reformatted again

Author: Richard Kuo (Onyx)

Diff for: .pre-commit-config.yaml

@@ -1,12 +1,12 @@
 repos:
   - repo: https://github.com/psf/black
-    rev: 23.3.0
+    rev: 25.1.0
     hooks:
     - id: black
       language_version: python3.11
 
   - repo: https://github.com/asottile/reorder_python_imports
-    rev: v3.9.0
+    rev: v3.14.0
     hooks:
     - id: reorder-python-imports
       args: ['--py311-plus', '--application-directories=backend/']
@@ -18,14 +18,14 @@ repos:
   # These settings will remove unused imports with side effects
   # Note: The repo currently does not and should not have imports with side effects
   - repo: https://github.com/PyCQA/autoflake
-    rev: v2.2.0
+    rev: v2.3.1
     hooks:
       - id: autoflake
         args: [ '--remove-all-unused-imports', '--remove-unused-variables', '--in-place' , '--recursive']
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
     # Ruff version.
-    rev: v0.0.286
+    rev: v0.11.4
     hooks:
       - id: ruff
   - repo: https://github.com/pre-commit/mirrors-prettier

Diff for: backend/ee/onyx/server/oauth/google_drive.py

@@ -195,9 +195,9 @@ def handle_google_drive_oauth_callback(
         credential_dict: dict[str, str] = {}
         credential_dict[DB_CREDENTIALS_DICT_TOKEN_KEY] = oauth_creds_sanitized_json_str
         credential_dict[DB_CREDENTIALS_PRIMARY_ADMIN_KEY] = session.email
-        credential_dict[
-            DB_CREDENTIALS_AUTHENTICATION_METHOD
-        ] = GoogleOAuthAuthenticationMethod.OAUTH_INTERACTIVE.value
+        credential_dict[DB_CREDENTIALS_AUTHENTICATION_METHOD] = (
+            GoogleOAuthAuthenticationMethod.OAUTH_INTERACTIVE.value
+        )
 
         credential_info = CredentialBase(
             credential_json=credential_dict,

Diff for: backend/onyx/agents/agent_search/deep_search/main/nodes/generate_validate_refined_answer.py

@@ -244,9 +244,7 @@ def generate_validate_refined_answer(
     revision_question_efficiency = (
         len(total_answered_questions) / len(initial_answered_sub_questions)
         if initial_answered_sub_questions
-        else 10.0
-        if refined_answered_sub_questions
-        else 1.0
+        else 10.0 if refined_answered_sub_questions else 1.0
     )
 
     sub_question_answer_str = "\n\n------\n\n".join(

Diff for: backend/onyx/agents/agent_search/deep_search/main/states.py

@@ -112,9 +112,9 @@ class SubQuestionResultsUpdate(LoggerUpdate):
         list[InferenceSection], dedup_inference_sections
     ] = []
     context_documents: Annotated[list[InferenceSection], dedup_inference_sections] = []
-    cited_documents: Annotated[
-        list[InferenceSection], dedup_inference_sections
-    ] = []  # cited docs from sub-answers are used for answer context
+    cited_documents: Annotated[list[InferenceSection], dedup_inference_sections] = (
+        []
+    )  # cited docs from sub-answers are used for answer context
     sub_question_results: Annotated[
         list[SubQuestionAnswerResults], dedup_question_answer_results
     ] = []

Diff for: backend/onyx/agents/agent_search/deep_search/shared/expanded_retrieval/states.py

@@ -44,9 +44,9 @@ class DocVerificationUpdate(LoggerUpdate, BaseModel):
 
 class DocRetrievalUpdate(LoggerUpdate, BaseModel):
     query_retrieval_results: Annotated[list[QueryRetrievalResult], add] = []
-    retrieved_documents: Annotated[
-        list[InferenceSection], dedup_inference_sections
-    ] = []
+    retrieved_documents: Annotated[list[InferenceSection], dedup_inference_sections] = (
+        []
+    )
 
 
 class DocRerankingUpdate(LoggerUpdate, BaseModel):
@@ -64,9 +64,9 @@ class ExpandedRetrievalUpdate(LoggerUpdate, BaseModel):
 class ExpandedRetrievalOutput(LoggerUpdate, BaseModel):
     expanded_retrieval_result: QuestionRetrievalResult = QuestionRetrievalResult()
     base_expanded_retrieval_result: QuestionRetrievalResult = QuestionRetrievalResult()
-    retrieved_documents: Annotated[
-        list[InferenceSection], dedup_inference_sections
-    ] = []
+    retrieved_documents: Annotated[list[InferenceSection], dedup_inference_sections] = (
+        []
+    )
 
