serv_benchmark/backend/migrate_file_organization.py

#!/usr/bin/env python3
"""
Migrate existing uploads to organized structure
Moves files from uploads/ to uploads/{hostname}/images or uploads/{hostname}/files
"""

import os
import shutil
import sys
from pathlib import Path

# Add parent directory to path
sys.path.insert(0, str(Path(__file__).parent))

from sqlalchemy.orm import Session
from app.db.session import SessionLocal
from app.core.config import settings
from app.models.device import Device
from app.models.document import Document
from app.utils.file_organizer import (
    sanitize_hostname,
    is_image_file,
    ensure_device_directories
)


def migrate_files(dry_run: bool = True):
    """
    Migrate existing files to organized structure

    Args:
        dry_run: If True, only print what would be done
    """
    db: Session = SessionLocal()

    try:
        # Get all documents
        documents = db.query(Document).all()

        print(f"Found {len(documents)} documents to migrate")
        print(f"Mode: {'DRY RUN' if dry_run else 'ACTUAL MIGRATION'}")
        print("-" * 80)

        migrated_count = 0
        error_count = 0
        skipped_count = 0

        for doc in documents:
            # Get device
            device = db.query(Device).filter(Device.id == doc.device_id).first()

            if not device:
                print(f"❌ Document {doc.id}: Device {doc.device_id} not found - SKIPPING")
                error_count += 1
                continue

            # Check if file exists
            if not os.path.exists(doc.stored_path):
                print(f"⚠️  Document {doc.id}: File not found at {doc.stored_path} - SKIPPING")
                skipped_count += 1
                continue

            # Determine if image
            is_image = is_image_file(doc.filename, doc.mime_type)
            file_type = "image" if is_image else "file"

            # Get new path
            sanitized_hostname = sanitize_hostname(device.hostname)
            subdir = "images" if is_image else "files"
            filename = os.path.basename(doc.stored_path)

            new_path = os.path.join(
                settings.UPLOAD_DIR,
                sanitized_hostname,
                subdir,
                filename
            )

            # Check if already in correct location
            if doc.stored_path == new_path:
                print(f"✓  Document {doc.id}: Already in correct location")
                skipped_count += 1
                continue

            print(f"📄 Document {doc.id} ({file_type}):")
            print(f"   Device: {device.hostname} (ID: {device.id})")
            print(f"   From: {doc.stored_path}")
            print(f"   To:   {new_path}")

            if not dry_run:
                try:
                    # Create target directory
                    os.makedirs(os.path.dirname(new_path), exist_ok=True)

                    # Move file
                    shutil.move(doc.stored_path, new_path)

                    # Update database
                    doc.stored_path = new_path
                    db.add(doc)

                    print(f"   ✅ Migrated successfully")
                    migrated_count += 1

                except Exception as e:
                    print(f"   ❌ Error: {e}")
                    error_count += 1
            else:
                print(f"   [DRY RUN - would migrate]")
                migrated_count += 1

            print()

        if not dry_run:
            db.commit()
            print("Database updated")

        print("-" * 80)
        print(f"Summary:")
        print(f"  Migrated: {migrated_count}")
        print(f"  Skipped: {skipped_count}")
        print(f"  Errors: {error_count}")
        print(f"  Total: {len(documents)}")

        if dry_run:
            print()
            print("This was a DRY RUN. To actually migrate files, run:")
            print("  python backend/migrate_file_organization.py --execute")

    finally:
        db.close()


def cleanup_empty_directories(base_dir: str):
    """Remove empty directories after migration"""
    for root, dirs, files in os.walk(base_dir, topdown=False):
        for dir_name in dirs:
            dir_path = os.path.join(root, dir_name)
            try:
                if not os.listdir(dir_path):  # Directory is empty
                    os.rmdir(dir_path)
                    print(f"Removed empty directory: {dir_path}")
            except Exception as e:
                print(f"Could not remove {dir_path}: {e}")


if __name__ == "__main__":
    import argparse

    parser = argparse.ArgumentParser(description="Migrate uploads to organized structure")
    parser.add_argument(
        "--execute",
        action="store_true",
        help="Actually perform the migration (default is dry-run)"
    )
    parser.add_argument(
        "--cleanup",
        action="store_true",
        help="Clean up empty directories after migration"
    )

    args = parser.parse_args()

    print("=" * 80)
    print("File Organization Migration")
    print("=" * 80)
    print()

    migrate_files(dry_run=not args.execute)

    if args.execute and args.cleanup:
        print()
        print("=" * 80)
        print("Cleaning up empty directories")
        print("=" * 80)
        cleanup_empty_directories(settings.UPLOAD_DIR)

    print()
    print("Done!")