baboleevan / crypto-js Goto Github PK

Currently it seems that symmetric ciphers like AES can work with CBC mode but 
only using random iv. 

Could you add ability to provide iv along with mode in options to AES?

What steps will reproduce the problem?
1. run Crypto.SHA256("á") in firebug

What is the expected output?
fb9778719d93551b1c88df5f1ab229f83ea30af112fc51017c9f0a383dcf6cb3

What do you see instead?
eb3e85d7fbf56104c8021681e68d771b9b9205b93e9219c8f81f60af703de381

What version of the product are you using?
2.0.0

On what operating system?
Mac/Chrome 7

What steps will reproduce the problem?
1. compute has using Crypto.HMAC(Crypto.SHA1, "Message", "Key", { asBytes: true 
}); in IE
2. Repeat step (1) in Safari
3. Compare array of bytes

What is the expected output? What do you see instead?
Output should be the same, but they're different.

What version of the product are you using? On what operating system?
Windows Vista Home Edition, IE 8, Safari 5

When using the HMAC SHA1 functions, the functions do not return the same
data as provided by the RFC2202 (http://www.faqs.org/rfcs/rfc2202.html).

Problem: Browsers will stop and display a warning when crypto operations takes 
too long to run (like PBKDF2 with large iteration). Means of detection browsers 
use varies, some measure actual time, others measuring # of statements.

Suggested solution: divide up crypto operations into smaller units and run each 
unit through a callback argument. With browsers, callback can be implemented 
using setTimeout or Web worker API. With Node.js, callback can call 
process.nextTick.

What steps will reproduce the problem?
1. in an HTML file, include scripts/crypto-md5/crypto-md5.js (from the 
latest downloadable zip file)
2. also include scripts/crypto-sha1/crypto-sha1.js
3. the Crypto.MD5 function will not be available (you get an error if you 
try to use it)
4. remove scripts/crypto-sha1/crypto-sha1.js
5. the Crypto.MD5 function is now available

What is the expected output? What do you see instead?
I expect to be able to use both the Crypto.MD5 and Crypto.SHA1 (and 
Crypto.SHA256)  in the same page.

What version of the product are you using? On what operating system?
crypto-js 2.0.0
OS is Linux (Ubuntu 9.10)
Browser = Firefox 3.5.8

Please provide any additional information below.
It seems crypto-sha1.js overwrites the Crypto object, thus removing the 
Crypto.MD5 function
This is probably true for all the Crypto.* methods

What steps will reproduce the problem?

Use the latest Chrome dev (6.0.495.0)

Call:
Crypto.util.bytesToBase64(Crypto.HMAC(Crypto.SHA1, "GA", "GA", {asBytes: true}))

What is the expected output? What do you see instead?

Expected: bBg1Ebw0ajGh8E47JCjrZ63PdKI=

Instead I see: o4iKvmV1Ahll2gbgpSLys/9e5pY=

What version of the product are you using? On what operating system?

Chrome  (6.0.495.0) on OSX

Please provide any additional information below.

What steps will reproduce the problem?
1. Start the Testsuite on Opera 10.53

What is the expected output? What do you see instead?

All tests should pass but these are failing

test util.gt: failed.
Value should be true.
Expected: true (boolean)
Actual: false (boolean)

test util.lt: failed.
Value should be true.
Expected: true (boolean)
Actual: false (boolean)

test util.endian: failed.
Values should be equal.
Expected: 4041261184 (number)
Actual: -253706112 (number)

test enc.UTF8.decode: failed.
Values should be equal.
Expected: 1105365410,3465620992 (string)
Actual: 1105365410,-829346304 (string)

test enc.Base64.decode: failed.
Values should be equal.
Expected: 352033795,3648913408 (string)
Actual: 352033795,-646053888 (object)

test types.WordArray.concat: failed.
Values should be equal.
Expected: 4294967295 (string)
Actual: -1 (string)

test SHA256: failed.
Values should be equal.
Expected: f7846f55cf23e14eebeab5b4e1550cad5b509e3348fbc4efa3a1413d393cb650 
(string)
Actual: 8a19c396a89853b8baf0050e3b4dfeb4f19bae586f6b491249a1fe38ff98faa5 
(string)

Opera
Version 10.53
Build 3374
Platform Win32

I needed compatibility with node.js crypto, so I had to hack AES.js to pass IV 
in. At the same time, IV had to be packaged separately and not part of the 
encrypted result.

My modified version of AES.js is attached for your review.

Note that options.iv is assumed to be an array in my version.

Thx.

What steps will reproduce the problem?
1. Replace YUI Compressor with Google Closure compiler
2. See build fail

The following cause errors and/or warnings:

in CBC.js:
while (c.pop() != 0x80) ;

The compiler wants {} instead of ;

in AES.js (twice):
mode = options && options.mode || C.mode.OFB;

The variable mode is not declared.  Prepending "var " to the lines solves this.

While using crypto-js we have experienced some problems in converting a byte 
string into a list of integers. We occasionally get values bigger than 0xff. 
Which obviously should not happen when converting a bytestring into integers.

We were able to fix the problem by doing a bitwise AND against 0xff before 
outputting the value (see code below). This seemed to fix the issue for us. We 
feel this might be a bug in the library. In case we got it wrong, we'd like to 
gain better understanding about what is happening.

Crypto.charenc.Binary.stringToBytes = function (str) {
  for (var bytes = [], i = 0; i < str.length; i++) {
    bytes.push(str.charCodeAt(i) & 0xff);
  }
  return bytes;
};

What steps will reproduce the problem?
1. Call Crypto.MD5() with a string larger than 4000 characters

What is the expected output? What do you see instead?
The returned MD5 hash string is the correct length, however every character
is a zero.

What version of the product are you using? On what operating system?
Crypto-JS v1.1.0, Firefox 3.5, Windows XP Pro

We are trying to decrypt existing AES crypto that uses the CBC mode. The 
problem is that the upper layers always make sure that the plain text is of 
correct length, and thus no padding mechanism is needed from the encryption 
library. crypto-js however, assumes in CBC mode that the data has been padded 
in a specific way.

We are trying to write a library that is compatible with a piece of existing 
software, so we cannot change that software to add a crypto-js compatible 
padding. Also the other way around, while encrypting the data, the padding 
added by crypto-js CBC code might break the original application while it tries 
to decrypt items created by crypto-js.

We were able to work around this by making our own copy of the cbc mode and 
commenting out the padding related code. However we are hoping the library 
would gain an option for disabling the padding.