fuzzy-search-lib/dist/algorithms/dice-coefficient.js

A flexible fuzzy search library supporting both MongoDB and PostgreSQL

"use strict";
Object.defineProperty(exports, "__esModule", { value: true });
exports.diceCoefficient = void 0;
/**
 * Generates bigrams (pairs of consecutive characters) from a string.
 *
 * @param str - Input string
 * @returns Array of bigrams
 */
function getBigrams(str) {
    const bigrams = [];
    for (let i = 0; i < str.length - 1; i++) {
        bigrams.push(str.substring(i, i + 2));
    }
    return bigrams;
}
/**
 * Calculates the Sørensen-Dice coefficient between two strings.
 *
 * @param s1 - First string
 * @param s2 - Second string
 * @returns Sørensen-Dice coefficient between 0 and 1
 */
function diceCoefficient(s1, s2) {
    // Handle edge cases
    if (s1 === s2)
        return 1;
    if (s1.length === 0 || s2.length === 0)
        return 0;
    if (s1.length === 1 && s2.length === 1)
        return s1 === s2 ? 1 : 0;
    if (s1.length < 2 || s2.length < 2)
        return 0;
    const bigrams1 = getBigrams(s1.toLowerCase());
    const bigrams2 = getBigrams(s2.toLowerCase());
    const set1 = new Set(bigrams1);
    const set2 = new Set(bigrams2);
    const intersection = new Set([...set1].filter(x => set2.has(x)));
    const totalBigrams = set1.size + set2.size;
    if (totalBigrams === 0)
        return 1;
    return (2 * intersection.size) / totalBigrams;
}
exports.diceCoefficient = diceCoefficient;
//# sourceMappingURL=dice-coefficient.js.map