To reply, what we already discussed on irc.
This will do it:
Class<RequestHandler> clazz = (Class<RequestHandler>) tmp;
InjectionTarget<RequestHandler> it =
manager.createInjectionTarget(manager.createAnnotatedType(clazz));
CreationalContext<RequestHandler> cc =
manager.createCreationalContext(null);
// handler = it.produce(cc); // no need for this, if you already have
an instance
it.inject(handler, cc);
https://github.com/capedwarf/capedwarf-green/blob/master/server-api/src/main/java/org/jboss/capedwarf/server/api/servlet/WeldServlet.java
On Dec 7, 2011, at 1:21 PM, José Rodolfo Freitas wrote:
> Ales and All,
>
> when
>
> @Inject
> GenericQuery<String> gqs;
>
> my producer do:
>
> new GenericQuery(clazz)
>
> And then @Inject inside GenericQuery does not work (since I'm using new).
>
> What's the best approach in this situation?
>
> Using BeanManagerProvider.getInstance() inside GenericQuery constructor to
> provide instances for the objects I want to inject?
>
>
> On Tue, Dec 6, 2011 at 9:29 PM, José Rodolfo Freitas
> <[email protected]> wrote:
> that´s an awesome solution Ales!
> I didn´t thought that I could obtain the parametizedType that way!
> Testing this will be my first task tomorrow morning.
>
> thank you all for your ideas.
>
>
> On Tue, Dec 6, 2011 at 8:55 PM, Jason Porter <[email protected]> wrote:
> Nice solution Ales, wish I'd thought of it :)
>
>
> On Tue, Dec 6, 2011 at 15:45, Ales Justin <[email protected]> wrote:
> You could do something like this:
>
> @SuppressWarnings({"unchecked"})
> @Produces
> public GenericQuery injectClass(InjectionPoint ip) {
> Annotated annotated = ip.getAnnotated();
> Class clazz = Object.class;
> Type type = annotated.getBaseType();
> if (type instanceof ParameterizedType) {
> ParameterizedType pt = (ParameterizedType) type;
> clazz = (Class) pt.getActualTypeArguments()[0];
> }
> return new GenericQuery(clazz);
> }
>
> --
>
> @Inject
> GenericQuery<String> gqs;
> @Inject
> GenericQuery<Number> gqn;
>
> --
>
> System.out.println("GQS: " + gqs.getClazz());
> System.out.println("GQN: " + gqn.getClazz());
>
> --
>
> GQS: class java.lang.String
> GQN: class java.lang.Number
>
> --
>
> https://github.com/alesj/cdi-arq-workshop/commit/64860f15197be40a7714dfd8a7da931c1db11411
>
> HTH
>
> -Ales
>
>> Hey guys, how you doing?
>>
>> I'm trying to achieve something that might be impossible, but before
>> concluding that, I'd like to ask you, CDI gurus!
>>
>> I have the following class:
>>
>> public class Foo<T> {
>>
>> public TypedQuery<T> getQuery(){
>>
>> }
>>
>> }
>>
>> As you can Imagine, inside my getQuery method, I'd have to use "T.class" to
>> make it TypedQuery. which is impossible due java generics type erasure.
>>
>> so I'd have to build a private field to hold the t.class for me.
>>
>> public class Foo<T> {
>>
>> private Class<T> klass;
>>
>> public TypedQuery<T> getQuery(){
>>
>> }
>>
>> public void setKlass(Class<T> klass){
>> this.klass = klass;
>> }
>> }
>>
>> The problem is that forcing this 'setKlass' feels very ugly to the api, and
>> it's not very error prone, since one could easily forget to set this
>> configuration.
>>
>> So I had an Idea: force the setKlass inside the constructor:
>>
>> public class Foo<T> {
>>
>> private Class<T> klass;
>>
>> public Foo(Class<T> klass){
>> this.klass = klass;
>> }
>>
>> public TypedQuery<T> getQuery(){
>>
>> }
>>
>> }
>>
>> Unfortunatelly, this breaks cdi, since it cannot inject it anymore. At least
>> AFAIK.
>>
>> So, is there a way out of this? maybe using a secret solder feature?
>> _______________________________________________
>> seam-dev mailing list
>> [email protected]
>> https://lists.jboss.org/mailman/listinfo/seam-dev
>
>
> _______________________________________________
> seam-dev mailing list
> [email protected]
> https://lists.jboss.org/mailman/listinfo/seam-dev
>
>
>
>
> --
> Jason Porter
> http://lightguard-jp.blogspot.com
> http://twitter.com/lightguardjp
>
> Software Engineer
> Open Source Advocate
> Author of Seam Catch - Next Generation Java Exception Handling
>
> PGP key id: 926CCFF5
> PGP key available at: keyserver.net, pgp.mit.edu
>
>
_______________________________________________
seam-dev mailing list
[email protected]
https://lists.jboss.org/mailman/listinfo/seam-dev
