#pragma GCC optimize("Ofast")

#include <bits/stdc++.h>
using namespace std;

// #define int long long

char bf[11000000];
int bfp=0;

void wc(char c) {
	bf[bfp++] = c;
}

void wu(int i) {
	int x = 1;
	int c = 1;
	while (x <= i / 10) {
		x *= 10;
		c++;
	}
	while (c--) {
		int d = i / x;
		wc('0' + d);
		i -= d * x;
		i *= 10;
	}
}

alignas(64) int tk[20][1000001],th[20][1000001];

signed main(){
	ios::sync_with_stdio(false);
	cin.tie(nullptr);
	int N;
	cin>>N;
	string S;
	cin>>S;
	vector<int>c(N+1),vk(N+1),vh(N+1),w(N+1),f(1);
	for(int i=0;i<N;i++){
		if(S[i]=='M'){
			f.push_back(i);
		}
		c[i+1]=c[i]+(S[i]!='K');
		vk[i+1]=(S[i]!='H'?vk[i]+1:vk[i]-1);
		vh[i+1]=(S[i]!='K'?vh[i]-1:vh[i]+1);
		w[i+1]=w[i]+(S[i]=='M');
	}
	f.push_back(N);
	for(int i=0;i<N+1;i++){
		tk[0][i]=vk[i];
		th[0][i]=vh[i];
	}
	for(int j=1;j<20;j++){
		for(int l=0,r=1<<j-1;r<N+1;l++,r++){
			tk[j][l]=min(tk[j-1][l],tk[j-1][r]);
		}
	}
	for(int j=1;j<20;j++){
		for(int l=0,r=1<<j-1;r<N+1;l++,r++){
			th[j][l]=min(th[j-1][l],th[j-1][r]);
		}
	}
	auto qk=[&](int l,int r)->int{
		int i=__lg(r-l);
		return min(tk[i][l],tk[i][r-(1<<i)]);
	};
	auto qh=[&](int l,int r)->int{
		int i=__lg(r-l);
		return min(th[i][l],th[i][r-(1<<i)]);
	};
	for(int i=0;i<N;i++){
		int b=qk(i+1,N+1)-vk[i]>=0;
		if(b){
			int l=1,h=w[N]-w[i]+1;
			while(l<h){
				int m=(l+h)/2;
				int j=f[w[i]+m];
				int x=vk[j]-vk[i];
				if(qh(j,N+1)-vh[j]+x>=0){
					h=m;
				}
				else{
					l=m+1;
				}
			}
			wu(c[N]-c[i]-(l-1));wc(' ');
		}
		else{
			int l=i+1,h=N;
			while(l<h){
				int m=(l+h)/2;
				if(qk(i+1,m+1)-vk[i]>=0){
					l=m+1;
				}
				else{
					h=m;
				}
			}
			wu((l-1-i)/2);wc(' ');
		}
	}
	wc('\n');
	fwrite(bf,sizeof(char),bfp,stdout);
	return 0;
}