Victor and Proposition

Time Limit: 6000ms
Memory Limit: 524288KB

This problem will be judged on HDU. Original ID: 5420
64-bit integer IO format: %I64d      Java class name: Main

At the very beginning, Victor has a proposition, then this proposition procudes many propositions. Then every proposition procudes more propositions...... Finally there are n propositions. These propositions can be regarded as a tree whose root is 1.

We assume that the first proposition, whose number is 1, belongs to the 0-th generation, and those propositions produced by the x-th generation belong to the x+1-th generation. We also assume that all of the propositions in the x-th generation are in level x. Specially, Victor has discovered that the proposition whose number is i can infer the proposition whose number is xi and all of the propositions in xi's subtree, whose levels are not greater than xi's level + di.

Notice : a is b's father does not show that either a can infer b or b can infer a.

Now please determine the number of such ordered pairs (i,j), that 1≤i<j≤n, the proposition i can infer the proposition j, and the proposition j can also infer the proposition i.

Input
The first line of the input contains an integer T, denoting the number of test cases.

In every test case, there is an integer n in the first line, denoting the number of the propositions.

The second line contains n−1 integers, the i-th integer fi+1(fi<i) denotes that the proposition i+1 is produced by the proposition fi+1.

Then there are n lines, the i-th line contains two integers xi and di.

1≤T≤5.

2≤n≤100000.

0≤di<n.

Output
Your program should print T lines : the i-th of these should contain a single integer, denoting the number of such ordered pairs (i,j).

Sample Input
1
4
1 2 1
2 1
1 0
4 0
2 0

Sample Output
6

Source
BestCoder Round #52 (div.2)

解题:线段树优化建图,妙哉,内存开得好凶残,吓呆本宝宝了

 #include <bits/stdc++.h>
using namespace std;
typedef long long LL;
typedef pair<int,int> pii;
const int INF = 0x3f3f3f3f;
const int maxn = ;
struct arc {
int to,next;
arc(int x = ,int y = -) {
to = x;
next = y;
}
} e[(+)**];
int head[maxn],L[maxn],R[maxn],tot,clk;
int dep[maxn],hs[maxn],st[maxn],n;
vector<pii>order[maxn];
void add(int u,int v) {
e[tot] = arc(v,head[u]);
head[u] = tot++;
}
void init() {
tot = ;
memset(head,-,sizeof head);
}
void dfs(int u,int depth) {
hs[L[u] = ++clk] = u;
dep[u] = depth;
for(int i = head[u]; ~i; i = e[i].next) dfs(e[i].to,depth + );
R[u] = clk;
}
void build(int L,int R,int v) {
order[v].resize(R - L + );
if(L == R) {
st[v] = ++n;
order[v][] = pii(dep[hs[L]],hs[L]);
add(n,hs[L]);
return;
}
int mid = (L + R)>>;
build(L,mid,v<<);
build(mid + ,R,v<<|);
st[v] = n + ;
merge(order[v<<].begin(),order[v<<].end(),order[v<<|].begin(),order[v<<|].end(),order[v].begin());
for(int i = ; i <= R - L; ++i)
add(n + i + , n + i);
for(int i = ; i <= R - L; ++i)
add(n + i + ,order[v][i].second);
n += R - L + ;
}
void connect(int L,int R,int lt,int rt,int u,int d,int v) {
if(lt <= L && rt >= R) {
int pos = lower_bound(order[v].begin(),order[v].end(),pii(d,INF)) - order[v].begin() - ;
if(~pos) add(u,st[v] + pos);
return;
}
int mid = (L + R)>>;
if(lt <= mid) connect(L,mid,lt,rt,u,d,v<<);
if(rt > mid) connect(mid + ,R,lt,rt,u,d,v<<|);
}
int dfn[maxn],low[maxn],cnt[maxn],scc,ct;
bool instack[maxn];
stack<int>stk;
void tarjan(int u) {
dfn[u] = low[u] = ++ct;
instack[u] = true;
stk.push(u);
for(int i = head[u]; ~i; i = e[i].next) {
if(!dfn[e[i].to]) {
tarjan(e[i].to);
low[u] = min(low[u],low[e[i].to]);
} else if(instack[e[i].to]) low[u] = min(low[u],dfn[e[i].to]);
}
if(low[u] == dfn[u]) {
int v;
cnt[++scc] = ;
do {
instack[v = stk.top()] = false;
stk.pop();
cnt[scc] += (v <= clk);
} while(v != u);
}
}
int main() {
int kase,u,v;
scanf("%d",&kase);
while(kase--) {
scanf("%d",&n);
init();
clk = scc = ct = ;
memset(dfn,,sizeof dfn);
memset(instack,false,sizeof instack);
for(int i = ; i <= n; ++i) {
scanf("%d",&u);
add(u,i);
}
dfs(,);
init();
build(,clk,);
for(int i = ; i <= clk; ++i) {
scanf("%d%d",&u,&v);
connect(,clk,L[u],R[u],i,dep[u] + v,);
}
for(int i = ; i <= n; ++i)
if(!dfn[i]) tarjan(i);
LL ret = ;
for(int i = ; i <= scc; ++i)
ret += (LL)cnt[i]*(cnt[i]-)/;
printf("%I64d\n",ret);
}
return ;
}

