POJ 3415 Common Substrings 后缀数组 + 线段树 + dfs

题意：给出两个字符串A，B和一个数字K。

计算S = {(i,j,k) | k >= K ,A(i,k) == B(j,k)}这个集合的元素个数。

思路： 首先后缀数组处理出 high[]，然后对high[]建立线段树，线段树的每个节点记录两个信息：对应区间的最小值和最小值的位置(下标)，继而对线段树进行dfs。

dfs的过程都写在注释里了。

void dfs(int l,int r,LL &anw,int len,int Low)
{
    if(l > r)
        return ;
        
    N tmp = Query(1,1,len,l,r);//询问此段区间内的最小值及其位置。

    if(tmp.Min >= Low)
    {
        //此段区间内分属A，B串的个数相乘然后与可取长度的个数相乘。
        anw += (ans1[r]-ans1[l-2])*(ans2[r]-ans2[l-2])*(tmp.Min-Low+1);
        dfs(l,tmp.site-1,anw,len,tmp.Min+1);//从最小值处分开继续dfs，[l,r]内的tmp.Min均以计算，故加一。
        dfs(tmp.site+1,r,anw,len,tmp.Min+1);
        return ;
    }

    dfs(l,tmp.site-1,anw,len,Low);
    dfs(tmp.site+1,r,anw,len,Low);
}

下面为全部代码

#include <algorithm>
#include <iostream>
#include <cstring>
#include <cstdlib>
#include <cstdio>
#include <queue>
#include <cmath>
#include <stack>
#include <map>
#include <ctime>
#include <iomanip>

#pragma comment(linker, "/STACK:1024000000");
#define EPS (1e-6)
#define LL long long
#define ULL unsigned long long
#define _LL __int64
#define INF 0x3f3f3f3f
#define Mod 1000000007

using namespace std;

const int MAXN = 200510;

char s[MAXN];

int Rank[2*MAXN],sa[2*MAXN],tr[2*MAXN],high[MAXN];

struct EDGE
{
    int v,next;
}edge[2*MAXN];

int tail[MAXN],Top;

inline void Link(int u,int v)
{
    edge[Top].v = v;
    edge[Top].next = -1;
    edge[tail[u]].next = Top;
    tail[u] = Top++;
}

void Get_SA(char *s,int n,int m)
{
    memset(Rank,0,sizeof(Rank));
    memset(sa,0,sizeof(sa));

    int i,j,k,ans,site;

    for(i = max(n,m);i >= 0; --i)
        tail[i] = i,edge[i].next = -1;

    Top = max(n,m)+1;

    for(i = 1; i <= n; ++i)
        Link(s[i]-'A',i);

    ans = 1,site = 1;

    for(i = 0; i <= m; ++i)
    {
        for(j = edge[i].next; j != -1; j = edge[j].next)
            sa[site++] = edge[j].v,Rank[edge[j].v] = ans;
        if(edge[i].next != -1)
            ans++;
        tail[i] = i,edge[i].next = -1;
    }

    for(k = 1;k <= n; k <<= 1)
    {
        Top = n+1;
        for(i = 1;i <= n; ++i)
            Link(Rank[sa[i]+k],sa[i]);

        site = 1;

        for(i = 0;i <= n; ++i)
        {
            for(j = edge[i].next;j != -1; j = edge[j].next)
                sa[site++] = edge[j].v;
            tail[i] = i,edge[i].next = -1;
        }

        Top = n+1;

        for(i = 1;i <= n; ++i)
            Link(Rank[sa[i]],sa[i]);

        site = 1;

        for(i = 1;i <= n; ++i)
        {
            for(j = edge[i].next;j != -1; j = edge[j].next)
                sa[site++] = edge[j].v;
            tail[i] = i,edge[i].next = -1;
        }

        for(tr[sa[1]] = 1,i = 2,ans = 1;i <= n; ++i)
        {
            if(Rank[sa[i]] != Rank[sa[i-1]] || Rank[sa[i]+k] != Rank[sa[i-1]+k])
                ans++;
            tr[sa[i]] = ans;
        }

        for(i = 1;i <= n; ++i)
            Rank[i] = tr[i];
        if(ans >= n)
            break;
    }

    for(i = 1,k = 1;i <= n; ++i)
    {
        if(k) k--;
        if(Rank[i] == 1) {k = 0;high[1] = n-sa[1]+1;continue;}
        j = sa[Rank[i]-1];
        while(i+k <= n && j+k <= n && s[i+k] == s[j+k])
            k++;
        high[Rank[i]] = k;
    }
//
//    for(i = 1;i <= n; ++i)
//        printf("i = %2d SA = %2d Rank = %2d high = %2d\n",i,sa[i],Rank[i],high[i]);

    //以上为Rank，SA，HIGH的构造过程
}

LL ans1[MAXN],ans2[MAXN];

struct N
{
    int site,Min;
}st[4*MAXN];

void Init(int site,int l,int r)
{
    if(l == r)
    {
        st[site].Min = high[l],st[site].site = l;
        return ;
    }
    int mid = (l+r)>>1;

    Init(site<<1,l,mid);
    Init(site<<1|1,mid+1,r);

    st[site] = st[site<<1].Min < st[site<<1|1].Min ? st[site<<1] : st[site<<1|1];
}

N Query(int site,int L,int R,int l,int r)
{
    if(L == l && R == r)
        return st[site];
    int mid = (L+R)>>1;
    if(r <= mid)
        return Query(site<<1,L,mid,l,r);
    if(mid < l)
        return Query(site<<1|1,mid+1,R,l,r);
    N t1 = Query(site<<1,L,mid,l,mid);
    N t2 = Query(site<<1|1,mid+1,R,mid+1,r);
    if(t1.Min < t2.Min)
        return t1;
    return t2;
}

//l,r为左右区间端点。anw为最终答案。len为high[]的size,Low为此次dfs中符合要求的最小值，初始时为输入的K。
void dfs(int l,int r,LL &anw,int len,int Low)
{
    if(l > r)
        return ;
        
    N tmp = Query(1,1,len,l,r);//询问此段区间内的最小值及其位置。

    if(tmp.Min >= Low)
    {
        //此段区间内分属A，B串的个数相乘然后与可取长度的个数相乘。
        anw += (ans1[r]-ans1[l-2])*(ans2[r]-ans2[l-2])*(tmp.Min-Low+1);
        dfs(l,tmp.site-1,anw,len,tmp.Min+1);//从最小值处分开继续dfs，[l,r]内的tmp.Min均以计算，故加一。
        dfs(tmp.site+1,r,anw,len,tmp.Min+1);
        return ;
    }

    dfs(l,tmp.site-1,anw,len,Low);
    dfs(tmp.site+1,r,anw,len,Low);
}

int main()
{
    int n,k,len,i;

    while(scanf("%d",&k) && k)
    {
        scanf("%s",s+1);
        n = strlen(s+1);
        scanf("%s",s+n+2);
        s[n+1] = 'z'+1;
        Get_SA(s,len = strlen(s+1),200);
        for(ans1[0] = 0,ans2[0] = 0, i = 1;i <= len; ++i)
        {
            ans1[i] = ans1[i-1],ans2[i] = ans2[i-1];
            if(sa[i] <= n) ans1[i]++;
            if(sa[i] > n+1) ans2[i]++;
        }
        Init(1,1,len);

        LL anw = 0;

        dfs(2,len,anw,len,k);

        printf("%I64d\n",anw);
    }

    return 0;
}