DeDRM_tools/DeDRM_plugin/ineptepub.py

467 lines
19 KiB
Python
Raw Normal View History

#!/usr/bin/env python3
2010-07-24 17:55:24 +06:00
# -*- coding: utf-8 -*-
2009-02-19 18:39:43 +06:00
# ineptepub.py
# Copyright © 2009-2022 by i♥cabbages, Apprentice Harper et al.
2009-02-19 18:39:43 +06:00
# Released under the terms of the GNU General Public Licence, version 3
# <http://www.gnu.org/licenses/>
2009-02-19 18:39:43 +06:00
# Revision history:
# 1 - Initial release
2013-10-03 00:59:40 +06:00
# 2 - Rename to INEPT, fix exit code
# 5 - Version bump to avoid (?) confusion;
# Improve OS X support by using OpenSSL when available
# 5.1 - Improve OpenSSL error checking
# 5.2 - Fix ctypes error causing segfaults on some systems
# 5.3 - add support for OpenSSL on Windows, fix bug with some versions of libcrypto 0.9.8 prior to path level o
# 5.4 - add support for encoding to 'utf-8' when building up list of files to decrypt from encryption.xml
# 5.5 - On Windows try PyCrypto first, OpenSSL next
# 5.6 - Modify interface to allow use with import
# 5.7 - Fix for potential problem with PyCrypto
# 5.8 - Revised to allow use in calibre plugins to eliminate need for duplicate code
# 5.9 - Fixed to retain zip file metadata (e.g. file modification date)
# 6.0 - moved unicode_argv call inside main for Windows DeDRM compatibility
# 6.1 - Work if TkInter is missing
# 6.2 - Handle UTF-8 file names inside an ePub, fix by Jose Luis
# 6.3 - Add additional check on DER file sanity
2016-01-14 23:15:43 +06:00
# 6.4 - Remove erroneous check on DER file sanity
2016-01-15 12:30:54 +06:00
# 6.5 - Completely remove erroneous check on DER file sanity
# 6.6 - Import tkFileDialog, don't assume something else will import it.
# 7.0 - Add Python 3 compatibility for calibre 5.0
2021-12-23 16:29:58 +06:00
# 7.1 - Add ignoble support, dropping the dedicated ignobleepub.py script
# 7.2 - Only support PyCryptodome; clean up the code
# 8.0 - Add support for "hardened" Adobe DRM (RMSDK >= 10)
2012-11-07 19:14:25 +06:00
2009-02-19 18:39:43 +06:00
"""
2013-10-03 00:59:40 +06:00
Decrypt Adobe Digital Editions encrypted ePub books.
2009-02-19 18:39:43 +06:00
"""
__license__ = 'GPL v3'
__version__ = "8.0"
2009-02-19 18:39:43 +06:00
import sys
import os
2013-10-03 00:59:40 +06:00
import traceback
2021-12-23 16:29:58 +06:00
import base64
2009-02-19 18:39:43 +06:00
import zlib
2013-10-03 00:59:40 +06:00
import zipfile
from zipfile import ZipInfo, ZipFile, ZIP_STORED, ZIP_DEFLATED
from zeroedzipinfo import ZeroedZipInfo
2013-10-03 00:59:40 +06:00
from contextlib import closing
from lxml import etree
from uuid import UUID
import hashlib
try:
from Cryptodome.Cipher import AES, PKCS1_v1_5
from Cryptodome.PublicKey import RSA
except ImportError:
from Crypto.Cipher import AES, PKCS1_v1_5
from Crypto.PublicKey import RSA
2022-03-19 15:14:45 +06:00
def unpad(data, padding=16):
if sys.version_info[0] == 2:
pad_len = ord(data[-1])
else:
pad_len = data[-1]
return data[:-pad_len]
2022-08-07 00:09:30 +06:00
from utilities import SafeUnbuffered
2022-08-07 00:19:18 +06:00
from argv_utils import unicode_argv
2009-02-19 18:39:43 +06:00
2010-07-24 17:55:24 +06:00
class ADEPTError(Exception):
pass
2009-02-19 18:39:43 +06:00
class ADEPTNewVersionError(Exception):
pass
META_NAMES = ('mimetype', 'META-INF/rights.xml')
2013-10-03 00:59:40 +06:00
NSMAP = {'adept': 'http://ns.adobe.com/adept',
'enc': 'http://www.w3.org/2001/04/xmlenc#'}
class Decryptor(object):
def __init__(self, bookkey, encryption):
enc = lambda tag: '{%s}%s' % (NSMAP['enc'], tag)
self._aes = AES.new(bookkey, AES.MODE_CBC, b'\x00'*16)