 
 ## Graph State

Diff for: backend/onyx/agents/agent_search/shared_graph_utils/operators.py

@@ -23,9 +23,9 @@ def dedup_question_answer_results(
     question_answer_results_1: list[SubQuestionAnswerResults],
     question_answer_results_2: list[SubQuestionAnswerResults],
 ) -> list[SubQuestionAnswerResults]:
-    deduped_question_answer_results: list[
-        SubQuestionAnswerResults
-    ] = question_answer_results_1
+    deduped_question_answer_results: list[SubQuestionAnswerResults] = (
+        question_answer_results_1
+    )
     utilized_question_ids: set[str] = set(
         [x.question_id for x in question_answer_results_1]
     )

Diff for: backend/onyx/background/indexing/run_indexing.py

@@ -423,9 +423,9 @@ def _run_indexing(
                 unresolved_only=True,
                 db_session=db_session_temp,
             )
-            doc_id_to_unresolved_errors: dict[
-                str, list[IndexAttemptError]
-            ] = defaultdict(list)
+            doc_id_to_unresolved_errors: dict[str, list[IndexAttemptError]] = (
+                defaultdict(list)
+            )
             for error in unresolved_errors:
                 if error.document_id:
                     doc_id_to_unresolved_errors[error.document_id].append(error)

Diff for: backend/onyx/chat/answer.py

@@ -212,9 +212,9 @@ def citations(self) -> list[CitationInfo]:
         return citations
 
     def citations_by_subquestion(self) -> dict[SubQuestionKey, list[CitationInfo]]:
-        citations_by_subquestion: dict[
-            SubQuestionKey, list[CitationInfo]
-        ] = defaultdict(list)
+        citations_by_subquestion: dict[SubQuestionKey, list[CitationInfo]] = (
+            defaultdict(list)
+        )
         basic_subq_key = SubQuestionKey(level=BASIC_KEY[0], question_num=BASIC_KEY[1])
         for packet in self.processed_streamed_output:
             if isinstance(packet, CitationInfo):

Diff for: backend/onyx/chat/process_message.py

@@ -172,9 +172,9 @@ def _translate_citations(
     citation_to_saved_doc_id_map: dict[int, int] = {}
     for citation in citations_list:
         if citation.citation_num not in citation_to_saved_doc_id_map:
-            citation_to_saved_doc_id_map[
-                citation.citation_num
-            ] = doc_id_to_saved_doc_id_map[citation.document_id]
+            citation_to_saved_doc_id_map[citation.citation_num] = (
+                doc_id_to_saved_doc_id_map[citation.document_id]
+            )
 
     return MessageSpecificCitations(citation_map=citation_to_saved_doc_id_map)
 

Diff for: backend/onyx/connectors/confluence/onyx_confluence.py

@@ -397,9 +397,9 @@ def __getattr__(self, name: str) -> Any:
             return attr
 
         # wrap the method with our retry handler
-        rate_limited_method: Callable[
-            ..., Any
-        ] = self._make_rate_limited_confluence_method(name, self._credentials_provider)
+        rate_limited_method: Callable[..., Any] = (
+            self._make_rate_limited_confluence_method(name, self._credentials_provider)
+        )
 
         def wrapped_method(*args: Any, **kwargs: Any) -> Any:
             return rate_limited_method(*args, **kwargs)

Diff for: backend/onyx/connectors/connector_runner.py

@@ -88,9 +88,7 @@ def __init__(
 
         self.doc_batch: list[Document] = []
 
-    def run(
-        self, checkpoint: CT
-    ) -> Generator[
+    def run(self, checkpoint: CT) -> Generator[
         tuple[list[Document] | None, ConnectorFailure | None, CT | None],
         None,
         None,

Diff for: backend/onyx/connectors/egnyte/connector.py

@@ -100,9 +100,9 @@ def _process_egnyte_file(
 
     # Add lock info if present
     if lock_info := file_metadata.get("lock_info"):
-        metadata[
-            "lock_owner"
-        ] = f"{lock_info.get('first_name', '')} {lock_info.get('last_name', '')}"
+        metadata["lock_owner"] = (
+            f"{lock_info.get('first_name', '')} {lock_info.get('last_name', '')}"
+        )
 
