pytryagain 0.1.3

A lightweight, zero-dependency retry decorator for sync and async Python functions.

pytryagain

A lightweight, zero-dependency retry decorator for sync and async Python functions.

---

Table of Contents

• Installation
• Quick Start
• Examples
  • Basic usage
  • Async functions
  • Decorator factory
  • Limit retried exceptions
  • Backoff strategies
  • Timeout
  • Conditional retry
  • Callbacks
  • Async callbacks
• API Reference
• Backoff Strategies
• License

---

Installation

pip install pytryagain

---

Quick Start

from pytryagain import retry

@retry
def fetch_data(url: str) -> bytes:
    ...  # retried up to 3 times on any Exception

---

Examples

Basic usage

Use @retry as a plain decorator — 3 attempts with exponential jitter backoff by default:

from pytryagain import retry

@retry
def connect_to_db() -> Connection:
    return db.connect()

Customise the number of attempts with tries:

@retry(tries=5)
def connect_to_db() -> Connection:
    return db.connect()

---

Async functions

Works identically with async def — uses asyncio.sleep between attempts instead of time.sleep:

@retry(tries=5)
async def fetch_user(user_id: int) -> dict:
    async with httpx.AsyncClient() as client:
        response = await client.get(f"/users/{user_id}")
        response.raise_for_status()
        return response.json()

---

Decorator factory

Apply a shared retry policy across multiple functions:

from pytryagain import retry
from pytryagain.backoff import ConstantBackoff

http_retry = retry(tries=4, backoff=ConstantBackoff(delay=1.0))

@http_retry
def get_orders() -> list:
    ...

@http_retry
def get_inventory() -> list:
    ...

---

Limit retried exceptions

By default all Exception subclasses trigger a retry. Use exceptions to narrow this:

@retry(tries=4, exceptions=(TimeoutError, ConnectionError))
def connect(host: str) -> None:
    ...  # ValueError and others propagate immediately without retrying

---

Backoff strategies

Control the delay between attempts with any backoff strategy:

from pytryagain.backoff import (
    ConstantBackoff,
    LinearBackoff,
    ExponentialBackoff,
    ExponentialJitterBackoff,
)

# Wait 2 s between every attempt
@retry(tries=5, backoff=ConstantBackoff(delay=2.0))
def fetch() -> None: ...

# Wait 1 s, 2 s, 3 s, …
@retry(tries=5, backoff=LinearBackoff(base=1.0))
def fetch() -> None: ...

# Wait 2 s, 4 s, 8 s, … (doubles each time)
@retry(tries=5, backoff=ExponentialBackoff(base=2.0, initial=1.0))
def fetch() -> None: ...

# Exponential with random jitter — avoids thundering herd
@retry(tries=5, backoff=ExponentialJitterBackoff(base=2.0, initial=1.0))
def fetch() -> None: ...

---

Timeout

Stop retrying once a total wall-clock budget is exhausted, regardless of tries:

@retry(tries=10, timeout=30.0)
def call_api() -> dict:
    ...  # gives up after 30 seconds even if not all tries are used

timeout and tries work together — whichever limit is hit first wins.

---

Conditional retry

Use retry_if to inspect the exception and decide whether to retry:

# Retry only on HTTP 503 Service Unavailable
@retry(tries=5, retry_if=lambda exc: getattr(exc, "status_code", None) == 503)
def call_api() -> dict:
    ...

# Retry only on transient database errors
@retry(tries=3, retry_if=lambda exc: isinstance(exc, OperationalError) and exc.is_transient)
def query_db() -> list:
    ...

Returning False from retry_if immediately re-raises the exception (after calling on_giveup_callback if set).

---

Callbacks

Run a function after each failed attempt, or once when all retries are exhausted:

import logging

logger = logging.getLogger(__name__)

def log_attempt(exc: BaseException, attempt: int) -> None:
    logger.warning("attempt %d failed: %s", attempt, exc)

def alert_on_giveup(exc: BaseException, attempt: int) -> None:
    logger.error("gave up after %d attempts: %s", attempt, exc)

@retry(
    tries=4,
    on_exception_callback=log_attempt,
    on_giveup_callback=alert_on_giveup,
)
def send_payment(amount: float) -> None:
    ...

Both callbacks receive (exc, attempt) where attempt is 1-based.

---

Async callbacks

Async functions accept both sync and async callbacks:

async def notify_slack(exc: BaseException, attempt: int) -> None:
    await slack.post(f"Retry #{attempt} failed: {exc}")

async def page_oncall(exc: BaseException, attempt: int) -> None:
    await pagerduty.trigger(f"All retries exhausted: {exc}")

@retry(
    tries=5,
    on_exception_callback=notify_slack,
    on_giveup_callback=page_oncall,
)
async def process_job(job_id: str) -> None:
    ...

Note: async callbacks cannot be used with sync functions — a ValueError is raised at decoration time.

---

API Reference

retry

retry(
    func=...,
    tries=3,
    backoff=ExponentialJitterBackoff(),
    exceptions=(Exception,),
    on_exception_callback=...,
    on_giveup_callback=...,
    retry_if=...,
    timeout=...,
)

Parameter	Type	Default	Description
func	Callable	—	Function to wrap. Omit to use as a decorator factory.
tries	int	3	Total attempts including the first call. tries=1 means no retries.
backoff	BackOff	ExponentialJitterBackoff()	Delay strategy between attempts.
exceptions	tuple[type[BaseException], ...]	(Exception,)	Exception types that trigger a retry.
on_exception_callback	Callable[[BaseException, int], None]	—	Called after each failed attempt except the last.
on_giveup_callback	Callable[[BaseException, int], None]	—	Called once when all attempts are exhausted.
retry_if	Callable[[BaseException], bool]	—	Predicate to decide whether to retry. False re-raises immediately.
timeout	float	—	Total time budget in seconds across all attempts.

---

Backoff Strategies

Strategy	Description	Parameters
ConstantBackoff	Fixed delay every attempt	delay=1.0
LinearBackoff	Grows linearly: base * attempt	base=1.0
ExponentialBackoff	Doubles each attempt: initial * base ** attempt	base=2.0, initial=1.0
ExponentialJitterBackoff	Exponential with random jitter in [0, exp]	base=2.0, initial=1.0
FullJitterBackoff	Random in [0, min(cap, base ** attempt)]	cap=60.0, base=2.0
EqualJitterBackoff	Half fixed, half random	cap=60.0, base=2.0
DecorrelatedJitterBackoff	Each delay based on previous sleep	base=1.0, cap=60.0
FibonacciBackoff	Fibonacci sequence scaled by base	base=1.0
TruncatedExponentialBackoff	Exponential capped at a maximum	base=2.0, initial=1.0, cap=60.0

All strategies implement the BackOff protocol — you can supply your own:

def my_backoff(attempt: int) -> float:
    return min(attempt * 0.5, 10.0)

@retry(tries=5, backoff=my_backoff)
def fetch() -> None: ...

---

License

MIT — see LICENSE.