MemDB

A lightweight, in-memory JSON database for Node.js designed for speed and simplicity. Intended for single-process environments only.

memdb keeps all data strictly in memory. There is no persistence.

Features

• Mongo-style Queries: $gt, $lt, $gte, $lte, $ne, $in, $nin, $regex
• In-memory Only: All reads and writes operate on live memory
• Date Support: Automatic normalization and comparison of Date values
• Sorting: Multi-key ascending / descending sort
• Pagination: Built-in limit and skip
• Minimal API: Only get() and set()
• Synchronous Writes: set() is fully synchronous

Installation

npm i memdb

Usage

var memdb = require('memdb')

// Single in-memory collection
var db = memdb()

db.set({ type: 'user', name: 'Alice' })
db.set({ type: 'user', name: 'Bob' })

var users = db.get({ type: 'user' })

Usage Examples

1. Inserting Data

Passing a single object inserts a new document. A UUID id is generated automatically if missing.

db.set({
  name: 'Project Alpha',
  status: 'pending',
  priority: 1,
  createdAt: new Date()
})

2. Querying with Operators

// Numeric and inequality operators
var tasks = db.get({
  priority: { $gte: 5 },
  status: { $ne: 'archived' }
})

// Regex and array operators
var results = db.get({
  name: { $regex: /^Project/i },
  tags: { $in: ['urgent', 'active'] }
})

3. Updating Data

// Update all pending tasks
db.set({ status: 'pending' }, { status: 'active' })

// Update by ID
db.set({ id: 'some-uuid' }, { progress: 100 })

4. Deleting Data

// Delete a single record
db.set({ id: 'some-uuid' }, null)

// Delete all completed tasks
db.set({ status: 'completed' }, null)

// Clear entire database
db.set({}, null)

5. Pagination and Sorting

var page = db.get(
  { type: 'log' },
  { sort: { createdAt: -1 }, limit: 10, skip: 10 }
)

API Reference

| Method | Description | | ----------------------- | ------------------------------------------------------------- | | get(query, [options]) | Returns matching documents. Supports { limit, skip, sort }. | | set(query, [values]) | Insert, update, or delete documents. |

Execution Model

• All data lives in memory
• No filesystem access
• No async behavior
• No background work
• Process exit discards all data

Limitations

• Single process only
• No persistence
• No transactions
• No indexes
• Linear scan queries

Designed for embedded, ephemeral, test, cache, and control-plane use cases.

License

ISC.

Acknowledgements

Created by Vidar Eldøy