1 # Test hashlib module 2 # 3 # $Id$ 4 # 5 # Copyright (C) 2005-2010 Gregory P. Smith (greg (at] krypto.org) 6 # Licensed to PSF under a Contributor Agreement. 7 # 8 9 import array 10 import hashlib 11 import itertools 12 import sys 13 try: 14 import threading 15 except ImportError: 16 threading = None 17 import unittest 18 import warnings 19 from test import test_support 20 from test.test_support import _4G, precisionbigmemtest 21 22 # Were we compiled --with-pydebug or with #define Py_DEBUG? 23 COMPILED_WITH_PYDEBUG = hasattr(sys, 'gettotalrefcount') 24 25 26 def hexstr(s): 27 import string 28 h = string.hexdigits 29 r = '' 30 for c in s: 31 i = ord(c) 32 r = r + h[(i >> 4) & 0xF] + h[i & 0xF] 33 return r 34 35 36 class HashLibTestCase(unittest.TestCase): 37 supported_hash_names = ( 'md5', 'MD5', 'sha1', 'SHA1', 38 'sha224', 'SHA224', 'sha256', 'SHA256', 39 'sha384', 'SHA384', 'sha512', 'SHA512' ) 40 41 _warn_on_extension_import = COMPILED_WITH_PYDEBUG 42 43 def _conditional_import_module(self, module_name): 44 """Import a module and return a reference to it or None on failure.""" 45 try: 46 exec('import '+module_name) 47 except ImportError, error: 48 if self._warn_on_extension_import: 49 warnings.warn('Did a C extension fail to compile? %s' % error) 50 return locals().get(module_name) 51 52 def __init__(self, *args, **kwargs): 53 algorithms = set() 54 for algorithm in self.supported_hash_names: 55 algorithms.add(algorithm.lower()) 56 self.constructors_to_test = {} 57 for algorithm in algorithms: 58 self.constructors_to_test[algorithm] = set() 59 60 # For each algorithm, test the direct constructor and the use 61 # of hashlib.new given the algorithm name. 62 for algorithm, constructors in self.constructors_to_test.items(): 63 constructors.add(getattr(hashlib, algorithm)) 64 def _test_algorithm_via_hashlib_new(data=None, _alg=algorithm): 65 if data is None: 66 return hashlib.new(_alg) 67 return hashlib.new(_alg, data) 68 constructors.add(_test_algorithm_via_hashlib_new) 69 70 _hashlib = self._conditional_import_module('_hashlib') 71 if _hashlib: 72 # These two algorithms should always be present when this module 73 # is compiled. If not, something was compiled wrong. 74 assert hasattr(_hashlib, 'openssl_md5') 75 assert hasattr(_hashlib, 'openssl_sha1') 76 for algorithm, constructors in self.constructors_to_test.items(): 77 constructor = getattr(_hashlib, 'openssl_'+algorithm, None) 78 if constructor: 79 constructors.add(constructor) 80 81 _md5 = self._conditional_import_module('_md5') 82 if _md5: 83 self.constructors_to_test['md5'].add(_md5.new) 84 _sha = self._conditional_import_module('_sha') 85 if _sha: 86 self.constructors_to_test['sha1'].add(_sha.new) 87 _sha256 = self._conditional_import_module('_sha256') 88 if _sha256: 89 self.constructors_to_test['sha224'].add(_sha256.sha224) 90 self.constructors_to_test['sha256'].add(_sha256.sha256) 91 _sha512 = self._conditional_import_module('_sha512') 92 if _sha512: 93 self.constructors_to_test['sha384'].add(_sha512.sha384) 94 self.constructors_to_test['sha512'].add(_sha512.sha512) 95 96 super(HashLibTestCase, self).__init__(*args, **kwargs) 97 98 def test_hash_array(self): 99 a = array.array("b", range(10)) 100 constructors = self.constructors_to_test.itervalues() 101 for cons in itertools.chain.from_iterable(constructors): 102 c = cons(a) 103 c.hexdigest() 104 105 def test_algorithms_attribute(self): 106 self.assertEqual(hashlib.algorithms, 107 tuple([_algo for _algo in self.supported_hash_names if 108 _algo.islower()])) 109 110 def test_unknown_hash(self): 111 try: 112 hashlib.new('spam spam spam spam spam') 113 except ValueError: 114 pass 115 else: 116 self.assertTrue(0 == "hashlib didn't reject bogus hash name") 117 118 def test_get_builtin_constructor(self): 119 get_builtin_constructor = hashlib.