// Oldham, Jeffrey D.
// 2000Apr18
// CS1321

// Program to Find Duplicate URLs

// Given a list of URLs, the program finds URLS with duplicate
// contents.  Every URL with exactly the same contents as another URL
// is mentioned as a duplicate.  However, we do not guarantee listing
// all pairs of duplicates.  For example, if three URLs have
// exactly the same contents, we will list two, not three pairs.

// Command-Line Argument
// 1. name of a file containing the names of the URLs

#include <iostream>
#include <fstream>
#include <string>
#include <stdlib.h>		// has EXIT_SUCCESS
#include "hashTable.h"
#include "URL.h"

int main(int argc, char *argv[])
{
  if (argc != 2) {
    cerr << argv[0] << ": file-of-URLs\n";
    throw "incorrect number of command-line arguments";
  }

  ifstream URLs(argv[1]);
  if (!URLs) {
    cerr << argv[0] << ": " << argv[1] << " not readable\n";
    throw "missing URL file";
  }

  hashTable URLTable;
  string currentURL;
  while (URLs >> currentURL) {
    URL u(currentURL);
    hashTable::insertPair insertAnswer = URLTable.insert(u);
    if (insertAnswer.first == false) // Found a duplicate.
      cout << currentURL << " matches " << insertAnswer.second << endl;
  }

  URLs.close();
  return EXIT_SUCCESS;
}