[204] | 1 | #!/usr/bin/env python |
---|
| 2 | |
---|
[210] | 3 | ## @package pyregfi |
---|
| 4 | # Python interface to the regfi library. |
---|
| 5 | # |
---|
| 6 | |
---|
[221] | 7 | ## @mainpage API Documentation |
---|
| 8 | # |
---|
| 9 | # The pyregfi module provides a Python interface to the @ref regfi Windows |
---|
| 10 | # registry library. |
---|
| 11 | # |
---|
| 12 | # The library operates on registry hives, each of which is contained within a |
---|
| 13 | # single file. To get started, one must first open the registry hive file with |
---|
| 14 | # the open() or file() Python built-in functions (or equivalent) and then pass |
---|
| 15 | # the resulting file object to pyregfi. For example: |
---|
| 16 | # @code |
---|
| 17 | # >>> import pyregfi |
---|
| 18 | # >>> fh = open('/mnt/win/c/WINDOWS/system32/config/system', 'rb') |
---|
| 19 | # >>> myHive = pyregfi.Hive(fh) |
---|
| 20 | # @endcode |
---|
| 21 | # |
---|
| 22 | # Using this Hive object, one can begin investigating what top-level keys |
---|
| 23 | # exist by starting with the root Key attribute: |
---|
| 24 | # @code |
---|
| 25 | # >>> for key in myHive.root.subkeys: |
---|
| 26 | # ... print(key.name) |
---|
| 27 | # ControlSet001 |
---|
| 28 | # ControlSet003 |
---|
| 29 | # LastKnownGoodRecovery |
---|
| 30 | # MountedDevices |
---|
| 31 | # Select |
---|
| 32 | # Setup |
---|
| 33 | # WPA |
---|
| 34 | # @endcode |
---|
| 35 | # |
---|
| 36 | # From there, accessing subkeys and values by name is a simple matter of: |
---|
| 37 | # @code |
---|
| 38 | # >>> myKey = myHive.root.subkeys['Select'] |
---|
| 39 | # >>> myValue = myKey.values['Current'] |
---|
| 40 | # @endcode |
---|
| 41 | # |
---|
| 42 | # The data associated with a Value can be obtained through the fetch_data() |
---|
| 43 | # method: |
---|
| 44 | # @code |
---|
| 45 | # >>> print(myValue.fetch_data()) |
---|
| 46 | # 1 |
---|
| 47 | # @endcode |
---|
| 48 | # |
---|
| 49 | # While useful for simple exercises, using the subkeys object for deeply nested |
---|
| 50 | # paths is not efficient and doesn't make for particularly attractive code. |
---|
| 51 | # Instead, a special-purpose HiveIterator class is provided for simplicity of |
---|
| 52 | # use and fast access to specific known paths: |
---|
| 53 | # @code |
---|
| 54 | # >>> myIter = pyregfi.HiveIterator(myHive) |
---|
| 55 | # >>> myIter.descend(['ControlSet001','Control','NetworkProvider','HwOrder']) |
---|
| 56 | # >>> myKey = myIter.current_key() |
---|
| 57 | # >>> print(myKey.values['ProviderOrder'].fetch_data()) |
---|
| 58 | # RDPNP,LanmanWorkstation,WebClient |
---|
| 59 | # @endcode |
---|
| 60 | # |
---|
| 61 | # The first two lines above can be simplified in some "syntactic sugar" provided |
---|
| 62 | # by the Hive.subtree() method. Also, as one might expect, the HiveIterator |
---|
| 63 | # also acts as an iterator, producing keys in a depth-first order. |
---|
| 64 | # For instance, to traverse all keys under the ControlSet003\\Services key, |
---|
| 65 | # printing their names as we go, we could do: |
---|
| 66 | # @code |
---|
| 67 | # >>> for key in Hive.subtree(['ControlSet003','Services']): |
---|
| 68 | # >>> print(key.name) |
---|
| 69 | # Services |
---|
| 70 | # Abiosdsk |
---|
| 71 | # abp480n5 |
---|
| 72 | # Parameters |
---|
| 73 | # PnpInterface |
---|
| 74 | # ACPI |
---|
| 75 | # [...] |
---|
| 76 | # @endcode |
---|
| 77 | # |
---|
| 78 | # Note that "Services" was printed first, since the subtree is traversed as a |
---|
| 79 | # "preordering depth-first" search starting with the HiveIterator's current_key(). |
---|
| 80 | # As one might expect, traversals of subtrees stops when all elements in a |
---|
| 81 | # specific subtree (and none outside of it) have been traversed. |
---|
| 82 | # |
---|
| 83 | # For more information, peruse the various attributes and methods available on |
---|
| 84 | # the Hive, HiveIterator, Key, Value, and Security classes. |
---|
| 85 | # |
---|
| 86 | # @note @ref regfi is a read-only library by design and there |
---|
| 87 | # are no plans to implement write support. |
---|
| 88 | # |
---|
| 89 | # @note At present, pyregfi has been tested with Python versions 2.6 and 3.1 |
---|
| 90 | # |
---|
| 91 | # @note Developers strive to make pyregfi thread-safe. |
---|
| 92 | # |
---|
| 93 | # @note Key and Value names are case-sensitive in regfi and pyregfi |
---|
| 94 | # |
---|
[204] | 95 | import sys |
---|
[219] | 96 | import time |
---|
[204] | 97 | from pyregfi.structures import * |
---|
| 98 | |
---|
| 99 | import ctypes |
---|
| 100 | import ctypes.util |
---|
| 101 | |
---|
[221] | 102 | ## An enumeration of registry Value data types |
---|
[210] | 103 | # |
---|
[221] | 104 | # @note This is a static class, there is no need to instantiate it. |
---|
| 105 | # Just access its attributes directly as DATA_TYPES.SZ, etc |