self._encryption = etree.fromstring(encryption)
2013-10-03 00:59:40 +06:00
self._encrypted = encrypted = set()
self._encryptedForceNoDecomp = encryptedForceNoDecomp = set()
self._otherData = otherData = set()
self._json_elements_to_remove = json_elements_to_remove = set()
self._has_remaining_xml = False
2013-10-03 00:59:40 +06:00
expr = './%s/%s/%s' % (enc('EncryptedData'), enc('CipherData'),
enc('CipherReference'))
for elem in self._encryption.findall(expr):
2013-10-03 00:59:40 +06:00
path = elem.get('URI', None)
encryption_type_url = (elem.getparent().getparent().find("./%s" % (enc('EncryptionMethod'))).get('Algorithm', None))
2013-10-03 00:59:40 +06:00
if path is not None:
if (encryption_type_url == "http://www.w3.org/2001/04/xmlenc#aes128-cbc"):
# Adobe
path = path.encode('utf-8')
encrypted.add(path)
json_elements_to_remove.add(elem.getparent().getparent())
elif (encryption_type_url == "http://ns.adobe.com/adept/xmlenc#aes128-cbc-uncompressed"):
# Adobe uncompressed, for stuff like video files
path = path.encode('utf-8')
encryptedForceNoDecomp.add(path)
json_elements_to_remove.add(elem.getparent().getparent())
else:
path = path.encode('utf-8')
otherData.add(path)
self._has_remaining_xml = True
for elem in json_elements_to_remove:
elem.getparent().remove(elem)
def check_if_remaining(self):
return self._has_remaining_xml
def get_xml(self):
return "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n" + etree.tostring(self._encryption, encoding="utf-8", pretty_print=True, xml_declaration=False).decode("utf-8")
2013-10-03 00:59:40 +06:00
def decompress(self, bytes):
dc = zlib.decompressobj(-15)
try:
decompressed_bytes = dc.decompress(bytes)
ex = dc.decompress(b'Z') + dc.flush()
if ex:
decompressed_bytes = decompressed_bytes + ex
except:
# possibly not compressed by zip - just return bytes
return bytes
return decompressed_bytes
2013-10-03 00:59:40 +06:00
def decrypt(self, path, data):
if path.encode('utf-8') in self._encrypted or path.encode('utf-8') in self._encryptedForceNoDecomp:
2013-10-03 00:59:40 +06:00
data = self._aes.decrypt(data)[16:]
if type(data[-1]) != int:
place = ord(data[-1])
else:
place = data[-1]
data = data[:-place]
if not path.encode('utf-8') in self._encryptedForceNoDecomp:
data = self.decompress(data)
2013-04-05 22:44:48 +06:00
return data
2013-10-03 00:59:40 +06:00
# check file to make check whether it's probably an Adobe Adept encrypted ePub
def adeptBook(inpath):
with closing(ZipFile(open(inpath, 'rb'))) as inf:
namelist = set(inf.namelist())
if 'META-INF/rights.xml' not in namelist or \
'META-INF/encryption.xml' not in namelist:
return False
try:
2013-10-03 00:59:40 +06:00
rights = etree.fromstring(inf.read('META-INF/rights.xml'))
adept = lambda tag: '{%s}%s' % (NSMAP['adept'], tag)
expr = './/%s' % (adept('encryptedKey'),)
bookkey = ''.join(rights.findtext(expr))
2021-12-23 16:29:58 +06:00
if len(bookkey) in [192, 172, 64]:
2013-10-03 00:59:40 +06:00
return True
except:
# if we couldn't check, assume it is
return True
return False
2021-12-23 16:29:58 +06:00
def isPassHashBook(inpath):
# If this is an Adobe book, check if it's a PassHash-encrypted book (B&N)
with closing(ZipFile(open(inpath, 'rb'))) as inf:
namelist = set(inf.namelist())
if 'META-INF/rights.xml' not in namelist or \
'META-INF/encryption.xml' not in namelist:
return False
try:
rights = etree.fromstring(inf.read('META-INF/rights.xml'))
adept = lambda tag: '{%s}%s' % (NSMAP['adept'], tag)
expr = './/%s' % (adept('encryptedKey'),)
bookkey = ''.join(rights.findtext(expr))
if len(bookkey) == 64:
return True
except:
pass
2021-12-23 16:29:58 +06:00
return False
# Checks the license file and returns the UUID the book is licensed for.
2021-11-15 16:59:56 +06:00
# This is used so that the Calibre plugin can pick the correct decryption key
# first try without having to loop through all possible keys.
def adeptGetUserUUID(inpath):
2021-11-15 16:59:56 +06:00
with closing(ZipFile(open(inpath, 'rb'))) as inf:
try:
rights = etree.fromstring(inf.read('META-INF/rights.xml'))
adept = lambda tag: '{%s}%s' % (NSMAP['adept'], tag)
expr = './/%s' % (adept('user'),)
user_uuid = ''.join(rights.findtext(expr))
if user_uuid[:9] != "urn:uuid:":
2021-11-15 16:59:56 +06:00
return None
return user_uuid[9:]
except:
return None
def removeHardening(rights, keytype, keydata):
adept = lambda tag: '{%s}%s' % (NSMAP['adept'], tag)
textGetter = lambda name: ''.join(rights.findtext('.//%s' % (adept(name),)))
# Gather what we need, and generate the IV
resourceuuid = UUID(textGetter("resource"))
deviceuuid = UUID(textGetter("device"))
fullfillmentuuid = UUID(textGetter("fulfillment")[:36])
kekiv = UUID(int=resourceuuid.int ^ deviceuuid.int ^ fullfillmentuuid.int).bytes
# Derive kek from just "keytype"
rem = int(keytype, 10) % 16
H = hashlib.sha256(keytype.encode("ascii")).digest()
kek = H[2*rem : 16 + rem] + H[rem : 2*rem]
return unpad(AES.new(kek, AES.MODE_CBC, kekiv).decrypt(keydata), 16) # PKCS#7
2013-10-03 00:59:40 +06:00
def decryptBook(userkey, inpath, outpath):
with closing(ZipFile(open(inpath, 'rb'))) as inf:
namelist = inf.namelist()
2013-10-03 00:59:40 +06:00
if 'META-INF/rights.xml' not in namelist or \
'META-INF/encryption.xml' not in namelist:
print("{0:s} is DRM-free.".format(os.path.basename(inpath)))
2013-10-03 00:59:40 +06:00
return 1
for name in META_NAMES:
namelist.remove(name)
2013-04-05 22:44:48 +06:00
try:
2013-10-03 00:59:40 +06:00
rights = etree.fromstring(inf.read('META-INF/rights.xml'))
adept = lambda tag: '{%s}%s' % (NSMAP['adept'], tag)
expr = './/%s' % (adept('encryptedKey'),)
bookkeyelem = rights.find(expr)
bookkey = bookkeyelem.text
keytype = bookkeyelem.attrib.get('keyType', '0')
if len(bookkey) >= 172 and int(keytype, 10) > 2:
print("{0:s} is a secure Adobe Adept ePub with hardening.".format(os.path.basename(inpath)))
elif len(bookkey) == 172:
2021-12-23 16:29:58 +06:00
print("{0:s} is a secure Adobe Adept ePub.".format(os.path.basename(inpath)))
elif len(bookkey) == 64:
print("{0:s} is a secure Adobe PassHash (B&N) ePub.".format(os.path.basename(inpath)))
else:
print("{0:s} is not an Adobe-protected ePub!".format(os.path.basename(inpath)))
2013-10-03 00:59:40 +06:00
return 1
2021-12-23 16:29:58 +06:00
if len(bookkey) != 64:
# Normal or "hardened" Adobe ADEPT
rsakey = RSA.importKey(userkey) # parses the ASN1 structure
bookkey = base64.b64decode(bookkey)
if int(keytype, 10) > 2:
bookkey = removeHardening(rights, keytype, bookkey)
try:
bookkey = PKCS1_v1_5.new(rsakey).decrypt(bookkey, None) # automatically unpads
except ValueError:
bookkey = None
if bookkey is None:
print("Could not decrypt {0:s}. Wrong key".format(os.path.basename(inpath)))
return 2
else:
2021-12-23 16:29:58 +06:00
# Adobe PassHash / B&N
key = base64.b64decode(userkey)[:16]
bookkey = base64.b64decode(bookkey)
bookkey = unpad(AES.new(key, AES.MODE_CBC, b'\x00'*16).decrypt(bookkey), 16) # PKCS#7
2021-12-23 16:29:58 +06:00
2022-01-02 21:23:36 +06:00
if len(bookkey) > 16:
2021-01-28 18:06:59 +06:00
bookkey = bookkey[-16:]
2021-12-23 16:29:58 +06:00
2013-10-03 00:59:40 +06:00
encryption = inf.read('META-INF/encryption.xml')
2020-11-22 21:03:45 +06:00
decryptor = Decryptor(bookkey, encryption)
2013-10-03 00:59:40 +06:00
kwds = dict(compression=ZIP_DEFLATED, allowZip64=False)
with closing(ZipFile(open(outpath, 'wb'), 'w', **kwds)) as outf:
for path in (["mimetype"] + namelist):
2013-10-03 00:59:40 +06:00
data = inf.read(path)
zi = ZipInfo(path)
zi.compress_type=ZIP_DEFLATED
if path == "mimetype":
zi.compress_type = ZIP_STORED
elif path == "META-INF/encryption.xml":
# Check if there's still something in there
if (decryptor.check_if_remaining()):
data = decryptor.get_xml()
print("Adding encryption.xml for the remaining embedded files.")
