pydepends 0.1.4

A python dependency injection system

PyDepends

A lightweight Python dependency injection library that simplifies managing dependencies in both synchronous and asynchronous code, using decorators in a FastAPI-like style.

Features

• Dependency injection with support for synchronous and asynchronous dependencies.
• Dependency overrides via a provider for easy testing and configuration.
• Context-managed dependency lifecycles with support for generators and async generators.
• Simple API with decorators and wrappers for clean, maintainable code.

---

Installation

pip install pydepends

Quickstart

Define dependencies in a simple way.

from asyncio import run
from pydepends import inject, Depends, Provider

async def dependency() -> int:
    return 5

provider = Provider()

@inject(provider)
async def main(x: int, y: int, z: int = Depends(dependency)) -> int:
    return x + y + z

assert run(main(1, 2)) == 8

Here’s a simple example showing how to define dependencies as a dependency tree:

from pydepends import inject, Depends 
 
def left_leaf_dependency():
    return 2

async def right_leaf_dependency():
    yield 3 #generators supported
 
async def right_node_dependency(leaf = Depends(right_leaf_dependency)):
    return leaf * 5
 
async def root_dependency(left: int = Depends(left_leaf_dependency), right: int = Depends(right_node_dependency)) -> int:
    return left * right * 7

Inject the dependencies in a sync function:

from pydepends import Provider
 
provider = Provider()

@inject(provider)
def handle_dependency(root: int = Depends(root_dependency)) -> int:
    return root * 11
 
value = handle_dependency()

assert value == 2 * 3 * 5 * 7 * 11
print(f"Computed value: {value}")  # Output: Computed value: 2310

Or inject them in an async function. Sync dependencies are put into an asyncio thread to avoid blocking the event loop.

@inject(provider)
async def async_handle_dependency(root: int = Depends(root_dependency)):
    return root*11

async def main():
    value = await async_handle_dependency()
    assert value == 2*3*5*7*11 
    print(f"Computed value: {value}")  # Output: Computed value: 2310


import asyncio
asyncio.run(main())

How it works

• Depends wraps a callable into a Dependency object.
• Provider allows overriding dependencies (useful for testing or different environments).
• The @inject decorator resolves and injects dependencies based on the provider.
• Supports both synchronous and asynchronous callables, managing contexts as needed.

License

This project is licensed under the Apache License, Version 2.0 — see the LICENSE file for details.

© 2025 Eric Hermosis. All rights reserved.

---

Contributing

Contributions are welcome! Please open issues or pull requests for bugs, features, or enhancements.

---