opfsdb
v1.0.8
Published
Origin private file system DB
Downloads
1
Maintainers
Readme
OPFSDB
OPFSDB (Origin Private File System Database) is a TypeScript library that provides a lightweight and efficient in-browser database solution using the File System Access API and Web Workers. It utilizes the B+ Tree data structure from the serializable-bptree library for indexing and querying data, allowing for fast and scalable storage and retrieval of records.
You can find demo with table for data browsing and querying here. You can also load custom CSV file but for now "id" is required column.
Table of Contents
Features
- File System Access API: OPFSDB uses the File System Access API to create a private, sandboxed database for your application, ensuring data privacy and security.
- Web Workers: Utilizes Web Workers to allow for sync OPFS API that drastically improves read and write performance.
- Shared Worker API: OPFSDB leverages the Shared Worker API to enable communication between multiple Web Workers, allowing for multi-tab interaction with one database.
- Efficient Indexing: OPFSDB uses serializable-bptree, a high-performance B+ Tree implementation, for indexing and querying data.
- TypeScript Support: Written in TypeScript for better type safety and developer experience.
- Encoding/Decoding: OPFSDB encodes data in CBOR format using cbor-x library, also storing the structures alongside.
Architecture
OPFSDB has three main entry points:
Main Page Code: This is the code that runs in the main browser context. It is responsible for initializing WorkerManager and interacting with the database through the
sendCommand
method.Shared Worker: This is a special type of Web Worker that can be shared across multiple browser contexts. It acts as a communication bridge between the main page code and the dedicated Web Workers.
Dedicated Web Workers: These are the actual workers that perform the database operations. They can either execute queries directly or communicate with a master worker through the shared worker.
Installation
npm install opfsdb
pnpm add opfsdb
yarn add opfsdb
Getting Started
Main Thread
In your main application code, you can interact with OPFSDB using the sendCommand
method from the WorkerManager
class:
import { WorkerManager } from 'opfsdb/WorkerManager'
const workerManager = new WorkerManager('worker.js', 'shared-worker.js')
// Send a command to create a new table
workerManager.sendCommand({
name: 'createTable',
tableName: 'users',
keys: {
id: { type: 'string' },
name: { type: 'string', indexed: true },
age: { type: 'number', indexed: true },
},
})
// Insert data into the table
workerManager.sendCommand({
name: 'insert',
tableName: 'users',
record: { id: '1', name: 'John Doe', age: 30 },
})
// Query data from the table
workerManager
.sendCommand({
name: 'query',
tableName: 'users',
query: { age: { gte: 25 }, name: { like: 'John%' } },
})
.then(results => {
console.log('Query results:', results)
})
Shared Worker
The Shared Worker acts as a communication channel between different Web Workers. It connects with each Web Worker and tracks the master one, when one of the slave Web Workers receives a command from the Main Thread, it will communicate with the master Web Worker through the Shared Web Worker.
You can create a new Shared Worker file (shared-worker.js
) and import the SharedWorkerController
class from OPFSDB:
import { SharedWorkerController } from 'opfsdb/workers/SharedWorkerController'
const sharedWorkerController = new SharedWorkerController()
Web Worker
The Web Workers handle the actual execution of commands. They receive commands from the Main thread and, depending on whether it's the master Worker or not, they will either execute the command and return the results or send the command to the Shared Worker.
You can create a new Web Worker file (worker.js
) and import the DedicatedWorkerController
class from OPFSDB:
import { DedicatedWorkerController } from 'opfsdb/workers/DedicatedWorkerController'
const dedicatedWorkerController = new DedicatedWorkerController()
Docs
You can find typedoc in the docs folder Here