turkish-profanity-filter/src/utils.js

A configurable Turkish profanity filter for text content

// src/utils.js
// Helper functions for the library
exports.escapeRegExp = (string) => {
    return string.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
};

exports.createRegExp = (wordList, options = {}) => {
    const { wholeWords = true, caseSensitive = false } = options;

    // Filter out any empty strings to avoid regex errors
    const filteredList = wordList.filter(word => word && typeof word === 'string');

    if (filteredList.length === 0) {
        // Return a regex that will never match if the word list is empty
        return new RegExp('(?!)');
    }

    const escapedWords = filteredList.map(word => exports.escapeRegExp(word));

    // For non-ASCII characters like Turkish special characters, 
    // we need a different approach for whole words
    let pattern;
    if (wholeWords) {
        // This pattern matches whole words by checking for spaces, punctuation,
        // or string boundaries before and after the word
        pattern = `(^|[^a-zA-ZğüşıöçĞÜŞİÖÇ])(${escapedWords.join('|')})($|[^a-zA-ZğüşıöçĞÜŞİÖÇ])`;
    } else {
        pattern = `(${escapedWords.join('|')})`;
    }

    return new RegExp(pattern, caseSensitive ? 'g' : 'gi');
};