vosonandCursor
f5b7011c45
Players get a fast TTL-backed homepage (local profile / Cloudflare D1) with dense UI polish; login unlocks /home without blocking on every OpenDota refresh. Co-authored-by: Cursor <cursoragent@cursor.com>
1.9 KiB
Climperor Cloudflare player data layer
Production multi-user storage for Steam-login player pages:
| Resource | Name | Role |
|---|---|---|
| D1 | climperor-users |
users, stats, heroes, match index, sync jobs |
| R2 | climperor-player-data |
private full match JSON (matches/{id}.json) |
| Queue | climperor-player-sync |
async OpenDota/Steam refresh |
| DLQ | climperor-player-sync-dlq |
failed sync messages |
| Worker | climperor-player-sync |
queue consumer |
Provision
# via keyzoo refining/cloudflare asset
python web/cloudflare/provision.py
Then deploy the worker:
cd web/cloudflare/player-sync
npx wrangler@3 deploy
npx wrangler@3 secret put STEAM_API_KEY
# optional:
npx wrangler@3 secret put OPENDOTA_API_KEY
Pages project climperor-relations needs bindings DB, MATCHES, SYNC_QUEUE
(see web/frontend/wrangler.toml). provision.py binds production and preview
with matching fail_open. Secrets already used by auth:
STEAM_API_KEY, SESSION_SECRET, optional PLAYER_PAGES_PUBLISH_SECRET.
Worker also needs STEAM_API_KEY (python web/cloudflare/put_worker_secrets.py
after staging via _stage_steam_key.py).
R2: if create returns “enable R2”, open
https://dash.cloudflare.com/?to=/:account/r2 once, then re-run provision.py
and uncomment the [[r2_buckets]] block in player-sync/wrangler.toml, redeploy Worker.
Until then, match list/stats still work via D1; full match JSON download is deferred.
Ids are recorded in .resources.json (no secrets).
Local
python web/serve_relations.py continues to use pc/player_pages/ JSON + enrich
(GET /api/players/me runs enrich_profile_recent).
Production /home uses GET /api/players/me (D1 + queue).
Tests
python -m unittest pc.tests.test_player_stats -v
node web/cloudflare/player-sync/test_stats.mjs