hf_utils.py

from huggingface_hub import snapshot_download
import os
import shutil

def download_filtered_space_files(space_id: str, local_dir: str = "repo_files", file_extensions: list = None):
    """
    Downloads only files with specified extensions from a Hugging Face Space repository.

    Args:
        space_id (str): The ID of the Hugging Face Space (e.g., "naman1102/Final_Assignment_Template").
        local_dir (str): Local directory to store the downloaded files.
        file_extensions (list): List of file extensions to include (e.g., ['.py', '.md']).
                                If None, no filtering is applied (all files are downloaded).
    """
    if not file_extensions:
        raise ValueError("You must specify a list of file extensions to filter by.")

    print(f"Downloading Space '{space_id}' and filtering for: {', '.join(file_extensions)}")

    # Clear out local_dir if it already exists
    if os.path.exists(local_dir):
        shutil.rmtree(local_dir)

    # Convert file extensions to allow_patterns format (e.g., ['.py', '.md'] -> ['*.py', '*.md'])
    allow_patterns = [f"*{ext}" for ext in file_extensions]
    
    # Download directly to local_dir with filtering during download
    repo_path = snapshot_download(
        repo_id=space_id,
        repo_type="space",
        local_dir=local_dir,
        allow_patterns=allow_patterns
    )
    
    # Count downloaded files for feedback
    copied_files = 0
    for root, _, files in os.walk(local_dir):
        for file in files:
            if any(file.endswith(ext) for ext in file_extensions):
                copied_files += 1

    print(f"Downloaded {copied_files} filtered file(s) to: {local_dir}")

# Example usage
# download_filtered_space_files("finegrain/finegrain-image-enhancer", file_extensions=['.py', '.md', '.txt'])  # Downloads only .py, .md, and .txt files

from huggingface_hub import list_spaces

def search_top_spaces(query: str, limit: int = 5):
    """
    Search and return top Hugging Face Space repo IDs based on a keyword.

    Args:
        query (str): The keyword to search for (e.g., "image", "chatbot").
        limit (int): Maximum number of results to return.

    Returns:
        List of repo IDs.
    """
    results = list(list_spaces(search=query, sort="likes", direction=-1))  # Convert generator to list
    top_spaces = [space.id for space in results[:limit]]
    
    return top_spaces

# Example usage
# top_image_spaces = search_top_spaces("tic tac toe", limit=10)
# print("Top games-related Spaces:")
# for space_id in top_image_spaces:
#     print("-", space_id)