troml 0.3.0

Suggests classifiers for Python packages.

troml

troml provides a list of potential classifiers that could be added to a Python package.

It supports modern Python packages that use the pyproject.toml standard (aka PEP 621 and PEP 639). troml follows PEP 561 to determine typing information.

Usage

uvx

uv is an extremely fast Python package and project manager, written in Rust. Self-contained library using uv. uvx is an alias for uv tool run ....

1. Install uv
2. Go to a directory with source code for a Python package
3. uvx troml

pipx

pipx is a way to run install and run Python applications in isolated environments.

1. Install pipx
2. pipx install troml
3. Go to a directory with source code for a Python package
4. troml

pip --user

Install troml to the Python user install directory. More details in the docs.

1. pip install --user troml
2. Go to a directory with source code for a Python package
3. troml

CLI arguments

Specify a path

troml also accepts a path as the first argument.

uvx troml /path/to/pypyproject.toml

Supported classifiers

Python version

troml will read project.python-requires and suggest classifiers based on it.

Legacy license

troml will suggest removing the legacy license classifiers.

Dependencies

troml will suggest classifiers based on the dependencies in project.dependencies, project.dependency-groups, and tool.uv.constraint-dependencies.

Typing

troml will suggest the "Typing :: Typed" classifier based on the existence of the py.typed file in the same directory as the pyproject.toml file.

What's with the name?

• The classifiers for Python are called "Trove classifiers"
• Modern Python packages use TOML for configuration

"trove" 🤝 "TOML"

Yes, I agree, the name is a little weird.

Development

Run from source

uv run troml [PATH-TO-PYPROJECT-TOML]

Commands

• unit tests (via pytest): just test
• linting (via ruff): just lint
• type checking (via mypy): just type
• unit test coverage (via coverage.py): just coverage

Inspiration

• https://indieweb.social/@adamghill/115174743670090084
• another approach from @kfdm: https://codeberg.org/kfdm/add-classifiers
• post about add-classifers: https://paultraylor.net/blog/2025/add-classifiers/
• https://github.com/jvllmr/trove-setup
• https://pypi.org/project/typer/