DeDRM_tools/Kindle_Mobi_Tools/lib/mobihuff.py
2015-02-25 07:00:09 +00:00

168 lines
No EOL
4.9 KiB
Python

# This is a python script. You need a Python interpreter to run it.
# For example, ActiveState Python, which exists for windows.
#
# Big Thanks to Igor SKOCHINSKY for providing me with all his information
# and source code relating to the inner workings of this compression scheme.
# Without it, I wouldn't be able to solve this as easily.
#
# Changelog
# 0.01 - Initial version
# 0.02 - Fix issue with size computing
# 0.03 - Fix issue with some files
import struct, sys
class BitReader:
def __init__(self, data):
self.data, self.pos, self.nbits = data + "\x00\x00\x00\x00", 0, len(data) * 8
def peek(self, n):
r, g = 0, 0
while g < n:
r, g = (r << 8) | ord(self.data[(self.pos+g)>>3]), g + 8 - ((self.pos+g) & 7)
return (r >> (g - n)) & ((1 << n) - 1)
def eat(self, n):
self.pos += n
return self.pos <= self.nbits
def left(self):
return self.nbits - self.pos
class HuffReader:
def __init__(self, huffs):
self.huffs = huffs
h = huffs[0]
if huffs[0][0:4] != 'HUFF' or huffs[0][4:8] != '\x00\x00\x00\x18':
raise ValueError('invalid huff1 header')
if huffs[1][0:4] != 'CDIC' or huffs[1][4:8] != '\x00\x00\x00\x10':
raise ValueError('invalid huff2 header')
self.entry_bits, = struct.unpack('>L', huffs[1][12:16])
off1,off2 = struct.unpack('>LL', huffs[0][16:24])
self.dict1 = struct.unpack('<256L', huffs[0][off1:off1+256*4])
self.dict2 = struct.unpack('<64L', huffs[0][off2:off2+64*4])
self.dicts = huffs[1:]
self.r = ''
def _unpack(self, bits, depth = 0):
if depth > 32:
raise ValueError('corrupt file')
while bits.left():
dw = bits.peek(32)
v = self.dict1[dw >> 24]
codelen = v & 0x1F
assert codelen != 0
code = dw >> (32 - codelen)
r = (v >> 8)
if not (v & 0x80):
while code < self.dict2[(codelen-1)*2]:
codelen += 1
code = dw >> (32 - codelen)
r = self.dict2[(codelen-1)*2+1]
r -= code
assert codelen != 0
if not bits.eat(codelen):
return
dicno = r >> self.entry_bits
off1 = 16 + (r - (dicno << self.entry_bits)) * 2
dic = self.dicts[dicno]
off2 = 16 + ord(dic[off1]) * 256 + ord(dic[off1+1])
blen = ord(dic[off2]) * 256 + ord(dic[off2+1])
slice = dic[off2+2:off2+2+(blen&0x7fff)]
if blen & 0x8000:
self.r += slice
else:
self._unpack(BitReader(slice), depth + 1)
def unpack(self, data):
self.r = ''
self._unpack(BitReader(data))
return self.r
class Sectionizer:
def __init__(self, filename, ident):
self.contents = file(filename, 'rb').read()
self.header = self.contents[0:72]
self.num_sections, = struct.unpack('>H', self.contents[76:78])
if self.header[0x3C:0x3C+8] != ident:
raise ValueError('Invalid file format')
self.sections = []
for i in xrange(self.num_sections):
offset, a1,a2,a3,a4 = struct.unpack('>LBBBB', self.contents[78+i*8:78+i*8+8])
flags, val = a1, a2<<16|a3<<8|a4
self.sections.append( (offset, flags, val) )
def loadSection(self, section):
if section + 1 == self.num_sections:
end_off = len(self.contents)
else:
end_off = self.sections[section + 1][0]
off = self.sections[section][0]
return self.contents[off:end_off]
def getSizeOfTrailingDataEntry(ptr, size):
bitpos, result = 0, 0
while True:
v = ord(ptr[size-1])
result |= (v & 0x7F) << bitpos
bitpos += 7
size -= 1
if (v & 0x80) != 0 or (bitpos >= 28) or (size == 0):
return result
def getSizeOfTrailingDataEntries(ptr, size, flags):
num = 0
flags >>= 1
while flags:
if flags & 1:
num += getSizeOfTrailingDataEntry(ptr, size - num)
flags >>= 1
return num
def unpackBook(input_file):
sect = Sectionizer(input_file, 'BOOKMOBI')
header = sect.loadSection(0)
crypto_type, = struct.unpack('>H', header[0xC:0xC+2])
if crypto_type != 0:
raise ValueError('The book is encrypted. Run mobidedrm first')
if header[0:2] != 'DH':
raise ValueError('invalid compression type')
extra_flags, = struct.unpack('>L', header[0xF0:0xF4])
records, = struct.unpack('>H', header[0x8:0x8+2])
huffoff,huffnum = struct.unpack('>LL', header[0x70:0x78])
huffs = [sect.loadSection(i) for i in xrange(huffoff, huffoff+huffnum)]
huff = HuffReader(huffs)
def decompressSection(nr):
data = sect.loadSection(nr)
trail_size = getSizeOfTrailingDataEntries(data, len(data), extra_flags)
return huff.unpack(data[0:len(data)-trail_size])
r = ''
for i in xrange(1, records+1):
r += decompressSection(i)
return r
print "MobiHuff v0.03"
print " Copyright (c) 2008 The Dark Reverser <dark.reverser@googlemail.com>"
if len(sys.argv)!=3:
print ""
print "Description:"
print " Unpacks the new mobipocket huffdic compression."
print " This program works with unencrypted files only."
print "Usage:"
print " mobihuff.py infile.mobi outfile.html"
else:
infile = sys.argv[1]
outfile = sys.argv[2]
try:
print "Decompressing...",
result = unpackBook(infile)
file(outfile, 'wb').write(result)
print "done"
except ValueError, e:
print
print "Error: %s" % e