后缀数组练习题若干

POJ 1743    不可重叠最长重复子串

二分答案。 即子串的长度，假设为k时。

利用height数组，将排序后的后缀分为若干组。

每组内的height值都不小于k。

然后只需查看组内是否有满足要求的两个不会产生重叠的子串即可。

#include <iostream>
#include <cstdio>
#include <cstring>
#include <vector>
#include <set>
#include <queue>
#include <algorithm>
#define MAXN 22222
#define MAXM 111
#define INF 1000000000
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
int n, a[MAXN];
bool check(int mid, int n)
{
    int flag = 0;
    int mx = -1, mi = n;
    for(int i = 2; i <= n + 1; i++)
    {
        if((i == n + 1 && flag) || (height[i] < mid && flag))
        {
            flag = 0;
            mx = max(mx, sa[i - 1]);
            mi = min(mi, sa[i - 1]);
            if(mx - mi >= mid) return true;
            mi = n, mx = -1;
        }
        else if(height[i] >= mid)
        {
            flag = 1;
            mx = max(mx, sa[i - 1]);
            mi = min(mi, sa[i - 1]);
        }
    }
    return false;
}
int main()
{
    while(scanf("%d", &n) != EOF && n)
    {
        for(int i = 0; i < n; i++) scanf("%d", &a[i]);
        for(int i = 0; i < n - 1; i++) r[i] = a[i + 1] - a[i] + 89;
        r[--n] = 0;
        da(r, sa, n + 1, 200);
        calheight(r, sa, n);
        int low = 4, high = n / 2, ans = 0;
        while(low <= high)
        {
            int mid = (low + high) >> 1;
            if(check(mid, n))
            {
                low = mid + 1;
                ans = max(ans, mid);
            }
            else high = mid - 1;
        }
        if(ans < 4) printf("0\n");
        else printf("%d\n", ans + 1);
    }
    return 0;
}

POJ 3261 可重叠的出现K次的最长重复子串

还是二分子串长度。 后缀分为若干组，然后判断是否有一个组的size不小于k

#include <iostream>
#include <cstdio>
#include <cstring>
#include <vector>
#include <set>
#include <queue>
#include <algorithm>
#define MAXN 22222
#define MAXM 111
#define INF 1000000000
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
int n, k;
bool check(int mid)
{
    int cnt = 1;
    for(int i = 2; i <= n; i++)
    {
        if(height[i] < mid) cnt = 1;
        else cnt++;
        if(cnt >= k) return 1;
    }
    return 0;
}
int main()
{
    int m = 0;
    scanf("%d%d", &n, &k);
    for(int i = 0; i < n; i++)
    {
        scanf("%d", &r[i]);
        r[i]++;
        m = max(r[i], m);
    }
    r[n] = 0;
    da(r, sa, n + 1, m + 1);
    calheight(r, sa, n);
    int low = 1, high = n;
    int ans = 0;
    while(low <= high)
    {
        int mid = (low + high) >> 1;
        if(check(mid))
        {
            ans = max(ans, mid);
            low = mid + 1;
        }
        else high = mid - 1;
    }
    printf("%d\n", ans);
    return 0;
}

SPOJ SUBST1 求一个串中不同子串的个数

每个子串都是某个后缀的前缀

对于一个后缀。 它将产生n - sa[k]个前缀

但是有height[k]个前缀是跟前一个字符串的前缀相同。

故每个后缀的贡献是n - sa[k] - height[k]

求和即可

#include <iostream>
#include <cstdio>
#include <cstring>
#include <vector>
#include <set>
#include <queue>
#include <algorithm>
#define MAXN 55555
#define MAXM 111
#define INF 1000000000
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
char s[MAXN];
int main()
{
    int T;
    scanf("%d", &T);
    while(T--)
    {
        scanf("%s", s);
        int n = strlen(s);
        int m = 0;
        for(int i = 0; i < n; i++)
        {
            r[i] = (int)s[i];
            m = max(m, r[i]);
        }
        r[n] = 0;
        da(r, sa, n + 1, m + 1);
        calheight(r, sa, n);
        long long ans = 0;
        for(int i = 1; i <= n; i++) ans += n - sa[i] - height[i];
        printf("%lld\n", ans);
    }
    return 0;
}

URAL 1297  求最长回文串

假设原串为S，将原串倒置后是T。

建立一个新串S+“~”+T

然后对新串做后缀数组。

然后我们枚举的是回文串的中心。

假设中心的位置为i。

有两种情况

回文为奇数

那么求lcp(i, n - i - 1)

