retry

library0.9.2pypypi✓ verified 27d ago

The `retry` library (version 0.9.2) provides a simple decorator for adding retry logic to Python functions. It allows configuration of exceptions to catch, number of attempts, delay between retries, backoff strategy, and optional jitter. This library is largely unmaintained, with its last release in 2016, and is superseded by more actively developed alternatives like `tenacity`.

http-networking

pip install retry

github ↗package ↗

INSTALL

IMPORT

SIG · RETRY

retry

http-networkingpythonv0.9.2

Install

1.7s avg

Import

53ms

Disk

17MB

Pass rate

10/ 10

Env Coverage10 / 10

glibc

3.9–3.13

musl

3.9–3.13

Install & Compatibility

Where this runs

tested against v0.9.2 · pip install

no network on importno background threads

Install × environment matrix

Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux

musl

py 3.10–3.95 runs

installs and imports cleanly · install 0.0s · import 0.054s · 18.8MB

glibc

py 3.10–3.95 runs

installs and imports cleanly · install 1.7s · import 0.052s · 19MB

17MB installed

● package 17MB

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

retry

✓ from retry import retry

✗ from retrying import retry

The 'retrying' library is a different, also unmaintained, project that `tenacity` forked from. This `retry` library has a distinct import path.

This quickstart demonstrates how to apply the `@retry` decorator to a function, configuring it to retry on specific exceptions, with a maximum number of attempts, an initial delay, exponential backoff, and a maximum delay between retries. It will print logging messages for each retry attempt.

import random from retry import retry import logging logging.basicConfig(level=logging.INFO) @retry(exceptions=(IOError, ValueError), tries=3, delay=2, backoff=2, max_delay=10) def unreliable_function(): if random.randint(0, 5) > 2: raise IOError("Simulated network error") elif random.randint(0, 5) < 1: raise ValueError("Simulated data error") else: print("Function succeeded!") return "Success" try: result = unreliable_function() print(f"Final result: {result}") except (IOError, ValueError) as e: print(f"Function failed after all retries: {e}")

Debug

Known issues

breakingThis 'retry' library (invl/retry) is largely unmaintained, with its last release in 2016. Its API is not compatible with 'tenacity', which is the actively maintained successor to the 'retrying' library (another distinct, unmaintained project).

fix

Consider migrating to `tenacity` (pip install tenacity) for active development, better features, and bug fixes. Tenacity offers a more robust and flexible API for retry strategies.

affects: <=0.9.2

gotchaBy default, the `@retry` decorator will retry indefinitely if the `tries` parameter is not explicitly set or set to -1.

fix

Always set a finite `tries` value (e.g., `tries=5`) or a `max_delay` to prevent infinite loops in production environments.

affects: <=0.9.2

gotchaThe `retry` decorator does not inherently support asynchronous functions (async/await syntax).

fix

For asynchronous code, you will need to use a different retry library that explicitly supports `async/await`, such as `tenacity` with `AsyncRetrying` or `stamina`.

affects: <=0.9.2

gotchaWhen using `jitter`, if a tuple `(min, max)` is provided, the extra delay is chosen randomly. If a single number is provided, the jitter is fixed.

fix

Be aware of the type of `jitter` parameter passed. For true random jitter to prevent 'thundering herd' problems, provide a tuple (e.g., `jitter=(0, 1)`).

affects: <=0.9.2

Errors

Common errors & fixes

ModuleNotFoundError: No module named 'retry'

The 'retry' library has not been installed in your Python environment.

fix

Run 'pip install retry' in your terminal.

TypeError: retry() missing 1 required positional argument: 'f'

The `@retry` decorator was called with arguments but without the necessary parentheses for the decorator factory, causing it to be interpreted as if it's decorating a function directly.

fix

Always use parentheses when providing arguments to the decorator, even if empty: `@retry(tries=3)` instead of `@retry(tries=3)`.

AttributeError: module 'retry' has no attribute 'retry'

After `import retry`, the user attempted to access `retry.retry`. The `retry` module itself is the decorator function, so you should use `retry` directly.

fix

If you `import retry`, use `@retry(...)` or `retry(func, ...)`. If you `from retry import retry`, then use `@retry(...)` or `retry(func, ...)`.

TypeError: retry() got an unexpected keyword argument 'wait'

The 'retry' library does not recognize the keyword argument 'wait' (commonly used in other retry libraries like 'tenacity'). It expects 'delay' for the wait time.

fix

Replace unsupported arguments like 'wait' with the correct argument 'delay' (e.g., `@retry(delay=2)` instead of `@retry(wait=2)`).

Upgrade

Version history

0.9.2latest on PyPI · released May 11, 2016

Audit

Dependencies

decoratoroptionalUsed to optionally preserve function signatures when decorating, not strictly required for core functionality.

Agent activity

16 hits · last 30 days

node

Resources

githubhttps://github.com/invl/retry ↗

packagehttps://pypi.org/project/retry/ ↗