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.

refactor(api): enable strict typing for RAG extractor, embedding and retrieval tests (#43913)

Eddy ZHANG2e8b23c47b55
diff --git a/api/tests/unit_tests/core/rag/embedding/test_embedding_base.py b/api/tests/unit_tests/core/rag/embedding/test_embedding_base.py
index 493a6f5374f..83ddc10739c 100644
--- a/api/tests/unit_tests/core/rag/embedding/test_embedding_base.py
+++ b/api/tests/unit_tests/core/rag/embedding/test_embedding_base.py
@@ -29,81 +29,101 @@ class ConcreteEmbeddings(Embeddings):
return [1.0] * 10
+class _SuperCallingEmbeddings(Embeddings):
+ """Overrides each abstract method by delegating to the Embeddings base implementation."""
+
+ @override
+ def embed_documents(self, texts: list[str]) -> list[list[float]]:
+ return super().embed_documents(texts)
+
+ @override
+ def embed_multimodal_documents(self, multimodel_documents: list[dict[str, Any]]) -> list[list[float]]:
+ return super().embed_multimodal_documents(multimodel_documents)
+
+ @override
+ def embed_query(self, text: str) -> list[float]:
+ return super().embed_query(text)
+
+ @override
+ def embed_multimodal_query(self, multimodel_document: dict[str, Any]) -> list[float]:
+ return super().embed_multimodal_query(multimodel_document)
+
+
class TestEmbeddingsBase:
"""Test suite for the abstract Embeddings base class."""
- def test_embeddings_is_abc(self):
+ def test_embeddings_is_abc(self) -> None:
"""Test that Embeddings is an abstract base class."""
assert hasattr(Embeddings, "__abstractmethods__")
assert len(Embeddings.__abstractmethods__) > 0
- def test_embed_documents_is_abstract(self):
+ def test_embed_documents_is_abstract(self) -> None:
"""Test that embed_documents is an abstract method."""
assert "embed_documents" in Embeddings.__abstractmethods__
- def test_embed_multimodal_documents_is_abstract(self):
+ def test_embed_multimodal_documents_is_abstract(self) -> None:
"""Test that embed_multimodal_documents is an abstract method."""
assert "embed_multimodal_documents" in Embeddings.__abstractmethods__
- def test_embed_query_is_abstract(self):
+ def test_embed_query_is_abstract(self) -> None:
"""Test that embed_query is an abstract method."""
assert "embed_query" in Embeddings.__abstractmethods__
- def test_embed_multimodal_query_is_abstract(self):
+ def test_embed_multimodal_query_is_abstract(self) -> None:
"""Test that embed_multimodal_query is an abstract method."""
assert "embed_multimodal_query" in Embeddings.__abstractmethods__
- def test_embed_documents_raises_not_implemented(self):
+ def test_embed_documents_raises_not_implemented(self) -> None:
"""Test that embed_documents raises NotImplementedError in its body."""
source = inspect.getsource(Embeddings.embed_documents)
assert "raise NotImplementedError" in source
- def test_embed_multimodal_documents_raises_not_implemented(self):
+ def test_embed_multimodal_documents_raises_not_implemented(self) -> None:
"""Test that embed_multimodal_documents raises NotImplementedError in its body."""
source = inspect.getsource(Embeddings.embed_multimodal_documents)
assert "raise NotImplementedError" in source
- def test_embed_query_raises_not_implemented(self):
+ def test_embed_query_raises_not_implemented(self) -> None:
"""Test that embed_query raises NotImplementedError in its body."""
source = inspect.getsource(Embeddings.embed_query)
assert "raise NotImplementedError" in source
- def test_embed_multimodal_query_raises_not_implemented(self):
+ def test_embed_multimodal_query_raises_not_implemented(self) -> None:
"""Test that embed_multimodal_query raises NotImplementedError in its body."""
source = inspect.getsource(Embeddings.embed_multimodal_query)
assert "raise NotImplementedError" in source
- def test_aembed_documents_raises_not_implemented(self):
+ def test_aembed_documents_raises_not_implemented(self) -> None:
"""Test that aembed_documents raises NotImplementedError in its body."""