回文为偶数那么求lcp(i, n - i)

然后更新最优解即可

用手画一画就知道是什么意思了。

#include <iostream>
#include <cstdio>
#include <cstring>
#include <vector>
#include <set>
#include <queue>
#include <algorithm>
#define MAXN 111111
#define MAXM 111
#define INF 1000000000
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
int Log[MAXN];
int mi[MAXN][20];
void rmqinit(int n)
{
    for(int i = 1; i <= n; i++) mi[i][0] = height[i];
    int m = Log[n];
    for(int i = 1; i <= m; i++)
        for(int j = 1; j <= n; j++)
        {
            mi[j][i] = mi[j][i - 1];
            if(j + (1 << (i - 1)) <= n) mi[j][i] = min(mi[j][i], mi[j + (1 << (i - 1))][i - 1]);
        }
}
int lcp(int a, int b)
{
    a = rank[a];    b = rank[b];
    if(a > b) swap(a,b);
    a ++;
    int t = Log[b - a + 1];
    return min(mi[a][t] , mi[b - (1<<t) + 1][t]);
}
char s[MAXN * 2];
int main()
{
    Log[1] = 0;
    for(int i = 2; i < MAXN; i++) Log[i] = Log[i >> 1] + 1;
    while(scanf("%s", s) != EOF)
    {
        int len = strlen(s);
        for(int i = 0; i < len; i++) r[i] = (int)s[i];
        r[len] = 128;
        for(int i = 0; i < len; i++) r[len + 1 + i] = (int)s[len - 1 - i];
        int n = 2 * len + 1;
        r[n] = 0;
        da(r, sa, n + 1, 130);
        calheight(r, sa, n);
        rmqinit(n);
        int ans = 0;
        int pos;
        for(int i = 0; i < len; i++)
        {
            int tmp = lcp(i, n - i - 1); //奇数
            if(tmp * 2 - 1 > ans)
            {
                ans= tmp * 2 - 1;
                pos = i - tmp + 1;
            }
            tmp = lcp(i, n - i); //偶数
            if(tmp * 2 > ans)
            {
                ans = tmp * 2;
                pos = i - tmp;
            }
        }
        for(int i = 0; i < ans; i++) putchar(s[pos + i]);
        puts("");
    }
    return 0;
}

