读BitTorrent码日记1

# Written by Petru Paler
# see LICENSE.txt for license information
"""

读码日记: 2004-8-20 笔者：
笔者: zfive5(醉马不肖 之 [孤舟蓑笠翁， 独钓寒江雪])
注 释：文件是写的函数主要完成对bt种子文件的操作函数，在开始前大家可以
使用make做一个.torrent文件看看其中的内容，在我的e:/1/目录下有两个文件,
dir为:

2004-08-31 15:33

.
2004-08-31 15:33 ..
2004-08-31 14:44 10 1.txt
2004-08-31 15:35 11 2.txt

经过使用maketorrent做出来的.torrent文件内容如下(这里由于太长,加入的换行符和空格注释):
d
8:announce
30:http://127.0.0.1:6969/announce
13:creation date
i1093944351e
4:info
d
5:files
l
d
6:length
i10e
4:path
l
5:1.txt
e
e
d
6:length
i11e
4:path
l
5:2.txt
e
e
e
4:name
1:1
12:piece length
i32768e
6:pieces
20:訢^滢g?覷'最姠痲胴(D4 44 5e e4 de 67 d1 18 03 d3 55 27 d7 ee 8a a2 af 71 eb d8)(这是bt文件块的一个文件块的20 位SHA消息摘要）
e
3:nfo
d
e
e

写的直接些就是这样的：
announce:http://127.0.0.1:6969/announce
creation date:1093944351(秒数)
info
files
{
length=10(字节)
path=1.txt
,
length=11(字节)
path=2.txt
}
name=1
piece length=32768(每一块文件大小)
pieces=D4 44 5e e4 de 67 d1 18 03 d3 55 27 d7 ee 8a a2 af 71 eb d8
nfo=null(空)

这里有几种类型的数据结构要说明一下:
整型 : i<d>e 例如: i1093944351e<br>字符串: n:<str> 例如: 30:http://127.0.0.1:6969/announce<br>列表： l&lt;内容&gt;e 例如: l5:1.txte<br>字典： d&lt;内容&gt;e 例如: d6:lengthi10e4:pathl5:1.txteed6:lengthi11e4:pathl5:2.txteee</str></d>

"""
#bt源码 胜过一切解释
import re
from cStringIO import StringIO

int_filter = re.compile('(0|-?[1-9][0-9]*)e')

#解码整型函数
def decode_int(x, f):
m = int_filter.match(x, f)
if m is None:
raise ValueError
return (long(m.group(1)), m.end())

string_filter = re.compile('(0|[1-9][0-9]*):')

#解码字符串函数
def decode_string(x, f):
m = string_filter.match(x, f)
if m is None:
raise ValueError
l = int(m.group(1))
s = m.end()
return (x[s:s+l], s + l)

#解码列表函数
def decode_list(x, f):
r = []
while x[f] != 'e':
v, f = bdecode_rec(x, f)
r.append(v)
return (r, f + 1)

#解码字典函数
def decode_dict(x, f):
r = {}
lastkey = None
while x[f] != 'e':
k, f = decode_string(x, f)
if lastkey is not None and lastkey >= k:
raise ValueError
lastkey = k
v, f = bdecode_rec(x, f)
r[k] = v
return (r, f + 1)

#对内的解码函数
def bdecode_rec(x, f):
t = x[f]
if t == 'i':
return decode_int(x, f + 1)
elif t == 'l':
return decode_list(x, f + 1)
elif t == 'd':
return decode_dict(x, f + 1)
else:
return decode_string(x, f)

#对外的解码函数
def bdecode(x):
try:
r, l = bdecode_rec(x, 0)
except IndexError:
raise ValueError
if l != len(x):
raise ValueError
return r

