In Python, context manager is a great way to handle the closing of resources. It makes sure that resources are correctly closed after their use.

Basically, a context manager consists of two parts: enter and exit. You can create a class to function as a context manager, but it is kind of verbose (see here). In Python3, we can use the contextlib to create custom context managers easily using the decorator. A good example is to measure the running time of a code block. Here is the code:

from contextlib import contextmanager
import time

@contextmanager
def report_time(name):
    start = time.time()
    yield
    end = time.time()
    print(f"Time needed for {name}: {end-start}")

In the above code, we define a context manager report_time, when the block it manages starts, it will set the start time in start. Then it yields, which means that the block it manages starts running. When the block finishes, report_time resumes running and prints the running time of the measured code block.

Here is an example to use this code:

import requests

with report_time("httpbin request"):
    r = requests.get("https://httpbin.org/get")

Ref:

• https://jeffknupp.com/blog/2016/03/07/python-with-context-managers/
• https://www.youtube.com/watch?v=U2t2t_cpvoc