fast-python-events 1.0.0

A high-performance domain events library for Python applications.

Domain Events ⚡

Ultra-fast events system for Python, optimized for high performance and production-ready.

Features

• 🚀 7-8x faster than Django Signals
• 💾 Memory-efficient 60% less usage compared to Django Signals
• 🔧 Type-safe with DTOs
• 🎯 Simple and robust
• 🔌 Celery integration

Installation

Core dependencies

pip install -r requirements/requirements.txt

Cython acceleration (recommended for production)

Build and install the Cython module in editable mode so it's available throughout your environment:

pip install cython
pip install -e ./cython_core

This will compile and link the accelerated module automatically. No need to modify PYTHONPATH.

Development dependencies

Includes everything needed for testing, benchmarks and development:

pip install -r requirements/requirements-dev.txt

Quick Start

from domain_events import DomainEventSystem, DomainEvent
from dataclasses import dataclass

@dataclass(frozen=True)
class UserRegisteredEvent(DomainEvent):
    event_type = "user.registered"
    user_id: int
    email: str

system = DomainEventSystem()

def send_welcome_email(event: UserRegisteredEvent):
    print(f"Welcome {event.email}!")

system.register_event(
    "user.registered",
    UserRegisteredEvent,
    [(send_welcome_email, {"async": True})]
)

event = UserRegisteredEvent(user_id=1, email="[e-mail]")
system.publish(event)

Python Integration

In your code:

from domain_events import domain_events

domain_events.publish(user_event)

Celery Integration

from domain_events import DomainEventSystem
from celery import Celery

app = Celery()
event_system = DomainEventSystem(celery_app=app)

Performance

• Lookup: 40-60ns (8x faster than Django Signals)
• Dispatch (Cython): 90-120ns per handler (typically 2-10x faster than pure Python)
• Bulk dispatch (Cython): up to 10x faster for large batches
• Memory: 60% less usage compared to Django Signals
• Benchmark results:
  • Cython event dispatch: ~110ns per handler (10 handlers, single event)
  • Python event dispatch: ~170ns per handler (10 handlers, single event)
  • Cython bulk dispatch: ~25μs for 1000 events
  • Python bulk dispatch: ~88μs for 1000 events

For best results in production, always enable Cython acceleration.