---|
| 106 | class DATA_TYPES(object): |
---|
| 107 | ## None / Unknown |
---|
| 108 | NONE = 0 |
---|
| 109 | ## String |
---|
| 110 | SZ = 1 |
---|
| 111 | ## String with %...% expansions |
---|
| 112 | EXPAND_SZ = 2 |
---|
| 113 | ## Binary buffer |
---|
| 114 | BINARY = 3 |
---|
| 115 | ## 32 bit integer (little endian) |
---|
| 116 | DWORD = 4 # DWORD, little endian |
---|
| 117 | ## 32 bit integer (little endian) |
---|
| 118 | DWORD_LE = 4 |
---|
| 119 | ## 32 bit integer (big endian) |
---|
| 120 | DWORD_BE = 5 # DWORD, big endian |
---|
| 121 | ## Symbolic link |
---|
| 122 | LINK = 6 |
---|
| 123 | ## List of strings |
---|
| 124 | MULTI_SZ = 7 |
---|
| 125 | ## Unknown structure |
---|
| 126 | RESOURCE_LIST = 8 |
---|
| 127 | ## Unknown structure |
---|
| 128 | FULL_RESOURCE_DESCRIPTOR = 9 |
---|
| 129 | ## Unknown structure |
---|
| 130 | RESOURCE_REQUIREMENTS_LIST = 10 |
---|
| 131 | ## 64 bit integer |
---|
| 132 | QWORD = 11 # 64-bit little endian |
---|
[205] | 133 | |
---|
| 134 | |
---|
[208] | 135 | def _buffer2bytearray(char_pointer, length): |
---|
| 136 | if length == 0 or char_pointer == None: |
---|
| 137 | return None |
---|
| 138 | |
---|
| 139 | ret_val = bytearray(length) |
---|
| 140 | for i in range(0,length): |
---|
| 141 | ret_val[i] = char_pointer[i][0] |
---|
| 142 | |
---|
| 143 | return ret_val |
---|
| 144 | |
---|
| 145 | |
---|
[215] | 146 | def _strlist2charss(str_list): |
---|
| 147 | ret_val = [] |
---|
| 148 | for s in str_list: |
---|
| 149 | ret_val.append(s.encode('utf-8', 'replace')) |
---|
| 150 | |
---|
[220] | 151 | ret_val = (ctypes.c_char_p*(len(str_list)+1))(*ret_val) |
---|
[215] | 152 | # Terminate the char** with a NULL pointer |
---|
| 153 | ret_val[-1] = 0 |
---|
| 154 | |
---|
| 155 | return ret_val |
---|
| 156 | |
---|
| 157 | |
---|
[209] | 158 | def _charss2strlist(chars_pointer): |
---|
| 159 | ret_val = [] |
---|
| 160 | i = 0 |
---|
| 161 | s = chars_pointer[i] |
---|
| 162 | while s != None: |
---|
[213] | 163 | ret_val.append(s.decode('utf-8', 'replace')) |
---|
[209] | 164 | i += 1 |
---|
| 165 | s = chars_pointer[i] |
---|
[208] | 166 | |
---|
[209] | 167 | return ret_val |
---|
[208] | 168 | |
---|
[210] | 169 | |
---|
[221] | 170 | ## Retrieves messages produced by regfi during parsing and interpretation |
---|
| 171 | # |
---|
| 172 | # The regfi C library may generate log messages stored in a special thread-safe |
---|
| 173 | # global data structure. These messages should be retrieved periodically or |
---|
| 174 | # after each major operation by callers to determine if any errors or warnings |
---|
| 175 | # should be reported to the user. Failure to retrieve these could result in |
---|
| 176 | # excessive memory consumption. |
---|
| 177 | def GetLogMessages(): |
---|
| 178 | msgs = regfi.regfi_log_get_str() |
---|
| 179 | if msgs == None: |
---|
| 180 | return '' |
---|
| 181 | return msgs.decode('utf-8') |
---|
| 182 | |
---|
| 183 | |
---|
| 184 | ## Abstract class for most objects returned by the library |
---|
[212] | 185 | class _StructureWrapper(object): |
---|
[214] | 186 | _hive = None |
---|
| 187 | _base = None |
---|
[206] | 188 | |
---|
[207] | 189 | def __init__(self, hive, base): |
---|
[215] | 190 | if not hive: |
---|
| 191 | raise Exception("Could not create _StructureWrapper," |
---|
| 192 | + " hive is NULL. Current log:\n" |
---|
| 193 | + GetLogMessages()) |
---|
| 194 | if not base: |
---|
| 195 | raise Exception("Could not create _StructureWrapper," |
---|
| 196 | + " base is NULL. Current log:\n" |
---|
| 197 | + GetLogMessages()) |
---|
[214] | 198 | self._hive = hive |
---|
| 199 | self._base = base |
---|
[206] | 200 | |
---|
[224] | 201 | |
---|
[221] | 202 | # Memory management for most regfi structures is taken care of here |
---|
[206] | 203 | def __del__(self): |
---|
[214] | 204 | regfi.regfi_free_record(self._base) |
---|
[206] | 205 | |
---|
[224] | 206 | |
---|
[221] | 207 | # Any attribute requests not explicitly defined in subclasses gets passed |
---|
| 208 | # to the equivalent REGFI_* structure defined in structures.py |
---|
[206] | 209 | def __getattr__(self, name): |
---|
[214] | 210 | return getattr(self._base.contents, name) |
---|
[224] | 211 | |
---|
[221] | 212 | |
---|
| 213 | ## Test for equality |
---|
| 214 | # |
---|
| 215 | # Records returned by pyregfi may be compared with one another. For example: |
---|
| 216 | # @code |
---|
| 217 | # >>> key2 = key1.subkeys['child'] |
---|
| 218 | # >>> key1 == key2 |
---|
| 219 | # False |
---|
| 220 | # >>> key1 != key2 |
---|
| 221 | # True |
---|
| 222 | # >>> key1 == key2.get_parent() |
---|
| 223 | # True |
---|
| 224 | # @endcode |
---|
[206] | 225 | def __eq__(self, other): |
---|
| 226 | return (type(self) == type(other)) and (self.offset == other.offset) |
---|
| 227 | |
---|
[224] | 228 | |
