VectorDatabase

VectorDatabase is a Swift library for embedding and storing text data using SQLite and USearch. It provides functionality to insert, lookup, and search text embeddings.

Installation

To use VectorDatabase in your project, add the following dependency to your Package.swift file:

.package(url: "https://github.com/adamwulf/VectorDatabase", .branch("main"))

Then, run swift package update to fetch the package.

Usage

Importing the Library

To use VectorDatabase, import it at the top of your Swift file:

import VectorDatabase

Initializing the Database

Create an instance of VectorDatabase by specifying the path where the database files will be stored:

do {
    let vectorDB = try VectorDatabase(at: "/path/to/database")
} catch {
    print("Failed to initialize VectorDatabase: \(error)")
}

Inserting Text

Insert text into the database and get a unique key for the text:

do {
    let key = try await vectorDB.insert(text: "example text")
    print("Inserted text with key: \(key)")
} catch {
    print("Failed to insert text: \(error)")
}

Looking Up Text

Lookup text by its content or by its unique key:

do {
    let result = try vectorDB.lookup(text: "example text")
    print("Found text: \(result.text), embedding: \(result.embedding)")
} catch {
    print("Text not found: \(error)")
}

do {
    let result = try vectorDB.lookup(key: someKey)
    print("Found text: \(result.text), embedding: \(result.embedding)")
} catch {
    print("Text not found: \(error)")
}

Searching for Similar Texts

Search for texts similar to a given text or embedding:

do {
    let results = try vectorDB.search(text: "example text", count: 10)
    for result in results {
        print("Found key: \(result.key), distance: \(result.distance)")
    }
} catch {
    print("Search failed: \(error)")
}

do {
    let embedding: [Double] = // some embedding
    let results = try vectorDB.search(embedding: embedding, count: 10)
    for result in results {
        print("Found key: \(result.key), distance: \(result.distance)")
    }
} catch {
    print("Search failed: \(error)")
}

Example Command Line Tool

The package also includes a command line tool vecdb for embedding and storing input text into a database. You can use it as follows:

$ swift run vecdb --help
$ swift run vecdb example /path/to/database

This will prime the database with some example texts and perform a search.

License

This project is licensed under the MIT License. See the LICENSE.md file for details.