DataMate/runtime/datamate-python/app/module/annotation/interface/project.py

from typing import Optional
import math
import uuid

from fastapi import APIRouter, Depends, HTTPException, Query, Path
from sqlalchemy.ext.asyncio import AsyncSession

from app.db.session import get_db
from app.db.models import LabelingProject
from app.module.shared.schema import StandardResponse, PaginatedData
from app.module.dataset import DatasetManagementService
from app.core.logging import get_logger

from ..service.mapping import DatasetMappingService
from ..service.template import AnnotationTemplateService
from ..schema import (
    DatasetMappingCreateRequest,
    DatasetMappingCreateResponse,
    DeleteDatasetResponse,
    DatasetMappingResponse,
)

router = APIRouter(
    prefix="/project",
    tags=["annotation/project"]
)
logger = get_logger(__name__)

@router.get("/{mapping_id}/login")
async def login_label_studio(
    mapping_id: str,
    db: AsyncSession = Depends(get_db)
):
    raise HTTPException(status_code=410, detail="当前为内嵌编辑器模式，不再支持 Label Studio 登录代理接口")

@router.post("", response_model=StandardResponse[DatasetMappingCreateResponse], status_code=201)
async def create_mapping(
    request: DatasetMappingCreateRequest,
    db: AsyncSession = Depends(get_db)
):
    """
    创建数据集映射

    在 DataMate 中创建标注项目（t_dm_labeling_projects），用于内嵌 Label Studio 编辑器。

    注意：一个数据集可以创建多个标注项目

    支持通过 template_id 指定标注模板，如果提供了模板ID，则使用模板的配置
    """
    try:
        dm_client = DatasetManagementService(db)
        mapping_service = DatasetMappingService(db)
        template_service = AnnotationTemplateService()

        logger.info(f"Create dataset mapping request: {request.dataset_id}")

        # 从DM服务获取数据集信息
        dataset_info = await dm_client.get_dataset(request.dataset_id)
        if not dataset_info:
            raise HTTPException(
                status_code=404,
                detail=f"Dataset not found in DM service: {request.dataset_id}"
            )

        project_name = request.name or \
                       dataset_info.name or \
                       "A new project from DataMate"

        project_description = request.description or \
                              dataset_info.description or \
                              f"Imported from DM dataset {dataset_info.name} ({dataset_info.id})"

        # 如果提供了模板ID，获取模板配置
        label_config = None
        if request.template_id:
            logger.info(f"Using template: {request.template_id}")
            template = await template_service.get_template(db, request.template_id)
            if not template:
                raise HTTPException(
                    status_code=404,
                    detail=f"Template not found: {request.template_id}"
                )
            label_config = template.label_config
            logger.debug(f"Template label config loaded for template: {template.name}")

        # 如果直接提供了 label_config (自定义或修改后的)，则覆盖模板配置
        if request.label_config:
            label_config = request.label_config
            logger.debug("Using custom label config from request")

        # DataMate-only：不再创建/依赖 Label Studio Server 项目。
        # 为兼容既有 schema 字段（labeling_project_id 长度 8），生成一个 8 位数字 ID。
        labeling_project_id = str(uuid.uuid4().int % 10**8).zfill(8)

        project_configuration = {}
        if label_config:
            project_configuration["label_config"] = label_config
        if project_description:
            project_configuration["description"] = project_description

        labeling_project = LabelingProject(
                id=str(uuid.uuid4()),  # Generate UUID here
                dataset_id=request.dataset_id,
                labeling_project_id=labeling_project_id,
                name=project_name,
                template_id=request.template_id,  # Save template_id to database
                configuration=project_configuration or None,
            )

        # 创建映射关系，包含项目名称（先持久化映射以获得 mapping.id）
        mapping = await mapping_service.create_mapping(labeling_project)

        response_data = DatasetMappingCreateResponse(
            id=mapping.id,
            labeling_project_id=str(mapping.labeling_project_id),
            labeling_project_name=mapping.name or project_name
        )

        return StandardResponse(
            code=201,
            message="success",
            data=response_data
        )

    except HTTPException:
        raise
    except Exception as e:
        logger.error(f"Error while creating dataset mapping: {e}")
        raise HTTPException(status_code=500, detail="Internal server error")

@router.get("", response_model=StandardResponse[PaginatedData[DatasetMappingResponse]])
async def list_mappings(
    page: int = Query(1, ge=1, description="页码（从1开始）"),
    page_size: int = Query(20, ge=1, le=100, description="每页记录数", alias="pageSize"),
    include_template: bool = Query(False, description="是否包含模板详情", alias="includeTemplate"),
    db: AsyncSession = Depends(get_db)
):
    """
    查询所有映射关系（分页）

    返回所有有效的数据集映射关系（未被软删除的），支持分页查询。
    可选择是否包含完整的标注模板信息（默认不包含，以提高列表查询性能）。

    参数:
    - page: 页码（从1开始）
    - pageSize: 每页记录数
    - includeTemplate: 是否包含模板详情（默认false）
    """
    try:
        service = DatasetMappingService(db)

        # 计算 skip
        skip = (page - 1) * page_size

        logger.info(f"List mappings: page={page}, page_size={page_size}, include_template={include_template}")

        # 获取数据和总数
        mappings, total = await service.get_all_mappings_with_count(
            skip=skip,
            limit=page_size,
            include_deleted=False,
            include_template=include_template
        )

