I need to convert a large array (Uint8Array(224337596)) inside my code. Apparently the size is to big and makes the browser crash.
Is there any workaround to maybe to this in chunks?
var encrypted = convertUint8ArrayToWordArray(mergedArray)
function convertUint8ArrayToWordArray(u8Array) {
var words = [], i = 0, len = u8Array.length;
while (i < len) {
words.push(
(u8Array[i++] << 24) |
(u8Array[i++] << 16) |
(u8Array[i++] << 8) |
(u8Array[i++])
);
}
return {
sigBytes: words.length * 4,
words: words
};
}
If your browser window freezes, than you can process the array asynchronously in batches. I've put the example into snippet
function pause() {
return new Promise(r => setTimeout(r, 0))
}
async function convertUint8ArrayToWordArray(u8Array) {
var words = [], i = 0, len = u8Array.length;
while (i < len) {
words.push(
(u8Array[i++] << 24) |
(u8Array[i++] << 16) |
(u8Array[i++] << 8) |
(u8Array[i++])
);
if (i % 100000 == 0) {
await pause();
}
}
return {
sigBytes: words.length * 4,
words: words
};
}
const bigArray = new Uint8Array(224337596);
for (let idx = 0; idx < bigArray.length; ++idx) {
bigArray[idx] = Math.floor(Math.random() * 256);
}
convertUint8ArrayToWordArray(bigArray).then((res) => {
console.log(res.words[0])
});
Related
Why do I get different result in my encryption on TriplesDes using c# and JavaScript cryptojs? Please see my code below.
c#
public static string EncryptTxt()
{
SHA512CryptoServiceProvider sha = new SHA512CryptoServiceProvider();
using (var tdes = new TripleDESCryptoServiceProvider())
{
var msg = 'jvofs:JCV XXXXX:201911141547:12345678';
var key = 'jjvofs';
var keyOffset = 10;
System.Text.UTF8Encoding UTF8 = new System.Text.UTF8Encoding();
TripleDESCryptoServiceProvider TDESAlgorithm = new TripleDESCryptoServiceProvider();
byte[] Results;
byte[] newKeyx = new byte[24];
byte[] keybyte = sha.ComputeHash(Encoding.UTF8.GetBytes(key));
Array.Copy(keybyte, keyOffset, newKeyx, 0, newKeyx.Length);
TDESAlgorithm.Key = newKeyx;
TDESAlgorithm.Mode = CipherMode.ECB;
TDESAlgorithm.Padding = PaddingMode.PKCS7;
byte[] DataToEncrypt = UTF8.GetBytes(msg);
try
{
ICryptoTransform Encryptor = TDESAlgorithm.CreateEncryptor();
Results = Encryptor.TransformFinalBlock(DataToEncrypt, 0, DataToEncrypt.Length);
}
finally
{
TDESAlgorithm.Clear();
}
var a = Convert.ToBase64String(DataToEncrypt);
var b = Convert.ToBase64String(newKeyx);
var c = Convert.ToBase64String(Results);
return Convert.ToBase64String(Results);
}
}
JavaScript using cryptojs
txtEncrypter = () => {
const msg = 'jvofs:JCV XXXXX:201911141547:12345678';
const key = 'jjvofs';
const keyOffset = 10;
const keybyte: any = this.wordArrayToByteArray(crypto.SHA512(key), 100);
// For message
const dataToEncrypt = crypto.enc.Utf8.parse(msg);
const dte = this.wordArrayToByteArray(dataToEncrypt, 100);
const dataToEncryptx = this._arrayBufferToBase64(dte);
const dataToEncryptxx = crypto.enc.Utf8.parse(dataToEncryptx);
// For key
let newKeyx = keybyte.slice(keyOffset, 34);
const newKeyxB4Splice = newKeyx;
const newKeyxB4Splicex = this._arrayBufferToBase64(newKeyx);
newKeyx = crypto.enc.Utf8.parse(newKeyx);
const options = {
mode: crypto.mode.ECB,
padding: crypto.pad.Pkcs7
};
const encrypted = crypto.TripleDES.encrypt(dataToEncrypt, newKeyx, options);
const base64String = encrypted.toString();
console.log(base64String);
}
wordArrayToByteArray(wordArray, length) {