---|
[206] | 229 | def __ne__(self, other): |
---|
| 230 | return (not self.__eq__(other)) |
---|
| 231 | |
---|
[208] | 232 | |
---|
[221] | 233 | class Key(): |
---|
[206] | 234 | pass |
---|
| 235 | |
---|
[221] | 236 | |
---|
| 237 | class Value(): |
---|
[206] | 238 | pass |
---|
| 239 | |
---|
[221] | 240 | |
---|
| 241 | ## Registry security record and descriptor |
---|
| 242 | # XXX: Access to security descriptors not yet implemented |
---|
[206] | 243 | class Security(_StructureWrapper): |
---|
| 244 | pass |
---|
| 245 | |
---|
[221] | 246 | ## Abstract class for ValueList and SubkeyList |
---|
[212] | 247 | class _GenericList(object): |
---|
[214] | 248 | _hive = None |
---|
[224] | 249 | _key_base = None |
---|
[214] | 250 | _length = None |
---|
| 251 | _current = None |
---|
[207] | 252 | |
---|
[221] | 253 | # implementation-specific functions for SubkeyList and ValueList |
---|
[214] | 254 | _fetch_num = None |
---|
| 255 | _find_element = None |
---|
| 256 | _get_element = None |
---|
| 257 | _constructor = None |
---|
[208] | 258 | |
---|
[207] | 259 | def __init__(self, key): |
---|
[224] | 260 | if not key: |
---|
| 261 | raise Exception("Could not create _GenericList; key is NULL." |
---|
| 262 | + "Current log:\n" + GetLogMessages()) |
---|
| 263 | |
---|
| 264 | if not regfi.regfi_reference_record(key._base): |
---|
| 265 | raise Exception("Could not create _GenericList; memory error." |
---|
| 266 | + "Current log:\n" + GetLogMessages()) |
---|
| 267 | self._key_base = key._base |
---|
| 268 | self._length = self._fetch_num(self._key_base) |
---|
[214] | 269 | self._hive = key._hive |
---|
| 270 | |
---|
[207] | 271 | |
---|
[224] | 272 | def __del__(self): |
---|
| 273 | regfi.regfi_free_record(self._key_base) |
---|
[221] | 274 | |
---|
[224] | 275 | |
---|
[221] | 276 | ## Length of list |
---|
[207] | 277 | def __len__(self): |
---|
[214] | 278 | return self._length |
---|
[207] | 279 | |
---|
[221] | 280 | |
---|
| 281 | ## Retrieves a list element by name |
---|
| 282 | # |
---|
| 283 | # @return the first element whose name matches, or None if the element |
---|
| 284 | # could not be found |
---|
[207] | 285 | def __getitem__(self, name): |
---|
[220] | 286 | index = ctypes.c_uint32() |
---|
[208] | 287 | if isinstance(name, str): |
---|
| 288 | name = name.encode('utf-8') |
---|
| 289 | |
---|
[209] | 290 | if name != None: |
---|
| 291 | name = create_string_buffer(bytes(name)) |
---|
| 292 | |
---|
[224] | 293 | if self._find_element(self._hive.file, self._key_base, |
---|
[220] | 294 | name, byref(index)): |
---|
| 295 | return self._constructor(self._hive, |
---|
[214] | 296 | self._get_element(self._hive.file, |
---|
[224] | 297 | self._key_base, |
---|
[214] | 298 | index)) |
---|
[207] | 299 | raise KeyError('') |
---|
| 300 | |
---|
[209] | 301 | def get(self, name, default): |
---|
| 302 | try: |
---|
| 303 | return self[name] |
---|
| 304 | except KeyError: |
---|
| 305 | return default |
---|
| 306 | |
---|
[207] | 307 | def __iter__(self): |
---|
[214] | 308 | self._current = 0 |
---|
[207] | 309 | return self |
---|
| 310 | |
---|
| 311 | def __next__(self): |
---|
[214] | 312 | if self._current >= self._length: |
---|
[207] | 313 | raise StopIteration('') |
---|
| 314 | |
---|
[224] | 315 | elem = self._get_element(self._hive.file, self._key_base, |
---|
[220] | 316 | ctypes.c_uint32(self._current)) |
---|
[214] | 317 | self._current += 1 |
---|
| 318 | return self._constructor(self._hive, elem) |
---|
[207] | 319 | |
---|
[212] | 320 | # For Python 2.x |
---|
[214] | 321 | next = __next__ |
---|
[207] | 322 | |
---|
[212] | 323 | |
---|
[221] | 324 | ## The list of subkeys associated with a Key |
---|
| 325 | # |
---|
| 326 | # This attribute is both iterable: |
---|
| 327 | # @code |
---|
| 328 | # for k in myKey.subkeys: |
---|
| 329 | # ... |
---|
| 330 | # @endcode |
---|
| 331 | # and accessible as a dictionary: |
---|
| 332 | # @code |
---|
| 333 | # mySubkey = myKey.subkeys["keyName"] |
---|
| 334 | # @endcode |
---|
| 335 | # |
---|
| 336 | # @note SubkeyLists should never be accessed directly and only exist |
---|
| 337 | # in association with a parent Key object. Do not retain references to |
---|
| 338 | # SubkeyLists. Instead, access them via their parent Key at all times. |
---|
| 339 | class SubkeyList(_GenericList): |
---|
[214] | 340 | _fetch_num = regfi.regfi_fetch_num_subkeys |
---|
| 341 | _find_element = regfi.regfi_find_subkey |
---|
| 342 | _get_element = regfi.regfi_get_subkey |
---|
[208] | 343 | |
---|
| 344 | |
---|
[221] | 345 | ## The list of values associated with a Key |
---|
| 346 | # |
---|
| 347 | # This attribute is both iterable: |
---|
| 348 | # @code |
---|
| 349 | # for v in myKey.values: |
---|
| 350 | # ... |
---|
| 351 | # @endcode |
---|
| 352 | # and accessible as a dictionary: |
---|
