LRU cache的简单实现

来源：互联网发布：下载xlsx软件编辑：程序博客网时间：2024/05/21 10:29

从昨天想起这个东西，中午写好了，但是leetcode上还木有ac:

就是用双链表存储，同时用一个map存储链表：

LRUCache(int capacity):maxSize(capacity), curSize(0){pHead = new LinkedList(-1, -1);pHead->pre = pHead;pHead->next = pHead;}~LRUCache(){for (MapList::iterator iter = mapList.begin(); iter != mapList.end(); ++iter){delete iter->second;}delete pHead;}void set(int key,int val){MapList::iterator iter = mapList.find(key);if (mapList.end() == iter){if (maxSize == curSize){LinkedList* pToDel = pHead->pre;pHead->pre = pToDel->pre;pToDel->pre->next = pHead;mapList.erase(pToDel->key);pToDel->key = key;pToDel->val = val;mapList[key] = pToDel; }else{LinkedList* pCurList = new LinkedList(key, val);pHead->next->pre = pCurList;pCurList->next = pHead->next;pHead->next = pCurList;pCurList->pre = pHead;mapList[key] = pCurList;++curSize;}}else{LinkedList* pCurList = iter->second;pCurList->pre->next = pCurList->next;pCurList->next->pre = pCurList->pre;pCurList->next = pHead->next;pCurList->pre = pHead;pHead->next = pCurList;}}int get(int key){MapList::iterator iter = mapList.find(key);if (mapList.end() == iter){printf("can't find key:%d\n", key);return -1;}else{LinkedList* pCurList = iter->second;pCurList->pre->next = pCurList->next;pCurList->next->pre = pCurList->pre;pCurList->next = pHead->next;pCurList->pre = pHead;pHead->next = pCurList;}return (iter->second)->val;}private:struct LinkedList{int key;int val;LinkedList* pre;LinkedList* next;LinkedList(int key, int val):key(key), val(val), pre(NULL), next(NULL){}};const int maxSize;int curSize;LinkedList* pHead;typedef std::map<int, LinkedList*> MapList;MapList mapList;

这里实现的思路是set时超过了size就删掉最早插入的那个节点。

貌似木有用到get这个查询的用法，想了下还是保存查询的次数纪录，可以用位图的办法。下午吃饭的时候再写。

0 0