Linear-time anagram detector

gistfile1.py

import string

# The first 26 primes (corresponding to A-Z)
ALPHABET_PRIMES = [2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97, 101]

def get_letter_prime(letter):
    """Given a letter A-Z (equivalent to a-z), return a unique prime number
    corresponding to that letter. Given a whitespace character, return 0. Given
    anything else, raise an error."""
    if letter in string.whitespace:
        return 0 # whitespace characters are valueless

    letter_ord = ord(letter.upper())
    
    if letter_ord < ord("A") or letter_ord > ord("Z"):
        raise ValueError("letter must be alphabetical or space")
    
    index = ord(letter.upper()) - ord("A")
    
    return ALPHABET_PRIMES[index]

def get_prime_letter_sum(letters):
    """Return a unique sum corresponding to the number of occurrences of
    each letter A-Z (equivalent to a-z) in the given input. Whitespace
    characters are ignored, and non-whitespace, non-alphabetical characters
    raise an error."""
    return reduce(lambda t, l: t + get_letter_prime(l), letters, 0)

def is_anagram_primality(a, b):
    """Returns whether or not a and b are anagrams. All whitespace
    characters are ignored, and non-whitespace, non-alphabetical
    characters raise an error."""
    return get_prime_letter_sum(a) == get_prime_letter_sum(b)

Has a few restrictions, sure, but it's pretty nice. Could be extended to all of ASCII if we really cared enough.