| 353 | # @code |
---|
| 354 | # myValue = myKey.values["valueName"] |
---|
| 355 | # @endcode |
---|
| 356 | # |
---|
| 357 | # @note ValueLists should never be accessed directly and only exist |
---|
| 358 | # in association with a parent Key object. Do not retain references to |
---|
| 359 | # ValueLists. Instead, access them via their parent Key at all times. |
---|
| 360 | class ValueList(_GenericList): |
---|
[214] | 361 | _fetch_num = regfi.regfi_fetch_num_values |
---|
| 362 | _find_element = regfi.regfi_find_value |
---|
| 363 | _get_element = regfi.regfi_get_value |
---|
[208] | 364 | |
---|
| 365 | |
---|
[215] | 366 | ## Registry key |
---|
[221] | 367 | # These represent registry keys (@ref REGFI_NK records) and provide |
---|
| 368 | # access to their subkeys, values, and other metadata. |
---|
| 369 | # |
---|
| 370 | # @note Value instances may provide access to more than the attributes |
---|
| 371 | # documented here. However, undocumented attributes may change over time |
---|
| 372 | # and are not officially supported. If you need access to an attribute |
---|
| 373 | # not shown here, see pyregfi.structures. |
---|
[207] | 374 | class Key(_StructureWrapper): |
---|
[221] | 375 | ## A @ref ValueList object representing the list of Values |
---|
| 376 | # stored on this Key |
---|
[207] | 377 | values = None |
---|
[221] | 378 | |
---|
| 379 | ## A @ref SubkeyList object representing the list of subkeys |
---|
| 380 | # stored on this Key |
---|
[208] | 381 | subkeys = None |
---|
[207] | 382 | |
---|
[221] | 383 | ## The raw Key name as an uninterpreted bytearray |
---|
| 384 | name_raw = (b"...") |
---|
| 385 | |
---|
| 386 | ## The name of the Key as a (unicode) string |
---|
| 387 | name = "..." |
---|
| 388 | |
---|
| 389 | ## The absolute file offset of the Key record's cell in the Hive file |
---|
| 390 | offset = 0xCAFEBABE |
---|
| 391 | |
---|
| 392 | ## This Key's last modified time represented as the number of seconds |
---|
| 393 | # since the UNIX epoch in UTC; similar to what time.time() returns |
---|
| 394 | modified = 1300000000.123456 |
---|
| 395 | |
---|
| 396 | ## The NK record's flags field |
---|
| 397 | flags = 0x10110001 |
---|
| 398 | |
---|
[207] | 399 | def __init__(self, hive, base): |
---|
| 400 | super(Key, self).__init__(hive, base) |
---|
[221] | 401 | self.values = ValueList(self) |
---|
| 402 | self.subkeys = SubkeyList(self) |
---|
[207] | 403 | |
---|
[208] | 404 | def __getattr__(self, name): |
---|
| 405 | if name == "name": |
---|
[219] | 406 | ret_val = super(Key, self).__getattr__(name) |
---|
| 407 | |
---|
[209] | 408 | if ret_val == None: |
---|
| 409 | ret_val = self.name_raw |
---|
| 410 | else: |
---|
[213] | 411 | ret_val = ret_val.decode('utf-8', 'replace') |
---|
[209] | 412 | |
---|
[208] | 413 | elif name == "name_raw": |
---|
[219] | 414 | ret_val = super(Key, self).__getattr__(name) |
---|
[208] | 415 | length = super(Key, self).__getattr__('name_length') |
---|
| 416 | ret_val = _buffer2bytearray(ret_val, length) |
---|
| 417 | |
---|
[219] | 418 | elif name == "modified": |
---|
| 419 | ret_val = regfi.regfi_nt2unix_time(byref(self._base.contents.mtime)) |
---|
| 420 | |
---|
| 421 | else: |
---|
| 422 | ret_val = super(Key, self).__getattr__(name) |
---|
| 423 | |
---|
[208] | 424 | return ret_val |
---|
| 425 | |
---|
[221] | 426 | |
---|
| 427 | ## Retrieves the Security properties for this key |
---|
[207] | 428 | def fetch_security(self): |
---|
[214] | 429 | return Security(self._hive, |
---|
[215] | 430 | regfi.regfi_fetch_sk(self._hive.file, self._base)) |
---|
[207] | 431 | |
---|
[221] | 432 | |
---|
| 433 | ## Retrieves the class name for this key |
---|
| 434 | # |
---|
| 435 | # Class names are typically stored as UTF-16LE strings, so these are decoded |
---|
| 436 | # into proper python (unicode) strings. However, if this fails, a bytearray |
---|
| 437 | # is instead returned containing the raw buffer stored for the class name. |
---|
| 438 | # |
---|
| 439 | # @return The class name as a string or bytearray. None if a class name |
---|
| 440 | # doesn't exist or an unrecoverable error occurred during retrieval. |
---|
[219] | 441 | def fetch_classname(self): |
---|
| 442 | ret_val = None |
---|
| 443 | cn_p = regfi.regfi_fetch_classname(self._hive.file, self._base) |
---|
| 444 | if cn_p: |
---|
| 445 | cn_struct = cn_p.contents |
---|
| 446 | if cn_struct.interpreted: |
---|
| 447 | ret_val = cn_struct.interpreted.decode('utf-8', 'replace') |
---|
| 448 | else: |
---|
| 449 | ret_val = _buffer2bytearray(cn_struct.raw, |
---|
| 450 | cn_struct.size) |
---|
| 451 | regfi.regfi_free_record(cn_p) |
---|
| 452 | |
---|
| 453 | return ret_val |
---|
| 454 | |
---|
[221] | 455 | |
---|
| 456 | ## Retrieves this key's parent key |
---|
| 457 | # |
---|
| 458 | # @return The parent's Key instance or None if current key is root |
