resurrecting an old thread. I have submitted a patch[1] that 
> > implements simple wildcarding to bean ids used in 
> configuration. The 
> > patch can provide the basis for a solution to the original problem 
> > described in this thread[2], namely portName clashes vs 
> long bean ids.
> > As it stands it provides a simple way to apply default config 
> > behaviour across a range of configured entities which solves my 
> > problem of easily applying a property to all http 
> destinations in a server.
> >
> > Would a committer have a moment to check it out?
> >
> > The JIRA[3] contains complete details and some requests for 
> comment. 
> >
> > In short, configuration can change from:
> >
> >     <bean
> > 
> name="{http://cxf.apache.org/factory_pattern}NumberPortOne.http-destin
> > at
> > ion" abstract="true"> 
> >       <property name="multiplexWithAddress" value="true"/> 
> >     </bean> 
> >     <bean
> > 
> name="{http://cxf.apache.org/factory_pattern}NumberPortTwo.http-destin
> > at
> > ion" abstract="true"> 
> >       <property name="multiplexWithAddress" value="true"/> 
> >     </bean>  
> >     ...
> >
> > to: 
> >     <bean name="*.http-destination" abstract="true"
> > class="JettyHTTPDestination"> 
> >       <property name="multiplexWithAddress" value="true"/> 
> >     </bean>
> >
> > or:
> >     <bean name="*" abstract="true" class="JettyHTTPDestination"> 
> >       <property name="multiplexWithAddress" value="true"/> 
> >     </bean>
> >
> > Gary.
> >
> > [1]
> > 
> https://issues.apache.org/jira/secure/attachment/12356655/patc
> h-cxf-623.
> > r534554.patch
> > [2]
> > 
> http://mail-archives.apache.org/mod_mbox/incubator-cxf-user/200704.mbo
> > x/
> > %3c46149E97.8080500@iona.com%3e
> > [3] https://issues.apache.org/jira/browse/CXF-623
> >   
> 
> 

Mime
View raw message