geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From David Jencks <>
Subject Re: ClassFinder questions/problems -- annotation processing
Date Sat, 17 Feb 2007 23:50:47 GMT
I think the only annotations geronimo itself has to pay attention to  

javax.annotation.Resource(s)* (maybe)

there are some more I'm really not sure of but probably are taken  
care of by the jaxws impl:

These last might require the appropriate sub-builder (web service  
builder) to receive a ClassFinder from whatever's calling them.

I've mostly thought about how this interacts with NamingBuilders  
rather than other stuff like the web service builders.

I'm not sure I understand what you are proposing, but I thought about  
this and think we can do it fairly easily within the framework of  
pluggable naming builders while also allowing the default  
environments to be added to the environment correctly.

Some observations on who knows what and thus who should do what:

the module builder knows which parts of the ear are accessible to the  
module, so it should construct a ClassFinder reflecting this access.   
This can only be done after copying files out of the source into the  
car location.  If we can construct a classloader that reaches into an  
unpacked ear that would be ideal, we wouldn't have to copy the j2ee  
app at all then.

As part of this the module builder has to find all the component- 
declaring annotations and reflect that in the ClassFinder if  
appropriate (this is ejbs only, right? Or WebEndpoint as well?)

the naming builders each know what kind of (non-component-declaring)  
annotations are relevant so they should look for them in the  
ClassFinder provided by the module builder.

Only the module builder understands the schema for the spec dd, so it  
has to add the xml for the annotations into the dd to make it  
metadata-complete.  So the naming builders have to return the found  
annotations in some form the module builder can deal with.

A classloader that includes the contributions of the naming builders  
should only be needed by the naming builder itself when it's  
constructing the reference.  It might not be needed even then.


So here's how I see the interaction of the module builders and the  
naming builders:

ModuleBuilder.createModule doesn't call naming builders.

ModuleBuilder.installModule copies stuff into the target and  
constructs a ClassFinder, then calls NamingBuilder.buildEnvironment  
with the ClassFinder.  It stuffs whatever is returned back into the  
spec dd at the appropriate location.  The NamingBuilders add stuff to  
the environment, so after this we can construct a real classloader.

NamingBuilder.buidEnvironment looks at the xml and extracts whatever  
annotations are relevant for it and returns them: if also adds stuff  
to the environment if appropriate (typically if it found something in  
the xml or annotations)

ModuleBuilder.initContext calls NamingBuilder.initContext (although  
this seems a bit fishy to me)

ModuleBuilder.addGBeans calls NamingBuilder.buildNaming which  
actually constructs the references and stuffs them in a map at the  
appropriate location.

I think it might be appropriate for the naming builder to make a list  
of _all_ the stuff it will need to deal with including from the  
original dd and return that: we can pass that in to the initContext  
and buildNaming methods so it doesn't have to look at the xml again.

The really nasty part of the spec here appears to me to be the  
extreme overloading of the @Resource annotation.  I think they should  
have gone the rest of the way and said ejb and web service refs were  
also resources.  What I outlined above provides no good way to check  
that all the annotations are handled by some NamingBuilder.  This  
might be OK.  An alternative might be to have the ModuleBuilder  
extract all the possibly relevant annotations and have the  
NamingBuilders remove the ones they process.   If there's anything  
left at the end, something didn't get processed.  At the moment I'm  
not too worried by the possibility that something didn't get processed.


At least the AbstractWebModuleBuilder is going to need to look at the  
security annotations.  I think the same ClassFinder that the  
NamingBuilders use would work for this.  We probably want to stash it  
in the Module so it can be used whenever necessary.

I just started looking at this annotation stuff recently so it's  
entirely possible I missed the boat  completely.... don't be shy  
about letting me know :-)

david jencks

On Feb 16, 2007, at 8:59 PM, Tim McConnell wrote:

