Mock command line utils by replaying recorded invocations
This package provides two command line utilities: crecord and creplay. The former records the output (stdout and stderr) and the return code of a command and the latter replays the command invocation by reproducing its output and return code:
$ crecord echo foo foo $ creplay echo foo foo $ crecord ls foo ls: foo: No such file or directory $ creplay ls foo ls: foo: No such file or directory $ echo $? 1
This could be used to mock slow and environment-dependent command invocations for testing purposes.
There’s also a pytest plugin contained in cli_mock package. It exposes popen_controller fixture that can be used to replay crecord logs in response to subprocess.Popen invocations (and the APIs that call it under the hood):
def test_foo(popen_controller): popen_controller.set_replay_log(my_log) output = subprocess.check_output(['foo']) assert output == b'bar\n'
After the replay log is activated calls to subprocess.Popen and friends with the commands that are contained in the log will replay from the log. Commands that are not in the log will trigger an AssertionError.
It’s possible to allow executing commands that are not in the log via activating non-strict mode:
In non-strict mode any command that is not in the log will be passed to subprocess.Popen without modifications and will be executed in a usual fashion.