#pragma GCC optimize("Ofast")
#pragma GCC optimize("unroll-loops")
#pragma GCC target("avx,avx2,fma")
#include <bits/stdc++.h>
using namespace std;
typedef long long ll;
typedef long double ld;
int N, K, a[100001];
ll pref[100001] = {0};
ll dp[100001][2] = {0};
int last_split[100001][202];
int cht[100001], fr = 0, bk = -1;
#define m(j) (pref[j])
#define c(j) (dp[j][0]-pref[j]*pref[N])
inline bool bad(const int &x, const int &y, const int &z)
{
return (c(z) - c(y)) * (m(z) - m(x)) <= (c(z) - c(x)) * (m(z) - m(y));
}
inline ll eval(const int &i, const ll &x)
{
return m(i) * x + c(i);
}
int main()
{
cin.tie(0)->sync_with_stdio(0);
cin >> N >> K;
for (int i = 1; i <= N; i++)
{
cin >> a[i];
pref[i] = pref[i-1] + a[i];
}
for (int k = 1; k <= K + 1; k++)
{
fr = 0, bk = -1;
for (int i = k; i <= N; i++)
{
while (fr < bk && bad(cht[bk-1], cht[bk], i-1))
bk--;
cht[++bk] = i-1;
while (fr < bk && eval(cht[fr], pref[i]) <= eval(cht[fr+1], pref[i]))
fr++;
dp[i][1] = eval(cht[fr], pref[i]) + pref[i] * (pref[N] - pref[i]);
last_split[i][k] = cht[fr];
}
for (int i = 1; i <= N; i++)
dp[i][0] = dp[i][1];
}
cout << dp[N][0] << '\n';
int i = N;
do
{
i = last_split[i][K+1];
cout << i << ' ';
} while (--K);
cout << '\n';
return 0;
}
