doitlater 0.1.2

When you have to do something... later.

DoItLater

A simple python library to schedule work in the future with ability to loop infinitely. Does not depend on any 3rd party libraries.

• Installation
• Usage
• API

---

Installation

The package can be simply installed using on of these methods:

• pip: pip install doitlater
• poetry: poetry add doitlater
• poetry: poetry add git+https://github.com/evalkaz/doitlater.git

---

Usage

Simplest example:

from datetime import datetime, timedelta
from doitlater import Later

later = Later()

@later.on(datetime(2021, 1, 1))
def say_hello():
    print("Happy new years in 2021!")

if __name__ == "__main__":
    later.do()

To call the same function every 10 seconds with 30 seconds cold start:

@later.on(
    datetime.now() + timedelta(seconds=30),
    timedelta(seconds=10),
    loop=True
)
def repeatable_work():
    print("Hello every 10 seconds!")

The later.on() is stackable so this will work too (will be executed every 5 and every 7 days):

@later.on(datetime(2021, 1, 1), timedelta(days=5))
@later.on(datetime(2021, 1, 1), timedelta(weeks=1))
def say_hello():
    print("Happy new years in 2021!")

You can also pass a list of datetime or timedelta (or a mixed one) when to execute the function:

@later.on(datetime(2021, 1, 1), [
    datetime(2021, 2, 1),
    datetime(2021, 3, 1),
    timedelta(days=31)], loop=False)
def say_hello():
    print("Happy new years in 2021!")

If you want to stop executing the same function just return False:

@later.on(datetime(2021, 1, 1), timedelta(seconds=5), loop=True)
def only_one_hello():
    print("This will appear only once!")
    return False

If you have multiple functions and one of them throws an exception but you don't want to stop the work, pass ignore_errors=True to Later() object:

later = Later(ignore_errors=True)

By default, library will output logs on errors. To change logging, use dictConfig:

from datetime import datetime, timedelta
from doitlater import Later
from logging.config import dictConfig

dictConfig(
    {
        "version": 1,
        "formatters": {
            "default": {
                "format": "[%(asctime)s] %(levelname)s in %(module)s: %(message)s",
            }
        },
        "handlers": {
            "default": {"class": "logging.StreamHandler", "formatter": "default",}
        },
        "root": {"level": "INFO", "handlers": ["default"]},
    }
)

later = Later()

@later.on(datetime(2021, 1, 1))
def say_hello():
    print("Happy new years in 2021!")

if __name__ == "__main__":
    later.do()

---

API

Later(workers, ignore_errors) takes these parameters:

• workers - number of threads to use, defaults to maximum number of threads supported by CPU.
• ignore_errors - False will exit when one of the function throws an error, True - ignores exceptions and will resume the work.

later.on(exactly, repeat, loop) function takes these parameters:

• exactly - on which time perform the first function call.
• repeat - a single value or a list of datetime, timedelta or both on when to repeat the call. If None is passed the function will not be called again. Defaults to None.
• loop - whether repeat calling times from repeat. Defaults to True.

later.do() does not take any parameters.