if (wordArray.hasOwnProperty('sigBytes') && wordArray.hasOwnProperty('words')) {
length = wordArray.sigBytes;
wordArray = wordArray.words;
}
const result = [];
let bytes: any;
let i = 0;
while (length > 0) {
bytes = this.wordToByteArray(wordArray[i], Math.min(4, length));
length -= bytes.length;
result.push(bytes);
i++;
}
return [].concat.apply([], result);
}
wordToByteArray(word: any, length: any) {
const ba = [], xFF = 0xFF;
if (length > 0) {
// tslint:disable-next-line:no-bitwise
ba.push(word >>> 24);
}
if (length > 1) {
// tslint:disable-next-line:no-bitwise
ba.push((word >>> 16) & xFF);
}
if (length > 2) {
// tslint:disable-next-line:no-bitwise
ba.push((word >>> 8) & xFF);
}
if (length > 3) {
// tslint:disable-next-line:no-bitwise
ba.push(word & xFF);
}
return ba;
}
byteArrayToWordArray(ba) {
const wa = [];
let i = 0;
for (i = 0; i < ba.length; i++) {
// tslint:disable-next-line:no-bitwise
wa[(i / 4) | 0] |= ba[i] << (24 - 8 * i);
}
return crypto.lib.WordArray.create(wa);
}
toUTF8Array(str) {
const utf8 = [];
for (let i = 0; i < str.length; i++) {
let charcode = str.charCodeAt(i);
if (charcode < 0x80) { utf8.push(charcode); } else if (charcode < 0x800) {
// tslint:disable-next-line:no-bitwise
utf8.push(0xc0 | (charcode >> 6),
// tslint:disable-next-line: no-bitwise
0x80 | (charcode & 0x3f));
} else if (charcode < 0xd800 || charcode >= 0xe000) {
// tslint:disable-next-line: no-bitwise
utf8.push(0xe0 | (charcode >> 12),
// tslint:disable-next-line: no-bitwise
0x80 | ((charcode>>6) & 0x3f),
// tslint:disable-next-line: no-bitwise
0x80 | (charcode & 0x3f));
} else {
i++;
// UTF-16 encodes 0x10000-0x10FFFF by
// subtracting 0x10000 and splitting the
// 20 bits of 0x0-0xFFFFF into two halves
// tslint:disable-next-line:no-bitwise
charcode = 0x10000 + (((charcode & 0x3ff)<<10)
// tslint:disable-next-line:no-bitwise
| (str.charCodeAt(i) & 0x3ff));
// tslint:disable-next-line:no-bitwise
utf8.push(0xf0 | (charcode >>18),
// tslint:disable-next-line:no-bitwise
0x80 | ((charcode>>12) & 0x3f),
// tslint:disable-next-line: no-bitwise
0x80 | ((charcode>>6) & 0x3f),
// tslint:disable-next-line: no-bitwise
0x80 | (charcode & 0x3f));
}
}
return utf8;
}
_arrayBufferToBase64( buffer ) {
let binary = '';
const bytes = new Uint8Array( buffer );
const len = bytes.byteLength;
for (let i = 0; i < len; i++) {
binary += String.fromCharCode( bytes[ i ] );
}
return window.btoa( binary );
}
When parsing the data for message and key, both c# and JavaScript are the same:
Message:
[C#] anZvZnM6SkNWIFhYWFhYOjIwMTkxMTE0MTU0NzoxMjM0NTY3OA==
[cryptojs] anZvZnM6SkNWIFhYWFhYOjIwMTkxMTE0MTU0NzoxMjM0NTY3OA==
Key:
[C#] smbkkmDrCBdRev7S4hLaWE16Nvym+9gW
[cryptojs] smbkkmDrCBdRev7S4hLaWE16Nvym+9gW
But as soon as the "crypto.TripleDES.encrypt(....)" runs, I get different result for c# and javascript:
[C#] 1pjvBOB81iAOqsskZ+cM080yDU37XBoCwMhbYULwva/Nql5vbEMiPQ==
[cryptojs] ncCcCYNy3jVsB/95SaC2N1rH5Q+hX04WvScMvwmtkPkrnL7Ki1bmPg==
After the key has been determined from the byte-array of the hash as a sub-array (incl. index 10 to excl. index 34), this sub-array must be converted back into a WordArray with equivalent content, i.e. the line:
newKeyx = crypto.enc.Utf8.parse(newKeyx);
must be replaced by:
newKeyx = byteArrayToWordArray(newKeyx);
With this change, the NodeJS-code returns the same result as the C#-code.
