Old, see https://github.com/duarteocarmo/boilerplate

Python boilerplate project structure

Python is easy, right? Wrong! I can never remember how to structure a project in a way that works. Hopefully this saves me some time in the future...

The goal is to be able to, without modifying any files:

run all tests
debug a test
run & debug the cli module

Note that the packages under pboil aren't intended to be deployed independently. I don't know how package publishing/deployment works.

python -m venv .venv  # if you've just cloned the repo
. .venv/Scripts/activate
# pip install -r requirements.txt  # none here, but if you've done pip freeze > requirements.txt
./run_tests.sh
python run_cli.py

Debugging is via vscode. Run the Python: Current file configuration. You can also use vscode's test runner to run all tests, single tests, and debug tests. To debug a test, right click on the play button next to the test.

To run & debug the cli, run python run_cli.py in this dir. Note that you can't debug the cli.py directly with vscode, as the cli package and all other packages are not 'self aware': they need to be run from an external context so that pboil's init file sets up package paths to allow inter-package dependencies.

uozuAho/python_boilerplate

Python boilerplate project structure