> Thanks for the info David. At times it's not obvious to me which  
> annotations have already been implemented as part of other projects  
> and which require Geronimo implementation changes. So I've been  
> keeping track of them on the wiki below. The ones with a red-X or  
> green-Checkmark in the far-right-hand column are those that I'm  
> assuming will require Geronimo changes. I'm working on the JSR-250  
> set now so if you if you have a chance to review and/or verify my  
> assumptions that would be helpful to me.
> +Annotations
> Thanks,
> Tim McConnell
> David Blevins wrote:
>> On Feb 15, 2007, at 9:25 PM, Tim McConnell wrote:
>>> Hi David/Dain, I finally see what's going on here now--and it  
>>> really makes a lot of sense. I'm not so sure it's a bug with the  
>>> classloading process so much as it's just the way the current  
>>> code functions. I can't easily show a sequence diagram here but  
>>> can briefly explain. It appears the the "EarContext" for a  
>>> deployed EAR file is aggregated across successive calls from  
>>> EARConfigBuilder.buildConfiguration() to the installModule()  
>>> method on first the WebModuleBuilder class, and then second on  
>>> the EjbModuleBuilder class. This explains why ClassFinder was  
>>> working correctly in EjbRefBuilder (i.e. the deployed module's  
>>> EarContext had been fully aggregated) but failed for me in  
>>> AbstractWebModuleBuilder (i.e., the deployed module's EarContext  
>>> had not yet been fully aggregated).
>> That would explain a lot!   Though, this does seem like an issue  
>> that should be fixed.  I know DJ isn't fond of some of the hacks  
>> we've had to add in the builder process.  Likely this may be the  
>> straw that broke the camels back.
>>> So, rather than fixing something that's not really broken in  
>>> AbstractWebModuleBuilder, the best solution in my view is to push  
>>> the Annotation processing out of the installModule processing of  
>>> the module builder(s) up into the configuration builder. This  
>>> would allow us to encapsulate the Annotation processing for  
>>> deployed EJB applications, Web applications, App Client  
>>> applications, and Connectors (not sure these would be annotated  
>>> though) into a single class: EARConfigBuilder. Additionally, it  
>>> would guarantee that we always have access to a fully aggregated  
>>> EarContext, and thus a fully populated classloader to pass to  
>>> ClassFinder. And finally, I think it would encapsulate most of  
>>> the Geronimo annotation processing except for Web Services. This  
>>> approach is somewhat in line with my original proposed plan for  
>>> Annotation Processing for Geronimo, it's just the conduit has  
>>> changed somewhat. Do either of you (or anyone else) have any  
>>> thoughts, comments and/or concerns ??
>> That'd be fine for Web modules and App Clients -- there are no  
>> Connector annotations and EJB annotations are taken care of by  
>> OpenEJB.  I know you keep wanting to do all the EJB-specific  
>> annotations, but there's no real reuse there.  Web modules and  
>> Connectors pretty much both have the same stuff, which is really  
>> only the five or so JSR-250 annotations plus these from javax.ejb:  
>> @EJB(s), @PersistenceUnit, and @PersistenceContext.
>> You can cross the rest off your list: i.e. javax.ejb @Remote,  
>> @RemoteHome, @Local, @LocalHome, @Stateless, @Stateful,  
>> @MessageDriven, @PrePassivate, @PreActivate, @Init, @Remove,  
>> @Timeout, etc.  These are EJB specific and already implemented for  
>> the most part.
>> -David
>>> Thanks,
>>> Tim McConnell
>>> Tim McConnell wrote:
>>>> Hi Dain, What you suggest does make a lot of sense. But for the  
>>>> stateless-calculator ear file (i.e., calculator-stateless-ear-2 
>>>> [1].0-M2.ear) I would then expect to find the calculator- 
>>>> stateless-ear-2[1].0-M2.jar file on one of the parent  
>>>> classloaders for the WAR classloader in  
>>>> AbstractWebModuleBuilder. It's not, so I suspect there is a bug  
>>>> somewhere as you suggest. I shall investigate further tomorrow.  
>>>> Thanks for the pointer....
>>>> Dain Sundstrom wrote:
>>>>> On Feb 6, 2007, at 12:49 PM, David Blevins wrote:
>>>>>> On Feb 4, 2007, at 7:19 PM, Tim McConnell wrote:
>>>>>>> Hi again Dave, after your recommendation below to do all the
>>>>>>> annotation discovery during the installModule phase of  
>>>>>>> deployment ClassFinder is working much better for me. I do  
>>>>>>> still have another scenario I'd appreciate some advice on. It
>>>>>>> seems that when an EJB EAR file (with embedded JAR and WAR  
>>>>>>> files) gets deployed in Geronimo, there are two builders  
>>>>>>> invoked: e.g., TomcatModuleBuilder/AbstractWebModuleBuilder 

>>>>>>> and EJBModuleBuilder such that the embedded JAR file is not 

>>>>>>> added to the ClassPath/ClassLoader when the WAR is deployed 

>>>>>>> (I assume this is the way it should work since I haven't  
>>>>>>> changed it--yet). So, if there are annotations in the WAR  
>>>>>>> class files pointing to classes in the JAR file, we'll still
>>>>>>> encounter NoClassDefException(s). I can just add the JAR  
>>>>>>> files in the EAR to the classpath of the WAR, which is what 

>>>>>>> I've done to get around the problem, but I'm not sure this is
>>>>>>> the best alternative. Do you have any thoughts?? Thanks much
>>>>>> Those should be added automatically via the deployment  
>>>>>> system.  Very puzzling.  Dain, did you see anything like this  
>>>>>> when you did that hack for @EJB annotation support in Servlets?
>>>>> Nope.  The WAR class loader is a child of the EAR class loader  
>>>>> so the WAR should "see" all of the jars in the ear.  If that is  
>>>>> not the case, then there is a bug.
>>>>> -dain

View raw message