chore(lint): enable perflint ruff rules (#14025)

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>

backend/alembic/env.py

@@ -131,10 +131,11 @@ def filter_tenants_by_range(
     filtered_tenant_schemas = sorted_tenant_schemas[start_idx:end_idx]
 
     # Combine with non-tenant schemas and preserve original order
-    filtered_tenants = []
-    for tenant_id in tenant_ids:
-        if tenant_id in filtered_tenant_schemas or tenant_id in non_tenant_schemas:
-            filtered_tenants.append(tenant_id)
+    filtered_tenants = [
+        tenant_id
+        for tenant_id in tenant_ids
+        if tenant_id in filtered_tenant_schemas or tenant_id in non_tenant_schemas
+    ]
 
     return filtered_tenants
 

backend/alembic/versions/12635f6655b7_drive_canonical_ids.py

@@ -100,9 +100,7 @@ def get_google_drive_documents_from_database() -> list[dict]:
         """)
     )
 
-    documents = []
-    for row in result:
-        documents.append({"document_id": row.id})
+    documents = [{"document_id": row.id} for row in result]
 
     return documents
 

backend/alembic/versions/a3795dce87be_migration_confluence_to_be_explicit.py

@@ -103,9 +103,9 @@ def upgrade() -> None:
             "is_cloud": is_cloud,
         }
 
-        for key, value in config.items():
-            if key not in ["wiki_page_url"]:
-                new_config[key] = value
+        new_config.update(
+            {key: value for key, value in config.items() if key != "wiki_page_url"}
+        )
 
         op.execute(
             connector.update()

backend/ee/onyx/background/celery/tasks/doc_permission_syncing/tasks.py

@@ -220,9 +220,11 @@ def check_for_doc_permissions_sync(self: Task, *, tenant_id: str) -> bool | None
         with get_session_with_current_tenant() as db_session:
             cc_pairs = get_all_auto_sync_cc_pairs(db_session)
 
-            for cc_pair in cc_pairs:
-                if _is_external_doc_permissions_sync_due(cc_pair):
-                    cc_pair_ids_to_sync.append(cc_pair.id)
+            cc_pair_ids_to_sync.extend(
+                cc_pair.id
+                for cc_pair in cc_pairs
+                if _is_external_doc_permissions_sync_due(cc_pair)
+            )
 
         # Tenant-work-gating hook: refresh this tenant's active-set membership
         # whenever doc-permission sync has any due cc_pairs to dispatch.

backend/ee/onyx/background/celery/tasks/external_group_syncing/tasks.py

@@ -206,9 +206,11 @@ def check_for_external_group_sync(self: Task, *, tenant_id: str) -> bool | None:
                         if cc_pair.id != cc_pair_to_remove.id
                     ]
 
-            for cc_pair in cc_pairs:
-                if _is_external_group_sync_due(cc_pair):
-                    cc_pair_ids_to_sync.append(cc_pair.id)
+            cc_pair_ids_to_sync.extend(
+                cc_pair.id
+                for cc_pair in cc_pairs
+                if _is_external_group_sync_due(cc_pair)
+            )
 
         # Tenant-work-gating hook: refresh this tenant's active-set membership
         # whenever external-group sync has any due cc_pairs to dispatch.

backend/ee/onyx/external_permissions/github/utils.py

@@ -179,7 +179,6 @@ def fetch_repository_team_slugs(
 ) -> List[str]:
     """Fetch team slugs with access to the repository."""
     logger.info("Fetching team slugs for repository %s", repo.full_name)
-    teams_data: List[str] = []
 
     team_objs: PaginatedList[Team] | list[Team] = (
         _run_with_retry(
@@ -190,8 +189,7 @@ def fetch_repository_team_slugs(
         or []
     )
 
-    for team in team_objs:
-        teams_data.append(team.slug)
+    teams_data: List[str] = [team.slug for team in team_objs]
 
     logger.info(
         "Fetched %s team slugs for repository %s", len(teams_data), repo.full_name

backend/ee/onyx/external_permissions/post_query_censoring.py

@@ -81,11 +81,12 @@ def _post_query_chunk_censoring(
             final_chunk_dict[censored_chunk.unique_id] = censored_chunk
 