__dict__[ 120 '__get_builtin_constructor'] 121 self.assertRaises(ValueError, get_builtin_constructor, 'test') 122 try: 123 import _md5 124 except ImportError: 125 pass 126 # This forces an ImportError for "import _md5" statements 127 sys.modules['_md5'] = None 128 try: 129 self.assertRaises(ValueError, get_builtin_constructor, 'md5') 130 finally: 131 if '_md5' in locals(): 132 sys.modules['_md5'] = _md5 133 else: 134 del sys.modules['_md5'] 135 136 def test_hexdigest(self): 137 for name in self.supported_hash_names: 138 h = hashlib.new(name) 139 self.assertTrue(hexstr(h.digest()) == h.hexdigest()) 140 141 def test_large_update(self): 142 aas = 'a' * 128 143 bees = 'b' * 127 144 cees = 'c' * 126 145 abcs = aas + bees + cees 146 147 for name in self.supported_hash_names: 148 m1 = hashlib.new(name) 149 m1.update(aas) 150 m1.update(bees) 151 m1.update(cees) 152 153 m2 = hashlib.new(name) 154 m2.update(abcs) 155 self.assertEqual(m1.digest(), m2.digest(), name+' update problem.') 156 157 m3 = hashlib.new(name, abcs) 158 self.assertEqual(m1.digest(), m3.digest(), name+' new problem.') 159 160 def check(self, name, data, digest): 161 constructors = self.constructors_to_test[name] 162 # 2 is for hashlib.name(...) and hashlib.new(name, ...) 163 self.assertGreaterEqual(len(constructors), 2) 164 for hash_object_constructor in constructors: 165 computed = hash_object_constructor(data).hexdigest() 166 self.assertEqual( 167 computed, digest, 168 "Hash algorithm %s constructed using %s returned hexdigest" 169 " %r for %d byte input data that should have hashed to %r." 170 % (name, hash_object_constructor, 171 computed, len(data), digest)) 172 173 def check_unicode(self, algorithm_name): 174 # Unicode objects are not allowed as input. 175 expected = hashlib.new(algorithm_name, str(u'spam')).hexdigest() 176 self.check(algorithm_name, u'spam', expected) 177 178 def test_unicode(self): 179 # In python 2.x unicode is auto-encoded to the system default encoding 180 # when passed to hashlib functions. 181 self.check_unicode('md5') 182 self.check_unicode('sha1') 183 self.check_unicode('sha224') 184 self.check_unicode('sha256') 185 self.check_unicode('sha384') 186 self.check_unicode('sha512') 187 188 def test_case_md5_0(self): 189 self.check('md5', '', 'd41d8cd98f00b204e9800998ecf8427e') 190 191 def test_case_md5_1(self): 192 self.check('md5', 'abc', '900150983cd24fb0d6963f7d28e17f72') 193 194 def test_case_md5_2(self): 195 self.check('md5', 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789', 196 'd174ab98d277d9f5a5611c2c9f419d9f') 197 198 @precisionbigmemtest(size=_4G + 5, memuse=1) 199 def test_case_md5_huge(self, size): 200 if size == _4G + 5: 201 try: 202 self.check('md5', 'A'*size, 'c9af2dff37468ce5dfee8f2cfc0a9c6d') 203 except OverflowError: 204 pass # 32-bit arch 205 206 @precisionbigmemtest(size=_4G - 1, memuse=1) 207 def test_case_md5_uintmax(self, size): 208 if size == _4G - 1: 209 try: 210 self.check('md5', 'A'*size, '28138d306ff1b8281f1a9067e1a1a2b3') 211 except OverflowError: 212 pass # 32-bit arch 213 214 # use the three examples from Federal Information Processing Standards 215 # Publication 180-1, Secure Hash Standard, 1995 April 17 216 # http://www.itl.nist.gov/div897/pubs/fip180-1.htm 217 218 def test_case_sha1_0(self): 219 self.check('sha1', "", 220 "da39a3ee5e6b4b0d3255bfef95601890afd80709") 221 222 def test_case_sha1_1(self): 223 self.check('sha1', "abc", 224 "a9993e364706816aba3e25717850c26c9cd0d89d") 225 226 def test_case_sha1_2(self): 227 self.check('sha1', "abcdbcdecdefdefgefghfghighijhijkijkljklmklmnlmnomnopnopq", 228 "84983e441c3bd26ebaae4aa1f95129e5e54670f1") 229 230 def test_case_sha1_3(self): 231 self.check('sha1', "a" * 1000000, 232 "34aa973cd4c4daa4f61eeb2bdbad27316534016f") 233 234 235 # use the examples from Federal Information Processing Standards 236 # Publication 180-2, Secure Hash Standard, 2002 August 1 237 # http://csrc.nist.gov/publications/fips/fips180-2/fips180-2.pdf 238 239 def test_case_sha224_0(self): 240 self.check('sha224', "", 241 "d14a028c2a3a2bc9476102bb288234c415a2b01f828ea62ac5b3e42f") 242 243 def test_case_sha224_1(self): 244 self.check('sha224', "abc", 245 "23097d223405d8228642a477bda255b32aadbce4bda0b3f7e36c9da7") 246 247 def test_case_sha224_2(self): 248 self.check('sha224', 249 "abcdbcdecdefdefgefghfghighijhijkijkljklmklmnlmnomnopnopq", 