# We removed DRM, but there's still stuff like obfuscated fonts.
else:
continue
try:
2013-10-03 00:59:40 +06:00
# get the file info, including time-stamp
oldzi = inf.getinfo(path)
# copy across useful fields
zi.date_time = oldzi.date_time
zi.comment = oldzi.comment
zi.extra = oldzi.extra
zi.internal_attr = oldzi.internal_attr
# external attributes are dependent on the create system, so copy both.
zi.external_attr = oldzi.external_attr
zi.volume = oldzi.volume
2013-10-03 00:59:40 +06:00
zi.create_system = oldzi.create_system
zi.create_version = oldzi.create_version
if any(ord(c) >= 128 for c in path) or any(ord(c) >= 128 for c in zi.comment):
# If the file name or the comment contains any non-ASCII char, set the UTF8-flag
zi.flag_bits |= 0x800
2013-10-03 00:59:40 +06:00
except:
pass
# Python 3 has a bug where the external_attr is reset to `0o600 << 16`
# if it's NULL, so we need a workaround:
if zi.external_attr == 0:
zi = ZeroedZipInfo(zi)
if path == "META-INF/encryption.xml":
outf.writestr(zi, data)
else:
outf.writestr(zi, decryptor.decrypt(path, data))
except:
print("Could not decrypt {0:s} because of an exception:\n{1:s}".format(os.path.basename(inpath), traceback.format_exc()))
return 2
return 0
2013-03-26 22:38:18 +06:00
def cli_main():
2013-04-05 22:44:48 +06:00
sys.stdout=SafeUnbuffered(sys.stdout)
sys.stderr=SafeUnbuffered(sys.stderr)
2022-08-07 00:19:18 +06:00
argv=unicode_argv("ineptepub.py")
progname = os.path.basename(argv[0])
if len(argv) != 4:
print("usage: {0} <keyfile.der> <inbook.epub> <outbook.epub>".format(progname))
return 1
keypath, inpath, outpath = argv[1:]
userkey = open(keypath,'rb').read()
result = decryptBook(userkey, inpath, outpath)
if result == 0:
print("Successfully decrypted {0:s} as {1:s}".format(os.path.basename(inpath),os.path.basename(outpath)))
return result
2009-02-19 18:39:43 +06:00
def gui_main():
2013-04-05 22:44:48 +06:00
try:
import tkinter
2020-11-22 21:03:45 +06:00
import tkinter.constants
import tkinter.filedialog
import tkinter.messagebox
2013-04-05 22:44:48 +06:00
import traceback
except:
return cli_main()
class DecryptionDialog(tkinter.Frame):
def __init__(self, root):
tkinter.Frame.__init__(self, root, border=5)
self.status = tkinter.Label(self, text="Select files for decryption")
2020-11-22 21:03:45 +06:00
self.status.pack(fill=tkinter.constants.X, expand=1)
body = tkinter.Frame(self)
2020-11-22 21:03:45 +06:00
body.pack(fill=tkinter.constants.X, expand=1)
sticky = tkinter.constants.E + tkinter.constants.W
body.grid_columnconfigure(1, weight=2)
tkinter.Label(body, text="Key file").grid(row=0)
self.keypath = tkinter.Entry(body, width=30)
self.keypath.grid(row=0, column=1, sticky=sticky)
if os.path.exists("adeptkey.der"):
self.keypath.insert(0, "adeptkey.der")
button = tkinter.Button(body, text="...", command=self.get_keypath)
button.grid(row=0, column=2)
tkinter.Label(body, text="Input file").grid(row=1)
self.inpath = tkinter.Entry(body, width=30)
self.inpath.grid(row=1, column=1, sticky=sticky)
button = tkinter.Button(body, text="...", command=self.get_inpath)
button.grid(row=1, column=2)
tkinter.Label(body, text="Output file").grid(row=2)
self.outpath = tkinter.Entry(body, width=30)
