Commit | Line | Data |
---|---|---|
00b3cd10 PG |
1 | #!/usr/bin/env python3 |
2 | ||
3 | """ | |
83f2d71e | 4 | Intra2net 2017 |
00b3cd10 PG |
5 | |
6 | =============================================================================== | |
704ceaa5 | 7 | crypto -- Encryption Layer for the Deltatar Backup |
00b3cd10 PG |
8 | =============================================================================== |
9 | ||
10 | Crypto stack: | |
11 | ||
12 | - AES-GCM for the symmetric encryption; | |
13 | - Scrypt as KDF. | |
14 | ||
15 | References: | |
16 | ||
17 | - NIST Recommendation for Block Cipher Modes of Operation: Galois/Counter | |
18 | Mode (GCM) and GMAC | |
19 | http://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-38d.pdf | |
20 | ||
21 | - AES-GCM v1: | |
22 | https://cryptome.org/2014/01/aes-gcm-v1.pdf | |
23 | ||
24 | - Authentication weaknesses in GCM | |
25 | http://csrc.nist.gov/groups/ST/toolkit/BCM/documents/comments/CWC-GCM/Ferguson2.pdf | |
26 | ||
83f2d71e PG |
27 | Trouble with python-cryptography packages: authentication tags can only be |
28 | passed in advance: https://github.com/pyca/cryptography/pull/3421 | |
29 | ||
6d08915c PG |
30 | Errors |
31 | ------------------------------------------------------------------------------- | |
32 | ||
33 | Errors fall into roughly three categories: | |
34 | ||
704ceaa5 | 35 | - Cryptographical errors or invalid data. |
6d08915c PG |
36 | |
37 | - ``InvalidGCMTag`` (decryption failed on account of an invalid GCM | |
38 | tag), | |
39 | - ``InvalidIVFixedPart`` (IV fixed part of object not found in list), | |
f6cd676f | 40 | - ``DuplicateIV`` (the IV of an encrypted object already occurred), |
704ceaa5 PG |
41 | - ``DecryptionError`` (used in CLI decryption for presenting error |
42 | conditions to the user). | |
6d08915c PG |
43 | |
44 | - Incorrect usage of the library. | |
45 | ||
46 | - ``InvalidParameter`` (non-conforming user supplied parameter), | |
47 | - ``InvalidHeader`` (data passed for reading not parsable into header), | |
48 | - ``FormatError`` (cannot handle header or parameter version), | |
49 | - ``RuntimeError``. | |
50 | ||
51 | - Bad internal state. If one of these is encountered it means that a state | |
52 | was reached that shouldn’t occur during normal processing. | |
53 | ||
54 | - ``InternalError``, | |
55 | - ``Unreachable``. | |
56 | ||
57 | Also, ``EndOfFile`` is used as a sentinel to communicate that a stream supplied | |
58 | for reading is exhausted. | |
59 | ||
f6cd676f PG |
60 | Initialization Vectors |
61 | ------------------------------------------------------------------------------- | |
62 | ||
63 | Initialization vectors are checked reuse during the lifetime of a decryptor. | |
704ceaa5 PG |
64 | The fixed counters for metadata files cannot be reused and attempts to do so |
65 | will cause a DuplicateIV error. This means the length of objects encrypted with | |
66 | a metadata counter is capped at 63 GB. | |
67 | ||
68 | For ordinary, non-metadata payload, there is an optional mode with strict IV | |
69 | checking that causes a crypto context to fail if an IV encountered or created | |
70 | was already used for decrypting or encrypting, respectively, an earlier object. | |
71 | Note that this mode can trigger false positives when decrypting non-linearly, | |
72 | e. g. when traversing the same object multiple times. Since the crypto context | |
73 | has no notion of a position in a PDT encrypted archive, this condition must be | |
74 | sorted out downstream. | |
75 | ||
76 | Command Line Utility | |
77 | ------------------------------------------------------------------------------- | |
78 | ||
79 | ``crypto.py`` may be invoked as a script for decrypting, validating, and | |
80 | splitting PDT encrypted files. Consult the usage message for details. | |
81 | ||
82 | Usage examples: | |
83 | ||
84 | Decrypt from stdin using the password ‘foo’: :: | |
85 | ||
86 | $ crypto.py process foo -i - -o - <some-file.tar.gz.pdtcrypt >some-file.tar.gz | |
87 | ||
88 | Output verbose information about the encrypted objects in the archive: :: | |
89 | ||
90 | $ crypto.py process foo -v -i some-file.tar.gz.pdtcrypt -o /dev/null | |
91 | PDT: decrypt from some-file.tar.gz.pdtcrypt | |
92 | PDT: decrypt to /dev/null | |
93 | PDT: source: file some-file.tar.gz.pdtcrypt | |
94 | PDT: sink: file /dev/null | |
95 | PDT: 0 hdr | |
96 | PDT: · version = 1 : 0100 | |
97 | PDT: · paramversion = 1 : 0100 | |
98 | PDT: · nacl : d270 b031 00d1 87e2 c946 610d 7b7f 7e5f | |
99 | PDT: · iv : 02ee 3dd7 a963 1eb1 0100 0000 | |
100 | PDT: · ctsize = 591 : 4f02 0000 0000 0000 | |
101 | PDT: · tag : 5b2d 6d8b 8f82 4842 12fd 0b10 b6e3 369b | |
102 | PDT: 64 decrypt obj no. 1, 591 B | |
103 | PDT: · [64] 0% done, read block (591 B of 591 B remaining) | |
104 | PDT: · decrypt ciphertext 591 B | |
105 | PDT: · decrypt plaintext 591 B | |
106 | PDT: 655 finalize | |
107 | … | |
108 | ||
109 | Also, the mode *scrypt* allows deriving encryption keys. To calculate the | |
110 | encryption key from the password ‘foo’ and the salt of the first object in a | |
111 | PDT encrypted file: :: | |
112 | ||
113 | $ crypto.py scrypt foo -i some-file.pdtcrypt | |
4f6405d6 | 114 | {"paramversion": 1, "salt": "Cqzbk48e3peEjzWto8D0yA==", "key": "JH9EkMwaM4x9F5aim5gK/Q=="} |
704ceaa5 PG |
115 | |
116 | The computed 16 byte key is given in hexadecimal notation in the value to | |
117 | ``hash`` and can be fed into Python’s ``binascii.unhexlify()`` to obtain the | |
118 | corresponding binary representation. | |
119 | ||
120 | Note that in Scrypt hashing mode, no data integrity checks are being performed. | |
121 | If the wrong password is given, a wrong key will be derived. Whether the password | |
122 | was indeed correct can only be determined by decrypting. Note that since PDT | |
123 | archives essentially consist of a stream of independent objects, the salt and | |
124 | other parameters may change. Thus a key derived using above method from the | |
125 | first object doesn’t necessarily apply to any of the subsequent objects. | |
f6cd676f | 126 | |
00b3cd10 PG |
127 | """ |
128 | ||
7b3940e5 | 129 | import base64 |
00b3cd10 | 130 | import binascii |
50710d86 | 131 | import bisect |
00b3cd10 PG |
132 | import ctypes |
133 | import io | |
c46c8670 | 134 | from functools import reduce, partial |
f41973a6 | 135 | import mmap |
00b3cd10 PG |
136 | import os |
137 | import struct | |
a808459e | 138 | import stat |
00b3cd10 PG |
139 | import sys |
140 | import time | |
da82bc58 | 141 | import types |
00b3cd10 PG |
142 | try: |
143 | import enum34 | |
144 | except ImportError as exn: | |
145 | pass | |
146 | ||
147 | if __name__ == "__main__": ## Work around the import mechanism’s lest Python’s | |
148 | pwd = os.getcwd() ## preference for local imports causes a cyclical | |
149 | ## import (crypto → pylibscrypt → […] → ./tarfile → crypto). | |
150 | sys.path = [ p for p in sys.path if p.find ("deltatar") < 0 ] | |
151 | ||
152 | import pylibscrypt | |
153 | from cryptography.hazmat.primitives.ciphers import Cipher, algorithms, modes | |
154 | from cryptography.hazmat.backends import default_backend | |
15d3eefd | 155 | import cryptography |
00b3cd10 PG |
156 | |
157 | ||
a64085a8 | 158 | __all__ = [ "hdr_make", "hdr_read", "hdr_fmt", "hdr_fmt_pretty" |
b360b772 | 159 | , "scrypt_hashfile" |
3031b7ae PG |
160 | , "PDTCRYPT_HDR_SIZE", "AES_GCM_IV_CNT_DATA" |
161 | , "AES_GCM_IV_CNT_INFOFILE", "AES_GCM_IV_CNT_INDEX" | |
2d6fd8c8 | 162 | ] |
00b3cd10 | 163 | |
a393d9cb PG |
164 | |
165 | ############################################################################### | |
15d3eefd PG |
166 | ## exceptions |
167 | ############################################################################### | |
168 | ||
169 | class EndOfFile (Exception): | |
170 | """Reached EOF.""" | |
ae3d0f2a PG |
171 | remainder = 0 |
172 | msg = 0 | |
8a8ac469 | 173 | def __init__ (self, n=None, msg=None): |
5d394c0d PG |
174 | if n is not None: |
175 | self.remainder = n | |
176 | self.msg = msg | |
15d3eefd | 177 | |
b0078f26 | 178 | |
b12110dd PG |
179 | class InvalidParameter (Exception): |
180 | """Inputs not valid for PDT encryption.""" | |
181 | pass | |
182 | ||
b0078f26 | 183 | |
15d3eefd PG |
184 | class InvalidHeader (Exception): |
185 | """Header not valid.""" | |
186 | pass | |
187 | ||
b0078f26 PG |
188 | |
189 | class InvalidGCMTag (Exception): | |
190 | """ | |
191 | The GCM tag calculated during decryption differs from that in the object | |
192 | header. | |
193 | """ | |
194 | pass | |
195 | ||
196 | ||
26b42ad4 | 197 | class InvalidIVFixedPart (Exception): |
89ec6e2f PG |
198 | """ |
199 | IV fixed part not in supplied list: either the backup is corrupt or the | |
200 | current object does not belong to it. | |
201 | """ | |
26b42ad4 PG |
202 | pass |
203 | ||
b0078f26 | 204 | |
be124bca | 205 | class IVFixedPartError (Exception): |
89ec6e2f PG |
206 | """ |
207 | Error creating a unique IV fixed part: repeated calls to system RNG yielded | |
208 | the same sequence of bytes as the last IV used. | |
209 | """ | |
be124bca PG |
210 | pass |
211 | ||
212 | ||
fac2cfe1 | 213 | class InvalidFileCounter (Exception): |
89ec6e2f PG |
214 | """ |
215 | When encrypting, an attempted reuse of a dedicated counter (info file, | |
216 | index file) was caught. | |
217 | """ | |
fac2cfe1 PG |
218 | pass |
219 | ||
220 | ||
ee6aa239 | 221 | class DuplicateIV (Exception): |
89ec6e2f PG |
222 | """ |
223 | During encryption, the current IV fixed part is identical to an already | |
224 | existing IV (same prefix and file counter). This indicates tampering or | |
225 | programmer error and cannot be recovered from. | |
226 | """ | |
ee6aa239 PG |
227 | pass |
228 | ||
229 | ||
230 | class NonConsecutiveIV (Exception): | |
89ec6e2f PG |
231 | """ |
232 | IVs not numbered consecutively. This is a hard error with strict IV | |
233 | checking. Precludes random access to the encrypted objects. | |
234 | """ | |
ee6aa239 PG |
235 | pass |
236 | ||
237 | ||
b12110dd PG |
238 | class FormatError (Exception): |
239 | """Unusable parameters in header.""" | |
240 | pass | |
241 | ||
b0078f26 | 242 | |
15d3eefd | 243 | class DecryptionError (Exception): |
89ec6e2f | 244 | """Error during decryption with ``crypto.py`` on the command line.""" |
15d3eefd PG |
245 | pass |
246 | ||
b0078f26 | 247 | |
70ad9458 | 248 | class Unreachable (Exception): |
89ec6e2f PG |
249 | """ |
250 | Makeshift __builtin_unreachable(); always a programmer error if | |
251 | thrown. | |
252 | """ | |
70ad9458 PG |
253 | pass |
254 | ||
b0078f26 | 255 | |
b12110dd PG |
256 | class InternalError (Exception): |
257 | """Errors not ascribable to bad user inputs or cryptography.""" | |
258 | pass | |
259 | ||
15d3eefd PG |
260 | |
261 | ############################################################################### | |
a393d9cb PG |
262 | ## crypto layer version |
263 | ############################################################################### | |
264 | ||
265 | ENCRYPTION_PARAMETERS = \ | |
c46c8670 | 266 | { 0: \ |
dd23cbc9 PG |
267 | { "kdf": ("dummy", 16) |
268 | , "enc": "passthrough" } | |
c46c8670 | 269 | , 1: \ |
dd23cbc9 PG |
270 | { "kdf": ( "scrypt" |
271 | , { "dkLen" : 16 | |
272 | , "N" : 1 << 16 | |
273 | , "r" : 8 | |
274 | , "p" : 1 | |
275 | , "NaCl_LEN" : 16 }) | |
276 | , "enc": "aes-gcm" } } | |
a393d9cb | 277 | |
00b3cd10 PG |
278 | ############################################################################### |
279 | ## constants | |
280 | ############################################################################### | |
281 | ||
dd47d6a2 | 282 | PDTCRYPT_HDR_MAGIC = b"PDTCRYPT" |
00b3cd10 | 283 | |
dd47d6a2 PG |
284 | PDTCRYPT_HDR_SIZE_MAGIC = 8 # 8 |
285 | PDTCRYPT_HDR_SIZE_VERSION = 2 # 10 | |
286 | PDTCRYPT_HDR_SIZE_PARAMVERSION = 2 # 12 | |
287 | PDTCRYPT_HDR_SIZE_NACL = 16 # 28 | |
288 | PDTCRYPT_HDR_SIZE_IV = 12 # 40 | |
289 | PDTCRYPT_HDR_SIZE_CTSIZE = 8 # 48 | |
290 | PDTCRYPT_HDR_SIZE_TAG = 16 # 64 GCM auth tag | |
00b3cd10 | 291 | |
dd47d6a2 PG |
292 | PDTCRYPT_HDR_SIZE = PDTCRYPT_HDR_SIZE_MAGIC + PDTCRYPT_HDR_SIZE_VERSION \ |
293 | + PDTCRYPT_HDR_SIZE_PARAMVERSION + PDTCRYPT_HDR_SIZE_NACL \ | |
294 | + PDTCRYPT_HDR_SIZE_IV + PDTCRYPT_HDR_SIZE_CTSIZE \ | |
295 | + PDTCRYPT_HDR_SIZE_TAG # = 64 | |
00b3cd10 PG |
296 | |
297 | # precalculate offsets since Python can’t do constant folding over names | |
dd47d6a2 PG |
298 | HDR_OFF_VERSION = PDTCRYPT_HDR_SIZE_MAGIC |
299 | HDR_OFF_PARAMVERSION = HDR_OFF_VERSION + PDTCRYPT_HDR_SIZE_VERSION | |
300 | HDR_OFF_NACL = HDR_OFF_PARAMVERSION + PDTCRYPT_HDR_SIZE_PARAMVERSION | |
301 | HDR_OFF_IV = HDR_OFF_NACL + PDTCRYPT_HDR_SIZE_NACL | |
302 | HDR_OFF_CTSIZE = HDR_OFF_IV + PDTCRYPT_HDR_SIZE_IV | |
303 | HDR_OFF_TAG = HDR_OFF_CTSIZE + PDTCRYPT_HDR_SIZE_CTSIZE | |
00b3cd10 PG |
304 | |
305 | FMT_UINT16_LE = "<H" | |
306 | FMT_UINT64_LE = "<Q" | |
50710d86 | 307 | FMT_I2N_IV = "<8sL" # 8 random bytes ‖ 32 bit counter |
83f2d71e PG |
308 | FMT_I2N_HDR = ("<" # host byte order |
309 | "8s" # magic | |
310 | "H" # version | |
311 | "H" # paramversion | |
312 | "16s" # sodium chloride | |
313 | "12s" # iv | |
3b53fb98 PG |
314 | "Q" # size |
315 | "16s") # GCM tag | |
00b3cd10 PG |
316 | |
317 | # aes+gcm | |
cb7a3911 PG |
318 | AES_GCM_MAX_SIZE = (1 << 36) - (1 << 5) # 2^39 - 2^8 b ≅ 64 GB |
319 | PDTCRYPT_MAX_OBJ_SIZE_DEFAULT = 63 * (1 << 30) # 63 GB | |
320 | PDTCRYPT_MAX_OBJ_SIZE = PDTCRYPT_MAX_OBJ_SIZE_DEFAULT | |
00b3cd10 | 321 | |
3031b7ae PG |
322 | # index and info files are written on-the fly while encrypting so their |
323 | # counters must be available inadvance | |
cb7a3911 PG |
324 | AES_GCM_IV_CNT_INFOFILE = 1 # constant |
325 | AES_GCM_IV_CNT_INDEX = AES_GCM_IV_CNT_INFOFILE + 1 | |
326 | AES_GCM_IV_CNT_DATA = AES_GCM_IV_CNT_INDEX + 1 # also for multivolume | |
327 | AES_GCM_IV_CNT_MAX_DEFAULT = 0xffFFffFF | |
328 | AES_GCM_IV_CNT_MAX = AES_GCM_IV_CNT_MAX_DEFAULT | |
2d6fd8c8 | 329 | |
be124bca PG |
330 | # IV structure and generation |
331 | PDTCRYPT_IV_GEN_MAX_RETRIES = 10 # × | |
332 | PDTCRYPT_IV_FIXEDPART_SIZE = 8 # B | |
333 | PDTCRYPT_IV_COUNTER_SIZE = 4 # B | |
39accaaa | 334 | |
00b3cd10 | 335 | ############################################################################### |
39accaaa | 336 | ## header, trailer |
00b3cd10 PG |
337 | ############################################################################### |
338 | # | |
339 | # Interface: | |
340 | # | |
341 | # struct hdrinfo | |
342 | # { version : u16 | |
343 | # , paramversion : u16 | |
344 | # , nacl : [u8; 16] | |
345 | # , iv : [u8; 12] | |
704ceaa5 PG |
346 | # , ctsize : usize |
347 | # , tag : [u8; 16] } | |
83f2d71e | 348 | # |
00b3cd10 | 349 | # fn hdr_read (f : handle) -> hdrinfo; |
c2d1c3ec | 350 | # fn hdr_make (f : handle, h : hdrinfo) -> IOResult<usize>; |
00b3cd10 PG |
351 | # fn hdr_fmt (h : hdrinfo) -> String; |
352 | # | |
353 | ||
83f2d71e | 354 | def hdr_read (data): |
704ceaa5 PG |
355 | """ |
356 | Read bytes as header structure. | |
357 | ||
358 | If the input could not be interpreted as a header, fail with | |
359 | ``InvalidHeader``. | |
360 | """ | |
83f2d71e | 361 | |
00b3cd10 | 362 | try: |
3b53fb98 | 363 | mag, version, paramversion, nacl, iv, ctsize, tag = \ |
83f2d71e PG |
364 | struct.unpack (FMT_I2N_HDR, data) |
365 | except Exception as exn: | |
15d3eefd PG |
366 | raise InvalidHeader ("error unpacking header from [%r]: %s" |
367 | % (binascii.hexlify (data), str (exn))) | |
00b3cd10 | 368 | |
dd47d6a2 | 369 | if mag != PDTCRYPT_HDR_MAGIC: |
15d3eefd | 370 | raise InvalidHeader ("bad magic in header: expected [%s], got [%s]" |
dd47d6a2 | 371 | % (PDTCRYPT_HDR_MAGIC, mag)) |
00b3cd10 | 372 | |
15d3eefd | 373 | return \ |
00b3cd10 PG |
374 | { "version" : version |
375 | , "paramversion" : paramversion | |
376 | , "nacl" : nacl | |
377 | , "iv" : iv | |
378 | , "ctsize" : ctsize | |
3b53fb98 | 379 | , "tag" : tag |
00b3cd10 PG |
380 | } |
381 | ||
382 | ||
39accaaa | 383 | def hdr_read_stream (instr): |
704ceaa5 PG |
384 | """ |
385 | Read header from stream at the current position. | |
386 | ||
387 | Fail with ``InvalidHeader`` if insufficient bytes were read from the | |
388 | stream, or if the content could not be interpreted as a header. | |
389 | """ | |
dd47d6a2 | 390 | data = instr.read(PDTCRYPT_HDR_SIZE) |
ae3d0f2a | 391 | ldata = len (data) |
8a8ac469 PG |
392 | if ldata == 0: |
393 | raise EndOfFile | |
394 | elif ldata != PDTCRYPT_HDR_SIZE: | |
395 | raise InvalidHeader ("hdr_read_stream: expected %d B, received %d B" | |
396 | % (PDTCRYPT_HDR_SIZE, ldata)) | |
47e27926 | 397 | return hdr_read (data) |
39accaaa PG |
398 | |
399 | ||
3b53fb98 | 400 | def hdr_from_params (version, paramversion, nacl, iv, ctsize, tag): |
704ceaa5 PG |
401 | """ |
402 | Assemble the necessary values into a PDTCRYPT header. | |
403 | ||
404 | :type version: int to fit uint16_t | |
405 | :type paramversion: int to fit uint16_t | |
406 | :type nacl: bytes to fit uint8_t[16] | |
407 | :type iv: bytes to fit uint8_t[12] | |
408 | :type size: int to fit uint64_t | |
409 | :type tag: bytes to fit uint8_t[16] | |
410 | """ | |
dd47d6a2 | 411 | buf = bytearray (PDTCRYPT_HDR_SIZE) |
83f2d71e | 412 | bufv = memoryview (buf) |
00b3cd10 | 413 | |
00b3cd10 | 414 | try: |
83f2d71e | 415 | struct.pack_into (FMT_I2N_HDR, bufv, 0, |
dd47d6a2 | 416 | PDTCRYPT_HDR_MAGIC, |
3b53fb98 | 417 | version, paramversion, nacl, iv, ctsize, tag) |
83f2d71e | 418 | except Exception as exn: |
a83fa4ed | 419 | return False, "error assembling header: %s" % str (exn) |
00b3cd10 | 420 | |
83f2d71e | 421 | return True, bytes (buf) |
00b3cd10 | 422 | |
00b3cd10 | 423 | |
8a990744 PG |
424 | def hdr_make_dummy (s): |
425 | """ | |
426 | Create a header sized block of bytes initialized to a value derived from a | |
427 | string. Used to verify we’ve jumped back correctly to the actual position | |
428 | of the object header. | |
429 | """ | |
430 | c = reduce (lambda a, c: a + ord(c), s, 0) % 0xFF | |
dd47d6a2 | 431 | return bytes (bytearray (struct.pack ("B", c)) * PDTCRYPT_HDR_SIZE) |
8a990744 PG |
432 | |
433 | ||
a393d9cb | 434 | def hdr_make (hdr): |
704ceaa5 PG |
435 | """ |
436 | Assemble a header from the given header structure. | |
437 | """ | |
a393d9cb PG |
438 | return hdr_from_params (version=hdr.get("version"), |
439 | paramversion=hdr.get("paramversion"), | |
440 | nacl=hdr.get("nacl"), iv=hdr.get("iv"), | |
3b53fb98 | 441 | ctsize=hdr.get("ctsize"), tag=hdr.get("tag")) |
a393d9cb PG |
442 | |
443 | ||
83f2d71e | 444 | HDR_FMT = "I2n_header { version: %d, paramversion: %d, nacl: %s[%d]," \ |
89131745 | 445 | " iv: %s[%d], ctsize: %d, tag: %s[%d] }" |
00b3cd10 | 446 | |
83f2d71e | 447 | def hdr_fmt (h): |
704ceaa5 | 448 | """Format a header structure into readable output.""" |
83f2d71e PG |
449 | return HDR_FMT % (h["version"], h["paramversion"], |
450 | binascii.hexlify (h["nacl"]), len(h["nacl"]), | |
451 | binascii.hexlify (h["iv"]), len(h["iv"]), | |
db1f3ac7 PG |
452 | h["ctsize"], |
453 | binascii.hexlify (h["tag"]), len(h["tag"])) | |
00b3cd10 | 454 | |
00b3cd10 | 455 | |
83f2d71e | 456 | def hex_spaced_of_bytes (b): |
704ceaa5 | 457 | """Format bytes object, hexdump style.""" |
83f2d71e PG |
458 | return " ".join ([ "%.2x%.2x" % (c1, c2) |
459 | for c1, c2 in zip (b[0::2], b[1::2]) ]) \ | |
460 | + (len (b) | 1 == len (b) and " %.2x" % b[-1] or "") # odd lengths | |
00b3cd10 | 461 | |
591a722f | 462 | |
3031b7ae PG |
463 | def hdr_iv_counter (h): |
464 | """Extract the variable part of the IV of the given header.""" | |
465 | _fixed, cnt = struct.unpack (FMT_I2N_IV, h ["iv"]) | |
466 | return cnt | |
467 | ||
468 | ||
469 | def hdr_iv_fixed (h): | |
470 | """Extract the fixed part of the IV of the given header.""" | |
471 | fixed, _cnt = struct.unpack (FMT_I2N_IV, h ["iv"]) | |
472 | return fixed | |
473 | ||
474 | ||
83f2d71e | 475 | hdr_dump = hex_spaced_of_bytes |
00b3cd10 | 476 | |
00b3cd10 | 477 | |
15d3eefd PG |
478 | HDR_FMT_PRETTY = \ |
479 | """version = %-4d : %s | |
480 | paramversion = %-4d : %s | |
481 | nacl : %s | |
482 | iv : %s | |
483 | ctsize = %-20d : %s | |
484 | tag : %s | |
83f2d71e | 485 | """ |
00b3cd10 | 486 | |
83f2d71e | 487 | def hdr_fmt_pretty (h): |
704ceaa5 PG |
488 | """ |
489 | Format header structure into multi-line representation of its contents and | |
490 | their raw representation. (Omit the implicit “PDTCRYPT” magic bytes that | |
491 | precede every header.) | |
492 | """ | |
83f2d71e PG |
493 | return HDR_FMT_PRETTY \ |
494 | % (h["version"], | |
495 | hex_spaced_of_bytes (struct.pack (FMT_UINT16_LE, h["version"])), | |
496 | h["paramversion"], | |
497 | hex_spaced_of_bytes (struct.pack (FMT_UINT16_LE, h["paramversion"])), | |
498 | hex_spaced_of_bytes (h["nacl"]), | |
499 | hex_spaced_of_bytes (h["iv"]), | |
500 | h["ctsize"], | |
15d3eefd PG |
501 | hex_spaced_of_bytes (struct.pack (FMT_UINT64_LE, h["ctsize"])), |
502 | hex_spaced_of_bytes (h["tag"])) | |
00b3cd10 | 503 | |
f6cd676f PG |
504 | IV_FMT = "((f %s) (c %d))" |
505 | ||
506 | def iv_fmt (iv): | |
704ceaa5 | 507 | """Format the two components of an IV in a readable fashion.""" |
f6cd676f PG |
508 | fixed, cnt = struct.unpack (FMT_I2N_IV, iv) |
509 | return IV_FMT % (binascii.hexlify (fixed), cnt) | |
510 | ||
00b3cd10 | 511 | |
00b3cd10 | 512 | ############################################################################### |
f41973a6 PG |
513 | ## restoration |
514 | ############################################################################### | |
515 | ||
516 | class Location (object): | |
517 | n = 0 | |
518 | offset = 0 | |
519 | ||
520 | def restore_loc_fmt (loc): | |
521 | return "%d off:%d" \ | |
522 | % (loc.n, loc.offset) | |
523 | ||
524 | def locate_hdr_candidates (fd): | |
525 | """ | |
526 | Walk over instances of the magic string in the payload, collecting their | |
527 | positions. If the offset of the first found instance is not zero, the file | |
528 | begins with leading garbage. | |
529 | ||
530 | :return: The list of offsets in the file. | |
531 | """ | |
532 | cands = [] | |
533 | ||
534 | mm = mmap.mmap(fd, 0, mmap.MAP_SHARED, mmap.PROT_READ) | |
535 | pos = 0 | |
536 | while True: | |
537 | pos = mm.find (PDTCRYPT_HDR_MAGIC, pos) | |
538 | if pos == -1: | |
539 | break | |
540 | cands.append (pos) | |
541 | pos += 1 | |
542 | ||
543 | return cands | |
544 | ||
545 | ||
6c8073ab PG |
546 | HDR_CAND_GOOD = 0 # header marks begin of valid object |
547 | HDR_CAND_FISHY = 1 # inconclusive (tag mismatch, obj overlap etc.) | |
548 | HDR_CAND_JUNK = 2 # not a header / object unreadable | |
549 | ||
550 | ||
551 | def inspect_hdr (fd, off): | |
552 | """ | |
553 | Attempt to parse a header in *fd* at position *off*. | |
554 | ||
555 | Returns a verdict about the quality of that header plus the parsed header | |
556 | when readable. | |
557 | """ | |
558 | ||
559 | _ = os.lseek (fd, off, os.SEEK_SET) | |
560 | ||
561 | if os.lseek (fd, 0, os.SEEK_CUR) != off: | |
562 | if PDTCRYPT_VERBOSE is True: | |
563 | noise ("PDT: %d → dismissed (lseek() past EOF)" % off) | |
564 | return HDR_CAND_JUNK, None | |
565 | ||
566 | raw = os.read (fd, PDTCRYPT_HDR_SIZE) | |
567 | if len (raw) != PDTCRYPT_HDR_SIZE: | |
568 | if PDTCRYPT_VERBOSE is True: | |
569 | noise ("PDT: %d → dismissed (EOF inside header)" % off) | |
570 | return HDR_CAND_JUNK, None | |
571 | ||
572 | try: | |
573 | hdr = hdr_read (raw) | |
574 | except InvalidHeader as exn: | |
575 | if PDTCRYPT_VERBOSE is True: | |
576 | noise ("PDT: %d → dismissed (invalid: [%s])" % (off, str (exn))) | |
577 | return HDR_CAND_JUNK, None | |
578 | ||
579 | obj0 = off + PDTCRYPT_HDR_SIZE | |
580 | objX = obj0 + hdr ["ctsize"] | |
581 | ||
582 | eof = os.lseek (fd, 0, os.SEEK_END) | |
583 | if eof < objX: | |
584 | if PDTCRYPT_VERBOSE is True: | |
585 | noise ("PDT: %d → EOF inside object (%d≤%d≤%d); adjusting size to " | |
586 | "%d" % (off, obj0, eof, objX, (eof - obj0))) | |
587 | # try reading up to the end | |
588 | hdr ["ctsize"] = eof - obj0 | |
589 | return HDR_CAND_FISHY, hdr | |
590 | ||
591 | return HDR_CAND_GOOD, hdr | |
592 | ||
593 | ||
a808459e | 594 | def try_decrypt (ifd, off, hdr, secret, ofd=-1): |
6c8073ab | 595 | """ |
a808459e PG |
596 | Attempt to decrypt the object in the (seekable) descriptor *ifd* starting |
597 | at *off* using the metadata in *hdr* and *secret*. An output fd can be | |
598 | specified with *ofd*; if it is *-1* – the default –, the decrypted payload | |
599 | will be discarded. | |
70a33834 PG |
600 | |
601 | Always creates a fresh decryptor, so validation steps across objects don’t | |
602 | apply. | |
202104ed PG |
603 | |
604 | Errors during GCM tag validation are ignored. | |
6c8073ab | 605 | """ |
70a33834 PG |
606 | ctleft = hdr ["ctsize"] |
607 | pos = off | |
608 | ||
609 | ks = secret [0] | |
610 | if ks == PDTCRYPT_SECRET_PW: | |
611 | decr = Decrypt (password=secret [1]) | |
612 | elif ks == PDTCRYPT_SECRET_KEY: | |
613 | key = binascii.unhexlify (secret [1]) | |
614 | decr = Decrypt (key=key) | |
615 | else: | |
616 | raise RuntimeError | |
617 | ||
70a33834 PG |
618 | decr.next (hdr) |
619 | ||
620 | try: | |
a808459e | 621 | os.lseek (ifd, pos, os.SEEK_SET) |
70a33834 PG |
622 | while ctleft > 0: |
623 | cnksiz = min (ctleft, PDTCRYPT_BLOCKSIZE) | |
a808459e | 624 | cnk = os.read (ifd, cnksiz) |
70a33834 PG |
625 | ctleft -= cnksiz |
626 | pos += cnksiz | |
a808459e PG |
627 | pt = decr.process (cnk) |
628 | if ofd != -1: | |
629 | os.write (ofd, pt) | |
202104ed PG |
630 | try: |
631 | pt = decr.done () | |
632 | except InvalidGCMTag: | |
633 | noise ("PDT: GCM tag mismatch for object %d–%d" | |
634 | % (off, off + hdr ["ctsize"])) | |
a808459e PG |
635 | if len (pt) > 0 and ofd != -1: |
636 | os.write (ofd, pt) | |
70a33834 | 637 | |
70a33834 PG |
638 | except Exception as exn: |
639 | noise ("PDT: error decrypting object %d–%d@%d, %d B remaining [%s]" | |
640 | % (off, off + hdr ["ctsize"], pos, ctleft, exn)) | |
641 | raise | |
6c8073ab | 642 | |
70a33834 | 643 | return pos - off |
6c8073ab PG |
644 | |
645 | ||
6690f5e0 PG |
646 | def readable_objects_offsets (ifd, secret, cands): |
647 | """ | |
648 | From a list of candidates, locate the ones that mark the start of actual | |
649 | readable PDTCRYPT objects. | |
650 | """ | |
651 | good = [] | |
652 | nobj = 0 | |
653 | for cand in cands: | |
654 | nobj += 1 | |
655 | vdt, hdr = inspect_hdr (ifd, cand) | |
656 | if vdt == HDR_CAND_JUNK: | |
657 | pass # ignore unreadable ones | |
658 | elif vdt in [HDR_CAND_GOOD, HDR_CAND_FISHY]: | |
659 | off0 = cand + PDTCRYPT_HDR_SIZE | |
660 | ok = try_decrypt (ifd, off0, hdr, secret) == hdr ["ctsize"] | |
661 | if ok is True: | |
662 | good.append (cand) | |
663 | return good | |
664 | ||
665 | ||
666 | def reconstruct_offsets (fname, secret): | |
667 | ifd = os.open (fname, os.O_RDONLY) | |
668 | ||
669 | try: | |
670 | cands = locate_hdr_candidates (ifd) | |
671 | return readable_objects_offsets (ifd, secret, cands) | |
672 | finally: | |
673 | os.close (ifd) | |
674 | ||
675 | ||
f41973a6 | 676 | ############################################################################### |
6178061e PG |
677 | ## passthrough / null encryption |
678 | ############################################################################### | |
679 | ||
680 | class PassthroughCipher (object): | |
681 | ||
682 | tag = struct.pack ("<QQ", 0, 0) | |
683 | ||
684 | def __init__ (self) : pass | |
685 | ||
686 | def update (self, b) : return b | |
687 | ||
50710d86 | 688 | def finalize (self) : return b"" |
6178061e PG |
689 | |
690 | def finalize_with_tag (self, _) : return b"" | |
691 | ||
692 | ############################################################################### | |
a393d9cb | 693 | ## convenience wrapper |
00b3cd10 PG |
694 | ############################################################################### |
695 | ||
c46c8670 PG |
696 | |
697 | def kdf_dummy (klen, password, _nacl): | |
704ceaa5 PG |
698 | """ |
699 | Fake KDF for testing purposes that is called when parameter version zero is | |
700 | encountered. | |
701 | """ | |
c46c8670 PG |
702 | q, r = divmod (klen, len (password)) |
703 | if isinstance (password, bytes) is False: | |
704 | password = password.encode () | |
705 | return password * q + password [:r], b"" | |
706 | ||
707 | ||
708 | SCRYPT_KEY_MEMO = { } # static because needed for both the info file and the archive | |
709 | ||
710 | ||
711 | def kdf_scrypt (params, password, nacl): | |
704ceaa5 PG |
712 | """ |
713 | Wrapper for the Scrypt KDF, corresponds to parameter version one. The | |
714 | computation result is memoized based on the inputs to facilitate spawning | |
715 | multiple encryption contexts. | |
716 | """ | |
c46c8670 PG |
717 | N = params["N"] |
718 | r = params["r"] | |
719 | p = params["p"] | |
720 | dkLen = params["dkLen"] | |
721 | ||
722 | if nacl is None: | |
723 | nacl = os.urandom (params["NaCl_LEN"]) | |
724 | ||
725 | key_parms = (password, nacl, N, r, p, dkLen) | |
726 | global SCRYPT_KEY_MEMO | |
727 | if key_parms not in SCRYPT_KEY_MEMO: | |
728 | SCRYPT_KEY_MEMO [key_parms] = \ | |
729 | pylibscrypt.scrypt (password, nacl, N, r, p, dkLen) | |
730 | return SCRYPT_KEY_MEMO [key_parms], nacl | |
a64085a8 PG |
731 | |
732 | ||
da82bc58 | 733 | def kdf_by_version (paramversion=None, defs=None): |
704ceaa5 PG |
734 | """ |
735 | Pick the KDF handler corresponding to the parameter version or the | |
736 | definition set. | |
737 | ||
738 | :rtype: function (password : str, nacl : str) -> str | |
739 | """ | |
da82bc58 PG |
740 | if paramversion is not None: |
741 | defs = ENCRYPTION_PARAMETERS.get(paramversion, None) | |
a64085a8 | 742 | if defs is None: |
1ed44e7b PG |
743 | raise InvalidParameter ("no encryption parameters for version %r" |
744 | % paramversion) | |
a64085a8 | 745 | (kdf, params) = defs["kdf"] |
c46c8670 PG |
746 | fn = None |
747 | if kdf == "scrypt" : fn = kdf_scrypt | |
748 | if kdf == "dummy" : fn = kdf_dummy | |
749 | if fn is None: | |
a64085a8 | 750 | raise ValueError ("key derivation method %r unknown" % kdf) |
c46c8670 | 751 | return partial (fn, params) |
a64085a8 PG |
752 | |
753 | ||
b360b772 PG |
754 | ############################################################################### |
755 | ## SCRYPT hashing | |
756 | ############################################################################### | |
757 | ||
758 | def scrypt_hashsource (pw, ins): | |
759 | """ | |
760 | Calculate the SCRYPT hash from the password and the information contained | |
761 | in the first header found in ``ins``. | |
762 | ||
763 | This does not validate whether the first object is encrypted correctly. | |
764 | """ | |
c1ecc2e2 PG |
765 | if isinstance (pw, str) is True: |
766 | pw = str.encode (pw) | |
767 | elif isinstance (pw, bytes) is False: | |
768 | raise InvalidParameter ("password must be a string, not %s" | |
769 | % type (password)) | |
770 | if isinstance (ins, io.BufferedReader) is False and \ | |
771 | isinstance (ins, io.FileIO) is False: | |
772 | raise InvalidParameter ("file to hash must be opened in “binary” mode") | |
b360b772 PG |
773 | hdr = None |
774 | try: | |
775 | hdr = hdr_read_stream (ins) | |
776 | except EndOfFile as exn: | |
777 | noise ("PDT: malformed input: end of file reading first object header") | |
778 | noise ("PDT:") | |
779 | return 1 | |
780 | ||
781 | nacl = hdr ["nacl"] | |
782 | pver = hdr ["paramversion"] | |
783 | if PDTCRYPT_VERBOSE is True: | |
784 | noise ("PDT: salt of first object : %s" % binascii.hexlify (nacl)) | |
785 | noise ("PDT: parameter version of archive : %d" % pver) | |
786 | ||
787 | try: | |
788 | defs = ENCRYPTION_PARAMETERS.get(pver, None) | |
789 | kdfname, params = defs ["kdf"] | |
790 | if kdfname != "scrypt": | |
791 | noise ("PDT: input is not an SCRYPT archive") | |
792 | noise ("") | |
793 | return 1 | |
794 | kdf = kdf_by_version (None, defs) | |
795 | except ValueError as exn: | |
796 | noise ("PDT: object has unknown parameter version %d" % pver) | |
797 | ||
798 | hsh, _void = kdf (pw, nacl) | |
799 | ||
c1ecc2e2 | 800 | return hsh, nacl, hdr ["version"], pver |
b360b772 PG |
801 | |
802 | ||
803 | def scrypt_hashfile (pw, fname): | |
704ceaa5 PG |
804 | """ |
805 | Calculate the SCRYPT hash from the password and the information contained | |
806 | in the first header found in the given file. The header is read only at | |
807 | offset zero. | |
808 | """ | |
b360b772 | 809 | with deptdcrypt_mk_stream (PDTCRYPT_SOURCE, fname or "-") as ins: |
c1ecc2e2 | 810 | hsh, _void, _void, _void = scrypt_hashsource (pw, ins) |
b360b772 PG |
811 | return hsh |
812 | ||
813 | ||
814 | ############################################################################### | |
815 | ## AES-GCM context | |
816 | ############################################################################### | |
817 | ||
a393d9cb PG |
818 | class Crypto (object): |
819 | """ | |
820 | Encryption context to remain alive throughout an entire tarfile pass. | |
821 | """ | |
6178061e | 822 | enc = None |
a393d9cb PG |
823 | nacl = None |
824 | key = None | |
50710d86 PG |
825 | cnt = None # file counter (uint32_t != 0) |
826 | iv = None # current IV | |
30019abf PG |
827 | fixed = None # accu for 64 bit fixed parts of IV |
828 | used_ivs = None # tracks IVs | |
829 | strict_ivs = False # if True, panic on duplicate object IV | |
48db09ba PG |
830 | password = None |
831 | paramversion = None | |
633b18a9 PG |
832 | stats = { "in" : 0 |
833 | , "out" : 0 | |
834 | , "obj" : 0 } | |
fa47412e | 835 | |
fa47412e PG |
836 | ctsize = -1 |
837 | ptsize = -1 | |
3031b7ae PG |
838 | info_counter_used = False |
839 | index_counter_used = False | |
a393d9cb | 840 | |
a64085a8 | 841 | def __init__ (self, *al, **akv): |
30019abf | 842 | self.used_ivs = set () |
a64085a8 | 843 | self.set_parameters (*al, **akv) |
39accaaa PG |
844 | |
845 | ||
704ceaa5 | 846 | def next_fixed (self): |
be124bca | 847 | # NOP for decryption |
50710d86 PG |
848 | pass |
849 | ||
850 | ||
851 | def set_object_counter (self, cnt=None): | |
704ceaa5 PG |
852 | """ |
853 | Safely set the internal counter of encrypted objects. Numerous | |
854 | constraints apply: | |
855 | ||
856 | The same counter may not be reused in combination with one IV fixed | |
857 | part. This is validated elsewhere in the IV handling. | |
858 | ||
859 | Counter zero is invalid. The first two counters are reserved for | |
860 | metadata. The implementation does not allow for splitting metadata | |
861 | files over multiple encrypted objects. (This would be possible by | |
862 | assigning new fixed parts.) Thus in a Deltatar backup there is at most | |
863 | one object with a counter value of one and two. On creation of a | |
864 | context, the initial counter may be chosen. The globals | |
865 | ``AES_GCM_IV_CNT_INFOFILE`` and ``AES_GCM_IV_CNT_INDEX`` can be used to | |
866 | request one of the reserved values. If one of these values has been | |
867 | used, any further attempt of setting the counter to that value will | |
868 | be rejected with an ``InvalidFileCounter`` exception. | |
869 | ||
870 | Out of bounds values (i. e. below one and more than the maximum of 2³²) | |
871 | cause an ``InvalidParameter`` exception to be thrown. | |
872 | """ | |
50710d86 PG |
873 | if cnt is None: |
874 | self.cnt = AES_GCM_IV_CNT_DATA | |
875 | return | |
876 | if cnt == 0 or cnt > AES_GCM_IV_CNT_MAX + 1: | |
b12110dd PG |
877 | raise InvalidParameter ("invalid counter value %d requested: " |
878 | "acceptable values are from 1 to %d" | |
879 | % (cnt, AES_GCM_IV_CNT_MAX)) | |
50710d86 PG |
880 | if cnt == AES_GCM_IV_CNT_INFOFILE: |
881 | if self.info_counter_used is True: | |
fac2cfe1 PG |
882 | raise InvalidFileCounter ("attempted to reuse info file " |
883 | "counter %d: must be unique" % cnt) | |
50710d86 | 884 | self.info_counter_used = True |
3031b7ae PG |
885 | elif cnt == AES_GCM_IV_CNT_INDEX: |
886 | if self.index_counter_used is True: | |
fac2cfe1 PG |
887 | raise InvalidFileCounter ("attempted to reuse index file " |
888 | " counter %d: must be unique" % cnt) | |
3031b7ae | 889 | self.index_counter_used = True |
50710d86 PG |
890 | if cnt <= AES_GCM_IV_CNT_MAX: |
891 | self.cnt = cnt | |
892 | return | |
893 | # cnt == AES_GCM_IV_CNT_MAX + 1 → wrap | |
894 | self.cnt = AES_GCM_IV_CNT_DATA | |
704ceaa5 | 895 | self.next_fixed () |
50710d86 PG |
896 | |
897 | ||
1f3fd7b0 | 898 | def set_parameters (self, password=None, key=None, paramversion=None, |
be124bca | 899 | nacl=None, counter=None, strict_ivs=False): |
704ceaa5 PG |
900 | """ |
901 | Configure the internal state of a crypto context. Not intended for | |
902 | external use. | |
903 | """ | |
be124bca | 904 | self.next_fixed () |
50710d86 | 905 | self.set_object_counter (counter) |
30019abf PG |
906 | self.strict_ivs = strict_ivs |
907 | ||
a83fa4ed PG |
908 | if paramversion is not None: |
909 | self.paramversion = paramversion | |
910 | ||
1f3fd7b0 PG |
911 | if key is not None: |
912 | self.key, self.nacl = key, nacl | |
913 | return | |
914 | ||
a83fa4ed PG |
915 | if password is not None: |
916 | if isinstance (password, bytes) is False: | |
917 | password = str.encode (password) | |
918 | self.password = password | |
919 | if paramversion is None and nacl is None: | |
920 | # postpone key setup until first header is available | |
921 | return | |
922 | kdf = kdf_by_version (paramversion) | |
923 | if kdf is not None: | |
924 | self.key, self.nacl = kdf (password, nacl) | |
fa47412e | 925 | |
39accaaa | 926 | |
39accaaa | 927 | def process (self, buf): |
704ceaa5 PG |
928 | """ |
929 | Encrypt / decrypt a buffer. Invokes the ``.update()`` method on the | |
930 | wrapped encryptor or decryptor, respectively. | |
931 | ||
932 | The Cryptography exception ``AlreadyFinalized`` is translated to an | |
933 | ``InternalError`` at this point. It may occur in sound code when the GC | |
934 | closes an encrypting stream after an error. Everywhere else it must be | |
935 | treated as a bug. | |
936 | """ | |
cb7a3911 PG |
937 | if self.enc is None: |
938 | raise RuntimeError ("process: context not initialized") | |
939 | self.stats ["in"] += len (buf) | |
fac2cfe1 PG |
940 | try: |
941 | out = self.enc.update (buf) | |
942 | except cryptography.exceptions.AlreadyFinalized as exn: | |
943 | raise InternalError (exn) | |
cb7a3911 PG |
944 | self.stats ["out"] += len (out) |
945 | return out | |
39accaaa PG |
946 | |
947 | ||
30019abf | 948 | def next (self, password, paramversion, nacl, iv): |
704ceaa5 PG |
949 | """ |
950 | Prepare for encrypting another object: Reset the data counters and | |
951 | change the configuration in case one of the variable parameters differs | |
952 | from the last object. Also check the IV for duplicates and error out | |
953 | if strict checking was requested. | |
954 | """ | |
fa47412e PG |
955 | self.ctsize = 0 |
956 | self.ptsize = 0 | |
957 | self.stats ["obj"] += 1 | |
30019abf PG |
958 | |
959 | self.check_duplicate_iv (iv) | |
960 | ||
6178061e PG |
961 | if ( self.paramversion != paramversion |
962 | or self.password != password | |
963 | or self.nacl != nacl): | |
1f3fd7b0 | 964 | self.set_parameters (password=password, paramversion=paramversion, |
30019abf PG |
965 | nacl=nacl, strict_ivs=self.strict_ivs) |
966 | ||
967 | ||
968 | def check_duplicate_iv (self, iv): | |
704ceaa5 PG |
969 | """ |
970 | Add an IV (the 12 byte representation as in the header) to the list. With | |
971 | strict checking enabled, this will throw a ``DuplicateIV``. Depending on | |
972 | the context, this may indicate a serious error (IV reuse). | |
973 | """ | |
30019abf PG |
974 | if self.strict_ivs is True and iv in self.used_ivs: |
975 | raise DuplicateIV ("iv %s was reused" % iv_fmt (iv)) | |
976 | # vi has not been used before; add to collection | |
977 | self.used_ivs.add (iv) | |
fa47412e PG |
978 | |
979 | ||
633b18a9 | 980 | def counters (self): |
704ceaa5 PG |
981 | """ |
982 | Access the data counters. | |
983 | """ | |
633b18a9 PG |
984 | return self.stats ["obj"], self.stats ["in"], self.stats ["out"] |
985 | ||
986 | ||
8de91f4f PG |
987 | def drop (self): |
988 | """ | |
989 | Clear the current context regardless of its finalization state. The | |
990 | next operation must be ``.next()``. | |
991 | """ | |
992 | self.enc = None | |
993 | ||
994 | ||
39accaaa PG |
995 | class Encrypt (Crypto): |
996 | ||
48db09ba PG |
997 | lastinfo = None |
998 | version = None | |
72a42219 | 999 | paramenc = None |
50710d86 | 1000 | |
1f3fd7b0 | 1001 | def __init__ (self, version, paramversion, password=None, key=None, nacl=None, |
30019abf | 1002 | counter=AES_GCM_IV_CNT_DATA, strict_ivs=True): |
704ceaa5 PG |
1003 | """ |
1004 | The ctor will throw immediately if one of the parameters does not conform | |
1005 | to our expectations. | |
1006 | ||
1007 | counter=AES_GCM_IV_CNT_DATA, strict_ivs=True): | |
1008 | :type version: int to fit uint16_t | |
1009 | :type paramversion: int to fit uint16_t | |
1010 | :param password: mutually exclusive with ``key`` | |
1011 | :type password: bytes | |
1012 | :param key: mutually exclusive with ``password`` | |
1013 | :type key: bytes | |
1014 | :type nacl: bytes | |
1015 | :type counter: initial object counter the values | |
1016 | ``AES_GCM_IV_CNT_INFOFILE`` and | |
1017 | ``AES_GCM_IV_CNT_INDEX`` are unique in each backup set | |
1018 | and cannot be reused even with different fixed parts. | |
1019 | :type strict_ivs: bool | |
1020 | """ | |
1f3fd7b0 PG |
1021 | if password is None and key is None \ |
1022 | or password is not None and key is not None : | |
1023 | raise InvalidParameter ("__init__: need either key or password") | |
1024 | ||
1025 | if key is not None: | |
1026 | if isinstance (key, bytes) is False: | |
1027 | raise InvalidParameter ("__init__: key must be provided as " | |
1028 | "bytes, not %s" % type (key)) | |
1029 | if nacl is None: | |
1030 | raise InvalidParameter ("__init__: salt must be provided along " | |
1031 | "with encryption key") | |
1032 | else: # password, no key | |
1033 | if isinstance (password, str) is False: | |
1034 | raise InvalidParameter ("__init__: password must be a string, not %s" | |
1035 | % type (password)) | |
1036 | if len (password) == 0: | |
1037 | raise InvalidParameter ("__init__: supplied empty password but not " | |
1038 | "permitted for PDT encrypted files") | |
36b9932a PG |
1039 | # version |
1040 | if isinstance (version, int) is False: | |
1041 | raise InvalidParameter ("__init__: version number must be an " | |
1042 | "integer, not %s" % type (version)) | |
1043 | if version < 0: | |
1044 | raise InvalidParameter ("__init__: version number must be a " | |
1045 | "nonnegative integer, not %d" % version) | |
1046 | # paramversion | |
1047 | if isinstance (paramversion, int) is False: | |
1048 | raise InvalidParameter ("__init__: crypto parameter version number " | |
1049 | "must be an integer, not %s" | |
1050 | % type (paramversion)) | |
1051 | if paramversion < 0: | |
1052 | raise InvalidParameter ("__init__: crypto parameter version number " | |
1053 | "must be a nonnegative integer, not %d" | |
1054 | % paramversion) | |
1055 | # salt | |
1056 | if nacl is not None: | |
1057 | if isinstance (nacl, bytes) is False: | |
1058 | raise InvalidParameter ("__init__: salt given, but of type %s " | |
1059 | "instead of bytes" % type (nacl)) | |
1060 | # salt length would depend on the actual encryption so it can’t be | |
1061 | # validated at this point | |
b12110dd | 1062 | self.fixed = [ ] |
48db09ba PG |
1063 | self.version = version |
1064 | self.paramenc = ENCRYPTION_PARAMETERS.get (paramversion) ["enc"] | |
72a42219 | 1065 | |
1f3fd7b0 | 1066 | super().__init__ (password, key, paramversion, nacl, counter=counter, |
30019abf | 1067 | strict_ivs=strict_ivs) |
a393d9cb PG |
1068 | |
1069 | ||
be124bca PG |
1070 | def next_fixed (self, retries=PDTCRYPT_IV_GEN_MAX_RETRIES): |
1071 | """ | |
1072 | Generate the next IV fixed part by reading eight bytes from | |
1073 | ``/dev/urandom``. The buffer so obtained is tested against the fixed | |
1074 | parts used so far to prevent accidental reuse of IVs. After a | |
1075 | configurable number of attempts to create a unique fixed part, it will | |
1076 | refuse to continue with an ``IVFixedPartError``. This is unlikely to | |
1077 | ever happen on a normal system but may detect an issue with the random | |
