SPOJ-SUBST1 New Distinct Substrings
题目大意
- 求出一个字符串中本质不同的子串个数
- n<=5e4
分析
- 用所有的子串个数减去重复的子串个数即可
- 所有的子串个数是n*(n+1)/2,重复的个数就是(画个图就明白了)
Code
#include<bits/stdc++.h>
#define rep(i,a,b) for(int i=(a);i<=(b);i++)
#define per(i,a,b) for(int i=(a);i>=(b);i--)
using namespace std;
const int N=1e5;
int height[N],tmp[N],bin[N],sa[N],rk[N],h[N],n,T;
string s;
void init(){
memset(height,0,sizeof(height));
memset(bin,0,sizeof(bin));
memset(tmp,0,sizeof(tmp));
memset(sa,0,sizeof(sa));
memset(rk,0,sizeof(rk));
memset(h,0,sizeof(h));
cin>>s; n=s.size();
}
void SA(){
int sz=max(n,300);
rep(i, 0, n-1 ) rk[i]=s[i];
rep(i, 0, n-1 ) ++bin[rk[i]];
rep(i, 1, sz-1) bin[i]+=bin[i-1];
rep(i, 0, n-1 ) sa[--bin[rk[i]]]=i;
for(int j=1;j<=n;j*=2)
{
int p=0;
per(i, n-1, n-j ) tmp[p++]=i;
rep(i, 0, n-1 ) if(sa[i]-j>=0)tmp[p++]=sa[i]-j;
rep(i, 0, sz-1) bin[i]=0;
rep(i, 0, n-1 ) ++bin[rk[i]];
rep(i, 1, sz-1) bin[i]+=bin[i-1];
per(i, n-1, 0 ) sa[--bin[rk[tmp[i]]]]=tmp[i];
p=tmp[sa[0]]=0;
rep(i,1,n-1){
int v0=sa[i-1],v1=sa[i],v00,v01;
if (v0 + j < n) v00 = rk[v0 + j]; else v00 = -1 ;
if (v1 + j < n) v01 = rk[v1 + j]; else v01 = -1 ;
if(rk[v0]==rk[v1] && v00==v01) tmp[v1]=p;
else tmp[v1]=++p;
}
rep( i, 0, n-1 ) rk[i]=tmp[i];
}
}
int getheight() {
int k=0;
rep(i,0,n-1) rk[sa[i]]=i;
rep(i,0,n-1) {
if (rk[i]==0) continue;
if (k) --k;
int j=sa[rk[i]-1];
while (j+k<n && i+k<n && s[i+k]==s[j+k]) ++k;
h[i]=height[rk[i]]=k;
}
}
int main()
{
scanf("%d",&T);
while(T--){
init();
SA();
getheight();
long long ans=1LL*n*(n+1)/2;
long long tot=0;
rep(i,0,n-1)tot+=height[i];
cout<<ans-tot<<"\n";
}
return 0;
}