source = inspect.getsource(Embeddings.aembed_documents)
assert "raise NotImplementedError" in source
- def test_aembed_query_raises_not_implemented(self):
+ def test_aembed_query_raises_not_implemented(self) -> None:
"""Test that aembed_query raises NotImplementedError in its body."""
source = inspect.getsource(Embeddings.aembed_query)
assert "raise NotImplementedError" in source
- def test_concrete_implementation_works(self):
+ def test_concrete_implementation_works(self) -> None:
"""Test that a concrete implementation of Embeddings works correctly."""
concrete = ConcreteEmbeddings()
result = concrete.embed_documents(["test1", "test2"])
assert len(result) == 2
assert all(len(emb) == 10 for emb in result)
- def test_concrete_implementation_embed_query(self):
+ def test_concrete_implementation_embed_query(self) -> None:
"""Test concrete implementation of embed_query."""
concrete = ConcreteEmbeddings()
result = concrete.embed_query("test query")
assert len(result) == 10
- def test_concrete_implementation_embed_multimodal_documents(self):
+ def test_concrete_implementation_embed_multimodal_documents(self) -> None:
"""Test concrete implementation of embed_multimodal_documents."""
concrete = ConcreteEmbeddings()
docs: list[dict[str, Any]] = [{"file_id": "file1"}, {"file_id": "file2"}]
result = concrete.embed_multimodal_documents(docs)
assert len(result) == 2
- def test_concrete_implementation_embed_multimodal_query(self):
+ def test_concrete_implementation_embed_multimodal_query(self) -> None:
"""Test concrete implementation of embed_multimodal_query."""
concrete = ConcreteEmbeddings()
result = concrete.embed_multimodal_query({"file_id": "test"})
@@ -113,111 +133,51 @@ class TestEmbeddingsBase:
class TestEmbeddingsNotImplemented:
"""Test that abstract methods raise NotImplementedError when called."""
- def test_embed_query_raises_not_implemented(self):
+ def test_embed_query_raises_not_implemented(self) -> None:
"""Test that embed_query raises NotImplementedError."""
- class PartialImpl:
- pass
-
- PartialImpl.embed_query = lambda self, text: Embeddings.embed_query(self, text)
- PartialImpl.embed_documents = lambda self, texts: Embeddings.embed_documents(self, texts)
- PartialImpl.embed_multimodal_documents = lambda self, docs: Embeddings.embed_multimodal_documents(self, docs)
- PartialImpl.embed_multimodal_query = lambda self, doc: Embeddings.embed_multimodal_query(self, doc)
- PartialImpl.aembed_documents = lambda self, texts: Embeddings.aembed_documents(self, texts)
- PartialImpl.aembed_query = lambda self, text: Embeddings.aembed_query(self, text)
-
- partial = PartialImpl()
+ partial = _SuperCallingEmbeddings()
with pytest.raises(NotImplementedError):
partial.embed_query("test")
- def test_embed_documents_raises_not_implemented(self):
+ def test_embed_documents_raises_not_implemented(self) -> None:
"""Test that embed_documents raises NotImplementedError."""
- class PartialImpl:
- pass
-
- PartialImpl.embed_query = lambda self, text: Embeddings.embed_query(self, text)
- PartialImpl.embed_documents = lambda self, texts: Embeddings.embed_documents(self, texts)
- PartialImpl.embed_multimodal_documents = lambda self, docs: Embeddings.embed_multimodal_documents(self, docs)
- PartialImpl.embed_multimodal_query = lambda self, doc: Embeddings.embed_multimodal_query(self, doc)
- PartialImpl.aembed_documents = lambda self, texts: Embeddings.aembed_documents(self, texts)
- PartialImpl.aembed_query = lambda self, text: Embeddings.aembed_query(self, text)
-
- partial = PartialImpl()
+ partial = _SuperCallingEmbeddings()
with pytest.raises(NotImplementedError):
partial.embed_documents(["test"])
- def test_embed_multimodal_documents_raises_not_implemented(self):
+ def test_embed_multimodal_documents_raises_not_implemented(self) -> None:
"""Test that embed_multimodal_documents raises NotImplementedError."""
