Telegram Chatbot for SymfonyLive Berlin 2023

Find it on Telegram: @SymfonyLiveBerlinBot

• PHP 8.2
• Doctrine compatible database engine
• Symfony CLI (or sth similar)
• Ngrok (or sth similar)
• Telegram Bot Token

The Telegram bot will communicate with our application to process messages and reply to them. This requires a websocket connection that is reachable from the outside world, i.e. the internet. This can be achieved by deploying the application, e.g. using Symfony Cloud, or exposing the local environment through an HTTPS tunnel service like serveo or ngrok.

1. Download and install the ngrok client
2. Sign up for a private ngrok account on https://ngrok.com
3. Go to the Auth section in the dashboard and follow the instructions for setting the auth token
4. Run ngrok: ngrok http https://localhost:8000 The URL https://localhost:8000 refers to our local Symfony Server, so be careful to update the port info if necessary.

When you run ngrok, you will see Forwarding addresses like https://abcd5678.ngrok.io. You will need this value for the WEBHOOK_BASE_URL environment variable. Alternatively you can copy .env as .env.local and set the value there.

You will need to set up a Telegram bot to interact with. You can do this through Telegram by sending a message to a bot @BotFather and following the instructions. Alternatively you can use the CLI as described in the link above.

Once you finished setting up the bot you will get a long message with an auth token that looks something like: 1234567890:ABCD7890efgh1234IJKL5678mnop11223-3.

This token is used for the environment variable TELEGRAM_TOKEN.

Set up a database using a platform supported by Doctrine ORM, e.g. MySQL, PostgreSQL or SQLite3. Set the database DSN using the environment variable DATABASE_URL or by setting the value in our .env.local. You can find an example for an SQLite3 DSN in the .env file, which should be suitable for first development steps.

To set up this demo we need to create a telegram bot and connect our local environment to it (see sections above).

1. Clone the repository and go into the project root:

   git clone [email protected]:chr-hertel/symfonylive-berlin-bot.git
   cd symfonylive-berlin-bot

2. Install the dependencies

   composer install

3. Set up ngrok and start the tunnel

   ngrok http https://localhost:8000

4. Set up Telegram bot (only necessary on first run)

5. Set up the database and environment variables

   • DATABASE_URL - see .env for examples
   • TELEGRAM_TOKEN - token provided by BotFather (eg. 1234567890:ABCD7890efgh1234IJKL5678mnop11223-3)
   • WEBHOOK_BASE_URL - ngrok base url (eg. https://abcd5678.ngrok.io)

6. Start the development server using the Symfony CLI

   symfony serve --detach

7. Set up the entities

   bin/console doctrine:database:create
   bin/console doctrine:schema:create
   bin/console doctrine:fixtures:load # Schedule of 2019

8. Load schedule of 2022 (optional)

   bin/console app:schedule:crawl

9. Set up the webhook & menu (only necessary once)

   bin/console app:webhook:register
   bin/console app:command:register

Relevant tools

• PHP CS Fixer

  php vendor/bin/php-cs-fixer fix
  

• PHPStan

  php vendor/bin/phpstan analyse
  

• PHPUnit

  # without coverage
  php bin/phpunit
  
  # with coverage, but without functional tests
  XDEBUG_MODE=coverage php bin/phpunit --coverage-html=public/phpunit --exclude=functional
  

• Infection (not executed in pipeline)

  php vendor/bin/infection
  

Pipeline is executed with GitHub Actions on Pull Requests.

You can simulate locally with:

bin/check

Deployment is executed with GitHub Action on push on main branch using Deployer.

If you're using PhpStorm and you want to play around with the WebHook controller or Telegram Bot API, see HTTP files:

• Create http-client-private.env.json file with your bot token & your chat id:

  {
    "dev": {
       "bot-token": "bot:TOKEN",
       "chat-id": "123456789"
    }
  }

• Now you can use the example files, to interact with both applications:

  • Telegram Bot API: telegram-bot-api.http
  • ChatBot Webhook: chatbot-webhook.http