---|
| 459 | # (or an error occured) |
---|
[215] | 460 | def get_parent(self): |
---|
[218] | 461 | if self.is_root(): |
---|
| 462 | return None |
---|
[215] | 463 | parent_base = regfi.regfi_get_parentkey(self._hive.file, self._base) |
---|
| 464 | if parent_base: |
---|
| 465 | return Key(self._hive, parent_base) |
---|
| 466 | return None |
---|
| 467 | |
---|
| 468 | def is_root(self): |
---|
[218] | 469 | return (self._hive.root == self) |
---|
[215] | 470 | |
---|
| 471 | |
---|
[210] | 472 | ## Registry value (metadata) |
---|
| 473 | # |
---|
| 474 | # These represent registry values (@ref REGFI_VK records) and provide |
---|
| 475 | # access to their associated data. |
---|
[221] | 476 | # |
---|
| 477 | # @note Value instances may provide access to more than the attributes |
---|
| 478 | # documented here. However, undocumented attributes may change over time |
---|
| 479 | # and are not officially supported. If you need access to an attribute |
---|
| 480 | # not shown here, see pyregfi.structures. |
---|
[208] | 481 | class Value(_StructureWrapper): |
---|
[221] | 482 | ## The raw Value name as an uninterpreted bytearray |
---|
| 483 | name_raw = (b"...") |
---|
| 484 | |
---|
| 485 | ## The name of the Value as a (unicode) string |
---|
| 486 | name = "..." |
---|
| 487 | |
---|
| 488 | ## The absolute file offset of the Value record's cell in the Hive file |
---|
| 489 | offset = 0xCAFEBABE |
---|
| 490 | |
---|
| 491 | ## The length of data advertised in the VK record |
---|
| 492 | data_size = 0xCAFEBABE |
---|
| 493 | |
---|
| 494 | ## An integer which represents the data type for this Value's data |
---|
| 495 | # Typically this value is one of 12 types defined in @ref DATA_TYPES, |
---|
| 496 | # but in some cases (the SAM hive) it may be used for other purposes |
---|
| 497 | type = DATA_TYPES.NONE |
---|
| 498 | |
---|
| 499 | ## The VK record's flags field |
---|
| 500 | flags = 0x10110001 |
---|
| 501 | |
---|
| 502 | ## Retrieves the Value's data according to advertised type |
---|
| 503 | # |
---|
| 504 | # Data is loaded from its cell(s) and then interpreted based on the data |
---|
| 505 | # type recorded in the Value. It is not uncommon for data to be stored with |
---|
| 506 | # the wrong type or even with invalid types. If you have difficulty |
---|
| 507 | # obtaining desired data here, use @ref fetch_raw_data(). |
---|
| 508 | # |
---|
| 509 | # @return The interpreted representation of the data as one of several |
---|
| 510 | # possible Python types, as listed below. None if any failure |
---|
| 511 | # occurred during extraction or conversion. |
---|
| 512 | # |
---|
| 513 | # @retval string for SZ, EXPAND_SZ, and LINK |
---|
| 514 | # @retval int for DWORD, DWORD_BE, and QWORD |
---|
| 515 | # @retval list(string) for MULTI_SZ |
---|
| 516 | # @retval bytearray for NONE, BINARY, RESOURCE_LIST, |
---|
| 517 | # FULL_RESOURCE_DESCRIPTOR, and RESOURCE_REQUIREMENTS_LIST |
---|
| 518 | # |
---|
[219] | 519 | def fetch_data(self): |
---|
[209] | 520 | ret_val = None |
---|
[219] | 521 | data_p = regfi.regfi_fetch_data(self._hive.file, self._base) |
---|
| 522 | if not data_p: |
---|
| 523 | return None |
---|
| 524 | data_struct = data_p.contents |
---|
[208] | 525 | |
---|
[219] | 526 | if data_struct.interpreted_size == 0: |
---|
| 527 | ret_val = None |
---|
[221] | 528 | elif data_struct.type in (DATA_TYPES.SZ, DATA_TYPES.EXPAND_SZ, DATA_TYPES.LINK): |
---|
[219] | 529 | # Unicode strings |
---|
| 530 | ret_val = data_struct.interpreted.string.decode('utf-8', 'replace') |
---|
[221] | 531 | elif data_struct.type in (DATA_TYPES.DWORD, DATA_TYPES.DWORD_BE): |
---|
[219] | 532 | # 32 bit integers |
---|
| 533 | ret_val = data_struct.interpreted.dword |
---|
[221] | 534 | elif data_struct.type == DATA_TYPES.QWORD: |
---|
[219] | 535 | # 64 bit integers |
---|
| 536 | ret_val = data_struct.interpreted.qword |
---|
[221] | 537 | elif data_struct.type == DATA_TYPES.MULTI_SZ: |
---|
[219] | 538 | ret_val = _charss2strlist(data_struct.interpreted.multiple_string) |
---|
[221] | 539 | elif data_struct.type in (DATA_TYPES.NONE, DATA_TYPES.RESOURCE_LIST, |
---|
| 540 | DATA_TYPES.FULL_RESOURCE_DESCRIPTOR, |
---|
| 541 | DATA_TYPES.RESOURCE_REQUIREMENTS_LIST, |
---|
| 542 | DATA_TYPES.BINARY): |
---|
[219] | 543 | ret_val = _buffer2bytearray(data_struct.interpreted.none, |
---|
| 544 | data_struct.interpreted_size) |
---|
[209] | 545 | |
---|
[219] | 546 | regfi.regfi_free_record(data_p) |
---|
| 547 | return ret_val |
---|
[221] | 548 | |
---|
| 549 | |
---|
| 550 | ## Retrieves raw representation of Value's data |
---|
| 551 | # |
---|
| 552 | # @return A bytearray containing the data |
---|
| 553 | # |
---|
[219] | 554 | def fetch_raw_data(self): |
---|
| 555 | ret_val = None |
---|
| 556 | # XXX: should we load the data without interpretation instead? |
