To assemble a bean, one simply has to use the constructs provided by Java:

@Bean(scope = DefaultScopes.SINGLETON)
public Person rod() {
  return new Person("Rod Johnson");
}

@Bean(scope = DefaultScopes.PROTOTYPE)
public Book book() {
  Book book = new Book("Expert One-on-One J2EE Design and Development");
  book.setAuthor(rod());  // rod() method is actually a bean reference !
  return book;
}

In the example above, the book author is using the return value of rod method. However, since both book and rod methods are marked with @Bean, the resulting beans, managed by Spring, will respect the container semantics: rod bean will be a singleton while book bean a prototype. When creating the configuration, Spring is aware of the annotation context and will replaces the rod() method invocation with a reference to the bean named 'rod'.

The container will return a new Book instance (prototype) each time book bean is request but will return the same instance (a singleton) for rod bean.

The code above is equivalent to:

<bean id="rod" class="Person" scope="singleton">
   <constructor-arg>Rod Johnson</constructor-arg>
</bean>

<bean id="book" class="Book" scope="prototype">
   <constructor-arg>Expert One-on-One J2EE Design and Development</constructor-arg>
   <property name="author" ref="rod"/>
</bean>

Note that while the examples above used two common scopes types, any type of scoping can be specified:

@Bean (scope = "customer")
public Bag shopingBag() {
  return new Basket();
}

@Bean (scope = "shift")
public Manager shopManager() {
  ...
}