     # IMPORTANT: make sure to retain the same ordering as the original `chunks` passed in
-    final_chunk_list: list[InferenceChunk] = []
-    for chunk in chunks:
-        # only if the chunk is in the final censored chunks, add it to the final list
-        # if it is missing, that means it was intentionally left out
-        if chunk.unique_id in final_chunk_dict:
-            final_chunk_list.append(final_chunk_dict[chunk.unique_id])
+    # only if the chunk is in the final censored chunks, add it to the final list
+    # if it is missing, that means it was intentionally left out
+    final_chunk_list: list[InferenceChunk] = [
+        final_chunk_dict[chunk.unique_id]
+        for chunk in chunks
+        if chunk.unique_id in final_chunk_dict
+    ]
 
     return final_chunk_list

backend/ee/onyx/external_permissions/sharepoint/permission_utils.py

@@ -893,9 +893,10 @@ def get_sharepoint_external_groups(
         return external_user_groups
 
     already_resolved = set(groups_and_members.groups_to_emails.keys())
-    for group in _enumerate_ad_groups_paginated(
-        get_access_token, already_resolved, graph_api_base
-    ):
-        external_user_groups.append(group)
+    external_user_groups.extend(
+        _enumerate_ad_groups_paginated(
+            get_access_token, already_resolved, graph_api_base
+        )
+    )
 
     return external_user_groups

backend/ee/onyx/server/analytics/api.py

@@ -267,15 +267,14 @@ def get_assistant_stats(
     all_dates = set(daily_messages_map.keys()) | set(daily_unique_users_map.keys())
 
     # Merge both sets of metrics by date
-    daily_results: list[AssistantDailyUsageResponse] = []
-    for date in sorted(all_dates):
-        daily_results.append(
-            AssistantDailyUsageResponse(
-                date=date,
-                total_messages=daily_messages_map.get(date, 0),
-                total_unique_users=daily_unique_users_map.get(date, 0),
-            )
+    daily_results: list[AssistantDailyUsageResponse] = [
+        AssistantDailyUsageResponse(
+            date=date,
+            total_messages=daily_messages_map.get(date, 0),
+            total_unique_users=daily_unique_users_map.get(date, 0),
         )
+        for date in sorted(all_dates)
+    ]
 
     # Now pull a single total distinct user count across the entire time range
     total_msgs = sum(d.total_messages for d in daily_results)

backend/ee/onyx/server/query_history/models.py

@@ -171,14 +171,13 @@ class QuestionAnswerPairSnapshot(BaseModel):
         cls,
         chat_session_snapshot: ChatSessionSnapshot,
     ) -> list["QuestionAnswerPairSnapshot"]:
-        message_pairs: list[tuple[MessageSnapshot, MessageSnapshot]] = []
-        for ind in range(1, len(chat_session_snapshot.messages), 2):
-            message_pairs.append(
-                (
-                    chat_session_snapshot.messages[ind - 1],
-                    chat_session_snapshot.messages[ind],
-                )
+        message_pairs: list[tuple[MessageSnapshot, MessageSnapshot]] = [
+            (
+                chat_session_snapshot.messages[ind - 1],
+                chat_session_snapshot.messages[ind],
             )
+            for ind in range(1, len(chat_session_snapshot.messages), 2)
+        ]
 
         return [
             cls(

backend/onyx/background/celery/celery_redis.py

@@ -214,7 +214,7 @@ def celery_inspect_get_reserved(worker_names: list[str], app: Celery) -> set[str
     # get the list of reserved tasks
     reserved_tasks: dict[str, list] | None = inspect.reserved()
     if reserved_tasks:
-        for _, task_list in reserved_tasks.items():
+        for task_list in reserved_tasks.values():
             for task in task_list:
                 reserved_task_ids.add(task["id"])
 