The conversion WordArray <-> byte-array (and thus all functions required for these conversions) aren't really necessary, because as an alternative the key can also be derived using only the CryptoJS-encoders:
...
const key = 'jjvofs';
const keyOffset = 10;
const keyLength = 24;
const keyHash = crypto.enc.Hex.stringify(crypto.SHA512(key));
const newKey = crypto.enc.Hex.parse(keyHash.slice(keyOffset * 2, (keyOffset + keyLength) * 2));
...
By the way: The ECB-mode is insecure and instead of TripleDES the more performant AES should be used.
I receive an id_token as part of my current href. It is encoded in base64. I try to decode it using atob(extractedIdToken), but get the following error:
Failed to execute 'atob' on 'Window': The string to be decoded is not correctly encoded
When I copy and paste the extracted id_token in my code and go to an online decoding site, it decodes correctly. Do you have suggestion?
I always use this to decode and encode in Base64, try it
var Base64 = {
_keyStr: "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=",
encode: function (e) {
var t = "";
var n, r, i, s, o, u, a;
var f = 0;
e = Base64._utf8_encode(e);
while (f < e.length) {
n = e.charCodeAt(f++);
r = e.charCodeAt(f++);
i = e.charCodeAt(f++);
s = n >> 2;
o = (n & 3) << 4 | r >> 4;
u = (r & 15) << 2 | i >> 6;
a = i & 63;
if (isNaN(r)) {
u = a = 64
} else if (isNaN(i)) {
a = 64
}
t = t + this._keyStr.charAt(s) + this._keyStr.charAt(o) + this._keyStr.charAt(u) + this._keyStr.charAt(a)
}
return t
},
decode: function (e) {
var t = "";
var n, r, i;
var s, o, u, a;
var f = 0;
e = e.replace(/[^A-Za-z0-9+/=]/g, "");
while (f < e.length) {
s = this._keyStr.indexOf(e.charAt(f++));
o = this._keyStr.indexOf(e.charAt(f++));
u = this._keyStr.indexOf(e.charAt(f++));
a = this._keyStr.indexOf(e.charAt(f++));
n = s << 2 | o >> 4;
r = (o & 15) << 4 | u >> 2;
i = (u & 3) << 6 | a;
t = t + String.fromCharCode(n);
if (u != 64) {
t = t + String.fromCharCode(r)
}
if (a != 64) {
t = t + String.fromCharCode(i)
}
}
t = Base64._utf8_decode(t);
return t
},
_utf8_encode: function (e) {
e = e.replace(/rn/g, "n");
var t = "";
for (var n = 0; n < e.length; n++) {
var r = e.charCodeAt(n);
if (r < 128) {
t += String.fromCharCode(r)
} else if (r > 127 && r < 2048) {
t += String.fromCharCode(r >> 6 | 192);
t += String.fromCharCode(r & 63 | 128)
} else {
t += String.fromCharCode(r >> 12 | 224);
t += String.fromCharCode(r >> 6 & 63 | 128);
t += String.fromCharCode(r & 63 | 128)
}
}
return t
},
_utf8_decode: function (e) {
var t = "";
var n = 0;
var r = c1 = c2 = 0;
while (n < e.length) {
r = e.charCodeAt(n);
if (r < 128) {
t += String.fromCharCode(r);
n++
} else if (r > 191 && r < 224) {
c2 = e.charCodeAt(n + 1);
t += String.fromCharCode((r & 31) << 6 | c2 & 63);
n += 2
} else {
c2 = e.charCodeAt(n + 1);
c3 = e.charCodeAt(n + 2);
t += String.fromCharCode((r & 15) << 12 | (c2 & 63) << 6 | c3 & 63);
n += 3
}
}
return t
}
};
Thanks all for your answers. I ended up moving the Base64 decoding process to the Java backend using the java package: java.util.Base64.
In my case, the issue was that the encoded string should have been treated as segments. Just separate the code with the separator "." or whatever your case is. Then, decode only the different parts separately. This resolved my issue and I was able to decode using window.atob(). Hope this helps.
