LCP DRM - Extracting Files

[JSWolf]

Originally Posted by DNSB

I took a quick pass through the code and nothing looked suspicious to me. OTOH, I am not near fluent in Python.

Edit: installed it in Calibre running in a VM, downloaded two books and both decrypted and imported into calibre. Nothing seems to have been written into non-standard locations. It was more fun adding the _init.diff file than anything else.

Would you mind sharing (even just temporarily) the plugin with the changes? Thanks.

[Martinoptic]

I imagine ep***r will be rubbing their hands in glee with this release.

[JSWolf]

Originally Posted by Martinoptic

I imagine ep***r will be rubbing their hands in glee with this release.

Actually no they won't they will hate this release. They are charging a subscription for removing LCP DRM. With this, there could be less people subscribing.

[vila]

Originally Posted by nnss

https://gitflic.ru/project/lcpdisliker/dedrmlcpatch
Download it Before it's deleted

It works!



The script first determines if the file is LCP-protected by searching for specific metadata files like META-INF/license.lcpl and META-INF/encryption.xml.

# From tryParseAsLCPType
if "id" in license and "encryption" in license and "profile" in license["encryption"]:
    return ("epub", license)

Then to verify a password, the script needs to "transform" the user's input into a cryptographic key.

# From LCPTransform.secret_transform_profile20
blk = blake(masterkey, 64)
crc = crc32bts(blk)
adlr = adler32bts(blk)
strn = blk + crc + adlr
hmac = hmac256(adlr, strn)
return binascii.hexlify(hashlib.sha256(hmac).digest()).decode("latin-1")

The script then verifies if the transformed key is correct by attempting to decrypt a small piece of data called the key_check.

# From decryptLCPbook
decrypted = dataDecryptLCP(key_check, transformed_hash)
if (decrypted is not None and decrypted.decode("ascii", errors="ignore") == license["id"]):
    correct_password_hash = transformed_hash

Finally, once the key is validated, the script uses it to decrypt the Content Key, which is then used to decrypt the actual ePub or PDF assets.

# From Decryptor.decrypt
aes = AES.new(self.book_key, AES.MODE_CBC, data[:16])
data = aes.decrypt(data[16:])
# Fix padding and decompress
data, was_decomp = self.decompress(data)

The data is decrypted using the AES Cipher Block Chaining mode, where the first 16 bytes of the data serve as the Initialization Vector. Finally, it also removes the padding and decompresses the content using zlib.

[vila]

Originally Posted by JSWolf

Actually no they won't they will hate this release. They are charging a subscription for removing LCP DRM. With this, there could be less people subscribing.

the fact that they are charging money for such a simple decryption is wild xD

[vila]

Originally Posted by vila

It works!



The script first determines if the file is LCP-protected by searching for specific metadata files like META-INF/license.lcpl and META-INF/encryption.xml.

# From tryParseAsLCPType
if "id" in license and "encryption" in license and "profile" in license["encryption"]:
    return ("epub", license)

Then to verify a password, the script needs to "transform" the user's input into a cryptographic key.

# From LCPTransform.secret_transform_profile20
blk = blake(masterkey, 64)
crc = crc32bts(blk)
adlr = adler32bts(blk)
strn = blk + crc + adlr
hmac = hmac256(adlr, strn)
return binascii.hexlify(hashlib.sha256(hmac).digest()).decode("latin-1")

The script then verifies if the transformed key is correct by attempting to decrypt a small piece of data called the key_check.

# From decryptLCPbook
decrypted = dataDecryptLCP(key_check, transformed_hash)
if (decrypted is not None and decrypted.decode("ascii", errors="ignore") == license["id"]):
    correct_password_hash = transformed_hash

Finally, once the key is validated, the script uses it to decrypt the Content Key, which is then used to decrypt the actual ePub or PDF assets.

# From Decryptor.decrypt
aes = AES.new(self.book_key, AES.MODE_CBC, data[:16])
data = aes.decrypt(data[16:])
# Fix padding and decompress
data, was_decomp = self.decompress(data)

The data is decrypted using the AES Cipher Block Chaining (CBC) mode, where the first 16 bytes of the data serve as the Initialization Vector (IV). Finally, it also removes the padding and decompresses the content using zlib.

Also the "key" or passphrase in this case is your email (if you downloaded from the internet archive) or library card number (if you downloaded from a public library) or order number (if you got it from an online bookstore)

[maddz]

Where is the key taken from? Unlike ADE, I don't have any Thorium app set up, I just have the DeLCPL plug-in for Calibre (similar to the DeASCM plug-in).

[DNSB]

Originally Posted by maddz

Where is the key taken from? Unlike ADE, I don't have any Thorium app set up, I just have the DeLCPL plug-in for Calibre (similar to the DeASCM plug-in).

You need to enter the passphrase you used for the download for the decryption. Most sites allow you to enter a passphrase used for that site while others use part of your email address

[JSWolf]

Originally Posted by vila

the fact that they are charging money for such a simple decryption is wild xD

Now they are not the only ones to be able to remove LCP DRM.

[ownedbycats]

I wonder if the person who posted on gitflic took it from epubor. if so, fair turnabout....

[PeterT]

Actually the python code added to the plugin is based on the original python source that was removed from DeDRM with just some additions to handle additional LCP versions. A diff shows many, many identical lines of code

Sent from my TB350FU using Tapatalk

[vila]

Originally Posted by PeterT

Actually the python code added to the plugin is based on the original python source that was removed from DeDRM with just some additions to handle additional LCP versions. A diff shows many, many identical lines of code

Sent from my TB350FU using Tapatalk

I just checked and was about to say haha
also found https://gitflic.ru/project/mikudaio/dedrm-lcp-extended

[DNSB]

Originally Posted by vila

I just checked and was about to say haha
also found https://gitflic.ru/project/mikudaio/dedrm-lcp-extended

That is a repackage of the original NoDRM release 10.0.3 with the lcpdedrm.py code in place.

[maddz]

Originally Posted by DNSB

You need to enter the passphrase you used for the download for the decryption. Most sites allow you to enter a passphrase used for that site while others use part of your email address

Don't think I was asked for one when I downloaded the book from NetGalley...

Well, I'll have a bash at installing it later on.

[pdurrant]

Extremely wide images converted to attachments.