@@ -235,7 +235,7 @@ def celery_inspect_get_active(worker_names: list[str], app: Celery) -> set[str]:
     # get the list of reserved tasks
     active_tasks: dict[str, list] | None = inspect.active()
     if active_tasks:
-        for _, task_list in active_tasks.items():
+        for task_list in active_tasks.values():
             for task in task_list:
                 active_task_ids.add(task["id"])
 

backend/onyx/background/celery/tasks/pruning/tasks.py

@@ -238,8 +238,7 @@ def check_for_pruning(self: Task, *, tenant_id: str) -> bool | None:
             cc_pair_ids: list[int] = []
             with get_session_with_current_tenant() as db_session:
                 cc_pairs = get_connector_credential_pairs(db_session)
-                for cc_pair_entry in cc_pairs:
-                    cc_pair_ids.append(cc_pair_entry.id)
+                cc_pair_ids.extend(cc_pair_entry.id for cc_pair_entry in cc_pairs)
 
             prune_dispatched = False
             for cc_pair_id in cc_pair_ids:

backend/onyx/background/celery/tasks/vespa/tasks.py

@@ -159,8 +159,7 @@ def check_for_vespa_sync_task(self: Task, *, tenant_id: str) -> bool | None:
                         db_session=db_session, only_up_to_date=False
                     )
 
-                    for usergroup in user_groups:
-                        usergroup_ids.append(usergroup.id)
+                    usergroup_ids.extend(usergroup.id for usergroup in user_groups)
 
                 for usergroup_id in usergroup_ids:
                     lock_beat.reacquire()

backend/onyx/chat/llm_loop.py

@@ -512,7 +512,6 @@ def construct_message_history(
     # Track dropped file messages so we can provide their metadata to the
     # FileReaderTool instead.
     truncated_history_before: list[ChatMessageSimple] = []
-    dropped_file_ids: list[str] = []
     current_token_count = 0
 
     for msg in reversed(history_before_last_user):
@@ -531,9 +530,11 @@ def construct_message_history(
     # recent messages, so the dropped ones are at the start of the original
     # list up to (len(history) - len(kept)).
     num_kept = len(truncated_history_before)
-    for msg in history_before_last_user[: len(history_before_last_user) - num_kept]:
-        if msg.file_id is not None:
-            dropped_file_ids.append(msg.file_id)
+    dropped_file_ids: list[str] = [
+        msg.file_id
+        for msg in history_before_last_user[: len(history_before_last_user) - num_kept]
+        if msg.file_id is not None
+    ]
 
     # Also treat "orphaned" metadata entries as dropped -- these are files
     # from messages removed by summary truncation (before convert_chat_history
@@ -663,10 +664,10 @@ def _create_file_tool_metadata_message(
         "read sections of any file. You MUST pass the file_id UUID (not the "
         "filename) to read_file:"
     ]
-    for meta in file_metadata:
-        lines.append(
-            f'- file_id="{meta.file_id}" filename="{meta.filename}" (~{meta.approx_char_count:,} chars)'
-        )
+    lines.extend(
+        f'- file_id="{meta.file_id}" filename="{meta.filename}" (~{meta.approx_char_count:,} chars)'
+        for meta in file_metadata
+    )
 
     message_content = "\n".join(lines)
     return ChatMessageSimple(

backend/onyx/connectors/axero/connector.py

@@ -212,8 +212,7 @@ def _get_forums(
         pages_fetched += len(contents)
         logger.debug("Fetched %s forums", pages_fetched)
 
-        for page in contents:
-            pages_to_return.append(page)
+        pages_to_return.extend(contents)
 
         if pages_fetched >= total_records:
             break

backend/onyx/connectors/canvas/connector.py

@@ -1021,33 +1021,31 @@ class CanvasConnector(
         start: SecondsSinceUnixEpoch | None,
         end: SecondsSinceUnixEpoch | None,
     ) -> list[SlimDocument]:
-        slim_docs: list[SlimDocument] = []
-
-        for page in self._list_pages(course_id):
-            slim_docs.append(
-                SlimDocument(
-                    id=f"canvas-page-{page.course_id}-{page.page_id}",
-                    external_access=self._get_item_permissions(course_id, page)
-                    or ExternalAccess.empty(),
-                )
+        slim_docs: list[SlimDocument] = [
+            SlimDocument(
+                id=f"canvas-page-{page.course_id}-{page.page_id}",
+                external_access=self._get_item_permissions(course_id, page)
+                or ExternalAccess.empty(),
             )
+            for page in self._list_pages(course_id)
+        ]
 
-        for assignment in self._list_assignments(course_id):
-            slim_docs.append(
-                SlimDocument(
-                    id=f"canvas-assignment-{assignment.course_id}-{assignment.id}",
-                    external_access=self._get_item_permissions(course_id, assignment)
-                    or ExternalAccess.empty(),
-                )
+        slim_docs.extend(
+            SlimDocument(
+                id=f"canvas-assignment-{assignment.course_id}-{assignment.id}",
+                external_access=self._get_item_permissions(course_id, assignment)
+                or ExternalAccess.empty(),
             )
+            for assignment in self._list_assignments(course_id)
+        )
 
