#pragma GCC optimize("Ofast,unroll-loops")
#pragma GCC target("avx2")
#include <bits/stdc++.h>
using namespace std;
const int MAX_N = 1e5 + 5;
const int64_t INF = 2e18;
vector<pair<int, int>> G[MAX_N];
int64_t _min(int64_t a, int64_t b) { return a < b ? a : b; }
int main() {
cin.tie(nullptr)->sync_with_stdio(false);
int N, K, M;
cin >> N >> K >> M;
vector<int> S(K);
for (int &v : S) {
cin >> v;
v--;
}
for (int i = 0; i < M; i++) {
int A, B, C;
cin >> A >> B >> C;
A--, B--;
G[A].push_back({B, C});
G[B].push_back({A, C});
}
vector<vector<int64_t>> DP(N, vector<int64_t>(1 << K, INF));
for (int i = 0; i < K; i++) {
DP[S[i]][1 << i] = 0;
}
for (int mask = 1; mask < (1 << K); mask++) {
for (int i = 0; i < N; i++) {
for (int submask = mask; submask * 2 >= mask; submask = (submask - 1) & mask) {
DP[i][mask] = _min(DP[i][mask], DP[i][submask] + DP[i][mask ^ submask]);
}
}
auto cmp = [&](pair<int64_t, int> a, pair<int64_t, int> b) { return a.first > b.first; };
priority_queue<pair<int64_t, int>, vector<pair<int64_t, int>>, decltype(cmp)> q(cmp);
for (int i = 0; i < N; i++) {
if (DP[i][mask] != INF) {
q.push({DP[i][mask], i});
}
}
while (!q.empty()) {
auto [d, u] = q.top();
q.pop();
if (d == DP[u][mask]) {
for (auto [v, w] : G[u]) {
if (DP[u][mask] + w < DP[v][mask]) {
q.push({DP[v][mask] = DP[u][mask] + w, v});
}
}
}
}
}
int64_t ans = INF;
for (int i = 0; i < N; i++) {
ans = min(ans, DP[i].back());
}
cout << ans << "\n";
}