- class PartialImpl:
- pass
-
- PartialImpl.embed_query = lambda self, text: Embeddings.embed_query(self, text)
- PartialImpl.embed_documents = lambda self, texts: Embeddings.embed_documents(self, texts)
- PartialImpl.embed_multimodal_documents = lambda self, docs: Embeddings.embed_multimodal_documents(self, docs)
- PartialImpl.embed_multimodal_query = lambda self, doc: Embeddings.embed_multimodal_query(self, doc)
- PartialImpl.aembed_documents = lambda self, texts: Embeddings.aembed_documents(self, texts)
- PartialImpl.aembed_query = lambda self, text: Embeddings.aembed_query(self, text)
-
- partial = PartialImpl()
+ partial = _SuperCallingEmbeddings()
with pytest.raises(NotImplementedError):
partial.embed_multimodal_documents([{"file_id": "test"}])
- def test_embed_multimodal_query_raises_not_implemented(self):
+ def test_embed_multimodal_query_raises_not_implemented(self) -> None:
"""Test that embed_multimodal_query raises NotImplementedError."""
- class PartialImpl:
- pass
-
- PartialImpl.embed_query = lambda self, text: Embeddings.embed_query(self, text)
- PartialImpl.embed_documents = lambda self, texts: Embeddings.embed_documents(self, texts)
- PartialImpl.embed_multimodal_documents = lambda self, docs: Embeddings.embed_multimodal_documents(self, docs)
- PartialImpl.embed_multimodal_query = lambda self, doc: Embeddings.embed_multimodal_query(self, doc)
- PartialImpl.aembed_documents = lambda self, texts: Embeddings.aembed_documents(self, texts)
- PartialImpl.aembed_query = lambda self, text: Embeddings.aembed_query(self, text)
-
- partial = PartialImpl()
+ partial = _SuperCallingEmbeddings()
with pytest.raises(NotImplementedError):
partial.embed_multimodal_query({"file_id": "test"})
- def test_aembed_documents_raises_not_implemented(self):
+ def test_aembed_documents_raises_not_implemented(self) -> None:
"""Test that aembed_documents raises NotImplementedError."""
- class PartialImpl:
- pass
-
- PartialImpl.embed_query = lambda self, text: Embeddings.embed_query(self, text)
- PartialImpl.embed_documents = lambda self, texts: Embeddings.embed_documents(self, texts)
- PartialImpl.embed_multimodal_documents = lambda self, docs: Embeddings.embed_multimodal_documents(self, docs)
- PartialImpl.embed_multimodal_query = lambda self, doc: Embeddings.embed_multimodal_query(self, doc)
- PartialImpl.aembed_documents = lambda self, texts: Embeddings.aembed_documents(self, texts)
- PartialImpl.aembed_query = lambda self, text: Embeddings.aembed_query(self, text)
+ partial = _SuperCallingEmbeddings()
- partial = PartialImpl()
-
- async def run_test():
+ async def run_test() -> None:
with pytest.raises(NotImplementedError):
await partial.aembed_documents(["test"])
asyncio.run(run_test())
- def test_aembed_query_raises_not_implemented(self):
+ def test_aembed_query_raises_not_implemented(self) -> None:
"""Test that aembed_query raises NotImplementedError."""
