applephrase 1.0.1

Generate memorable passphrases using pseudowords.

applephrase

applephrase is a commandline utility to generate memorable passphrases composed of a number of randomized pseudowords and a brief random string. The pseudowords can be generated for a number of languages: Dutch, French, English, German, Polish, Basque, Spanish, Italian, Serbocroatian Latin, Serbocroatian Cyrillic, Vietnamese, Estonian.

Some examples:

Odešišćima-Krškrme-aKpmvO
Ушоут-Џкићак-ON420G
Brühölde-Beruhrem-dg6fgP
Toevenboot-Rijswormen-yOrofH
Kindati-Mirdez-tmGS5k
Curpina-Beonangica-JVMK0t
Démièrea-Rammardée-nhlpwd
Naaled-Läägalide-R6qy04
Oằnẻo-Namchất-KaRrOm
Settenti-Ivonendo-kfOC8m
Trazonents-Oluviungly-nBwpng
Ulywcał-Ścipczyk-6hmSx7

The idea is that the use of pseudowords generates enough entropy in fewer words compared to regular passphrases, and protect against dictionary attacks. A terminating number of random alphanumeric characters helps get around the most common password requirements, and of course adds to the entropy. Since the pseudowords are (hopefully) pronouncable and the random characters limited in number, the resulting passphrase is memorable and still quite short.

Install

You can download applephrase by installing it from PyPI!

pip install applephrase

Alternatively, install or run it from a clone of this repo.

Usage

Run applephrase with desired iso-code of the language you wish to generate passwords for, e.g.:

$ applephrase nl

Use the options to generate "true" correct-horse-battery-staple-type passwords:

$ applephrase nl --nw 4 --nc 0 --sep "" --nocaps

For a list of all options and languages, type applephrase --help.

Wuggy

This project makes heavy use of Wuggy, the software that generates the pseudowords. The upstream Python package does not install anymore, but this is easily resolved by removing the unnecessarily strict versions from the requirements file. I made some more modifications to it and added the Estonian language found in one of the forks. Wuggy is not very fast but it is fast enough and is the only project I know of to be available in Python and generate pseudowords in languages other than English.

For Serbocroatian Cyrilic and German I had to add a workaround that increases runtime, so don't be worried when it takes a minute.

I include a copy of my modified Wuggy in applephrase/wuggy. If anyone is interested, I can break it out again (or help upsteam update their package).

Alternatives

https://github.com/chrplr/openlexicon/tree/master/apps/unipseudo is fast, but written in R/Shiny. Sources seem available, but it's unclear what license the lexicon is provided under.