Hackathon Ponts ENPC

Install

• when opening VSCode, install the suggested extensions (Python, Black Formatter and Pylance)
• create your python environment python3 -m venv .venv
• copy the .env.example file to a .env file
• replace the OPENAI_API_KEY and OPENAI_ORGANIZATION env variables with the real values
• activate your environment with
• download necessary data with python -m nltk.downloader all
• run the server with flask --app main run --debug

The server should answer on http://localhost:5000

You can deactivate the environment with deactivate.

Adding librairies

if you need to use a new librairies, you can do it with pip pip install [library name] or pip3 install [library name]

Project structure

Top-level layout:

• main.py — Flask application and routes
• requirements.txt — Python dependencies
• templates/ — HTML templates (contains index.html)
• static/ — static files (JS/CSS)
• src/utils/ — utility modules used by the app
  • ask_question_to_pdf.py — functions to ask questions to PDF/text and interface with GPT
  • ask_question_to_txt.py — (alternate) text-based helper
  • test_unitaire.py — small unit test scaffolding
  • histoire.txt, test.txt — sample text files

Endpoints

The Flask app exposes the following endpoints (see main.py):

• GET / - returns a small HTML/text ("Hello, World!")
• GET /hello/ - renders the index.html template
• POST /prompt - expects a form field prompt and returns a JSON object { 'answer': ... } from ask_question_to_pdf
• GET /question - returns a generated question via gpt3_question()
• POST /answer - expects a form field prompt and returns { 'answer': ... } using gpt3_answer

Example curl (POST to /prompt):

curl -X POST -F "prompt=Quel est le résumé ?" http://127.0.0.1:5000/prompt

Development notes

• The Flask app imports functions from src/utils/ask_question_to_pdf.py. If you change those utilities, restart the Flask server to pick changes.
• If you enable OpenAI features, make sure OPENAI_API_KEY is set and valid. The project currently expects the key to be available via environment variables.

Adding dependencies

Install via pip and update requirements.txt:

pip install <package>
pip freeze > requirements.txt