#pragma GCC optimize(2)
#pragma GCC optimize(3)
#pragma GCC optimize("Ofast","inline")
#include <bits/stdc++.h>
using namespace std;
#define int long long
const int M=5e5+5,N=5005,inf=1e9;
struct node{
    int to,sm,val;
}p[M];
vector<int>e[M];
int idx=1,st,ed,d[M],lst[M],ls[M],res,n,m,k,ans,w[M],g[M],cnt,nxt[N][N],f[M];
priority_queue<pair<int,int>,vector<pair<int,int> >,greater<pair<int,int> > >q;
bitset<M>mk;
void add(int x,int y,int z,int val){
    e[x].push_back(++idx);
    p[idx]=(node){y,z,val};
    e[y].push_back(++idx);
    p[idx]=(node){x,0,-val};
}
bool SPFA(){
    mk.reset();
    for(int i=1;i<=ed;i++)d[i]=1e9;
    d[st]=0;ls[st]=inf;
    q.push({0,st});
    while(q.size()){
        int x=q.top().second;
        q.pop();
        if(mk[x])continue;
        mk[x]=1;
        for(auto v:e[x]){
            int c=p[v].to;
            if(!p[v].sm)continue;
            if(d[c]>d[x]+p[v].val+f[x]-f[c]){
                d[c]=d[x]+p[v].val+f[x]-f[c];
                lst[c]=v;
                ls[c]=min(ls[x],p[v].sm);
                q.push({d[c],c});
            }
        }
    }
    if(d[ed]>=1e9)return 0;
    return 1;
}
void solve(){
    int x=ed,t=ls[ed];
    res+=(d[ed]+f[ed]-f[st])*t;ans+=t;
    while(x!=st){
        p[lst[x]].sm-=t;p[lst[x]^1].sm+=t;
        x=p[lst[x]^1].to;
    }
    for(int i=1;i<=ed;i++)f[i]+=d[i];
}
void Solve(){
    while(SPFA())solve();
}
signed main(){
    cin>>n>>k>>m;
    for(int i=1;i<=n;i++)cin>>w[i];
    for(int i=1;i<=m;i++)cin>>g[i];
    for(int i=1;i<=n;i++)nxt[m][i]=m+1;
    for(int i=m-1;i>=0;i--){
        for(int j=1;j<=n;j++)nxt[i][j]=nxt[i+1][j];
        nxt[i][g[i+1]]=i+1;
    }
    for(int i=1;i<=m;i++){
        add(i+m,i,1,w[g[i]]);
        if(i!=m)add(i+1,nxt[i][g[i]]+m,1,0);
    }
    ed=2*m+1;st=2*m+2;
    for(int i=1;i<=n;i++)add(st,nxt[0][i]+m,1,0);
    for(int i=2;i<=m;i++)add(i-1,i,k,0);
    add(m,ed,k,0);
    Solve();
    cout<<res;
}