Fuzzy Search Library

A flexible fuzzy search library supporting both MongoDB and PostgreSQL databases with multiple algorithms.

Features

• 🔍 Multiple Algorithms: Levenshtein, Jaro-Winkler, and Sørensen-Dice coefficient
• 🔄 Cross-Database Support: Works with both MongoDB and PostgreSQL
• ⚙️ Configurable: Customizable thresholds, limits, and sorting options
• 📦 TypeScript Support: Full TypeScript definitions included
• 🚀 Easy Integration: Simple API with adapter pattern
• 📊 Performance Optimized: Caching and query optimization

Installation

npm install fuzzy-search-lib

Quick Start

With MongoDB

import { FuzzySearch } from 'fuzzy-search-lib';

const searcher = new FuzzySearch({
  database: 'mongodb',
  connectionString: 'mongodb://localhost:27017/mydb',
  collection: 'users'
});

await searcher.connect();

const results = await searcher.search('john', {
  fields: ['name', 'email'],
  threshold: 0.6,
  limit: 10
});

await searcher.disconnect();

With PostgreSQL

import { FuzzySearch } from 'fuzzy-search-lib';

const searcher = new FuzzySearch({
  database: 'postgresql',
  connectionString: 'postgresql://user:pass@localhost:5432/mydb',
  table: 'users'
});

await searcher.connect();

const results = await searcher.search('john', {
  fields: ['first_name', 'last_name'],
  threshold: 0.7
});

await searcher.disconnect();

API

FuzzySearch Constructor

new FuzzySearch(config: FuzzySearchConfig)

FuzzySearchConfig Properties:

• database: 'mongodb' | 'postgresql'
• connectionString: Database connection string
• collection: Collection name (required for MongoDB)
• table: Table name (required for PostgreSQL)

Search Method

search(query: string, options: SearchOptions): Promise<SearchResult<T>>

SearchOptions Properties:

• fields: Array of field names to search in
• threshold: Minimum similarity score (0-1, default: 0.5)
• limit: Maximum number of results (default: 10)
• offset: Number of results to skip (default: 0)
• algorithm: Algorithm to use ('levenshtein' | 'jaro-winkler' | 'dice-coefficient', default: 'levenshtein')

Algorithms

Levenshtein Distance

Measures the minimum number of single-character edits required to change one word into another.

Jaro-Winkler Distance

Designed for short strings like person names, gives more favorable ratings to strings that match from the beginning.

Sørensen-Dice Coefficient

Measures the similarity between two strings based on the number of common bigrams.

License

MIT