Quick flask guide

• Sign up for OpenAI and get your secret key. This is rotated periodically so needs to be updated.
• Recommended to instead use a Python virtual environment. Download the following dependencies with requirements.txt instead. Else:
• Install Python
• Install Flask pip install Flask
• Install RabbitMQ pip install pika
• Install OpenAI pip install requests
• Install dotenv to enable Environment Variables pip install python-dotenv

To run the application, use the flask command or python -m flask. You need to tell the Flask where your application is with the --app option.

0. Create .env file (ask Pan for credentials) (OpenAI console):

OPENAI_SECRET_KEY="Get your own from the OpenAI console"
RABBITMQ_HOSTNAME='Ask Pan'
RABBITMQ_USER='Ask Pan'
RABBITMQ_PORT='5672'
RABBITMQ_PASSWORD='Ask Pan'
RABBITMQ_PREFIX='amqps://'
ENVIRONMENT='prod'
SENDER="Ask Pan"

1. Run RabbitMQ locally: docker run -it --rm --name rabbitmq -p 5672:5672 -p 15672:15672 rabbitmq:3.11-management
2. flask --app app.py run (py -m flask --app app.py run)
   • To run in debug mode (automatic changes): py -m flask --app app run --debug
3. Go to root of localhost site to start the RabbitMQ connection (http://127.0.0.1:5000)

RabbitMQ dashboard to manage messages:

• http://localhost:15672/
• user pass are both guest

• Python app deployed with Vercel at https://quizgpt-service.vercel.app/
• For this to work, need the following versions installed:

Flask==1.1.4
MarkupSafe==2.0.1

• Dashboard: https://vercel.com/fryingpannn/quizgpt-service
• To redeploy, push changes to the repo or vercel --prod.

------- PYTHONANYWHERE DOESN'T WORK WITH NON-PUBLIC APIs -> Use the Vercel deployed URL.

• The Python app is deployed using PythonAnywhere at http://fryingpannn.pythonanywhere.com/
  • https://www.pythonanywhere.com/user/fryingpannn/webapps/#tab_id_fryingpannn_pythonanywhere_com
  • Files must be changed directly from within the dashboard source code.
  • Use bash console on the dashboard to install new libraries.

• The RabbitMQ service is deployed using CloudAMQP at https://api.cloudamqp.com/console/eabc96f0-a3bb-456b-b3f2-04b798d7e228/details. Messages inside can be managed from the console.
• Ask Pan for the URL of this service.

Receiving message format:

• number: number of questions to generate
• text: text to generate questions from
• id: unique identifier for quiz
• sender: has to be an allowed sender

Sending message format:

• id: unique identifier for quiz
• number: number of questions generated
• results: list of question-answer pairs
• sender: has to be an allowed sender

E.g.:

{
    "id": 1,
    "number": 3,
    "text": "Data structures and algorithms. Medium difficulty.",
    "sender": SENDER from env file
}
{
    "id": 2,
    "number": 3,
    "text": "Vegetables and fruits. Medium difficulty.",
    "sender": SENDER from env file
}

• If PythonAnywhere hosted instance doesn't connect to hosted RabbitMQ instance. Simply run the Python service locally to the hosted RabbitMQ instance.