Reference

waiter

class waiter.waiter(delays, timeout=inf)

An iterable which sleeps for given delays.

Parameters

• delays – any iterable of seconds, or a scalar which is repeated endlessly

• timeout – optional timeout for iteration

Stats

Stats class attribute

stats

Stats instance attribute

__add__(step) → waiter.waiter

Generate incremental backoff.

__aiter__()

Asynchronously generate a slow loop of elapsed time.

__ge__(floor) → waiter.waiter

Limit minimum delay generated.

__getitem__(slc: slice) → waiter.waiter

Slice delays, e.g., to limit attempt count.

__init__(delays, timeout=inf)

Initialize self. See help(type(self)) for accurate signature.

__iter__()

Generate a slow loop of elapsed time.

__le__(ceiling) → waiter.waiter

Limit maximum delay generated.

__mul__(factor) → waiter.waiter

Generate exponential backoff.

classmethod accumulate(*args, **kwargs) → waiter.waiter

Create waiter based on itertools.accumulate (requires Python 3).

classmethod count(*args, **kwargs) → waiter.waiter

Create waiter based on itertools.count.

classmethod exponential(base, **kwargs) → waiter.waiter

Create waiter with exponential backoff.

classmethod fibonacci(delay, **kwargs) → waiter.waiter

Create waiter with fibonacci backoff.

filtered(predicate: Callable, func: Callable, iterable: Iterable) → Iterator[tuple]

Provisionally generate arg, func(arg) pairs while predicate evaluates to true.

map(func: Callable, *iterables: Iterable) → waiter.waiter

Return new waiter with function mapped across delays.

poll(predicate, func, *args, **kwargs)

Repeat function call until predicate evaluates to true.

polling(predicate: Callable)

Return a decorator for poll.

classmethod polynomial(exp, **kwargs) → waiter.waiter

Create waiter with polynomial backoff.

random(start, stop) → waiter.waiter

Add random jitter within given range.

repeat(func, *args, **kwargs)

Repeat function call.

repeating(func: Callable)

A decorator for repeat.

retry(exception, func, *args, **kwargs)

Repeat function call until exception isn’t raised.

retrying(exception: Exception)

Return a decorator for retry.

stream(queue: Sequence, size: int = None) → Iterator

Generate chained values in groups from an iterable.

The queue can be extended while in use.

suppressed(exception, func: Callable, iterable: Iterable) → Iterator[tuple]

Provisionally generate arg, func(arg) pairs while exception isn’t raised.

throttle(iterable)

Delay iteration.

suppress

waiter.suppress(*exceptions)

Variant of contextlib.suppress, which also records exception.

first

waiter.first(predicate: Callable, iterable: Iterable, *default)

Return first item which evaluates to true, like any with filtering.

Stats

class waiter.Stats(**kwds)

Bases: collections.Counter

Mapping of attempt counts.

Note

The Stats interface is considered provisional. It can be customized by overriding at the class or instance level.

add(attempt, elapsed)

Record attempt and return next value.

property failures

number of repeat attempts

property total

total number of attempts