RC4-variants

Sample Python 3 implementation of various RC4 variants, including regular RC4 (ARC4), RC4A, and RC4-drop[n] (MARK-4). RC4A uses two state arrays and keys instead as one, whereas RC4-drop[n] discards the first n bytes of the keystream after key scheduling to improve security. A n in the range 768 <= n <= 3072 is recommended, while n = 0 corresponds to regular RC4.

Currently, supported forms of input and output include plain text, hexadecimal text (without an 0x), and files.

Since RC4 (including RC4A and RC4-drop[n]) is widely believed to be insecure, using this implementation for non-educational purposes is not recommended. If you must, RC4-drop[n] with n = 3072 or RC4A should be more secure than RC4.

All included variants of RC4 are symmetric encryption algorithms, so you can decrypt the ciphertext by encrypting it again.

Usage

import RC4

rc4 = RC4.RC4("0102030405", 768) # key, offset
ciphertext_1 = rc4.cipher("0102030405", "hex", "plain") # hex plaintext, ascii ciphertext
ciphertext_2 = rc4.cipher("foo", "plain", "file") # ascii plaintext, file ciphertext
ciphertext_3 = rc4.cipher("bar", "file", "hex") # file plaintext, hex ciphertext
print(ciphertext_1, ciphertext_2, ciphertext_3)
rc4a = RC4.RC4A("0102030405") # key
ciphertext_4 = rc4a.cipher("0102030405", "hex", "plain") # hex plaintext, ascii ciphertext
ciphertext_5 = rc4a.cipher("foo", "plain", "file") # ascii plaintext, file ciphertext
ciphertext_6 = rc4a.cipher("bar", "file", "hex") # file plaintext, hex ciphertext
print(ciphertext_4, ciphertext_5, ciphertext_6)