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

"""
Label Studio Editor（前端嵌入式）接口

说明：
- 不依赖 Label Studio Server；仅复用其“编辑器”前端库
- DataMate 负责提供 tasks/annotations 数据与保存能力
- 当前支持 dataset_type=TEXT/IMAGE 的项目
"""

from __future__ import annotations

from typing import Optional

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

from app.core.logging import get_logger
from app.db.session import get_db
from app.module.annotation.schema.editor import (
    EditorProjectInfo,
    EditorTaskListResponse,
    EditorTaskSegmentResponse,
    EditorTaskResponse,
    FileVersionCheckResponse,
    UseNewVersionResponse,
    UpsertAnnotationRequest,
    UpsertAnnotationResponse,
)
from app.module.annotation.service.editor import AnnotationEditorService
from app.module.annotation.security import (
    RequestUserContext,
    get_request_user_context,
)
from app.module.shared.schema import StandardResponse

logger = get_logger(__name__)

router = APIRouter(
    prefix="/editor",
    tags=["annotation/editor"],
)


@router.get(
    "/projects/{project_id}",
    response_model=StandardResponse[EditorProjectInfo],
)
async def get_editor_project_info(
    project_id: str = Path(..., description="标注项目ID（t_dm_labeling_projects.id）"),
    db: AsyncSession = Depends(get_db),
    user_context: RequestUserContext = Depends(get_request_user_context),
):
    service = AnnotationEditorService(db, user_context)
    info = await service.get_project_info(project_id)
    return StandardResponse(code=200, message="success", data=info)


@router.get(
    "/projects/{project_id}/tasks",
    response_model=StandardResponse[EditorTaskListResponse],
)
async def list_editor_tasks(
    project_id: str = Path(..., description="标注项目ID（t_dm_labeling_projects.id）"),
    page: int = Query(0, ge=0, description="页码（从0开始）"),
    size: int = Query(50, ge=1, le=200, description="每页大小"),
    exclude_source_documents: Optional[bool] = Query(
        None,
        alias="excludeSourceDocuments",
        description="是否排除已被转换为TXT的源文档文件（PDF/DOC/DOCX，仅文本数据集生效）",
    ),
    db: AsyncSession = Depends(get_db),
    user_context: RequestUserContext = Depends(get_request_user_context),
):
    service = AnnotationEditorService(db, user_context)
    result = await service.list_tasks(
        project_id,
        page=page,
        size=size,
        exclude_source_documents=exclude_source_documents,
    )
    return StandardResponse(code=200, message="success", data=result)


@router.get(
    "/projects/{project_id}/tasks/{file_id}",
    response_model=StandardResponse[EditorTaskResponse],
)
async def get_editor_task(
    project_id: str = Path(..., description="标注项目ID（t_dm_labeling_projects.id）"),
    file_id: str = Path(..., description="文件ID（t_dm_dataset_files.id）"),
    segment_index: Optional[int] = Query(
        None, alias="segmentIndex", description="段落索引（分段模式下使用）"
    ),
    db: AsyncSession = Depends(get_db),
    user_context: RequestUserContext = Depends(get_request_user_context),
):
    service = AnnotationEditorService(db, user_context)
    task = await service.get_task(project_id, file_id, segment_index=segment_index)
    return StandardResponse(code=200, message="success", data=task)


@router.get(
    "/projects/{project_id}/tasks/{file_id}/segments",
    response_model=StandardResponse[EditorTaskSegmentResponse],
)
async def get_editor_task_segment(
    project_id: str = Path(..., description="标注项目ID（t_dm_labeling_projects.id）"),
    file_id: str = Path(..., description="文件ID（t_dm_dataset_files.id）"),
    segment_index: int = Query(
        ..., ge=0, alias="segmentIndex", description="段落索引（从0开始）"
    ),
    db: AsyncSession = Depends(get_db),
    user_context: RequestUserContext = Depends(get_request_user_context),
):
    service = AnnotationEditorService(db, user_context)
    result = await service.get_task_segment(project_id, file_id, segment_index)
    return StandardResponse(code=200, message="success", data=result)


@router.put(
    "/projects/{project_id}/tasks/{file_id}/annotation",
    response_model=StandardResponse[UpsertAnnotationResponse],
)
async def upsert_editor_annotation(
    request: UpsertAnnotationRequest,
    project_id: str = Path(..., description="标注项目ID（t_dm_labeling_projects.id）"),
    file_id: str = Path(..., description="文件ID（t_dm_dataset_files.id）"),
    db: AsyncSession = Depends(get_db),
    user_context: RequestUserContext = Depends(get_request_user_context),
):
    service = AnnotationEditorService(db, user_context)
    result = await service.upsert_annotation(project_id, file_id, request)
    return StandardResponse(code=200, message="success", data=result)


@router.get(
    "/projects/{project_id}/files/{file_id}/version",
    response_model=StandardResponse[FileVersionCheckResponse],
)
async def check_file_version(
    project_id: str = Path(..., description="标注项目ID（t_dm_labeling_projects.id）"),
    file_id: str = Path(..., description="文件ID（t_dm_dataset_files.id）"),
    db: AsyncSession = Depends(get_db),
    user_context: RequestUserContext = Depends(get_request_user_context),
):
    """
    检查文件是否有新版本
    """
    service = AnnotationEditorService(db, user_context)
    result = await service.check_file_version(project_id, file_id)
    return StandardResponse(code=200, message="success", data=result)


@router.post(
    "/projects/{project_id}/files/{file_id}/use-new-version",
    response_model=StandardResponse[UseNewVersionResponse],
)
async def use_new_version(
    project_id: str = Path(..., description="标注项目ID（t_dm_labeling_projects.id）"),
    file_id: str = Path(..., description="文件ID（t_dm_dataset_files.id）"),
    db: AsyncSession = Depends(get_db),
    user_context: RequestUserContext = Depends(get_request_user_context),
):
    """
    使用文件新版本并清空标注
    """
    service = AnnotationEditorService(db, user_context)
    result = await service.use_new_version(project_id, file_id)
    return StandardResponse(code=200, message="success", data=result)