A basic example of getting input from the user and using a promise to continue program execution after the line of input is received.

$ node example.js
Enter some text and press enter: I am a great beer!
You entered: I am a great beer!

Take a look at the code:

var Promise = require('promise');

var theyPressedEnter = function(input) {
  return input.indexOf('\n') != -1;
};

var getInput = function() {
  return new Promise(function(resolve, reject) {
    var input = '';

    process.stdin.resume();
    process.stdin.setEncoding('utf8');

    process.stdin.on('data', function(chunk) {
      input += chunk;

      if (theyPressedEnter(chunk)) {
        input = input.replace('\n', '');
        process.stdin.pause();
        resolve(input);
      }
    });
  });
};

process.stdout.write("Enter some text and press enter: ");

getInput().then(function(theInput) {
  console.log("You entered: " + theInput);
});

Note that process.stdin has a end event however it is triggered with the end of stream or CTRL-D sequence. We want only a single line of input. So we check the chunk that is provided by the data event to see if it contains \n and if it does, we resolve the promise with the full input. The program execution then continues with the then function passing the input to it.

Notice in the above code we one of the parameters at the top of the function is reject yet it is not used. Well this allows us to reject the promise. Say something bad happened like we couldn't read from process.stdin. In that case, we could handle the error by rejecting the promise with an error message. The execution would then go to the second function in the then call. A really basic example of rejecting:

var Promise = require('promise');

new Promise(function(resolve, reject) {
    reject('something really bad happened');
}).then(
    function() {
        console.log('I never get called -- nobody resolved!');
    },
    function(error) {
        console.log('I got called with', error);
    }
);

Try running that (code in reject.js)...

It's complicated. Try them out for a bit and then take a look at the source for the Promise module and at the Promise/A+ specification.

So another way to implement this would be (code from no-promise.js):

var theyPressedEnter = function(input) {
  return input.indexOf('\n') != -1;
};

var getInput = function(callback) {
  var input = '';

  process.stdin.resume();
  process.stdin.setEncoding('utf8');

  process.stdin.on('data', function(chunk) {
    input += chunk;

    if (theyPressedEnter(chunk)) {
      input = input.replace('\n', '');
      process.stdin.pause();
      callback(input);
    }
  });
};

process.stdout.write("Enter some text and press enter: ");

getInput(function(theInput) {
  console.log("You entered: " + theInput);
});

The difference here is our code is tied to knowing it has a single callback. It is just a little messier. However, in the end, it accomplishes the exact same thing. It is important to note that promsises are not a magic bullet. They are just a way to have better separation between your code and potentially handle more complicated scenarious (multiple promises) with a syntax that is easier to read and understand. Sometimes, a simple callback is going to make more sense.