TIL: Python has an assertLogs function for unit testing log statements

Today I learned how to unit test log statements with Python. I updated @bcdevexbot to check the status of an opportunity. The bot logs an error message if it encounters an unknown status. To test this scenario I used the assertLogs function.

The assertsLogs(logger=None, level=None) tests that at least one statement is logged at the given level. The test fails if no statements are logged at that level. If statements were logged then the function returns an object containing them.

Below is a very simple example of using the assertsLogs function. The logging_example.py file contains the code to test. The check_status(status) function is a contrived example that performs some logging.


import logging

logger = logging.getLogger('demo_logger')

def check_status(status):
    if status == "Open":
        logger.debug("Good status")
        logger.error("Unknown status: {0}".format(status))

if __name__ == '__main__':

The test_logging_example.py file contains the unit test. The assertsLogs(logger=None, level=None) takes two optional parameters.

The first parameter is the name of the logger. It defaults to the root logger if one is not provided. In the example below, the logger used in the logging_example.py file is used.

The second parameter is the log level. It defaults to INFO if a level is not provided. A different level can be specified as in the example below.


import unittest
import logging_example

class MyTestCase(unittest.TestCase):
    def test_logging(self):
        with self.assertLogs('demo_logger', level='DEBUG') as lc:
            self.assertEqual(['DEBUG:demo_logger:Good status'], lc.output)
            self.assertEqual(['DEBUG:demo_logger:Good status','ERROR:demo_logger:Unknown status: Bad'], lc.output)

if __name__ == '__main__':

To learn more see the Python Documentation.