~bzr-pqm/bzr/bzr.dev

2220.3.1 by Martin Pool
add bencode utility
1
# bencode structured encoding
2
#
3
# Written by Petru Paler
4
#
5
# Permission is hereby granted, free of charge, to any person
6
# obtaining a copy of this software and associated documentation files
7
# (the "Software"), to deal in the Software without restriction,
8
# including without limitation the rights to use, copy, modify, merge,
9
# publish, distribute, sublicense, and/or sell copies of the Software,
10
# and to permit persons to whom the Software is furnished to do so,
11
# subject to the following conditions:
12
# 
13
# The above copyright notice and this permission notice shall be
14
# included in all copies or substantial portions of the Software.
15
# 
16
# The Software is provided "AS IS", without warranty of any kind,
17
# express or implied, including but not limited to the warranties of
18
# merchantability,  fitness for a particular purpose and
19
# noninfringement. In no event shall the  authors or copyright holders
20
# be liable for any claim, damages or other liability, whether in an
21
# action of contract, tort or otherwise, arising from, out of or in
22
# connection with the Software or the use or other dealings in the
23
# Software.
24
25
def decode_int(x, f):
26
    f += 1
27
    newf = x.index('e', f)
28
    try:
29
        n = int(x[f:newf])
30
    except (OverflowError, ValueError):
31
        n = long(x[f:newf])
32
    if x[f] == '-':
33
        if x[f + 1] == '0':
34
            raise ValueError
35
    elif x[f] == '0' and newf != f+1:
36
        raise ValueError
37
    return (n, newf+1)
38
39
def decode_string(x, f):
40
    colon = x.index(':', f)
41
    try:
42
        n = int(x[f:colon])
43
    except (OverflowError, ValueError):
44
        n = long(x[f:colon])
45
    if x[f] == '0' and colon != f+1:
46
        raise ValueError
47
    colon += 1
48
    return (x[colon:colon+n], colon+n)
49
50
def decode_list(x, f):
51
    r, f = [], f+1
52
    while x[f] != 'e':
53
        v, f = decode_func[x[f]](x, f)
54
        r.append(v)
55
    return (r, f + 1)
56
57
def decode_dict(x, f):
58
    r, f = {}, f+1
59
    lastkey = None
60
    while x[f] != 'e':
61
        k, f = decode_string(x, f)
62
        if lastkey >= k:
63
            raise ValueError
64
        lastkey = k
65
        r[k], f = decode_func[x[f]](x, f)
66
    return (r, f + 1)
67
68
decode_func = {}
69
decode_func['l'] = decode_list
70
decode_func['d'] = decode_dict
71
decode_func['i'] = decode_int
72
decode_func['0'] = decode_string
73
decode_func['1'] = decode_string
74
decode_func['2'] = decode_string
75
decode_func['3'] = decode_string
76
decode_func['4'] = decode_string
77
decode_func['5'] = decode_string
78
decode_func['6'] = decode_string
79
decode_func['7'] = decode_string
80
decode_func['8'] = decode_string
81
decode_func['9'] = decode_string
82
83
def bdecode(x):
84
    try:
85
        r, l = decode_func[x[0]](x, 0)
86
    except (IndexError, KeyError):
87
        raise ValueError
88
    if l != len(x):
89
        raise ValueError
90
    return r
91
92
def test_bdecode():
93
    try:
94
        bdecode('0:0:')
95
        assert 0
96
    except ValueError:
97
        pass
98
    try:
99
        bdecode('ie')
100
        assert 0
101
    except ValueError:
102
        pass
103
    try:
104
        bdecode('i341foo382e')
105
        assert 0
106
    except ValueError:
107
        pass
108
    assert bdecode('i4e') == 4L
109
    assert bdecode('i0e') == 0L
110
    assert bdecode('i123456789e') == 123456789L
111
    assert bdecode('i-10e') == -10L
112
    try:
113
        bdecode('i-0e')
114
        assert 0
115
    except ValueError:
116
        pass
117
    try:
118
        bdecode('i123')
119
        assert 0
120
    except ValueError:
121
        pass
122
    try:
123
        bdecode('')
124
        assert 0
125
    except ValueError:
126
        pass
127
    try:
128
        bdecode('i6easd')
129
        assert 0
130
    except ValueError:
131
        pass
132
    try:
133
        bdecode('35208734823ljdahflajhdf')
134
        assert 0
135
    except ValueError:
136
        pass
137
    try:
138
        bdecode('2:abfdjslhfld')
139
        assert 0
140
    except ValueError:
141
        pass
142
    assert bdecode('0:') == ''
143
    assert bdecode('3:abc') == 'abc'
144
    assert bdecode('10:1234567890') == '1234567890'
145
    try:
146
        bdecode('02:xy')
147
        assert 0
148
    except ValueError:
149
        pass
150
    try:
151
        bdecode('l')
152
        assert 0
153
    except ValueError:
154
        pass
155
    assert bdecode('le') == []
