eventsource-py 0.1.2

Production-ready event sourcing library for Python

eventsource-py

A production-ready event sourcing library for Python 3.11+.

Features

• Event Store - PostgreSQL, SQLite, and In-Memory backends with optimistic locking
• Domain Events - Immutable event classes with Pydantic validation and versioning
• Event Registry - Thread-safe event type registration for serialization/deserialization
• Aggregate Pattern - Base classes for event-sourced aggregates with state reconstruction
• Repository Pattern - Clean abstractions for loading and saving aggregates
• Projection System - Checkpoint tracking, retry logic, and dead letter queue support
• Event Bus - In-Memory and Redis Streams backends for event distribution
• Transactional Outbox - Reliable event publishing pattern
• Multi-tenancy - Built-in tenant isolation support
• Observability - Optional OpenTelemetry integration

Installation

# Basic installation
pip install eventsource-py

# With PostgreSQL support
pip install eventsource-py[postgresql]

# With SQLite support
pip install eventsource-py[sqlite]

# With Redis support
pip install eventsource-py[redis]

# With OpenTelemetry support
pip install eventsource-py[telemetry]

# All optional dependencies
pip install eventsource-py[all]

Requirements

• Python 3.11+
• pydantic >= 2.0
• sqlalchemy >= 2.0 (for PostgreSQL backend)

Quick Start

1. Define Your Events

from uuid import UUID
from eventsource import DomainEvent, register_event

@register_event
class OrderCreated(DomainEvent):
    """Event emitted when an order is created."""
    event_type: str = "OrderCreated"
    aggregate_type: str = "Order"
    customer_id: UUID
    total_amount: float

@register_event
class OrderShipped(DomainEvent):
    """Event emitted when an order is shipped."""
    event_type: str = "OrderShipped"
    aggregate_type: str = "Order"
    tracking_number: str

2. Define Your Aggregate State

from pydantic import BaseModel

class OrderState(BaseModel):
    """State of an Order aggregate."""
    order_id: UUID
    customer_id: UUID | None = None
    total_amount: float = 0.0
    status: str = "draft"
    tracking_number: str | None = None

3. Create Your Aggregate

from eventsource import AggregateRoot

class OrderAggregate(AggregateRoot[OrderState]):
    """Event-sourced Order aggregate."""
    aggregate_type = "Order"

    def _get_initial_state(self) -> OrderState:
        return OrderState(order_id=self.aggregate_id)

    def _apply(self, event: DomainEvent) -> None:
        if isinstance(event, OrderCreated):
            self._state = OrderState(
                order_id=self.aggregate_id,
                customer_id=event.customer_id,
                total_amount=event.total_amount,
                status="created",
            )
        elif isinstance(event, OrderShipped):
            if self._state:
                self._state = self._state.model_copy(
                    update={
                        "status": "shipped",
                        "tracking_number": event.tracking_number,
                    }
                )

    def create(self, customer_id: UUID, total_amount: float) -> None:
        """Command: Create the order."""
        if self.version > 0:
            raise ValueError("Order already created")

        event = OrderCreated(
            aggregate_id=self.aggregate_id,
            customer_id=customer_id,
            total_amount=total_amount,
            aggregate_version=self.get_next_version(),
        )
        self.apply_event(event)

    def ship(self, tracking_number: str) -> None:
        """Command: Ship the order."""
        if not self.state or self.state.status != "created":
            raise ValueError("Order must be created before shipping")

        event = OrderShipped(
            aggregate_id=self.aggregate_id,
            tracking_number=tracking_number,
            aggregate_version=self.get_next_version(),
        )
        self.apply_event(event)

4. Use the Repository

import asyncio
from uuid import uuid4
from eventsource import InMemoryEventStore, AggregateRepository

async def main():
    # Set up event store and repository
    event_store = InMemoryEventStore()
    repo = AggregateRepository(
        event_store=event_store,
        aggregate_factory=OrderAggregate,
        aggregate_type="Order",
    )

    # Create and save a new order
    order_id = uuid4()
    order = repo.create_new(order_id)
    order.create(customer_id=uuid4(), total_amount=99.99)
    await repo.save(order)

    # Load the order and ship it
    loaded_order = await repo.load(order_id)
    loaded_order.ship(tracking_number="TRACK-123")
    await repo.save(loaded_order)

    # Verify the state
    final_order = await repo.load(order_id)
    print(f"Order status: {final_order.state.status}")
    print(f"Tracking: {final_order.state.tracking_number}")

asyncio.run(main())

Architecture

+-------------------+     +-------------------+     +-------------------+
|                   |     |                   |     |                   |
|    Commands       |---->|    Aggregates     |---->|   Event Store     |
|                   |     |                   |     |                   |
+-------------------+     +-------------------+     +--------+----------+
                                                             |
                                                             v
+-------------------+     +-------------------+     +-------------------+
|                   |     |                   |     |                   |
|   Read Models     |<----|   Projections     |<----|   Event Bus       |
|                   |     |                   |     |                   |
+-------------------+     +-------------------+     +-------------------+

Core Concepts

• Events - Immutable facts that capture state changes. Events are never deleted or modified.
• Event Store - Persists events with ordering and optimistic locking guarantees.
• Aggregates - Consistency boundaries that reconstruct state from event streams.
• Repository - Abstracts loading/saving aggregates from/to the event store.
• Projections - Build read-optimized views from event streams.
• Event Bus - Distributes events to subscribers for async processing.

Documentation

For comprehensive documentation, see the Documentation Index.

Getting Started

• Getting Started Guide - Installation and first steps
• Architecture Overview - System design and concepts
• FAQ - Frequently asked questions

Guides

• Multi-tenant Setup - Tenant isolation patterns and configuration
• Error Handling - Exception handling best practices
• Authentication - Security integration patterns
• Production Deployment - Production readiness checklist

API Reference

• Events API - DomainEvent and EventRegistry
• Event Stores - EventStore interface and implementations
• Aggregates - AggregateRoot and Repository
• Projections - Projection system
• Event Bus - Event distribution

Examples

• Basic Usage - Simple order example
• Multi-tenant Setup - Multi-tenancy configuration
• Projections - Building read models
• Testing Patterns - Unit and integration testing

Architecture Decision Records

• ADR Index - All architectural decisions
• ADR-0001: Async-First Design
• ADR-0002: Pydantic Event Models
• ADR-0003: Optimistic Locking

Developer Documentation

• Code Structure - Project organization
• Testing Guide - Testing strategies and patterns

Development

# Install in development mode
pip install -e ".[dev]"

# Run tests
pytest

# Run tests with coverage
pytest --cov=eventsource --cov-report=html

# Run type checking
mypy src/eventsource

# Run linting
ruff check src/eventsource

# Format code
ruff format src/eventsource

License

This project is licensed under the MIT License - see the LICENSE file for details.