Skip to content

armgit / dify

Public Imported from GitHub
0

Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.

test: migrate RAG residual entities to real ORM models (#40629)

Asuka Minatoa958c2cadbc9
diff --git a/api/tests/unit_tests/core/rag/extractor/test_extract_processor.py b/api/tests/unit_tests/core/rag/extractor/test_extract_processor.py
index 0b73d488fa0..369e63e57e3 100644
--- a/api/tests/unit_tests/core/rag/extractor/test_extract_processor.py
+++ b/api/tests/unit_tests/core/rag/extractor/test_extract_processor.py
@@ -1,12 +1,35 @@
+from datetime import datetime
from pathlib import Path
from types import SimpleNamespace
import pytest
+from sqlalchemy.orm import Session
import core.rag.extractor.extract_processor as processor_module
from core.rag.extractor.entity.datasource_type import DatasourceType
from core.rag.extractor.extract_processor import ExtractProcessor
from core.rag.models.document import Document
+from extensions.storage.storage_type import StorageType
+from models.enums import CreatorUserRole
+from models.model import UploadFile
+
+
+def _upload_file(*, key: str, file_id: str = "upload-file-1") -> UploadFile:
+ upload_file = UploadFile(
+ tenant_id="tenant-1",
+ storage_type=StorageType.LOCAL,
+ key=key,
+ name=Path(key).name,
+ size=1,
+ extension=Path(key).suffix.lstrip("."),
+ mime_type="application/octet-stream",
+ created_by_role=CreatorUserRole.ACCOUNT,
+ created_by="user-1",
+ created_at=datetime(2025, 1, 1),
+ used=True,
+ )
+ upload_file.id = file_id
+ return upload_file
class _ExtractorFactory:
@@ -68,7 +91,7 @@ class TestExtractProcessorLoaders:
],
)
- upload_file = SimpleNamespace(key="file.txt")
+ upload_file = _upload_file(key="file.txt")
docs = ExtractProcessor.load_from_upload_file(upload_file=upload_file, return_text=False)
text = ExtractProcessor.load_from_upload_file(upload_file=upload_file, return_text=True)
@@ -154,12 +177,7 @@ class TestExtractProcessorFileRouting:
setting = SimpleNamespace(
datasource_type=DatasourceType.FILE,
- upload_file=SimpleNamespace(
- id="upload-file-1",
- key=f"uploaded{extension}",
- tenant_id="tenant-1",
- created_by="user-1",
- ),
+ upload_file=_upload_file(key=f"uploaded{extension}"),
)
docs = ExtractProcessor.extract(setting, is_automatic=is_automatic, session=session)
@@ -229,13 +247,13 @@ class TestExtractProcessorFileRouting:
@pytest.mark.parametrize("extension", [".pdf", ".docx"])
def test_extract_passes_session_to_database_backed_file_extractors(
- self, monkeypatch: pytest.MonkeyPatch, extension: str
+ self, monkeypatch: pytest.MonkeyPatch, extension: str, unbound_session: Session
):
- session = object()
-
- _, _, kwargs = self._run_extract_for_extension(monkeypatch, extension, etl_type="SelfHosted", session=session)
+ _, _, kwargs = self._run_extract_for_extension(
+ monkeypatch, extension, etl_type="SelfHosted", session=unbound_session
+ )
- assert kwargs["session"] is session
+ assert kwargs["session"] is unbound_session
def test_extract_requires_upload_file_when_file_path_not_provided(self):
setting = SimpleNamespace(datasource_type=DatasourceType.FILE, upload_file=None)
diff --git a/api/tests/unit_tests/core/rag/retrieval/test_dataset_retrieval_attachment_entry.py b/api/tests/unit_tests/core/rag/retrieval/test_dataset_retrieval_attachment_entry.py
index 1e41827d316..a5a91a01545 100644
--- a/api/tests/unit_tests/core/rag/retrieval/test_dataset_retrieval_attachment_entry.py
+++ b/api/tests/unit_tests/core/rag/retrieval/test_dataset_retrieval_attachment_entry.py
@@ -1,14 +1,22 @@
"""Focused tests for attachment-aware dataset retrieval entry behavior."""
-from unittest.mock import MagicMock, patch
+from unittest.mock import patch
+
+from sqlalchemy.orm import Session
from core.rag.retrieval.dataset_retrieval import DatasetRetrieval
from core.workflow.nodes.knowledge_retrieval.retrieval import KnowledgeRetrievalRequest
+from models.dataset import Dataset
-def test_knowledge_retrieval_allows_attachment_only_requests() -> None:
+def test_knowledge_retrieval_allows_attachment_only_requests(unbound_session: Session) -> None:
retrieval = DatasetRetrieval()
- available_dataset = MagicMock(id="dataset-1")
+ available_dataset = Dataset(
+ id="dataset-1",
+ tenant_id="tenant-1",
+ name="Dataset",
+ created_by="user-1",
+ )
request = KnowledgeRetrievalRequest(
tenant_id="tenant-1",
@@ -30,7 +38,7 @@ def test_knowledge_retrieval_allows_attachment_only_requests() -> None:
patch.object(retrieval, "_get_available_datasets", return_value=[available_dataset]),
patch.object(retrieval, "multiple_retrieve", return_value=[]) as mock_multiple,
):
- result = retrieval.knowledge_retrieval(MagicMock(), request)
+ result = retrieval.knowledge_retrieval(unbound_session, request)
assert result == []
mock_multiple.assert_called_once()