Skip to content

A fire-tested template for production grade python libraries and packages.

License

Notifications You must be signed in to change notification settings

opcode81/pymetrius

 
 

Repository files navigation

Pymetrius - cookiecutter for python libraries

This repository contains a cookiecutter template that can be used for library development. The template contains several well-known "best-practices" for libraries (poetry, poethepoet, mypy, ruff, nbqa) and also some tools inspired by projects of ours that we consider generally useful - build and release scripts, auto-generation of documentation files, and others. Earlier versions of this template were used in several industry projects as well as for open source libraries.

Build, install and tests of the library are run by default poethepoet tasks, the documentation is built with Jupyter-Book. The template includes CI/CD pipelines based on github actions. The documentation will be published to GitHub pages using an action.

In the documentation links to source code will be created, therefore you will be prompted to give the project's url.

See the resulting repository's contributing guidelines for further details. Some examples for projects following the general style of the template are tianshou and armscan_env

Usage

Prerequisites

The template supports python 3.11 and higher. For a smooth project generation you need to have

  1. Cookiecutter. Install it e.g. with pip install cookiecutter
  2. Poetry (for using the new repository)

Creating a new project

Call

cookiecutter https://github.com/appliedAI-Initiative/pymetrius -o path/to/directory

and walk through the questions. You can also clone this repository, adjust the template and call cookiecutter on the local file.

You will get a repo in <path/to/directory>/<project_name>. For finalizing the setup, you can cd into it, and call e.g.,

git init
poetry shell
poetry install --with dev
poe format
git add . && git commit -m "Initial commit from pymetrius"

Note: if you want to use sphinx-spelling, as is configured by default in the poe tasks, you may need to install the enchant library, see installation instructions. Otherwise, you can just remove the spellcheck from the tasks.

Push to your branch and enjoy the fully setup pipelines and documentation!

Contributing

The core maintainers are Michael Panchenko and Carlo Cagnetta at appliedAI. Initially, the project was also supported by Adrian Rumpold. We are happy about any contribution to this project, feel free to contact us directly or to open an issue or a pull request.

About

A fire-tested template for production grade python libraries and packages.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 54.4%
  • Jupyter Notebook 25.2%
  • Shell 12.3%
  • Dockerfile 5.8%
  • TeX 2.3%