-        for announcement in self._list_announcements(course_id, start, end):
-            slim_docs.append(
-                SlimDocument(
-                    id=f"canvas-announcement-{announcement.course_id}-{announcement.id}",
-                    external_access=self._get_item_permissions(course_id, announcement)
-                    or ExternalAccess.empty(),
-                )
+        slim_docs.extend(
+            SlimDocument(
+                id=f"canvas-announcement-{announcement.course_id}-{announcement.id}",
+                external_access=self._get_item_permissions(course_id, announcement)
+                or ExternalAccess.empty(),
             )
+            for announcement in self._list_announcements(course_id, start, end)
+        )
 
         return slim_docs

backend/onyx/connectors/coda/connector.py

@@ -231,9 +231,7 @@ class CodaConnector(LoadConnector, PollConnector):
             else:
                 raise
 
-        values = {}
-        for col_name, col_value in response.get("values", {}).items():
-            values[col_name] = col_value
+        values = dict(response.get("values", {}))
 
         return CodaRow(
             id=response["id"],
@@ -408,17 +406,17 @@ class CodaConnector(LoadConnector, PollConnector):
                     raise
 
             items = response.get("items", [])
-            for item in items:
-                tables.append(
-                    CodaTable(
-                        id=item["id"],
-                        browser_link=item["browserLink"],
-                        name=item["name"],
-                        created_at=item["createdAt"],
-                        updated_at=item["updatedAt"],
-                        doc_id=doc_id,
-                    )
+            tables.extend(
+                CodaTable(
+                    id=item["id"],
+                    browser_link=item["browserLink"],
+                    name=item["name"],
+                    created_at=item["createdAt"],
+                    updated_at=item["updatedAt"],
+                    doc_id=doc_id,
                 )
+                for item in items
+            )
 
             next_page_token = response.get("nextPageToken")
             if not next_page_token:
@@ -453,9 +451,7 @@ class CodaConnector(LoadConnector, PollConnector):
 
             items = response.get("items", [])
             for item in items:
-                values = {}
-                for col_name, col_value in item.get("values", {}).items():
-                    values[col_name] = col_value
+                values = dict(item.get("values", {}))
 
                 rows.append(
                     CodaRow(

backend/onyx/connectors/confluence/connector.py

@@ -570,8 +570,10 @@ class ConfluenceConnector(
             # Extract labels
             labels = []
             if "metadata" in page and "labels" in page["metadata"]:
-                for label in page["metadata"]["labels"].get("results", []):
-                    labels.append(label.get("name", ""))
+                labels.extend(
+                    label.get("name", "")
+                    for label in page["metadata"]["labels"].get("results", [])
+                )
             if labels:
                 metadata["labels"] = labels
 
@@ -722,10 +724,12 @@ class ConfluenceConnector(
                         )
                     labels: list[str] = []
                     if "metadata" in attachment and "labels" in attachment["metadata"]:
-                        for label in attachment["metadata"]["labels"].get(
-                            "results", []
-                        ):
-                            labels.append(label.get("name", ""))
+                        labels.extend(
+                            label.get("name", "")
+                            for label in attachment["metadata"]["labels"].get(
+                                "results", []
+                            )
+                        )
                     if labels:
                         attachment_metadata["labels"] = labels
                     page_url = page_url or build_confluence_document_id(
@@ -1187,8 +1191,7 @@ class ConfluenceConnector(
             )
 
         # Yield space hierarchy nodes first
-        for node in self._yield_space_hierarchy_nodes():
-            doc_metadata_list.append(node)
+        doc_metadata_list.extend(self._yield_space_hierarchy_nodes())
 