        # 计算总页数
        total_pages = math.ceil(total / page_size) if total > 0 else 0

        # 构造分页响应
        paginated_data = PaginatedData(
            page=page,
            size=page_size,
            total_elements=total,
            total_pages=total_pages,
            content=mappings
        )

        logger.info(f"List mappings: page={page}, returned {len(mappings)}/{total}, templates_included: {include_template}")

        return StandardResponse(
            code=200,
            message="success",
            data=paginated_data
        )

    except Exception as e:
        logger.error(f"Error listing mappings: {e}")
        raise HTTPException(status_code=500, detail="Internal server error")

@router.get("/{mapping_id}", response_model=StandardResponse[DatasetMappingResponse])
async def get_mapping(
    mapping_id: str,
    db: AsyncSession = Depends(get_db)
):
    """
    根据 UUID 查询单个映射关系（包含关联的标注模板详情）

    返回数据集映射关系以及关联的完整标注模板信息，包括：
    - 映射基本信息
    - 数据集信息
    - Label Studio 项目信息
    - 完整的标注模板配置（如果存在）
    """
    try:
        service = DatasetMappingService(db)

        logger.info(f"Get mapping with template details: {mapping_id}")

        # 获取映射，并包含完整的模板信息
        mapping = await service.get_mapping_by_uuid(mapping_id, include_template=True)

        if not mapping:
            raise HTTPException(
                status_code=404,
                detail=f"Mapping not found: {mapping_id}"
            )

        logger.info(f"Found mapping: {mapping.id}, template_included: {mapping.template is not None}")

        return StandardResponse(
            code=200,
            message="success",
            data=mapping
        )

    except HTTPException:
        raise
    except Exception as e:
        logger.error(f"Error getting mapping: {e}")
        raise HTTPException(status_code=500, detail="Internal server error")

@router.get("/by-source/{dataset_id}", response_model=StandardResponse[PaginatedData[DatasetMappingResponse]])
async def get_mappings_by_source(
    dataset_id: str,
    page: int = Query(1, ge=1, description="页码（从1开始）"),
    page_size: int = Query(20, ge=1, le=100, description="每页记录数", alias="pageSize"),
    include_template: bool = Query(True, description="是否包含模板详情", alias="includeTemplate"),
    db: AsyncSession = Depends(get_db)
):
    """
    根据源数据集 ID 查询所有映射关系（分页，包含模板详情）

    返回该数据集创建的所有标注项目（不包括已删除的），支持分页查询。
    默认包含关联的完整标注模板信息。

    参数:
    - dataset_id: 数据集ID
    - page: 页码（从1开始）
    - pageSize: 每页记录数
    - includeTemplate: 是否包含模板详情（默认true）
    """
    try:
        service = DatasetMappingService(db)

        # 计算 skip
        skip = (page - 1) * page_size

        logger.info(f"Get mappings by source dataset id: {dataset_id}, page={page}, page_size={page_size}, include_template={include_template}")

        # 获取数据和总数（包含模板信息）
        mappings, total = await service.get_mappings_by_source_with_count(
            dataset_id=dataset_id,
            skip=skip,
            limit=page_size,
            include_template=include_template
        )

        # 计算总页数
        total_pages = math.ceil(total / page_size) if total > 0 else 0

        # 构造分页响应
        paginated_data = PaginatedData(
            page=page,
            size=page_size,
            total_elements=total,
            total_pages=total_pages,
            content=mappings
        )

        logger.info(f"Found {len(mappings)} mappings on page {page}, total: {total}, templates_included: {include_template}")

        return StandardResponse(
            code=200,
            message="success",
            data=paginated_data
        )

    except HTTPException:
        raise
    except Exception as e:
        logger.error(f"Error getting mappings: {e}")
        raise HTTPException(status_code=500, detail="Internal server error")

@router.delete("/{project_id}", response_model=StandardResponse[DeleteDatasetResponse])
async def delete_mapping(
    project_id: str = Path(..., description="映射UUID（path param）"),
    db: AsyncSession = Depends(get_db)
):
    """
    删除映射关系（软删除）

    通过 path 参数 `project_id` 指定要删除的映射（映射的 UUID）。

    内嵌编辑器模式下仅软删除 DataMate 标注项目记录，不再删除/依赖 Label Studio Server 项目。
    """
    try:
        logger.debug(f"Delete mapping request received: project_id={project_id!r}")

        service = DatasetMappingService(db)

        # 使用 mapping UUID 查询映射记录
        logger.debug(f"Deleting by mapping UUID: {project_id}")
        mapping = await service.get_mapping_by_uuid(project_id)

        logger.debug(f"Mapping lookup result: {mapping}")

        if not mapping:
            raise HTTPException(
                status_code=404,
                detail=f"Mapping either not found or not specified."
            )

        id = mapping.id
        logger.debug(f"Found mapping: {id}")

        # 软删除映射记录
        soft_delete_success = await service.soft_delete_mapping(id)
        logger.debug(f"Soft delete result for mapping {id}: {soft_delete_success}")

        if not soft_delete_success:
            raise HTTPException(
                status_code=500,
                detail="Failed to delete mapping record"
            )

        logger.info(f"Successfully deleted mapping: {id}")

        return StandardResponse(
            code=200,
            message="success",
            data=DeleteDatasetResponse(
                id=id,
                status="success"
            )
        )

    except HTTPException:
        raise
    except Exception as e:
        logger.error(f"Error deleting mapping: {e}")
        raise HTTPException(status_code=500, detail="Internal server error")