Why would we call cin.clear() and cin.ignore() after reading input?

Google Code University's C++ tutorial used to have this code:

// Description: Illustrate the use of cin to get input
// and how to recover from errors.

#include <iostream>
using namespace std;

int main()
{
  int input_var = 0;
  // Enter the do while loop and stay there until either
  // a non-numeric is entered, or -1 is entered.  Note that
  // cin will accept any integer, 4, 40, 400, etc.
  do {
    cout << "Enter a number (-1 = quit): ";
    // The following line accepts input from the keyboard into
    // variable input_var.
    // cin returns false if an input operation fails, that is, if
    // something other than an int (the type of input_var) is entered.
    if (!(cin >> input_var)) {
      cout << "Please enter numbers only." << endl;
      cin.clear();
      cin.ignore(10000,'\n');
    }
    if (input_var != -1) {
      cout << "You entered " << input_var << endl;
    }
  }
  while (input_var != -1);
  cout << "All done." << endl;

  return 0;
}

What is the significance of cin.clear() and cin.ignore()? Why are the 10000 and \n parameters necessary?

Solution 1:

The cin.clear() clears the error flag on cin (so that future I/O operations will work correctly), and then cin.ignore(10000, '\n') skips to the next newline (to ignore anything else on the same line as the non-number so that it does not cause another parse failure). It will only skip up to 10000 characters, so the code is assuming the user will not put in a very long, invalid line.

Solution 2:

You enter the

if (!(cin >> input_var))

statement if an error occurs when taking the input from cin. If an error occurs then an error flag is set and future attempts to get input will fail. That's why you need

cin.clear();

to get rid of the error flag. Also, the input which failed will be sitting in what I assume is some sort of buffer. When you try to get input again, it will read the same input in the buffer and it will fail again. That's why you need

cin.ignore(10000,'\n');

It takes out 10000 characters from the buffer but stops if it encounters a newline (\n). The 10000 is just a generic large value.