     # Create the document owners
     primary_owner = None

Diff for: backend/onyx/connectors/gong/connector.py

@@ -280,9 +280,9 @@ def _fetch_calls(
                     speaker_id = segment.get("speakerId", "")
                     if speaker_id not in speaker_to_name:
                         if self.hide_user_info:
-                            speaker_to_name[
-                                speaker_id
-                            ] = f"User {len(speaker_to_name) + 1}"
+                            speaker_to_name[speaker_id] = (
+                                f"User {len(speaker_to_name) + 1}"
+                            )
                         else:
                             speaker_to_name[speaker_id] = id_to_name_map.get(
                                 speaker_id, "Unknown"

Diff for: backend/onyx/connectors/google_drive/connector.py

@@ -141,8 +141,7 @@ def __call__(
         checkpoint: GoogleDriveCheckpoint,
         start: SecondsSinceUnixEpoch | None = None,
         end: SecondsSinceUnixEpoch | None = None,
-    ) -> Iterator[RetrievedDriveFile]:
-        ...
+    ) -> Iterator[RetrievedDriveFile]: ...
 
 
 def add_retrieval_info(

Diff for: backend/onyx/connectors/google_utils/google_kv.py

@@ -147,9 +147,9 @@ def build_service_account_creds(
     if primary_admin_email:
         credential_dict[DB_CREDENTIALS_PRIMARY_ADMIN_KEY] = primary_admin_email
 
-    credential_dict[
-        DB_CREDENTIALS_AUTHENTICATION_METHOD
-    ] = GoogleOAuthAuthenticationMethod.UPLOADED.value
+    credential_dict[DB_CREDENTIALS_AUTHENTICATION_METHOD] = (
+        GoogleOAuthAuthenticationMethod.UPLOADED.value
+    )
 
     return CredentialBase(
         credential_json=credential_dict,

Diff for: backend/onyx/connectors/notion/connector.py

@@ -576,9 +576,9 @@ def _recursive_load(self) -> Generator[list[Document], None, None]:
 
     def load_credentials(self, credentials: dict[str, Any]) -> dict[str, Any] | None:
         """Applies integration token to headers"""
-        self.headers[
-            "Authorization"
-        ] = f'Bearer {credentials["notion_integration_token"]}'
+        self.headers["Authorization"] = (
+            f'Bearer {credentials["notion_integration_token"]}'
+        )
         return None
 
     def load_from_state(self) -> GenerateDocumentsOutput:

Diff for: backend/onyx/db/auth.py

@@ -29,10 +29,10 @@
 def get_default_admin_user_emails() -> list[str]:
     """Returns a list of emails who should default to Admin role.
     Only used in the EE version. For MIT, just return empty list."""
-    get_default_admin_user_emails_fn: Callable[
-        [], list[str]
-    ] = fetch_versioned_implementation_with_fallback(
-        "onyx.auth.users", "get_default_admin_user_emails_", lambda: list[str]()
+    get_default_admin_user_emails_fn: Callable[[], list[str]] = (
+        fetch_versioned_implementation_with_fallback(
+            "onyx.auth.users", "get_default_admin_user_emails_", lambda: list[str]()
+        )
     )
     return get_default_admin_user_emails_fn()
 

Diff for: backend/onyx/db/models.py

@@ -702,12 +702,12 @@ class Connector(Base):
         back_populates="connector",
         cascade="all, delete-orphan",
     )
-    documents_by_connector: Mapped[
-        list["DocumentByConnectorCredentialPair"]
-    ] = relationship(
-        "DocumentByConnectorCredentialPair",
-        back_populates="connector",
-        passive_deletes=True,
+    documents_by_connector: Mapped[list["DocumentByConnectorCredentialPair"]] = (
+        relationship(
+            "DocumentByConnectorCredentialPair",
+            back_populates="connector",
+            passive_deletes=True,
+        )
     )
 
     # synchronize this validation logic with RefreshFrequencySchema etc on front end
@@ -760,12 +760,12 @@ class Credential(Base):
         back_populates="credential",
         cascade="all, delete-orphan",
     )
-    documents_by_credential: Mapped[
-        list["DocumentByConnectorCredentialPair"]
-    ] = relationship(
-        "DocumentByConnectorCredentialPair",
-        back_populates="credential",
-        passive_deletes=True,
+    documents_by_credential: Mapped[list["DocumentByConnectorCredentialPair"]] = (
+        relationship(
+            "DocumentByConnectorCredentialPair",
+            back_populates="credential",
+            passive_deletes=True,
+        )
     )
 
     user: Mapped[User | None] = relationship("User", back_populates="credentials")
@@ -2194,11 +2194,11 @@ class UserGroup(Base):
         secondary=UserGroup__ConnectorCredentialPair.__table__,
         viewonly=True,
     )
-    cc_pair_relationships: Mapped[
-        list[UserGroup__ConnectorCredentialPair]
-    ] = relationship(
-        "UserGroup__ConnectorCredentialPair",
-        viewonly=True,
+    cc_pair_relationships: Mapped[list[UserGroup__ConnectorCredentialPair]] = (
+        relationship(
+            "UserGroup__ConnectorCredentialPair",
+            viewonly=True,
+        )
     )
     personas: Mapped[list[Persona]] = relationship(
         "Persona",

