#pragma GCC optimize("Ofast")
#pragma GCC optimize("unroll-loops")
#pragma GCC target("sse,sse2,sse3,ssse3,sse4,popcnt,abm,mmx,avx,tune=native")
#include <bits/stdc++.h>
using namespace std;
#define rep(i, j, k) for (int i = j; i < k; i++)
#define pb push_back
#define for_all(x) x.begin(), x.end()
constexpr int N = 1e5 + 10;
typedef long long ll;
typedef pair<int, int> pii;
typedef complex<long double> point;
template <typename S, typename T>
inline void smin(S &l, T r) {
l = l < r ? l : r;
}
int n, k;
ll dp[2][N], a[N], A[N], B[N];
int upt[N][202];
inline ll f(int l, int r) { return (A[r] - A[l - 1]) * (A[r] - A[l - 1]); }
inline double g(int x, int y) {
return double(dp[y] - dp[x] - (A[y] - A[x]) * (A[y] - A[x])) /
(2 * (A[y] - A[x]));
}
int deq[N], L, R;
bool mark[N];
int main() {
ios::sync_with_stdio(0);
cin.tie(0);
cout.tie(0);
cin >> n >> k;
rep(i, 1, n + 1) {
cin >> a[i];
A[i] = A[i - 1] + a[i];
B[i] = A[i] * A[i];
dp[0][i] = B[i];
}
rep(i, 1, k + 1) {
L = R = 0;
deq[R] = 0;
rep(j, 1, n + 1) {
if (a[j] == 0) {
dp[1][j] = dp[1][j - 1];
upt[j][i] = upt[j - 1][i];
continue;
}
while (L < R && dp[0][deq[L]] + f(deq[L] + 1, j) >=
dp[0][deq[L + 1]] + f(deq[L + 1] + 1, j))
L++;
upt[j][i] = deq[L];
dp[1][j] = dp[0][deq[L]] + f(deq[L] + 1, j);
while (L < R && g(deq[R], j) <= g(deq[R - 1], deq[R])) R--;
deq[++R] = j;
}
memcpy(dp[0], dp[1], (n + 1) * sizeof(long long));
}
cout << (B[n] - dp[0][n]) / 2 << endl;
vector<int> res;
int me = upt[n][k];
for (int i = k; i; i--) {
if (me) {
res.pb(me);
mark[me] = true;
}
me = upt[me][i - 1];
}
int ptr = 1;
while (res.size() < k) {
if (!mark[ptr]) res.pb(ptr);
ptr++;
}
for (auto e : res) cout << e << ' ';
//system("pause");
return 0;
}
