src/gaia/utils/cli/organize_files.py

#!/usr/bin/env python
"""
GAIA Project File Organizer

This script organizes files in the GAIA project by moving them to appropriate directories:
1. Moves deployment-related files to the deployment directory
2. Moves example notebooks to the examples directory
3. Moves utility scripts to the utils directory

Usage:
    python src/gaia/utils/cli/organize_files.py [--dry-run]
"""

import os
import sys
import shutil
import logging
import argparse
from pathlib import Path
from typing import Dict, List, Tuple

# Configure logging
logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s - %(levelname)s - %(message)s",
    handlers=[
        logging.StreamHandler()
    ]
)
logger = logging.getLogger("organize_files")

# Files to move to deployment directory
DEPLOYMENT_FILES = [
    "prepare_for_hf_deployment.py",
    "hf_deployment_instructions.md",
    "huggingface_space_deployment_guide.md",
    "huggingface_space_update_guide.md",
    "huggingface_space_update_guide.md",
    "huggingface_space_cleanup.md",
    "huggingface_space_deployment_summary.md",
    "space_config.json",
    "deploy_to_hf.py",
    "deploy_to_huggingface_credential.py",
    "deploy_to_huggingface_dotenv.py",
    "deploy_to_huggingface_env.py",
    "deploy_to_huggingface.py",
    "deploy_with_token.py",
    "deployment_checklist.md",
    "deployment_readiness_report.json",
    "deployment_readiness_report.md",
    "deployment_verification.py",
    "final_deployment_guide.md",
    "complete_deployment_guide.md",
    "complete_deployment_solution.md",
    "get_huggingface_token.py",
    "test_deployment_readiness.py",
    "verify_deployment_readiness.py",
    "production_cleanup.py",
    "production_readiness_check.py",
    "production_test.py",
    "production_verification_report.md",
]

# Example notebooks to move to examples directory
EXAMPLE_FILES = [
    "mail_sorting.ipynb",
    "agent.ipynb"
]

# Utility scripts to move to utils directory
UTIL_FILES = [
    "generate_env_template.py",
    "fix_encoding.py",
    "fix_graph_imports.py",
    "remove_debug_statements.py",
    "remove_duplicate_tests.py",
    "run_verification_tests.py",
    "simple_memory_test.py",
    "simple_test.py",
    "validate_app_simple.py",
    "validate_app.py",
    "update_agent.py",
    "verify_components.py",
    "verify_minimal.py",
    "create_env_and_test.py"
]

def move_file(src: str, dest_dir: str, dry_run: bool = False) -> bool:
    """
    Move a file to the specified directory
    
    Args:
        src: Source file path
        dest_dir: Destination directory
        dry_run: If True, only show what would be moved without making changes
        
    Returns:
        bool: True if successful or would be successful, False otherwise
    """
    if not os.path.exists(src):
        logger.warning(f"Source file does not exist: {src}")
        return False
        
    if not os.path.exists(dest_dir):
        if dry_run:
            logger.info(f"DRY RUN: Would create directory: {dest_dir}")
        else:
            os.makedirs(dest_dir, exist_ok=True)
            logger.info(f"Created directory: {dest_dir}")
    
    dest_path = os.path.join(dest_dir, os.path.basename(src))
    
    if dry_run:
        logger.info(f"DRY RUN: Would move {src} -> {dest_path}")
        return True
        
    try:
        shutil.move(src, dest_path)
        logger.info(f"Moved {src} -> {dest_path}")
        return True
    except Exception as e:
        logger.error(f"Error moving {src} to {dest_path}: {str(e)}")
        return False

def main():
    """Main function to organize files."""
    parser = argparse.ArgumentParser(description="GAIA Project File Organizer")
    parser.add_argument("--dry-run", action="store_true", help="Show what would be moved without making changes")
    args = parser.parse_args()
    
    if args.dry_run:
        logger.info("Running in DRY RUN mode - no files will be moved")
    
    # Track results
    results = {
        "deployment": {"success": 0, "failed": 0, "missing": 0},
        "examples": {"success": 0, "failed": 0, "missing": 0},
        "utils": {"success": 0, "failed": 0, "missing": 0}
    }
    
    # Move deployment files
    logger.info("\n=== Moving deployment files ===")
    for file in DEPLOYMENT_FILES:
        if os.path.exists(file):
            success = move_file(file, "deployment", args.dry_run)
            if success:
                results["deployment"]["success"] += 1
            else:
                results["deployment"]["failed"] += 1
        else:
            logger.info(f"File not found (skipping): {file}")
            results["deployment"]["missing"] += 1
    
    # Move example files
    logger.info("\n=== Moving example files ===")
    for file in EXAMPLE_FILES:
        if os.path.exists(file):
            success = move_file(file, "examples", args.dry_run)
            if success:
                results["examples"]["success"] += 1
            else:
                results["examples"]["failed"] += 1
        else:
            logger.info(f"File not found (skipping): {file}")
            results["examples"]["missing"] += 1
    
    # Move utility files
    logger.info("\n=== Moving utility files ===")
    for file in UTIL_FILES:
        if os.path.exists(file):
            success = move_file(file, "utils", args.dry_run)
            if success:
                results["utils"]["success"] += 1
            else:
                results["utils"]["failed"] += 1
        else:
            logger.info(f"File not found (skipping): {file}")
            results["utils"]["missing"] += 1
    
    # Print summary
    logger.info("\n=== File Organization Summary ===")
    logger.info(f"Deployment Files: {results['deployment']['success']} moved, {results['deployment']['failed']} failed, {results['deployment']['missing']} not found")
    logger.info(f"Example Files: {results['examples']['success']} moved, {results['examples']['failed']} failed, {results['examples']['missing']} not found")
    logger.info(f"Utility Files: {results['utils']['success']} moved, {results['utils']['failed']} failed, {results['utils']['missing']} not found")
    
    # Return success code if no failures
    total_failures = sum(result["failed"] for result in results.values())
    return 0 if total_failures == 0 else 1

if __name__ == "__main__":
    sys.exit(main())