         # Per-page mode only: collapse shared ancestors to one GET each.
         ancestor_restrictions_cache: dict[str, dict[str, Any] | None] = {}
@@ -1223,8 +1226,7 @@ class ConfluenceConnector(
             limit=_SLIM_DOC_BATCH_SIZE,
         ):
             # Yield ancestor hierarchy nodes for this page
-            for node in self._yield_ancestor_hierarchy_nodes(page):
-                doc_metadata_list.append(node)
+            doc_metadata_list.extend(self._yield_ancestor_hierarchy_nodes(page))
 
             page_restrictions = page.get("restrictions") or {}
             page_space_key = page.get("space", {}).get("key")

backend/onyx/connectors/document360/connector.py

@@ -96,20 +96,20 @@ class Document360Connector(LoadConnector, PollConnector):
 
         for category in all_categories:
             if not self.categories or category["name"] in self.categories:
-                for article in category["articles"]:
-                    articles_with_category.append(
-                        {"id": article["id"], "category_name": category["name"]}
-                    )
+                articles_with_category.extend(
+                    {"id": article["id"], "category_name": category["name"]}
+                    for article in category["articles"]
+                )
                 for child_category in category["child_categories"]:
                     all_nested_categories = flatten_child_categories(child_category)
                     for nested_category in all_nested_categories:
-                        for article in nested_category["articles"]:
-                            articles_with_category.append(
-                                {
-                                    "id": article["id"],
-                                    "category_name": nested_category["name"],
-                                }
-                            )
+                        articles_with_category.extend(
+                            {
+                                "id": article["id"],
+                                "category_name": nested_category["name"],
+                            }
+                            for article in nested_category["articles"]
+                        )
 
         return articles_with_category
 

backend/onyx/connectors/fireflies/connector.py

@@ -108,10 +108,11 @@ def _create_doc_from_transcript(transcript: dict) -> Document | None:
     meeting_organizer_email = transcript["organizer_email"]
     organizer_email_user_info = [BasicExpertInfo(email=meeting_organizer_email)]
 
-    meeting_participants_email_list = []
-    for participant in transcript.get("participants", []):
-        if participant != meeting_organizer_email and participant:
-            meeting_participants_email_list.append(BasicExpertInfo(email=participant))
+    meeting_participants_email_list = [
+        BasicExpertInfo(email=participant)
+        for participant in transcript.get("participants", [])
+        if participant != meeting_organizer_email and participant
+    ]
 
     return Document(
         id=fireflies_id,

backend/onyx/connectors/gmail/connector.py

@@ -427,15 +427,16 @@ class GmailConnector(
 
         try:
             admin_service = get_admin_service(self.creds, self.primary_admin_email)
-            emails = []
-            for user in execute_paginated_retrieval(
-                retrieval_function=admin_service.users().list,  # ty: ignore[unresolved-attribute]
-                list_key="users",
-                fields=USER_FIELDS,
-                domain=self.google_domain,
-            ):
-                if email := user.get("primaryEmail"):
-                    emails.append(email)
+            emails = [
+                email
+                for user in execute_paginated_retrieval(
+                    retrieval_function=admin_service.users().list,  # ty: ignore[unresolved-attribute]
+                    list_key="users",
+                    fields=USER_FIELDS,
+                    domain=self.google_domain,
+                )
+                if (email := user.get("primaryEmail"))
+            ]
             return emails
 
         except HttpError as e:

backend/onyx/connectors/google_drive/connector.py

@@ -2096,14 +2096,18 @@ class GoogleDriveConnector(
             )
 
             # Build slim documents
-            for file in files_batch:
-                if doc := build_slim_document(
-                    self.creds,
-                    file.drive_file,
-                    permission_sync_context,
-                    retriever_email=file.user_email,
-                ):
-                    slim_batch.append(doc)
+            slim_batch.extend(
+                doc
+                for file in files_batch
+                if (
+                    doc := build_slim_document(
+                        self.creds,
+                        file.drive_file,
+                        permission_sync_context,
+                        retriever_email=file.user_email,
+                    )
+                )
+            )
 
             # Combine: hierarchy nodes first, then slim docs
             result: list[SlimDocument | HierarchyNode] = []

backend/onyx/connectors/hubspot/connector.py

@@ -718,8 +718,9 @@ class HubSpotConnector(LoadConnector, PollConnector):
             associated_notes = self._get_associated_notes(
                 api_client, ticket.id, "tickets"
             )
-            for note in associated_notes:
-                sections.append(self._create_object_section(note, "notes"))
+            sections.extend(
+                self._create_object_section(note, "notes") for note in associated_notes
+            )
 