I have what I thought was a pretty simply question. I'm using this code to generate a SHA1 uuid in Golang:
namespace := uuid.Parse("b9cfdb9d-f741-4e1f-89ae-fac6b2a5d740")
sha := uuid.NewSHA1(namespace, []byte("something"))
fmt.Println(sha.String())
Now I want to generate the same UUID in javascript, and I thought it would be as easy as something like this:
var hash = CryptoJS.SHA1("b9cfdb9d-f741-4e1f-89ae-fac6b2a5d740" + "something")
// chomp the hash into a UUID string
However, I'm running into serious issues. It seems that the uuid.Parse function in Golang is running this parsing function that converts the namespace to a 16-byte array, so even though I use the same SHA1 algorithm in Javascript, I'm not getting the same output.
I'v been messing around with doing the same in JS, but I'm stumped.
Any smart crypto people in here that can help me?
Well, that only took me a month.
var SHA1Generator = {
hex_chr: "0123456789abcdef",
hex: function (num) {
var str = "";
for (var j = 7; j >= 0; j--)
str += this.hex_chr.charAt((num >> (j * 4)) & 0x0F);
return str;
},
str2blks_SHA1: function (str) {
var nblk = ((str.length + 8) >> 6) + 1;
var blks = new Array(nblk * 16);
for (var i = 0; i < nblk * 16; i++) blks[i] = 0;
for (i = 0; i < str.length; i++)
blks[i >> 2] |= str.charCodeAt(i) << (24 - (i % 4) * 8);
blks[i >> 2] |= 0x80 << (24 - (i % 4) * 8);
blks[nblk * 16 - 1] = str.length * 8;
return blks;
},
add: function (x, y) {
var lsw = (x & 0xFFFF) + (y & 0xFFFF);
var msw = (x >> 16) + (y >> 16) + (lsw >> 16);
return (msw << 16) | (lsw & 0xFFFF);
},
rol: function (num, cnt) {
return (num << cnt) | (num >>> (32 - cnt));
},
ft: function (t, b, c, d) {
if (t < 20) return (b & c) | ((~b) & d);
if (t < 40) return b ^ c ^ d;
if (t < 60) return (b & c) | (b & d) | (c & d);
return b ^ c ^ d;
},
kt: function (t) {
return (t < 20) ? 1518500249 : (t < 40) ? 1859775393 :
(t < 60) ? -1894007588 : -899497514;
},
calcSHA1FromByte: function(byteArr) {
var str = '';
for(var i=0; i<byteArr.length; i++)
str += String.fromCharCode(byteArr[i]);
return this.calcSHA1(str);
},
calcSHA1: function (str) {
if (str != '') {
var x = this.str2blks_SHA1(str);
var w = new Array(80);
var a = 1732584193;
var b = -271733879;
var c = -1732584194;
var d = 271733878;
var e = -1009589776;
for (var i = 0; i < x.length; i += 16) {
var olda = a;
var oldb = b;
var oldc = c;
var oldd = d;
var olde = e;
for (var j = 0; j < 80; j++) {
if (j < 16) w[j] = x[i + j];
else w[j] = this.rol(w[j - 3] ^ w[j - 8] ^ w[j - 14] ^ w[j - 16], 1);
t = this.add(this.add(this.rol(a, 5), this.ft(j, b, c, d)), this.add(this.add(e, w[j]), this.kt(j)));
e = d;
d = c;
c = this.rol(b, 30);
b = a;
a = t;
}
a = this.add(a, olda);
b = this.add(b, oldb);
c = this.add(c, oldc);
d = this.add(d, oldd);
e = this.add(e, olde);
}
return this.hex(a) + this.hex(b) + this.hex(c) + this.hex(d) + this.hex(e);
}
else {
return '';
}
}
};
function stringToByteArray(str) {
var bytes = [];
for (var i = 0; i < str.length; ++i) {
bytes.push(str.charCodeAt(i));
}
return bytes;
}
function uuidToByteArray(hex) {
// If this is a uuid, remove the dashes
hex = hex.replace(/-/g, "");
// convert each hex number into a string representation
// of the byte integer.
