题意

给你 \(n\) 个 \(01\) 串 \(S\) ,其中有些位置可能为 \(?\) 表示能任意填 \(0/1\) 。问对于所有填法,把所有串插入到 \(Trie\) 的节点数之和(空串看做根节点)。

\(n \le 20, 1 \le |S_i| \le 50\)

题解

直接算显然不太好算的。

\(Trie\) 的节点数其实就是本质不同的前缀个数,可以看做 \(n\) 个串的所有前缀的并集的大小。

我们可以套用容斥原理最初的式子。

\[\left| \bigcup_{i=1}^n A_i \right| = \sum_{k = 1}^{n} (-1)^{k - 1} \sum_{1 \le i_1 < i_2 < \cdots < i_k \le n} |A_{i_1} \cap A_{i_2} \cap \cdots \cap A_{i_k}|
\]

这样的话,我们就可以转化成对于每个子集的交集了,也就是公共前缀的个数。

我们设 \(f(S)\) 为 \(S\) 集合内的所有子串对于 所有填的方案 的公共前缀的个数。

那么答案为 \(ans = \sum_{S \subseteq T} (-1)^{|S| - 1} f(S)\)

如何得到呢?由于 \(n\) 很小我们可以暴力枚举集合,然后枚举当前前缀的长度,直接计数。

  1. 如果当前所有的都是 \(?\) 那么意味着可以任意填 \(0/1\) 。
  2. 如果存在一种数字,其他都是 \(?\) ,那么意味着只能填这种数字。
  3. 如果存在两种数字,那么之后都不可能为公共前缀了,直接退出即可。

直接实现是 \(O(2^n n |S|)\) 的。可以把状态集合合并一下优化到 \(O(2^n |S|)\) 。(但是我太懒了)

代码

实现的时候不要忘记是所有填的方案。

#include <bits/stdc++.h>

#define For(i, l, r) for (register int i = (l), i##end = (int)(r); i <= i##end; ++i)
#define Fordown(i, r, l) for (register int i = (r), i##end = (int)(l); i >= i##end; --i)
#define Rep(i, r) for (register int i = (0), i##end = (int)(r); i < i##end; ++i)
#define Set(a, v) memset(a, v, sizeof(a))
#define Cpy(a, b) memcpy(a, b, sizeof(a))
#define debug(x) cout << #x << ": " << (x) << endl using namespace std; template<typename T> inline bool chkmin(T &a, T b) { return b < a ? a = b, 1 : 0; }
template<typename T> inline bool chkmax(T &a, T b) { return b > a ? a = b, 1 : 0; } inline int read() {
int x(0), sgn(1); char ch(getchar());
for (; !isdigit(ch); ch = getchar()) if (ch == '-') sgn = -1;
for (; isdigit(ch); ch = getchar()) x = (x * 10) + (ch ^ 48);
return x * sgn;
} void File() {
#ifdef zjp_shadow
freopen ("1646.in", "r", stdin);
freopen ("1646.out", "w", stdout);
#endif
} const int N = 21, L = 51, Mod = 998244353; int n, len[1 << N], Pow[N * L]; char str[N][L]; int main () { File(); n = read(); Set(len, 0x3f); int tot = 0; Rep (i, n) {
scanf ("%s", str[i] + 1);
len[1 << i] = strlen(str[i] + 1);
For (j, 1, strlen(str[i] + 1))
if (str[i][j] == '?') ++ tot;
}
Pow[0] = 1;
For (i, 1, tot)
Pow[i] = 2ll * Pow[i - 1] % Mod; Rep (i, 1 << n)
chkmin(len[i], min(len[i ^ (i & -i)], len[i & -i])); int ans = 0;
Rep (i, 1 << n) if (i) {
int res = 0, sum = tot, pre = 0;
For (j, 1, len[i]) {
int flag = 0, now = 0;
Rep (k, n) if (i >> k & 1) {
if (str[k][j] == '?') ++ now;
else flag |= (str[k][j] - '0' + 1);
}
sum -= now;
if (flag == 3) break;
if (!flag) ++ pre;
res = (res + Pow[pre + sum]) % Mod;
}
ans = (ans + (__builtin_popcount(i) & 1 ? 1 : -1) * res) % Mod;
}
ans += Pow[tot]; if (ans < 0) ans += Mod;
printf ("%d\n", ans); return 0; }

hihoCoder #1646 : Rikka with String II(容斥原理)的更多相关文章

  1. HDU 5831 Rikka with Parenthesis II(六花与括号II)

    31 Rikka with Parenthesis II (六花与括号II) Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 65536 ...

  2. HDU 5831 Rikka with Parenthesis II (栈+模拟)

    Rikka with Parenthesis II 题目链接: http://acm.hdu.edu.cn/showproblem.php?pid=5831 Description As we kno ...