POJ 2406  

给定一个字符串S，已知该串是由某串重复K次 连接得到的。

求最大的k

这题的话。 貌似POJ上暴力跑的很快。

用后缀数组需要的求是枚举子串的长度。

假设长度为len， 那么检查lcp(0, len)是否等于n - len即可

倍增在这里被卡掉了

用的DC3

#include <iostream>
#include <cstdio>
#include <cstring>
#include <vector>
#include <set>
#include <queue>
#include <cmath>
#include <algorithm>
#define MAXN 1111111
#define MAXM 111
#define INF 1000000000
#define F(x) ((x)/3+((x)%3==1?0:tb))
#define G(x) ((x)<tb?(x)*3+1:((x)-tb)*3+2)
using namespace std;
int wa[MAXN] , wb[MAXN] , wv[MAXN] , tmp[MAXN];
int c0(int *r, int a, int b){
    return r[a] == r[b] && r[a + 1] == r[b + 1] && r[a + 2] == r[b + 2];
}
int c12(int k, int *r, int a, int b){
    if (k == 2)
    return r[a] < r[b] || r[a] == r[b] && c12(1, r, a + 1, b + 1);
    else return r[a] < r[b] || r[a] == r[b] && wv[a + 1] < wv[b + 1];
}
void sort(int *r, int *a, int *b, int n, int m)
{
    int i;
    for (i = 0; i < n; i++) wv[i] = r[a[i]];
    for (i = 0; i < m; i++) tmp[i] = 0;
    for (i = 0; i < n; i++) tmp[wv[i]]++;
    for (i = 1; i < m; i++) tmp[i] += tmp[i-1];
    for (i = n-1; i >= 0; i--) b[--tmp[wv[i]]] = a[i];
}
void dc3(int *r, int *sa, int n, int m)
{
    int i, j, *rn = r + n;
    int *san = sa + n, ta = 0, tb = (n + 1) / 3, tbc = 0, p;
    r[n] = r[n + 1] = 0;
    for (i = 0; i < n; i++) if (i % 3 != 0) wa[tbc++] = i;
    sort(r + 2, wa, wb, tbc, m);
    sort(r + 1, wb, wa, tbc, m);
    sort(r, wa, wb, tbc, m);
    for (p = 1, rn[F(wb[0])] = 0, i = 1; i < tbc; i++)
        rn[F(wb[i])] = c0(r, wb[i-1], wb[i]) ? p-1 : p++;
    if (p < tbc) dc3(rn, san, tbc, p);
    else for (i = 0; i < tbc; i++) san[rn[i]] = i;
    for (i = 0; i < tbc; i++) if (san[i] < tb) wb[ta++] = san[i] * 3;
    if (n % 3 == 1) wb[ta++] = n-1;
    sort(r, wb, wa, ta, m);
    for (i = 0; i < tbc; i++) wv[wb[i] = G(san[i])] = i;
    for (i = 0, j = 0, p = 0; i < ta && j < tbc; p++)
        sa[p] = c12(wb[j] % 3, r, wa[i], wb[j]) ? wa[i++] : wb[j++];
    for (; i < ta; p++) sa[p] = wa[i++];
    for (; j < tbc; p++) sa[p] = wb[j++];
}
void da(int str[], int sa[], int rank[], int height[], int n, int m)
{
//       for (int i = n; i < n * 3; i++)
//        str[i] = 0;
    dc3 (str , sa , n + 1 , m);
    int i, j, k;
    for (i = 0; i < n; i++){
        sa[i] = sa[i + 1];
        rank[sa[i]] = i;
    }
    for (i = 0, j = 0, k = 0; i < n; height[rank[i ++]] = k)
        if (rank[i] > 0)
            for (k ? k--: 0 , j = sa[rank[i]-1];
            i + k < n && j + k < n && str[i + k] == str[j + k];
            k++);
}
int lcp[MAXN];
int r[MAXN];
int  sa[MAXN], rank[MAXN] , height[MAXN];
int n;
void getlcp()
{
    int k = rank[0];
    lcp[k] = n;
    for(int i = k; i >= 2; i--)
        lcp[i - 1] = min(lcp[i], height[i]);
    for(int i = k + 1; i <= n; i++)
        lcp[i] = min(lcp[i - 1], height[i]);
}
char s[MAXN];
bool ok(int k)
{
    int rk = rank[k];
    if(lcp[rk] == n - k) return true;
    return false;
}
int main()
{
    while(gets(s))
    {
        if(s[0] == '.') break;
        n = strlen(s);
        for(int i = 0; i <= n; i++) r[i] = s[i];
        da(r, sa, rank, height, n + 1, 130);
        getlcp();
        int tmp = (int)sqrt(n + 0.5);
        int ans = 0;
        for(int i = 1; i <= tmp; i++)
        {
            if(n % i != 0) continue;
            if(ok(i)) ans = max(ans, n / i);
            if(ok(n / i)) ans = max(ans, i);
        }
        printf("%d\n", ans);
    }
    return 0;
}

POJ 3693

题目大意就是求重复次数最多的连续重复子串。例如abababc 答案就是ababab  因为ab连续出现的次数最多

并且题目还要求输出字典序最小的

比如abababcdcdcd 

ababab和cdcdcd都符合要求

但是ababab字典序小

具体做法参见罗穗骞的论文

穷举子串的长度L，然后求长度为L的子串最多出现几次

首先连续出现一次是肯定的，所以只考虑出现两次及以上的情况

假设在字符串中出现了两次，记这个重复了两次L长度子串的子串为S。

那么S肯定包含了字符r[0], r[L], r[L*2], r[3 * L]....中的某相邻的两个。

所以就看r[L*i]和r[L*(i + 1)]往前往后分别匹配到多远，记这个长度为K（具体匹配方式看代码），那么就连续出现了K/L+1次，最后看最大值多少

注意每次求这个k要分为两种情况，一是公共前缀恰好模L等于0，另一种是模L不等于0

模L不等于0时还要计算一个值，假如公共前缀%L等于t,就求lcp(i - (L - t), i - (L - t) + L);

