// Helene Martin, CSE 143
// This program counts the number of words in a large file.                                               
// It demonstrates the use of a Set collection.                                              

import java.io.*;
import java.util.*;

// time taken in ms
//             smallmoby      mobydick
// ArrayList   1528           4122
// LinkedList  2785           6816
// TreeSet     374            454
// HashSet     332            450

public class WordCount {
   public static void main(String[] args) throws FileNotFoundException {
      // List<String> words = new LinkedList<String>();
      Set<String> words = new HashSet<String>();

      System.out.println("Reading file...");
      Scanner input = new Scanner(new File("smallmoby.txt"));
      // Scanner input = new Scanner(new File("mobydick.txt"));

      long start = System.currentTimeMillis();

      while (input.hasNext()) {
         String word = input.next();
         // if (!words.contains(word)) { // not needed for Set
         words.add(word);
         // }
      }

      long end = System.currentTimeMillis();
      long elapsed = end - start;

      System.out.println("The file has " + words.size() + " words.");
      System.out.println("Took " + elapsed + " ms.");

      // display contents of set. Can't use a regular for loop
      // since sets don't have indexes
      for (String word : words) {
         System.out.println(word);
      }
   }
}