---|
| 557 | data_p = regfi.regfi_fetch_data(self._hive.file, self._base) |
---|
| 558 | if not data_p: |
---|
| 559 | return None |
---|
[209] | 560 | |
---|
[219] | 561 | data_struct = data_p.contents |
---|
| 562 | ret_val = _buffer2bytearray(data_struct.raw, |
---|
| 563 | data_struct.size) |
---|
| 564 | regfi.regfi_free_record(data_p) |
---|
[208] | 565 | return ret_val |
---|
| 566 | |
---|
[221] | 567 | |
---|
[219] | 568 | def __getattr__(self, name): |
---|
| 569 | ret_val = super(Value, self).__getattr__(name) |
---|
| 570 | if name == "name": |
---|
| 571 | if ret_val == None: |
---|
| 572 | ret_val = self.name_raw |
---|
| 573 | else: |
---|
| 574 | ret_val = ret_val.decode('utf-8', 'replace') |
---|
[208] | 575 | |
---|
[219] | 576 | elif name == "name_raw": |
---|
| 577 | length = super(Value, self).__getattr__('name_length') |
---|
| 578 | ret_val = _buffer2bytearray(ret_val, length) |
---|
| 579 | |
---|
| 580 | return ret_val |
---|
| 581 | |
---|
| 582 | |
---|
[208] | 583 | # Avoids chicken/egg class definitions. |
---|
| 584 | # Also makes for convenient code reuse in these lists' parent classes. |
---|
[221] | 585 | SubkeyList._constructor = Key |
---|
| 586 | ValueList._constructor = Value |
---|
[208] | 587 | |
---|
| 588 | |
---|
| 589 | |
---|
[210] | 590 | ## Represents a single registry hive (file) |
---|
| 591 | class Hive(): |
---|
[204] | 592 | file = None |
---|
| 593 | raw_file = None |
---|
[218] | 594 | _root = None |
---|
| 595 | |
---|
[221] | 596 | ## The root Key of this Hive |
---|
| 597 | root = None |
---|
| 598 | |
---|
| 599 | ## This Hives's last modified time represented as the number of seconds |
---|
| 600 | # since the UNIX epoch in UTC; similar to what time.time() returns |
---|
| 601 | modified = 1300000000.123456 |
---|
| 602 | |
---|
| 603 | ## First sequence number |
---|
| 604 | sequence1 = 12345678 |
---|
| 605 | |
---|
| 606 | ## Second sequence number |
---|
| 607 | sequence2 = 12345678 |
---|
| 608 | |
---|
| 609 | ## Major version |
---|
| 610 | major_version = 1 |
---|
| 611 | |
---|
| 612 | ## Minor version |
---|
| 613 | minor_version = 5 |
---|
| 614 | |
---|
| 615 | # XXX: Possibly add a second or factory function which opens a |
---|
| 616 | # hive file for you |
---|
| 617 | |
---|
| 618 | ## Constructor |
---|
| 619 | # |
---|
| 620 | # @param fh A Python file object. The constructor first looks for a valid |
---|
| 621 | # fileno attribute on this object and uses it if possible. |
---|
| 622 | # Otherwise, the seek and read methods are used for file |
---|
| 623 | # access. |
---|
| 624 | # |
---|
| 625 | # @note Supplied file must be seekable |
---|
[204] | 626 | def __init__(self, fh): |
---|
[205] | 627 | try: |
---|
[221] | 628 | # The fileno method may not exist, or it may throw an exception |
---|
| 629 | # when called if the file isn't backed with a descriptor. |
---|
[205] | 630 | if hasattr(fh, 'fileno'): |
---|
[213] | 631 | self.file = regfi.regfi_alloc(fh.fileno(), REGFI_ENCODING_UTF8) |
---|
[205] | 632 | return |
---|
| 633 | except: |
---|
| 634 | pass |
---|
[204] | 635 | |
---|
[225] | 636 | fh.seek(0) |
---|
[205] | 637 | self.raw_file = structures.REGFI_RAW_FILE() |
---|
| 638 | self.raw_file.fh = fh |
---|
| 639 | self.raw_file.seek = seek_cb_type(self.raw_file.cb_seek) |
---|
| 640 | self.raw_file.read = read_cb_type(self.raw_file.cb_read) |
---|
[225] | 641 | self.file = regfi.regfi_alloc_cb(pointer(self.raw_file), REGFI_ENCODING_UTF8) |
---|
| 642 | if not self.file: |
---|
| 643 | # XXX: switch to non-generic exception |
---|
| 644 | raise Exception("Could not open registry file. Current log:\n" |
---|
| 645 | + GetLogMessages()) |
---|
[204] | 646 | |
---|
| 647 | def __getattr__(self, name): |
---|
[218] | 648 | if name == "root": |
---|
[224] | 649 | # XXX: This creates reference loops. Need to cache better inside regfi |
---|
| 650 | #if self._root == None: |
---|
| 651 | # self._root = Key(self, regfi.regfi_get_rootkey(self.file)) |
---|
| 652 | #return self._root |
---|
| 653 | return Key(self, regfi.regfi_get_rootkey(self.file)) |
---|
[218] | 654 | |
---|
[221] | 655 | elif name == "modified": |
---|
| 656 | return regfi.regfi_nt2unix_time(byref(self._base.contents.mtime)) |
---|
| 657 | |
---|
[204] | 658 | return getattr(self.file.contents, name) |
---|
[221] | 659 | |
---|
[205] | 660 | |
---|
[210] | 661 | def __del__(self): |
---|
[205] | 662 | regfi.regfi_free(self.file) |
---|
| 663 | if self.raw_file != None: |
---|
[213] | 664 | self.raw_file = None |
---|
[204] | 665 | |
---|
[221] | 666 | |
---|
[205] | 667 | def __iter__(self): |
---|
| 668 | return HiveIterator(self) |
---|
[204] | 669 | |
---|
[215] | 670 | |
---|
[210] | 671 | ## Creates a @ref HiveIterator initialized at the specified path in |
---|
[221] | 672 | # the hive. |
---|
[210] | 673 | # |
