该如何实践State模式【转自www.bitsCN.com】

State模式将对象行为的变化封装成具有统一接口的状态，它与Strategy模式的区别在于任何状态的改变都是定义好的，即“改变行为”的动作由自己来做。

         
下面是自动门的状态图，它描述了“开”与“关”两个状态之间的变化条件。
         feedom.net
        
抽象状态DoorState提供open()和close()两个方法，
package com.zj.doorstate;
 
public interface DoorState {
    void open();
 
    void close();
}

Context为AutoDoor，它关联一个DoorState state。
package com.zj.doorstate;
 
public class AutoDoor {
    private DoorState state;
    private DoorState open = new Open(this);
    private DoorState close = new Close(this);
 
    public AutoDoor() {
       this.state = open;
    }
 
    public void setSate(DoorState state) {
       this.state = state;
    }
 
    public DoorState getOpenState() {
       return open;
    }
 
    public DoorState getCloseState() {
       return close;
    }
 
    public void open() {
       state.open();
    }
 
    public void close() {
       state.close();
    }
}

ConcreteState有两个。
     状态Open
package com.zj.doorstate;
 
public class Open implements DoorState {
    private AutoDoor door;
 
    public Open(AutoDoor door) {
       this.door = door;
    }
 
    public void close() {
       System.out.println("O.K. Door closed.");
       door.setSate(door.getCloseState());
    }
 
    public void open() {
       System.out.println("Door already opened.");
    }
}

状态Close
package com.zj.doorstate;
 
public class Close implements DoorState {
    private AutoDoor door;
 
    public Close(AutoDoor door) {
       this.door = door;
    }
 
    public void close() {
       System.out.println("Door already closed.");
    }
 
    public void open() {
       System.out.println("O.K. Door opened.");
       door.setSate(door.getOpenState());
    }
}

由两个具体状态类可知，具体状态之间的转换由它们自身定义。
【转自】