zoukankan      html  css  js  c++  java
  • hdu 3553 Just a String (后缀数组)

    hdu 3553 Just a String (后缀数组)

    题意:很简单,问一个字符串的第k大的子串是谁。

    解题思路:后缀数组。先预处理一遍,把能算的都算出来。将后缀按sa排序,假如我们知道答案在那个区间范围内了(假设为[l,r]),那么我们算下这个区间内的lcp的最小值(设最小值的位置为mid,大小为x),如果x*(r-l+1)>=k,那么,答案就是这个区间的lcp的最小值的某一部分(具体是哪一部分,画个图稍微算下就出来了)。如果x * ( r - l + 1 ) < k 那么我们分两种情况考虑,如果[l,mid]区间范围内的字符串总数大于等于k,那么把区间范围缩小到[l,mid],否则范围缩小到[mid+1,r]。一点点的逼近答案就可以了。

    #include<stdio.h>
    #include<string.h>
    #include<algorithm>
    #define ll __int64
    using namespace std ;
    const int maxn = 111111 ;
    
    int min ( int a , int b ) { return a < b ? a : b ; }
    int f[maxn] ;
    int dp[22][maxn] ;
    ll sum[maxn] ;
    
    char s1[maxn] ;
    int s[maxn] ;
    
    struct Suf{
        int wa[maxn] , wb[maxn] , ws[maxn] , wv[maxn] ;
        int rank[maxn] , hei[maxn] , sa[maxn] ;
    
        int cmp ( int *r , int i , int j , int l ){ return r[i] == r[j] && r[i+l] == r[j+l] ; }
    
        void da ( int *r , int n , int m ){
                int *x = wa , *y = wb , *t ;
            int i , j , k , p ;
            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 < m ; i ++ ) ws[i] = 0 ;
                for ( i = 0 ; i < n ; i ++ ) ws[x[i]] ++ ;
                for ( i = 1 ; i < m ; i ++ ) ws[i] += ws[i-1] ;
                for ( i = n - 1 ; i >= 0 ; i -- ) sa[--ws[x[y[i]]]] = y[i] ;
                for ( t = x , x = y , y = t ,x[sa[0]] = 0 , p = 1 , i = 1 ; i < n ; i ++ )
                    x[sa[i]] = cmp ( y , sa[i-1] , sa[i] , j ) ? p - 1 : p ++ ;
            }
            k = 0 ;
            for ( i = 1 ; i < n ; i ++ ) rank[sa[i]] = i ;
            for ( i = 0 ; i < n - 1 ; hei[rank[i++]] = k )
                for ( k ? k -- : 0 , j = sa[rank[i]-1] ; r[i+k] == r[j+k] ; k ++ ) ;
        }
    
    	int min_hei ( int x , int y ) {
    		return ( hei[x] < hei[y] ? x : y ) ;
    	}
    
        void rmq ( int n ) {
            int i , j ;
            for ( i = 1 ; i <= n ; i ++ ) dp[0][i] = i ;
            for ( i = 1 ; i <= 20 ; i ++ )
                for ( j = 1 ; j + ( 1 << i ) - 1 <= n ; j ++ )
                    dp[i][j] = min_hei ( dp[i-1][j] , dp[i-1][j+(1<<(i-1))] ) ;
        }
    
        int query ( int l , int r ) {
            if ( l > r ) swap ( l , r ) ;
            l ++ ;//要从height[l+1]到height[r]之间求最小值
            if ( l == r ) return dp[0][l] ;
            int k = r - l + 1 ;
            return min_hei ( dp[f[k]][l] , dp[f[k]][r-(1<<f[k])+1] ) ;
        }
    
        void solve ( int n , ll k ) {
    		rmq ( n ) ;
    		int l = 1 , r = n  , i;
    		sum[0] = 0 ;
    		for ( i = 1 ; i <= n ; i ++ )
    			sum[i] = sum[i-1] + n - sa[i] ;
    		int h = 0 ;
    		int pos = 0 , len ;
    		while ( l < r ) {
    			int mid = query ( l , r ) - 1 ;
    //			printf ( "l = %d , r = %d mid = %d , k = %I64d , fuck = %d
    " , l , r , mid , k , ( hei[mid] - h ) * ( r - l + 1 ) ) ;
    			if ( k <= (ll) ( hei[mid+1] - h ) * ( r - l + 1 ) ) {
    				pos = l ;
    				len = h + k / ( r - l + 1 ) + ( k % ( r - l + 1 ) != 0 ) ;
    //				printf ( "pos = %d , l = %d
    " , pos , len ) ;
    				break ;
    			}
    			k -= (ll) (hei[mid+1] - h ) * ( r - l + 1 ) ;
    			if ( k <= sum[mid] - sum[l-1] - (ll) hei[mid+1] * ( mid - l + 1 ) ) {
    				r = mid ;
    			}
    			else {
    				k -= sum[mid] - sum[l-1] - (ll) hei[mid+1] * ( mid - l + 1 ) ;
    				l = mid + 1 ;
    			}
    			h = hei[mid+1] ;
    		}
    		if ( !pos ) pos = l , len = h + k ;
    		for ( i = 0 ; i < len ; i ++ )
    			printf ( "%c" , s[sa[pos]+i] ) ;
    		puts ( "" ) ;
        }
    
    } arr ;
    
    int main () {
        int cas , i , j , ca = 0 ;
        ll m ;
        j = 0 ;
        for ( i = 1 ; i < maxn - 1111 ; i ++ ) {
            if ( i > 1 << j + 1 ) j ++ ;
            f[i] = j ;
        }
        scanf ( "%d" , &cas ) ;
        while ( cas -- ) {
            scanf ( "%s" , s1 ) ;
            scanf ( "%I64d" , &m ) ;
            int len = strlen ( s1 ) ;
            for ( i = 0 ; i < len ; i++ ) s[i] = s1[i] ;
            s[len] = 0 ;
            arr.da ( s , len + 1 , 411 ) ;
            printf ( "Case %d: " , ++ ca ) ;
            arr.solve ( len , m ) ;
        }
    }
    /*
    10000
    ddff 9
    */


  • 相关阅读:
    iOS开发那些事-iOS6苹果地图有用开发
    Partition List -- LeetCode
    C++11:强类型枚举(enum)
    数据绑定(八)使用Binding的RelativeSource
    对于流程优化的处理器架构
    Linux/UNIX流程关系
    《炉石传说》建筑设计欣赏(6):卡&amp;在执行数据时,组织能力
    采用truelicense进行Java规划license控制 扩展可以验证后,license 开始结束日期,验证绑定一个给定的mac住址
    [Linux]scp 命令远程复制
    从头开始编写项目Makefile(八):型号规则
  • 原文地址:https://www.cnblogs.com/riskyer/p/3354336.html
Copyright © 2011-2022 走看看