---|
[221] | 674 | # @param path A list of Key names which represent an absolute path within |
---|
| 675 | # the Hive |
---|
| 676 | # |
---|
| 677 | # @return A @ref HiveIterator which is positioned at the specified path. |
---|
| 678 | # |
---|
| 679 | # @exception Exception If the path could not be found/traversed |
---|
[206] | 680 | def subtree(self, path): |
---|
| 681 | hi = HiveIterator(self) |
---|
| 682 | hi.descend(path) |
---|
| 683 | return hi |
---|
[205] | 684 | |
---|
[206] | 685 | |
---|
[210] | 686 | ## A special purpose iterator for registry hives |
---|
| 687 | # |
---|
| 688 | # Iterating over an object of this type causes all keys in a specific |
---|
| 689 | # hive subtree to be returned in a depth-first manner. These iterators |
---|
| 690 | # are typically created using the @ref Hive.subtree() function on a @ref Hive |
---|
| 691 | # object. |
---|
| 692 | # |
---|
| 693 | # HiveIterators can also be used to manually traverse up and down a |
---|
| 694 | # registry hive as they retain information about the current position in |
---|
| 695 | # the hive, along with which iteration state for subkeys and values for |
---|
| 696 | # every parent key. See the @ref up and @ref down methods for more |
---|
| 697 | # information. |
---|
[205] | 698 | class HiveIterator(): |
---|
[220] | 699 | _hive = None |
---|
| 700 | _iter = None |
---|
| 701 | _iteration_root = None |
---|
[205] | 702 | |
---|
| 703 | def __init__(self, hive): |
---|
[220] | 704 | self._iter = regfi.regfi_iterator_new(hive.file, REGFI_ENCODING_UTF8) |
---|
| 705 | if self._iter == None: |
---|
[205] | 706 | raise Exception("Could not create iterator. Current log:\n" |
---|
| 707 | + GetLogMessages()) |
---|
[214] | 708 | self._hive = hive |
---|
[205] | 709 | |
---|
| 710 | def __getattr__(self, name): |
---|
| 711 | return getattr(self.file.contents, name) |
---|
| 712 | |
---|
| 713 | def __del__(self): |
---|
[220] | 714 | regfi.regfi_iterator_free(self._iter) |
---|
[205] | 715 | |
---|
| 716 | def __iter__(self): |
---|
[220] | 717 | self._iteration_root = None |
---|
[205] | 718 | return self |
---|
| 719 | |
---|
| 720 | def __next__(self): |
---|
[220] | 721 | if self._iteration_root == None: |
---|
| 722 | self._iteration_root = self.current_key() |
---|
| 723 | elif not regfi.regfi_iterator_down(self._iter): |
---|
| 724 | up_ret = regfi.regfi_iterator_up(self._iter) |
---|
[206] | 725 | while (up_ret and |
---|
[220] | 726 | not regfi.regfi_iterator_next_subkey(self._iter)): |
---|
| 727 | if self._iteration_root == self.current_key(): |
---|
| 728 | self._iteration_root = None |
---|
[206] | 729 | raise StopIteration('') |
---|
[220] | 730 | up_ret = regfi.regfi_iterator_up(self._iter) |
---|
[205] | 731 | |
---|
| 732 | if not up_ret: |
---|
[221] | 733 | self._iteration_root = None |
---|
[205] | 734 | raise StopIteration('') |
---|
| 735 | |
---|
[210] | 736 | # XXX: Use non-generic exception |
---|
[220] | 737 | if not regfi.regfi_iterator_down(self._iter): |
---|
[205] | 738 | raise Exception('Error traversing iterator downward.'+ |
---|
| 739 | ' Current log:\n'+ GetLogMessages()) |
---|
| 740 | |
---|
[220] | 741 | regfi.regfi_iterator_first_subkey(self._iter) |
---|
[206] | 742 | return self.current_key() |
---|
[205] | 743 | |
---|
[212] | 744 | # For Python 2.x |
---|
[214] | 745 | next = __next__ |
---|
[212] | 746 | |
---|
[221] | 747 | # XXX: Should add sanity checks on some of these traversal functions |
---|
| 748 | # to throw exceptions if a traversal/retrieval *should* have worked |
---|
| 749 | # but failed for some reason. |
---|
| 750 | |
---|
| 751 | ## Descends the iterator to a subkey |
---|
| 752 | # |
---|
| 753 | # Descends the iterator one level to the current subkey, or a subkey |
---|
| 754 | # specified by name. |
---|
| 755 | # |
---|
| 756 | # @param subkey_name If specified, locates specified subkey by name |
---|
| 757 | # (via find_subkey()) and descends to it. |
---|
| 758 | # |
---|
| 759 | # @return True if successful, False otherwise |
---|
[220] | 760 | def down(self, subkey_name=None): |
---|
| 761 | if subkey_name == None: |
---|
| 762 | return regfi.regfi_iterator_down(self._iter) |
---|
| 763 | else: |
---|
| 764 | if name != None: |
---|
| 765 | name = name.encode('utf-8') |
---|
| 766 | return (regfi.regfi_iterator_find_subkey(self._iter, name) |
---|
| 767 | and regfi.regfi_iterator_down(self._iter)) |
---|
[206] | 768 | |
---|
[221] | 769 | |
---|
| 770 | ## Causes the iterator to ascend to the current Key's parent |
---|
| 771 | # |
---|
| 772 | # @return True if successful, False otherwise |
---|
| 773 | # |
---|
| 774 | # @note The state of current subkeys and values at this level in the tree |
---|
| 775 | # is lost as a side effect. That is, if you go up() and then back |
---|
| 776 | # down() again, current_subkey() and current_value() will return |
