Fix issue 2540: Synchronise concurrent command calls to single-client to single-client mode

[Vivanov98]

Pull Request check-list

• Does $ tox pass with this change (including linting)?
• Do the CI tests pass with this change (enable it first in your forked repo and wait for the github action build to finish)?
• Is the new or changed code fully tested?
• Is a documentation update included (if this change modifies existing APIs, or introduces new ones)?
• Is there an example added to the examples folder (if applicable)?
• Was the change added to CHANGES file?

Description of change

Fixing issue #2540 (example script in issue tested and bug no longer occurs)

The underlying cause is that the connection in single-client mode isn't protected by a lock, resulting in multiple coroutines reading/writing to a connection stream when multiple commands are called in parallel (forexample with asyncio.gather)

Additionally, redundant connections can be made if the first time the client is used, it is used in parallel (forexample with asyncio.gather).

A simple lock mechanism has been added to the execute_command function to ensure that single-client mode is always executing commands in sequence. This lock is also used each time the initialize method is called, ensuring that multiple connections aren't spun up if the first time the client is called happens to be with a parallel mechanism.

Worth mentioning that the use cases for using parallel calling mechanisms in single-client mode are probably limited, though this still seems useful to have in terms of consistency.

[codecov-commenter]

Codecov Report

Base: 92.26% // Head: 92.27% // Increases project coverage by +0.00% 🎉

Coverage data is based on head (6660ed5) compared to base (e39c7ba).
Patch coverage: 97.29% of modified lines in pull request are covered.

📣 This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #2568   +/-   ##
=======================================
  Coverage   92.26%   92.27%           
=======================================
  Files         115      115           
  Lines       29610    29645   +35     
=======================================
+ Hits        27320    27354   +34     
- Misses       2290     2291    +1     

Impacted Files	Coverage Δ
tests/test_asyncio/test_connection.py	97.39% <96.42%> (-0.31%)	⬇️
redis/asyncio/client.py	93.27% <100.00%> (+0.76%)	⬆️
tests/test_asyncio/test_cluster.py	97.48% <0.00%> (-0.13%)	⬇️
tests/test_cluster.py	96.84% <0.00%> (-0.11%)	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[Vivanov98]

The only deterministic test to validate the locking does what we want that I could think of required patching the connection and call_with_retry objects. This test is consistent, and would fail without the client changes.