2019独角兽企业重金招聘Python工程师标准>>>
编程人员经常误用各个集合类提供的拷贝构造函数作为克隆List
,Set
,ArrayList
,HashSet
或者其他集合实现的方法。需要记住的是,Java集合的拷贝构造函数只提供浅拷贝而不是深拷贝,这意味着存储在原始List和克隆List中的对象是相同的,指向Java堆内存中相同的位置。增加了这个误解的原因之一是对于不可变对象集合的浅克隆。由于不可变性,即使两个集合指向相同的对象是可以的。字符串池包含的字符串就是这种情况,更改一个不会影响到另一个。使用ArrayList
的拷贝构造函数创建雇员List的拷贝时就会出现问题,Employee
类不是不可变的。在这种情况下,如果原始集合修改了雇员信息,这个变化也将反映到克隆集合。同样如果克隆集合雇员信息发生变化,原始集合也会被更改。绝大多数情况下,这种变化不是我们所希望的,克隆对象应该与原始对象独立。解决这个问题的方法是深克隆集合,深克隆将递归克隆对象直到基本数据类型或者不可变类。本文将了解一下深拷贝ArrayList
或者HashSet
等集合类的一种方法。如果你了解深拷贝与浅拷贝之间的区别,那么理解集合深克隆的方法就会很简单。
Java集合的深克隆
下面例子有一个Employee
集合,Employee是可变对象,成员变量name
和designation
。它们存储在HashSet
中。使用java.util.Collection
接口的addAll()
方法创建集合拷贝。然后修改存储在原始集合每个Employee
对象的designation
值。理想情况下这个改变不会影响克隆集合,因为克隆集合和原始集合应该相互独立,但是克隆集合也被改变了。修正这个问题的方法是对存储在Collection
类中的元素深克隆。
package javaBasic;import java.util.Collection;
import java.util.HashSet;
import java.util.Iterator;/*** Java program to demonstrate copy constructor of Collection provides shallow* copy and techniques to deep clone Collection by iterating over them.* * @author http://javarevisited.blogspot.com*/
public class CollectionCloningTest {public static void main(String args[]) {// deep cloning Collection in JavaCollection<Employee> org = new HashSet<Employee>();org.add(new Employee("Joe", "Manager"));org.add(new Employee("Tim", "Developer"));org.add(new Employee("Frank", "Developer"));// creating copy of Collection using copy constructorCollection<Employee> copy = new HashSet<Employee>(org);System.out.println("Original Collection {} " + org);System.out.println("Copy of Collection {} " + copy);Iterator<Employee> itr = org.iterator();while (itr.hasNext()) {itr.next().setDesignation("staff");}System.out.println("Original Collection after modification {} " + org);System.out.println("Copy of Collection without modification {} " + copy);// deep Cloning List in Java}
}class Employee {private String name;private String designation;public Employee(String name, String designation) {this.name = name;this.designation = designation;}public String getDesignation() {return designation;}public void setDesignation(String designation) {this.designation = designation;}public String getName() {return name;}public void setName(String name) {this.name = name;}@Overridepublic String toString() {return String.format("%s: %s", name, designation);}
}
输出:
1 2 3 4 | - Original Collection [Joe: Manager, Frank: Developer, Tim: Developer] - Copy of Collection [Joe: Manager, Frank: Developer, Tim: Developer] - Original Collection after modification [Joe: staff, Frank: staff, Tim: staff] - Copy of Collection without modification [Joe: staff, Frank: staff, Tim: staff] |
可以看到改变原始Collection
中Employee
对象(改变designation为”staff
“)在克隆集合中也有所反映,因为克隆是浅拷贝,指向堆中相同的Employee
对象。为了修正这个问题,需要遍历集合,深克隆Employee
对象,在这之前,要重写Employee
对象的clone方法。
1)Employee
实现Cloneable
接口
2)为Employee
类增加下面的clone()
方法
1 2 3 4 5 6 7 8 9 10 11 12 | @Override protected Employee clone() { Employee clone = null ; try { clone = (Employee) super .clone(); } catch (CloneNotSupportedException e){ throw new RuntimeException(e); // won't happen } return clone; } |
3)不使用拷贝构造函数,使用下面的代码来深拷贝集合
1 2 3 4 5 6 | Collection<Employee> copy = new HashSet<Employee>(org.size()); Iterator<Employee> iterator = org.iterator(); while (iterator.hasNext()){ copy.add(iterator.next().clone()); } |
Code
package javaBasic;import java.util.Collection;
import java.util.HashSet;
import java.util.Iterator;/*** Java program to demonstrate copy constructor of Collection provides shallow* copy and techniques to deep clone Collection by iterating over them.* * @author http://javarevisited.blogspot.com*/
public class CollectionCloningTest {public static void main(String args[]) {// deep cloning Collection in JavaCollection<Employee> org = new HashSet<Employee>();org.add(new Employee("Joe", "Manager"));org.add(new Employee("Tim", "Developer"));org.add(new Employee("Frank", "Developer"));// creating copy of Collection using copy constructor// Collection<Employee> copy = new HashSet<Employee>(org);/*** 不使用拷贝构造函数,使用下面的代码来深拷贝集合*/Collection<Employee> copy = new HashSet<Employee>(org.size());Iterator<Employee> iterator = org.iterator();while (iterator.hasNext()) {copy.add(iterator.next().clone());}System.out.println("Original Collection {} " + org);System.out.println("Copy of Collection {} " + copy);Iterator<Employee> itr = org.iterator();while (itr.hasNext()) {itr.next().setDesignation("staff");}System.out.println("Original Collection after modification {} " + org);System.out.println("Copy of Collection without modification {} " + copy);// deep Cloning List in Java}
}class Employee implements Cloneable {private String name;private String designation;public Employee(String name, String designation) {this.name = name;this.designation = designation;}public String getDesignation() {return designation;}public void setDesignation(String designation) {this.designation = designation;}public String getName() {return name;}public void setName(String name) {this.name = name;}@Overridepublic String toString() {return String.format("%s: %s", name, designation);}@Overrideprotected Employee clone() {Employee clone = null;try {clone = (Employee) super.clone();} catch (CloneNotSupportedException e) {throw new RuntimeException(e); // won't happen}return clone;}
}
4)运行相同的代码更改原始集合,克隆集合不会也被更改。
1 2 | - Original Collection after modification [Joe: staff, Tim: staff, Frank: staff] - Copy of Collection without modification [Frank: Developer, Joe: Manager, Tim: Developer] |
可以看到克隆集合和原始集合相互独立,它们指向不同的对象。
这就是Java中如何克隆集合的内容。现在我们知道拷贝构造函数或者List
或Set
等各种集合类的addAll()
方法仅仅创建了集合的浅拷贝,而且原始集合和克隆集合指向相同的对象。为避免这个问题,应该深克隆集合,遍历集合克隆每个元素。尽管这要求集合中的对象必须支持深克隆操作。