github_storage.py

import os
from github import Github
import base64
import shutil
import zipfile
from io import BytesIO


# Global Variables

# HF ------------ 
hf_folder_path = '/home/user/app/embeddings'
zip_name = 'embeddings'

# Github -------
github_token = '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'  # Replace with your GitHub token
repo_name = 'AdityaMetkar/Patseer-Database'  # Replace with your repository, e.g., 'octocat/Hello-World'
folder_path = 'Manual Database/embeddings.zip'  # Replace with the path to the folder in the repository

# Authenticate to GitHub
g = Github(github_token)
repo = g.get_repo(repo_name)



# Functions -------------------------------
def zip_folder():
    shutil.make_archive(zip_name, 'zip', hf_folder_path)
    return zip_name + '.zip'


def update_db():

    try:
        # Check if the file already exists in the repository
        existing_file = repo.get_contents(folder_path)

        compressed_zip = zip_folder()
        with open(compressed_zip, 'rb') as file:
            file_content = file.read()

        # Update the existing file
        repo.update_file(existing_file.path, "New DB Update", file_content, existing_file.sha)
        print(f"Updated {folder_path} in GitHub repository.")

    except Exception as e:
        print(f"Error: {e}")


def download_db():
    if not os.path.exists(hf_folder_path):
        os.makedirs(hf_folder_path)
        
    file_content = repo.get_contents(folder_path)

    try:
        # Download the zip file content from GitHub
        file_content = repo.get_contents(folder_path)
        zip_data = base64.b64decode(file_content.content)

        # Extract the downloaded zip file directly to hf_folder_path using shutil
        with zipfile.ZipFile(BytesIO(zip_data)) as zip_ref:
            for file in zip_ref.namelist():
                zip_ref.extract(file, hf_folder_path)

        print(f"Successfully unzipped files to {hf_folder_path}")

    except Exception as e:
        print(f"Error: {e}")




# Download the folder
# download_folder()
# update_db()