- class PartialImpl:
- pass
-
- PartialImpl.embed_query = lambda self, text: Embeddings.embed_query(self, text)
- PartialImpl.embed_documents = lambda self, texts: Embeddings.embed_documents(self, texts)
- PartialImpl.embed_multimodal_documents = lambda self, docs: Embeddings.embed_multimodal_documents(self, docs)
- PartialImpl.embed_multimodal_query = lambda self, doc: Embeddings.embed_multimodal_query(self, doc)
- PartialImpl.aembed_documents = lambda self, texts: Embeddings.aembed_documents(self, texts)
- PartialImpl.aembed_query = lambda self, text: Embeddings.aembed_query(self, text)
-
- partial = PartialImpl()
+ partial = _SuperCallingEmbeddings()
- async def run_test():
+ async def run_test() -> None:
with pytest.raises(NotImplementedError):
await partial.aembed_query("test")
diff --git a/api/tests/unit_tests/core/rag/extractor/test_extractor_base.py b/api/tests/unit_tests/core/rag/extractor/test_extractor_base.py
index 1d5f27181b7..5ce57f3403b 100644
--- a/api/tests/unit_tests/core/rag/extractor/test_extractor_base.py
+++ b/api/tests/unit_tests/core/rag/extractor/test_extractor_base.py
@@ -4,23 +4,23 @@ from core.rag.extractor.extractor_base import BaseExtractor
class _CallsBaseExtractor(BaseExtractor):
- def extract(self):
+ def extract(self) -> object:
return super().extract()
class _ConcreteExtractor(BaseExtractor):
- def extract(self):
+ def extract(self) -> list[str]:
return ["ok"]
class TestBaseExtractor:
- def test_extract_default_raises_not_implemented(self):
+ def test_extract_default_raises_not_implemented(self) -> None:
extractor = _CallsBaseExtractor()
with pytest.raises(NotImplementedError):
extractor.extract()
- def test_concrete_extractor_can_override(self):
+ def test_concrete_extractor_can_override(self) -> None:
extractor = _ConcreteExtractor()
assert extractor.extract() == ["ok"]
diff --git a/api/tests/unit_tests/core/rag/extractor/test_jina_reader_extractor.py b/api/tests/unit_tests/core/rag/extractor/test_jina_reader_extractor.py
index 79fcac2f88b..803cd9e5aae 100644
--- a/api/tests/unit_tests/core/rag/extractor/test_jina_reader_extractor.py
+++ b/api/tests/unit_tests/core/rag/extractor/test_jina_reader_extractor.py
@@ -8,7 +8,7 @@ from services.data_source.website_service import WebsiteService
class TestJinaReaderWebExtractor:
- def test_extract_crawl_mode_returns_document(self, mocker: MockerFixture):
+ def test_extract_crawl_mode_returns_document(self, mocker: MockerFixture) -> None:
mocker.patch(
"core.rag.extractor.jina_reader_extractor.WebsiteService.get_crawl_url_data",
return_value={
@@ -36,7 +36,7 @@ class TestJinaReaderWebExtractor:
"title": "title",
}
- def test_extract_crawl_mode_with_missing_data_returns_empty(self, mocker: MockerFixture):
+ def test_extract_crawl_mode_with_missing_data_returns_empty(self, mocker: MockerFixture) -> None:
mocker.patch(
"core.rag.extractor.jina_reader_extractor.WebsiteService.get_crawl_url_data",
return_value=None,
@@ -52,7 +52,7 @@ class TestJinaReaderWebExtractor:
assert extractor.extract() == []
- def test_extract_non_crawl_mode_returns_empty(self, mocker: MockerFixture):
+ def test_extract_non_crawl_mode_returns_empty(self, mocker: MockerFixture) -> None:
mock_get_crawl = mocker.patch(
"core.rag.extractor.jina_reader_extractor.WebsiteService.get_crawl_url_data",
return_value={"content": "unused"},
diff --git a/api/tests/unit_tests/core/rag/extractor/test_notion_extractor.py b/api/tests/unit_tests/core/rag/extractor/test_notion_extractor.py
index 2eac243de8c..1bbeb00f260 100644
--- a/api/tests/unit_tests/core/rag/extractor/test_notion_extractor.py
+++ b/api/tests/unit_tests/core/rag/extractor/test_notion_extractor.py
@@ -26,7 +26,7 @@ def persisted_document(
sqlite_engine: Engine,
) -> Iterator[tuple[sessionmaker[Session], DocumentModel]]:
"""Persist a Notion document and bind the extractor's ``db.session`` to SQLite."""
