graph_rag_server/scripts/cleanup_old_code.py

#!/usr/bin/env python3
"""
旧代码清理脚本

此脚本用于安全地删除已废弃的旧代码。
在运行此脚本之前，请确保：
1. 新代码已经过充分测试
2. 所有导入引用已更新
3. 已创建数据备份

警告：此操作不可逆！
"""

import os
import shutil
from pathlib import Path
from typing import List
import argparse


# 要删除的旧代码目录
OLD_CODE_DIRS = [
    'src/api/sdk',
    'src/api/db',
    'src/api/dataset',
    'src/api/mcp',
]

# 要保留的文件（用于向后兼容）
KEEP_FILES = [
    'src/api/__init__.py',  # 包含废弃警告
    'src/api/DEPRECATED.md',  # 废弃说明文档
]


def list_files_to_delete(dirs: List[str]) -> List[Path]:
    """
    列出将要删除的文件
    
    Args:
        dirs: 要删除的目录列表
    
    Returns:
        文件路径列表
    """
    files_to_delete = []
    
    for dir_path in dirs:
        path = Path(dir_path)
        if path.exists() and path.is_dir():
            for file_path in path.rglob('*'):
                if file_path.is_file():
                    files_to_delete.append(file_path)
    
    return files_to_delete


def calculate_size(files: List[Path]) -> int:
    """
    计算文件总大小
    
    Args:
        files: 文件路径列表
    
    Returns:
        总大小（字节）
    """
    total_size = 0
    for file_path in files:
        try:
            total_size += file_path.stat().st_size
        except Exception:
            pass
    return total_size


def format_size(size_bytes: int) -> str:
    """
    格式化文件大小
    
    Args:
        size_bytes: 字节数
    
    Returns:
        格式化的大小字符串
    """
    for unit in ['B', 'KB', 'MB', 'GB']:
        if size_bytes < 1024.0:
            return f"{size_bytes:.2f} {unit}"
        size_bytes /= 1024.0
    return f"{size_bytes:.2f} TB"


def preview_deletion(dry_run: bool = True):
    """
    预览将要删除的内容
    
    Args:
        dry_run: 是否为演练模式
    """
    print("=" * 60)
    print("Old Code Cleanup Script")
    print("=" * 60)
    print()
    
    if dry_run:
        print("🔍 DRY RUN MODE - No files will be deleted")
    else:
        print("⚠️  DELETION MODE - Files will be permanently deleted!")
    
    print()
    print("Directories to clean:")
    for dir_path in OLD_CODE_DIRS:
        status = "✓ exists" if Path(dir_path).exists() else "✗ not found"
        print(f"  - {dir_path} ({status})")
    
    print()
    print("Files to keep:")
    for file_path in KEEP_FILES:
        status = "✓ exists" if Path(file_path).exists() else "✗ not found"
        print(f"  - {file_path} ({status})")
    
    print()
    
    # 列出将要删除的文件
    files_to_delete = list_files_to_delete(OLD_CODE_DIRS)
    
    # 过滤掉要保留的文件
    files_to_delete = [
        f for f in files_to_delete 
        if str(f) not in KEEP_FILES
    ]
    
    if not files_to_delete:
        print("✅ No files to delete")
        return
    
    total_size = calculate_size(files_to_delete)
    
    print(f"Files to delete: {len(files_to_delete)}")
    print(f"Total size: {format_size(total_size)}")
    print()
    
    # 显示前 10 个文件
    print("Sample files (first 10):")
    for file_path in files_to_delete[:10]:
        size = format_size(file_path.stat().st_size)
        print(f"  - {file_path} ({size})")
    
    if len(files_to_delete) > 10:
        print(f"  ... and {len(files_to_delete) - 10} more files")
    
    print()
    
    if not dry_run:
        # 确认删除
        response = input("⚠️  Are you sure you want to delete these files? (yes/no): ")
        if response.lower() != 'yes':
            print("❌ Deletion cancelled")
            return
        
        print()
        print("Deleting files...")
        
        deleted_count = 0
        for dir_path in OLD_CODE_DIRS:
            path = Path(dir_path)
            if path.exists() and path.is_dir():
                try:
                    shutil.rmtree(path)
                    deleted_count += 1
                    print(f"  ✓ Deleted {dir_path}")
                except Exception as e:
                    print(f"  ✗ Error deleting {dir_path}: {e}")
        
        print()
        print(f"✅ Cleanup complete! Deleted {deleted_count} directories")
        print()
        print("Next steps:")
        print("  1. Run tests to ensure everything still works")
        print("  2. Commit the changes")
        print("  3. Update documentation if needed")


def main():
    """主函数"""
    parser = argparse.ArgumentParser(
        description='Clean up old deprecated code',
        formatter_class=argparse.RawDescriptionHelpFormatter,
        epilog="""
Examples:
  # Preview what will be deleted (safe)
  python scripts/cleanup_old_code.py --dry-run
  
  # Actually delete the files (dangerous!)
  python scripts/cleanup_old_code.py --delete

Warning: Deletion is permanent! Make sure you have backups.
        """
    )
    
    parser.add_argument(
        '--dry-run',
        action='store_true',
        default=True,
        help='Preview what will be deleted without actually deleting (default)'
    )
    
    parser.add_argument(
        '--delete',
        action='store_true',
        help='Actually delete the files (use with caution!)'
    )
    
    args = parser.parse_args()
    
    # 如果指定了 --delete，则不是 dry run
    dry_run = not args.delete
    
    preview_deletion(dry_run=dry_run)


if __name__ == '__main__':
    main()