Question

I have a answer to another guy question here How to count string occurrence in string?
So I was playing with algorithms here, and after benchmarking some functions I was wondering why a backwards loop was significantly slower than forward.

Benchmark test here

---

NOTE: This code below does not work as supposed to be, there are
others that work (thats not the point of this question), be aware
before Copying>Pasting it

Forward

function occurrences(string, substring) {

  var n = 0;
  var c = 0;
  var l = substring.length;

  for (var i = 0, len = string.length; i < len; i++) {

    if (string.charAt(i) == substring.charAt(c)) {
      c++;
    } else {
      c = 0;
    }

    if (c == l) {
      c = 0;
      n++;
    }
  }
  return n;
}

Backwards

function occurrences(string, substring) {

  var n = 0;
  var l = substring.length - 1;
  var c = l;

  for (i = string.length; i > 1; i--) {

    if (string.charAt(i) == substring.charAt(c)) {
      c--;
    } else {
      c = l;
    }

    if (c < 0) {
      c = l;
      n++;
    }
  }
  return n;
}

Answer 1

I found the bottle-neck myself.

when I did this

for (i = string.length; i > 1; i--) {

I accidentaly deleted the "var" from var i, so I’ve made i global.
After fixing it I got the expected results.

for (var i = string.length; i > 1; i--) {

I never though that this may be a HUGE difference, so pay attention guys.

Fixed Benckmark test here

Before:

After:

PS: for practical use, do NOT use this functions, the indexOf version is much faster.