Data Maintenance
This project keeps historical Bahamut Anime data on the accumulate-data branch.
Branches
main contains the application code, documentation, tests, and release metadata.
accumulate-data contains the generated data and plot assets used by GitHub Pages. The
large anime.db file intentionally remains on that branch for now.
gh-pages contains the deployed static site generated by MkDocs.
Daily Collection
Daily data collection is currently triggered outside this repository. There is no scheduled
GitHub Actions workflow on main that records new daily data.
Until that process is moved into the repository, treat the external collector as the owner of:
- updating
anime.dbonaccumulate-data - committing generated data changes
- pushing
accumulate-data - triggering plot publication through the existing branch workflow
Manual Plot Regeneration
Use the Trigger plot manually workflow when plot assets need to be regenerated from the
current accumulate-data branch.
That workflow:
- checks out
accumulate-data - installs the project dependencies
- regenerates plot HTML and trend JSON files
- commits the changed
docs/assets if there are changes - builds and publishes GitHub Pages
Recovery Notes
If daily collection fails, do not rewrite accumulate-data history unless a corrupted commit has
not been shared. Prefer a forward fix commit that restores or regenerates the missing data.
If generated plot assets are stale but anime.db is correct, rerun the manual plot workflow.
If anime.db itself is stale, rerun the external collector first, then regenerate plots.
Release Check
Python package releases are tag driven. A version is published only after the corresponding tag exists on GitHub and the PyPI publish workflow succeeds.
Useful checks:
git ls-remote --tags origin
curl --max-time 10 --fail --silent --show-error https://pypi.org/pypi/bahamut-ani-stat/json