Lib/hashlib.py

   1 # $Id$
   2 #
   3 #  Copyright (C) 2005-2007   Gregory P. Smith (greg@krypto.org)
   4 #  Licensed to PSF under a Contributor Agreement.
   5 #
   6
   7 __doc__ = """hashlib module - A common interface to many hash functions.
   8
   9 new(name, data=b'') - returns a new hash object implementing the
  10                       given hash function; initializing the hash
  11                       using the given binary data.
  12
  13 Named constructor functions are also available, these are faster
  14 than using new(name):
  15
  16 md5(), sha1(), sha224(), sha256(), sha384(), and sha512()
  17
  18 More algorithms may be available on your platform but the above are
  19 guaranteed to exist.
  20
  21 NOTE: If you want the adler32 or crc32 hash functions they are available in
  22 the zlib module.
  23
  24 Choose your hash function wisely.  Some have known collision weaknesses.
  25 sha384 and sha512 will be slow on 32 bit platforms.
  26
  27 Hash objects have these methods:
  28  - update(arg): Update the hash object with the bytes in arg. Repeated calls
  29                 are equivalent to a single call with the concatenation of all
  30                 the arguments.
  31  - digest():    Return the digest of the bytes passed to the update() method
  32                 so far.
  33  - hexdigest(): Like digest() except the digest is returned as a unicode
  34                 object of double length, containing only hexadecimal digits.
  35  - copy():      Return a copy (clone) of the hash object. This can be used to
  36                 efficiently compute the digests of strings that share a common
  37                 initial substring.
  38
  39 For example, to obtain the digest of the string 'Nobody inspects the
  40 spammish repetition':
  41
  42     >>> import hashlib
  43     >>> m = hashlib.md5()
  44     >>> m.update(b"Nobody inspects")
  45     >>> m.update(b" the spammish repetition")
  46     >>> m.digest()
  47     b'\xbbd\x9c\x83\xdd\x1e\xa5\xc9\xd9\xde\xc9\xa1\x8d\xf0\xff\xe9'
  48
  49 More condensed:
  50
  51     >>> hashlib.sha224(b"Nobody inspects the spammish repetition").hexdigest()
  52     'a4337bc45a8fc544c03f52dc550cd6e1e87021bc896588bd79e901e2'
  53
  54 """
  55
  56
  57 def __get_builtin_constructor(name):
  58     if name in ('SHA1', 'sha1'):
  59         import _sha1
  60         return _sha1.sha1
  61     elif name in ('MD5', 'md5'):
  62         import _md5
  63         return _md5.md5
  64     elif name in ('SHA256', 'sha256', 'SHA224', 'sha224'):
  65         import _sha256
  66         bs = name[3:]
  67         if bs == '256':
  68             return _sha256.sha256
  69         elif bs == '224':
  70             return _sha256.sha224
  71     elif name in ('SHA512', 'sha512', 'SHA384', 'sha384'):
  72         import _sha512
  73         bs = name[3:]
  74         if bs == '512':
  75             return _sha512.sha512
  76         elif bs == '384':
  77             return _sha512.sha384
  78
  79     raise ValueError("unsupported hash type")
  80
  81
  82 def __py_new(name, data=b''):
  83     """new(name, data=b'') - Return a new hashing object using the named algorithm;
  84     optionally initialized with data (which must be bytes).
  85     """
  86     return __get_builtin_constructor(name)(data)
  87
  88
  89 def __hash_new(name, data=b''):
  90     """new(name, data=b'') - Return a new hashing object using the named algorithm;
  91     optionally initialized with data (which must be bytes).
  92     """
  93     try:
  94         return _hashlib.new(name, data)
  95     except ValueError:
  96         # If the _hashlib module (OpenSSL) doesn't support the named
  97         # hash, try using our builtin implementations.
  98         # This allows for SHA224/256 and SHA384/512 support even though
  99         # the OpenSSL library prior to 0.9.8 doesn't provide them.
 100         return __get_builtin_constructor(name)(data)
 101
 102
 103 try:
 104     import _hashlib
 105     # use the wrapper of the C implementation
 106     new = __hash_new
 107
 108     for opensslFuncName in filter(lambda n: n.startswith('openssl_'), dir(_hashlib)):
 109         funcName = opensslFuncName[len('openssl_'):]
 110         try:
 111             # try them all, some may not work due to the OpenSSL
 112             # version not supporting that algorithm.
 113             f = getattr(_hashlib, opensslFuncName)
 114             f()
 115             # Use the C function directly (very fast)
 116             exec(funcName + ' = f')
 117         except ValueError:
 118             try:
 119                 # Use the builtin implementation directly (fast)
 120                 exec(funcName + ' = __get_builtin_constructor(funcName)')
 121             except ValueError:
 122                 # this one has no builtin implementation, don't define it
 123                 pass
 124     # clean up our locals
 125     del f
 126     del opensslFuncName
 127     del funcName
 128
 129 except ImportError:
 130     # We don't have the _hashlib OpenSSL module?
 131     # use the built in legacy interfaces via a wrapper function
 132     new = __py_new
 133
 134     # lookup the C function to use directly for the named constructors
 135     md5 = __get_builtin_constructor('md5')
 136     sha1 = __get_builtin_constructor('sha1')
 137     sha224 = __get_builtin_constructor('sha224')
 138     sha256 = __get_builtin_constructor('sha256')
 139     sha384 = __get_builtin_constructor('sha384')
 140     sha512 = __get_builtin_constructor('sha512')