asyncsqlpy 0.1.3

Async SQL + Redis wrapper with Mongo-like interface, fire-and-forget SQL background updates.

asyncsqlpy

asyncsqlpy is an ultra-fast, asynchronous SQL layer that mimics MongoDB-style operations with full Redis caching and background write queue support.
It’s designed for high-performance systems that need MongoDB-like flexibility over PostgreSQL.

---

🚀 Features

• ⚡ Async I/O – Fully asynchronous using aiopg and aioredis
• 🧠 MongoDB-like syntax – Use familiar methods like find_one, insert_one, update_many, etc.
• 🔁 Redis caching – Automatic query caching with TTL
• 🧩 Background SQL Worker Queue – Batched inserts/updates/deletes to reduce I/O overhead
• 🔎 Pub/Sub Watcher – Real-time change streaming via Redis channels
• 💥 Automatic Retry + Reconnect – Fault-tolerant retry for transient SQL/Redis errors
• 🧰 Dynamic Collections – Access collections as attributes or subscripts (e.g. db.users or db["users"])
• 🧹 Safe Shutdown – Waits for all pending operations before closing

---

🧑‍💻 Installation

pip install asyncsqlpy

---

⚙️ Quick Start

import asyncio
from asyncsqlpy import AsyncSqlDB

DSN = "postgres://user:password@hostname:port/dbname?sslmode=require"
REDIS_URL = "redis://localhost:6379"

async def main():
    db = AsyncSqlDB(DSN, REDIS_URL)
    await db.connect()

    users = db["users"]

    # Insert
    await users.insert_one({"name": "Alice", "age": 25})

    # Find
    user = await users.find_one({"name": "Alice"})
    print(user)

    # Update
    await users.update_many({"name": "Alice"}, {"$set": {"age": 26}})

    # Delete
    await users.delete_one({"name": "Alice"})

    # Close safely (waits for background queue to finish)
    await db.close()

asyncio.run(main())

---

🔄 Watcher (Real-Time Stream)

import asyncio

async def listener(data):
    print("Database change detected:", data)

async def watch_changes():
    db = AsyncSqlDB(DSN, REDIS_URL)
    await db.connect()
    users = db["users"]
    await users.watch(listener)

asyncio.run(watch_changes())

---

🧮 API Reference

AsyncSqlDB

Method	Description
connect()	Initialize PostgreSQL + Redis connection
close()	Gracefully close and flush all background tasks
__getitem__ / __getattr__	Access collection dynamically

AsyncMongoLikeCollection

Method	Description
find_one(filter)	Fetch one document
find(filter)	Return list of matching documents
insert_one(doc)	Insert a new document
insert_many(docs)	Insert multiple documents
update_one(filter, update)	Update a single document
update_many(filter, update)	Update multiple documents
delete_one(filter)	Delete a single document
delete_many(filter)	Delete multiple documents
count_documents(filter)	Count matching documents
watch(callback)	Listen for real-time change events via Redis

---

🧱 Background Worker Queue

All write operations (insert/update/delete) are queued and flushed in batches to PostgreSQL, improving throughput dramatically under high load.

await db.close()  # ensures all batched operations are written before shutdown

---

⚠️ Error Handling

• Automatic retries with exponential backoff for SQL and Redis operations
• Transparent reconnects for transient connection failures
• Warnings are logged via Python’s logging module

---

🧰 Example Architecture

         ┌────────────────┐
         │   Your App     │
         └──────┬─────────┘
                │
        Async API Calls
                │
         ┌──────▼─────────┐
         │  asyncsqlpy    │
         │  (Mongo-like)  │
         └──────┬─────────┘
                │
     ┌──────────┴───────────┐
     │                      │
┌────▼─────┐           ┌────▼────┐
│PostgreSQL│           │ Redis   │
│(storage) │           │(cache + │
│          │           │ pub/sub)│
└──────────┘           └─────────┘

---

🧾 License

MIT License © 2025 Sathishzus

---

💬 Author

Sathishzus – Open Source Systems & Cloud Performance Tools
🔗 GitHub | 🌐 Website