geMoldInsight/src/moldinsight/services/file_history_service.py

# services/file_history_service.py
"""文件历史查询视图——按文件名分组的多版本上传历史。

批次 3 自 storage_integration_rustfs.py 按职责拆分（见 task_storage_service.py 头注）。
"""
from typing import Optional

from sqlalchemy import select, func, desc
from sqlalchemy.ext.asyncio import AsyncSession
from sqlalchemy.orm import joinedload

from moldinsight.models import STPFile, ProcessingTask
from shared.utils.logger import get_logger

logger = get_logger(__name__)


class FileHistoryService:
    """按文件名聚合的上传历史查询"""

    async def get_file_history_by_filename(
        self,
        session: AsyncSession,
        filename: str,
        user_id: Optional[int] = None,
        limit: int = 50
    ) -> list:
        """获取同一文件名的所有上传历史记录"""

        query = select(STPFile).options(
            joinedload(STPFile.processing_tasks)
        ).where(
            STPFile.original_filename == filename
        ).order_by(STPFile.upload_time.desc())

        if user_id:
            query = query.where(STPFile.user_id == user_id)

        query = query.limit(limit)

        result = await session.execute(query)
        files = result.unique().scalars().all()

        return [
            {
                'id': f.id,
                'task_id': f.processing_tasks[0].task_id if f.processing_tasks else None,
                'upload_batch': f.upload_batch,
                'upload_time': f.upload_time.strftime('%Y-%m-%d %H:%M:%S') if f.upload_time else None,
                'file_size': f.file_size,
                'status': f.status,
                'volume': f.volume,
                'surface_area': f.surface_area,
                'product_weight': f.product_weight,
                'has_analysis': f.status == 'completed'
            }
            for f in files
        ]

    async def get_all_file_groups(
        self,
        session: AsyncSession,
        user_id: Optional[int] = None,
        limit: int = 100
    ) -> list:
        """获取所有文件分组（按文件名分组），包含每个文件的最新分析结果"""

        # 子查询：获取每个文件名的最新上传
        subquery = (
            select(
                STPFile.original_filename,
                func.max(STPFile.upload_time).label('latest_upload')
            )
            .group_by(STPFile.original_filename)
            .order_by(desc('latest_upload'))
            .limit(limit)
        )

        if user_id:
            subquery = subquery.where(STPFile.user_id == user_id)

        subquery = subquery.subquery()

        # 主查询：获取最新记录和统计信息
        query = (
            select(STPFile).options(
                joinedload(STPFile.processing_tasks)
            )
            .join(
                subquery,
                (STPFile.original_filename == subquery.c.original_filename) &
                (STPFile.upload_time == subquery.c.latest_upload)
            )
            .order_by(STPFile.upload_time.desc())
        )

        result = await session.execute(query)
        latest_files = result.unique().scalars().all()

        # 一次性聚合每个文件名的上传次数（替代逐文件 count 的 N+1 查询）
        count_subquery = (
            select(STPFile.original_filename, func.count().label("upload_count"))
            .group_by(STPFile.original_filename)
        )
        if user_id:
            count_subquery = count_subquery.where(STPFile.user_id == user_id)
        count_result = await session.execute(count_subquery)
        upload_counts = {
            row.original_filename: row.upload_count for row in count_result
        }

        # 获取每个文件名的上传次数
        file_groups = []
        for f in latest_files:
            task_id = f.processing_tasks[0].task_id if f.processing_tasks else None
            upload_count = upload_counts.get(f.original_filename, 1)

            file_groups.append({
                'filename': f.original_filename,
                'latest_id': f.id,
                'latest_task_id': task_id,
                'latest_upload_time': f.upload_time.strftime('%Y-%m-%d %H:%M:%S') if f.upload_time else None,
                'latest_status': f.status,
                'upload_count': upload_count,
                'file_size': f.file_size,
                'volume': f.volume,
                'surface_area': f.surface_area,
                'product_weight': f.product_weight
            })

        return file_groups