Fast Int Math + Bitwise Hacks For JavaScript

bitwise-hacks.js

// http://michalbe.blogspot.com.br/2013/03/javascript-less-known-parts-bitwise.html
// http://jsperf.com/bitwise-vs-math-object
// http://united-coders.com/christian-harms/results-for-game-for-forfeits-and-the-winner-is/
// https://mudcu.be/journal/2011/11/bitwise-gems-and-other-optimizations/
// https://dreaminginjavascript.wordpress.com/2009/02/09/bitwise-byte-foolish/
// http://jsperf.com/math-min-max-vs-ternary-vs-if/24

"use strict";

var PI = Math.PI;
var HALF_PI = PI * 0.5;
var TWO_PI = PI * 2;
var RADIAN = 180 / PI;

// Max int for simple bitwise operations
// https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Bitwise_Operators#Signed_32-bit_integers
var MAX_SIGNED_32_BIT_INT = Math.pow(2, 31) - 1;
var MIN_SIGNED_32_BIT_INT = ~MAX_SIGNED_32_BIT_INT;

// ECMAScript 6 - MIN/MAX safe integer
if (Number.MAX_SAFE_INTEGER === void 0) {
  Number.MIN_SAFE_INTEGER = -( Number.MAX_SAFE_INTEGER = Math.pow(2, 53) - 1 );
}

function ABS_INT(n) {
  return (n ^ (n >> 31)) - (n >> 31);
}

function MAX_INT(a, b) {
  return a - ((a - b) & ((a - b) >> 31));
  //return a ^ ((a ^ b) & -(a < b));
  //  var c = a - b;
  //  return a - ((c >> 31) & 0x1) * c;  
}

function MIN_INT(a, b) {
  return a - ((a - b) & ((b - a) >> 31));
  // return b ^ ((a ^ b) & -(a < b));
}

function CLAMP_INT( x, min, max ) {
  x = x - ((x - max) & ((max - x) >> 31));
  return x - ((x - min) & ((x - min) >> 31));
  //return ( n > max ) ? max : ( n < min ) ? n : min;
}

function IS_INT_POWER_OF_TWO( value ) {
  return ( value & ( value - 1 ) ) === 0 && value !== 0;
}

function PLUS_ONE_INT(n) { // slower than ++
  return -~n;
}

function MINUS_ONE_INT() { // slower than --
  return ~-n;
}

function IS_ODD_INT(n) {
  return (n & 1) === 1;
}

function ARRAY_SWAP_INT(array, i, j) {
  array[i] ^= array[j];
  array[j] ^= array[i];
  array[i] ^= array[j];
}

function HAS_SAME_SIGN(a, b) {
  return (a ^ b) >= 0;
}

function POW_OF_2_INT(n) { // === Math.pow(2, n)
  return 2 << (n - 1);
}

function AVG_INT(a, b) { // a + b / 2
  return (a + b) >> 1;
}

function TOGGLE_A_B_INT(n, a, b) { // if(x==a) x=b; if(x==b) x=a;
  return a ^ b ^ n;
}


function SET_BIT(n, bit) {
  return n | (1 << bit);
}

function CLEAR_BIT(n, bit) {
  return n & ~(1 << bit);
}

function MODULO_INT(numerator, divisor) { // 600% faster out of function than %
  return numerator & (divisor - 1);
} 

May be var MIN_SIGNED_32_BIT_INT = ~MAX_SIGNED_32_BIT_INT + 1 ?

Looks like MODULO_INT doesn't work as intended as

numerator & (divisor - 1) != numerator % divisor

Or did I misunderstand the function of MODULO_INT?

@AugLuk yeah, it could be better documented what the limitations are.
The divisor has to be a power of 2.
When it is, you get the same result as % always (even for huge floats, it's 0 :v) but the correct result until a divisor of 2**53 aka 0x20000000000000

MODULO_INT only works for divisors of power 2 e.g. 2,4,8,16, ...

another one: https://gist.github.com/everget/320499f197bc27901b90847bf9159164