RockyDB

Simple NoSQL database written in Python. It relies on rocksdb as its storage engine. This is more of a Proof-of-concept than a production-ready database.

Installation

pip install rockydb

Contents

• RockyDB
• Installation
• Contents
• Features
• Documentation
  • Create collection
  • Insert doucment
  • Get document
  • Delete document
  • Query

Features

Currently under active development, however here is the feature list so far:

• Create collections
• Insert, get and delete documents
• REST API
• Query language
• Indexes
• Full-text Search [IN-DEVELOPMENT]

Performance

Dataset: NBA Players Dataset. Computer: MacBook Pro (13-inch, 2019). RockyDB is still in its early days, these results will likely get better in the future.

Documentation

Full Documentation. Below are the basics:

Create collection

from rockydb import RockyDB

db = RockyDB(path="database/")
news = db.collection("news")

Insert document

Supported data types: str, int, float, bool and list. Will support more later.

doc_id = news.insert({
  "title": "Can store strings",
  "year": 2022,
  "people": ["lists", "are", "fine", "too"],
  "pi": 3.14,
  "real": True
})

The insert method will return a unique document _id. _id will be created if document does not contain it.

Get document

news.get(doc_id)

Delete document

news.delete(doc_id)

Query

news.find({"pi?lt": 3.14, "real": True}, limit=10)

The limit arg is optional, default is 10. Supports exact, lte, lt, gt and gte queries.