🌀 Retry Tools

A Python decorator for retrying functions with exponential backoff, optional jitter, and full support for both synchronous and asynchronous code.

🔍 About

This library provides a simple yet powerful @retry decorator that retries a function when specified exceptions are raised. It supports:

• Synchronous and asynchronous functions
• Exponential backoff with optional jitter
• Custom exception handling
• Logging or custom log output

Useful for handling flaky APIs, intermittent database errors, or any transient issues where retrying helps.

📦 Installation

pip install retrytools

🚀 Quick Start

Basic Usage

from retrytools import retry

@retry(catch_errors=ValueError, tries=3, delay=1)
def flaky_function():
    # Simulates an error-prone operation
    if random.random() < 0.7:
        raise ValueError("Transient issue!")
    return "Success!"

result = flaky_function()
print(result)

With Async Function

import asyncio
from retrytools import retry

@retry(catch_errors=ConnectionError, tries=5, delay=0.5)
async def unstable_async_api():
    if random.random() < 0.5:
        raise ConnectionError("Temporary network glitch")
    return "Fetched!"

asyncio.run(unstable_async_api())

⚙️ Parameters

Argument	Type	Description
catch_errors	Type[Exception] or tuple	Exception(s) to retry on
tries	int	Max attempts (default: 3)
delay	float	Initial delay in seconds
throw_error	Exception (optional)	Custom error to raise after final failure
logger	Logger or Callable[[str]]	Logging handler (e.g., print or logging.Logger)
jitter	bool or float	Random delay variation: True for full jitter, float for additive

📝 Example with Logging

import logging
from retrytools import retry

logger = logging.getLogger(__name__)
logging.basicConfig(level=logging.INFO)

@retry(catch_errors=(RuntimeError, ConnectionError), tries=4, throw_error=Exception("Custom"), delay=2, logger=logger, jitter=True)
def sometimes_fails():
    if random.randint(0, 1):
        raise RuntimeError("Oops!")
    return "Got it!"

sometimes_fails()

📌 Notes

• Jitter is helpful to avoid retry storms in distributed systems.
• Supports Python 3.7+.
• Works transparently for both sync and async code.

🛠️ License

MIT License