  3. 【Hihocoder1413】Rikka with String(后缀自动机)

    [Hihocoder1413]Rikka with String(后缀自动机) 题面 Hihocoder 给定一个小写字母串,回答分别把每个位置上的字符替换为'#'后的本质不同的子串数. 题解 首先横 ...

  4. hdu 5831 Rikka with Parenthesis II 线段树

    Rikka with Parenthesis II 题目连接: http://acm.hdu.edu.cn/showproblem.php?pid=5831 Description As we kno ...

  5. HDU 5831 Rikka with Parenthesis II (贪心)

    Rikka with Parenthesis II Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 65536/65536 K (Jav ...

  6. hdu 5831 Rikka with Parenthesis II 括号匹配+交换

    Rikka with Parenthesis II Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 65536/65536 K (Jav ...

  7. hdu.5202.Rikka with string(贪心)

    Rikka with string Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 65536/65536 K (Java/Others ...

  8. [LeetCode] 344 Reverse String && 541 Reverse String II

    原题地址: 344 Reverse String: https://leetcode.com/problems/reverse-string/description/ 541 Reverse Stri ...

  9. leetcode 344. Reverse String 、541. Reverse String II 、796. Rotate String

    344. Reverse String 最基础的旋转字符串 class Solution { public: void reverseString(vector<char>& s) ...

随机推荐

  1. vue: WebStorm设置快速编译运行

    WebSorm是一款优秀的前端开发工具,而Vue项目可以使用Node进行编译运行,平常我们可以通过命令行部署项目进行调试. 本文介绍设置Webstorm进行快速部署Vue项目. 第一步 点击启动快捷按 ...

  2. redis中的hash、列表、集合操作

    一.hash操作 数据结构:key:{k1:v1, k2:v2, k3:v3} 类似Python中的字典 如:info : {name: lina, age: 22, sex: F} hset key ...

  3. 软工网络15团队作业4——Alpha阶段敏捷冲刺

    Deadline: 2018-4-29 10:00PM,以提交至班级博客时间为准. 根据以下要求,团队在日期区间[4.16,4.29]内,任选8天进行冲刺,冲刺当天晚10点前发布一篇随笔,共八篇. 另 ...

  4. No enclosing instance of type is accessible. Must qualify the allocation with an enclosing instance of type LeadRestControllerTest (e.g. x.new A() where x is an instance of ).

    java - No enclosing instance is accessible. Must qualify the allocation with an enclosing instance o ...

  5. 配置nginx反向代理服务器,解决浏览器跨域调用接口的限制问题

    配置nginx反向代理服务器,解决浏览器跨域调用接口的限制问题 - 大venn的博客 - CSDN博客https://blog.csdn.net/u011135260/article/details/ ...

  6. 使用node写一个简单的页面操作

    let http = require('http'); let urlStr = require('url'); let fs = require('fs'); let path = require( ...

  7. IdentityServer4【Introduction】之包和项目构建

    包和项目构建 IdentityServer包含了以下的nuget包: IdentityServer4 nuget | github 这个包包含了IdentityServer核心的组成部分,有对象模型, ...

  8. CSS3圆角详解(border-radius)

    1.CSS3圆角的优点 传统的圆角生成方案,必须使用多张图片作为背景图案.CSS3的出现,使得我们再也不必浪费时间去制作这些图片了,而且还有其他多个优点: 减少维护的工作量.图片文件的生成.更新.编写 ...

  9. Maven 项目 查找指定包的引用位置

  10. python数据结构与算法第十天【插入排序】

    1.插入排序的原理 2.代码实现 def insert_sort(alist): # 从第二个位置,即下标为1的元素开始向前插入 for i in range(1, len(alist)): # 从第 ...