feat(data-management): 实现数据集文件版本管理和内部路径保护

- 将数据集文件查询方法替换为只查询可见文件的版本
- 引入文件状态管理（ACTIVE/ARCHIVED）和内部目录结构
- 实现文件重复处理策略，支持版本控制模式而非覆盖
- 添加内部数据目录保护，防止访问.datamate等系统目录
- 重构文件上传流程，引入暂存目录和事务后清理机制
- 实现文件版本归档功能，保留历史版本到专用存储位置
- 优化文件路径规范化和安全验证逻辑
- 修复文件删除逻辑，确保归档文件不会被错误移除
- 更新数据集压缩下载功能以排除内部系统文件

runtime/python-executor/datamate/auto_annotation_worker.py

@@ -372,15 +372,15 @@ def _register_output_dataset(
         )
         return
 
-    insert_file_sql = text(
-        """
-        INSERT INTO t_dm_dataset_files (
-            id, dataset_id, file_name, file_path, file_type, file_size, status
-        ) VALUES (
-            :id, :dataset_id, :file_name, :file_path, :file_type, :file_size, :status
-        )
-        """
-    )
+    insert_file_sql = text(
+        """
+        INSERT INTO t_dm_dataset_files (
+            id, dataset_id, file_name, file_path, logical_path, version, file_type, file_size, status
+        ) VALUES (
+            :id, :dataset_id, :file_name, :file_path, :logical_path, :version, :file_type, :file_size, :status
+        )
+        """
+    )
     update_dataset_stat_sql = text(
         """
         UPDATE t_dm_datasets
@@ -393,37 +393,43 @@ def _register_output_dataset(
     with SQLManager.create_connect() as conn:
         added_count = 0
 
-        for file_name, file_path, file_size in image_files:
-            ext = os.path.splitext(file_name)[1].lstrip(".").upper() or None
-            conn.execute(
-                insert_file_sql,
-                {
-                    "id": str(uuid.uuid4()),
-                    "dataset_id": output_dataset_id,
-                    "file_name": file_name,
-                    "file_path": file_path,
-                    "file_type": ext,
-                    "file_size": int(file_size),
-                    "status": "ACTIVE",
-                },
-            )
-            added_count += 1
-
-        for file_name, file_path, file_size in annotation_files:
-            ext = os.path.splitext(file_name)[1].lstrip(".").upper() or None
-            conn.execute(
-                insert_file_sql,
-                {
-                    "id": str(uuid.uuid4()),
-                    "dataset_id": output_dataset_id,
-                    "file_name": file_name,
-                    "file_path": file_path,
-                    "file_type": ext,
-                    "file_size": int(file_size),
-                    "status": "ACTIVE",
-                },
-            )
-            added_count += 1
+        for file_name, file_path, file_size in image_files:
+            ext = os.path.splitext(file_name)[1].lstrip(".").upper() or None
+            logical_path = os.path.relpath(file_path, output_dir).replace("\\", "/")
+            conn.execute(
+                insert_file_sql,
+                {
+                    "id": str(uuid.uuid4()),
+                    "dataset_id": output_dataset_id,
+                    "file_name": file_name,
+                    "file_path": file_path,
+                    "logical_path": logical_path,
+                    "version": 1,
+                    "file_type": ext,
+                    "file_size": int(file_size),
+                    "status": "ACTIVE",
+                },
+            )
+            added_count += 1
+
+        for file_name, file_path, file_size in annotation_files:
+            ext = os.path.splitext(file_name)[1].lstrip(".").upper() or None
+            logical_path = os.path.relpath(file_path, output_dir).replace("\\", "/")
+            conn.execute(
+                insert_file_sql,
+                {
+                    "id": str(uuid.uuid4()),
+                    "dataset_id": output_dataset_id,
+                    "file_name": file_name,
+                    "file_path": file_path,
+                    "logical_path": logical_path,
+                    "version": 1,
+                    "file_type": ext,
+                    "file_size": int(file_size),
+                    "status": "ACTIVE",
+                },
+            )
+            added_count += 1
 
         if added_count > 0:
             conn.execute(