author | Martin von Zweigbergk <martinvonz@google.com> |
Sat, 09 Feb 2019 16:46:11 -0800 | |
changeset 41668 | c4f023b656cf |
parent 36191 | 80301c90a2dc |
child 43076 | 2372284d9457 |
permissions | -rw-r--r-- |
7701 | 1 |
# base85.py: pure python base85 codec |
2 |
# |
|
3 |
# Copyright (C) 2009 Brendan Cully <brendan@kublai.com> |
|
4 |
# |
|
8225
46293a0c7e9f
updated license to be explicit about GPL version 2
Martin Geisler <mg@lazybytes.net>
parents:
7881
diff
changeset
|
5 |
# This software may be used and distributed according to the terms of the |
10263 | 6 |
# GNU General Public License version 2 or any later version. |
7701 | 7 |
|
27334
9007f697e8ef
base85: use absolute_import
Gregory Szorc <gregory.szorc@gmail.com>
parents:
16598
diff
changeset
|
8 |
from __future__ import absolute_import |
9007f697e8ef
base85: use absolute_import
Gregory Szorc <gregory.szorc@gmail.com>
parents:
16598
diff
changeset
|
9 |
|
7701 | 10 |
import struct |
11 |
||
35944
01b4d88ccb24
py3: use pycompat.bytestr to convert _b85chars to bytes
Pulkit Goyal <7895pulkit@gmail.com>
parents:
27334
diff
changeset
|
12 |
from .. import pycompat |
01b4d88ccb24
py3: use pycompat.bytestr to convert _b85chars to bytes
Pulkit Goyal <7895pulkit@gmail.com>
parents:
27334
diff
changeset
|
13 |
|
01b4d88ccb24
py3: use pycompat.bytestr to convert _b85chars to bytes
Pulkit Goyal <7895pulkit@gmail.com>
parents:
27334
diff
changeset
|
14 |
_b85chars = pycompat.bytestr("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdef" |
01b4d88ccb24
py3: use pycompat.bytestr to convert _b85chars to bytes
Pulkit Goyal <7895pulkit@gmail.com>
parents:
27334
diff
changeset
|
15 |
"ghijklmnopqrstuvwxyz!#$%&()*+-;<=>?@^_`{|}~") |
7835
2505e9f84153
Optimization of pure.base85.b85encode
Mads Kiilerich <mads@kiilerich.com>
parents:
7701
diff
changeset
|
16 |
_b85chars2 = [(a + b) for a in _b85chars for b in _b85chars] |
7701 | 17 |
_b85dec = {} |
18 |
||
19 |
def _mkb85dec(): |
|
8632
9e055cfdd620
replace "i in range(len(xs))" with "i, x in enumerate(xs)"
Martin Geisler <mg@lazybytes.net>
parents:
8225
diff
changeset
|
20 |
for i, c in enumerate(_b85chars): |
9e055cfdd620
replace "i in range(len(xs))" with "i, x in enumerate(xs)"
Martin Geisler <mg@lazybytes.net>
parents:
8225
diff
changeset
|
21 |
_b85dec[c] = i |
7701 | 22 |
|
23 |
def b85encode(text, pad=False): |
|
24 |
"""encode text in base85 format""" |
|
25 |
l = len(text) |
|
26 |
r = l % 4 |
|
27 |
if r: |
|
28 |
text += '\0' * (4 - r) |
|
29 |
longs = len(text) >> 2 |
|
30 |
words = struct.unpack('>%dL' % (longs), text) |
|
31 |
||
9029
0001e49f1c11
compat: use // for integer division
Alejandro Santos <alejolp@alejolp.com>
parents:
8632
diff
changeset
|
32 |
out = ''.join(_b85chars[(word // 52200625) % 85] + |
0001e49f1c11
compat: use // for integer division
Alejandro Santos <alejolp@alejolp.com>
parents:
8632
diff
changeset
|
33 |
_b85chars2[(word // 7225) % 7225] + |
7835
2505e9f84153
Optimization of pure.base85.b85encode
Mads Kiilerich <mads@kiilerich.com>
parents:
7701
diff
changeset
|
34 |
_b85chars2[word % 7225] |
2505e9f84153
Optimization of pure.base85.b85encode
Mads Kiilerich <mads@kiilerich.com>
parents:
7701
diff
changeset
|
35 |
for word in words) |
7701 | 36 |
|
37 |
if pad: |
|
38 |
return out |
|
39 |
||
40 |
# Trim padding |
|
41 |
olen = l % 4 |
|
42 |
if olen: |
|
43 |
olen += 1 |
|
9029
0001e49f1c11
compat: use // for integer division
Alejandro Santos <alejolp@alejolp.com>
parents:
8632
diff
changeset
|
44 |
olen += l // 4 * 5 |
7701 | 45 |
return out[:olen] |
46 |
||
47 |
def b85decode(text): |
|
48 |
"""decode base85-encoded text""" |
|
49 |
if not _b85dec: |
|
50 |
_mkb85dec() |
|
51 |
||
52 |
l = len(text) |
|
53 |
out = [] |
|
54 |
for i in range(0, len(text), 5): |
|
10282
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
10263
diff
changeset
|
55 |
chunk = text[i:i + 5] |
36191
80301c90a2dc
py3: converts bytes to pycompat.bytestr to get bytechrs while enumerating
Pulkit Goyal <7895pulkit@gmail.com>
parents:
35944
diff
changeset
|
56 |
chunk = pycompat.bytestr(chunk) |
7701 | 57 |
acc = 0 |
8632
9e055cfdd620
replace "i in range(len(xs))" with "i, x in enumerate(xs)"
Martin Geisler <mg@lazybytes.net>
parents:
8225
diff
changeset
|
58 |
for j, c in enumerate(chunk): |
7701 | 59 |
try: |
8632
9e055cfdd620
replace "i in range(len(xs))" with "i, x in enumerate(xs)"
Martin Geisler <mg@lazybytes.net>
parents:
8225
diff
changeset
|
60 |
acc = acc * 85 + _b85dec[c] |
7701 | 61 |
except KeyError: |
16598
20a9d823f242
pure/base85: align exception type/msg on base85.c
Patrick Mezard <patrick@mezard.eu>
parents:
10282
diff
changeset
|
62 |
raise ValueError('bad base85 character at position %d' |
20a9d823f242
pure/base85: align exception type/msg on base85.c
Patrick Mezard <patrick@mezard.eu>
parents:
10282
diff
changeset
|
63 |
% (i + j)) |
7701 | 64 |
if acc > 4294967295: |
16598
20a9d823f242
pure/base85: align exception type/msg on base85.c
Patrick Mezard <patrick@mezard.eu>
parents:
10282
diff
changeset
|
65 |
raise ValueError('Base85 overflow in hunk starting at byte %d' % i) |
7701 | 66 |
out.append(acc) |
67 |
||
68 |
# Pad final chunk if necessary |
|
69 |
cl = l % 5 |
|
70 |
if cl: |
|
71 |
acc *= 85 ** (5 - cl) |
|
72 |
if cl > 1: |
|
73 |
acc += 0xffffff >> (cl - 2) * 8 |
|
74 |
out[-1] = acc |
|
75 |
||
76 |
out = struct.pack('>%dL' % (len(out)), *out) |
|
77 |
if cl: |
|
78 |
out = out[:-(5 - cl)] |
|
79 |
||
80 |
return out |