#测试解码函数
def test_bdecode():
try:
bdecode('0:0:')
assert 0
except ValueError:
pass
try:
bdecode('ie')
assert 0
except ValueError:
pass
try:
bdecode('i341foo382e')
assert 0
except ValueError:
pass
assert bdecode('i4e') == 4L
assert bdecode('i0e') == 0L
assert bdecode('i123456789e') == 123456789L
assert bdecode('i-10e') == -10L
try:
bdecode('i-0e')
assert 0
except ValueError:
pass
try:
bdecode('i123')
assert 0
except ValueError:
pass
try:
bdecode('')
assert 0
except ValueError:
pass
try:
bdecode('i6easd')
assert 0
except ValueError:
pass
try:
bdecode('35208734823ljdahflajhdf')
assert 0
except ValueError:
pass
try:
bdecode('2:abfdjslhfld')
assert 0
except ValueError:
pass
assert bdecode('0:') == ''
assert bdecode('3:abc') == 'abc'
assert bdecode('10:1234567890') == '1234567890'
try:
bdecode('02:xy')
assert 0
except ValueError:
pass
try:
bdecode('l')
assert 0
except ValueError:
pass
assert bdecode('le') == []
try:
bdecode('leanfdldjfh')
assert 0
except ValueError:
pass
assert bdecode('l0:0:0:e') == ['', '', '']
try:
bdecode('relwjhrlewjh')
assert 0
except ValueError:
pass
assert bdecode('li1ei2ei3ee') == [1, 2, 3]
assert bdecode('l3:asd2:xye') == ['asd', 'xy']
assert bdecode('ll5:Alice3:Bobeli2ei3eee') == [['Alice', 'Bob'], [2, 3]]
try:
bdecode('d')
assert 0
except ValueError:
pass
try:
bdecode('defoobar')
assert 0
except ValueError:
pass
assert bdecode('de') == {}
assert bdecode('d3:agei25e4:eyes4:bluee') == {'age': 25, 'eyes': 'blue'}
assert bdecode('d8:spam.mp3d6:author5:Alice6:lengthi100000eee') == {'spam.mp3': {'author': 'Alice', 'length': 100000}}
try:
bdecode('d3:fooe')
assert 0
except ValueError:
pass
try:
bdecode('di1e0:e')
assert 0
except ValueError:
pass
try:
bdecode('d1:b0:1:a0:e')
assert 0
except ValueError:
pass
try:
bdecode('d1:a0:1:a0:e')
assert 0
except ValueError:
pass
try:
bdecode('i03e')
assert 0
except ValueError:
pass
try:
bdecode('l01:ae')
assert 0
except ValueError:
pass
try:
bdecode('9999:x')
assert 0
except ValueError:
pass
try:
bdecode('l0:')
assert 0
except ValueError:
pass
try:
bdecode('d0:0:')
assert 0
except ValueError:
pass
try:
bdecode('d0:')
assert 0
except ValueError:
pass

#对内编码函数
def bencode_rec(x, b):
t = type(x)
if t in (int, long, bool):
b.write('i%de' % x)
elif t is str:
b.write('%d:%s' % (len(x), x))
elif t in (list, tuple):
b.write('l')
for e in x:
bencode_rec(e, b)
b.write('e')
elif t is dict:
b.write('d')
keylist = x.keys()
keylist.sort()
for k in keylist:
assert type(k) is str
bencode_rec(k, b)
bencode_rec(x[k], b)
b.write('e')
else:
assert 0

#对外编码函数
def bencode(x):
b = StringIO()
bencode_rec(x, b)
return b.getvalue()

#测试编码函数
def test_bencode():
assert bencode(4) == 'i4e'
assert bencode(0) == 'i0e'
assert bencode(-10) == 'i-10e'
assert bencode(12345678901234567890L) == 'i12345678901234567890e'
assert bencode('') == '0:'
assert bencode('abc') == '3:abc'
assert bencode('1234567890') == '10:1234567890'
assert bencode([]) == 'le'
assert bencode([1, 2, 3]) == 'li1ei2ei3ee'
assert bencode([['Alice', 'Bob'], [2, 3]]) == 'll5:Alice3:Bobeli2ei3eee'
assert bencode({}) == 'de'
assert bencode({'age': 25, 'eyes': 'blue'}) == 'd3:agei25e4:eyes4:bluee'
assert bencode({'spam.mp3': {'author': 'Alice', 'length': 100000}}) == 'd8:spam.mp3d6:author5:Alice6:lengthi100000eee'
try:
bencode({1: 'foo'})
assert 0
except AssertionError:
pass