JamesFincher/nbAPI

An api that parses an uploaded text file into Jupyter Notebook.

FastAPI Notebook Creator

A FastAPI web service that creates Jupyter notebooks from text input or files, and commits the generated notebooks to a specified GitHub repository.

Table of contents

• Description
• Installation
• Configuration
• Usage
• API Endpoints

Description

The service accepts input text (with specific formatting) or a file and generates a Jupyter notebook from it. It then commits the generated notebook to a specified GitHub repository, returning the GitHub URL of the notebook.

The input text should have the following format:

• # followed by a space, for an <h1> markdown cell.
• ## followed by a space, for an <h2> markdown cell.
• Triple backticks followed by "python" to start a code cell, and triple backticks again to close it.

Installation

1. Clone the repository.

   git clone https://github.com/your-username/repository-name.git
   

2. Create a virtual environment and activate it.

   python -m venv venv
   source venv/bin/activate
   

3. Install the required packages from requirements.txt using pip.

   pip install -r requirements.txt
   

4. Rename the .env.sample file to .env and fill in your GitHub token, repository owner, repository name, and branch.

Configuration

Set environment variables in the starter .env file. Fill in the following environment variable values:

GITHUB_TOKEN=your_github_token_here
REPO_OWNER=your_repo_owner_here
REPO_NAME=your_repo_name_here
BRANCH_NAME=your_branch_name_here

Usage

1. Start the FastAPI server.

   uvicorn main:app --reload
   

2. Access the interactive documentation at http://localhost:8000/docs.

API Endpoints

• GET /: Get a status message to check if the API is working fine
• POST /create_notebook/: Create a Jupyter notebook from an uploaded file
  • Input: FormData with file containing input text
  • Output: JSON with a message indicating the success result, and the GitHub URL of the created notebook
• POST /create_notebook_text/: Create a Jupyter notebook from a text input
  • Input: Form input containing text (content type: application/x-www-form-urlencoded)
  • Output: JSON with a message indicating the success result, and the GitHub URL of the created notebook