Scraper automation (docker-compose):
- Add scraper service to docker-compose.yml using the same image and
shared park_data volume; overrides CMD to run scrape-schedule.sh
- scripts/scrape-schedule.sh: runs an initial scrape on container start,
then sleeps until 3:00 AM (respects TZ env var) and repeats nightly;
logs timestamps and next-run countdown; non-fatal on scrape errors
Staleness window: 7 days → 72 hours in lib/db.ts so data refreshes
more frequently with the automated schedule in place
Remove favicon: delete app/icon.tsx and public/logo.svg
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Buyout days are now treated as closed unless they carry a Passholder
Preview event, in which case they surface as a distinct purple cell
in the UI showing "Passholder" + hours
- DB gains a special_type column (auto-migrated on next startup)
- scrape.ts threads specialType through to upsertDay
- debug.ts now shows events, isBuyout, isPassholderPreview, and
specialType in the parsed result section
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Creates debug/ folder (txt files gitignored). Output is printed to
the terminal and written to the file simultaneously.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
npm run debug -- --park greatadventure --date 2026-07-04
Prints the raw API response for that day alongside the parsed result
so mismatched or missing hours can be traced to their source.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Each park prints a row of █ (fetched) and · (skipped) as months
complete, then ends with open day count, "up to date", or error count.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Next.js 15 + Tailwind CSS v4 week calendar showing Six Flags park hours.
Scrapes the internal CloudFront API, stores results in SQLite.
Includes Dockerfile (Debian/Playwright-compatible), docker-compose, and
Gitea Actions pipeline that builds and pushes to the container registry.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
