pyOpenSci: Create a Python Package Workshop Repository
This repository supports pyOpenSci workshops focused on Python packaging.
Getting started
Before the workshop, be sure that you read the setup instructions that you received in an email.
You will need a GitHub and a testPyPI account to follow along.
GitHub Codespaces
We have GitHub Codespaces configured for this workshop. To use this repository and the dev container associated with it in a workshop, instructor your learners to create a new repository
as a template using this repo.

Open a codespace prior to attending
the workshop to ensure the environment loads properly. The initial codespace build can take up to 15 minutes.
The link above provides you with instructions
on how to do this and how to setup the accounts you need for the workshop.
Creating your package
To use the copier template, follow the tutorial here that walks you through the steps.
The command to run the copier template is:
copier copy gh:pyopensci/pyos-package-template .
Cheat Sheet
Below is a command cheat sheet that will help you keep track of all of hte various commands used in our tutorials.
Ship It: Command Cheat Sheet
Codespace setup (one-time)
# Shorten your terminal prompt
export PS1="$ "
# Disable keyring. You need to do this in codesppaces only when publishing to PyPI using Hatch.
export PYTHON_KEYRING_BACKEND=keyring.backends.null.Keyring
Create your package
# Call the pyOpenSci package template
copier copy gh:pyOpenSci/pyos-package-template .
uv commands
# Launch Python in your project environment
uv run python
# Add package dependency
uv add numpy
# Remove package dependency
uv remove numpy
# Add a development dependency
uv add --dev pytest
Hatch environments
# View all Hatch environments and scripts
hatch env show
# build your package; check your package distribution files using twine
hatch run build:check
# Run test suite with coverage (across Python versions)
hatch run test:run
# Build sphinx documentation
hatch run docs:build
# Serve sphinx docs locally with live reload
hatch run docs:serve
# Check docstrings & code style
hatch run style:check
# Format code with ruff
hatch run style:format
# Check dependencies for vulnerabilities
hatch run audit:check
Build & publish
# build sdist + wheel into dist/ using native hatch in your current environment
hatch build
# Publish with hatch to test pypi (r = repository)
hatch publish -r test
Install a package
# Install from Test PyPI
uv pip install --index-url https://test.pypi.org/simple/ your-package-name
pip install -i https://test.pypi.org/simple/ your-package-name
# Install from GitHub
pip install git+https://github.com/username/repo-name
Run tests directly with pytest (vs hatch)
# Run all tests in your active Python envt with pytest
pytest # Run tests using pytest with coverage
pytest --cov=pyospackage --cov-report=term-missing
Contributors ✨
Thanks goes to these wonderful people (emoji key):
This project follows the all-contributors specification. Contributions of any kind welcome!