1078 | generator. | |
1079 | ||
1080 | The list of fixed parts that were used by the context at hand can be | |
1081 | accessed through the ``.fixed`` list. Its last element is the fixed | |
1082 | part currently in use. | |
1083 | """ | |
1084 | i = 0 | |
1085 | while i < retries: | |
1086 | fp = os.urandom (PDTCRYPT_IV_FIXEDPART_SIZE) | |
1087 | if fp not in self.fixed: | |
1088 | self.fixed.append (fp) | |
1089 | return | |
1090 | i += 1 | |
1091 | raise IVFixedPartError ("error obtaining a unique IV fixed part from " | |
1092 | "/dev/urandom; giving up after %d tries" % i) | |
1093 | ||
1094 | ||
a393d9cb | 1095 | def iv_make (self): |
704ceaa5 PG |
1096 | """ |
1097 | Construct a 12-bytes IV from the current fixed part and the object | |
1098 | counter. | |
1099 | """ | |
b12110dd | 1100 | return struct.pack(FMT_I2N_IV, self.fixed [-1], self.cnt) |
a393d9cb PG |
1101 | |
1102 | ||
cb7a3911 | 1103 | def next (self, filename=None, counter=None): |
704ceaa5 PG |
1104 | """ |
1105 | Prepare for encrypting the next incoming object. Update the counter | |
1106 | and put together the IV, possibly changing prefixes. Then create the | |
1107 | new encryptor. | |
1108 | ||
1109 | The argument ``counter`` can be used to specify a file counter for this | |
1110 | object. Unless it is one of the reserved values, the counter of | |
1111 | subsequent objects will be computed from this one. | |
1112 | ||
1113 | If this is the first object in a series, ``filename`` is required, | |
1114 | otherwise it is reused if not present. The value is used to derive a | |
1115 | header sized placeholder to use until after encryption when all the | |
1116 | inputs to construct the final header are available. This is then | |
1117 | matched in ``.done()`` against the value found at the position of the | |
1118 | header. The motivation for this extra check is primarily to assist | |
1119 | format debugging: It makes stray headers easy to spot in malformed | |
1120 | PDTCRYPT files. | |
1121 | """ | |
cb7a3911 PG |
1122 | if filename is None: |
1123 | if self.lastinfo is None: | |
1124 | raise InvalidParameter ("next: filename is mandatory for " | |
1125 | "first object") | |
1126 | filename, _dummy = self.lastinfo | |
1127 | else: | |
1128 | if isinstance (filename, str) is False: | |
1129 | raise InvalidParameter ("next: filename must be a string, no %s" | |
1130 | % type (filename)) | |
3031b7ae PG |
1131 | if counter is not None: |
1132 | if isinstance (counter, int) is False: | |
1133 | raise InvalidParameter ("next: the supplied counter is of " | |
1134 | "invalid type %s; please pass an " | |
1135 | "integer instead" % type (counter)) | |
1136 | self.set_object_counter (counter) | |
fac2cfe1 | 1137 | |
50710d86 | 1138 | self.iv = self.iv_make () |
72a42219 | 1139 | if self.paramenc == "aes-gcm": |
6178061e PG |
1140 | self.enc = Cipher \ |
1141 | ( algorithms.AES (self.key) | |
1142 | , modes.GCM (self.iv) | |
1143 | , backend = default_backend ()) \ | |
1144 | .encryptor () | |
72a42219 | 1145 | elif self.paramenc == "passthrough": |
6178061e PG |
1146 | self.enc = PassthroughCipher () |
1147 | else: | |
b12110dd PG |
1148 | raise InvalidParameter ("next: parameter version %d not known" |
1149 | % self.paramversion) | |
48db09ba PG |
1150 | hdrdum = hdr_make_dummy (filename) |
1151 | self.lastinfo = (filename, hdrdum) | |
30019abf | 1152 | super().next (self.password, self.paramversion, self.nacl, self.iv) |
72a42219 | 1153 | |
3031b7ae | 1154 | self.set_object_counter (self.cnt + 1) |
48db09ba | 1155 | return hdrdum |
a393d9cb | 1156 | |
a393d9cb | 1157 | |
cd77dadb | 1158 | def done (self, cmpdata): |
704ceaa5 PG |
1159 | """ |
1160 | Complete encryption of an object. After this has been called, attempts | |
1161 | of encrypting further data will cause an error until ``.next()`` is | |
1162 | invoked properly. | |
1163 | ||
1164 | Returns a 64 bytes buffer containing the object header including all | |
1165 | values including the “late” ones e. g. the ciphertext size and the | |
1166 | GCM tag. | |
1167 | """ | |
36b9932a PG |
1168 | if isinstance (cmpdata, bytes) is False: |
1169 | raise InvalidParameter ("done: comparison input expected as bytes, " | |
1170 | "not %s" % type (cmpdata)) | |
cb7a3911 PG |
1171 | if self.lastinfo is None: |
1172 | raise RuntimeError ("done: encryption context not initialized") | |
48db09ba PG |
1173 | filename, hdrdum = self.lastinfo |
1174 | if cmpdata != hdrdum: | |
b12110dd PG |
1175 | raise RuntimeError ("done: bad sync of header for object %d: " |
1176 | "preliminary data does not match; this likely " | |
1177 | "indicates a wrongly repositioned stream" | |
1178 | % self.cnt) | |
6178061e | 1179 | data = self.enc.finalize () |
633b18a9 | 1180 | self.stats ["out"] += len (data) |
cd77dadb | 1181 | self.ctsize += len (data) |
48db09ba PG |
1182 | ok, hdr = hdr_from_params (self.version, self.paramversion, self.nacl, |
1183 | self.iv, self.ctsize, self.enc.tag) | |
8a990744 | 1184 | if ok is False: |
b12110dd PG |
1185 | raise InternalError ("error constructing header: %r" % hdr) |
1186 | return data, hdr, self.fixed | |
a393d9cb | 1187 | |
a393d9cb | 1188 | |
cd77dadb | 1189 | def process (self, buf): |
704ceaa5 PG |
1190 | """ |
1191 | Encrypt a chunk of plaintext with the active encryptor. Returns the | |
1192 | size of the input consumed. This **must** be checked downstream. If the | |
1193 | maximum possible object size has been reached, the current context must | |
1194 | be finalized and a new one established before any further data can be | |
1195 | encrypted. The second argument is the remainder of the plaintext that | |
1196 | was not encrypted for the caller to use immediately after the new | |
1197 | context is ready. | |
1198 | """ | |
36b9932a PG |
1199 | if isinstance (buf, bytes) is False: |
1200 | raise InvalidParameter ("process: expected byte buffer, not %s" | |
1201 | % type (buf)) | |
cb7a3911 PG |
1202 | bsize = len (buf) |
1203 | newptsize = self.ptsize + bsize | |
1204 | diff = newptsize - PDTCRYPT_MAX_OBJ_SIZE | |
1205 | if diff > 0: | |
1206 | bsize -= diff | |
1207 | newptsize = PDTCRYPT_MAX_OBJ_SIZE | |
1208 | self.ptsize = newptsize | |
1209 | data = super().process (buf [:bsize]) | |
cd77dadb | 1210 | self.ctsize += len (data) |
cb7a3911 | 1211 | return bsize, data |
cd77dadb PG |
1212 | |
1213 | ||
39accaaa | 1214 | class Decrypt (Crypto): |
a393d9cb | 1215 | |
3031b7ae | 1216 | tag = None # GCM tag, part of header |
3031b7ae | 1217 | last_iv = None # check consecutive ivs in strict mode |
39accaaa | 1218 | |
1f3fd7b0 | 1219 | def __init__ (self, password=None, key=None, counter=None, fixedparts=None, |
ee6aa239 | 1220 | strict_ivs=False): |
704ceaa5 PG |
1221 | """ |
1222 | Sanitizing ctor for the decryption context. ``fixedparts`` specifies a | |
1223 | list of IV fixed parts accepted during decryption. If a fixed part is | |
1224 | encountered that is not in the list, decryption will fail. | |
1225 | ||
1226 | :param password: mutually exclusive with ``key`` | |
1227 | :type password: bytes | |
1228 | :param key: mutually exclusive with ``password`` | |
1229 | :type key: bytes | |
1230 | :type counter: initial object counter the values | |
1231 | ``AES_GCM_IV_CNT_INFOFILE`` and | |
1232 | ``AES_GCM_IV_CNT_INDEX`` are unique in each backup set | |
1233 | and cannot be reused even with different fixed parts. | |
1234 | :type fixedparts: bytes list | |
1235 | """ | |
1f3fd7b0 PG |
1236 | if password is None and key is None \ |
1237 | or password is not None and key is not None : | |
1238 | raise InvalidParameter ("__init__: need either key or password") | |
1239 | ||
1240 | if key is not None: | |
1241 | if isinstance (key, bytes) is False: | |
1242 | raise InvalidParameter ("__init__: key must be provided as " | |
1243 | "bytes, not %s" % type (key)) | |
1244 | else: # password, no key | |
1245 | if isinstance (password, str) is False: | |
1246 | raise InvalidParameter ("__init__: password must be a string, not %s" | |
1247 | % type (password)) | |
1248 | if len (password) == 0: | |
1249 | raise InvalidParameter ("__init__: supplied empty password but not " | |
1250 | "permitted for PDT encrypted files") | |
36b9932a | 1251 | # fixed parts |
50710d86 | 1252 | if fixedparts is not None: |
36b9932a PG |
1253 | if isinstance (fixedparts, list) is False: |
1254 | raise InvalidParameter ("__init__: IV fixed parts must be " | |
1255 | "supplied as list, not %s" | |
1256 | % type (fixedparts)) | |
b12110dd PG |
1257 | self.fixed = fixedparts |
1258 | self.fixed.sort () | |
ee6aa239 | 1259 | |
a83fa4ed PG |
1260 | super().__init__ (password=password, key=key, counter=counter, |
1261 | strict_ivs=strict_ivs) | |
39accaaa PG |
1262 | |
1263 | ||
b12110dd | 1264 | def valid_fixed_part (self, iv): |
704ceaa5 PG |
1265 | """ |
1266 | Check if a fixed part was already seen. | |
1267 | """ | |
50710d86 | 1268 | # check if fixed part is known |
b12110dd PG |
1269 | fixed, _cnt = struct.unpack (FMT_I2N_IV, iv) |
1270 | i = bisect.bisect_left (self.fixed, fixed) | |
1271 | return i != len (self.fixed) and self.fixed [i] == fixed | |
50710d86 PG |
1272 | |
1273 | ||
ee6aa239 | 1274 | def check_consecutive_iv (self, iv): |
704ceaa5 PG |
1275 | """ |
1276 | Check whether the counter part of the given IV is indeed the successor | |
1277 | of the currently present counter. This should always be the case for | |
1278 | the objects in a well formed PDT archive but should not be enforced | |
1279 | when decrypting out-of-order. | |
1280 | """ | |
ee6aa239 | 1281 | fixed, cnt = struct.unpack (FMT_I2N_IV, iv) |
3031b7ae PG |
1282 | if self.strict_ivs is True \ |
1283 | and self.last_iv is not None \ | |
ee6aa239 PG |
1284 | and self.last_iv [0] == fixed \ |
1285 | and self.last_iv [1] != cnt - 1: | |
f6cd676f | 1286 | raise NonConsecutiveIV ("iv %s counter not successor of " |
ee6aa239 | 1287 | "last object (expected %d, found %d)" |
f6cd676f | 1288 | % (iv_fmt (self.last_iv [1]), cnt)) |
ee6aa239 PG |
1289 | self.last_iv = (iv, cnt) |
1290 | ||
1291 | ||
79782fa9 | 1292 | def next (self, hdr): |
704ceaa5 PG |
1293 | """ |
1294 | Start decrypting the next object. The PDTCRYPT header for the object | |
1295 | can be given either as already parsed object or as bytes. | |
1296 | """ | |
dccfe104 PG |
1297 | if isinstance (hdr, bytes) is True: |
1298 | hdr = hdr_read (hdr) | |
36b9932a PG |
1299 | elif isinstance (hdr, dict) is False: |
1300 | # this won’t catch malformed specs though | |
1301 | raise InvalidParameter ("next: wrong type of parameter hdr: " | |
1302 | "expected bytes or spec, got %s" | |
fbfda3d4 | 1303 | % type (hdr)) |
36b9932a PG |
1304 | try: |
1305 | paramversion = hdr ["paramversion"] | |
1306 | nacl = hdr ["nacl"] | |
1307 | iv = hdr ["iv"] | |
1308 | tag = hdr ["tag"] | |
1309 | except KeyError: | |
1310 | raise InvalidHeader ("next: not a header %r" % hdr) | |
1311 | ||
30019abf | 1312 | super().next (self.password, paramversion, nacl, iv) |
b12110dd | 1313 | if self.fixed is not None and self.valid_fixed_part (iv) is False: |
f6cd676f PG |
