Library helps to easily write concurrent executed code blocks.
Quick example:
import asyncio
from concurrently import concurrently
async def amain(loop):
"""
How to fetch some web pages with concurrently.
"""
urls = [ # define pages urls
'https://test/page_1',
'https://test/page_2',
'https://test/page_3',
'https://test/page_4',
]
results = {}
# immediately run wrapped function concurrent
# in 2 thread (asyncio coroutines)
@concurrently(2)
async def fetch_urls():
for url in urls:
# some function for download page
page = await fetch_page(url)
results[url] = page
# wait until all concurrent threads finished
await fetch_urls()
print(results)
if __name__ == '__main__':
loop = asyncio.get_event_loop()
loop.run_until_complete(amain(loop))- Decorator @concurrently makes to main thinks:
- starts concurrent execution specified count of decorated function
- returns special waiter object to control the running functions
By default, the code runs as asyncio coroutines, but there are other supported ways to execute, by specifying the argument engine.
See https://concurrently.readthedocs.io/
Now support only Python 3.5 and above.
From PyPi:
$ pip install concurrentlyFrom local:
# update setuptools
$ pip install 'setuptools >= 30.4'
# do install
$ make install
# or
$ pip install .Prepare and activate virtual environment like:
$ python3 -m venv .env
# for bash
$ source .env/bin/activate
# for fish
$ . .env/bin/activate.fishUpdate pre-install dependencies:
$ pip install 'setuptools >= 30.4'Install:
$ make install_dev
# or
$ pip install --editable .[develop]Run tests:
$ make test
# or
$ pytest tests/