LRU结构(采用hashmap + 双向链表实现)

采用HashMap加上双向链表可以做到get和put的时间复杂度达到O(1)。map里面的value存的就是双向链表的节点（key,value），是相互连接在一起的。

　　

import java.util.HashMap;
//双向链表节点
class dulNode{ 
    int key;
    int val;
    dulNode next;
    dulNode pre;
    public dulNode(int key, int val){
        this.key = key;
        this.val = val;
        this.next = null;
        this.pre = null;
    }
}
//双向链表
class dulList{
    dulNode head;
    dulNode tail;
    int size;
    public dulList(){
        head = new dulNode(0, 0);
        tail = new dulNode(0, 0);
        head.next = tail;
        tail.pre = head;
        size = 0;
    }
    public void addFirst(dulNode node){
        node.next = head.next;
        head.next.pre = node;
        head.next = node;
        node.pre = head;
        size ++;
    }
    public void remove(dulNode node){
        node.next.pre = node.pre;
        node.pre.next = node.next;
        size --;
    }
    public dulNode removeLast(){
        if(tail.pre == head)
            return null;
        else{
            dulNode last = tail.pre;
            remove(tail.pre);
            return last;
        }
    }
    public int size(){
        return size;
    }
}

class LRUCache {
    int cap;

    HashMap<Integer, dulNode> map;    
    dulList list;
    public LRUCache(int capacity) {
        this.cap = capacity;
        map = new HashMap<Integer, dulNode>();
        list = new dulList();
    }
    
    public int get(int key) {
        if(!map.containsKey(key)){
            return -1;
        }
        int val = map.get(key).val;
        put(key, val);
        return val;
    }
    
    public void put(int key, int value) {
        dulNode node = new dulNode(key, value);
        if(map.containsKey(key)){
            list.remove(map.get(key));
            list.addFirst(node);
            map.put(key, node);
        }
        else{
            if(cap == list.size()){
                dulNode last = list.removeLast();
                map.remove(last.key);    //同时在map里面把这个key删除
            }
            list.addFirst(node);
            map.put(key, node);
        }
    }
}