Test uses the real index.html file and scripts in a headless browser using Playwright. Output is compared against known valid (golden) files. This way we are testing the output of the actual tool, not a reimplementation.

Setup and running

python3 -m venv .venv
source .venv/bin/activate
pip install -r tests/requirements.txt
playwright install chromium

To run:

python3 tests/test_conversions.py

Prints ok/FAIL per scenario and exits non-zero if anything doesn't match.

Updating golden files and test image

To generate the test image again (only needed if we changed the image):

python3 tests/generate_test_image.py

After an intentional change to the conversion logic, regenerate the golden output and optionally review the diff before committing:

python3 tests/test_conversions.py --update
git diff tests/golden