var bytes = [];
for(var i = 0; i < hex.length; i += 2) {
bytes.push(parseInt(hex.substring(i, i+2),16));
}
return bytes;
}
function sha1ToUUID5(hash) {
var uuid = hash.substring(0, 8) +
'-' + hash.substring(8, 12) +
// four most significant bits holds version number 5
'-' + ((parseInt(hash.substring(12, 16), 16) & 0x0fff) | 0x5000).toString(16) +
// two most significant bits holds zero and one for variant DCE1.1
'-' + ((parseInt(hash.substring(16, 20), 16) & 0x3fff) | 0x8000).toString(16) +
'-' + hash.substring(20, 32); //12 digits
return uuid;
}
var namespace = "e75a36a9-3323-40dd-a7d1-1c57ad2aa3cd"
var id = "event154"
var namespaceBytes = uuidToByteArray(namespace);
var idBytes = stringToByteArray(id);
var allBytes = namespaceBytes.concat(idBytes);
console.log("ORG 4505612c-c323-5d6f-b5cc-b7f362b9ba55")
console.log("NEW " + sha1ToUUID5(SHA1Generator.calcSHA1FromByte(allBytes)))
What's the easiest way to convert a floating point number to its binary representation in Javascript? (e.g. 1.0 -> 0x3F800000).
I have tried to do it manually, and this works to some extent (with usual numbers), but it fails for very big or very small numbers (no range checking) and for special cases (NaN, infinity, etc.):
function floatToNumber(flt)
{
var sign = (flt < 0) ? 1 : 0;
flt = Math.abs(flt);
var exponent = Math.floor(Math.log(flt) / Math.LN2);
var mantissa = flt / Math.pow(2, exponent);
return (sign << 31) | ((exponent + 127) << 23) | ((mantissa * Math.pow(2, 23)) & 0x7FFFFF);
}
Am I reinventing the wheel?
EDIT: I've improved my version, now it handles special cases.
function assembleFloat(sign, exponent, mantissa)
{
return (sign << 31) | (exponent << 23) | (mantissa);
}
function floatToNumber(flt)
{
if (isNaN(flt)) // Special case: NaN
return assembleFloat(0, 0xFF, 0x1337); // Mantissa is nonzero for NaN
var sign = (flt < 0) ? 1 : 0;
flt = Math.abs(flt);
if (flt == 0.0) // Special case: +-0
return assembleFloat(sign, 0, 0);
var exponent = Math.floor(Math.log(flt) / Math.LN2);
if (exponent > 127 || exponent < -126) // Special case: +-Infinity (and huge numbers)
return assembleFloat(sign, 0xFF, 0); // Mantissa is zero for +-Infinity
var mantissa = flt / Math.pow(2, exponent);
return assembleFloat(sign, exponent + 127, (mantissa * Math.pow(2, 23)) & 0x7FFFFF);
}
I'm still not sure if this works 100% correctly, but it seems to work good enough.
(I'm still looking for existing implementations).
new technologies are making this easy and probably also more forward-compatible. I love extending built in prototypes, not everyone does. So feel free to modify following code to classical procedural approach:
(function() {
function NumberToArrayBuffer() {
// Create 1 entry long Float64 array
return [new Float64Array([this]).buffer];
}
function NumberFromArrayBuffer(buffer) {
// Off course, the buffer must be at least 8 bytes long, otherwise this is a parse error
return new Float64Array(buffer, 0, 1)[0];
}
if(Number.prototype.toArrayBuffer) {
console.warn("Overriding existing Number.prototype.toArrayBuffer - this can mean framework conflict, new WEB API conflict or double inclusion.");
}
Number.prototype.toArrayBuffer = NumberToArrayBuffer;
Number.prototype.fromArrayBuffer = NumberFromArrayBuffer;
// Hide this methods from for-in loops
Object.defineProperty(Number.prototype, "toArrayBuffer", {enumerable: false});
Object.defineProperty(Number.prototype, "fromArrayBuffer", {enumerable: false});
})();
Test:
(function() {
function NumberToArrayBuffer() {
// Create 1 entry long Float64 array
return new Float64Array([this.valueOf()]).buffer;
}
function NumberFromArrayBuffer(buffer) {
// Off course, the buffer must be ar least 8 bytes long, otherwise this is a parse error
return new Float64Array(buffer, 0, 1)[0];
}
if(Number.prototype.toArrayBuffer) {
console.warn("Overriding existing Number.prototype.toArrayBuffer - this can mean framework conflict, new WEB API conflict or double inclusion.");
}
Number.prototype.toArrayBuffer = NumberToArrayBuffer;
Number.fromArrayBuffer = NumberFromArrayBuffer;
// Hide this methods from for-in loops
Object.defineProperty(Number.prototype, "toArrayBuffer", {enumerable: false});
Object.defineProperty(Number, "fromArrayBuffer", {enumerable: false});
})();
var test_numbers = [0.00000001, 666666666666, NaN, Infinity, -Infinity,0,-0];
console.log("Conversion symethry test: ");
test_numbers.forEach(
function(num) {
console.log(" ", Number.fromArrayBuffer((num).toArrayBuffer()));
}
);
console.log("Individual bytes of a Number: ",new Uint8Array((666).toArrayBuffer(),0,8));
<script src="https://getfirebug.com/firebug-lite-debug.js"></script>
Here's a function that works on everything I've tested it on, except it doesn't distinguish -0.0 and +0.0.
