Codeforces Round #129 (Div. 1)E. Little Elephant and Strings
题意:给出n个字符串,问每个字符串有多少个子串(不同位置,相同的子串视作不同)至少出现在这n个字符串中的k个当中。
解法:这题学到了一个SAM的新技能,对于这多个串,建SAM的时候,不是把它们连在一起,建立SAM,而是先给它们建立Trie树,然后广搜这棵Trie树,对于Trie树上的V节点,在建SAM的时候,它应该接在Trie树上他的父亲节点后面,我们用TtoM[U]表示Trie树上的U节点映射到SAM上的标号。这样建立SAM的优点是,我找任何一个字符串的任何一个前缀,它匹配的的SAM上的节点的代表串必然是这个前缀。我们先记住这个东西,怎么用等会儿看。我们要求的是每个字符串的所有子串至少出现在K个字符串中,那么我们先看看所有的子串中,有哪些子串是出现在了k个字符串中,表达在SAM上就是有哪些节点被K个字符串匹配到过。我们用cnt[u]表示u节点被几个字符串匹配过。我们每次拿出一个字符串,它能给一些节点的cnt[]值贡献1,这些节点,就是这个字符串的每个前缀在sam中的节点到根的链的并集,这个用LCA求就好了。统计完cnt[]之后,看每个节点的cnt值是否大于等于k,是的话,这个节点u上就有val[u]-val[fa[u]]个子串是被k个以上字符串包含的,用add[u]表示这个值。最后,算每个字符串的答案的时候,就是这个字符串的每个前缀映射到SAM上的节点到根的链上的add之和。
代码:
#include<stdio.h> #include<algorithm> #include<string.h> #include<vector> #include<queue> #define ll __int64 using namespace std ; const int maxn = 111111 ; const int N = maxn << 1; struct Edge { int to , next ; } edge[N] ; int head[N] , tot , f[N<<1] ; void new_edge ( int a , int b ) { edge[tot].to = b ; edge[tot].next = head[a] ; head[a] = tot ++ ; } char s[maxn] , s1[maxn] ; int l[maxn] , len ; int TtoM[maxn<<1] ; struct LCA { int dp[22][N<<1] ; int to[N] , tim[N] ; int tot , n ; int MIN ( int a , int b ) { return tim[a] < tim[b] ? a : b ; } void init () { tot = 0 ; n = 0 ; } void dfs ( int u , int fa ) { tim[u] = ++ tot ; for ( int i = head[u] ; i != -1 ; i = edge[i].next ) { int v = edge[i].to ; if ( v == fa ) continue ; dfs ( v , u ) ; dp[0][++n] = u ; } dp[0][++n] = u ; to[u] = n ; } void rmq () { for ( int i = 1 ; i <= 20 ; i ++ ) { for ( int j = 1 ; j + (1<<i) - 1 <= n ; j ++ ) { dp[i][j] = MIN ( dp[i-1][j] , dp[i-1][j+(1<<i-1)] ) ; } } } int query ( int a , int b ) { a = to[a] , b = to[b] ; if ( a > b ) swap ( a , b ) ; int k = b - a + 1 ; return MIN ( dp[f[k]][a] , dp[f[k]][b-(1<<f[k])+1] ) ; } } lca ; namespace SAM { int fa[N] , val[N] , c[26][N] ; int cnt[N] ; int tot , last ; int ws[N] , wv[N] ; ll add[N] ; vector<int> vec[N] ; void init () ; void solve ( int , int ) ; inline int new_node ( int _val ) { val[++tot] = _val ; for ( int i = 0 ; i < 26 ; i ++ ) c[i][tot] = 0 ; cnt[tot] = fa[tot] = add[tot] = 0 ; vec[tot].clear () ; return tot ; } int ADD ( int k , int p ) { int i ; int np = new_node ( val[p] + 1 ) ; while ( p && !c[k][p] ) c[k][p] = np , p = fa[p] ; if ( !p ) fa[np] = 1 ; else { int q = c[k][p] ; if ( val[q] == val[p] + 1 ) fa[np] = q ; else { int nq = new_node ( val[p] + 1 ) ; for ( i = 0 ; i < 26 ; i ++ ) c[i][nq] = c[i][q] ; fa[nq] = fa[q] ; fa[q] = fa[np] = nq ; while ( p && c[k][p] == q ) c[k][p] = nq , p = fa[p] ; } } return np ; } void SORT () { for ( int i = 0 ; i < maxn ; i ++ ) wv[i] = 0 ; for ( int i = 1 ; i <= tot ; i ++ ) wv[val[i]] ++ ; for ( int i = 1 ; i < maxn ; i ++ ) wv[i] += wv[i-1] ; for ( int i = 1 ; i <= tot ; i ++ ) ws[wv[val[i]]--] = i ; } } namespace Trie { int c[26][maxn] , tot ; int new_node () { for ( int i = 0 ; i < 26 ; i ++ ) c[i][tot] = 0 ; return tot ++ ; } void init () { tot = 0 ; new_node () ; } void insert ( int n ) { for ( int i = 1 ; i <= n ; i ++ ) { int now = 0 ; for ( int j = l[i] ; j < l[i+1] ; j ++ ) { int k = s[j] - 'a' ; if ( !c[k][now] ) c[k][now] = new_node () ; now = c[k][now] ; } } } } queue<int> Q ; void SAM::init () { tot = 0 ; TtoM[0] = new_node ( 0 ) ; Q.push ( 0 ) ; #define v Trie::c[k][u] while ( !Q.empty () ) { int u = Q.front () ; Q.pop () ; for ( int k = 0 ; k < 26 ; k ++ ) if ( v ){ TtoM[v]=ADD(k,TtoM[u]) ; Q.push ( v ) ; } } } int cmp ( int a , int b ) { return lca.tim[a] < lca.tim[b] ; } int sta[maxn] ; void SAM::solve ( int n , int k ) { SORT () ; for ( int i = 2 ; i <= tot ; i ++ ) { new_edge ( fa[i] , i ) ; } lca.dfs ( 1 , 0 ) ; lca.rmq () ; for ( int i = 1 ; i <= n ; i ++ ) { int u = 0 ; int top = 0 ; for ( int j = l[i] ; j < l[i+1] ; j ++ ) { int k = s[j] - 'a' ; u = v ; sta[++top] = TtoM[u]; cnt[TtoM[u]] ++ ; } sort ( sta + 1 , sta + top + 1 , cmp ) ; for ( int j = 2 ; j <= top ; j ++ ) { int w = lca.query ( sta[j-1] , sta[j] ) ; cnt[w] -- ; } } for ( int i = tot ; i >= 1 ; i -- ) { int p = ws[i] ; cnt[fa[p]] += cnt[p] ; if ( cnt[p] >= k ) add[p] = val[p] - val[fa[p]] ; } for ( int i = 1 ; i <= tot ; i ++ ) { int u = ws[i] ; for ( int j = head[u] ; j != -1 ; j = edge[j].next ) { int to = edge[j].to ; add[to] += add[u] ; } } for ( int i = 1 ; i <= n ; i ++ ) { int u = 0 ; ll ans = 0 ; for ( int j = l[i] ; j < l[i+1] ; j ++ ) { int k = s[j] - 'a' ; u = v ; ans += add[TtoM[u]] ; } printf ( "%I64d " , ans ) ; } puts ( "" ) ; } #undef v void init () { tot = 0 ; memset ( head , -1 , sizeof ( head ) ) ; lca.init () ; Trie::init () ; } int main () { f[0] = -1 ; for ( int i = 1 ; i < maxn << 2 ; i ++ ) f[i] = f[i>>1] + 1 ; int n , k ; while ( scanf ( "%d%d" , &n , &k ) != EOF ) { init () ; len = 0 ; for ( int i = 1 ; i <= n ; i ++ ) { scanf ( "%s" , s1 ) ; int k = strlen ( s1 ) ; l[i] = len ; for ( int j = 0 ; j < k ; j ++ ) s[len++] = s1[j] ; } l[n+1] = len ; Trie::insert (n) ; SAM::init () ; SAM::solve ( n , k ) ; } return 0 ; } /* 3 2 abc bc ab 3 2 abc ac ab 2 2 abc bc 1 1 bc 2 2 ab b 4 4 abab baba aaabbbababa abababababa 2 2 abab baba 2 2 aba bab 2 2 ab ba */
Codeforces Round #129 (Div. 1)E. Little Elephant and Strings,布布扣,bubuko.com
Codeforces Round #129 (Div. 1)E. Little Elephant and Strings
原文地址:http://blog.csdn.net/no__stop/article/details/38611209