             # Add association IDs to metadata
             if associated_contact_ids:
@@ -874,8 +875,9 @@ class HubSpotConnector(LoadConnector, PollConnector):
             associated_notes = self._get_associated_notes(
                 api_client, company.id, "companies"
             )
-            for note in associated_notes:
-                sections.append(self._create_object_section(note, "notes"))
+            sections.extend(
+                self._create_object_section(note, "notes") for note in associated_notes
+            )
 
             # Add association IDs to metadata
             if associated_contact_ids:
@@ -1028,8 +1030,9 @@ class HubSpotConnector(LoadConnector, PollConnector):
 
             # Get associated notes
             associated_notes = self._get_associated_notes(api_client, deal.id, "deals")
-            for note in associated_notes:
-                sections.append(self._create_object_section(note, "notes"))
+            sections.extend(
+                self._create_object_section(note, "notes") for note in associated_notes
+            )
 
             # Add association IDs to metadata
             if associated_contact_ids:
@@ -1202,8 +1205,9 @@ class HubSpotConnector(LoadConnector, PollConnector):
             associated_notes = self._get_associated_notes(
                 api_client, contact.id, "contacts"
             )
-            for note in associated_notes:
-                sections.append(self._create_object_section(note, "notes"))
+            sections.extend(
+                self._create_object_section(note, "notes") for note in associated_notes
+            )
 
             # Add association IDs to metadata
             if associated_company_ids:

backend/onyx/connectors/jira/connector.py

@@ -955,25 +955,28 @@ class JiraConnector(
 
                 # Yield hierarchy nodes BEFORE the slim document (parent-before-child)
                 # 1. Yield project hierarchy node (if not already yielded)
-                for node in self._yield_project_hierarchy_node(
-                    project_key, project_name, seen_hierarchy_node_ids
-                ):
-                    slim_doc_batch.append(node)
+                slim_doc_batch.extend(
+                    self._yield_project_hierarchy_node(
+                        project_key, project_name, seen_hierarchy_node_ids
+                    )
+                )
 
                 # 2. If parent is an Epic, yield hierarchy node for it
                 parent = best_effort_get_field_from_issue(issue, _FIELD_PARENT)
                 if parent:
-                    for node in self._yield_parent_hierarchy_node_if_epic(
-                        parent, project_key, seen_hierarchy_node_ids
-                    ):
-                        slim_doc_batch.append(node)
+                    slim_doc_batch.extend(
+                        self._yield_parent_hierarchy_node_if_epic(
+                            parent, project_key, seen_hierarchy_node_ids
+                        )
+                    )
 
                 # 3. If this issue IS an Epic, yield it as hierarchy node
                 if self._is_epic(issue):
-                    for node in self._yield_epic_hierarchy_node(
-                        issue, project_key, seen_hierarchy_node_ids
-                    ):
-                        slim_doc_batch.append(node)
+                    slim_doc_batch.extend(
+                        self._yield_epic_hierarchy_node(
+                            issue, project_key, seen_hierarchy_node_ids
+                        )
+                    )
 
                 # Now add the slim document
                 issue_key = best_effort_get_field_from_issue(issue, _FIELD_KEY)

backend/onyx/connectors/linear/connector.py

@@ -329,13 +329,13 @@ class LinearConnector(LoadConnector, PollConnector, OAuthConnector):
                 ]
 
                 # Add comment sections
-                for comment in node["comments"]["nodes"]:
-                    sections.append(
-                        TextSection(
-                            link=node["url"],
-                            text=comment["body"] or "",
-                        )
+                sections.extend(
+                    TextSection(
+                        link=node["url"],
+                        text=comment["body"] or "",
                     )
+                    for comment in node["comments"]["nodes"]
+                )
 
                 # Cast the sections list to the expected type
                 typed_sections = cast(list[TextSection | ImageSection], sections)

backend/onyx/connectors/notion/connector.py

@@ -879,7 +879,7 @@ class NotionConnector(LoadConnector, PollConnector, SlimConnector):
         page_title = None
         if hasattr(page, "database_name") and page.database_name:
             return page.database_name
-        for _, prop in page.properties.items():
+        for prop in page.properties.values():
             if prop["type"] == "title" and len(prop["title"]) > 0:
                 page_title = " ".join([t["plain_text"] for t in prop["title"]]).strip()
                 break

backend/onyx/connectors/outline/connector.py

@@ -192,10 +192,9 @@ class OutlineConnector(LoadConnector, PollConnector):
                 )
 
