Qaviton IO

Qaviton IO
is a package with a simple API, making use of python's async & multiprocessing
to enable fast execution of many asyncable operations.

Installation

pip install qaviton-io -U

Requirements

• Python 3.6+

Features

• async task manager
• process task manager
• task logger

Usage

async manager:

from time import time
from typing import List
from requests import get, Response  # lets make use of requests to make async http calls
from qaviton_io import AsyncManager, Log

# we can save the responses here
rs: List[Response] = []

# let's create an async manager & log
manager = AsyncManager()
log = Log()

# first we make a simple function to make an http call
# we want to log the result
# and make sure that in case of an exception
# the manager won't stop
@log.task(exceptions=Exception)
def task():
    r = get("https://qaviton.com")
    r.raise_for_status()
    rs.append(r)

# let's run our task once and see how long it takes
t = time()
manager.run([task for _ in range(1)])
t = time() - t
print(f'took {round(t, 3)}s')

# now let's run our task 20 times and see how long it takes
t = time()
manager.run([task for _ in range(20)])
t = time() - t
print(f'took {round(t, 3)}s')

# let's view the results in the log report
log.report()

process manager:

from time import time
from requests import get
from qaviton_io.types import Tasks
from qaviton_io import ProcessManager, Log
from traceback import format_exc

# make sure your tasks are defined at the module level,
# so they can be pickled by multiprocessing

# first we need to make a log registry
log = Log()


# now we make some tasks
# this is a nested task
# we don't want to handle any exceptions
# so in case of failure the parent will not proceed
@log.task()
def task(url):
    r = get(url)
    r.raise_for_status()


# this is the prent task
# we want to handle all exceptions
# so in case of failure the next task will execute
@log.task(exceptions=Exception)
def multi_task():
    for url in [
        "https://qaviton.com",
        "https://qaviton.co.il",  # make sure you enter a valid address
        "https://qaviton.com1",  # make sure you enter a valid address
    ]:
        task(url)


# let's create a function to execute tasks
def execute_tasks(tasks: Tasks, timeout):
    manager = ProcessManager()
    t = time()
    try:
        manager.run_until_complete(tasks, timeout=timeout)
        timed_out = None
    except TimeoutError:
        timed_out = format_exc()
    t = time() - t
    manager.log.report()
    print(f'took {round(t, 3)}s\n')
    manager.log.clear()
    return timed_out


# now all that's left is to run the tasks
if __name__ == "__main__":
    timeouts = [
        execute_tasks([multi_task for _ in range(1)], timeout=3),
        execute_tasks([multi_task for _ in range(20)], timeout=6),
        execute_tasks([multi_task for _ in range(80)], timeout=9),
    ]
    for timeout in timeouts:
        if timeout:
            print(timeout)

notes:

• for good performance and easy usage
  you should probably stick with using the AsyncManager

• The ProcessManager uses async operations as well as multi-processing.
  It distributes tasks across cpus, and those tasks are executed using the AsyncManager
  if you want maximum efficiency you should consider using the ProcessManager

• The ProcessManager uses the multiprocessing module
  and should be treated with it's restrictions & limitations accordingly

• The ProcessManager gets stuck easily,
  make sure to use timeouts when using it