It's based on code from http://jsfromhell.com/classes/binary-parser, but it's specialized for 32-bit floats and returns an integer instead of a string. I also modified it to make it faster and (slightly) more readable.
// Based on code from Jonas Raoni Soares Silva
// http://jsfromhell.com/classes/binary-parser
function encodeFloat(number) {
var n = +number,
status = (n !== n) || n == -Infinity || n == +Infinity ? n : 0,
exp = 0,
len = 281, // 2 * 127 + 1 + 23 + 3,
bin = new Array(len),
signal = (n = status !== 0 ? 0 : n) < 0,
n = Math.abs(n),
intPart = Math.floor(n),
floatPart = n - intPart,
i, lastBit, rounded, j, exponent;
if (status !== 0) {
if (n !== n) {
return 0x7fc00000;
}
if (n === Infinity) {
return 0x7f800000;
}
if (n === -Infinity) {
return 0xff800000
}
}
i = len;
while (i) {
bin[--i] = 0;
}
i = 129;
while (intPart && i) {
bin[--i] = intPart % 2;
intPart = Math.floor(intPart / 2);
}
i = 128;
while (floatPart > 0 && i) {
(bin[++i] = ((floatPart *= 2) >= 1) - 0) && --floatPart;
}
i = -1;
while (++i < len && !bin[i]);
if (bin[(lastBit = 22 + (i = (exp = 128 - i) >= -126 && exp <= 127 ? i + 1 : 128 - (exp = -127))) + 1]) {
if (!(rounded = bin[lastBit])) {
j = lastBit + 2;
while (!rounded && j < len) {
rounded = bin[j++];
}
}
j = lastBit + 1;
while (rounded && --j >= 0) {
(bin[j] = !bin[j] - 0) && (rounded = 0);
}
}
i = i - 2 < 0 ? -1 : i - 3;
while(++i < len && !bin[i]);
(exp = 128 - i) >= -126 && exp <= 127 ? ++i : exp < -126 && (i = 255, exp = -127);
(intPart || status !== 0) && (exp = 128, i = 129, status == -Infinity ? signal = 1 : (status !== status) && (bin[i] = 1));
n = Math.abs(exp + 127);
exponent = 0;
j = 0;
while (j < 8) {
exponent += (n % 2) << j;
n >>= 1;
j++;
}
var mantissa = 0;
n = i + 23;
for (; i < n; i++) {
mantissa = (mantissa << 1) + bin[i];
}
return ((signal ? 0x80000000 : 0) + (exponent << 23) + mantissa) | 0;
}
I'm trying to base64 encode a utf8 string containing Thai characters. I'm using the browser's built in btoa function. It works for ascii text, however Thai is causing it to throw a INVALID_CHARACTER_ERR: DOM Exception 5 exception.
Here's a sample that fails (the character that looks like an "n" is Thai)
btoa('aก')
What do I need to do to base64 encode non-ascii strings?
var Base64 = {
encode: function(s) {
return btoa(unescape(encodeURIComponent(s)));
},
decode: function(s) {
return decodeURIComponent(escape(atob(s)));
}
};
Unfortunately btoa/atob aren't specified in any standard, but the implementations in firefox and webkit both fail on multibyte characters so even if they were now specified those builtin functions would not be able to support multibyte characters (as the input and output strings would necessarily change).
It would seem your only option would be to roll your own base64 encode+decode routines
check this workaround
http://ecmanaut.blogspot.com/2006/07/encoding-decoding-utf8-in-javascript.html
I know this is old, but I was recently looking for a UTF8-to-Base64 encoder as well. I found a handy little script at http://www.webtoolkit.info/javascript-base64.html, and a performance improved version at http://jsbase64.codeplex.com/.
