simple_scheduler

• Does not miss future events.
• Works even when period < execution time
• Tested on only linux platform
• Uses light weight multiprocessing to schedule jobs.
• This package uses a 24-hour clock, only.
• Simultaneously schedule any number of jobs.
• Recurring jobs to be precisely scheduled.
• Event jobs to be executed within the minute.
• Schedule the same function, again, with a different job_name
• On execution failure, set the number of reattempts.
• Control the duration between each re-attempt.
• Works only in background, hence easy to intergrate across platforms(eg. django, flask)
• Set start and stop timestamps if you want to ensure that your function executes only within this window

Install

From PyPi :

pip install simple-scheduler

Quick-start

There are 2 different schedulers:

1. Event scheduler (event occurs at an exact time)

from simple_scheduler.event import event_scheduler
from time import sleep, ctime, time

def print_args(a, b=1, *args, **kwargs):
    print(ctime(time()), a, b, args, kwargs)

event_scheduler.add_job(
    target=print_args,
    args=(0,),
    kwargs={"b":2},
    tz="Asia/Kolkata",
    job_name="print-args-1",
    number_of_reattempts=2,
    when=['fri|14:28', '*|14:**'],
    reattempt_duration_in_seconds=5
    )
event_scheduler.add_job(
    target=print_args,
    args=(0, 2, "arg1", "arg2"),
    kwargs={
        "key1":"value1",
        "key2":"value2"
        },
    when=['fri|14:28', '*|14:**'],
    tz="Asia/Kolkata",
    start="Jul 23 13:00:00 2021",
    stop="Jul 23 15:00:00 2021",
    job_name="print-args-2",
    number_of_reattempts=2,
    reattempt_duration_in_seconds=5
    )
event_scheduler.run()

2. Recurring scheduler (tasks occur after every "x"-seconds)

from simple_scheduler.recurring import recurring_scheduler
from time import sleep, ctime, time

def wait_10_secs(t): wait_X_secs(10)

def wait_X_secs(t):
    began_at = ctime(time()); sleep(t)
    print(f"I waited {t} seconds. [From: {began_at} to {ctime(time())}]")
    
recurring_scheduler.add_job(
    target=wait_10_secs,
    period_in_seconds=5,                              # period < execution time
    start="Jul 23 12:18:00 2021",
    stop="Jul 23 12:19:00 2021",
    job_name="ten",
    number_of_reattempts=0,
    reattempt_duration_in_seconds=0,
    kwargs={"t":10},
    tz="Asia/Kolkata"
    )
recurring_scheduler.add_job(
    target=wait_X_secs,
    kwargs={"t":3},                                   # period > execution time
    period_in_seconds=5,
    job_name="three"
    )
recurring_scheduler.run()

APIs

Toggle verbose

event_scheduler.verbose=False
recurring_scheduler.verbose=True

Job summary

event_scheduler.job_summary()
recurring_scheduler.job_summary()

Number of reattempts in case event fails [fallback]

event_scheduler.add_job(number_of_reattempts=3)
recurring_scheduler.add_job(number_of_reattempts=0)

Reattempt duration(in seconds) between each reattempt [fallback]

event_scheduler.add_job(reattempt_duration_in_seconds=10)
recurring_scheduler.add_job(reattempt_duration_in_seconds=10)

Start time (keep the scheduler running but postpone execution until this time)

event_scheduler.add_job(start="Dec 31 23:59:59 2021")
recurring_scheduler.add_job(start="Dec 31 23:59:59 2021")

Stop time (time when a job expires)

event_scheduler.add_job(stop="Dec 31 23:59:59 2021")
recurring_scheduler.add_job(stop="Dec 31 23:59:59 2021")

Remove a single job

event_scheduler.remove_job(job_name)
recurring_scheduler.remove_job(job_name)

Clear schedule (remove all jobs)

event_scheduler.clear()
recurring_scheduler.clear()

The parameter - when, in event scheduler

# its List[str], collection of "day|HH:MM"
# can take values like
    day|HH:MM,
    *|HH:MM,
    *|*H:MM,
    *|*H:MM,
    *|**:MM,
    *|**:*M,
    *|**:M*
# day:str = mon/tue/wed/thu/fri/sat/sun

Docker with gunicorn

In app.py ensure that scheduler is started globally and not within main()

event_scheduler.run() # global variable

if __name__ == "__main__":
   app.run(host="0.0.0.0", port="5000")

Also, in gunicorn config use the --preload argument. This will ensure that
only 1 instance of scheduler is running.