#include <bits/stdc++.h>
using namespace std;
#define ar array
#define sz(v) int(std::size(v))
using i64 = long long;
using u64 = unsigned long long;
using pii = pair<int, int>;
bool mem_st;
const int N = 500, M = 1e5, Q = 1e6;
struct edge {
int i, j, w;
bool operator<(const edge e) const {
return w < e.w;
}
};
struct arr : ar<int, N-1> {
int sz = 0;
void pb(int x) { at(sz++) = x; }
auto end() { return begin()+sz; }
};
int n, m, q;
edge e[M];
arr pf[M+1], sf[M+1];
struct {
int p[N];
void init() {
for (int i = 0; i < n; i++) p[i] = i;
}
int find(int i) {
return p[i] == i ? i : p[i] = find(p[i]);
}
bool join(int i, int j) {
i = find(i), j = find(j);
if (i == j) return false;
p[i] = j;
return true;
}
} dsu;
mt19937 rng(1234);
int ran(int l, int r) { return l+rng()%(r-l+1); }
void log_time() {
cerr << "Time: " << double(clock()) / CLOCKS_PER_SEC << '\n';
}
u64 hsh[M], hh[Q];
int xx[Q], tt[Q]; i64 ans[Q];
void solve(int qq) {
int x = xx[qq];
int t = tt[qq];
dsu.init();
int one = 0, two = 0;
#define ins(h) if (dsu.join(e[h].i, e[h].j)) ans[qq] += abs(e[h].w-x), hh[qq] ^= hsh[h]
while (one < pf[t].sz && two < sf[t].sz)
if (2*x-e[pf[t][one]].w < e[sf[t][two]].w) {ins(pf[t][one]); one++;}
else {ins(sf[t][two]); two++;}
while (one < pf[t].sz) {ins(pf[t][one]); one++;}
while (two < sf[t].sz) {ins(sf[t][two]); two++;}
#undef ins
}
void solve(int l, int r) {
if (r-l <= 1) return;
if (hh[l] == hh[r]) {
static arr w, u;
w.sz = 0;
u.sz = 0;
int x = xx[l];
int t = tt[l];
dsu.init();
int one = 0, two = 0;
#define ins(ww,h) if (dsu.join(e[h].i, e[h].j)) ww.pb(e[h].w)
while (one < pf[t].sz && two < sf[t].sz)
if (2*x-e[pf[t][one]].w < e[sf[t][two]].w) {ins(w,pf[t][one]); one++;}
else {ins(u,sf[t][two]); two++;}
while (one < pf[t].sz) {ins(w,pf[t][one]); one++;}
while (two < sf[t].sz) {ins(u,sf[t][two]); two++;}
#undef ins
{
i64 suf = 0;
for (int x : w) suf += x;
i64 pre = 0; int h = 0;
for (int i = l+1; i < r; i++) {
while (h < w.sz && w[w.sz-h-1] <= xx[i]) {
pre += w[w.sz-h-1];
suf -= w[w.sz-h-1];
h++;
}
ans[i] = i64(xx[i]) * h - pre + suf - i64(xx[i]) * (w.sz-h);
}
}
{
i64 suf = 0;
for (int x : u) suf += x;
i64 pre = 0; int h = 0;
for (int i = l+1; i < r; i++) {
while (h < u.sz && u[h] <= xx[i]) {
pre += u[h];
suf -= u[h];
h++;
}
ans[i] += i64(xx[i]) * h - pre + suf - i64(xx[i]) * (u.sz-h);
}
}
} else {
int m = (l+r)/2;
solve(m);
solve(l,m), solve(m,r);
}
}
bool mem_en;
// #define MAX
int main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
#ifdef MAX
n = N;
m = N-1;
// m = 1000;
#else
cin >> n >> m;
#endif
mt19937_64 rng_big(1234);
for (int h = 0; h < m; h++) {
hsh[h] = rng_big();
#ifdef MAX
if (h < n-1) e[h].i = h+1, e[h].j = h+2;
else e[h].i = ran(1,n), e[h].j = ran(1,n);
e[h].w = ran(1,1e9);
#else
cin >> e[h].i >> e[h].j >> e[h].w;
#endif
e[h].i--, e[h].j--;
}
sort(e, e+m);
for (int t = 1; t <= m; t++) {
auto ins = [&](int h) {
if (dsu.join(e[h].i, e[h].j))
pf[t].pb(h);
};
dsu.init();
ins(t-1);
for (int h : pf[t-1]) ins(h);
}
for (int t = m-1; t >= 0; t--) {
auto ins = [&](int h) {
if (dsu.join(e[h].i, e[h].j))
sf[t].pb(h);
};
dsu.init();
ins(t);
for (int h : sf[t+1]) ins(h);
}
#ifdef LOCAL
cerr << "AFTER BUILD" << endl;
log_time();
#endif
i64 work = 0;
#ifdef MAX
q = Q;
#else
cin >> q;
#endif
int t = 0;
for (int qq = 0; qq < q; qq++) {
#ifdef MAX
xx[qq] = (1e9/q) * (qq+1);
#else
cin >> xx[qq];
#endif
while (t < m && e[t].w <= xx[qq]) t++;
tt[qq] = t;
}
// for (int h = 0; h < q; h++) {
// solve(h);
// cout << hh[h] << endl;
// }
solve(0);
solve(q-1);
solve(0, q-1);
for (int h = 0; h < q; h++) cout << ans[h] << '\n';
#ifdef LOCAL
cerr << "Work: " << work << endl;
cerr << "Memory: " << double(&mem_en - &mem_st) / (1 << 20) << '\n';
log_time();
#endif
}
