dorfsmay / python-logging-impact

Demonstration of how python logging can unexpectdly impact your code performance

python-logging-impact

Demonstration of possible impact performances when passing functions to logging.

Python logging statements executed regardless of logging level

Logging statements (logging.info, logging.debug, etc...) are executed, that's how they determine the log level and decide to proceed further or not.

Function composition eagerness

Python is eager when doing function composition, logme() is executed before being passed to logging.debug.

https://stackoverflow.com/questions/53782625/why-are-logging-statement-in-python-evaluated-regardless-of-level/53783577

Solutions

1. Use the Python % string formating + variables: logging.debug = debug(msg, *args, **kwargs)
2. Make the argument lazylly evaluated? Still exploring...

Original Problem statement

If you run the script in this repo without options, "I was executed" is still printed. Compare: logging_impacts_me.py vs logging_impacts_me.py -d

This can have a huge performance impact on your code. I discovered this issue while implementing an object based double linked list to solve Advent of Code 2018 day 9 problem. I originally had an __str__ method which allowed to print a helpful representation of each object in the list while debugging, calling the str() functions which made my solution too slow to solve part 2 of the problem in a reasonable amount of time.