import java.util.Date;
import java.util.HashMap;
import java.util.Map;
import java.util.Comparator;
import java.util.Arrays;
import java.io.FileReader;
import java.io.BufferedReader;

public class WebStat {
  public static class IntVal {
    private int iv;
    public IntVal(int iv){
      this.iv = iv;
    }
    public void add(int c){
      iv += c;
    }
    public int intValue(){
      return iv;
    }
  }
  public static void incCount(HashMap<String, IntVal> hm, String key){
    if (hm.containsKey(key)){
      IntVal count = hm.get(key);
      count.add(1);
	} else {
      hm.put(key, new IntVal(1));
    }
  }
  public static void print_top20(String label, HashMap<String, IntVal> h){
    Map.Entry<String, IntVal>[] entries = (Map.Entry<String, IntVal>[])new Map.Entry[h.size()];
    entries = h.entrySet().toArray(entries);
    Arrays.sort(entries, new Comparator<Map.Entry<String, IntVal> >(){
	  public int compare(Map.Entry<String, IntVal> a, Map.Entry<String, IntVal> b){
        return -1*(a.getValue().intValue() - b.getValue().intValue());
      }
    });

    System.out.println("------------ " + label + " -------------");
    for (int i = 0; i < 20; i++){
      System.out.printf("%2d. %s (%d)\n", i, entries[i].getKey(), entries[i].getValue().intValue());
    }
    System.out.println();
  }
  public static void main(String[] args) throws Exception {
    if (args.length < 1){
      System.out.println("Usage:: [ruby] webstat.rb <inpfile>");
      System.exit(0);
    }
    BufferedReader br = new BufferedReader(new FileReader(args[0]));
    long t1 = System.currentTimeMillis();
    int nlines = 0;
    Date start_date = null, end_date = null;
    LogEntry le = null;
    HashMap<String, IntVal> hosts = new HashMap<String, IntVal>();
    HashMap<String, IntVal> urls = new HashMap<String, IntVal>();
    HashMap<String, IntVal> referrers = new HashMap<String, IntVal>();
    HashMap<String, IntVal> uastrings = new HashMap<String, IntVal>();
    String line;
    long st = System.currentTimeMillis();
    while ((line = br.readLine()) != null){
      try {
        le = new LogEntry(line);
        if (start_date == null)
          start_date = le.getDate();
        incCount(hosts, le.getHost());
        incCount(urls, le.getURL());
        incCount(referrers, le.getReferrer());
        incCount(uastrings, le.getUa());
      } catch (Exception exc){
        System.out.println("Log entry parse failed at line: " + (nlines + 1) + ", error: " + exc);
        System.out.println("LINE: " + line);
      }
      nlines += 1;
      if (nlines % 4096 == 0){
        long et = System.currentTimeMillis();
        System.out.println("processed " + nlines + " lines ... (" + (et - st)/1000.0 + " seconds)");
        st = et;
	  }
    }
    end_date = le.getDate();
    long t2 = System.currentTimeMillis();

    System.out.printf("start_date:%s, end_date:%s\n", start_date, end_date);
    System.out.printf("lines:%d, hosts:%d, urls:%d, referrers:%d, uastrings:%d\n",
      nlines, hosts.size(), urls.size(), referrers.size(), uastrings.size());
    System.out.println("Processing time : " + (t2 - t1)/1000.0 + " seconds");

    //System.err.println("Hit Enter to continue ...");
    //System.in.read();
    //System.in.skip(32);

    // Do the sorting and display of top 20
    t1 = System.currentTimeMillis();
    print_top20("Top 20 Hosts", hosts);
    print_top20("Top 20 URLs", urls);
    print_top20("Top 20 Referrers", referrers);
    print_top20("Top 20 UA Strings", uastrings);
    t2 = System.currentTimeMillis();

    System.out.println("Sort and Display time : " + (t2 - t1)/1000.0 + " seconds");

    //System.err.println("Hit Enter to continue ...");
    //System.in.read();
  }
}