CODEWITHSUNDEEP
pythonvigenere
Guimo
Guimo

Reputation: 732

Vigenere Cipher in Python bug

I'm trying to implement Vigenere's Cipher. I want to be able to obfuscate every single character in a file, not just alphabetic characters.

I think I'm missing something with the different types of encoding. I have made some test cases and some characters are getting replaced badly in the final result.

This is one test case:

,.-´`1234678abcde^*{}"¿?!"·$%&/\º

end

And this is the result I'm getting:

).-4`1234678abcde^*{}"??!"7$%&/:

end

As you can see, ',' is being replaced badly with ')' as well as some other characters.

My guess is that the others (for example, '¿' being replaced with '?') come from the original character not being in the range of [0, 127], so its normal those are changed. But I don't understand why ',' is failing.

My intent is to obfuscate CSV files, so the ',' problem is the one I'm mainly concerned about.

In the code below, I'm using modulus 128, but I'm not sure if that's correct. To execute it, put a file named "OriginalFile.txt" in the same folder with the content to cipher and run the script. Two files will be generated, Ciphered.txt and Deciphered.txt.

"""
Attempt to implement Vigenere cipher in Python.
"""

import os

key = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"

fileOriginal = "OriginalFile.txt"
fileCiphered = "Ciphered.txt"
fileDeciphered = "Deciphered.txt"

# CIPHER PHASE

if os.path.isfile(fileCiphered):
    os.remove(fileCiphered)

keyToUse = 0

with open(fileOriginal, "r") as original:
    with open(fileCiphered, "a") as ciphered:
        while True:
            c = original.read(1) # read char

            if not c:
                break

            k = key[keyToUse]
            protected = chr((ord(c) + ord(k))%128)
            ciphered.write(protected)
            keyToUse = (keyToUse + 1)%len(key)

print("Cipher successful")

# DECIPHER PHASE

if os.path.isfile(fileDeciphered):
    os.remove(fileDeciphered)

keyToUse = 0

with open(fileCiphered, "r") as ciphered:
    with open(fileDeciphered, "a") as deciphered:
        while True:
            c = ciphered.read(1) # read char

            if not c:
                break

            k = key[keyToUse]
            unprotected = chr((128 + ord(c) - ord(k))%128) # +128 so that we don't get into negative numbers
            deciphered.write(unprotected)
            keyToUse = (keyToUse + 1)%len(key)

print("Decipher successful")

Upvotes: 0

Views: 102

Answers (1)

dcrosta
dcrosta

Reputation: 26278

Assumption: you're trying to produce a new, valid CSV with the contents of cells enciphered via Vigenere, not to encipher the whole file.

In that case, you should check out the csv module, which will handle properly reading and writing CSV files for you (including cells that contain commas in the value, which might happen after you encipher a cell's contents, as you see). Very briefly, you can do something like:

with open("...", "r") as fpin, open("...", "w") as fpout:
    reader = csv.reader(fpin)
    writer = csv.writer(fpout)
    for row in reader:
        # row will be a list of strings, one per column in the row
        ciphered = [encipher(cell) for cell in row]
        writer.writerow(ciphered)

When using the csv module you should be aware of the notion of "dialects" -- ways that different programs (usually spreadsheet-like things, think Excel) handle CSV data. csv.reader() usually does a fine job of inferring the dialect you have in the input file, but you might need to tell csv.writer() what dialect you want for the output file. You can get the list of built-in dialects with csv.list_dialects() or you can make your own by creating a custom Dialect object.

Upvotes: 1

Related Questions