Calulate byte length of string as UTF-8.

utf8length.js

/*
 * Fair License (Fair)
 *
 * Copyright (c) 2012, IWAMURO Motonori
 * 
 * Usage of the works is permitted provided that this instrument is
 * retained with the works, so that any entity that uses the works is
 * notified of this instrument.
 * 
 * DISCLAIMER: THE WORKS ARE WITHOUT WARRANTY.
 */

// Calulate byte length of string as UTF-8.
function utf8length(s) {
  var len = s.length;
  var u8len = 0;
  for (var i = 0; i < len; i++) {
    var c = s.charCodeAt(i);
    if (c <= 0x007f) { // ASCII
      u8len++;
    } else if (c <= 0x07ff) {
      u8len += 2;
    } else if (c <= 0xd7ff || 0xe000 <= c) {
      u8len += 3;
    } else if (c <= 0xdbff) { // high-surrogate code
      c = s.charCodeAt(++i);
      if (c < 0xdc00 || 0xdfff < c) // Is trailing char low-surrogate code?
        throw "Error: Invalid UTF-16 sequence. Missing low-surrogate code.";
      u8len += 4;
    } else /* if (c <= 0xdfff) */ { // low-surrogate code
      throw "Error: Invalid UTF-16 sequence. Missing high-surrogate code.";
    }
  }
  return u8len;
}