NTT多项式乘法模板（HUD 1402）

FFT虽然能快速处理卷积，但是它也有很大的弊端。精度问题有时会导致一些错误。而且，有许多题目涉及了取模，比如 998244353，复数域下的 DFT精度更是暴露无遗。于是考虑是否有模意义下的这种算法。于是，便出现了快速数论变换（Fast Number−Theoretic Transform, FNT）。

NTT和FFT基本上一样，只不过用原根代替了复数根。

参考博客：https://www.cnblogs.com/zhouzhendong/p/8831887.html

NTT大数乘法代码：

#pragma GCC optimize(2)
#pragma GCC optimize(3)
#pragma GCC optimize(4)
#pragma comment(linker, "/STACK:102400000,102400000")
#include<unordered_map>
#include<unordered_set>
#include<algorithm>
#include<iostream>
#include<fstream>
#include<complex>
#include<cstdlib>
#include<cstring>
#include<cassert>
#include<iomanip>
#include<string>
#include<cstdio>
#include<bitset>
#include<vector>
#include<cctype>
#include<cmath>
#include<ctime>
#include<stack>
#include<queue>
#include<deque>
#include<list>
#include<set>
#include<map>
using namespace std;
#define pt(a) cout<<a<<endl
#define debug test
#define mst(ss,b) memset((ss),(b),sizeof(ss))
#define rep(i,a,n) for (int i=a;i<=n;i++)
#define per(i,a,n) for (int i=n-1;i>=a;i--)
#define all(x) (x).begin(),(x).end()
#define fi first
#define se second
#define SZ(x) ((int)(x).size())
#define ll long long
#define ull unsigned long long
#define pb push_back
#define mp make_pair
#define inf 0x3f3f3f3f
#define eps 1e-10
#define PI acos(-1.0)
typedef pair<int,int> PII;
const ll mod = 998244353;
const int N = 1e6+10;

ll gcd(ll p,ll q){return q==0?p:gcd(q,p%q);}
ll qp(ll a,ll b) {ll res=1;a%=mod; assert(b>=0); for(;b;b>>=1){if(b&1)res=res*a%mod;a=a*a%mod;}return res;}
int to[4][2]={{-1,0},{1,0},{0,-1},{0,1}};

ll A,B,a[N],b[N],R[N],g[N],n,L;
char s1[N],s2[N];

void NTT(ll a[N],int n){
    for (int i=0;i<n;i++) if (i<R[i]) swap(a[i],a[R[i]]);
    for (int d=1,t=n>>1;d<n;d<<=1,t>>=1)
        for (int i=0;i<n;i+=(d<<1))
            for (int j=0;j<d;j++){
                ll tmp=g[t*j]*a[i+j+d]%mod;
                a[i+j+d]=(a[i+j]-tmp+mod)%mod;
                a[i+j]=(a[i+j]+tmp)%mod;
            }
}

int main() {
    ios::sync_with_stdio(0),cin.tie(0),cout.tie(0);
    while(cin>>s1>>s2) {
        mst(a,0),mst(b,0);
        A=strlen(s1),B=strlen(s2);
        for (int i=0;i<A;i++) a[A-i-1]=s1[i]-'0';
        for (int i=0;i<B;i++) b[B-i-1]=s2[i]-'0';
        for (n=1,L=0;n<A+B;n<<=1,L++);
        for (int i=0;i<n;i++) R[i]=(R[i>>1]>>1)|((i&1)<<(L-1));
        g[0]=1,g[1]=qp(3,(mod-1)/n);
        for (int i=2;i<n;i++) g[i]=g[i-1]*g[1]%mod;
        NTT(a,n),NTT(b,n);
        for (int i=0;i<n;i++) a[i]=a[i]*b[i]%mod;
        g[0]=1,g[1]=qp(g[1],mod-2);
        for (int i=2;i<n;i++) g[i]=g[i-1]*g[1]%mod;
        NTT(a,n);
        ll Inv=qp(n,mod-2);
        for (int i=0;i<n;i++) a[i]=a[i]*Inv%mod;
        for (int i=0;i<n-1;i++) a[i+1]+=a[i]/10,a[i]%=10;
        int d;
        for (d=n-1;d&&!a[d];d--);
        for (int i=d;i>=0;i--) cout<<a[i];
        cout<<endl;
    }
    return 0;
}