Abilian Development Tools (ADT) is a curated collection of Python development tools that includes formatters, testing frameworks, style checkers, type checkers, and supply chain audit tools. By adding abilian-devtools = '*' to your project's requirements.in or pyproject.toml, you can access over 40+ curated projects and plugins.
Proper configuration and usage in your project is still required.
Additionally, the package provides a command-line interface (CLI) called adt to help users get started with various development tasks such as running tests, security audits, and code formatting.
ADT was developped at Abilian as a tool to help manage dozens of Python projects (open source or not). We hope it can be useful to others too.
This is a curated, and opiniated, collection of best-of-breed Python development tools:
- Formatters (
ruff format,docformatter) - Testing frameworks (
pytestand friends,nox) - Style checkers (
ruff) - Type checkers (
ty) - Supply chain audit (
reuse,vulture,deptry) - And more.
Obviously, all the credit goes to the creators and maintainers of these wonderful projects. You have all our gratitude!
Instead of having to track all the 40+ projects and plugins we have curated, you just need to add abilian-devtools = '*' in your project's requirements.in or pyproject.toml.
You still need to properly configure and call them in your own projects.
For example configuration, see, for instance, https://github.com/abilian/nua (Makefile, pyproject.toml, setup.cfg, tasks.py, noxfile.py...).
As a bonus, we're providing a CLI called adt which can help you get started:
$ adt
adt (0.9.x)
Usage:
adt <command> [options] [arguments]
Options:
-V Show version and exit
-d Enable debug mode
-v Increase verbosity
Available commands:
all Run everything (linters and tests).
audit Run security audit.
bump-version Bump version in pyproject.toml, commit & apply tag.
check Run checker/linters on specified files or directories.
clean Cleanup cruft.
cruft Run cruft audit.
format Format code in specified files or directories.
help-make Helper to generate the `make help` message.
seed Seed project with configuration files from profiles.
test Run tests.
typecheck Run type checker on specified files or directories.
-
adt all— Run linters, type checker, and tests in sequence. Handy as a pre-commit or CI sanity check. -
adt audit— Run security and supply chain audits (pip-audit,bandit,reuse lint). -
adt bump-version <part>— Bump the version inpyproject.toml, create a commit, and tag the release.<part>is one ofmajor,minor,patch, ordaily. -
adt check [paths...]— Run static checkers (ruff check,vulture) on the given paths, or onsrcandtestsby default. -
adt clean— Remove build artifacts, caches, and other cruft (__pycache__,.pytest_cache,dist/,build/, etc.). -
adt cruft— Audit the project against a set of expected standard files (e.g.LICENSE,README,pyproject.toml). -
adt format [paths...]— Format code withruff format(and friends) on the given paths, or onsrcandtestsby default. -
adt help-make— Print amake helpmessage generated from the targets in yourMakefile. -
adt seed— Drop opinionated configuration files into a project from a profile (a reusable bundle of templates, variables, and post-seed scripts). This is what we use to keep dozens of projects aligned on the same conventions without depending on a heavyweight templating tool.adt seed -l— list available profiles (looked up in~/.config/adt/profiles/or sources declared in~/.config/adt/config.toml).adt seed -i <profile>— show a profile's description, variables, template files, and scripts.adt seed -p <profile>— apply a profile (uses the configured default profile when-pis omitted; falls back to built-in templates if no profile is configured).adt seed -v key=value— override or provide template variables; repeat the flag to set several.adt seed -n— dry run: show what would be created or overwritten without touching the filesystem.adt seed -o— overwrite existing files (you'll still be prompted per file unless-yis passed).adt seed -y— skip all confirmation prompts (useful in CI or when re-seeding a known-good profile).
Profiles support inheritance (one profile can
extendanother), conditional templates, Jinja2 rendering, and post-seed scripts that run withADT_*variables exported in the environment. -
adt test [paths...]— Runpyteston the given paths, or onsrcandtestsby default. -
adt typecheck [paths...]— Run thetytype checker on the given paths, or onsrcby default.
We have created Abilian DevTools to help us maintain our own projects, and we thought it could be useful to others.
Here are some of the reasons why we have created this project:
-
Streamlined Tool Collection: Abilian Devtools brings together a wide range of Python development tools in a single, curated package. This allows developers to focus on their work without spending time searching for and integrating individual tools.
-
Consistency: By using a curated set of best-of-breed tools, our team can achieve a consistent level of code quality, style, and security across their projects.
-
Simplified Dependency Management: Instead of managing individual dependencies for each tool, developers only need to add abilian-devtools to their project's
requirements.inorpyproject.toml. This makes it easier to maintain and update dependencies over time. -
Easy-to-use CLI: The
adtcommand-line utility simplifies common development tasks such as running tests, code formatting, and security audits. This can save time and effort, especially for those new to these tools. -
Up-to-date Toolset: Abilian Devtools aims to provide an up-to-date collection of tools, ensuring that developers have access to the latest features and improvements without having to manually track and update each tool.
Here are some ideas for future improvements:
-
Support for additional tools: for instance, tools that deal with changelogs (via Conventional Commits), versioning, documentation...
-
Monorepo support: Better support for monorepos.
-
Customization: The curated nature of Abilian Devtools means that it comes with a predefined set of tools. Your project may require additional or alternative tools, or different settings. ADT could help managing (and updating) settings according to your own tastes and needs.
-
Generating configuration and supporting files: Currently our projects are generated from a template by third-party tools (Cruft or Cookiecutter, using this template). ADT could help generating configuration files for the various tools (
pyproject.toml,setup.cfg,noxfile.py,Makefile,tasks.py...). -
Updating configuration and supporting files: As tools and best practices evolves, and for long-running projects, configuration need to be adapted over time, which can become quite time-consuming, specially when you are working on many small projects (or monorepos). Tools like Cruft or Medikit) can help, but in our experience are too fragile. ADT could help with this.
-
CI/CD: ADT could help with CI/CD integration.
- On GitHub (evergreen)
- On Reddit (May 2023)
- On AFPY Forum (May 2023, in French)
This presentation from 2017 was given at the Paris Open Source Summit (POSS). Many tools have evolved or appeared since then, but the general principles are still valid.
This presentation from 2005 was given (in French) at the "Rencontres Mondiales du Logiciel Libre" in Bordeaux. It's obviously outdated, but kept for nostalgic reasons ;)