                 # Apply time filtering if specified
-                filtered_batch: list[Document | HierarchyNode] = []
-                for doc in doc_batch:
-                    if time_filter is None or time_filter(doc):
-                        filtered_batch.append(doc)
+                filtered_batch: list[Document | HierarchyNode] = [
+                    doc for doc in doc_batch if time_filter is None or time_filter(doc)
+                ]
 
                 start_ind += num_results
                 if filtered_batch:

backend/onyx/connectors/salesforce/doc_conversion.py

@@ -104,8 +104,7 @@ def _json_to_natural_language(data: dict | list, indent: int = 0) -> str:
             else:
                 result.append(f"{indent_str}{key}: {value}")
     elif isinstance(data, list):
-        for item in data:
-            result.append(_json_to_natural_language(item, indent + 2))
+        result.extend(_json_to_natural_language(item, indent + 2) for item in data)
 
     return "\n".join(result)
 

backend/onyx/connectors/salesforce/sqlite_functions.py

@@ -466,7 +466,7 @@ class OnyxSalesforceSQLite:
             parent_relationship_fields = parent_relationship_fields_by_type[
                 changed_type
             ]
-            for field_name, _ in parent_relationship_fields.items():
+            for field_name in parent_relationship_fields:
                 if field_name not in sf_object.data:
                     logger.warning(
                         "field_name=%r not in data for changed_type=%r!",

backend/onyx/connectors/xenforo/connector.py

@@ -40,16 +40,13 @@ def get_title(soup: BeautifulSoup) -> str:
 
 def get_pages(soup: BeautifulSoup, url: str) -> list[str]:
     page_tags = soup.select("li.pageNav-page")
-    page_numbers = []
-    for button in page_tags:
-        if re.match(r"^\d+$", button.text):
-            page_numbers.append(button.text)
+    page_numbers = [
+        button.text for button in page_tags if re.match(r"^\d+$", button.text)
+    ]
 
     max_pages = int(max(page_numbers, key=int)) if page_numbers else 1
 
-    all_pages = []
-    for x in range(1, int(max_pages) + 1):
-        all_pages.append(f"{url}page-{x}")
+    all_pages = [f"{url}page-{x}" for x in range(1, int(max_pages) + 1)]
     return all_pages
 
 

backend/onyx/context/search/federated/slack_search.py

@@ -1152,23 +1152,23 @@ def slack_retrieval(
             ),  # ALWAYS apply exclude_channels
         }
 
-        for query_string in query_strings:
-            search_tasks.append(
+        search_tasks.extend(
+            (
+                query_slack,
                 (
-                    query_slack,
-                    (
-                        query_string,
-                        access_token,
-                        query_limit,
-                        allowed_private_channel,
-                        bot_token,
-                        include_dm,
-                        dm_entities,
-                        available_channels,
-                        channel_metadata_dict,
-                    ),
-                )
+                    query_string,
+                    access_token,
+                    query_limit,
+                    allowed_private_channel,
+                    bot_token,
+                    include_dm,
+                    dm_entities,
+                    available_channels,
+                    channel_metadata_dict,
+                ),
             )
+            for query_string in query_strings
+        )
 
     # Execute searches in parallel
     results = run_functions_tuples_in_parallel(search_tasks)

backend/onyx/context/search/pipeline.py

@@ -174,7 +174,7 @@ def merge_individual_chunks(
     chunk_to_section: dict[tuple[str, int], InferenceSection] = {}
 
     # Process each document's chunks
-    for _doc_id, doc_chunk_list in doc_chunks.items():
+    for doc_chunk_list in doc_chunks.values():
         if not doc_chunk_list:
             continue
 

backend/onyx/context/search/retrieval/search_runner.py

@@ -94,8 +94,6 @@ def search_chunks(
     embedding_model: EmbeddingModel | None = None,
     prefetched_federated_retrieval_infos: list[FederatedRetrievalInfo] | None = None,
 ) -> list[InferenceChunk]:
-    run_queries: list[tuple[Callable, tuple]] = []
-
     source_filters = (
         set(query_request.filters.source_type)
         if query_request.filters.source_type
@@ -121,10 +119,10 @@ def search_chunks(
         federated_retrieval_info.source.to_non_federated_source()
         for federated_retrieval_info in federated_retrieval_infos
     )
-    for federated_retrieval_info in federated_retrieval_infos:
-        run_queries.append(
-            (federated_retrieval_info.retrieval_function, (query_request,))
-        )
+    run_queries: list[tuple[Callable, tuple]] = [
+        (federated_retrieval_info.retrieval_function, (query_request,))
+        for federated_retrieval_info in federated_retrieval_infos
+    ]
 