为什么呢，我们画一画就知道了，这样的做法，实际上两个公共前缀往前延伸了几个位置，使得前缀的长度加上延伸的长度是L的倍数

然后求lcp，会发现，他是有可能比原来的k大的，那么连续出现的次数也有可能改变。所以这种情况不能遗漏

这里用到了lcp,既求任意两个后缀的最长公共前缀，使用RMQ实现。

最后输出的时候，由于要按字典序输出，就枚举sa数组

#include <iostream>
#include <cstdio>
#include <cstring>
#include <vector>
#include <set>
#include <queue>
#include <cmath>
#include <algorithm>
#define MAXN 1111111
#define MAXM 111
#define INF 1000000000
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
int Log[MAXN];
int mi[MAXN][20];
void rmqinit(int n)
{
    for(int i = 1; i <= n; i++) mi[i][0] = height[i];
    int m = Log[n];
    for(int i = 1; i <= m; i++)
        for(int j = 1; j <= n; j++)
        {
            mi[j][i] = mi[j][i - 1];
            if(j + (1 << (i - 1)) <= n) mi[j][i] = min(mi[j][i], mi[j + (1 << (i - 1))][i - 1]);
        }
}
int lcp(int a, int b)
{
    a = rank[a];    b = rank[b];
    if(a > b) swap(a,b);
    a ++;
    int t = Log[b - a + 1];
    return min(mi[a][t] , mi[b - (1<<t) + 1][t]);
}
char s[MAXN];
int ans[MAXN];
int main()
{
    int cas = 0;
    Log[1] = 0;
    for(int i = 2; i < MAXN; i++) Log[i] = Log[i >> 1] + 1;
    while(scanf("%s", s) != EOF)
    {
        if(s[0] == '#') break;
        int n = strlen(s);
        for(int i = 0; i < n; i++) r[i] = s[i];
        r[n] = 0;
        da(r, sa, n + 1, 130);
        calheight(r, sa, n);
        rmqinit(n);
        int cnt = 0;
        int mx = -1;
        int l;
        for(l = 1; l < n; l++)
        {
            for(int i = 0; i + l < n; i += l)
            {
                int k = lcp(i, i + l);
                int p = k / l + 1;
                int t = l - k % l;
                t = i - t;
                if(t >= 0 && k % l)
                {
                    int tk = lcp(t, t + l);
                    if(tk / l + 1 > p) p = tk / l + 1;
                }
                if(p > mx)
                {
                    cnt = 0;
                    mx = p;
                    ans[cnt++] = l;
                }
                if(p == mx) ans[cnt++] = l;
            }
        }
        int pos = 0;
        int flag = 0;
        for(int i = 1; i <= n && !flag; i++)
        {
            for(int j = 0; j < cnt; j++)
            {
                int k = ans[j];
                if(lcp(sa[i], sa[i] + k) >= (mx - 1) * k)
                {
                    pos = sa[i];
                    l = mx * k;
                    flag = 1;
                    break;
                }
            }
        }
        printf("Case %d: ", ++cas);
        for(int i = 0; i < l; i++) printf("%c", s[pos + i]);
        printf("\n");
    }
    return 0;
}

SPOJ REPEATS

这题跟前边的一题一样。

POJ 2774

最长公共子串。

非常简单

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <algorithm>
#include <cstdlib>
#define MAXN 1000005
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
char s1[MAXN], s2[MAXN];
int main()
{
    gets(s1);
    gets(s2);

    int len = strlen(s1);
    strcat(s1, "$");
    strcat(s1, s2);
    int n = strlen(s1), m = 0;
    for(int i = 0; i < n; i++)
    {
        m = max(m, (int)s1[i]);
        r[i] = s1[i];
    }
    r[n] = 0;
    da(r, sa, n + 1, m + 1);
    calheight(r, sa, n);
    int res = 0;
    for(int i = 1; i <= n; i++)
        if((sa[i - 1] < len && sa[i] >= len) || (sa[i - 1] >= len && sa[i] < len)) res = max(res, height[i]);
    printf("%d\n", res);

    return 0;
}

URAL 1517

还是最长公共子串

POJ 3415

比较厉害的一题

要用到单调栈

我抄了爱酱的代码

求出长度不小于k的公共子串个数

计算A的某个后缀与B的某个后缀的最长公共前缀长度，如果长度L大于k，则加上L-k+1组。
将两个字符串连接起来，中间用一个没有出现的字符分开

然后通过height数组分组，某个组内的height都是大于等于k的，也就是任意两个后缀的最长公共前缀都至少为k。
扫描一遍，遇到一个B的后缀就与之前的A后缀进行统计，求出所有的满足的组数。但是这样的做法便是n^2的。
可以发现两个后缀的最长公共前缀为这一段的height值的最小值。
可以通过一个单调栈来维护一下，当前要入栈元素如果小于栈底元素，说明之后加入的B后缀与栈底的最长公共前缀是小于等于入栈的。这样就保证了单调栈内的height值是绝对递增的，逐渐合并
然后扫描两遍即可

