README.md

Profanity 🧼

A multi-language profanity filter with full TypeScript support

Getting Started

Install the package

npm i @2toad/profanity

If you're using Node 11.x or older, you'll need to install Profanity 1.x

Usage

import { profanity, CensorType } from '@2toad/profanity';
// or
const { profanity, CensorType } = require('@2toad/profanity');

profanity.exists('I like big butts and I cannot lie');
// true

profanity.exists('I like big glutes and I cannot lie');
// false

profanity.censor('I like big butts (aka arses) and I cannot lie');
// I like big @#$%&! (aka @#$%&!) and I cannot lie

profanity.censor('I like big butts (aka arses) and I cannot lie', CensorType.FirstChar);
// I like big *utts (aka *rses) and I cannot lie

Options

Create an instance of the Profanity class to change the default options:

import { Profanity } from '@2toad/profanity';

const profanity = new Profanity({
    languages: ['de'],
    wholeWord: false,
    grawlix: '*****',
    grawlixChar: '$',
});

languages

By default, this is set to ['en'] (English). You can change the default to any supported language, including multiple languages:

const profanity = new Profanity({
    languages: ["en", "de"],
});

You can override this option by specifying the languages in exists or censor:

profanity.exists('Je suis un connard', ["fr"]);
// true

profanity.censor('I like big butts and je suis un connard', CensorType.Word, ["en", "de", "fr"]);
// I like big @#$%&! and je suis un @#$%&!

If no languages are specified in the method call, it will use the languages specified in the options.

wholeWord

By default, this is set to true so profanity only matches on whole words:

profanity.exists('Arsenic is poisonous but not profane');
// false

Setting this to false, results in partial word matches:

profanity.exists('Arsenic is poisonous but not profane');
// true (matched on arse)

Compound Words

Profanity detection works on parts of compound words, rather than treating hyphenated or underscore-separated words as indivisible.

When wholeWord is true, each portion of a compound word is analyzed for a match:

profanity.exists("Don't be an arsenic-monster");
// false

profanity.exists("Don't be an arse-monster");
// true (matched on arse)

Setting wholeWord to false, results in partial word matches on each portion of a compound word:

profanity.exists("Don't be an arsenic-monster");
// true (matched on arse)

grawlix

By default this is set to @#$%&!:

profanity.censor('I like big butts and I cannot lie');
// I like big @#$%&! and I cannot lie

Setting this to ****, results in:

profanity.censor('I like big butts and I cannot lie');
// I like big **** and I cannot lie

grawlixChar

When specifying a CensorType other than CensorType.Word, this is the character used by the censor function.

By default this is set to *:

profanity.censor('I like big butts and I cannot lie', CensorType.AllVowels);
// I like big b*tts and I cannot lie

Setting this to $, results in:

profanity.censor('I like big butts and I cannot lie', CensorType.AllVowels);
// I like big b$tts and I cannot lie

Customize the word list

Add words:

profanity.addWords(['aardvark', 'zebra']);

Remove words:

profanity.removeWords(['butt', 'arse']);

Whitelist

The whitelist allows you to specify words that are always ignored by the profanity filter.

This can be useful if you want to enable partial word matching (wholeWord = false), so combined words are caught (e.g., arselicker), while specific words you add to the whitelist are ignored (e.g., arsenic).

Add words to the whitelist:

profanity.whitelist.addWords(['arsenic', 'buttress']);

Remove words from the whitelist:

profanity.whitelist.removeWords(['arsenic', 'buttress']);

Benchmarking ⏱️

To see how Profanity performs, check out our benchmark results.

Contributing 🤝

So you want to contribute to the Profanity project? Fantastic! Please read the Contribute doc to get started.