1314 | raise InvalidIVFixedPart ("iv %s has invalid fixed part" |
1315 | % iv_fmt (iv)) | |
3031b7ae | 1316 | self.check_consecutive_iv (iv) |
ee6aa239 | 1317 | |
36b9932a | 1318 | self.tag = tag |
b12110dd PG |
1319 | defs = ENCRYPTION_PARAMETERS.get (paramversion, None) |
1320 | if defs is None: | |
1321 | raise FormatError ("header contains unknown parameter version %d; " | |
1322 | "maybe the file was created by a more recent " | |
1323 | "version of Deltatar" % paramversion) | |
50710d86 | 1324 | enc = defs ["enc"] |
6178061e PG |
1325 | if enc == "aes-gcm": |
1326 | self.enc = Cipher \ | |
1327 | ( algorithms.AES (self.key) | |
36b9932a | 1328 | , modes.GCM (iv, tag=self.tag) |
6178061e PG |
1329 | , backend = default_backend ()) \ |
1330 | . decryptor () | |
1331 | elif enc == "passthrough": | |
1332 | self.enc = PassthroughCipher () | |
1333 | else: | |
b12110dd PG |
1334 | raise InternalError ("encryption parameter set %d refers to unknown " |
1335 | "mode %r" % (paramversion, enc)) | |
f484f2d1 | 1336 | self.set_object_counter (self.cnt + 1) |
39accaaa PG |
1337 | |
1338 | ||
db1f3ac7 | 1339 | def done (self, tag=None): |
704ceaa5 PG |
1340 | """ |
1341 | Stop decryption of the current object and finalize it with the active | |
1342 | context. This will throw an *InvalidGCMTag* exception to indicate that | |
1343 | the authentication tag does not match the data. If the tag is correct, | |
1344 | the rest of the plaintext is returned. | |
1345 | """ | |
633b18a9 | 1346 | data = b"" |
db1f3ac7 PG |
1347 | try: |
1348 | if tag is None: | |
f484f2d1 | 1349 | data = self.enc.finalize () |
db1f3ac7 | 1350 | else: |
36b9932a PG |
1351 | if isinstance (tag, bytes) is False: |
1352 | raise InvalidParameter ("done: wrong type of parameter " | |
1353 | "tag: expected bytes, got %s" | |
1354 | % type (tag)) | |
f484f2d1 | 1355 | data = self.enc.finalize_with_tag (self.tag) |
b0078f26 | 1356 | except cryptography.exceptions.InvalidTag: |
f08c604b | 1357 | raise InvalidGCMTag ("done: tag mismatch of object %d: %s " |
b0078f26 | 1358 | "rejected by finalize ()" |
f08c604b | 1359 | % (self.cnt, binascii.hexlify (self.tag))) |
50710d86 | 1360 | self.ctsize += len (data) |
633b18a9 | 1361 | self.stats ["out"] += len (data) |
b0078f26 | 1362 | return data |
00b3cd10 PG |
1363 | |
1364 | ||
47e27926 | 1365 | def process (self, buf): |
704ceaa5 PG |
1366 | """ |
1367 | Decrypt the bytes object *buf* with the active decryptor. | |
1368 | """ | |
36b9932a PG |
1369 | if isinstance (buf, bytes) is False: |
1370 | raise InvalidParameter ("process: expected byte buffer, not %s" | |
1371 | % type (buf)) | |
47e27926 PG |
1372 | self.ctsize += len (buf) |
1373 | data = super().process (buf) | |
1374 | self.ptsize += len (data) | |
1375 | return data | |
1376 | ||
1377 | ||
00b3cd10 | 1378 | ############################################################################### |
770173c5 PG |
1379 | ## testing helpers |
1380 | ############################################################################### | |
1381 | ||
cb7a3911 | 1382 | def _patch_global (glob, vow, n=None): |
770173c5 PG |
1383 | """ |
1384 | Adapt upper file counter bound for testing IV logic. Completely unsafe. | |
1385 | """ | |
1386 | assert vow == "I am fully aware that this will void my warranty." | |
cb7a3911 PG |
1387 | r = globals () [glob] |
1388 | if n is None: | |
1389 | n = globals () [glob + "_DEFAULT"] | |
1390 | globals () [glob] = n | |
770173c5 PG |
1391 | return r |
1392 | ||
cb7a3911 PG |
1393 | _testing_set_AES_GCM_IV_CNT_MAX = \ |
1394 | partial (_patch_global, "AES_GCM_IV_CNT_MAX") | |
1395 | ||
1396 | _testing_set_PDTCRYPT_MAX_OBJ_SIZE = \ | |
1397 | partial (_patch_global, "PDTCRYPT_MAX_OBJ_SIZE") | |
1398 | ||
a808459e PG |
1399 | def open2_dump_file (fname, dir_fd, force=False): |
1400 | outfd = -1 | |
1401 | ||
1402 | oflags = os.O_CREAT | os.O_WRONLY | |
6690f5e0 | 1403 | if force is True: |
a808459e PG |
1404 | oflags |= os.O_TRUNC |
1405 | else: | |
1406 | oflags |= os.O_EXCL | |
1407 | ||
1408 | try: | |
1409 | outfd = os.open (fname, oflags, | |
1410 | stat.S_IRUSR | stat.S_IWUSR, dir_fd=dir_fd) | |
1411 | except FileExistsError as exn: | |
1412 | noise ("PDT: refusing to overwrite existing file %s" % fname) | |
1413 | noise ("") | |
1414 | raise RuntimeError ("destination file %s already exists" % fname) | |
1415 | if PDTCRYPT_VERBOSE is True: | |
1416 | noise ("PDT: new output file %s (fd=%d)" % (fname, outfd)) | |
1417 | ||
1418 | return outfd | |
1419 | ||
770173c5 | 1420 | ############################################################################### |
00b3cd10 PG |
1421 | ## freestanding invocation |
1422 | ############################################################################### | |
1423 | ||
da82bc58 PG |
1424 | PDTCRYPT_SUB_PROCESS = 0 |
1425 | PDTCRYPT_SUB_SCRYPT = 1 | |
f41973a6 | 1426 | PDTCRYPT_SUB_SCAN = 2 |
da82bc58 PG |
1427 | |
1428 | PDTCRYPT_SUB = \ | |
1429 | { "process" : PDTCRYPT_SUB_PROCESS | |
f41973a6 PG |
1430 | , "scrypt" : PDTCRYPT_SUB_SCRYPT |
1431 | , "scan" : PDTCRYPT_SUB_SCAN } | |
da82bc58 | 1432 | |
a83fa4ed PG |
1433 | PDTCRYPT_SECRET_PW = 0 |
1434 | PDTCRYPT_SECRET_KEY = 1 | |
1435 | ||
e3abcdf0 PG |
1436 | PDTCRYPT_DECRYPT = 1 << 0 # decrypt archive with password |
1437 | PDTCRYPT_SPLIT = 1 << 1 # split archive into individual objects | |
da82bc58 | 1438 | PDTCRYPT_HASH = 1 << 2 # output scrypt hash for file and given password |
e3abcdf0 | 1439 | |
a808459e PG |
1440 | PDTCRYPT_SPLITNAME = "pdtcrypt-object-%d.bin" |
1441 | PDTCRYPT_RESCUENAME = "pdtcrypt-rescue-object-%0.5d.bin" | |
e3abcdf0 | 1442 | |
70ad9458 | 1443 | PDTCRYPT_VERBOSE = False |
ee6aa239 | 1444 | PDTCRYPT_STRICTIVS = False |
b07633d3 | 1445 | PDTCRYPT_OVERWRITE = False |
15d3eefd | 1446 | PDTCRYPT_BLOCKSIZE = 1 << 12 |
70ad9458 PG |
1447 | PDTCRYPT_SINK = 0 |
1448 | PDTCRYPT_SOURCE = 1 | |
1449 | SELF = None | |
1450 | ||
77058bab PG |
1451 | PDTCRYPT_DEFAULT_VER = 1 |
1452 | PDTCRYPT_DEFAULT_PVER = 1 | |
1453 | ||
7b3940e5 PG |
1454 | # scrypt hashing output control |
1455 | PDTCRYPT_SCRYPT_INTRANATOR = 0 | |
1456 | PDTCRYPT_SCRYPT_PARAMETERS = 1 | |
4f6405d6 | 1457 | PDTCRYPT_SCRYPT_DEFAULT = PDTCRYPT_SCRYPT_INTRANATOR |
7b3940e5 PG |
1458 | |
1459 | PDTCRYPT_SCRYPT_FORMAT = \ | |
1460 | { "i2n" : PDTCRYPT_SCRYPT_INTRANATOR | |
1461 | , "params" : PDTCRYPT_SCRYPT_PARAMETERS } | |
1462 | ||
4c62ddc0 | 1463 | PDTCRYPT_TT_COLUMNS = 80 # assume standard terminal |
15d3eefd PG |
1464 | |
1465 | class PDTDecryptionError (Exception): | |
1466 | """Decryption failed.""" | |
1467 | ||
e3abcdf0 PG |
1468 | class PDTSplitError (Exception): |
1469 | """Decryption failed.""" | |
1470 | ||
15d3eefd PG |
1471 | |
1472 | def noise (*a, **b): | |
591a722f | 1473 | print (file=sys.stderr, *a, **b) |
15d3eefd PG |
1474 | |
1475 | ||
89e1073c PG |
1476 | class PassthroughDecryptor (object): |
1477 | ||
1478 | curhdr = None # write current header on first data write | |
1479 | ||
1480 | def __init__ (self): | |
1481 | if PDTCRYPT_VERBOSE is True: | |
1482 | noise ("PDT: no encryption; data passthrough") | |
1483 | ||
1484 | def next (self, hdr): | |
1485 | ok, curhdr = hdr_make (hdr) | |
1486 | if ok is False: | |
1487 | raise PDTDecryptionError ("bad header %r" % hdr) | |
1488 | self.curhdr = curhdr | |
1489 | ||
1490 | def done (self): | |
1491 | if self.curhdr is not None: | |
1492 | return self.curhdr | |
1493 | return b"" | |
1494 | ||
1495 | def process (self, d): | |
1496 | if self.curhdr is not None: | |
1497 | d = self.curhdr + d | |
1498 | self.curhdr = None | |
1499 | return d | |
1500 | ||
1501 | ||
a83fa4ed | 1502 | def depdtcrypt (mode, secret, ins, outs): |
15d3eefd | 1503 | """ |
a83fa4ed PG |
1504 | Remove PDTCRYPT layer from all objects encrypted with the secret. Used on a |
1505 | Deltatar backup this will yield a (possibly Gzip compressed) tarball. | |
15d3eefd PG |
1506 | """ |
1507 | ctleft = -1 # length of ciphertext to consume | |
1508 | ctcurrent = 0 # total ciphertext of current object | |
15d3eefd PG |
1509 | total_obj = 0 # total number of objects read |
1510 | total_pt = 0 # total plaintext bytes | |
1511 | total_ct = 0 # total ciphertext bytes | |
1512 | total_read = 0 # total bytes read | |
e3abcdf0 PG |
1513 | outfile = None # Python file object for output |
1514 | ||
89e1073c | 1515 | if mode & PDTCRYPT_DECRYPT: # decryptor |
a83fa4ed PG |
1516 | ks = secret [0] |
1517 | if ks == PDTCRYPT_SECRET_PW: | |
1518 | decr = Decrypt (password=secret [1], strict_ivs=PDTCRYPT_STRICTIVS) | |
1519 | elif ks == PDTCRYPT_SECRET_KEY: | |
1520 | key = binascii.unhexlify (secret [1]) | |
1521 | decr = Decrypt (key=key, strict_ivs=PDTCRYPT_STRICTIVS) | |
1522 | else: | |
1523 | raise InternalError ("‘%d’ does not specify a valid kind of secret" | |
1524 | % ks) | |
89e1073c PG |
1525 | else: |
1526 | decr = PassthroughDecryptor () | |
1527 | ||
e3abcdf0 PG |
1528 | def nextout (_): |
1529 | """Dummy for non-split mode: output file does not vary.""" | |
1530 | return outs | |
1531 | ||
1532 | if mode & PDTCRYPT_SPLIT: | |
1533 | def nextout (outfile): | |
1534 | """ | |
1535 | We were passed an fd as outs for accessing the destination | |
1536 | directory where extracted archive components are supposed | |
1537 | to end up in. | |
1538 | """ | |
1539 | ||
1540 | if outfile is None: | |
1541 | if PDTCRYPT_VERBOSE is True: | |
1542 | noise ("PDT: no output file to close at this point") | |
77058bab PG |
1543 | else: |
1544 | if PDTCRYPT_VERBOSE is True: | |
1545 | noise ("PDT: release output file %r" % outfile) | |
e3abcdf0 PG |
1546 | # cleanup happens automatically by the GC; the next |
1547 | # line will error out on account of an invalid fd | |
1548 | #outfile.close () | |
1549 | ||
1550 | assert total_obj > 0 | |
1551 | fname = PDTCRYPT_SPLITNAME % total_obj | |
1552 | try: | |
a808459e PG |
1553 | outfd = open2_dump_file (fname, outs, force=PDTCRYPT_OVERWRITE) |
1554 | except RuntimeError as exn: | |
1555 | raise PDTSplitError (exn) | |
e3abcdf0 PG |
1556 | return os.fdopen (outfd, "wb", closefd=True) |
1557 | ||
15d3eefd | 1558 | |
47d22679 | 1559 | def tell (s): |
b09a99eb | 1560 | """ESPIPE is normal on non-seekable stdio stream.""" |
47d22679 PG |
1561 | try: |
1562 | return s.tell () | |
1563 | except OSError as exn: | |
b09a99eb | 1564 | if exn.errno == os.errno.ESPIPE: |
47d22679 PG |
1565 | return -1 |
1566 | ||
e3abcdf0 | 1567 | def out (pt, outfile): |
15d3eefd PG |
1568 | npt = len (pt) |
1569 | nonlocal total_pt | |
1570 | total_pt += npt | |
70ad9458 | 1571 | if PDTCRYPT_VERBOSE is True: |
15d3eefd PG |
1572 | noise ("PDT:\t· decrypt plaintext %d B" % (npt)) |
1573 | try: | |
e3abcdf0 | 1574 | nn = outfile.write (pt) |
15d3eefd PG |
1575 | except OSError as exn: # probably ENOSPC |
1576 | raise DecryptionError ("error (%s)" % exn) | |
1577 | if nn != npt: | |
1578 | raise DecryptionError ("write aborted after %d of %d B" % (nn, npt)) | |
1579 | ||
1580 | while True: | |
1581 | if ctleft <= 0: | |
1582 | # current object completed; in a valid archive this marks either | |
1583 | # the start of a new header or the end of the input | |
