hf_space_sync.py

"""
HuggingFace Space Synchronization Utilities

This module handles synchronization between GitHub repositories and HuggingFace Spaces.
"""

import os
import shutil
from pathlib import Path
from typing import List, Dict, Any, Optional
from huggingface_hub import HfApi, Repository, create_repo
from dotenv import load_dotenv
import yaml

load_dotenv()


class HFSpaceSync:
    """
    Manages synchronization between local files and HuggingFace Spaces.
    """

    def __init__(self, config_path: str = "config.yaml"):
        """
        Initialize HuggingFace Space synchronization.

        Args:
            config_path: Path to configuration file
        """
        self.config = self._load_config(config_path)
        self.hf_token = os.getenv("HF_TOKEN")
        self.hf_username = os.getenv("HF_USERNAME")
        self.api = HfApi(token=self.hf_token)

        if not self.hf_token:
            raise ValueError("HF_TOKEN not found in environment variables")

    def _load_config(self, config_path: str) -> Dict[str, Any]:
        """Load configuration from YAML file."""
        try:
            with open(config_path, 'r') as f:
                return yaml.safe_load(f)
        except FileNotFoundError:
            print(f"Config file {config_path} not found, using defaults")
            return {}

    def create_space(self, space_name: str, space_type: str = "streamlit",
                     private: bool = False) -> str:
        """
        Create a new HuggingFace Space.

        Args:
            space_name: Name of the space to create
            space_type: Type of space (streamlit, gradio, static)
            private: Whether the space should be private

        Returns:
            Space repository ID
        """
        try:
            repo_id = f"{self.hf_username}/{space_name}"
            create_repo(
                repo_id=repo_id,
                token=self.hf_token,
                repo_type="space",
                space_sdk=space_type,
                private=private
            )
            print(f"Created space: {repo_id}")
            return repo_id
        except Exception as e:
            if "already exists" in str(e).lower():
                print(f"Space {repo_id} already exists")
                return repo_id
            else:
                raise

    def upload_files(self, space_name: str, local_path: str,
                     repo_path: str = "", commit_message: str = "Update files") -> None:
        """
        Upload files to a HuggingFace Space.

        Args:
            space_name: Name of the space
            local_path: Local file or directory path
            repo_path: Path in the repository (default: root)
            commit_message: Commit message
        """
        repo_id = f"{self.hf_username}/{space_name}"

        try:
            if os.path.isfile(local_path):
                self.api.upload_file(
                    path_or_fileobj=local_path,
                    path_in_repo=repo_path or os.path.basename(local_path),
                    repo_id=repo_id,
                    repo_type="space",
                    commit_message=commit_message
                )
                print(f"Uploaded {local_path} to {repo_id}")
            elif os.path.isdir(local_path):
                self.api.upload_folder(
                    folder_path=local_path,
                    path_in_repo=repo_path,
                    repo_id=repo_id,
                    repo_type="space",
                    commit_message=commit_message
                )
                print(f"Uploaded folder {local_path} to {repo_id}")
        except Exception as e:
            print(f"Error uploading to {repo_id}: {e}")
            raise

    def sync_directory(self, space_name: str, local_dir: str = ".",
                       exclude_patterns: List[str] = None) -> Dict[str, Any]:
        """
        Synchronize a local directory with a HuggingFace Space.

        Args:
            space_name: Name of the space
            local_dir: Local directory to sync
            exclude_patterns: Patterns to exclude from sync

        Returns:
            Dictionary with sync results
        """
        if exclude_patterns is None:
            exclude_patterns = self.config.get('sync', {}).get('exclude_patterns', [])

        repo_id = f"{self.hf_username}/{space_name}"

        # Ensure space exists
        try:
            self.api.repo_info(repo_id=repo_id, repo_type="space")
        except Exception:
            print(f"Space {repo_id} doesn't exist, creating...")
            self.create_space(space_name)

        # Upload directory
        uploaded_files = []
        skipped_files = []

        for root, dirs, files in os.walk(local_dir):
            # Filter directories
            dirs[:] = [d for d in dirs if not any(pattern in d for pattern in exclude_patterns)]

            for file in files:
                file_path = os.path.join(root, file)

                # Check if file should be excluded
                if any(pattern in file_path for pattern in exclude_patterns):
                    skipped_files.append(file_path)
                    continue

                try:
                    rel_path = os.path.relpath(file_path, local_dir)
                    self.api.upload_file(
                        path_or_fileobj=file_path,
                        path_in_repo=rel_path,
                        repo_id=repo_id,
                        repo_type="space",
                        commit_message=f"Sync: {rel_path}"
                    )
                    uploaded_files.append(rel_path)
                    print(f"Synced: {rel_path}")
                except Exception as e:
                    print(f"Error syncing {file_path}: {e}")
                    skipped_files.append(file_path)

        return {
            "space": repo_id,
            "uploaded": len(uploaded_files),
            "skipped": len(skipped_files),
            "uploaded_files": uploaded_files,
            "skipped_files": skipped_files
        }

    def list_spaces(self) -> List[Dict[str, Any]]:
        """
        List all spaces for the authenticated user.

        Returns:
            List of space information dictionaries
        """
        spaces = self.api.list_models(author=self.hf_username, filter="space")
        return [{"id": space.id, "private": space.private} for space in spaces]

    def download_space(self, space_name: str, local_dir: str) -> str:
        """
        Download a space to a local directory.

        Args:
            space_name: Name of the space
            local_dir: Local directory to download to

        Returns:
            Path to downloaded directory
        """
        repo_id = f"{self.hf_username}/{space_name}"

        from huggingface_hub import snapshot_download

        snapshot_download(
            repo_id=repo_id,
            repo_type="space",
            local_dir=local_dir,
            token=self.hf_token
        )

        print(f"Downloaded {repo_id} to {local_dir}")
        return local_dir


if __name__ == "__main__":
    # Example usage
    sync = HFSpaceSync()
    print("Available spaces:")
    for space in sync.list_spaces():
        print(f"  - {space['id']}")