HDU 5420 Victor and Proposition的更多相关文章

  1. ACM: HDU 5418 Victor and World - Floyd算法+dp状态压缩

    HDU 5418 Victor and World Time Limit:2000MS     Memory Limit:131072KB     64bit IO Format:%I64d & ...

  2. HDU 5417 Victor and Machine

    题目链接:http://acm.hdu.edu.cn/showproblem.php?pid=5417 Problem Description Victor has a machine. When t ...

  3. HDU 5418 Victor and World 允许多次经过的TSP

    题目链接: hdu: http://acm.hdu.edu.cn/showproblem.php?pid=5418 bestcoder(中文): http://bestcoder.hdu.edu.cn ...

  4. HDU 5418 Victor and World(状压DP+Floyed预处理)

    Victor and World Time Limit: 4000/2000 MS (Java/Others)    Memory Limit: 262144/131072 K (Java/Other ...

  5. HDU 5421 Victor and String

    Victor and String Time Limit: 1000ms Memory Limit: 262144KB This problem will be judged on HDU. Orig ...

  6. HDU - 5419 Victor and Toys(组合计数)

    http://acm.hdu.edu.cn/showproblem.php?pid=5419 题意 n个物品,标号1-n,物品i有权值wi.现在有m个区间[l,r],从中任意选三个区间i,j,k,求物 ...

  7. HDU 5421 Victor and String(回文树)

    Victor and String Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 524288/262144 K (Java/Othe ...

  8. HDU 5419——Victor and Toys——————【线段树|差分前缀和】

    Victor and Toys Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 262144/131072 K (Java/Others ...

  9. HDU 5418——Victor and World——————【状态压缩+floyd】

    Victor and World Time Limit: 4000/2000 MS (Java/Others)    Memory Limit: 262144/131072 K (Java/Other ...

随机推荐

  1. [POI2012]Vouchers

    Description 考虑正整数集合,现在有n组人依次来取数,假设第i组来了x人,他们每个取的数一定是x的倍数,并且是还剩下的最小的x个. 正整数中有m个数被标成了幸运数,问有哪些人取到了幸运数. ...

  2. poj 2506 Tiling 递推

    题目链接: http://poj.org/problem?id=2506 题目描述: 有2*1和2*2两种瓷片,问铺成2*n的图形有多少种方法? 解题思路: 利用递推思想,2*n可以由2*(n-1)的 ...

  3. ACM牛人博客

    ACM牛人博客 kuangbin kuangbin(新) wuyiqi wuyiqi(新) ACM!荣耀之路! 九野的博客 传说中的ACM大牛!!! read more

  4. 08 H5新增input元素

    <!doctype html> <html> <head> <meta charset="utf-8"> <title> ...

  5. 在spring data jpa中使用自定义转换器之使用枚举转换

    转载请注明http://www.cnblogs.com/majianming/p/8553217.html 在项目中,经常会出现这样的情况,一个实体的字段名是枚举类型的 我们在把它存放到数据库中是需要 ...

  6. c# -反射 初见

    反射是一个很强大的功能,不过好像有些消耗性能,大家慎重使用. 1.反射是干什么的? 通过反射,我们可与获取程序集中的原数据. 1.什么是程序集? dll.exe  这些将很多能实现具体功能的代码封装起 ...

  7. Android 在代码中安装 APK 文件

    废话不说,上代码 private void install(String filePath) { Log.i(TAG, "开始执行安装: " + filePath); File a ...

  8. sql server查看某个表上的触发器

    用企业管理器查看 在某个具体的表上点右键->“所有任务”->“管理触发器”,选择所要查看的触发器

  9. Asp.Net MVC之 自动装配、动态路径(链接)等

    一.Model层 using System; using System.Collections.Generic; using System.Linq; using System.Web; namesp ...

  10. Java基础知识强化98.01:Jsp和servlet有什么区别

    1. Jsp和servlet有什么区别 首先你先要弄懂什么是servlet,servlet是在服务器端执行的java程序,只不过它有专门的一套规则(就是我们平常所说的api):jsp说得简单点就是用另 ...