#pragma GCC optimize("O3,unroll-loops")
#pragma GCC target("avx,avx2,fma")
#include <bits/stdc++.h>
#include <ext/pb_ds/assoc_container.hpp>
#include <ext/pb_ds/tree_policy.hpp>
/* DEFINES */
#define S second
#define F first
#define ll long long
#define ull unsigned long long
#define ld long double
#define npos ULLONG_MAX
#define INF LLONG_MAX
#define vv(a) vector<a>
#define pp(a, b) pair<a, b>
#define pq(a) priority_queue<a>
#define qq(a) queue<a>
#define ss(a) set<a>
#define mm(a, b) map<a, b>
#define ump(a, b) unordered_map<a, b>
#define ANDROID \
ios_base::sync_with_stdio(0); \
cin.tie(0); \
cout.tie(0);
#define elif else if
#define endl "\n"
#define allc(a) begin(a), end(a)
#define all(a) a, a + sizeof(a) / sizeof(a[0])
#define pb push_back
#define logi(a) __lg(a)
#define sqrt(a) sqrtl(a)
#define mpr make_pair
#define ins insert
using namespace std;
using namespace __gnu_pbds;
using namespace __cxx11;
typedef char chr;
typedef basic_string<chr> str;
template <typename T, typename key = less<T>>
using ordered_set = tree<T, null_type, key, rb_tree_tag, tree_order_statistics_node_update>;
mt19937 rng(chrono::steady_clock::now().time_since_epoch().count());
void solve()
{
ll n, d, m;
cin >> n >> d >> m;
ll cnt[n], a[n];
vv(ll) v[n];
memset(cnt, 0, sizeof(cnt));
for (ll i = 0, x; i < m; i++)
cin >> x, cnt[--x]++, v[x].pb(i + 1);
memcpy(a, cnt, sizeof(a));
ll l = 0, r = m, ans = m;
while (l <= r)
{
ll mid = (l + r) >> 1, ptr = 0;
bool ok = true;
memcpy(cnt, a, sizeof(a));
for (ll i = 0; i < n; i++)
{
for (ll j = 0; j < mid and ptr < n; j++)
{
while (ptr < n and !cnt[ptr])
ptr++;
if (ptr < n)
cnt[ptr]--;
}
if (i >= d and cnt[i - d])
{
ok = false;
break;
}
}
if (ok)
r = mid - 1, ans = mid;
else
l = mid + 1;
}
ll ptr = 0;
cout << ans << endl;
memcpy(cnt, a, sizeof(a));
for (ll i = 0; i < n; i++)
{
for (ll j = 0; j < ans and ptr < n; j++)
{
while (ptr < n and !cnt[ptr])
ptr++;
if (ptr < n)
cnt[ptr]--, cout << v[ptr].back() << " ", v[ptr].pop_back();
}
cout << "0\n";
}
}
/*
*/
int main()
{
ANDROID
// precomp();
ll t = 1;
// cin >> t;
for (ll i = 1; i <= t; i++)
// cout << "Case #" << i << ": ",
solve();
cerr << "\nTime elapsed : " << clock() * 1000.0 / CLOCKS_PER_SEC << " ms\n";
}