Here is the script:
var B64 = {
alphabet: 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=',
lookup: null,
ie: /MSIE /.test(navigator.userAgent),
ieo: /MSIE [67]/.test(navigator.userAgent),
encode: function (s) {
var buffer = B64.toUtf8(s),
position = -1,
len = buffer.length,
nan0, nan1, nan2, enc = [, , , ];
if (B64.ie) {
var result = [];
while (++position < len) {
nan0 = buffer[position];
nan1 = buffer[++position];
enc[0] = nan0 >> 2;
enc[1] = ((nan0 & 3) << 4) | (nan1 >> 4);
if (isNaN(nan1))
enc[2] = enc[3] = 64;
else {
nan2 = buffer[++position];
enc[2] = ((nan1 & 15) << 2) | (nan2 >> 6);
enc[3] = (isNaN(nan2)) ? 64 : nan2 & 63;
}
result.push(B64.alphabet.charAt(enc[0]), B64.alphabet.charAt(enc[1]), B64.alphabet.charAt(enc[2]), B64.alphabet.charAt(enc[3]));
}
return result.join('');
} else {
var result = '';
while (++position < len) {
nan0 = buffer[position];
nan1 = buffer[++position];
enc[0] = nan0 >> 2;
enc[1] = ((nan0 & 3) << 4) | (nan1 >> 4);
if (isNaN(nan1))
enc[2] = enc[3] = 64;
else {
nan2 = buffer[++position];
enc[2] = ((nan1 & 15) << 2) | (nan2 >> 6);
enc[3] = (isNaN(nan2)) ? 64 : nan2 & 63;
}
result += B64.alphabet[enc[0]] + B64.alphabet[enc[1]] + B64.alphabet[enc[2]] + B64.alphabet[enc[3]];
}
return result;
}
},
decode: function (s) {
if (s.length % 4)
throw new Error("InvalidCharacterError: 'B64.decode' failed: The string to be decoded is not correctly encoded.");
var buffer = B64.fromUtf8(s),
position = 0,
len = buffer.length;
if (B64.ieo) {
var result = [];
while (position < len) {
if (buffer[position] < 128)
result.push(String.fromCharCode(buffer[position++]));
else if (buffer[position] > 191 && buffer[position] < 224)
result.push(String.fromCharCode(((buffer[position++] & 31) << 6) | (buffer[position++] & 63)));
else
result.push(String.fromCharCode(((buffer[position++] & 15) << 12) | ((buffer[position++] & 63) << 6) | (buffer[position++] & 63)));
}
return result.join('');
} else {
var result = '';
while (position < len) {
if (buffer[position] < 128)
result += String.fromCharCode(buffer[position++]);
else if (buffer[position] > 191 && buffer[position] < 224)
result += String.fromCharCode(((buffer[position++] & 31) << 6) | (buffer[position++] & 63));
else
result += String.fromCharCode(((buffer[position++] & 15) << 12) | ((buffer[position++] & 63) << 6) | (buffer[position++] & 63));
}
return result;
}
},
toUtf8: function (s) {
var position = -1,
len = s.length,
chr, buffer = [];
if (/^[\x00-\x7f]*$/.test(s)) while (++position < len)
buffer.push(s.charCodeAt(position));
else while (++position < len) {
chr = s.charCodeAt(position);
if (chr < 128)
buffer.push(chr);
else if (chr < 2048)
buffer.push((chr >> 6) | 192, (chr & 63) | 128);
else
buffer.push((chr >> 12) | 224, ((chr >> 6) & 63) | 128, (chr & 63) | 128);
}
return buffer;
},
fromUtf8: function (s) {
var position = -1,
len, buffer = [],
enc = [, , , ];
if (!B64.lookup) {
len = B64.alphabet.length;
B64.lookup = {};
while (++position < len)
B64.lookup[B64.alphabet.charAt(position)] = position;
position = -1;
}
len = s.length;
while (++position < len) {
enc[0] = B64.lookup[s.charAt(position)];
enc[1] = B64.lookup[s.charAt(++position)];
buffer.push((enc[0] << 2) | (enc[1] >> 4));
enc[2] = B64.lookup[s.charAt(++position)];
if (enc[2] == 64)
break;
buffer.push(((enc[1] & 15) << 4) | (enc[2] >> 2));
enc[3] = B64.lookup[s.charAt(++position)];
if (enc[3] == 64)
break;
buffer.push(((enc[2] & 3) << 6) | enc[3]);
}
return buffer;
}
};
Disclaimer: I haven't tested this with Thai characters specifically, but assume it will work.
Sav