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <algorithm>
#include <cstdlib>
#define MAXN 200005
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
char s1[MAXN], s2[MAXN];
int k;
int st[MAXN][2];
int main()
{
    while(scanf("%d", &k) != EOF && k)
    {
        getchar();
        gets(s1);
        gets(s2);
        int len = strlen(s1);
        strcat(s1, "$");
        strcat(s1, s2);
        int n = strlen(s1), m = 0;
        for(int i = 0; i < n; i++)
        {
            m = max(m, (int)s1[i]);
            r[i] = s1[i];
        }
        r[n] = 0;
        da(r, sa, n + 1, m + 1);
        calheight(r, sa, n);
        int top = 0;
        long long tot = 0;
        long long sum = 0;
        for(int i = 2; i <= n; i++)
        {
            if(height[i] < k) top = tot = 0;
            else
            {
                int cnt = 0;
                if(sa[i - 1] < len) cnt++, tot += height[i] - k + 1;
                while(top && height[i] <= st[top][0])
                {
                    tot -= st[top][1] * (st[top][0] - height[i]);
                    cnt += st[top][1];
                    top--;
                }
                st[++top][0] = height[i];
                st[top][1] = cnt;
                if(sa[i] > len) sum += tot;
            }
        }
        top = tot = 0;
        for(int i = 2; i <= n; i++)
        {
            if(height[i] < k) top = tot = 0;
            else
            {
                int cnt = 0;
                if(sa[i - 1] > len) cnt++, tot += height[i] - k + 1;
                while(top && height[i] <= st[top][0])
                {
                    tot -= st[top][1] * (st[top][0] - height[i]);
                    cnt += st[top][1];
                    top--;
                }
                st[++top][0] = height[i];
                st[top][1] = cnt;
                if(sa[i] < len) sum += tot;
            }
        }
        printf("%I64d\n", sum);
    }
    return 0;
}

POJ 3294

求多个字符串的最长公共子串。

如有多个，按字典序输出

还是用若干个互不相同且没出现的字符将这些字符串给衔接起来

然后二分进行分组。

看组内是不是所有字符串都有后缀在里面

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <algorithm>
#include <cstdlib>
#include <cmath>
#include <map>
#include <sstream>
#include <queue>
#include <vector>
#define MAXN 211111
#define MAXM 211111
#define INF 500000001
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
int m;
char s[MAXN];
char ts[1111];
int id[MAXN], n;
int vis[1111];
bool check(int mid)
{
    int flag = 0, cnt = 0;
    memset(vis, -1, sizeof(vis));
    for(int i = 1; i <= n; i++)
    {
        if(id[sa[i]] == -1)
        {
            flag++;
            cnt = 1;
            continue;
        }
        if(height[i] < mid)
        {
            cnt = 1;
            vis[id[sa[i]]] = ++flag;
        }
        else
        {
            if(vis[id[sa[i]]] != flag) cnt++;
            vis[id[sa[i]]] = flag;
        }
        if(cnt > m / 2) return true;
    }
    return false;
}
void getans(int x)
{
    int flag = 0, cnt = 0, bg, ok = 0;
    memset(vis, -1, sizeof(vis));
    for(int i = 1; i <= n; i++)
    {
        if(id[sa[i]] == -1)
        {
            flag++;
            cnt = 1;
            continue;
        }
        if(height[i] < x)
        {
            cnt = 1;
            vis[id[sa[i]]] = ++flag;
            ok = 0;
            bg = sa[i];
        }
        else
        {
            if(vis[id[sa[i]]] != flag) cnt++;
            vis[id[sa[i]]] = flag;
        }
        if(cnt > m / 2 && !ok)
        {
            ok = 1;
            for(int j = 0; j < x; j++) putchar(s[bg + j]);
            puts("");
        }
    }
}
int main()
{
    int cas = 0;
    while(scanf("%d", &m) != EOF && m)
    {
        if(cas++) puts("");
        scanf("%s", s);
        if(m == 1)
        {
            puts(s);
            continue;
        }
        int now = 130;
        for(int i = 1; i < m; i++)
        {
            scanf("%s", ts);
            strcat(s, "$");
            strcat(s, ts);
        }
        n = strlen(s);
        int cnt = 1;
        for(int i = 0; i < n; i++)
        {
            id[i] = cnt;
            r[i] = s[i];
            if(s[i] == '$') id[i] = -1, cnt++, r[i] = 130 + cnt;
        }
        r[n] = 0;

        da(r, sa, n + 1, 250);
        calheight(r, sa, n);
        int low = 0, high = n;
        int ans = 0;
        while(low <= high)
        {
            int mid = (low + high) >> 1;
            if(check(mid))
            {
                ans = max(ans, mid);
                low = mid + 1;
            }
            else high = mid - 1;
        }
        if(ans)
            getans(ans);
        else puts("?");
    }
    return 0;
}