1584 | if ctleft == 0: # current object requires finalization | |
70ad9458 | 1585 | if PDTCRYPT_VERBOSE is True: |
47d22679 | 1586 | noise ("PDT: %d finalize" % tell (ins)) |
5d394c0d PG |
1587 | try: |
1588 | pt = decr.done () | |
1589 | except InvalidGCMTag as exn: | |
f08c604b PG |
1590 | raise DecryptionError ("error finalizing object %d (%d B): " |
1591 | "%r" % (total_obj, len (pt), exn)) \ | |
1592 | from exn | |
e3abcdf0 | 1593 | out (pt, outfile) |
70ad9458 | 1594 | if PDTCRYPT_VERBOSE is True: |
15d3eefd PG |
1595 | noise ("PDT:\t· object validated") |
1596 | ||
70ad9458 | 1597 | if PDTCRYPT_VERBOSE is True: |
47d22679 | 1598 | noise ("PDT: %d hdr" % tell (ins)) |
15d3eefd PG |
1599 | try: |
1600 | hdr = hdr_read_stream (ins) | |
dd47d6a2 | 1601 | total_read += PDTCRYPT_HDR_SIZE |
ae3d0f2a PG |
1602 | except EndOfFile as exn: |
1603 | total_read += exn.remainder | |
dd47d6a2 | 1604 | if total_ct + total_obj * PDTCRYPT_HDR_SIZE != total_read: |
15d3eefd PG |
1605 | raise PDTDecryptionError ("ciphertext processed (%d B) plus " |
1606 | "overhead (%d × %d B) does not match " | |
1607 | "the number of bytes read (%d )" | |
dd47d6a2 | 1608 | % (total_ct, total_obj, PDTCRYPT_HDR_SIZE, |
15d3eefd PG |
1609 | total_read)) |
1610 | # the single good exit | |
1611 | return total_read, total_obj, total_ct, total_pt | |
1612 | except InvalidHeader as exn: | |
1613 | raise PDTDecryptionError ("invalid header at position %d in %r " | |
ee6aa239 | 1614 | "(%s)" % (tell (ins), exn, ins)) |
70ad9458 | 1615 | if PDTCRYPT_VERBOSE is True: |
15d3eefd PG |
1616 | pretty = hdr_fmt_pretty (hdr) |
1617 | noise (reduce (lambda a, e: (a + "\n" if a else "") + "PDT:\t· " + e, | |
1618 | pretty.splitlines (), "")) | |
1619 | ctcurrent = ctleft = hdr ["ctsize"] | |
89e1073c | 1620 | |
15d3eefd | 1621 | decr.next (hdr) |
e3abcdf0 PG |
1622 | |
1623 | total_obj += 1 # used in file counter with split mode | |
1624 | ||
1625 | # finalization complete or skipped in case of first object in | |
1626 | # stream; create a new output file if necessary | |
1627 | outfile = nextout (outfile) | |
15d3eefd | 1628 | |
70ad9458 | 1629 | if PDTCRYPT_VERBOSE is True: |
15d3eefd | 1630 | noise ("PDT: %d decrypt obj no. %d, %d B" |
47d22679 | 1631 | % (tell (ins), total_obj, ctleft)) |
15d3eefd PG |
1632 | |
1633 | # always allocate a new buffer since python-cryptography doesn’t allow | |
1634 | # passing a bytearray :/ | |
1635 | nexpect = min (ctleft, PDTCRYPT_BLOCKSIZE) | |
70ad9458 | 1636 | if PDTCRYPT_VERBOSE is True: |
15d3eefd | 1637 | noise ("PDT:\t· [%d] %d%% done, read block (%d B of %d B remaining)" |
47d22679 | 1638 | % (tell (ins), |
15d3eefd PG |
1639 | 100 - ctleft * 100 / (ctcurrent > 0 and ctcurrent or 1), |
1640 | nexpect, ctleft)) | |
1641 | ct = ins.read (nexpect) | |
1642 | nct = len (ct) | |
1643 | if nct < nexpect: | |
47d22679 | 1644 | off = tell (ins) |
ae3d0f2a PG |
1645 | raise EndOfFile (nct, |
1646 | "hit EOF after %d of %d B in block [%d:%d); " | |
15d3eefd PG |
1647 | "%d B ciphertext remaining for object no %d" |
1648 | % (nct, nexpect, off, off + nexpect, ctleft, | |
1649 | total_obj)) | |
1650 | ctleft -= nct | |
1651 | total_ct += nct | |
1652 | total_read += nct | |
1653 | ||
70ad9458 | 1654 | if PDTCRYPT_VERBOSE is True: |
15d3eefd PG |
1655 | noise ("PDT:\t· decrypt ciphertext %d B" % (nct)) |
1656 | pt = decr.process (ct) | |
e3abcdf0 | 1657 | out (pt, outfile) |
15d3eefd | 1658 | |
d6c15a52 | 1659 | |
70ad9458 | 1660 | def deptdcrypt_mk_stream (kind, path): |
d6c15a52 | 1661 | """Create stream from file or stdio descriptor.""" |
70ad9458 | 1662 | if kind == PDTCRYPT_SINK: |
d6c15a52 | 1663 | if path == "-": |
70ad9458 | 1664 | if PDTCRYPT_VERBOSE is True: noise ("PDT: sink: stdout") |
d6c15a52 PG |
1665 | return sys.stdout.buffer |
1666 | else: | |
70ad9458 | 1667 | if PDTCRYPT_VERBOSE is True: noise ("PDT: sink: file %s" % path) |
d6c15a52 | 1668 | return io.FileIO (path, "w") |
70ad9458 | 1669 | if kind == PDTCRYPT_SOURCE: |
d6c15a52 | 1670 | if path == "-": |
70ad9458 | 1671 | if PDTCRYPT_VERBOSE is True: noise ("PDT: source: stdin") |
d6c15a52 PG |
1672 | return sys.stdin.buffer |
1673 | else: | |
70ad9458 | 1674 | if PDTCRYPT_VERBOSE is True: noise ("PDT: source: file %s" % path) |
d6c15a52 PG |
1675 | return io.FileIO (path, "r") |
1676 | ||
1677 | raise ValueError ("bogus stream “%s” / %s" % (kind, path)) | |
1678 | ||
15d3eefd | 1679 | |
a83fa4ed | 1680 | def mode_depdtcrypt (mode, secret, ins, outs): |
da82bc58 PG |
1681 | try: |
1682 | total_read, total_obj, total_ct, total_pt = \ | |
a83fa4ed | 1683 | depdtcrypt (mode, secret, ins, outs) |
da82bc58 PG |
1684 | except DecryptionError as exn: |
1685 | noise ("PDT: Decryption failed:") | |
1686 | noise ("PDT:") | |
1687 | noise ("PDT: “%s”" % exn) | |
1688 | noise ("PDT:") | |
a83fa4ed | 1689 | noise ("PDT: Did you specify the correct key / password?") |
da82bc58 PG |
1690 | noise ("") |
1691 | return 1 | |
1692 | except PDTSplitError as exn: | |
1693 | noise ("PDT: Split operation failed:") | |
1694 | noise ("PDT:") | |
1695 | noise ("PDT: “%s”" % exn) | |
1696 | noise ("PDT:") | |
a83fa4ed | 1697 | noise ("PDT: Hint: target directory should be empty.") |
da82bc58 PG |
1698 | noise ("") |
1699 | return 1 | |
1700 | ||
1701 | if PDTCRYPT_VERBOSE is True: | |
1702 | noise ("PDT: decryption successful" ) | |
1703 | noise ("PDT: %.10d bytes read" % total_read) | |
1704 | noise ("PDT: %.10d objects decrypted" % total_obj ) | |
1705 | noise ("PDT: %.10d bytes ciphertext" % total_ct ) | |
1706 | noise ("PDT: %.10d bytes plaintext" % total_pt ) | |
1707 | noise ("" ) | |
1708 | ||
1709 | return 0 | |
1710 | ||
1711 | ||
7b3940e5 | 1712 | def mode_scrypt (pw, ins=None, nacl=None, fmt=PDTCRYPT_SCRYPT_INTRANATOR): |
77058bab | 1713 | hsh = None |
7b3940e5 | 1714 | paramversion = PDTCRYPT_DEFAULT_PVER |
77058bab PG |
1715 | if ins is not None: |
1716 | hsh, nacl, version, paramversion = scrypt_hashsource (pw, ins) | |
1717 | defs = ENCRYPTION_PARAMETERS.get(paramversion, None) | |
1718 | else: | |
1719 | nacl = binascii.unhexlify (nacl) | |
7b3940e5 | 1720 | defs = ENCRYPTION_PARAMETERS.get(paramversion, None) |
77058bab PG |
1721 | version = PDTCRYPT_DEFAULT_VER |
1722 | ||
1723 | kdfname, params = defs ["kdf"] | |
1724 | if hsh is None: | |
1725 | kdf = kdf_by_version (None, defs) | |
1726 | hsh, _void = kdf (pw, nacl) | |
da82bc58 PG |
1727 | |
1728 | import json | |
7b3940e5 PG |
1729 | |
1730 | if fmt == PDTCRYPT_SCRYPT_INTRANATOR: | |
1731 | out = json.dumps ({ "salt" : base64.b64encode (nacl).decode () | |
1732 | , "key" : base64.b64encode (hsh) .decode () | |
1733 | , "paramversion" : paramversion }) | |
1734 | elif fmt == PDTCRYPT_SCRYPT_PARAMETERS: | |
1735 | out = json.dumps ({ "salt" : binascii.hexlify (nacl).decode () | |
1736 | , "key" : binascii.hexlify (hsh) .decode () | |
1737 | , "version" : version | |
1738 | , "scrypt_params" : { "N" : params ["N"] | |
1739 | , "r" : params ["r"] | |
1740 | , "p" : params ["p"] | |
1741 | , "dkLen" : params ["dkLen"] } }) | |
1742 | else: | |
1743 | raise RuntimeError ("bad scrypt output scheme %r" % fmt) | |
1744 | ||
da82bc58 PG |
1745 | print (out) |
1746 | ||
1747 | ||
4c62ddc0 PG |
1748 | def noise_output_candidates (cands, indent=8, cols=PDTCRYPT_TT_COLUMNS): |
1749 | """ | |
1750 | Print a list of offsets without garbling the terminal too much. | |
1751 | ||
1752 | The indent is counted from column zero; if it is wide enough, the “PDT: ” | |
1753 | marker will be prepended, considered part of the indentation. | |
1754 | """ | |
1755 | wd = cols - 1 | |
1756 | nc = len (cands) | |
1757 | idt = " " * indent if indent < 5 else "PDT: " + " " * (indent - 5) | |
1758 | line = idt | |
1759 | lpos = indent | |
1760 | sep = "," | |
1761 | lsep = len (sep) | |
1762 | init = True # prevent leading separator | |
1763 | ||
1764 | if indent >= wd: | |
1765 | raise ValueError ("the requested indentation exceeds the line " | |
1766 | "width by %d" % (indent - wd)) | |
1767 | ||
1768 | for n in cands: | |
1769 | ns = "%d" % n | |
1770 | lns = len (ns) | |
1771 | if init is False: | |
1772 | line += sep | |
1773 | lpos += lsep | |
1774 | ||
1775 | lpos += lns | |
1776 | if lpos > wd: # line break | |
1777 | noise (line) | |
1778 | line = idt | |
1779 | lpos = indent + lns | |
1780 | elif init is True: | |
1781 | init = False | |
1782 | else: # space | |
1783 | line += ' ' | |
1784 | lpos += 1 | |
1785 | ||
1786 | line += ns | |
1787 | ||
1788 | if lpos != indent: | |
1789 | noise (line) | |
1790 | ||
1791 | ||
a808459e | 1792 | def mode_scan (secret, fname, outs=None, nacl=None): |
f41973a6 PG |
1793 | """ |
1794 | Dissect a binary file, looking for PDTCRYPT headers and objects. | |
a808459e PG |
1795 | |
1796 | If *outs* is supplied, recoverable data will be dumped into the specified | |
1797 | directory. | |
f41973a6 PG |
1798 | """ |
1799 | try: | |
a808459e | 1800 | ifd = os.open (fname, os.O_RDONLY) |
f41973a6 PG |
1801 | except FileNotFoundError: |
1802 | noise ("PDT: failed to open %s readonly" % fname) | |
1803 | noise ("") | |
1804 | usage (err=True) | |
1805 | ||
1806 | try: | |
1807 | if PDTCRYPT_VERBOSE is True: | |
1808 | noise ("PDT: scan for potential sync points") | |
a808459e | 1809 | cands = locate_hdr_candidates (ifd) |
f41973a6 PG |
1810 | if len (cands) == 0: |
1811 | noise ("PDT: scan complete: input does not contain potential PDT " | |
1812 | "headers; giving up.") | |
1813 | return -1 | |
1814 | if PDTCRYPT_VERBOSE is True: | |
4c62ddc0 PG |
1815 | noise ("PDT: scan complete: found %d candidates:" % len (cands)) |
1816 | noise_output_candidates (cands) | |
6c8073ab | 1817 | except: |
a808459e | 1818 | os.close (ifd) |
6c8073ab | 1819 | raise |
f41973a6 | 1820 | |
6c8073ab PG |
1821 | junk, todo = [], [] |
1822 | try: | |
a808459e | 1823 | nobj = 0 |
6c8073ab | 1824 | for cand in cands: |
a808459e PG |
1825 | nobj += 1 |
1826 | vdt, hdr = inspect_hdr (ifd, cand) | |
6c8073ab PG |
1827 | if vdt == HDR_CAND_JUNK: |
1828 | junk.append (cand) | |
1829 | else: | |
1830 | off0 = cand + PDTCRYPT_HDR_SIZE | |
1831 | if PDTCRYPT_VERBOSE is True: | |
a808459e | 1832 | noise ("PDT: obj %d: read payload @%d" % (nobj, off0)) |
70a33834 PG |
1833 | pretty = hdr_fmt_pretty (hdr) |
1834 | noise (reduce (lambda a, e: (a + "\n" if a else "") + "PDT:\t· " + e, | |
1835 | pretty.splitlines (), "")) | |
6c8073ab | 1836 | |
a808459e PG |
1837 | ofd = -1 |
1838 | if outs is not None: | |
1839 | ofname = PDTCRYPT_RESCUENAME % nobj | |
1840 | ofd = open2_dump_file (ofname, outs, force=PDTCRYPT_OVERWRITE) | |
1841 | ||
1842 | try: | |
1843 | ok = try_decrypt (ifd, off0, hdr, secret, ofd=ofd) == hdr ["ctsize"] | |
1844 | finally: | |
1845 | if ofd != -1: | |
1846 | os.close (ofd) | |
70a33834 | 1847 | if vdt == HDR_CAND_GOOD and ok is True: |
6c8073ab PG |
1848 | noise ("PDT: %d → ✓ valid object %d–%d" |
1849 | % (cand, off0, off0 + hdr ["ctsize"])) | |
70a33834 | 1850 | elif vdt == HDR_CAND_FISHY and ok is True: |
6c8073ab PG |
1851 | noise ("PDT: %d → × object %d–%d, corrupt header" |
1852 | % (cand, off0, off0 + hdr ["ctsize"])) | |
