Over a two week Computer Science course in Cambridge I was set the challenge to create an encryption engine. My encryption engine has 3 layers. The first is a software simulation of enigma, with some improvements of course. The second layer pseudo-randomly scrambles the order of the letters based on a list of numbers inputted. Lastly, a substitution cipher replaces the encrypted letters with 4-10 letter random strings.
The main file is encrypt.py, where there is both and encrypt and decrypt function. By default, the program will prompt the user for text to encrypt, then you will be given th choice to decrypt a message or end the program. Since it only uses the command line as the UI the program doesn't accept any spacing (e.g. \n or \t) apart from spaces. The encryption engine accepts all letters (both upper and lower case), symbols, punctuation and numbers, meaning there are 95 different inputs possiblities.
The first encryption layer is a software representation of the German enigma machine. It follows the same steps of passing the input through a plugboard, then a series of rotors, through a reflector and then back through the rotors and the plugboard. For more information: https://www.scienceabc.com/innovation/the-imitation-game-how-did-the-enigma-machine-work.html
However, the enigma machine was crackable because a letter couldn't be enconded as itself, this was due to hardware problems. Enigma 2.0 can encrypt a letter to itself. Furthermore, enigma 2.0 is not limited to 3 rotors like the normal enigma machines, you can have as many rotors as needed. Effectively making the machine have infinite possible parameters for encryption, this makes it uncrackable.
The second encryption layer effectively scrambles the orders of the letters. It is given a list of numbers, which represent number of times the string should be split, it then goes through each neighbouring split and swaps them round and mirrors each split, it repeats this for every number. If you pass a long enough list of numbers the letters become pseudo-randomly shuffled. However, it can easily be decrypted as long as you know the original list of numbers used.
This is a substitution cypher. For every letter passed in it is replaced by one of 4 randomly generated strings of from 4 to 10 letters long, seperated by as many potential dividers as wanted, can be set in the program.
The main improvement would be to get the ability to set the parameters for the engine through a GUI instead of having to go into the code. Also have the ability to set the parameters to decode messages from a different session, currently not possible as everything is randomly generated every time the program is run. Might add the GUI as a part of my site.
The main flaw with this encryption is that the key has to be transmitted to whoever wants to decrypt the message, and the key is also quite long. Would need to implement a public and private key for this to be a viable communication encryption engine.
These were class assignments, they get increasingly complex as we were allowed to use more data structures. Encryption 1&2 are generated by another python program as I didn't want to copy and paste all the code. (Work smarter not harder!)
For encryption 3 and 4 we were allowed lists and for encryption 5 we were allowed to use dictionaries.