fix: 完善数据预处理与 JSON 上传链路

backend/tests/test_data_process_store.py

@@ -18,6 +18,7 @@ from app.modules.data_process.store import (
     _preview_config_changed,
     _reasoning_output_is_valid,
     _source_storage_descriptor,
+    repeat_task_id,
 )
 
 
@@ -331,6 +332,140 @@ class _TaskDetailStore(DataProcessStore):
         yield self._conn
 
 
+class _RepeatConnection:
+    def __init__(self) -> None:
+        self.source_files = [
+            {
+                "id": "source-old",
+                "name": "source.jsonl",
+                "size_bytes": 12,
+                "record_count": 1,
+                "file_format": "jsonl",
+                "checksum_sha256": "a" * 64,
+                "content": '{"id":1}\n',
+                "content_preview": '{"id":1}',
+                "metadata": {"storage_backend": "local"},
+                "created_by": "user-1",
+            }
+        ]
+        self.source_previews = [
+            {
+                "id": "preview-old",
+                "source_file_id": "source-old",
+                "original_content": '{"id":1}',
+                "edited_content": '{"id":1,"checked":true}',
+                "source_start": 0,
+                "source_end": 8,
+                "source_start_line": 1,
+                "source_end_line": 1,
+                "token_count": 5,
+                "status": "modified",
+                "quality_score": {"overall": 90},
+            }
+        ]
+        self.created_task: dict[str, Any] | None = None
+        self.created_files: list[dict[str, Any]] = []
+        self.created_previews: list[dict[str, Any]] = []
+
+    def execute(self, sql: str, params: Any = None) -> _Result:
+        normalized = " ".join(sql.split())
+        if params is not None:
+            assert normalized.count("%s") == len(params)
+        if normalized.startswith("SELECT * FROM data_process_tasks WHERE id="):
+            return _Result(row=None)
+        if normalized.startswith("SELECT * FROM data_process_source_files"):
+            return _Result(rows=[dict(item) for item in self.source_files])
+        if normalized.startswith("SELECT * FROM data_process_preview_items"):
+            return _Result(rows=[dict(item) for item in self.source_previews])
+        if normalized.startswith("INSERT INTO data_process_tasks"):
+            self.created_task = {
+                "id": params[0],
+                "name": params[1],
+                "description": params[2],
+                "status": "pending",
+                "process_type": params[3],
+                "source_dataset_id": params[4],
+                "config": params[5],
+                "progress": 20,
+                "input_count": params[6],
+                "results_confirmed": False,
+                "workflow_step": "preview",
+                "preview_status": "completed",
+                "preview_progress": 100,
+                "preview_total_files": params[7],
+                "preview_completed_files": params[8],
+                "created_at": params[15],
+                "updated_at": params[16],
+            }
+            return _Result(row=dict(self.created_task))
+        if normalized.startswith("INSERT INTO data_process_source_files"):
+            self.created_files.append(
+                {
+                    "id": params[0],
+                    "task_id": params[1],
+                    "storage_object_id": params[2],
+                    "content": params[8],
+                }
+            )
+            return _Result()
+        if normalized.startswith("INSERT INTO data_process_preview_items"):
+            self.created_previews.append(
+                {
+                    "id": params[0],
+                    "task_id": params[1],
+                    "source_file_id": params[2],
+                    "edited_content": params[4],
+                }
+            )
+            return _Result()
+        if normalized.startswith("SELECT (SELECT COUNT(*) FROM data_process_source_files"):
+            return _Result(
+                row={
+                    "source_file_count": len(self.created_files),
+                    "preview_count": len(self.created_previews),
+                }
+            )
+        raise AssertionError(f"unexpected SQL: {normalized}")
+
+
+class _RepeatStore(DataProcessStore):
+    def __init__(self, conn: _RepeatConnection) -> None:
+        self._conn = conn
+
+    @contextmanager
+    def connect(self) -> Iterator[_RepeatConnection]:
+        yield self._conn
+
+    def _task_in_connection(
+        self,
+        conn: Any,
+        task_id: str,
+        *,
+        for_update: bool = False,
+    ) -> dict[str, Any]:
+        assert task_id == "task-source"
+        assert for_update is True
+        return {
+            "id": task_id,
+            "name": "原任务",
+            "description": "原描述",
+            "status": "completed",
+            "process_type": "structured",
+            "source_dataset_id": None,
+            "config": {
+                "temperature": 0.3,
+                "_regeneration_prepared": {"prepared": True},
+            },
+            "results_confirmed": True,
+            "preview_status": "completed",
+            "tenant_id": "tenant-1",
+            "project_id": "project-1",
+            "owner_id": "owner-1",
+            "created_by": "user-1",
+            "updated_at": "2026-07-28T12:00:00Z",
+        }
+
+
 class _TaskListConnection:
     def __init__(self) -> None:
         self.task = {
@@ -574,6 +709,47 @@ def test_decode_row_serializes_postgres_numeric_values_as_json_numbers() -> None
     assert decoded == {"progress": 100.0, "duration_seconds": 389.0}
 
 
+def test_repeat_task_copies_business_snapshot_with_new_resource_ids() -> None:
+    conn = _RepeatConnection()
+    store = _RepeatStore(conn)
+    request_id = "repeat-request-0001"
+    target_task_id = repeat_task_id("task-source", request_id)
+
+    repeated = store.repeat_task(
+        "task-source",
+        expected_updated_at="2026-07-28T12:00:00Z",
+        request_id=request_id,
+        file_copies={
+            "source-old": {
+                "id": "source-new",
+                "storage_object_id": (
+                    f"local://data-process/{target_task_id}/source-new/v1/source.jsonl"
+                ),
+            }
+        },
+    )
+
+    assert repeated["created"] is True
+    assert repeated["task"]["id"] == target_task_id
+    assert repeated["task"]["config"] == {"temperature": 0.3}
+    assert repeated["task"]["results_confirmed"] is False
+    assert repeated["copied_source_file_count"] == 1
+    assert repeated["copied_preview_count"] == 1
+    assert conn.created_files == [
+        {
+            "id": "source-new",
+            "task_id": target_task_id,
+            "storage_object_id": (
+                f"local://data-process/{target_task_id}/source-new/v1/source.jsonl"
+            ),
+            "content": '{"id":1}\n',
+        }
+    ]
+    assert conn.created_previews[0]["task_id"] == target_task_id
+    assert conn.created_previews[0]["source_file_id"] == "source-new"
+    assert conn.created_previews[0]["edited_content"] == '{"id":1,"checked":true}'
+
+
 def test_decode_row_decodes_aggregated_output_datasets_json() -> None:
     decoded = _decode_row(
         {