Writing English Words as Hexadecimal

I was curious as to what English words could be written solely with the 6 letters available in the hexadecimal numerical system, so I hammered together a short Python script which takes as its sole argument a filename of newline separated words and regurgitates the words that consist of only letters A–F.

import sys

if len(sys.argv) != 2:
    print('usage: onlyhex [FILE]')
    quit()

with open(sys.argv[1]) as fp:
    for word in fp.readlines():
        word = word.strip()
        if len(word) > 2 and set(word).issubset(set("abcdef")):
            print(word)

Running the script on some 400K English words in text files from a git repo returns numerous but obscure results. Instead you may want to run it on a list from another git repo of the 10-20K most common English words. This winnows the results into interesting words like cafe, ace, face, deaf, add, decaf, and fade. This means numbers like 0xfadedface and 0xdecafcafe are valid hexadecimal, corresponding to 67342564046 and 59805518590 in our familiar decimal system, respectively.

But why limit ourselves to only the letters A–F when we can represent letters such as O with the similarly looking digit zero—also valid in hexadecimal? The following Python script includes a small adjustment to the previous which allows words containing the letter "O" and converts them to the digit "0".

import sys

if len(sys.argv) != 2:
    print('usage: onlyhex [FILE]')
    quit()

with open(sys.argv[1]) as fp:
    for word in fp.readlines():
        word = word.strip()
        if len(word) > 2 and set(word).issubset(set("abcdefo")):
            print(word.translate(str.maketrans("o", "0")))

This script outputs new words such as dec0de and c0c0a. These words can be formed into phrases such as 0x0ddf00d, 0xafedface, 0xdefacedfacade, 0xdecafc0ffee, 0xdeafbee, 0xabadbed, 0xc0dedcab, all of which are valid hexadecimal. The following script is similar to the last one but it sorts the words into noun and/or adjective buckets and combines them into valid hexadecimal phrases.

import random
import sys

from nltk.corpus import wordnet

if len(sys.argv) != 2:
    print('usage: onlyhex [FILE]')
    quit()

nouns, adjectives = [], []

with open(sys.argv[1]) as fp:
    for word in fp.readlines():
        word = word.strip()
        if len(word) > 2 and set(word).issubset(set("abcdefo")):
            types = set(s.pos() for s in wordnet.synsets(word))
            word = word.translate(str.maketrans("o", "0"))
            if 'n' in types:
                nouns.append(word)
            if 'a' in types:
                adjectives.append(word)

for _ in range(10):
    print(f'0x{random.choice(adjectives)}{random.choice(nouns)}')

For example, running this on a corpus of words yields

$ python3 onlyhex.py words.txt
0xdeaffade
0x0fff00d
0xbeddedcaca0
0xdeadba0bab
0xdeaff00d
0xfacedface
0xdead0af
0xfacedabb
0xdeafba0bab
0x0ffdeaf