Alphanumeric ASCII homoglyphs / confusing character sets

homoglyph-notes.md

ascii subsets of visually similar characters

Nice answers for unicode in: https://security.stackexchange.com/questions/128286/list-of-visually-similar-characters-for-detecting-spoofing-and-social-engineeri

filtered from the unicode list on codebox/homoglyph

» # char_codes.txt is https://github.com/codebox/homoglyph/blob/3f61c31f4bb01c5312a7763dc6e226729417c8e7/raw_data/char_codes.txt
» python ~/to_ascii_char_codes.py < ~/Downloads/char_codes.txt 
['0', 'O', 'o']
['1', 'I', 'l']

:hwine suggested looking at the pwgen -B filter for ambiguous chars too:

const char *pw_ambiguous = "B8G6I1l0OQDS5Z2";

https://sourceforge.net/p/pwgen/code/ci/c25787fce93b2c99efd87219585c0531e80b6d1c/tree/pw_rand.c#l20

to_ascii_char_codes.py

import sys
import string


def main():
    alnum = frozenset(string.ascii_letters + string.digits)

    for line in sys.stdin.readlines():
        if line.startswith('#'):
            continue

        # filter for ascii code points
        code_points = [int(code_point, 16) for code_point in line.split(",") if int(code_point, 16) < 0x7f]
        if len(code_points) < 2:
            continue

        # filter for alphanumeric chars
        alnum_chars = [chr(c) for c in code_points if chr(c) in alnum]
        if not alnum_chars:
            continue

        # print sets of confusing chars
        print(alnum_chars)

        # file is sorted low to high and we're out of ascii range
        if code_points[0] > 0x7f:
            break


main()