Iterator Pattern

 
Iterator Pattern Defined:
provides a way to access the elements of an aggregate object sequentially without exposing its underlying represation;
个人理解：
对于在类中存在一些集合类（Vector,Array,LinkedList,HashTabled）等，当需要对集合类进行遍历等操作时，底层的寻址操作如（get（i），array[i])等，可以利用Iterator Pattern来隐藏集合类的遍历方法，提供对客户的一致接口；减少客户端的改动可能，客户端只需要调用一致的接口createIterator（）来获取一个Iterator接口对象，利用该接口的子类来实现具体的遍历操作方法，客户端执行一致的next（）等操作；另外，通过Iterator对象，很容易根据需要修改遍历行为的具体方法（可能不单单是顺序遍历，可能倒序等），这时不需要修改Aggregate类的实现，简化了Aggregate的接口，而Iterator与Aggregator也各自对自己负责（place the responsibility where it should be)
 
Single Responsibility Principle(SRP原则）
A class should have only one reason to change!!
"responsibility  means "a reason to change"";
SRP原则指出了一个类的设计原则是类应该附有单一的责任，而不应该持有多个毫不相干的责任，如果持有多种不相干的行为，则应该对该类进行分解，使不相干的方法行为属于不同的类或接口。原因是一个类如果持有过多的不相干的行为，当产生变化时，会导致类的大部分接口出现修改问题。在下面的例子中，Aggregate是集合类，只负责添加保存对象，不对遍历操作负责；而在Iterator中，只对集合类遍历操作负责，不关心具体的添加保存对象等行为；

 

import java.io.*; import java.util.*; class PersonInfo{ private String name; private String sex; private int age; public PersonInfo(String n,String s,int a){ name = n; sex =s; age = a; } public void print(){ System.out.println("name : " + name + ";sex : " + sex + ";age : " +age ); } } abstract class PersonAggregate{ abstract public zbhIterator createIterator(); } class ManagerAggregate extends PersonAggregate{ private PersonInfo [] personInfo; //Array[] private static int MAX_NUM = 10; private int num; public ManagerAggregate(){ personInfo = new PersonInfo[MAX_NUM]; num = 0; } public void addInfo(String name,String sex,int age){ if(num == MAX_NUM){ System.out.println("Fatal error :overflow"); }else{ PersonInfo tmp = new PersonInfo(name,sex,age); personInfo[num] = tmp; num++; } } public zbhIterator createIterator(){ return new ManagerIterator(personInfo); } } class StaffAggregate extends PersonAggregate{ private LinkedList personInfoList; //LinkedList public StaffAggregate(){ personInfoList = new LinkedList(); } public void addInfo(String name,String sex,int age){ PersonInfo ptmp = new PersonInfo(name,sex,age); personInfoList.add(ptmp); } public zbhIterator createIterator(){ return new StaffIterator(personInfoList); //return personInfoList.iterator(); //如果返回的是java里面实现的Iterator接口，可以更简单的返回，不用自己实现 } } interface zbhIterator{ public Object next(); public boolean hasNext(); public void remove(); } class StaffIterator implements zbhIterator{ private LinkedList lpersonInfo; private int pos; public StaffIterator(LinkedList ll){ lpersonInfo = ll; pos = 0; } public Object next(){ return lpersonInfo.get(pos++); } public boolean hasNext(){ if(pos >= lpersonInfo.size()) { return false; } return true; } public void remove(){ lpersonInfo.removeLast(); } } class ManagerIterator implements zbhIterator{ private PersonInfo [] pInfo; private int pos; public ManagerIterator(PersonInfo[] p){ pInfo = p; pos = 0; } public PersonInfo next(){ return (PersonInfo)pInfo[pos++]; } public boolean hasNext(){ if(pos >= pInfo.length || pInfo[pos] == null){ return false; } else return true; } public void remove(){ int i; for(i = 0;i<pInfo.length;i++){ if(pInfo[i] == null){ break; } } if(i>0) pInfo[i-1] = null; } } public class Client { /** * @param args */ private static void printAll(PersonAggregate pa){ //当增加新的Aggregate时，不需要修改客户代码 zbhIterator iterator = pa.createIterator(); //只需继承相关接口 while(iterator.hasNext()){ ((PersonInfo)iterator.next()).print(); } } public static void main(String[] args) { // TODO 自动生成方法存根 ManagerAggregate ma = new ManagerAggregate(); ma.addInfo("zbh","male",23); ma.addInfo("zbc","female",22); ma.addInfo("jjd","male",30); printAll(ma); //一致的接口 System.out.println("-----------------------------------"); StaffAggregate sa = new StaffAggregate(); sa.addInfo("mmm","female",10); sa.addInfo("dd","male",20); sa.addInfo("jjd","male",30); sa.addInfo("jk","female",40); printAll(sa); //一致接口 } }

 

CSDN上传图片好像出了问题，原来打算把UML传上来一起发，发不了，没办法！！