feat(data-management): 添加数据集父子层级结构功能

- 在OpenAPI规范中新增parentDatasetId字段用于层级过滤
- 实现数据集父子关系的创建、更新和删除逻辑
- 添加数据集移动时的路径重命名和文件路径前缀更新
- 增加子数据集数量验证防止误删父数据集
- 更新前端界面支持选择父数据集和导航显示
- 优化Python后端自动标注任务的路径处理逻辑
- 修改数据库表结构添加外键约束确保数据一致性

runtime/datamate-python/app/db/models/dataset_management.py

@@ -14,6 +14,7 @@ class Dataset(Base):
     __tablename__ = "t_dm_datasets"
     
     id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()), comment="UUID")
+    parent_dataset_id = Column(String(36), nullable=True, comment="父数据集ID（UUID）")
     name = Column(String(255), nullable=False, comment="数据集名称")
     description = Column(Text, nullable=True, comment="数据集描述")
     dataset_type = Column(String(50), nullable=False, comment="数据集类型：IMAGE/TEXT/QA/MULTIMODAL/OTHER")

runtime/python-executor/datamate/auto_annotation_worker.py

@@ -195,8 +195,8 @@ def _update_task_status(
         conn.execute(sql, params)
 
 
-def _load_dataset_files(dataset_id: str) -> List[Tuple[str, str, str]]:
-    """加载指定数据集下的所有已完成文件。"""
+def _load_dataset_files(dataset_id: str) -> List[Tuple[str, str, str]]:
+    """加载指定数据集下的所有已完成文件。"""
 
     sql = text(
         """
@@ -208,9 +208,45 @@ def _load_dataset_files(dataset_id: str) -> List[Tuple[str, str, str]]:
         """
     )
 
-    with SQLManager.create_connect() as conn:
-        rows = conn.execute(sql, {"dataset_id": dataset_id}).fetchall()
-        return [(str(r[0]), str(r[1]), str(r[2])) for r in rows]
+    with SQLManager.create_connect() as conn:
+        rows = conn.execute(sql, {"dataset_id": dataset_id}).fetchall()
+        return [(str(r[0]), str(r[1]), str(r[2])) for r in rows]
+
+
+def _load_dataset_meta(dataset_id: str) -> Optional[Dict[str, Any]]:
+    """加载数据集基础信息（含父ID与路径）。"""
+
+    sql = text(
+        """
+        SELECT id, name, parent_dataset_id, path
+        FROM t_dm_datasets
+        WHERE id = :dataset_id
+        """
+    )
+    with SQLManager.create_connect() as conn:
+        row = conn.execute(sql, {"dataset_id": dataset_id}).fetchone()
+        if not row:
+            return None
+        return dict(row._mapping)  # type: ignore[attr-defined]
+
+
+def _resolve_output_parent(source_dataset_id: str) -> Tuple[Optional[str], str]:
+    """根据源数据集确定产出数据集的父级与基路径（产出挂在父级下）。"""
+
+    base_path = DEFAULT_OUTPUT_ROOT.rstrip("/") or "/dataset"
+    source_meta = _load_dataset_meta(source_dataset_id)
+    if not source_meta:
+        return None, base_path
+
+    parent_dataset_id = source_meta.get("parent_dataset_id")
+    if not parent_dataset_id:
+        return None, base_path
+
+    parent_meta = _load_dataset_meta(str(parent_dataset_id))
+    parent_path = parent_meta.get("path") if parent_meta else None
+    if not parent_path:
+        return None, base_path
+    return str(parent_dataset_id), str(parent_path)
 
 
 def _load_files_by_ids(file_ids: List[str]) -> List[Tuple[str, str, str]]:
@@ -245,34 +281,35 @@ def _ensure_output_dir(output_dir: str) -> str:
     return output_dir
 
 
-def _create_output_dataset(
-    source_dataset_id: str,
-    source_dataset_name: str,
-    output_dataset_name: str,
-) -> Tuple[str, str]:
-    """为自动标注结果创建一个新的数据集并返回 (dataset_id, path)。"""
-
-    new_dataset_id = str(uuid.uuid4())
-    dataset_base_path = DEFAULT_OUTPUT_ROOT.rstrip("/") or "/dataset"
-    output_dir = os.path.join(dataset_base_path, new_dataset_id)
+def _create_output_dataset(
+    source_dataset_id: str,
+    source_dataset_name: str,
+    output_dataset_name: str,
+) -> Tuple[str, str]:
+    """为自动标注结果创建一个新的数据集并返回 (dataset_id, path)。"""
+
+    new_dataset_id = str(uuid.uuid4())
+    parent_dataset_id, dataset_base_path = _resolve_output_parent(source_dataset_id)
+    output_dir = os.path.join(dataset_base_path, new_dataset_id)
 
     description = (
         f"Auto annotations for dataset {source_dataset_name or source_dataset_id}"[:255]
     )
 
-    sql = text(
-        """
-        INSERT INTO t_dm_datasets (id, name, description, dataset_type, path, status)
-        VALUES (:id, :name, :description, :dataset_type, :path, :status)
-        """
-    )
-    params = {
-        "id": new_dataset_id,
-        "name": output_dataset_name,
-        "description": description,
-        "dataset_type": "IMAGE",
-        "path": output_dir,
-        "status": "ACTIVE",
+    sql = text(
+        """
+        INSERT INTO t_dm_datasets (id, parent_dataset_id, name, description, dataset_type, path, status)
+        VALUES (:id, :parent_dataset_id, :name, :description, :dataset_type, :path, :status)
+        """
+    )
+    params = {
+        "id": new_dataset_id,
+        "parent_dataset_id": parent_dataset_id,
+        "name": output_dataset_name,
+        "description": description,
+        "dataset_type": "IMAGE",
+        "path": output_dir,
+        "status": "ACTIVE",
     }
 
     with SQLManager.create_connect() as conn: