Contributing

Bug reports, fixes, features and documentation are all welcome. For anything bigger than a small fix, open an issue first so the scope is agreed before you write the code.

Set up

You need uv and Python 3.11 or newer.

git clone https://github.com/mftnakrsu/Automatic_Number_Plate_Recognition_YOLO_OCR.git
cd Automatic_Number_Plate_Recognition_YOLO_OCR
uv sync --all-extras --group dev
uv run pre-commit install   # optional: run the checks on every commit

The checks

CI runs these on Python 3.11, 3.12, 3.13 and 3.14, then builds the Docker image and calls its /health endpoint. Run them before you push:

make lint   # ruff check and ruff format --check
make type   # pyright
make test   # pytest

The default test suite uses fake models and needs no network. The end-to-end test with the real models downloads them on its first run (about 30 MB):

ANPR_RUN_REAL=1 uv run pytest tests/integration/test_pipeline_real.py

Conventions

Branch names start with feat/, fix/, docs/, chore/, ci/, test/ or refactor/.
Commit messages follow Conventional Commits: feat:, fix:, docs: and so on.
One logical change per pull request.
Code is formatted with ruff format at 100 columns and type-checked with pyright. Public functions have full type hints.
Never store plate text. Anything that persists a plate goes through storage.hashing.hash_plate.

The documentation

This site lives in docs-site/ and is built with Astro Starlight. Pages are Markdown files in docs-site/src/content/docs/, with Turkish translations under tr/.

cd docs-site
npm install
npm run dev   # http://localhost:4321

Every push to main that touches docs-site/ publishes the site.