---|
| 777 | # default selections. |
---|
[206] | 778 | def up(self): |
---|
[220] | 779 | return regfi.regfi_iterator_up(self._iter) |
---|
[206] | 780 | |
---|
[221] | 781 | |
---|
| 782 | ## Selects first subkey of current key |
---|
| 783 | # |
---|
| 784 | # @return A Key instance for the first subkey. |
---|
| 785 | # None on error or if the current key has no subkeys. |
---|
[220] | 786 | def first_subkey(self): |
---|
| 787 | if regfi.regfi_iterator_first_subkey(self._iter): |
---|
| 788 | return self.current_subkey() |
---|
| 789 | return None |
---|
| 790 | |
---|
[221] | 791 | |
---|
| 792 | ## Selects first value of current Key |
---|
| 793 | # |
---|
| 794 | # @return A Value instance for the first value. |
---|
| 795 | # None on error or if the current key has no values. |
---|
[220] | 796 | def first_value(self): |
---|
| 797 | if regfi.regfi_iterator_first_value(self._iter): |
---|
| 798 | return self.current_value() |
---|
| 799 | return None |
---|
| 800 | |
---|
[221] | 801 | |
---|
| 802 | ## Selects the next subkey in the current Key's list |
---|
| 803 | # |
---|
| 804 | # @return A Key instance for the next subkey. |
---|
| 805 | # None if there are no remaining subkeys or an error occurred. |
---|
[220] | 806 | def next_subkey(self): |
---|
| 807 | if regfi.regfi_iterator_next_subkey(self._iter): |
---|
| 808 | return self.current_subkey() |
---|
| 809 | return None |
---|
| 810 | |
---|
[221] | 811 | |
---|
| 812 | ## Selects the next value in the current Key's list |
---|
| 813 | # |
---|
| 814 | # @return A Value instance for the next value. |
---|
| 815 | # None if there are no remaining values or an error occurred. |
---|
[220] | 816 | def next_value(self): |
---|
| 817 | if regfi.regfi_iterator_next_value(self._iter): |
---|
| 818 | return self.current_value() |
---|
| 819 | return None |
---|
| 820 | |
---|
[221] | 821 | |
---|
| 822 | ## Selects the first subkey which has the specified name |
---|
| 823 | # |
---|
| 824 | # @return A Key instance for the selected key. |
---|
| 825 | # None if it could not be located or an error occurred. |
---|
[220] | 826 | def find_subkey(self, name): |
---|
| 827 | if name != None: |
---|
| 828 | name = name.encode('utf-8') |
---|
| 829 | if regfi.regfi_iterator_find_subkey(self._iter, name): |
---|
| 830 | return self.current_subkey() |
---|
| 831 | return None |
---|
| 832 | |
---|
[221] | 833 | |
---|
| 834 | ## Selects the first value which has the specified name |
---|
| 835 | # |
---|
| 836 | # @return A Value instance for the selected value. |
---|
| 837 | # None if it could not be located or an error occurred. |
---|
[220] | 838 | def find_value(self, name): |
---|
| 839 | if name != None: |
---|
| 840 | name = name.encode('utf-8') |
---|
| 841 | if regfi.regfi_iterator_find_value(self._iter, name): |
---|
| 842 | return self.current_value() |
---|
| 843 | return None |
---|
| 844 | |
---|
[221] | 845 | ## Retrieves the currently selected subkey |
---|
| 846 | # |
---|
| 847 | # @return A Key instance of the current subkey |
---|
[220] | 848 | def current_subkey(self): |
---|
| 849 | return Key(self._hive, regfi.regfi_iterator_cur_subkey(self._iter)) |
---|
| 850 | |
---|
[221] | 851 | ## Retrieves the currently selected value |
---|
| 852 | # |
---|
| 853 | # @return A Value instance of the current value |
---|
[220] | 854 | def current_value(self): |
---|
| 855 | return Value(self._hive, regfi.regfi_iterator_cur_value(self._iter)) |
---|
| 856 | |
---|
[221] | 857 | ## Retrieves the current key |
---|
| 858 | # |
---|
| 859 | # @return A Key instance of the current position of the iterator |
---|
[220] | 860 | def current_key(self): |
---|
| 861 | return Key(self._hive, regfi.regfi_iterator_cur_key(self._iter)) |
---|
| 862 | |
---|
[221] | 863 | |
---|
| 864 | ## Traverse downward multiple levels |
---|
| 865 | # |
---|
| 866 | # This is more efficient than calling down() multiple times |
---|
| 867 | # |
---|
| 868 | # @param path A list of Key names which represent the path to descend |
---|
| 869 | # |
---|
| 870 | # @exception Exception If path could not be located |
---|
[206] | 871 | def descend(self, path): |
---|
[215] | 872 | cpath = _strlist2charss(path) |
---|
[206] | 873 | |
---|
[210] | 874 | # XXX: Use non-generic exception |
---|
[220] | 875 | if not regfi.regfi_iterator_walk_path(self._iter, cpath): |
---|
[206] | 876 | raise Exception('Could not locate path.\n'+GetLogMessages()) |
---|
[221] | 877 | |
---|
| 878 | |
---|
| 879 | # Freeing symbols defined for the sake of documentation |
---|
| 880 | del Value.name,Value.name_raw,Value.offset,Value.data_size,Value.type,Value.flags |
---|
| 881 | del Key.name,Key.name_raw,Key.offset,Key.modified,Key.flags |
---|
| 882 | del Hive.root,Hive.modified,Hive.sequence1,Hive.sequence2,Hive.major_version,Hive.minor_version |
---|