swearing-filter

This Regex based TypeScript library provides robust and customizable filtering of swear words and offensive language from text.

Features

Multiple filtering rules: Regular expression-based filtering for highly flexible control.

Customizable word lists: Easily add, remove, or modify swear words to tailor the filter to your specific needs and context. Support for multiple languages is easily implemented by adding language-specific word lists.

Contextual awareness: Employ advanced techniques to reduce false positives by considering word context (e.g., "ass" in "assassin" would not be flagged). This feature may require additional configuration or external dependencies.

Performance optimization: Designed for efficiency, minimizing impact on application performance, even with large word lists.

Easy integration: Clean, well-documented API for seamless integration into any TypeScript project.

Zero dependency: It's built entirely from standard library components, ensuring minimal size, maximum portability, and reduced risk of conflicts with other libraries in your project.

Support languages

English, Russian, Finnish, Swedish, Chinese

Welcome to contribute more languages

Get started

npm i swearing-filter

Installation

// Using Node.js `require()`
const swearingFilter = require('swearing-filter');

// Using ES6 imports
import swearingFilter from 'swearing-filter';

Options

{
  // Placeholder for replacement swear words
  placeholder: "***",

  // Array of language for filtration
  languages: ['ru', 'en'],

 // In debug mode you can see a pattern, that was applied, and the original word
  debug: false
}

Usage

const swearingFilter = new swearingFilter({
  // options here
});

// Check if there any swear words in the text
swearingFilter.isBad('Original message with swear words'); 
// return: true

// Replace swear words with placeholder
swearingFilter.replace('Original message with swear words'); 
// return: "Original message with ***"

// Fixing swear words inside string (only for Russian language)
swearingFilter.fix('Original message with swear words'); 
// return: "Original message with fixed text"

Contribution

We are actively seeking new contributors to help expand and refine our filtering rules. Whether you have ideas for new regex patterns, improvements to existing ones, or support for additional languages, your contributions are welcome. Join us in making this library more robust and versatile for everyone.

Supported regex rules:

| Regex rule | ass | asset | compass | passed | |:----------- |:------:|:-------:|:---------:|:--------:| | ass | ✅ | ✅ | ✅ | ✅ | | ^ass | ✅ | ✅ | ❌ | ❌ | | ass$ | ✅ | ❌ | ✅ | ❌ | | ^ass$ | ✅ | ❌ | ❌ | ❌ | | ass[a-z]+ | ❌ | ✅ | ❌ | ✅ | | ^ass[a-z]+ | ❌ | ✅ | ❌ | ❌ |

Combine rules to increase speed:

^ass$|^asshole$

Escape regex special symbols: .+*?^$()[]{}|\ in words by using \\.

For example in rule ^\\$hit$, word "$hit" and $ was escaped