Skip to content

Latest commit

 

History

History
37 lines (25 loc) · 3.46 KB

README.md

File metadata and controls

37 lines (25 loc) · 3.46 KB

JingleBot

Jingle Jam Discord bot built on Cloudflare Workers, providing stats commands and automatic announcements

Development

  1. Create your test Discord application at https://discord.com/developers/applications (this does not need a bot account, just the application).
  2. Create your .dev.vars file.
    • Copy .dev.vars.example and fill out the information from your Discord application, plus the ID of your test server/guild where you'll use the bot.
    • Optionally, DISCORD_SUMMARY_CHANNEL + DISCORD_MILESTONE_CHANNEL can be set to channel IDs for testing the scheduled messages, as well as the DISCORD_BOT_TOKEN to send the messages as.
    • Optionally, DISCORD_CAUSES_EMOJI + DISCORD_REGULAR_EMOJI can be set to a JSON object mapping the cause names + regular emoji (see const regular in src/util/emoji.ts) to custom emoji Markdown (which can be uploaded directly to the Discord application).
  3. Authenticate with Wrangler by running npx wrangler login.
  4. Update wrangler.toml for your account.
    • Use npx wrangler whoami to get your account ID, update the value in wrangler.toml to match.
    • Use npx wrangler kv:namespace create "STORE" to create the KV namespace, update the id and preview_id in wrangler.toml to match.
  5. Develop with the worker by running npm run dev.
    • To test the CRON functionality, hit the /__scheduled?cron=*+*+*+*+* endpoint.
  6. (Optional) Start an HTTP tunnel to your local development server by running npm run tunnel, using cloudflared.
    • To use a custom hostname in your Cloudflare account, first make sure you're authenticated with cloudflared login, then run npm run tunnel -- --overwrite-dns --hostname <hostname> --name <name> (where <name> can be any convenient name for the tunnel).

Deployments

wrangler.toml and this repository is currently designed for a staging deployment and a production deployment.

Ensure that the environment in wrangler.toml has been updated with your chosen for the worker.

Ensure that the KV namespaces are created for staging/production environments and are configured in wrangler.toml. Use npx wrangler kv:namespace create "STORE" -e <staging/production>.

You'll also want to set DISCORD_CLIENT_ID + DISCORD_PUBLIC_KEY + STATS_API_ENDPOINT (optionally, DISCORD_SUMMARY_CHANNEL + DISCORD_MILESTONE_CHANNEL + DISCORD_BOT_TOKEN + DISCORD_CAUSES_EMOJI + DISCORD_REGULAR_EMOJI + WORKER_BASE_URL) as secrets for the worker, which you can do with npx wrangler secret put <var name> -e <staging/production> (the channel secrets can contain multiple IDs, separated by a comma).

If you're deploying for local, make sure that you've got the appropriate environment variables set for DISCORD_CLIENT_ID, DISCORD_CLIENT_SECRET + DISCORD_GUILD_ID (otherwise, they'll default to the values in .dev.vars).

To deploy from local, run npm run publish:staging to deploy to staging, and npm run publish:production to deploy to the production environment.

To deploy using GitHub, run make deploy-staging to force push and deploy to staging, and make deploy-production to force push and deploy to the production environment.

Live logs for both environments can be accessed with npm run logs:staging and npm run logs:production as needed.