hash.py - Dec 9, 2018

import hashlib
import os
import tkinter as tk
import tkinter.messagebox

# Cryptographic hashes are used to determine data integrity. A key feature of hashes
# is that they are designed to be collision resistant: nobody should be able to find
# two different input values that result in the same output hash.
def hash_file(file_name):
    BLOCKSIZE = 65536
    # Use the SHA256 hashing algorithm
    hasher = hashlib.sha256()
    with open(file_name, 'rb') as afile:
        buf = afile.read(BLOCKSIZE)
        while len(buf) > 0:
            hasher.update(buf)
            buf = afile.read(BLOCKSIZE)
    return hasher.digest()

words = ["apple", "bagel", "car", "daring", "electronic", "fax", "gigantic", "honey", "ice", "journal",
         "kite", "laser", "mask", "new", "obelisk", "painter", "queen", "red", "satellite", "tokyo",
         "university", "vampire", "wiggly", "xylophone", "yoga", "zebra"]

assert(len(words) == 26)

def number_to_base(n, b):
    if n == 0:
        return [0]
    digits = []
    while n:
        digits.append(int(n % b))
        n //= b
    return digits[::-1]

# Get the directory that the script is in, not the current working directory
script_dir = os.path.dirname(os.path.abspath(__file__))

# Hash every file in the current directory and all subdirectories
hashes = []
for subdir, dirs, files in os.walk(script_dir):
    for file in files:
        # Include the script in the hash, to ensure that the code remains unchanged
        path = os.path.join(subdir, file)
        hashes.append(hash_file(path))

overall_hasher = hashlib.sha256()
# Sort the hashes so that the order the files were processed in is irrelevant
overall_hasher.update(b''.join(sorted(hashes)))
# Use a Merkle-style hash to compute the overall hash
# See https://en.wikipedia.org/wiki/Merkle_tree
overall_hash = overall_hasher.hexdigest()

overall_hash_base26 = number_to_base(int(overall_hash, 16), len(words))
overall_hash_mnemonic = ' '.join([words[i] for i in overall_hash_base26])

root = tk.Tk()
root.withdraw()

# Show the hash to the user
tk.messagebox.showinfo("File integrity", "SHA256 of files: \n{}\n\n{}".format(overall_hash, overall_hash_mnemonic))

root.destroy()