250 "75388b16512776cc5dba5da1fd890150b0c6455cb4f58b1952522525") 251 252 def test_case_sha224_3(self): 253 self.check('sha224', "a" * 1000000, 254 "20794655980c91d8bbb4c1ea97618a4bf03f42581948b2ee4ee7ad67") 255 256 257 def test_case_sha256_0(self): 258 self.check('sha256', "", 259 "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855") 260 261 def test_case_sha256_1(self): 262 self.check('sha256', "abc", 263 "ba7816bf8f01cfea414140de5dae2223b00361a396177a9cb410ff61f20015ad") 264 265 def test_case_sha256_2(self): 266 self.check('sha256', 267 "abcdbcdecdefdefgefghfghighijhijkijkljklmklmnlmnomnopnopq", 268 "248d6a61d20638b8e5c026930c3e6039a33ce45964ff2167f6ecedd419db06c1") 269 270 def test_case_sha256_3(self): 271 self.check('sha256', "a" * 1000000, 272 "cdc76e5c9914fb9281a1c7e284d73e67f1809a48a497200e046d39ccc7112cd0") 273 274 275 def test_case_sha384_0(self): 276 self.check('sha384', "", 277 "38b060a751ac96384cd9327eb1b1e36a21fdb71114be07434c0cc7bf63f6e1da"+ 278 "274edebfe76f65fbd51ad2f14898b95b") 279 280 def test_case_sha384_1(self): 281 self.check('sha384', "abc", 282 "cb00753f45a35e8bb5a03d699ac65007272c32ab0eded1631a8b605a43ff5bed"+ 283 "8086072ba1e7cc2358baeca134c825a7") 284 285 def test_case_sha384_2(self): 286 self.check('sha384', 287 "abcdefghbcdefghicdefghijdefghijkefghijklfghijklmghijklmn"+ 288 "hijklmnoijklmnopjklmnopqklmnopqrlmnopqrsmnopqrstnopqrstu", 289 "09330c33f71147e83d192fc782cd1b4753111b173b3b05d22fa08086e3b0f712"+ 290 "fcc7c71a557e2db966c3e9fa91746039") 291 292 def test_case_sha384_3(self): 293 self.check('sha384', "a" * 1000000, 294 "9d0e1809716474cb086e834e310a4a1ced149e9c00f248527972cec5704c2a5b"+ 295 "07b8b3dc38ecc4ebae97ddd87f3d8985") 296 297 298 def test_case_sha512_0(self): 299 self.check('sha512', "", 300 "cf83e1357eefb8bdf1542850d66d8007d620e4050b5715dc83f4a921d36ce9ce"+ 301 "47d0d13c5d85f2b0ff8318d2877eec2f63b931bd47417a81a538327af927da3e") 302 303 def test_case_sha512_1(self): 304 self.check('sha512', "abc", 305 "ddaf35a193617abacc417349ae20413112e6fa4e89a97ea20a9eeee64b55d39a"+ 306 "2192992a274fc1a836ba3c23a3feebbd454d4423643ce80e2a9ac94fa54ca49f") 307 308 def test_case_sha512_2(self): 309 self.check('sha512', 310 "abcdefghbcdefghicdefghijdefghijkefghijklfghijklmghijklmn"+ 311 "hijklmnoijklmnopjklmnopqklmnopqrlmnopqrsmnopqrstnopqrstu", 312 "8e959b75dae313da8cf4f72814fc143f8f7779c6eb9f7fa17299aeadb6889018"+ 313 "501d289e4900f7e4331b99dec4b5433ac7d329eeb6dd26545e96e55b874be909") 314 315 def test_case_sha512_3(self): 316 self.check('sha512', "a" * 1000000, 317 "e718483d0ce769644e2e42c7bc15b4638e1f98b13b2044285632a803afa973eb"+ 318 "de0ff244877ea60a4cb0432ce577c31beb009c5c2c49aa2e4eadb217ad8cc09b") 319 320 @unittest.skipUnless(threading, 'Threading required for this test.') 321 @test_support.reap_threads 322 def test_threaded_hashing(self): 323 # Updating the same hash object from several threads at once 324 # using data chunk sizes containing the same byte sequences. 325 # 326 # If the internal locks are working to prevent multiple 327 # updates on the same object from running at once, the resulting 328 # hash will be the same as doing it single threaded upfront. 329 hasher = hashlib.sha1() 330 num_threads = 5 331 smallest_data = 'swineflu' 332 data = smallest_data*200000 333 expected_hash = hashlib.sha1(data*num_threads).hexdigest() 334 335 def hash_in_chunks(chunk_size, event): 336 index = 0 337 while index < len(data): 338 hasher.update(data[index:index+chunk_size]) 339 index += chunk_size 340 event.set() 341 342 events = [] 343 for threadnum in xrange(num_threads): 344 chunk_size = len(data) // (10**threadnum) 345 assert chunk_size > 0 346 assert chunk_size % len(smallest_data) == 0 347 event = threading.Event() 348 events.append(event) 349 threading.Thread(target=hash_in_chunks, 350 args=(chunk_size, event)).start() 351 352 for event in events: 353 event.wait() 354 355 self.assertEqual(expected_hash, hasher.hexdigest()) 356 357 def test_main(): 358 test_support.run_unittest(HashLibTestCase) 359 360 if __name__ == "__main__": 361 test_main() 362