thomdixon/pysha2

Pure Python implementation of SHA2 (i.e., SHA224, SHA256, SHA384, and SHA512).

pysha2

License

This software is distributed under the MIT License.

About

pysha2 is a pure Python implementation of the FIPS 180-2 secure hash standard. I originally wrote and published this on the web prior to the inclusion of hashlib into the Python standard library. Unfortunately, the original source was lost due to a hard drive failure. Fortunately, the library proved useful to some at the time (due to the prevalence of Python 2.3 and 2.4), and so I was able to recover all but the unit tests (which I later rewrote).

Usage

The library supports both the "old" hash interface of md5 and sha, as well as the "new" hash interface introduced by hashlib. This permits you to use pysha2 as a drop-in replacement for either interface.

A quick example of hashing a string:

import sha2

print sha2.sha256('Can you keep a secret?').hexdigest()

Testing

To run the included unit tests, execute:

python test.py

in the current directory or run with nosetests.

pysha2's Issues

Missing something?

Hello,
I used your code to hash the variable a
a= "00baf6626abc2df808da36a518c69f09b0d2ed0a79421ccfde4f559d2e42128" {#b"# CHAINS SHORTER at 2128_ }
and it gives the correct value (compared to hashlib),
but if I try to hash
a= "00baf6626abc2df808da36a518c69f09b0d2ed0a79421ccfde4f559d2e42128b" it provides a wrong hash value, got lost on the code so I thought I'd better ask you.
I am intending to use your code two 64 char concatenated variables.

Thanks a lot for your comments.

Typeerror issue on SHA256

sha256 type error
in this function a Typeerror is raised if m in not string BUT:

def update(self, m):
        if not m:
            return
        if type(m) is not str:
            raise TypeError, '%s() argument 1 must be string, not %s' % (sys._getframe().f_code.co_name, type(m).__name__)

In here you call the update function with bytes. therefore causing function to raise error

def digest(self):
        mdi = self._counter & 0x3F
        length = struct.pack('!Q', self._counter<<3)
        
        if mdi < 56:
            padlen = 55-mdi
        else:
            padlen = 119-mdi
        
        r = self.copy()
        r.update('\x80'+('\x00'*padlen)+length)
        return ''.join([struct.pack('!L', i) for i in r._h[:self._output_size]])

is it possible to backwards-hash(sha256) a known-size reversed input from a given hash?

Hi. You wrote the sha256 implementation in this repository which makes me assume that you know how it works (unlike me) and I was wondering if you could answer with YES/NO this question which I'll reiterate here:

Is possible to start from a hash and work backwards through the sha256 algorithm using the same input (but reversed, from the last byte to the first) and get to the same initial hash (which is the hash of empty input, or a 0 bytes file) ?

For example:
I have the input Hello World! which (sha256)hashes to 7f83b1657ff1fc53b92dc18148a1d65dfc2d4b1fa3d677284addd200126d9069
then can I use that very hash to go backwards via a reversed_sha256 algorithm(if it even can be reversed like that!) and coupled with the reversed input (!dlroW olleH) to get to the hash of the empty input which is e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855 ?

Thank you in advance for any reply, even if it's you don't know. It's much appreciated either way!
All the best to you!

Add a LICENSE please?

See http://choosealicense.com/