Learn more about our quantum error correction milestone

cirq.testing.assert_logs

A context manager for testing logging and warning events.

@contextlib.contextmanager
cirq.testing.assert_logs(
    *matches,
    count: Optional[int] = 1,
    min_level: int = logging.WARNING,
    max_level: int = logging.CRITICAL,
    capture_warnings: bool = True
) -> Iterator[List[logging.LogRecord]]

To use this one wraps the code that is to be tested for log events within the context of this method's return value:

with assert_logs(count=2, 'first_match', 'second_match') as logs:
    <code that produces python logs>

This captures the logging that occurs in the context of the with statement, asserts that the number of logs is equal to count, and then asserts that all of the strings in matches appear in the messages of the logs. Further, the captured logs are accessible as logs and further testing can be done beyond these simple asserts.

*matches Each of these is checked to see if they match, as a substring, any of the captures log messages.
count The expected number of messages in logs. Defaults to 1. If None is passed in counts are not checked.
min_level The minimum level at which to capture the logs. See the python logging module for valid levels. By default this captures at the logging.WARNING level and above, so this does not capture logging.INFO or logging.DEBUG logs by default.
max_level The maxium level at which to capture the logs. See the python logging module for valid levels. By default this captures to the logging.CRITICAL level thus, all the errors and critical messages will be captured as well.
capture_warnings Whether warnings from the python's warnings module are redirected to the logging system and captured.

A ContextManager that can be entered into which captures the logs for code executed within the entered context. This ContextManager checks that the asserts for the logs are true on exit.

ValueError If min_level is greater than max_level.