- TypeBase.metadata.create_all(sqlite_engine, tables=[DocumentModel.__table__])
+ TypeBase.metadata.create_all(sqlite_engine, tables=[TypeBase.metadata.tables[DocumentModel.__tablename__]])
session_maker = sessionmaker(bind=sqlite_engine, expire_on_commit=False)
document = DocumentModel(
id=str(uuid.uuid4()),
@@ -49,7 +49,7 @@ def persisted_document(
yield session_maker, document
-def _mock_response(data, status_code: int = 200, text: str = ""):
+def _mock_response(data: object, status_code: int = 200, text: str = "") -> mock.Mock:
response = mock.Mock()
response.status_code = status_code
response.text = text
@@ -58,7 +58,7 @@ def _mock_response(data, status_code: int = 200, text: str = ""):
class TestNotionExtractorInitAndPublicMethods:
- def test_init_with_explicit_token(self):
+ def test_init_with_explicit_token(self) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -70,7 +70,7 @@ class TestNotionExtractorInitAndPublicMethods:
assert extractor._notion_access_token == "token"
- def test_init_uses_injected_token_loader(self):
+ def test_init_uses_injected_token_loader(self) -> None:
loader = mock.Mock(return_value="resolved-token")
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
@@ -84,7 +84,7 @@ class TestNotionExtractorInitAndPublicMethods:
assert extractor._notion_access_token == "resolved-token"
loader.assert_called_once_with()
- def test_init_propagates_missing_token_from_loader(self):
+ def test_init_propagates_missing_token_from_loader(self) -> None:
loader = mock.Mock(side_effect=ValueError("Must specify `integration_token`"))
with pytest.raises(ValueError, match="Must specify `integration_token`"):
notion_extractor.NotionExtractor(
@@ -96,7 +96,7 @@ class TestNotionExtractorInitAndPublicMethods:
notion_token_loader=loader,
)
- def test_extract_updates_last_edited_and_loads_documents(self, monkeypatch: pytest.MonkeyPatch):
+ def test_extract_updates_last_edited_and_loads_documents(self, monkeypatch: pytest.MonkeyPatch) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -117,7 +117,7 @@ class TestNotionExtractorInitAndPublicMethods:
load_mock.assert_called_once_with("obj", "page")
assert len(docs) == 1
- def test_load_data_as_documents_page_database_and_invalid(self, monkeypatch: pytest.MonkeyPatch):
+ def test_load_data_as_documents_page_database_and_invalid(self, monkeypatch: pytest.MonkeyPatch) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -140,7 +140,7 @@ class TestNotionExtractorInitAndPublicMethods:
class TestNotionDatabase:
- def test_get_notion_database_data_parses_property_types_and_pagination(self, mocker: MockerFixture):
+ def test_get_notion_database_data_parses_property_types_and_pagination(self, mocker: MockerFixture) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -159,9 +159,9 @@ class TestNotionDatabase:
"multi_select": [{"name": "A"}, {"name": "B"}],
},
"title_prop": {"type": "title", "title": [{"plain_text": "Title"}]},
- "empty_title": {"type": "title", "title": []},
+ "empty_title": {"type": "title", "title": list[object]()},
"rich": {"type": "rich_text", "rich_text": [{"plain_text": "RichText"}]},
- "empty_rich": {"type": "rich_text", "rich_text": []},
+ "empty_rich": {"type": "rich_text", "rich_text": list[object]()},
"select_prop": {"type": "select", "select": {"name": "Selected"}},
"empty_select": {"type": "select", "select": None},
"status_prop": {"type": "status", "status": {"name": "Open"}},
@@ -175,7 +175,7 @@ class TestNotionDatabase:
"has_more": True,
"next_cursor": "cursor-2",
}
- second_page = {"results": [], "has_more": False, "next_cursor": None}
+ second_page = {"results": list[object](), "has_more": False, "next_cursor": None}
mock_post = mocker.patch("httpx.post", side_effect=[_mock_response(first_page), _mock_response(second_page)])
@@ -191,7 +191,7 @@ class TestNotionDatabase:
assert "Row Page URL:https://notion.so/page-1" in content
assert mock_post.call_count == 2
- def test_get_notion_database_data_joins_all_rich_text_segments(self, mocker: MockerFixture):
+ def test_get_notion_database_data_joins_all_rich_text_segments(self, mocker: MockerFixture) -> None:
"""Formatted text arrives as multiple segments; all of them must be kept."""
