/usr/lib/python2.7/dist-packages/spambayes/cdb.py is in spambayes 1.1b1-1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 | #! /usr/bin/env python
"""
Dan Bernstein's CDB implemented in Python
see http://cr.yp.to/cdb.html
"""
from __future__ import generators
import os
import struct
import mmap
def uint32_unpack(buf):
return struct.unpack('<L', buf)[0]
def uint32_pack(n):
return struct.pack('<L', n)
CDB_HASHSTART = 5381
def cdb_hash(buf):
h = CDB_HASHSTART
for c in buf:
h = (h + (h << 5)) & 0xffffffffL
h ^= ord(c)
return h
class Cdb(object):
def __init__(self, fp):
self.fp = fp
fd = fp.fileno()
self.size = os.fstat(fd).st_size
self.map = mmap.mmap(fd, self.size, access=mmap.ACCESS_READ)
self.eod = uint32_unpack(self.map[:4])
self.findstart()
self.loop = 0 # number of hash slots searched under this key
# initialized if loop is nonzero
self.khash = 0
self.hpos = 0
self.hslots = 0
# initialized if findnext() returns 1
self.dpos = 0
self.dlen = 0
def close(self):
self.map.close()
def __iter__(self, fn=None):
len = 2048
while len < self.eod:
klen, vlen = struct.unpack("<LL", self.map[len:len+8])
len += 8
key = self.map[len:len+klen]
len += klen
val = self.map[len:len+vlen]
len += vlen
if fn:
yield fn(key, val)
else:
yield (key, val)
def iteritems(self):
return self.__iter__()
def iterkeys(self):
return self.__iter__(lambda k, v: k)
def itervalues(self):
return self.__iter__(lambda k, v: v)
def items(self):
ret = []
for i in self.iteritems():
ret.append(i)
return ret
def keys(self):
ret = []
for i in self.iterkeys():
ret.append(i)
return ret
def values(self):
ret = []
for i in self.itervalues():
ret.append(i)
return ret
def findstart(self):
self.loop = 0
def read(self, n, pos):
# XXX add code for platforms without mmap
return self.map[pos:pos+n]
def match(self, key, pos):
if key == self.read(len(key), pos):
return 1
else:
return 0
def findnext(self, key):
if not self.loop:
u = cdb_hash(key)
buf = self.read(8, u << 3 & 2047)
self.hslots = uint32_unpack(buf[4:])
if not self.hslots:
raise KeyError
self.hpos = uint32_unpack(buf[:4])
self.khash = u
u >>= 8
u %= self.hslots
u <<= 3
self.kpos = self.hpos + u
while self.loop < self.hslots:
buf = self.read(8, self.kpos)
pos = uint32_unpack(buf[4:])
if not pos:
raise KeyError
self.loop += 1
self.kpos += 8
if self.kpos == self.hpos + (self.hslots << 3):
self.kpos = self.hpos
u = uint32_unpack(buf[:4])
if u == self.khash:
buf = self.read(8, pos)
u = uint32_unpack(buf[:4])
if u == len(key):
if self.match(key, pos + 8):
dlen = uint32_unpack(buf[4:])
dpos = pos + 8 + len(key)
return self.read(dlen, dpos)
raise KeyError
def __getitem__(self, key):
self.findstart()
return self.findnext(key)
def get(self, key, default=None):
self.findstart()
try:
return self.findnext(key)
except KeyError:
return default
def cdb_dump(infile):
"""dump a database in djb's cdbdump format"""
db = Cdb(infile)
for key, value in db.iteritems():
print "+%d,%d:%s->%s" % (len(key), len(value), key, value)
print
def cdb_make(outfile, items):
pos = 2048
tables = {} # { h & 255 : [(h, p)] }
# write keys and data
outfile.seek(pos)
for key, value in items:
outfile.write(uint32_pack(len(key)) + uint32_pack(len(value)))
h = cdb_hash(key)
outfile.write(key)
outfile.write(value)
tables.setdefault(h & 255, []).append((h, pos))
pos += 8 + len(key) + len(value)
final = ''
# write hash tables
for i in range(256):
entries = tables.get(i, [])
nslots = 2*len(entries)
final += uint32_pack(pos) + uint32_pack(nslots)
null = (0, 0)
table = [null] * nslots
for h, p in entries:
n = (h >> 8) % nslots
while table[n] is not null:
n = (n + 1) % nslots
table[n] = (h, p)
for h, p in table:
outfile.write(uint32_pack(h) + uint32_pack(p))
pos += 8
# write header (pointers to tables and their lengths)
outfile.flush()
outfile.seek(0)
outfile.write(final)
def test():
#db = Cdb(open("t"))
#print db['one']
#print db['two']
#print db['foo']
#print db['us']
#print db.get('ec')
#print db.get('notthere')
db = open('test.cdb', 'wb')
cdb_make(db,
[('one', 'Hello'),
('two', 'Goodbye'),
('foo', 'Bar'),
('us', 'United States'),
])
db.close()
db = Cdb(open("test.cdb", 'rb'))
print db['one']
print db['two']
print db['foo']
print db['us']
print db.get('ec')
print db.get('notthere')
if __name__ == '__main__':
test()
|