SPOJ PHRASES

求最长公共子串。

且该子串要在每个字符串中不重叠的出现至少两次

二分分组。

然后看每组内

记录一个位置的最大值，一个位置的最小值

观察差值即可

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <algorithm>
#include <cstdlib>
#include <cmath>
#include <map>
#include <sstream>
#include <queue>
#include <vector>
#define MAXN 111111
#define MAXM 211111
#define INF 500000001
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
int n, k;
char s[11111];
int id[MAXN];
int mi[111], mx[111];
bool check(int mid)
{
    for(int i = 1; i <= k; i++) mi[i] = INF, mx[i] = 0;
    for(int i = 1; i <= n; i++)
    {
        int t = id[sa[i]];
        if(height[i] < mid)
        {
            for(int j = 1; j <= k; j++) mi[j] = INF, mx[j] = 0;
            mx[t] = sa[i];
            mi[t] = sa[i];
        }
        else
        {
            int flag = 1;
            mx[t] = max(mx[t], sa[i]);
            mi[t] = min(mi[t], sa[i]);
            for(int j = 1; j <= k; j++)
                if(mx[j] - mi[j] < mid)
                    flag = 0;
            if(flag) return true;
        }
    }
    return false;
}
int main()
{
    int T;
    scanf("%d", &T);
    while(T--)
    {
        scanf("%d", &k);
        n = 0;
        int op = 130;
        memset(id, 0, sizeof(id));
        for(int i = 1; i <= k; i++)
        {
            scanf("%s", s);
            int len = strlen(s);
            for(int j = 0; j < len; j++)
            {
                r[n] = s[j];
                id[n++] = i;
            }
            r[n++] = op++;
        }
        r[--n] = 0;
        da(r, sa, n + 1, 150);
        calheight(r, sa, n);
        int low = 0, high = 10000, ans = 0;
        while(low <= high)
        {
            int mid = (low + high) >> 1;
            if(check(mid))
            {
                ans = max(ans, mid);
                low = mid + 1;
            }
            else high = mid - 1;
        }
        printf("%d\n", ans);
    }
    return 0;
}

POJ 3581

比较神的一题

给出一个序列

然后把这个序列拆成三部分

将每部分翻转后拼起来。

使得字典序最小

首先。我们将整个序列先翻转

然后找到排名第一的后缀，并且位置要>=2，因为不能使其他两个部分没数可用了，这个就是第一部分了

第二部分怎么搞呢。

将剩余的序列，复制一遍，添加到这个序列后面。

求一遍后缀数组。

最小的，并且位置>=1，不超过现在序列一般位置的即为第二部分的开始位置。

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <algorithm>
#include <cstdlib>
#include <cmath>
#include <map>
#include <sstream>
#include <queue>
#include <vector>
#define MAXN 211111
#define MAXM 211111
#define INF 500000001
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}

int a[MAXN], b[MAXN];
int n, m;

int main()
{
    scanf("%d", &n);
    for(int i = 0; i < n; i++)
    {
        scanf("%d", &a[n - 1 - i]);
        b[n - 1 - i] = a[n - 1 - i];
    }
    sort(b, b + n);
    int cnt = unique(b, b + n) - b;
    for(int i = 0; i < n; i++)
    {
        int pos = lower_bound(b, b + cnt, a[i]) - b + 1;
        r[i] = pos;
    }
    r[n] = 0;
    m = n + 1;
    da(r, sa, n + 1, m);
    int pos;
    for(int i = 1; i <= n; i++)
        if(sa[i] >= 2)
        {
            pos = sa[i];
            break;
        }
    for(int i = pos; i < n; i++) printf("%d\n", b[r[i] - 1]);
    n = pos;
    for(int i = 0; i < pos; i++) r[n + i] = r[i];
    n = n + n;
    r[n] = 0;
    da(r, sa, n + 1, m);
    for(int i = 1; i <= n; i++)
        if(sa[i] >= 1 && sa[i] < n / 2)
        {
            pos = sa[i];
            break;
        }
    for(int j = 0; j < n / 2; j++) printf("%d\n", b[r[j + pos] - 1]);
    return 0;
}

