UTF16 byte order, patch for enc-utf16.js

[GoogleCodeExporter]

The UTF16 parser is returning different byte order than my .NET pair.

I have this .NET test which works/describes behavior.
        [Test]
        public void TestSHA1()
        {
            var alg = new System.Security.Cryptography.SHA1Managed();
            byte[] bytes = new UnicodeEncoding().GetBytes("ax");
            Assert.AreEqual("61-00-78-00", BitConverter.ToString(bytes));
            byte[] hash = alg.ComputeHash(bytes);
            Assert.AreEqual("E7-77-35-54-8D-77-09-2C-87-B8-30-3E-9F-55-1A-3A-48-B8-0A-A6", BitConverter.ToString(hash));
            string base64String = Convert.ToBase64String(hash);
            Assert.AreEqual("53c1VI13CSyHuDA+n1UaOki4CqY=", base64String);
        }

This is the same data and expectations in Jasmine unit test

describe('CryptoJS test', function () {
    describe('SHA1', function () {
        it('should match .NET version', function () {
            var passwd = CryptoJS.enc.Utf16.parse('ax', true);
            expect(passwd.toString()).toBe('61007800');
            var hash = CryptoJS.SHA1(passwd);
            var hex = CryptoJS.enc.Hex.stringify(hash);
            expect(hex).toBe('e77735548d77092c87b8303e9f551a3a48b80aa6');
            var base64 = CryptoJS.enc.Base64.stringify(hash);
            expect(base64).toBe('53c1VI13CSyHuDA+n1UaOki4CqY=');
        });
    });
});

Most likely you would like to introduce another encoding or parameter to allow 
different byte order. Patch for enc-utf16.js

        /**
         * Converts a UTF-16 string to a word array.
         *
         * @param {string} utf16Str The UTF-16 string.
         * @param {bool} littleEngian byte order
         *
         * @return {WordArray} The word array.
         *
         * @static
         *
         * @example
         *
         *     var wordArray = CryptoJS.enc.Utf16.parse(utf16String);
         */
        parse: function (utf16Str, littleEngian) {
            // Shortcut
            var utf16StrLength = utf16Str.length;
            var le = (typeof littleEngian === "undefined") ? false : littleEngian;

            // Convert
            var words = [];
            for (var i = 0; i < utf16StrLength; i++) {
                  var ch=utf16Str.charCodeAt(i);
                  if(le){
                    ch=ch>>>8|((ch&0xFF)<<8);
                  }
                words[i >>> 1] |= ch << (16 - (i % 2) * 16);
            }

                        return f.create(words, utf16StrLength * 2);
        }

What version of the product are you using? On what operating system?
3.0.2, chrome 21

Original issue reported on code.google.com by pavel.savara on 7 Sep 2012 at 3:40

[GoogleCodeExporter]

I'll accept this as an enhancement to add little-endian UTF-16.

Original comment by Jeff.Mott.OR on 7 Sep 2012 at 3:56

• Added labels: Type-Enhancement
• Removed labels: Type-Defect

[GoogleCodeExporter]

The last line should be  
return WordArray.create(words, utf16StrLength * 2);

same as before. Sorry.

Original comment by pavel.savara on 10 Sep 2012 at 1:01

[GoogleCodeExporter]

Thank God I found this solution! 
Please apply this patch ASAP. It's really problem when trying connect to .NET 
servers.

Original comment by sdany...@gmail.com on 4 Dec 2012 at 2:44

[GoogleCodeExporter]

Latest release includes UTF-16 little endian.

var words = CryptoJS.enc.Utf16LE.parse('Hello, World!');

Original comment by Jeff.Mott.OR on 7 Jan 2013 at 1:56

• Changed state: Fixed