156
    try:
157
        bdecode('leanfdldjfh')
158
        assert 0
159
    except ValueError:
160
        pass
161
    assert bdecode('l0:0:0:e') == ['', '', '']
162
    try:
163
        bdecode('relwjhrlewjh')
164
        assert 0
165
    except ValueError:
166
        pass
167
    assert bdecode('li1ei2ei3ee') == [1, 2, 3]
168
    assert bdecode('l3:asd2:xye') == ['asd', 'xy']
169
    assert bdecode('ll5:Alice3:Bobeli2ei3eee') == [['Alice', 'Bob'], [2, 3]]
170
    try:
171
        bdecode('d')
172
        assert 0
173
    except ValueError:
174
        pass
175
    try:
176
        bdecode('defoobar')
177
        assert 0
178
    except ValueError:
179
        pass
180
    assert bdecode('de') == {}
181
    assert bdecode('d3:agei25e4:eyes4:bluee') == {'age': 25, 'eyes': 'blue'}
182
    assert bdecode('d8:spam.mp3d6:author5:Alice6:lengthi100000eee') == {'spam.mp3': {'author': 'Alice', 'length': 100000}}
183
    try:
184
        bdecode('d3:fooe')
185
        assert 0
186
    except ValueError:
187
        pass
188
    try:
189
        bdecode('di1e0:e')
190
        assert 0
191
    except ValueError:
192
        pass
193
    try:
194
        bdecode('d1:b0:1:a0:e')
195
        assert 0
196
    except ValueError:
197
        pass
198
    try:
199
        bdecode('d1:a0:1:a0:e')
200
        assert 0
201
    except ValueError:
202
        pass
203
    try:
204
        bdecode('i03e')
205
        assert 0
206
    except ValueError:
207
        pass
208
    try:
209
        bdecode('l01:ae')
210
        assert 0
211
    except ValueError:
212
        pass
213
    try:
214
        bdecode('9999:x')
215
        assert 0
216
    except ValueError:
217
        pass
218
    try:
219
        bdecode('l0:')
220
        assert 0
221
    except ValueError:
222
        pass
223
    try:
224
        bdecode('d0:0:')
225
        assert 0
226
    except ValueError:
227
        pass
228
    try:
229
        bdecode('d0:')
230
        assert 0
231
    except ValueError:
232
        pass
233
    try:
234
        bdecode('00:')
235
        assert 0
236
    except ValueError:
237
        pass
238
    try:
239
        bdecode('l-3:e')
240
        assert 0
241
    except ValueError:
242
        pass
243
    try:
244
        bdecode('i-03e')
245
        assert 0
246
    except ValueError:
247
        pass
248
    bdecode('d0:i3ee')
249
250
from types import StringType, IntType, LongType, DictType, ListType, TupleType
251
252
class Bencached(object):
253
    __slots__ = ['bencoded']
254
255
    def __init__(self, s):
256
        self.bencoded = s
257
258
def encode_bencached(x,r):
259
    r.append(x.bencoded)
260
261
def encode_int(x, r):
262
    r.extend(('i', str(x), 'e'))
263
264
def encode_string(x, r):
265
    r.extend((str(len(x)), ':', x))
266
267
def encode_list(x, r):
268
    r.append('l')
269
    for i in x:
270
        encode_func[type(i)](i, r)
271
    r.append('e')
272
273
def encode_dict(x,r):
274
    r.append('d')
275
    ilist = x.items()
276
    ilist.sort()
277
    for k, v in ilist:
278
        r.extend((str(len(k)), ':', k))
279
        encode_func[type(v)](v, r)
280
    r.append('e')
281
282
encode_func = {}
283
encode_func[type(Bencached(0))] = encode_bencached
284
encode_func[IntType] = encode_int
285
encode_func[LongType] = encode_int
286
encode_func[StringType] = encode_string
287
encode_func[ListType] = encode_list
288
encode_func[TupleType] = encode_list
289
encode_func[DictType] = encode_dict
290
291
try:
292
    from types import BooleanType
293
    encode_func[BooleanType] = encode_int
294
except ImportError:
295
    pass
296
297
def bencode(x):
298
    r = []
299
    encode_func[type(x)](x, r)
300
    return ''.join(r)
301
302
def test_bencode():
303
    assert bencode(4) == 'i4e'
304
    assert bencode(0) == 'i0e'
305
    assert bencode(-10) == 'i-10e'
306
    assert bencode(12345678901234567890L) == 'i12345678901234567890e'
307
    assert bencode('') == '0:'
308
    assert bencode('abc') == '3:abc'
309
    assert bencode('1234567890') == '10:1234567890'
310
    assert bencode([]) == 'le'
311
    assert bencode([1, 2, 3]) == 'li1ei2ei3ee'
312
    assert bencode([['Alice', 'Bob'], [2, 3]]) == 'll5:Alice3:Bobeli2ei3eee'
313
    assert bencode({}) == 'de'
314
    assert bencode({'age': 25, 'eyes': 'blue'}) == 'd3:agei25e4:eyes4:bluee'
315
    assert bencode({'spam.mp3': {'author': 'Alice', 'length': 100000}}) == 'd8:spam.mp3d6:author5:Alice6:lengthi100000eee'
316
    assert bencode(Bencached(bencode(3))) == 'i3e'
317
    try:
318
        bencode({1: 'foo'})
319
    except TypeError:
320
        return
321
    assert 0
322