POJ 3623

将字符串翻转后连接到原来的字符串后

中间用个没出现的字符隔开

然后我们拿的时候就要看他们的rank了

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <algorithm>
#include <cstdlib>
#include <cmath>
#include <map>
#include <sstream>
#include <queue>
#include <vector>
#define MAXN 111111
#define MAXM 211111
#define INF 500000001
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
char s[MAXN];
int n, m;
int main()
{
    scanf("%d", &n);
    m = n;
    for(int i = 0; i < n; i++) scanf("%s", &s[i]);
    for(int i = 0; i < n; i++) r[i] = s[i];
    r[n] = '$';
    for(int i = n - 1; i >= 0; i--) r[++n] = s[i];
    r[++n] = 0;
    da(r, sa, n + 1, 130);
    calheight(r, sa, n);
    int low = 0, high = m - 1, cnt = 0;
    for(int i = 0; i < m; i++)
    {
        int ra = rank[low], rb = rank[2 * m - high];
        if(ra < rb)
        {
            putchar(s[low]);
            low++;
        }
        else
        {
            putchar(s[high]);
            high--;
        }
        cnt++;
        if(cnt % 80 == 0) putchar('\n');
    }

    return 0;
}

POJ 3450

所有字符串的最长公共子串。

二分分组搞一下即可

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <algorithm>
#include <cstdlib>
#include <cmath>
#include <map>
#include <sstream>
#include <queue>
#include <vector>
#define MAXN 888888
#define MAXM 211111
#define INF 500000001
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
char s[2222];
int m, n;
int id[MAXN];
int vis[4005];
char res[2222];
bool check(int mid)
{
    int flag = 0, tot = 1, bg;
    memset(vis, -1, sizeof(vis));
    for(int i = 1; i <= n; i++)
    {
        int t = id[sa[i]];
        if(t == 0)
        {
            flag++;
            tot = 1;
            continue;
        }
        if(height[i] < mid)
        {
            vis[t] = ++flag;
            tot = 1;
            bg = sa[i];
        }
        else
        {
            if(vis[t] != flag) tot++;
            vis[t] = flag;
        }
        if(tot >= m)
        {
            //printf("%d\n", tot);
            for(int j = 0; j < mid; j++) res[j] = (char)r[bg + j];
            res[mid] = '\0';
            return 1;
        }
    }
    return 0;
}
int main()
{
    while(scanf("%d", &m) != EOF && m)
    {
        int op = 130;
        n = 0;
        memset(id, 0, sizeof(id));
        for(int i = 1; i <= m; i++)
        {
            scanf("%s", s);
            int len = strlen(s);
            for(int j = 0; j < len; j++)
            {
                r[n] = s[j];
                id[n++] = i;
            }
            r[n++] = op++;
        }
        n--;
        r[n] = 0;
        da(r, sa, n + 1, op + 1);
        calheight(r, sa, n);
        int low = 0, high = n, ans = 0;
        while(low <= high)
        {
            int mid = (low + high) >> 1;
            if(check(mid))
            {
                ans = max(ans, mid);
                low = mid + 1;
            }
            else high = mid - 1;
        }
        if(ans) printf("%s\n", res);
        else puts("IDENTITY LOST");
    }
    return 0;
}

