Use perftester as CLI

This is a similar approach that pytest takes. You simply run perftester in a directory of your choice and wait for the results. In the meantime, perftester will collect all the testing modules (which start with "perftester_" and have ".py" extension), then it will collect all perftester functions in each module (they start with "perftester_"), and then it will run all the functions. If you add any updates to pt.config, it will do the update. And that's it, you have the results printed to the console.

This approach uses the default state of pt.config object. You can change it before running the tests using the config_perftester.py Python script, which should be located in the directory from which you run the tests. That way, you can use multiple config_perftester.py, simply by putting them in different directories and running perftester from these directories.

The config_perftester.py file is not that necessary, as you can change any settings directly in testing files. You can find examples here and here.

Use perftester command for a directory

You can use perftester for a directory, by providing a path to it:

$ perftester ./tests/memory_tests/

perftester will look for config_perftester.py file in the directory in which you run the command, not in this provided directory. So, if you have a config file there, simply move to it and then run perftester.

perftester will recursively locate all perftester_ testing modules, and will run all the perftester tests from them.

Use perftester command for a single file

You can use perftester for a single file, by providing a path to this file:

$ perftester ./tests/perftester_time_tests.py

Remember that when you do so, perftester will look for config_perftester.py file.

Use perftester command without a path

If you ommit a path to the command, that is,

$ perftester

this will be the same as running the command with a path to the current directory, so

$ perftester .

Log results

When you use perftester as a CLI program, you can save the results to a file.

You cannot do that when you run perftester in a different way, like in an interactive session or with pytest. Of course, you can create then your own logger.

The default setting is to not log. In order to change that, you need to change pt.config's attribute log_to_file to True, and provide a log file as log_file, e.g.

pt.config.log_to_file = True
pt.config.log_file = pathlib.Path(".") / "perftester.log"

You must do that in the config_perftester.py file, not elsewhere.