学步园

题意：给定n（4<=n<=100000）个数字组成的环形序列，有m（4<=m<=100000）个操作p，v表示把p位置的数组更改为v，每次更改后求出最大子段和，

         但是要求不能所有的数字都被选择。

题解：最大子段和的维护，因为是环形，需要同时维护最小子段和，至于能不能全选，维护最小值判断下即可。

Sure原创，转载请注明出处

#include <iostream>
#include <cstdio>
#include <memory.h>
#define MIN(a , b) ((a) < (b) ? (a) : (b))
#define MAX(a , b) ((a) > (b) ? (a) : (b))
using namespace std;
const int inf = 1 << 29;
const int maxn = 100002;
struct T
{
    int l,r;
    int min,sum;
    int lminsum,rminsum,minsum;
    int lmaxsum,rmaxsum,maxsum;
}seg[maxn << 2];
int n,m;

void biuld(int l,int r,int num)
{
    seg[num].l = l;
    seg[num].r = r;
    if(l == r) return;
    int mid = (l + r) >> 1;
    biuld(l , mid , num << 1);
    biuld(mid + 1 , r , num << 1 | 1);
    return;
}

void UP(int num)
{
    seg[num].sum = seg[num << 1].sum + seg[num << 1 | 1].sum;
    seg[num].min = MIN(seg[num << 1].min , seg[num << 1 | 1].min);
    seg[num].lmaxsum = MAX(seg[num << 1].sum + seg[num << 1 | 1].lmaxsum , seg[num << 1].lmaxsum);
    seg[num].rmaxsum = MAX(seg[num << 1 | 1].sum + seg[num << 1].rmaxsum , seg[num << 1 | 1].rmaxsum);
    seg[num].maxsum = MAX(seg[num << 1].maxsum , seg[num << 1 | 1].maxsum);
    seg[num].maxsum = MAX(seg[num].maxsum , MAX(seg[num].lmaxsum , seg[num].rmaxsum));
    seg[num].maxsum = MAX(seg[num].maxsum , seg[num << 1].rmaxsum + seg[num << 1 | 1].lmaxsum);

    seg[num].lminsum = MIN(seg[num << 1].sum + seg[num << 1 | 1].lminsum , seg[num << 1].lminsum);
    seg[num].rminsum = MIN(seg[num << 1 | 1].sum + seg[num << 1].rminsum , seg[num << 1 | 1].rminsum);
    seg[num].minsum = MIN(seg[num << 1].minsum , seg[num << 1 | 1].minsum);
    seg[num].minsum = MIN(seg[num].minsum , MIN(seg[num].lminsum , seg[num].rminsum));
    seg[num].minsum = MIN(seg[num].minsum , seg[num << 1].rminsum + seg[num << 1 | 1].lminsum);
    return;
}

void update(int pos,int val,int num)
{
    if(seg[num].l == seg[num].r)
    {
        seg[num].min = seg[num].sum = val;
        seg[num].lmaxsum = seg[num].rmaxsum = seg[num].maxsum = MAX(0 , val);
        seg[num].lminsum = seg[num].rminsum = seg[num].minsum = MIN(0 , val);
        return;
    }
    int mid = (seg[num].l + seg[num].r) >> 1;
    if(pos <= mid) update(pos , val , num << 1);
    else update(pos , val , num << 1 | 1);
    UP(num);
    return;
}

void read()
{
    int val;
    for(int i=1;i<=n;i++)
    {
        scanf("%d",&val);
        update(i , val , 1);
    }
    return;
}

void solve()
{
    scanf("%d",&m);
    int p,v;
    while(m--)
    {
        scanf("%d %d",&p,&v);
        update(p , v , 1);
        if(seg[1].min >= 0) printf("%d\n",seg[1].sum - seg[1].min);
        else printf("%d\n",MAX(seg[1].maxsum , seg[1].sum - seg[1].minsum));
    }
    return;
}

int main()
{
    while(~scanf("%d",&n))
    {
        biuld(1,n,1);
        read();
        solve();
    }
    return 0;
}