数据结构设计问题

[air_snake]

[1] Design a layer in front of a system which cache the last n requests and
the responses to them from the system.
what data structure would you use to implement the cache in the later to
support following operations.
[a] When a request comes look it up in the cache and if it hits then return
the response from here and do not pass the request to the system
[b] If the request is not found in the cache then pass it on to the system
[c] Since cache can only store the last n requests, Insert the n+1th request
in the cache and delete one of the older requests from the cache

The objective is to achieve all the three operations in O(1).

[michaellufhl]

Hash based data structure can offer O(1), and you need to fix the size of the hash structure.

[dingshaofengbinbin]

循环队列。。。

[air_snake]

[Quote=引用 1 楼 michaellufhl 的回复:]

Hash based data structure can offer O(1), and you need to fix the size of the hash structure.
[/Quote]
只用hash怎么替换oldest的。

[air_snake]

[Quote=引用 2 楼 dingshaofengbinbin 的回复:]

循环队列。。。
[/Quote]
这个不是查询不是O(1)吧

[michaellufhl]

So we may mix up 2 solutions: hash + linked list(queue as dingshaofengbinbin mentioned) which has 2 points to head and tail.

Each element is stored both in hash set and linked list(same reference).

a) Lookup:
hash set offers O(1).

b) Put and remove oldest element:
Add new : add the new one by head point in linked list, and put it in hash set.
Remove oldest: find the one by tail point in linked list, remove it both in hash set and list.

Sorry, the solution offers O(2) when put/remove.

[michaellufhl]

Usually, we could chose binary search tree to solve the problem. such as red-black tree.

[zlntang]

下次能用中文不？

[showjim]

hash指向单链表目标元素的上一结点，记录链表尾结点与当前结点数

[yingkoz]

英文的！看不懂。

[xmx2009]

几个月没来，都变成了英语牛人了，佩服

[air_snake]

不好意思，我看一个面试网站上的，懒得翻译，就没改。
我觉得用FIFO加hash应该可以了，被替换出去的，可以去hash里面找，然后删除
FIFO用链表，添加删除头尾也是O(1)的了。

[超级大笨狼]

这个用一个先入先出的栈来做。Stack<T>

[绿色夹克衫]

循环队列+hash应该差不多，尤其是在实际应用当中，肯定可以，但如果细扣题目要求的话，可能还是稍微有些问题，但估计题目本意并不是非要一丝不差。

[zmr0469]

hash加忙链表，每次更新链表