ZetoSearch

A lightweight full-text search library for Node.js written in TypeScript — supporting multi-field indexing, TF-IDF / BM25 ranking, fuzzy matching, and auto-suggest.
Perfect for building fast, offline, or local search features without Elasticsearch or external dependencies.

🚀 Features

• 🔍 Full-text search across multiple fields
• 📊 TF-IDF + BM25 ranking for relevance-based results
• ✨ Fuzzy search with similarity scoring
• ⚡ Auto-suggest for predictive typing
• 🧩 Custom filters, sorting, and pagination
• 📘 Written in TypeScript with type definitions
• 🪶 Lightweight and dependency-free (only uses natural for NLP)

📦 Installation

npm install zetosearch

🧩 Quick Start

1️. Import and Initialize

import { ZetoSearch } from "zetosearch";

const searchEngine = new ZetoSearch({
  searchFields: ["title", "content"],
  resultFields: ["id", "title", "content"],
});

2️. Index Documents

searchEngine.indexDocuments([
  { id: 1, title: "JavaScript Basics", content: "Learn programming with JS" },
  { id: 2, title: "Node.js Advanced", content: "Explore backend development" },
  { id: 3, title: "TypeScript Guide", content: "Strong typing for JavaScript" },
]);

3️. Search

const results = searchEngine.search("javascript", {
  fuzzyFactor: 0.2,
  limit: 5,
  debug: true,
});

console.log(results);

Output Example:

{
  "results": [
    { "id": 1, "title": "JavaScript Basics", "content": "Learn programming with JS", "score": 2.341 },
    { "id": 3, "title": "TypeScript Guide", "content": "Strong typing for JavaScript", "score": 1.126 }
  ],
  "total": 2,
  "totalResults": 2,
  "query": "javascript",
  "searchTime": 2
}

💡 Auto-Suggest Example

const suggestions = searchEngine.autoSuggest("jav");
console.log(suggestions);

Output:

["javascript", "java", "js"]

⚙️ Options

ZetoSearch Options

| Option | Type | Default | Description | |---------|------|----------|-------------| | searchFields | string[] | Required | Fields to index for search | | resultFields | string[] | Required | Fields to return in results | | idAlias | string | "id" | Field name to use as the unique document ID | | stopWords | string[] | Common English stop words | Tokens to ignore | | fuzzyFactor | number | 0.1 | Controls fuzziness (0–1) | | minTokenLength | number | 1 | Minimum token length | | maxTokenLength | number | 50 | Maximum token length | | enableStemming | boolean | false | Use Porter stemmer |

Note on idAlias:
The default idAlias is "id". You can change it depending on your data source:

• MongoDB usually uses "_id"
• SQL databases or other sources may use custom IDs like userId, emp_id, etc.

Search Options

| Option | Type | Default | Description | |---------|------|----------|-------------| | fields | string[] | All search fields | Limit search to specific fields | | fuzzyFactor | number | Inherited from options | Override fuzziness per search | | filter | (doc) => boolean | None | Custom document filtering | | sort | { by: string, order?: "asc" \| "desc" } | { by: "score", order: "desc" } | Sorting options | | limit | number | 10 | Number of results | | offset | number | 0 | Pagination offset | | debug | boolean | false | Include debug info in response |

📊 Utility Methods

Get Stats

console.log(searchEngine.getStats());

Output:

{
  "totalDocuments": 3,
  "totalTokens": 42,
  "avgDocLength": 14,
  "indexSize": 5321,
  "fieldsIndexed": ["title", "content"]
}

🛠️ Document Management Methods

Add Document

searchEngine.addDocument(doc);

Adds a new document to the index and updates average document length. This allows you to add new documents on-the-fly without having to re-index all existing documents.

Update Document

searchEngine.updateDocument(doc);

Updates an existing document in the index. If the document doesn't exist, no changes are made. Useful for modifying documents without rebuilding the full index.

Remove Document

searchEngine.removeDocument(docId);

Removes a document by its ID and updates the index. Returns true if removed, otherwise false. Allows you to delete documents without affecting the rest of the index.

🧾 License

ISC License

© 2025 Pujan Jani
Open source and free to use for personal or commercial projects.