70a33834 | 1853 | elif vdt == HDR_CAND_GOOD and ok is False: |
6c8073ab PG |
1854 | noise ("PDT: %d → × object %d–%d, problematic payload" |
1855 | % (cand, off0, off0 + hdr ["ctsize"])) | |
70a33834 | 1856 | elif vdt == HDR_CAND_FISHY and ok is False: |
6c8073ab PG |
1857 | noise ("PDT: %d → × object %d–%d, corrupt header, problematic " |
1858 | "ciphertext" % (cand, off0, off0 + hdr ["ctsize"])) | |
1859 | else: | |
1860 | raise Unreachable | |
1861 | finally: | |
a808459e | 1862 | os.close (ifd) |
7b3940e5 | 1863 | |
70a33834 PG |
1864 | if len (junk) == 0: |
1865 | noise ("PDT: all headers ok") | |
1866 | else: | |
1867 | noise ("PDT: %d candidates not parseable as headers:" % len (junk)) | |
1868 | noise_output_candidates (junk) | |
1869 | ||
70ad9458 PG |
1870 | def usage (err=False): |
1871 | out = print | |
1872 | if err is True: | |
1873 | out = noise | |
5afcb45d | 1874 | indent = ' ' * len (SELF) |
da82bc58 | 1875 | out ("usage: %s SUBCOMMAND { --help" % SELF) |
5afcb45d | 1876 | out (" %s | [ -v ] { -p PASSWORD | -k KEY }" % indent) |
77058bab PG |
1877 | out (" %s [ { -i | --in } { - | SOURCE } ]" % indent) |
1878 | out (" %s [ { -n | --nacl } { SALT } ]" % indent) | |
1879 | out (" %s [ { -o | --out } { - | DESTINATION } ]" % indent) | |
1880 | out (" %s [ -D | --no-decrypt ] [ -S | --split ]" % indent) | |
7b3940e5 | 1881 | out (" %s [ -f | --format ]" % indent) |
70ad9458 PG |
1882 | out ("") |
1883 | out ("\twhere") | |
da82bc58 PG |
1884 | out ("\t\tSUBCOMMAND main mode: { process | scrypt }") |
1885 | out ("\t\t where:") | |
1886 | out ("\t\t process: extract objects from PDT archive") | |
1887 | out ("\t\t scrypt: calculate hash from password and first object") | |
a83fa4ed PG |
1888 | out ("\t\t-p PASSWORD password to derive the encryption key from") |
1889 | out ("\t\t-k KEY encryption key as 16 bytes in hexadecimal notation") | |
e3abcdf0 | 1890 | out ("\t\t-s enforce strict handling of initialization vectors") |
70ad9458 PG |
1891 | out ("\t\t-i SOURCE file name to read from") |
1892 | out ("\t\t-o DESTINATION file to write output to") | |
77058bab | 1893 | out ("\t\t-n SALT provide salt for scrypt mode in hex encoding") |
70ad9458 | 1894 | out ("\t\t-v print extra info") |
e3abcdf0 PG |
1895 | out ("\t\t-S split into files at object boundaries; this") |
1896 | out ("\t\t requires DESTINATION to refer to directory") | |
1897 | out ("\t\t-D PDT header and ciphertext passthrough") | |
7b3940e5 | 1898 | out ("\t\t-f format of SCRYPT hash output (“default” or “parameters”)") |
70ad9458 PG |
1899 | out ("") |
1900 | out ("\tinstead of filenames, “-” may used to specify stdin / stdout") | |
1901 | out ("") | |
1902 | sys.exit ((err is True) and 42 or 0) | |
1903 | ||
1904 | ||
a83fa4ed PG |
1905 | def bail (msg): |
1906 | noise (msg) | |
1907 | noise ("") | |
1908 | usage (err=True) | |
1909 | raise Unreachable | |
1910 | ||
1911 | ||
70ad9458 | 1912 | def parse_argv (argv): |
6690f5e0 | 1913 | global PDTCRYPT_OVERWRITE |
70ad9458 | 1914 | global SELF |
7b3940e5 PG |
1915 | mode = PDTCRYPT_DECRYPT |
1916 | secret = None | |
1917 | insspec = None | |
1918 | outsspec = None | |
a808459e | 1919 | outs = None |
7b3940e5 | 1920 | nacl = None |
4f6405d6 | 1921 | scrypt_format = PDTCRYPT_SCRYPT_DEFAULT |
70ad9458 PG |
1922 | |
1923 | argvi = iter (argv) | |
1924 | SELF = os.path.basename (next (argvi)) | |
1925 | ||
da82bc58 PG |
1926 | try: |
1927 | rawsubcmd = next (argvi) | |
1928 | subcommand = PDTCRYPT_SUB [rawsubcmd] | |
1929 | except StopIteration: | |
a83fa4ed | 1930 | bail ("ERROR: subcommand required") |
da82bc58 | 1931 | except KeyError: |
a83fa4ed | 1932 | bail ("ERROR: invalid subcommand “%s” specified" % rawsubcmd) |
da82bc58 | 1933 | |
59d74e2b PG |
1934 | def checked_arg (): |
1935 | nonlocal argvi | |
1936 | try: | |
1937 | return next (argvi) | |
1938 | except StopIteration: | |
1939 | bail ("ERROR: argument list incomplete") | |
1940 | ||
a83fa4ed PG |
1941 | def checked_secret (t, arg): |
1942 | nonlocal secret | |
1943 | if secret is None: | |
1944 | secret = (t, arg) | |
da82bc58 | 1945 | else: |
a83fa4ed | 1946 | bail ("ERROR: encountered “%s” but secret already given" % arg) |
da82bc58 | 1947 | |
70ad9458 PG |
1948 | for arg in argvi: |
1949 | if arg in [ "-h", "--help" ]: | |
1950 | usage () | |
1951 | raise Unreachable | |
1952 | elif arg in [ "-v", "--verbose", "--wtf" ]: | |
1953 | global PDTCRYPT_VERBOSE | |
1954 | PDTCRYPT_VERBOSE = True | |
1955 | elif arg in [ "-i", "--in", "--source" ]: | |
59d74e2b | 1956 | insspec = checked_arg () |
70ad9458 | 1957 | if PDTCRYPT_VERBOSE is True: noise ("PDT: decrypt from %s" % insspec) |
a83fa4ed | 1958 | elif arg in [ "-p", "--password" ]: |
59d74e2b | 1959 | arg = checked_arg () |
a83fa4ed PG |
1960 | checked_secret (PDTCRYPT_SECRET_PW, arg) |
1961 | if PDTCRYPT_VERBOSE is True: noise ("PDT: decrypting with password") | |
70ad9458 | 1962 | else: |
da82bc58 PG |
1963 | if subcommand == PDTCRYPT_SUB_PROCESS: |
1964 | if arg in [ "-s", "--strict-ivs" ]: | |
1965 | global PDTCRYPT_STRICTIVS | |
1966 | PDTCRYPT_STRICTIVS = True | |
77058bab PG |
1967 | elif arg in [ "-o", "--out", "--dest", "--sink" ]: |
1968 | outsspec = checked_arg () | |
1969 | if PDTCRYPT_VERBOSE is True: noise ("PDT: decrypt to %s" % outsspec) | |
da82bc58 | 1970 | elif arg in [ "-f", "--force" ]: |
da82bc58 PG |
1971 | PDTCRYPT_OVERWRITE = True |
1972 | if PDTCRYPT_VERBOSE is True: noise ("PDT: overwrite existing files") | |
1973 | elif arg in [ "-S", "--split" ]: | |
1974 | mode |= PDTCRYPT_SPLIT | |
1975 | if PDTCRYPT_VERBOSE is True: noise ("PDT: split files") | |
1976 | elif arg in [ "-D", "--no-decrypt" ]: | |
1977 | mode &= ~PDTCRYPT_DECRYPT | |
1978 | if PDTCRYPT_VERBOSE is True: noise ("PDT: not decrypting") | |
a83fa4ed | 1979 | elif arg in [ "-k", "--key" ]: |
59d74e2b | 1980 | arg = checked_arg () |
a83fa4ed PG |
1981 | checked_secret (PDTCRYPT_SECRET_KEY, arg) |
1982 | if PDTCRYPT_VERBOSE is True: noise ("PDT: decrypting with key") | |
da82bc58 | 1983 | else: |
a83fa4ed | 1984 | bail ("ERROR: unexpected positional argument “%s”" % arg) |
da82bc58 | 1985 | elif subcommand == PDTCRYPT_SUB_SCRYPT: |
77058bab PG |
1986 | if arg in [ "-n", "--nacl", "--salt" ]: |
1987 | nacl = checked_arg () | |
1988 | if PDTCRYPT_VERBOSE is True: noise ("PDT: salt key with %s" % nacl) | |
7b3940e5 PG |
1989 | elif arg in [ "-f", "--format" ]: |
1990 | arg = checked_arg () | |
1991 | try: | |
1992 | scrypt_format = PDTCRYPT_SCRYPT_FORMAT [arg] | |
1993 | except KeyError: | |
1994 | bail ("ERROR: invalid scrypt output format %s" % arg) | |
1995 | if PDTCRYPT_VERBOSE is True: | |
1996 | noise ("PDT: scrypt output format “%s”" % scrypt_format) | |
77058bab PG |
1997 | else: |
1998 | bail ("ERROR: unexpected positional argument “%s”" % arg) | |
f41973a6 | 1999 | elif subcommand == PDTCRYPT_SUB_SCAN: |
a808459e PG |
2000 | if arg in [ "-o", "--out", "--dest", "--sink" ]: |
2001 | outsspec = checked_arg () | |
2002 | if PDTCRYPT_VERBOSE is True: noise ("PDT: decrypt to %s" % outsspec) | |
2003 | elif arg in [ "-f", "--force" ]: | |
a808459e PG |
2004 | PDTCRYPT_OVERWRITE = True |
2005 | if PDTCRYPT_VERBOSE is True: noise ("PDT: overwrite existing files") | |
2006 | else: | |
2007 | bail ("ERROR: unexpected positional argument “%s”" % arg) | |
70ad9458 | 2008 | |
a83fa4ed | 2009 | if secret is None: |
ecb9676d | 2010 | if PDTCRYPT_VERBOSE is True: |
a83fa4ed | 2011 | noise ("ERROR: no password or key specified, trying $PDTCRYPT_PASSWORD") |
ecb9676d PG |
2012 | epw = os.getenv ("PDTCRYPT_PASSWORD") |
2013 | if epw is not None: | |
a83fa4ed PG |
2014 | checked_secret (PDTCRYPT_SECRET_PW, epw.strip ()) |
2015 | ||
2016 | if secret is None: | |
2017 | if PDTCRYPT_VERBOSE is True: | |
2018 | noise ("ERROR: no password or key specified, trying $PDTCRYPT_KEY") | |
2019 | ek = os.getenv ("PDTCRYPT_KEY") | |
2020 | if ek is not None: | |
2021 | checked_secret (PDTCRYPT_SECRET_KEY, ek.strip ()) | |
ecb9676d | 2022 | |
a83fa4ed | 2023 | if secret is None: |
da82bc58 | 2024 | if subcommand == PDTCRYPT_SUB_SCRYPT: |
a83fa4ed | 2025 | bail ("ERROR: scrypt hash mode requested but no password given") |
da82bc58 | 2026 | elif mode & PDTCRYPT_DECRYPT: |
a83fa4ed PG |
2027 | bail ("ERROR: encryption requested but no password given") |
2028 | ||
a808459e PG |
2029 | if mode & PDTCRYPT_SPLIT and outsspec is None: |
2030 | bail ("ERROR: split mode is incompatible with stdout sink " | |
2031 | "(the default)") | |
2032 | ||
2033 | if subcommand == PDTCRYPT_SUB_SCAN and outsspec is None: | |
2034 | pass # no output by default in scan mode | |
2035 | elif mode & PDTCRYPT_SPLIT or subcommand == PDTCRYPT_SUB_SCAN: | |
2036 | # destination must be directory | |
2037 | if outsspec == "-": | |
2038 | bail ("ERROR: mode is incompatible with stdout sink") | |
2039 | try: | |
2040 | try: | |
2041 | os.makedirs (outsspec, 0o700) | |
2042 | except FileExistsError: | |
2043 | # if it’s a directory with appropriate perms, everything is | |
2044 | # good; otherwise, below invocation of open(2) will fail | |
2045 | pass | |
2046 | outs = os.open (outsspec, os.O_DIRECTORY, 0o600) | |
2047 | except FileNotFoundError as exn: | |
2048 | bail ("ERROR: cannot create target directory “%s”" % outsspec) | |
2049 | except NotADirectoryError as exn: | |
2050 | bail ("ERROR: target path “%s” is not a directory" % outsspec) | |
2051 | else: | |
2052 | outs = deptdcrypt_mk_stream (PDTCRYPT_SINK, outsspec or "-") | |
2053 | ||
f41973a6 PG |
2054 | if subcommand == PDTCRYPT_SUB_SCAN: |
2055 | if insspec is None: | |
2056 | bail ("ERROR: please supply an input file for scanning") | |
2057 | if insspec == '-': | |
2058 | bail ("ERROR: input must be seekable; please specify a file") | |
a808459e | 2059 | return True, partial (mode_scan, secret, insspec, outs, nacl=nacl) |
f41973a6 | 2060 | |
77058bab PG |
2061 | if subcommand == PDTCRYPT_SUB_SCRYPT: |
2062 | if secret [0] == PDTCRYPT_SECRET_KEY: | |
2063 | bail ("ERROR: scrypt mode requires a password") | |
2064 | if insspec is not None and nacl is not None \ | |
2065 | or insspec is None and nacl is None : | |
2066 | bail ("ERROR: please supply either an input file or " | |
2067 | "the salt") | |
70ad9458 PG |
2068 | |
2069 | # default to stdout | |
77058bab PG |
2070 | ins = None |
2071 | if insspec is not None or subcommand != PDTCRYPT_SUB_SCRYPT: | |
2072 | ins = deptdcrypt_mk_stream (PDTCRYPT_SOURCE, insspec or "-") | |
da82bc58 PG |
2073 | |
2074 | if subcommand == PDTCRYPT_SUB_SCRYPT: | |
7b3940e5 PG |
2075 | return True, partial (mode_scrypt, secret [1].encode (), ins, nacl, |
2076 | fmt=scrypt_format) | |
da82bc58 | 2077 | |
a83fa4ed | 2078 | return True, partial (mode_depdtcrypt, mode, secret, ins, outs) |
15d3eefd PG |
2079 | |
2080 | ||
00b3cd10 | 2081 | def main (argv): |
da82bc58 | 2082 | ok, runner = parse_argv (argv) |
f08c604b | 2083 | |
da82bc58 | 2084 | if ok is True: return runner () |
15d3eefd | 2085 | |
da82bc58 | 2086 | return 1 |
f08c604b | 2087 | |
00b3cd10 PG |
2088 | |
2089 | if __name__ == "__main__": | |
2090 | sys.exit (main (sys.argv)) | |
2091 |