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
@@ -235,7 +235,7 @@ class TestNotionDatabase:
assert "rich:first second third" in content
assert "title_prop:Hello world" in content
- def test_get_notion_database_data_handles_missing_results_and_empty_content(self, mocker: MockerFixture):
+ def test_get_notion_database_data_handles_missing_results_and_empty_content(self, mocker: MockerFixture) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -248,7 +248,7 @@ class TestNotionDatabase:
mocker.patch("httpx.post", return_value=_mock_response({"results": None}))
assert extractor._get_notion_database_data("db-1") == []
- def test_requests_use_bounded_timeout(self, mocker: MockerFixture):
+ def test_requests_use_bounded_timeout(self, mocker: MockerFixture) -> None:
"""All outbound Notion API calls must carry a bounded timeout so a hanging endpoint cannot block extraction."""
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
@@ -269,7 +269,7 @@ class TestNotionDatabase:
extractor.get_notion_last_edited_time()
assert mock_request.call_args.kwargs["timeout"] == notion_extractor._REQUEST_TIMEOUT
- def test_get_notion_database_data_requires_access_token(self):
+ def test_get_notion_database_data_requires_access_token(self) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -285,7 +285,9 @@ class TestNotionDatabase:
class TestNotionBlocks:
- def test_get_notion_block_data_success_with_table_headings_children_and_pagination(self, mocker: MockerFixture):
+ def test_get_notion_block_data_success_with_table_headings_children_and_pagination(
+ self, mocker: MockerFixture
+ ) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -297,7 +299,7 @@ class TestNotionBlocks:
first_response = {
"results": [
- {"type": "table", "id": "tbl-1", "has_children": False, "table": {}},
+ {"type": "table", "id": "tbl-1", "has_children": False, "table": dict[str, object]()},
{
"type": "heading_1",
"id": "h1",
@@ -314,7 +316,7 @@ class TestNotionBlocks:
"type": "child_page",
"id": "cp1",
"has_children": True,
- "child_page": {"rich_text": []},
+ "child_page": {"rich_text": list[object]()},
},
],
"next_cursor": "cursor-2",
@@ -342,7 +344,7 @@ class TestNotionBlocks:
assert "Paragraph\nCHILD\n\n" in lines[2]
assert "## SubHeading" in lines[-1]
- def test_get_notion_block_data_handles_http_error_and_invalid_payload(self, mocker: MockerFixture):
+ def test_get_notion_block_data_handles_http_error_and_invalid_payload(self, mocker: MockerFixture) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -364,7 +366,7 @@ class TestNotionBlocks:
with pytest.raises(ValueError, match="Error fetching Notion block data: boom"):
extractor._get_notion_block_data("page-1")
- def test_read_block_supports_heading_table_and_recursion(self, mocker: MockerFixture):
+ def test_read_block_supports_heading_table_and_recursion(self, mocker: MockerFixture) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -388,7 +390,7 @@ class TestNotionBlocks:
"has_children": True,
"paragraph": {"rich_text": [{"text": {"content": "Parent"}}]},
},
- {"type": "table", "id": "tbl-1", "has_children": False, "table": {}},
+ {"type": "table", "id": "tbl-1", "has_children": False, "table": dict[str, object]()},
],
"next_cursor": None,
}
@@ -414,7 +416,7 @@ class TestNotionBlocks:
assert "Child" in content
assert "TABLE-MD" in content
- def test_read_block_breaks_on_missing_results(self, mocker: MockerFixture):