POJ 2758

http://hi.baidu.com/billdu/item/ebfe510f1b2f3fce90571807

这个讲的很明白。。

需要注意的就是

我们平常求lcp的时候默认是所求的两个位置是不一样的。

但是此题中会求相同位置的lcp

所以特殊处理一下

#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <algorithm>
#include <cstdlib>
#include <cmath>
#include <map>
#include <sstream>
#include <queue>
#include <vector>
#define MAXN 100001
#define MAXM 211111
#define INF 500000001
using namespace std;
int r[MAXN];
int wa[MAXN], wb[MAXN], wv[MAXN], tmp[MAXN];
int sa[MAXN]; //index range 1~n value range 0~n-1
int cmp(int *r, int a, int b, int l)
{
    return r[a] == r[b] && r[a + l] == r[b + l];
}
void da(int *r, int *sa, int n, int m)
{
    int i, j, p, *x = wa, *y = wb, *ws = tmp;
    for (i = 0; i < m; i++) ws[i] = 0;
    for (i = 0; i < n; i++) ws[x[i] = r[i]]++;
    for (i = 1; i < m; i++) ws[i] += ws[i - 1];
    for (i = n - 1; i >= 0; i--) sa[--ws[x[i]]] = i;
    for (j = 1, p = 1; p < n; j *= 2, m = p)
    {
        for (p = 0, i = n - j; i < n; i++) y[p++] = i;
        for (i = 0; i < n; i++)
            if (sa[i] >= j) y[p++] = sa[i] - j;
        for (i = 0; i < n; i++) wv[i] = x[y[i]];
        for (i = 0; i < m; i++) ws[i] = 0;
        for (i = 0; i < n; i++) ws[wv[i]]++;
        for (i = 1; i < m; i++) ws[i] += ws[i - 1];
        for (i = n - 1; i >= 0; i--) sa[--ws[wv[i]]] = y[i];
        for (swap(x, y), p = 1, x[sa[0]] = 0, i = 1; i < n; i++)
            x[sa[i]] = cmp(y, sa[i - 1], sa[i], j) ? p - 1 : p++;
    }
}
int rank[MAXN]; //index range 0~n-1 value range 1~n
int height[MAXN]; //index from 1   (height[1] = 0)
void calheight(int *r, int *sa, int n)
{
    int i, j, k = 0;
    for (i = 1; i <= n; ++i) rank[sa[i]] = i;
    for (i = 0; i < n; height[rank[i++]] = k)
        for (k ? k-- : 0, j = sa[rank[i] - 1]; r[i + k] == r[j + k]; ++k);
    return;
}
int Log[MAXN];
int mi[MAXN][20];
void rmqinit(int n)
{
    for(int i = 1; i <= n; i++) mi[i][0] = height[i];
    int m = Log[n];
    for(int i = 1; i <= m; i++)
        for(int j = 1; j <= n; j++)
        {
            mi[j][i] = mi[j][i - 1];
            if(j + (1 << (i - 1)) <= n) mi[j][i] = min(mi[j][i], mi[j + (1 << (i - 1))][i - 1]);
        }
}
int tlen;
int lcp(int a, int b)
{
    if(a == b) return tlen - a;
    a = rank[a];    b = rank[b];
    if(a > b) swap(a,b);
    a ++;
    int t = Log[b - a + 1];
    return min(mi[a][t] , mi[b - (1<<t) + 1][t]);
}
char s[MAXN], op[5], ch[5];
int len, m;
int nxt[MAXN], newpos[MAXN], prepos[MAXN];
void add(char c, int pos)
{
    if(pos > len) pos = len - 1;
    for(int i = len - 1; i >= pos; i--)
    {
        s[i + 1] = s[i];
        prepos[i + 1] = prepos[i];
        if(prepos[i] != -1) newpos[prepos[i]] = i + 1;
    }
    s[pos] = c;
    prepos[pos] = -1;
    s[++len] = 0;
    for(int i = pos - 1; i >= 0; i--)
    {
        if(prepos[i] == -1) break;
        nxt[prepos[i]] = pos - i;
    }
}
int query(int a, int b, int ret)
{
    int t = lcp(a, b);
    int dis = min(nxt[a], nxt[b]);
    if(t < dis) return ret + t;
    int na = newpos[a], nb = newpos[b];
    for(int i = dis; na + i <= len && nb + i <= len; i++)
    {
        if(na + i == len || nb + i == len || s[na + i] != s[nb + i]) return ret + i;
        else if(prepos[na + i] != -1 && prepos[nb + i] != -1)
            return query(prepos[na + i], prepos[nb + i], ret + i);
    }
}
int main()
{
    Log[1] = 0;
    for(int i = 2; i < MAXN; i++) Log[i] = Log[i >> 1] + 1;
    scanf("%s%d", s, &m);
    len = strlen(s);
    for(int i = 0; i < len; i++)
        r[i] = s[i];
    r[len] = 0;
    tlen = len;
    da(r, sa, len + 1, 130);
    calheight(r, sa, len);
    rmqinit(len);
    for(int i = 0; i <= len; i++) nxt[i] = INF, newpos[i] = prepos[i] = i;
    int A, B;
    s[len++] = 0;
    for(int i = 0; i < m; i++)
    {
        scanf("%s", op);
        if(op[0] == 'I')
        {
            scanf("%s%d", ch, &A);
            add(ch[0], A - 1);
            //cout << s << endl;
        }
        else
        {
            scanf("%d%d", &A, &B);
            printf("%d\n", query(A - 1, B - 1, 0));
        }
    }
    return 0;
}