> What about Dagger2? Dagger 2 does not modify existing classes, it will only generate code.
When you @Inject a member or a constructor in Dagger2, you're instructing it to generate the vanilla code you would have written to do manual dependency injection yourself. Dagger2 is much less magical than Google Guice in this regard. I think it will become clear if you take a look how the dependency injection bootstrapping occurs in Dagger2. >From the Dagger2 website - https://google.github.io/dagger/: > @Component(modules = DripCoffeeModule.class) > interface CoffeeShop { > CoffeeMaker maker(); > } > > // ... > > CoffeeShop coffeeShop = DaggerCoffeeShop.builder() > .dripCoffeeModule(new DripCoffeeModule()) > .build(); > CoffeeMaker = coffeeShop.maker(); In this example Dagger2 is generating a Component that matches the interface provided to expose the CoffeeMaker class. This component is basically used to bootstrap the object graph that Dagger2 has built up into your application. In the Dagger generated code, it knows about all the injectable types that can go into the constructor or members of the CoffeeMaker class, because you annotated those types with @Inject or created a provider for it DripCoffeeModule, and it will provide them, when newing it. Actual modification of the CoffeeMaker class by the annotation processor is not required to make this work. -- View this message in context: http://groovy.329449.n5.nabble.com/java-parser-usage-in-Groovy-tp5725492p5725534.html Sent from the Groovy Dev mailing list archive at Nabble.com.
