介绍
概念:运用共享技术有效地支持大量细粒度的对象。
应用场景:项目中某个类需要产生大量的对象,其中对象中的属性创建占用了大量的内存,如果这些属性对象可以被公用,那么可以利用享元模式把他们指向同一个地址.
这让我想起了字符串常量池的做法.
那么我们就用例子模仿一个字符串存储的逻辑.
代码:
//自创字符串类
package note.com.flyweight; /** * 自定义String类 * @author lxz * */ public class MyString { private String name = null; public MyString(String name){ this.name = name; } public String getName() { return name; } }
//自创字符串常量池
package note.com.flyweight; import java.util.HashMap; import java.util.Map; /** * MyString存放池,类似字符串常量池 * @author lxz * */ public class MyStringPool { private static Map<String,MyString> map = new HashMap<String,MyString>(); /* * 获取MyString */ public static synchronized MyString getMyString(String key){ MyString myString = map.get(key); if(myString==null){ myString = new MyString(key); map.put(key, myString); } return myString; } }
//人实体,其中名字属性用到我们自创的字符串来表示
package note.com.flyweight; public class Person { private MyString name = null; private Integer age = null; public Person(MyString name,Integer age){ this.name = name; this.age = age; } public MyString getName() { return name; } public void setName(MyString name) { this.name = name; } public Integer getAge() { return age; } public void setAge(Integer age) { this.age = age; } }
//测试类
package note.com.flyweight; public class FlyWeightTest { public static void main(String[] args) { Person one = new Person(MyStringPool.getMyString("小明"),5); Person two = new Person(MyStringPool.getMyString("小红"),6); Person three = new Person(MyStringPool.getMyString("小明"),7); Person four = new Person(MyStringPool.getMyString("小明"),8); System.out.println(one.getName()); System.out.println(two.getName()); System.out.println(three.getName()); System.out.println(four.getName()); } }
//结果:
note.com.flyweight.MyString@659e0bfd
note.com.flyweight.MyString@2a139a55
note.com.flyweight.MyString@659e0bfd
note.com.flyweight.MyString@659e0bfd
结果可见,三个同名的Person实体,他们的name属性指向了同一个地址,这说明我们省了两个MyString的对象空间.假如有一万个同明的人,我们就省了九千九百九十九个对象的空间,情景尽可能极端,这样做的效果越好.
革命尚未成功,同志们仍需努力.