#include <bits/stdc++.h>
using namespace std;
using ll = long long;
using vi = vector<int>;
using vll = vector<ll>;
using vvi = vector<vi>;
using vvll = vector<vll>;
using pii = pair<int,int>;
using pll = pair<ll,ll>;
using ui = unsigned int;
using ull = unsigned long long;
using pui = pair<ui,ui>;
using pull = pair<ull,ull>;
template<typename T>
using vc = std::vector<T>;
#define YES cout << "YES\n"
#define NO cout << "NO\n"
#define yes cout << "yes\n"
#define no cout << "no\n"
#define Yes cout << "Yes\n"
#define No cout << "No\n"
template<typename T> void sc(T &x) { cin >> x; }
template<typename T, typename U> void sc(pair<T,U> &p) { sc(p.first), sc(p.second); }
template<typename T> void sc(vector<T> &v) { for (auto &x : v) sc(x); }
template<typename... A> void sc(A&... a) { (sc(a), ...); }
template<typename T> void _pt(const T &x){cout << x;}
template<typename T, typename U> void _pt(const pair<T,U> &p){_pt(p.first); cout << ' '; _pt(p.second);}
template<typename T> void _pt(const vector<T> &v){bool f=1; for(auto &x:v){if(!f) cout<<' '; _pt(x); f=0;}}
template<typename... A> void pt(const A&... a){(_pt(a), ...);}
template<typename T, typename U> bool umin(T &a, const U &b) { return b < a ? (a = b, true) : false; }
template<typename T, typename U> bool umax(T &a, const U &b) { return b > a ? (a = b, true) : false; }
#define GET_MACRO(_1,_2,_3,_4,NAME,...) NAME
#define FOR1(n) for (int i=0; i<(n); ++i)
#define FOR2(i,n) for (int i=0; i<(n); ++i)
#define FOR3(i,l,r) for (int i=(l); i<(r); ++i)
#define FOR4(i,l,r,k) for (int i=(l); i<(r); i+=(k))
#define FOR(...) GET_MACRO(__VA_ARGS__, FOR4, FOR3, FOR2, FOR1)(__VA_ARGS__)
#define ROF1(n) for (int i=(n)-1; i>=0; --i)
#define ROF2(i,n) for (int i=(n)-1; i>=0; --i)
#define ROF3(i,l,r) for (int i=(r)-1; i>=(l); --i)
#define ROF4(i,l,r,k) for (int i=(r)-1; i>=(l); i-=(k))
#define ROF(...) GET_MACRO(__VA_ARGS__, ROF4, ROF3, ROF2, ROF1)(__VA_ARGS__)
#define all(c) (c).begin(), (c).end()
#define allr(c) (c).rbegin(), (c).rend()
#define eb emplace_back
#define mp make_pair
#define mt make_tuple
#define fi first
#define se second
#define pb push_back
#define trav(x,a) for (auto &x : (a))
#define sz(a) ((int)(a).size())
#define mem(a,v) memset(a, v, sizeof(a))
#define nl pt("\n")
#include "dungeons.h"
const int MN = 400000 + 5;
const int B = 11;
const int ML = 8;
ll pc[ML];
int n;
vi w, l, s, g;
const ll inf = LLONG_MAX;
struct LiftNode {
int to;
ll add;
ll need;
LiftNode(): to(0), add(0), need(inf) {}
LiftNode(int _to, ll _add, ll _need): to(_to), add(_add), need(_need) {}
};
// lift[level_index][binary_pow_index][node] => LiftNode
vc<vc<vc<LiftNode>>> lift;
void init(int N, vi s_, vi p_, vi w_, vi l_) {
n = N;
w = w_;
l = l_;
s = s_;
g = p_;
w.pb(n);
l.pb(n);
s.pb(0);
g.pb(0);
pc[0] = 1;
FOR(i, 1, ML) {
if (pc[i-1] > inf / B) pc[i] = inf / 2;
else pc[i] = pc[i-1] * B;
}
int logn = (n > 1) ? __lg(n) + 2 : 2;
lift.assign(ML, vc<vc<LiftNode>>(logn, vc<LiftNode>(n + 1)));
FOR(lvl, ML) {
lift[lvl][0][n] = LiftNode(n, 0, inf);
FOR(node, n) {
if ((ll)s[node] >= pc[lvl]) {
lift[lvl][0][node] = LiftNode(l[node], (ll)g[node], (ll)s[node]);
}
else {
lift[lvl][0][node] = LiftNode(w[node], (ll)s[node], inf);
}
}
}
FOR(lvl, ML) {
FOR(j, 1, logn) {
FOR(node, n + 1) {
const LiftNode &L = lift[lvl][j-1][node];
const LiftNode &R = lift[lvl][j-1][L.to];
int new_to = R.to;
ll new_add = L.add + R.add;
ll new_need;
if (L.need == inf) {
if (R.need == inf) new_need = inf;
else {
ll adjusted = R.need - L.add;
if (adjusted <= 0) new_need = 0;
else new_need = adjusted;
}
}
else {
if (R.need == inf) new_need = L.need;
else {
ll adjusted = R.need - L.add;
if (adjusted <= 0) new_need = min(L.need, 0LL);
else new_need = min(L.need, adjusted);
}
}
lift[lvl][j][node] = LiftNode(new_to, new_add, new_need);
}
}
}
}
ll simulate(int x, int Z) {
int cur = x;
ll ans = (ll)Z;
while (cur != n) {
int level = 0;
FOR(ML) {
if (pc[i] <= ans) level = i;
else break;
}
int logn = sz(lift[level]);
ll upperBound = (level + 1 < ML) ? pc[level + 1] : inf;
while (cur != n && ans < upperBound) {
ROF(j, logn) {
const LiftNode &node = lift[level][j][cur];
if (node.need > ans || node.need == inf) {
ans += node.add;
cur = node.to;
if (cur == n) break;
}
}
if (cur == n) break;
if (ans >= (ll)s[cur]) {
ans += (ll)s[cur];
cur = w[cur];
}
else {
ans += (ll)g[cur];
cur = l[cur];
}
}
}
return ans;
}
// int32_t main() {
// ios::sync_with_stdio(false);
// cin.tie(nullptr);
// init(3, {2, 6, 9}, {3, 1, 2}, {2, 2, 3}, {1, 0, 1});
// pt(simulate(0, 1)); nl;
// pt(simulate(2, 3)); nl;
// return 0;
// }
