author | Martin Geisler <mg@lazybytes.net> |
Tue, 19 Apr 2011 12:04:44 +0200 | |
changeset 13956 | ffb5c09ba822 |
parent 10282 | 08a0f04b56bd |
child 16598 | 20a9d823f242 |
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 |
|
8 |
import struct |
|
9 |
||
10 |
_b85chars = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ" \ |
|
11 |
"abcdefghijklmnopqrstuvwxyz!#$%&()*+-;<=>?@^_`{|}~" |
|
7835
2505e9f84153
Optimization of pure.base85.b85encode
Mads Kiilerich <mads@kiilerich.com>
parents:
7701
diff
changeset
|
12 |
_b85chars2 = [(a + b) for a in _b85chars for b in _b85chars] |
7701 | 13 |
_b85dec = {} |
14 |
||
15 |
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
|
16 |
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
|
17 |
_b85dec[c] = i |
7701 | 18 |
|
19 |
def b85encode(text, pad=False): |
|
20 |
"""encode text in base85 format""" |
|
21 |
l = len(text) |
|
22 |
r = l % 4 |
|
23 |
if r: |
|
24 |
text += '\0' * (4 - r) |
|
25 |
longs = len(text) >> 2 |
|
26 |
words = struct.unpack('>%dL' % (longs), text) |
|
27 |
||
9029
0001e49f1c11
compat: use // for integer division
Alejandro Santos <alejolp@alejolp.com>
parents:
8632
diff
changeset
|
28 |
out = ''.join(_b85chars[(word // 52200625) % 85] + |
0001e49f1c11
compat: use // for integer division
Alejandro Santos <alejolp@alejolp.com>
parents:
8632
diff
changeset
|
29 |
_b85chars2[(word // 7225) % 7225] + |
7835
2505e9f84153
Optimization of pure.base85.b85encode
Mads Kiilerich <mads@kiilerich.com>
parents:
7701
diff
changeset
|
30 |
_b85chars2[word % 7225] |
2505e9f84153
Optimization of pure.base85.b85encode
Mads Kiilerich <mads@kiilerich.com>
parents:
7701
diff
changeset
|
31 |
for word in words) |
7701 | 32 |
|
33 |
if pad: |
|
34 |
return out |
|
35 |
||
36 |
# Trim padding |
|
37 |
olen = l % 4 |
|
38 |
if olen: |
|
39 |
olen += 1 |
|
9029
0001e49f1c11
compat: use // for integer division
Alejandro Santos <alejolp@alejolp.com>
parents:
8632
diff
changeset
|
40 |
olen += l // 4 * 5 |
7701 | 41 |
return out[:olen] |
42 |
||
43 |
def b85decode(text): |
|
44 |
"""decode base85-encoded text""" |
|
45 |
if not _b85dec: |
|
46 |
_mkb85dec() |
|
47 |
||
48 |
l = len(text) |
|
49 |
out = [] |
|
50 |
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
|
51 |
chunk = text[i:i + 5] |
7701 | 52 |
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
|
53 |
for j, c in enumerate(chunk): |
7701 | 54 |
try: |
8632
9e055cfdd620
replace "i in range(len(xs))" with "i, x in enumerate(xs)"
Martin Geisler <mg@lazybytes.net>
parents:
8225
diff
changeset
|
55 |
acc = acc * 85 + _b85dec[c] |
7701 | 56 |
except KeyError: |
57 |
raise TypeError('Bad base85 character at byte %d' % (i + j)) |
|
58 |
if acc > 4294967295: |
|
59 |
raise OverflowError('Base85 overflow in hunk starting at byte %d' % i) |
|
60 |
out.append(acc) |
|
61 |
||
62 |
# Pad final chunk if necessary |
|
63 |
cl = l % 5 |
|
64 |
if cl: |
|
65 |
acc *= 85 ** (5 - cl) |
|
66 |
if cl > 1: |
|
67 |
acc += 0xffffff >> (cl - 2) * 8 |
|
68 |
out[-1] = acc |
|
69 |
||
70 |
out = struct.pack('>%dL' % (len(out)), *out) |
|
71 |
if cl: |
|
72 |
out = out[:-(5 - cl)] |
|
73 |
||
74 |
return out |