1. 前言

前文我们用java语言实现了无哨兵的单向链表.稍作修改即可实现有哨兵的单向链表.有哨兵的单向链表相较与无哨兵的而言，其对链表的头结点的增删操作更为方便.而在此我们实现了带有头节点和尾节点的双向链表(该头节点和尾节点都不存储有效的数据).

2. 带有头节点和尾节点的双向链表

例 : 

java">public class BinaryLinkedList implements Iterable<Integer> {//头指针指向头节点static Node prev = new Node(null, 0, null);//尾指针指向尾节点static Node tail = new Node(null, 0, null);public BinaryLinkedList() {//将头节点的next指针指向尾节点//将尾节点的prev指针指向头节点prev.next = tail;tail.prev = prev;}private static class Node {int value;Node prev;Node next;public Node(Node prev, int value, Node next) {this.prev = prev;this.value = value;this.next = next;}}//头插法public void addHead(int value) {Node p = new Node(prev, value, prev.next);prev.next.prev = p;prev.next = p;}//从头开始遍历public void Traverse1Head() {Node p;for (p = prev.next; p != tail; p = p.next) {//空链表时, p==nullif (p == null) {return;}System.out.println("该节点的值为" + p.value);}}//从尾开始遍历public void Traverse1Tail() {Node p;for (p = tail.prev; p != prev; p = p.prev) {if (p == null) {return;}System.out.println("该节点的值为" + p.value);}}//获取指定位置的值public static int get(int index) {Node p = findIndex(index);return p.value;}//从头指针开始找指定索引的节点的值private static Node findIndex(int index) {int count = -1;Node p = prev;//这里允许index==-1的原因是此时返回的是头节点//有实际意义, 因为此时insert函数就无需对插入位置为0时做出分析if (index < -1) {throw new RuntimeException("index输入不合法");}while (count < index) {p = p.next;//此时p可能为null, 所以需要判断if (p == null) {throw new RuntimeException("输入无效的index");}count++;}//如果p是尾节点, 将抛出异常if (p == tail) {throw new RuntimeException("尾节点不可操作");}return p;}//尾插法public static void addLast(int value) {Node p = new Node(tail.prev, value, tail);tail.prev.next = p;tail.prev = p;}public void Insert(int index, int value) {//如果index==0, 则p是头节点Node p = findIndex(index - 1);Node insert = new Node(p, value, p.next);p.next.prev = insert;p.next = insert;}public int remove(int index) {//找到要删除的节点Node p = findIndex(index);int value = p.value;p.next.prev = p.prev;p.prev.next = p.next;return value;}//迭代器迭代的数据类型是整形, 又由于泛型不能是基本数据类型, 所以用到包装类@Overridepublic Iterator<Integer> iterator() {//使用到了匿名内部类return new Iterator<Integer>() {//成员变量p(实例变量)Node p = prev.next;@Overridepublic boolean hasNext() {if (p != null && p != tail) {return true;}return false;}@Overridepublic Integer next() {int value = p.value;p = p.next;return value;}};}public void Traverse2(Consumer<Integer> consumer) {Node p;for (p = prev.next; p != tail; p = p.next) {if (p == null) {return;}consumer.accept(p.value);}}
}

3. 单元测试(测试案例)

测试案例供参考 : 

java">@Testpublic void test1() {BinaryLinkedList b = new BinaryLinkedList();b.addHead(12);b.addHead(23);b.addHead(34);b.addHead(45);b.addHead(56);b.addHead(67);b.addHead(78);
//        b.Traverse1Head();
//        b.Traverse1Tail();
//        System.out.println(b.get(0));}@Testpublic void test2() {BinaryLinkedList b = new BinaryLinkedList();b.addLast(12);b.addLast(23);b.addLast(34);b.addLast(45);b.addLast(56);b.addLast(67);b.addLast(78);//只有实现了Intrable接口的类才能使用foreach循环,//因为foreach底层就是使用迭代器不断调用hasNext(), next()方法//迭代出值赋值给临时变量elementfor (Integer element : b) {System.out.println("该节点的数据域是" + element);}}@Testpublic void test3() {BinaryLinkedList b = new BinaryLinkedList();b.addLast(12);b.addLast(23);b.addLast(34);b.addLast(45);b.addLast(56);b.addLast(67);b.addLast(78);//使用lambda表达式b.Traverse2(value -> System.out.println("该节点的数据域的值是" + value));System.out.println("*******************");//还可以使用方法引用b.Traverse2(System.out :: println);}