jangroth / sauerpod

A grumpy Telegram bot that converts video streams into audio tracks and publishes those as an individual podcast feed.

Sauerpod

A sauer (🇩🇪 for 'grumpy') Telegram bot that converts video streams into audio tracks and publishes those as an individual podcast feed. Implemented in Python using AWS cloud-native services. No servers required.

Implementation

• AWS Component Development Kit v2 (Python)
• Telegram API
• Serverless AWS:
  • Amazon API Gateway
  • AWS Step Functions
  • AWS Lambda
  • Amazon S3
  • Amazon DynamoDB
  • Amazon CloudFront
  • AWS Systems Manager

Logic Flow

Development setup (WIP)

Intial setup

Required once to set up the project:

Telegram

• Create bot
  • -> Note API token from botfather
• Initiate chat
  • -> Note chat id (find out with e.g. @RawDataBot)

AWS

• Store secrets in SSM
  • Change values & run ./scripts/bootstrap-secrets.sh
  • Revert changes in file (to not commit secrets to git)
    • git restore scripts/bootstrap-secrets.sh

Local development

Required for development work:

• Create/activate venv (. .venv/bin/activate)
• Configure AWS profile (export AWS_PROFILE=xyz)
• Deploy stack (cdk deploy --all or cdk watch)
• Point bot`s webhook to Lambda
  • Source(!) secrets (don't execute the script) . ./scripts/get-secrets-from-ssm.sh
  • Update web url at bot ./scripts/update-web-url-at-bot.sh

Resources

• https://core.telegram.org/bots/api
• https://xabaras.medium.com/setting-your-telegram-bot-webhook-the-easy-way-c7577b2d6f72

Disclaimer and Credits

• 'Grumpy bot' icons by www.flaticon.com.
• Sauerpod is meant as a POC to demonstrate the power and elegance of cloud-native Python development. It uses pytube to download audio streams of videos and temporarily store them in a private S3 bucket. It must not be used to share any copyright-protected media.