Skip to content

Latest commit

 

History

History
262 lines (204 loc) · 9.43 KB

README.md

File metadata and controls

262 lines (204 loc) · 9.43 KB
Raw

cookiecutter-caped-package

Cookiecutter template for authoring (caped-based) pypi packages.

NOTE: This repo is not meant to be cloned/forked directly! Please read "Getting Started" below

Getting Started

Create your package package

Install Cookiecutter and generate a new caped package project:

pip install cookiecutter
cookiecutter https://github.com/Kapoorlabs-CAPED/cookiecutter-template

Cookiecutter prompts you for information regarding your package (A new folder will be created in your current working directory):

full_name [Caped Developer]: Varun Kapoor
email [yourname@example.com]: randomaccessiblekapoor@gmail.com
github_username_or_organization [githubuser]: Kapoorlabs-CAPED
# NOTE: for packages whose primary purpose is to be a caped package, we
# recommend using the 'caped-' prefix in the package name.
# If your package provides functionality outside of caped, you may
# choose to leave caped out of the name.
package_name [caped-based-repobar]: caped-ai
Select github_repository_url:
1 - https://github.com/neuronz52/caped-ai
2 - provide later
Choose from 1, 2 [1]:
module_name [ai]: caped_growth_ai
display_name [caped based-repoBar]: AI
short_description [A simple package to use with caped library]:
# you can select from various package template examples
include_sample_data_package [y]:
use_git_tags_for_versioning [n]:
Select license:
1 - BSD-3
2 - MIT
3 - Mozilla Public License 2.0
4 - Apache Software License 2.0
5 - GNU LGPL v3.0
6 - GNU GPL v3.0
Choose from 1, 2, 3, 4, 5, 6 (1, 2, 3, 4, 5, 6) [1]:

You just created a minimal caped package, complete with tests and ready for automatic deployment!

For more detailed information on each prompt see the prompts reference.

caped-ai/
│
├── .github
│   └── workflows
│      └── test_and_deploy.yml
├── LICENSE
├── MANIFEST.in
├── caped_ai
│   ├── __init__.py
│   ├── caped.yaml
│   └── _tests
│       ├── __init__.py
│       
├── pyproject.toml
├── README.md
├── setup.cfg
└── tox.ini

Initialize a git repository in your package

NOTE: This is important not only for version management, but also if you want to pip install your package locally for testing with pip install -e .. (because the version of your package is managed using git tags, see below)

cd caped-ai
git init
git add .
git commit -m 'initial commit'

Upload it to github

  1. Create a new github repository

  2. Add your newly created github repo as a remote and push:

    # here, continuing with the example above...
# but replace with your own username and repo name

git remote add origin https://github.com/neuronz52/caped-ai.git
git push -u origin main

Monitor testing and coverage

The repository should already be setup to run your tests each time you push an update (configuration is in .github/workflows/test_and_deploy.yml). You can monitor them in the "Actions" tab of your github repository. If you're following along, go have a look... they should be running right now!

When the tests are done, test coverage will be viewable at codecov.io (assuming your repository is public): https://codecov.io/gh/<your-github-username>/<your-package-name>

Set up automatic deployments

Your new package is also nearly ready to automatically deploy to PyPI (whenever you create a tagged release), so that your users can simply pip install your package. You just need to create an API token to authenticate with PyPi, and then add it to your github repository:

  1. If you don't already have one, create an account at PyPI
  2. Verify your email address with PyPI, (if you haven't already)
  3. Generate an API token at PyPi: In your account settings go to the API tokens section and select "Add API token". Make sure to copy it somewhere safe!
  4. Create a new encrypted secret" in your github repository with the name "TWINE_API_KEY", and paste in your API token.

You are now setup for automatic deployment!

Automatic deployment and version management

Each time you want to deploy a new version, you just need to create a tagged commit, and push it to your main branch on github. Your package is set up to use setuptools_scm for version management, meaning you don't need to hard-code your version anywhere in your package. It will be inferred from the tag each time you release.

# the tag will be used as the version string for your package
# make it meaningful: https://semver.org/
git tag -a v0.1.0 -m "v0.1.0"

# make sure to use follow-tags so that the tag also gets pushed to github
git push --follow-tags

Note: as of git 2.4.1, you can set follow-tags as default with git config --global push.followTags true

Monitor the "actions" tab on your github repo for progress... and when the "deploy" step is finished, your new version should be visible on pypi:

https://pypi.org/project/<your-package-name>/

and available for pip install with:

# for example
pip install caped-ai

Running tests locally

Tests are automatically setup to run on github when you push to your repository.

You can run your tests locally with pytest. You'll need to make sure that your package is installed in your environment, along with testing requirements (specified in the setup.cfg extras_require section):

pip install -e ".[testing]"
pytest

Create your documentation

Documentation generation is not included in this template. We recommend following the getting started guides for one of the following documentation generation tools:

  1. Sphinx
  2. MkDocs
  3. JupyterBook

Pre-commit

This template includes a default yaml configuration for pre-commit. Among other things, it includes checks for best practices in caped packages. You may edit the config at .pre-commit-config.yaml

To use it run:

pip install pre-commit
pre-commit install

You can also have these checks run automatically for you when you push to github by installing pre-commit ci on your repository.

Features

Resources

Details on why this package template is using the src layout can be found here and here

Issues

If you encounter any problems with this cookiecutter template, please file an issue along with a detailed description.

License

Distributed under the terms of the BSD-3 license, cookiecutter-caped-package is free and open source software.