Diff for: backend/onyx/document_index/vespa/index.py

@@ -177,9 +177,9 @@ def __init__(
         self.index_to_large_chunks_enabled: dict[str, bool] = {}
         self.index_to_large_chunks_enabled[index_name] = large_chunks_enabled
         if secondary_index_name and secondary_large_chunks_enabled:
-            self.index_to_large_chunks_enabled[
-                secondary_index_name
-            ] = secondary_large_chunks_enabled
+            self.index_to_large_chunks_enabled[secondary_index_name] = (
+                secondary_large_chunks_enabled
+            )
 
     def ensure_indices_exist(
         self,
@@ -389,9 +389,9 @@ def index(
         new_document_id_to_original_document_id: dict[str, str] = {}
         for ind, chunk in enumerate(cleaned_chunks):
             old_chunk = chunks[ind]
-            new_document_id_to_original_document_id[
-                chunk.source_document.id
-            ] = old_chunk.source_document.id
+            new_document_id_to_original_document_id[chunk.source_document.id] = (
+                old_chunk.source_document.id
+            )
 
         existing_docs: set[str] = set()
 

Diff for: backend/onyx/indexing/indexing_pipeline.py

@@ -122,8 +122,7 @@ def __call__(
         self,
         document_batch: list[Document],
         index_attempt_metadata: IndexAttemptMetadata,
-    ) -> IndexingPipelineResult:
-        ...
+    ) -> IndexingPipelineResult: ...
 
 
 def _upsert_documents_in_db(
@@ -835,10 +834,10 @@ def index_doc_batch(
         doc_id_to_user_file_id: dict[str, int | None] = fetch_user_files_for_documents(
             document_ids=updatable_ids, db_session=db_session
         )
-        doc_id_to_user_folder_id: dict[
-            str, int | None
-        ] = fetch_user_folders_for_documents(
-            document_ids=updatable_ids, db_session=db_session
+        doc_id_to_user_folder_id: dict[str, int | None] = (
+            fetch_user_folders_for_documents(
+                document_ids=updatable_ids, db_session=db_session
+            )
         )
 
         doc_id_to_previous_chunk_cnt: dict[str, int | None] = {

Diff for: backend/onyx/server/features/persona/api.py

@@ -381,8 +381,9 @@ def delete_persona(
 
 @basic_router.get("/image-generation-tool")
 def get_image_generation_tool(
-    _: User
-    | None = Depends(current_user),  # User param not used but kept for consistency
+    _: User | None = Depends(
+        current_user
+    ),  # User param not used but kept for consistency
     db_session: Session = Depends(get_session),
 ) -> ImageGenerationToolStatus:  # Use bool instead of str for boolean values
     is_available = is_image_generation_available(db_session=db_session)

Diff for: backend/onyx/server/manage/llm/api.py

@@ -237,8 +237,9 @@ def set_provider_as_default(
 @admin_router.post("/provider/{provider_id}/default-vision")
 def set_provider_as_default_vision(
     provider_id: int,
-    vision_model: str
-    | None = Query(None, description="The default vision model to use"),
+    vision_model: str | None = Query(
+        None, description="The default vision model to use"
+    ),
     _: User | None = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
 ) -> None:

Diff for: backend/onyx/server/manage/slack_bot.py

@@ -96,13 +96,13 @@ def _form_channel_config(
     if follow_up_tags is not None:
         channel_config["follow_up_tags"] = follow_up_tags
 
-    channel_config[
-        "show_continue_in_web_ui"
-    ] = slack_channel_config_creation_request.show_continue_in_web_ui
+    channel_config["show_continue_in_web_ui"] = (
+        slack_channel_config_creation_request.show_continue_in_web_ui
+    )
 
-    channel_config[
-        "respond_to_bots"
-    ] = slack_channel_config_creation_request.respond_to_bots
+    channel_config["respond_to_bots"] = (
+        slack_channel_config_creation_request.respond_to_bots
+    )
 
     channel_config["is_ephemeral"] = slack_channel_config_creation_request.is_ephemeral