hf_scrapper.py

import os
import requests
import json
import urllib.request
from requests.exceptions import RequestException
from tqdm import tqdm

def get_system_proxies():
    try:
        proxies = urllib.request.getproxies()
        print("System proxies:", proxies)
        return {
            "http": proxies.get("http"),
            "https": proxies.get("http")
        }
    except Exception as e:
        print(f"Error getting system proxies: {e}")
        return {}

def download_and_cache_file(file_url, token, cache_path, proxies=None):
    print(f"Downloading file from URL: {file_url} to {cache_path} with proxies: {proxies}")
    try:
        response = requests.get(file_url, headers={'Authorization': f'Bearer {token}'}, proxies=proxies, stream=True)
        response.raise_for_status()

        # Get the total file size from the headers
        total_size = int(response.headers.get('content-length', 0))
        
        os.makedirs(os.path.dirname(cache_path), exist_ok=True)
        with open(cache_path, 'wb') as f:
            # Use tqdm to show download progress
            for chunk in tqdm(response.iter_content(chunk_size=8192), total=total_size//8192, unit='KB', unit_scale=True, unit_divisor=1024):
                if chunk:
                    f.write(chunk)

        print(f'File cached to {cache_path} successfully.')
        return True
    except RequestException as e:
        print(f"Error downloading file: {e}")
    except IOError as e:
        print(f"Error writing file {cache_path}: {e}")
    return False

def get_file_structure(repo, token, path="", proxies=None):
    api_url = f"https://huggingface.co/api/models/{repo}/tree/main/{path}"
    headers = {'Authorization': f'Bearer {token}'}
    print(f"Fetching file structure from URL: {api_url} with proxies: {proxies}")
    try:
        response = requests.get(api_url, headers=headers, proxies=proxies)
        response.raise_for_status()
        return response.json()
    except RequestException as e:
        print(f"Error fetching file structure: {e}")
        return []

def write_file_structure_to_json(file_structure, file_path):
    try:
        with open(file_path, 'w') as json_file:
            json.dump(file_structure, json_file, indent=2)
        print(f'File structure written to {file_path}')
    except IOError as e:
        print(f"Error writing file structure to JSON: {e}")

if __name__ == "__main__":
    file_url = "https://huggingface.co/Unicone-Studio/jellyfin_media/resolve/main/films/Funky%20Monkey%202004/Funky%20Monkey%20(2004)%20Web-dl%201080p.mp4"
    token = os.getenv("TOKEN")
    cache_path = "tmp/cache/films/Funky%20Monkey%202004/Funky%20Monkey%20(2004)%20Web-dl%201080p.mp4"
    proxy = get_system_proxies()
    download_and_cache_file(file_url, token, cache_path, proxies=proxy)