self.outpath.grid(row=2, column=1, sticky=sticky)
button = tkinter.Button(body, text="...", command=self.get_outpath)
button.grid(row=2, column=2)
buttons = tkinter.Frame(self)
buttons.pack()
botton = tkinter.Button(
buttons, text="Decrypt", width=10, command=self.decrypt)
2020-11-22 21:03:45 +06:00
botton.pack(side=tkinter.constants.LEFT)
tkinter.Frame(buttons, width=10).pack(side=tkinter.constants.LEFT)
button = tkinter.Button(
buttons, text="Quit", width=10, command=self.quit)
2020-11-22 21:03:45 +06:00
button.pack(side=tkinter.constants.RIGHT)
def get_keypath(self):
2020-11-22 21:03:45 +06:00
keypath = tkinter.filedialog.askopenfilename(
parent=None, title="Select Adobe Adept \'.der\' key file",
defaultextension=".der",
filetypes=[('Adobe Adept DER-encoded files', '.der'),
('All Files', '.*')])
if keypath:
keypath = os.path.normpath(keypath)
2020-11-22 21:03:45 +06:00
self.keypath.delete(0, tkinter.constants.END)
self.keypath.insert(0, keypath)
return
def get_inpath(self):
2020-11-22 21:03:45 +06:00
inpath = tkinter.filedialog.askopenfilename(
parent=None, title="Select ADEPT-encrypted ePub file to decrypt",
defaultextension=".epub", filetypes=[('ePub files', '.epub')])
if inpath:
inpath = os.path.normpath(inpath)
2020-11-22 21:03:45 +06:00
self.inpath.delete(0, tkinter.constants.END)
self.inpath.insert(0, inpath)
return
def get_outpath(self):
2020-11-22 21:03:45 +06:00
outpath = tkinter.filedialog.asksaveasfilename(
parent=None, title="Select unencrypted ePub file to produce",
defaultextension=".epub", filetypes=[('ePub files', '.epub')])
if outpath:
outpath = os.path.normpath(outpath)
2020-11-22 21:03:45 +06:00
self.outpath.delete(0, tkinter.constants.END)
self.outpath.insert(0, outpath)
return
def decrypt(self):
keypath = self.keypath.get()
inpath = self.inpath.get()
outpath = self.outpath.get()
if not keypath or not os.path.exists(keypath):
self.status['text'] = "Specified key file does not exist"
return
if not inpath or not os.path.exists(inpath):
self.status['text'] = "Specified input file does not exist"
return
if not outpath:
self.status['text'] = "Output file not specified"
return
if inpath == outpath:
self.status['text'] = "Must have different input and output files"
return
userkey = open(keypath,'rb').read()
self.status['text'] = "Decrypting..."
try:
decrypt_status = decryptBook(userkey, inpath, outpath)
except Exception as e:
self.status['text'] = "Error: {0}".format(e.args[0])
return
if decrypt_status == 0:
self.status['text'] = "File successfully decrypted"
else:
2020-11-22 21:03:45 +06:00
self.status['text'] = "There was an error decrypting the file."
root = tkinter.Tk()
root.title("Adobe Adept ePub Decrypter v.{0}".format(__version__))
2009-02-19 18:39:43 +06:00
root.resizable(True, False)
2013-10-03 00:59:40 +06:00
root.minsize(300, 0)
2020-11-22 21:03:45 +06:00
DecryptionDialog(root).pack(fill=tkinter.constants.X, expand=1)
2009-02-19 18:39:43 +06:00
root.mainloop()
2009-09-01 23:02:35 +07:00
return 0
2009-02-19 18:39:43 +06:00
if __name__ == '__main__':
2010-02-14 21:47:48 +06:00
if len(sys.argv) > 1:
sys.exit(cli_main())
2009-02-19 18:39:43 +06:00
sys.exit(gui_main())