mirror of
https://github.com/tiennm99/thptqg.git
synced 2026-10-12 22:14:59 +00:00
docs: expand README, add architecture / deployment / data-pipeline docs
- README: live-site URLs, features, scripts, project layout, quickstart - docs/system-architecture.md: data flow, 3-variant deploy mechanism, schema, parse-quirk matrix, score-tier model, admission-block model, frontend key behaviors (deep-link, share, keyboard shortcuts) - docs/data-pipeline.md: source Excel shape, score regex, why three parsers, overflow-sheet gotcha, audit tooling, refresh flow - docs/deployment-guide.md: CI workflow, adding a new variant, rollback, GH Pages file-size note - docs/README.md: index for the docs dir
This commit is contained in:
1 parent
1940e8936d
commit
8ca6bfa7a8
5 files changed
+336
-3
No files matched your search
@@ -0,0 +1,52 @@
|
||||
# Deployment Guide
|
||||
|
||||
Deploys to GitHub Pages via `.github/workflows/deploy.yml`. Every push to `main` rebuilds and redeploys all 3 variants.
|
||||
|
||||
## What the workflow does
|
||||
|
||||
1. Checkout, setup Node 20, `npm ci`
|
||||
2. `npm run build:db:all` — builds 3 SQLite DBs (main, old, old2)
|
||||
3. Gzip each DB at level 9
|
||||
4. `npm run build:all` — builds 3 Vite bundles into `dist/`, `dist/old/`, `dist/old2/`
|
||||
5. Remove uncompressed `.db` files from `dist/` (only `.gz` ships)
|
||||
6. `actions/upload-pages-artifact` + `actions/deploy-pages`
|
||||
|
||||
Total CI time ≈ 4–6 min (DB build dominates).
|
||||
|
||||
## Resulting URLs
|
||||
|
||||
- `https://<user>.github.io/thptqg2017/`
|
||||
- `https://<user>.github.io/thptqg2017/old/`
|
||||
- `https://<user>.github.io/thptqg2017/old2/`
|
||||
|
||||
## Local reproduction
|
||||
|
||||
```bash
|
||||
npm ci
|
||||
npm run build:db:all
|
||||
gzip -kf -9 public/thptqg2017.db public-old/thptqg2017.db public-old2/thptqg2017.db
|
||||
npm run build:all
|
||||
npx serve dist # or any static server
|
||||
```
|
||||
|
||||
## Adding a new variant
|
||||
|
||||
1. Drop source Excel files into a new `data-vN/` folder
|
||||
2. Add `public-vN/` to `.gitignore` patterns for the `.db` + `.db.gz`
|
||||
3. Copy `scripts/build-database-old.js` → `scripts/build-database-vN.js`; update `SRC_DIR` / `DB_PATH`; adjust parse logic for the new quirks (multi-sheet? blank rows? numeric guard?)
|
||||
4. Add variant to `VARIANT_CONFIG` in `vite.config.js`:
|
||||
```js
|
||||
vN: { base: "/thptqg2017/vN/", publicDir: "public-vN", outDir: "dist/vN" }
|
||||
```
|
||||
5. Add `build:db:vN` and `build:vN` npm scripts; wire them into `build:db:all` and `build:all`
|
||||
6. Update `.github/workflows/deploy.yml` — add gzip step + rm step for the new DB
|
||||
|
||||
## Notes
|
||||
|
||||
- **DB gzip is non-cacheable across deploys** — every rebuild produces a new `.db.gz` (not byte-identical due to SQLite page shuffling). First-visit users pay the 47 MB download; subsequent visits hit browser cache until the next deploy.
|
||||
- **GitHub Pages file size limit**: individual files ≤ 100 MB. Main gzipped DB is ~47 MB — safe. Uncompressed 159 MB DB would not fit; that's why we ship `.gz` and the frontend decompresses via `DecompressionStream`.
|
||||
- **No server-side compression assumption.** The app reads the `.gz` file directly (not `Content-Encoding: gzip`). GH Pages does not reliably gzip on-the-fly for arbitrary paths; shipping pre-gzipped bytes is deterministic.
|
||||
|
||||
## Rollback
|
||||
|
||||
Deploys are stateless snapshots. To rollback, revert the commit on `main` and push — the next workflow rebuilds the older state. There's no state to migrate.
|
||||
Reference in new issue
Block a user