Simplifying working Caesar cipher

Question

I'm learning Python 3 at the moment, so to test the skills I've learned, I am trying the puzzles at Python Challenge.

I've created some code to solve the 2nd puzzle here. it works, but I think that the way I've done it is very convoluted. Any suggestions on how to solve the puzzle in a simpler way?

(The code basically needs to substitute each letter in a message to the letter 2 spaces to the right of it, such as E->G.)

import string
alphabet = string.ascii_lowercase
letter=0
replaceLetter = 2
times=1
message = input()
newMessage=''
while times < 26:
 newMessage = message.replace(alphabet[letter], str(replaceLetter)+',')
 message = newMessage
 letter = letter + 1
 replaceLetter = replaceLetter + 1
 time = times + 1
 if letter == 26:
 times = 0
 break
newMessage = message.replace('26'+',', 'a')
message = newMessage
newMessage = message.replace('27'+',', 'b')
message = newMessage
number = 25
message = newMessage
while times < 26:
 newMessage = message.replace(str(number)+',', str(alphabet[number]))
 message = newMessage
 letter = letter + 1
 number = number - 1
 time = times + 1
 if number == -1:
 times = 0
 break
print(newMessage)

• Martin Beckett: Take a look at the python dictionary. Consider replacing characters?
• Mark Tolonen: There is a hint in the title of the challenge "What about making trans?". See maketrans. It can be only a couple of lines of code.

Answer 1

I would define a shift function that shifted the letters like so:

from string import whitespace, punctuation
def shift(c, shift_by = 2):
 if c in whitespace + punctuation: return c
 upper_ord, lower_ord, c_ord = ord('A'), ord('a'), ord(c)
 c_rel = (c_ord - lower_ord) if c_ord >= lower_ord else (c_ord - upper_ord)
 offset = lower_ord if c_ord >= lower_ord else upper_ord
 return chr(offset + (c_rel + shift_by) % 26)

Then, to translate a message:

msg = 'a quick brown fox jumped over the lazy dog'
encoded_msg = ''.join(shift(l) for l in msg)

Alternatively, combine Mark Tolonen's maketrans suggestion with g.d.d's deque suggestion to get:

import string
from collections import deque
alphabet = string.ascii_lowercase
alphabet_deque = deque(alphabet)
alphabet_deque.rotate(-2)
rotated_alphabet = ''.join(alphabet_deque)
tbl = string.maketrans(alphabet, rotated_alphabet)

Then, later in the code:

msg = 'a quick brown fox jumped over the lazy dog'
encoded_msg = string.translate(msg, tbl)

This second method only works for lowercase letters, but you can always create two separate translation tables -- one for uppercase letters and another for lowercase letters -- to account for case.

Offtopic note: sometimes I wish Python had Smalltalk-like cascaded message sends. Ah well, one can dream.

Answer 2

I would do a couple of things differently.

import string
from collections import deque
ascii1 = string.ascii_lowercase
# create a deque to simplify rotation.
d = deque(ascii1)
d.rotate(-2)
ascii2 = ''.join(d)
replacements = dict(zip(ascii1, ascii2))
oldmessage = 'This is a string that we want to run through the cipher.'
newmessage = ''.join(replacements.get(c.lower(), c) for c in oldmessage)
# results in 'vjku ku c uvtkpi vjcv yg ycpv vq twp vjtqwij vjg ekrjgt.'

Note that I didn't do anything here to account for casing.