download_repo.py

import urllib.parse  
import requests  
import io  
import json 


import logging

# Set up logging
logging.basicConfig(level=logging.ERROR, format='%(asctime)s - %(levelname)s - %(message)s')

def download_gitlab_repo_to_hfspace(api_url, project_id, version, target_folder, hf_api, hf_space_name):
    try:
        # Construct the URL for the release's zip file
        encoded_project_id = urllib.parse.quote(project_id, safe="")
        url = f"{api_url}/projects/{encoded_project_id}/repository/archive.zip?sha={version}"
        print("URL: ", url)

        logging.info(f"Constructed URL: {url}")

        # Send GET request to download the zip file
        response = requests.get(url)
        logging.info(f"HTTP GET Request sent to {url}, Status Code: {response.status_code}")

        if response.status_code == 200:
            content_disposition = response.headers.get('content-disposition')
            print("Content_disposition:", content_disposition)
            if content_disposition:
                filename = content_disposition.split('filename=')[-1].strip('\"')
                print("FILENAME: ", filename)
            else:
                filename = 'archive.zip'  # Fallback to a default name if not found

            logging.info(f"Filename determined: {filename}")
        else:
            logging.error(f"Failed to download the file. HTTP Status: {response.status_code}")
            return

        existing_files = hf_api.list_repo_files(repo_id=hf_space_name, repo_type='space')
        target_path = f"{target_folder}/{filename}"
        print("TARGET PATH: ", target_path)

        logging.info(f"Target Path: '{target_path}'")
        logging.info(f"Existing Files: {[repr(file) for file in existing_files]}")
        
        if target_path in existing_files:
            logging.warning(f"File '{target_path}' already exists in the repository. Skipping upload...")
            print(f"File '{target_path}' already exists in the repository. Skipping upload...")
        else:
            repo_content= io.BytesIO(response.content)
            print(repo_content)
            _upload_file_to_hfspace(repo_content, hf_api, target_path, hf_space_name)

    except FileNotFoundError:
        logging.error("The config.json file was not found. Please ensure it exists in the project directory.")
        print("The config.json file was not found. Please ensure it exists in the project directory.")
    except json.JSONDecodeError:
        logging.error("Failed to parse the config.json file. Please ensure it contains valid JSON.")
        print("Failed to parse the config.json file. Please ensure it contains valid JSON.")
    except Exception as e:
        logging.error(f"An unexpected error occurred: {e}")
        print(f"An unexpected error occurred: {e}")

def _upload_file_to_hfspace(content, hf_api, target_path, hf_space_name):
    try:
        logging.info(f"Preparing to upload file to '{target_path}'")
        hf_api.upload_file(
            path_or_fileobj=content,
            path_in_repo=target_path,
            repo_id=hf_space_name,
            repo_type="space"
        )
        logging.info(f"File successfully uploaded to '{target_path}'")
    except Exception as e:
        logging.error(f"Error during file upload: {e}")
        print(f"Error during file upload: {e}")