Question

I have just started reading “Spring In Action – Third edition” and am stuck up while experimenting with wiring concepts. I am unable to understand the lifecycle of a bean after writing this code :

    public class TestCase {
    public static void main(String[] args) {

        ApplicationContext context = new ClassPathXmlApplicationContext("test.xml");
        Test1 t1 = (Test1)context.getBean("test1");
        t1.setName1("Win");
        Test1 t2 = (Test1)context.getBean("test1");
        t2.setName1("Lin");
        Test2 t3 = (Test2)context.getBean("test2");
        Test1 t4 = t3.getName();
        System.out.println("End" +t4.getName1());

    }

}

public class Test1 {

    private String name1;

    public String getName1() {
        System.out.println("test1 - getter");
        return name1;
    }

    public void setName1(String name1) {
        System.out.println("test1 - setter");
        this.name1 = name1;
    }

    public void onStart()
    {
        System.out.println("start1");
    }

    public void onStop()
    {
        System.out.println("stop1");
    }
}


public class Test2 {

    private int age;
    private Test1 name;
    public int getAge() {
        return age;
    }
    public void setAge(int age) {
        this.age = age;
    }
    public Test1 getName() {
        System.out.println("test2 - getter");
        return name;
    }
    public void setName(Test1 name) {
        System.out.println("test2 - setter");
        this.name = name;
    }

    public void onStart()
    {
        System.out.println("start2");
    }

    public void onStop()
    {
        System.out.println("stop2");
    }
}

Here’s my test.xml :

<bean id="test1" class="springidol.Test1" init-method="onStart" destroy-method="onStop" >
</bean>

<bean id="test2" class="springidol.Test2" init-method="onStart" destroy-method="onStop">
<property name="name" ref="test1"></property>
</bean>

The output is :

   start1
test2 - setter
start2
test1 - setter
test1 - setter
test2 - getter
test1 - getter
End - Lin

If I change the Test1 scope to prototype I get :

start1
test2 - setter
start2
start1
test1 - setter
start1
test1 - setter
test2 - getter
test1 - getter
End - null

I know I am asking for too much, but can someone get me the steps involved here – I cant understand why test2 setter is getting called after loading of tst1 is done (and even before test2 is loaded) !

Secondly, why are the “End” outputs for prototype and default scopes different ?
Thanks.

Answer 1

Because Spring does set the references to other beans first. (Wiring of the beans)
After that the init methods will be invoked.

Since your test2 setter points to a reference of test1 within your application context configuration it will be invoked first.

———-edit————

The prototype scope is a little bit tricky, since it behaves different depending if your prototype scoped bean is a proxy or not.

If you get the prototyped bean from the application context directly, like you did, you will get a new instance each time. Setting a name to the first instance will not affect the name of the second instance you write out at the end.

But if a prototype scoped bean is referenced within the application context from another singleton bean, a proxy is injected. This proxy will even switch the actually invoked instance for every method call on it. This will be more irritating since:

Test1 t4 = t3.getName(); 
t4.setName1("lala");
assertEquals(null,t4.getName1()); //will be true

But please verify that, since thats what I suppose it is doing. I did not verified it by code.