This submission is migrated from previous version of oj.uz, which used different machine for grading. This submission may have different result if resubmitted.
#include "bits/stdc++.h"
#include <type_traits>
using namespace std;
#pragma GCC optimize("Ofast")
#pragma GCC target("avx,avx2,fma")
#pragma GCC optimization("unroll-loops")
// ============ Macros starts here ============
int recur_depth = 0;
#ifdef DEBUG
#define dbg(x) {++recur_depth; auto x_=x; --recur_depth; cerr<<string(recur_depth, '\t')<<"\e[91m"<<__func__<<":"<<__LINE__<<"\t"<<#x<<" = "<<x_<<"\e[39m"<<endl;}
#else
#define dbg(x)
#endif // DEBUG
template<typename Ostream, typename Cont>
typename enable_if<is_same<Ostream, ostream>::value, Ostream&>::type operator<<(Ostream& os, const Cont& v) {
    os << "[";
    for (auto& x : v) { os << x << ", "; }
    return os << "]";
}
template<typename Ostream, typename ...Ts>
Ostream& operator<<(Ostream& os, const pair<Ts...>& p) {
    return os << "{" << p.first << ", " << p.second << "}";
}
#define readFast                      \
    ios_base::sync_with_stdio(false); \
    cin.tie(0);                       \
    cout.tie(0);
#ifdef LOCAL
#define read() ifstream fin("date.in.txt")
#else
#define read() readFast
#endif // LOCAL
// ============ Macros ends here ============
#define fin cin
#define ll long long
#define sz(x) (int)(x).size()
#define all(v) v.begin(), v.end()
#define output(x) (((int)(x) && cout << "YES\n") || cout << "NO\n")
#define LSB(x) (x & (-x))
#define test cout << "WORKS\n";
const int N = 500 + 15;
const int MOD = 1e9 + 7;
int n, s;
vector<ll> v, k;
vector<vector<int>> w;
int main() {
    read();
    fin >> s >> n;
    v.resize(n);
    w.resize(s + 1);
    k.resize(n);
    for (int i = 0; i < n; ++i) {
        int weight;
        fin >> v[i] >> weight >> k[i];
        for (int j = 1; j <= k[i] && j * weight <= s; ++j) {
            w[weight].push_back(v[i]);
        }
    }
    for (int i = 1; i <= s; ++i) {
        sort(all(w[i]), greater<int>());
        int w_size = sz(w[i]);
        int left = (s + i - 1) / i;
        if (w_size > left) {
            w[i].erase(w[i].begin() + left, w[i].end());
        }
    }
    vector<int> val(s + 1, 0);
    for (int weight = 1; weight <= s; ++weight) {
        for (int i = 0; i < w[weight].size(); ++i) {
            for (int j = s - weight; j >= 0; --j) {
                val[j + weight] = max(val[j + weight], val[j] + w[weight][i]);
            }
        }
    }
    cout << *max_element(all(val));
    return 0;
} /*stuff you should look for !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
   * test the solution with the given example
   * int overflow, array bounds, matrix bounds
   * special cases (n=1?)
   * do smth instead of nothing and stay organized
   * WRITE STUFF DOWN
   * DON'T GET STUCK ON ONE APPROACH
~Benq~*/
Compilation message (stderr)
knapsack.cpp:7: warning: ignoring '#pragma GCC optimization' [-Wunknown-pragmas]
    7 | #pragma GCC optimization("unroll-loops")
      | 
knapsack.cpp: In function 'int main()':
knapsack.cpp:78:27: warning: comparison of integer expressions of different signedness: 'int' and 'std::vector<int>::size_type' {aka 'long unsigned int'} [-Wsign-compare]
   78 |         for (int i = 0; i < w[weight].size(); ++i) {
      |                         ~~^~~~~~~~~~~~~~~~~~| # | Verdict | Execution time | Memory | Grader output | 
|---|
| Fetching results... | 
| # | Verdict | Execution time | Memory | Grader output | 
|---|
| Fetching results... | 
| # | Verdict | Execution time | Memory | Grader output | 
|---|
| Fetching results... | 
| # | Verdict | Execution time | Memory | Grader output | 
|---|
| Fetching results... | 
| # | Verdict | Execution time | Memory | Grader output | 
|---|
| Fetching results... |