+ def test_read_block_breaks_on_missing_results(self, mocker: MockerFixture) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -427,7 +429,7 @@ class TestNotionBlocks:
assert extractor._read_block("root") == ""
- def test_read_table_rows_formats_markdown_with_pagination(self, mocker: MockerFixture):
+ def test_read_table_rows_formats_markdown_with_pagination(self, mocker: MockerFixture) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -464,7 +466,7 @@ class TestNotionBlocks:
"table_row": {
"cells": [
[{"text": {"content": "H1"}}],
- [],
+ list[object](),
]
}
},
@@ -489,7 +491,7 @@ class TestNotionBlocks:
assert "| H1 | |" in markdown
assert "| R2C1 | R2C2 |" in markdown
- def test_read_table_rows_joins_rich_text_segments_within_cell(self, mocker: MockerFixture):
+ def test_read_table_rows_joins_rich_text_segments_within_cell(self, mocker: MockerFixture) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -532,7 +534,7 @@ class TestNotionBlocks:
if line.startswith("|"):
assert line.count("|") == 3
- def test_read_table_rows_preserves_empty_data_cells_as_columns(self, mocker: MockerFixture):
+ def test_read_table_rows_preserves_empty_data_cells_as_columns(self, mocker: MockerFixture) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -557,7 +559,7 @@ class TestNotionBlocks:
"table_row": {
"cells": [
[{"text": {"content": "a1"}}],
- [],
+ list[object](),
[{"text": {"content": "c1"}}],
]
}
@@ -576,7 +578,7 @@ class TestNotionBlocks:
class TestNotionMetadataAndCredentialMethods:
- def test_update_last_edited_time_no_document_model(self):
+ def test_update_last_edited_time_no_document_model(self) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -592,7 +594,7 @@ class TestNotionMetadataAndCredentialMethods:
self,
monkeypatch: pytest.MonkeyPatch,
persisted_document: tuple[sessionmaker[Session], DocumentModel],
- ):
+ ) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -621,7 +623,7 @@ class TestNotionMetadataAndCredentialMethods:
self,
monkeypatch: pytest.MonkeyPatch,
persisted_document: tuple[sessionmaker[Session], DocumentModel],
- ):
+ ) -> None:
session_maker, document = persisted_document
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
@@ -639,7 +641,7 @@ class TestNotionMetadataAndCredentialMethods:
assert stored is not None
assert stored.data_source_info_dict["last_edited_time"] == "2025-01-01T00:00:00.000Z"
- def test_get_notion_last_edited_time_uses_page_and_database_urls(self, mocker: MockerFixture):
+ def test_get_notion_last_edited_time_uses_page_and_database_urls(self, mocker: MockerFixture) -> None:
extractor_page = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="page-id",
@@ -670,7 +672,7 @@ class TestNotionMetadataAndCredentialMethods:
assert extractor_db.get_notion_last_edited_time() == "2025-06-01T00:00:00.000Z"
assert "databases/db-id" in request_mock.call_args[0][1]
- def test_get_notion_last_edited_time_requires_access_token(self):
+ def test_get_notion_last_edited_time_requires_access_token(self) -> None:
extractor = notion_extractor.NotionExtractor(
notion_workspace_id="ws",
notion_obj_id="obj",
@@ -685,12 +687,12 @@ class TestNotionMetadataAndCredentialMethods:
extractor.get_notion_last_edited_time()
-def test_get_cell_text_uses_plain_text_for_mention_and_equation_segments():
+def test_get_cell_text_uses_plain_text_for_mention_and_equation_segments() -> None:
# Notion rich text segments of type mention or equation carry plain_text
# but no "text" object; they must not be dropped from the cell content.
cell = [
{"type": "text", "text": {"content": "see "}, "plain_text": "see "},
- {"type": "mention", "mention": {"type": "user", "user": {}}, "plain_text": "@alice"},
+ {"type": "mention", "mention": {"type": "user", "user": dict[str, object]()}, "plain_text": "@alice"},
{"type": "equation", "equation": {"expression": "e=mc^2"}, "plain_text": "e=mc^2"},
]
assert notion_extractor.NotionExtractor._get_cell_text(cell) == "see @alicee=mc^2"
diff --git a/api/tests/unit_tests/core/rag/retrieval/test_multi_dataset_function_call_router.py b/api/tests/unit_tests/core/rag/retrieval/test_multi_dataset_function_call_router.py
index 43c521dcfd4..8ef8e16b647 100644
--- a/api/tests/unit_tests/core/rag/retrieval/test_multi_dataset_function_call_router.py
+++ b/api/tests/unit_tests/core/rag/retrieval/test_multi_dataset_function_call_router.py
@@ -62,7 +62,7 @@ class TestFunctionCallMultiDatasetRouter:
router = FunctionCallMultiDatasetRouter()
response = Mock()
response.usage = LLMUsage.empty_usage()
- response.message.tool_calls = []
+ response.message.tool_calls = list[object]()
model_instance = Mock()
model_instance.invoke_llm.return_value = response
tool_1 = Mock()
diff --git a/api/tests/unit_tests/core/rag/retrieval/test_multi_dataset_react_route.py b/api/tests/unit_tests/core/rag/retrieval/test_multi_dataset_react_route.py
index bcc8ee8fcb5..562a49a4f11 100644
--- a/api/tests/unit_tests/core/rag/retrieval/test_multi_dataset_react_route.py
+++ b/api/tests/unit_tests/core/rag/retrieval/test_multi_dataset_react_route.py
@@ -1,3 +1,4 @@
+from collections.abc import Generator
from types import SimpleNamespace
from unittest.mock import Mock, patch
@@ -191,7 +192,10 @@ class TestReactMultiDatasetRouter:
delta = SimpleNamespace(message=SimpleNamespace(content="part"), usage=None)
chunk = SimpleNamespace(model="m1", prompt_messages=[Mock()], delta=delta)
- text, usage = router._handle_invoke_result(iter([chunk]))
+ def stream() -> Generator[SimpleNamespace]:
+ yield chunk
+
+ text, usage = router._handle_invoke_result(stream())
assert text == "part"
assert usage == LLMUsage.empty_usage()
@@ -229,7 +233,7 @@ class TestReactMultiDatasetRouter:
router = ReactMultiDatasetRouter()
model_config = Mock()
model_config.mode = "unknown-mode"
- model_config.parameters = {}
+ model_config.parameters = dict[str, object]()
tool = Mock()
tool.name = "dataset-1"
tool.description = "desc"
diff --git a/api/tests/unit_tests/pyrefly.toml b/api/tests/unit_tests/pyrefly.toml
index aec94de0e96..d7e43c75791 100644
--- a/api/tests/unit_tests/pyrefly.toml
+++ b/api/tests/unit_tests/pyrefly.toml
@@ -395,18 +395,14 @@ project-excludes = [
"core/rag/datasource/keyword/jieba/test_jieba_keyword_table_handler.py",
"core/rag/datasource/test_datasource_retrieval.py",
"core/rag/datasource/vdb/test_vector_factory.py",
- "core/rag/embedding/test_embedding_base.py",
"core/rag/embedding/test_embedding_service.py",
"core/rag/extractor/blob/test_blob.py",
"core/rag/extractor/test_csv_extractor.py",
"core/rag/extractor/test_excel_extractor.py",
"core/rag/extractor/test_extract_processor.py",
- "core/rag/extractor/test_extractor_base.py",
"core/rag/extractor/test_helpers.py",
"core/rag/extractor/test_html_extractor.py",
- "core/rag/extractor/test_jina_reader_extractor.py",
"core/rag/extractor/test_markdown_extractor.py",
- "core/rag/extractor/test_notion_extractor.py",
"core/rag/extractor/test_word_extractor.py",
"core/rag/extractor/unstructured/test_unstructured_extractors.py",
"core/rag/extractor/watercrawl/test_watercrawl.py",
@@ -417,8 +413,6 @@ project-excludes = [
"core/rag/rerank/test_reranker.py",
"core/rag/retrieval/test_dataset_retrieval.py",
"core/rag/retrieval/test_dataset_retrieval_methods.py",
- "core/rag/retrieval/test_multi_dataset_function_call_router.py",
- "core/rag/retrieval/test_multi_dataset_react_route.py",
"core/rag/splitter/test_text_splitter.py",
"core/repositories/test_celery_workflow_execution_repository.py",
"core/repositories/test_celery_workflow_node_execution_repository.py",