floatingnumber / bkcrack

Crack legacy zip encryption with Biham and Kocher's known plaintext attack.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

bkcrack

CI badge release badge license badge GitHub Sponsors badge

Crack legacy zip encryption with Biham and Kocher's known plaintext attack.

Overview

A ZIP archive may contain many entries whose content can be compressed and/or encrypted. In particular, entries can be encrypted with a password-based symmetric encryption algorithm referred to as traditional PKWARE encryption, legacy encryption or ZipCrypto. This algorithm generates a pseudo-random stream of bytes (keystream) which is XORed to the entry's content (plaintext) to produce encrypted data (ciphertext). The generator's state, made of three 32-bits integers, is initialized using the password and then continuously updated with plaintext as encryption goes on. This encryption algorithm is vulnerable to known plaintext attacks as shown by Eli Biham and Paul C. Kocher in the research paper A known plaintext attack on the PKZIP stream cipher. Given ciphertext and 12 or more bytes of the corresponding plaintext, the internal state of the keystream generator can be recovered. This internal state is enough to decipher ciphertext entirely as well as other entries which were encrypted with the same password. It can also be used to bruteforce the password with a complexity of nl-6 where n is the size of the character set and l is the length of the password.

bkcrack is a command-line tool which implements this known plaintext attack. The main features are:

  • Recover internal state from ciphertext and plaintext.
  • Change a ZIP archive's password using the internal state.
  • Recover the original password from the internal state.

Install

Precompiled packages

You can get the latest official release on GitHub.

Precompiled packages for Ubuntu, MacOS and Windows are available for download. Extract the downloaded archive wherever you like.

On Windows, Microsoft runtime libraries are needed for bkcrack to run. If they are not already installed on your system, download and install the latest Microsoft Visual C++ Redistributable package.

Compile from source

Alternatively, you can compile the project with CMake.

First, download the source files or clone the git repository. Then, running the following commands in the source tree will create an installation in the install folder.

cmake -S . -B build -DCMAKE_INSTALL_PREFIX=install
cmake --build build --config Release
cmake --build build --config Release --target install

Thrid-party packages

bkcrack is available in the package repositories listed below. Those packages are provided by external maintainers.

Packaging status

Usage

List entries

You can see a list of entry names and metadata in an archive named archive.zip like this:

bkcrack -L archive.zip

Entries using ZipCrypto encryption are vulnerable to a known-plaintext attack.

Recover internal keys

The attack requires at least 12 bytes of known plaintext. At least 8 of them must be contiguous. The larger the contiguous known plaintext, the faster the attack.

Load data from zip archives

Having a zip archive encrypted.zip with the entry cipher being the ciphertext and plain.zip with the entry plain as the known plaintext, bkcrack can be run like this:

bkcrack -C encrypted.zip -c cipher -P plain.zip -p plain

Load data from files

Having a file cipherfile with the ciphertext (starting with the 12 bytes corresponding to the encryption header) and plainfile with the known plaintext, bkcrack can be run like this:

bkcrack -c cipherfile -p plainfile

Offset

If the plaintext corresponds to a part other than the beginning of the ciphertext, you can specify an offset. It can be negative if the plaintext includes a part of the encryption header.

bkcrack -c cipherfile -p plainfile -o offset

Sparse plaintext

If you know little contiguous plaintext (between 8 and 11 bytes), but know some bytes at some other known offsets, you can provide this information to reach the requirement of a total of 12 known bytes. To do so, use the -x flag followed by an offset and bytes in hexadecimal.

bkcrack -c cipherfile -p plainfile -x 25 4b4f -x 30 21

Decipher

If the attack is successful, the deciphered data associated to the ciphertext used for the attack can be saved:

bkcrack -c cipherfile -p plainfile -d decipheredfile

If the keys are known from a previous attack, it is possible to use bkcrack to decipher data:

bkcrack -c cipherfile -k 12345678 23456789 34567890 -d decipheredfile

Decompress

The deciphered data might be compressed depending on whether compression was used or not when the zip file was created. If deflate compression was used, a Python 3 script provided in the tools folder may be used to decompress data.

python3 tools/inflate.py < decipheredfile > decompressedfile

Unlock encrypted archive

It is also possible to generate a new encrypted archive with the password of your choice:

bkcrack -C encrypted.zip -k 12345678 23456789 34567890 -U unlocked.zip password

The archive generated this way can be extracted using any zip file utility with the new password. It assumes that every entry was originally encrypted with the same password.

Recover password

Given the internal keys, bkcrack can try to find the original password. You can look for a password up to a given length using a given character set:

bkcrack -k 1ded830c 24454157 7213b8c5 -r 10 ?p

You can be more specific by specifying a minimal password length:

bkcrack -k 18f285c6 881f2169 b35d661d -r 11..13 ?p

Learn

A tutorial is provided in the example folder.

For more information, have a look at the documentation and read the source.

Contribute

Do not hesitate to suggest improvements or submit pull requests on GitHub.

If you would like to show your support to the project, you are welcome to make a donation or sponsor the project via Github Sponsors.

License

This project is provided under the terms of the zlib/png license.

About

Crack legacy zip encryption with Biham and Kocher's known plaintext attack.

License:zlib License


Languages

Language:C++ 95.7%Language:Python 2.4%Language:CMake 1.9%