     # Don't run normal hybrid search if there are no indexed sources to
     # search over

backend/onyx/db/credentials.py

@@ -63,14 +63,13 @@ def _relate_credential_to_user_groups__no_commit(
     credential_id: int,
     user_group_ids: list[int],
 ) -> None:
-    credential_user_groups = []
-    for group_id in user_group_ids:
-        credential_user_groups.append(
-            Credential__UserGroup(
-                credential_id=credential_id,
-                user_group_id=group_id,
-            )
+    credential_user_groups = [
+        Credential__UserGroup(
+            credential_id=credential_id,
+            user_group_id=group_id,
         )
+        for group_id in user_group_ids
+    ]
     db_session.add_all(credential_user_groups)
 
 

backend/onyx/document_index/opensearch/client.py

@@ -398,22 +398,21 @@ class OpenSearchClient(AbstractContextManager):
             A list of IndexInfo objects for each index.
         """
         response = self._client.cat.indices(format="json")
-        indices: list[IndexInfo] = []
-        for raw_index_info in response:
-            indices.append(
-                IndexInfo(
-                    name=raw_index_info.get("index", ""),
-                    health=raw_index_info.get("health", ""),
-                    status=raw_index_info.get("status", ""),
-                    num_primary_shards=raw_index_info.get("pri", ""),
-                    num_replica_shards=raw_index_info.get("rep", ""),
-                    docs_count=raw_index_info.get("docs.count", ""),
-                    docs_deleted=raw_index_info.get("docs.deleted", ""),
-                    created_at=raw_index_info.get("creation.date.string", ""),
-                    total_size=raw_index_info.get("store.size", ""),
-                    primary_shards_size=raw_index_info.get("pri.store.size", ""),
-                )
+        indices: list[IndexInfo] = [
+            IndexInfo(
+                name=raw_index_info.get("index", ""),
+                health=raw_index_info.get("health", ""),
+                status=raw_index_info.get("status", ""),
+                num_primary_shards=raw_index_info.get("pri", ""),
+                num_replica_shards=raw_index_info.get("rep", ""),
+                docs_count=raw_index_info.get("docs.count", ""),
+                docs_deleted=raw_index_info.get("docs.deleted", ""),
+                created_at=raw_index_info.get("creation.date.string", ""),
+                total_size=raw_index_info.get("store.size", ""),
+                primary_shards_size=raw_index_info.get("pri.store.size", ""),
             )
+            for raw_index_info in response
+        ]
         return indices
 
     @log_function_time(print_only=True, debug_only=True, include_args=True)
@@ -1377,16 +1376,15 @@ class OpenSearchIndexClient(OpenSearchClient):
             len(document_chunk_ids),
             self._index_name,
         )
-        data = []
-        for document_chunk_id in document_chunk_ids:
-            data.append(
-                {
-                    "_index": self._index_name,
-                    "_id": document_chunk_id,
-                    "_op_type": "update",
-                    "doc": properties_to_update,
-                }
-            )
+        data = [
+            {
+                "_index": self._index_name,
+                "_id": document_chunk_id,
+                "_op_type": "update",
+                "doc": properties_to_update,
+            }
+            for document_chunk_id in document_chunk_ids
+        ]
         # max_retries is the number of times to retry a request if we get a 429.
         # We do not raise on error (the default behavior of ``bulk`` is to
         # raise) because we want to attempt to retry certain failed chunks in

backend/onyx/document_index/vespa/shared_utils/vespa_request_builders.py

@@ -85,9 +85,10 @@ def build_vespa_filters(
                 return f'"{entity}"'
 
         if kg_entities:
-            filter_parts = []
-            for kg_entity in kg_entities:
-                filter_parts.append(f"(kg_entities contains {_build_kge(kg_entity)})")
+            filter_parts = [
+                f"(kg_entities contains {_build_kge(kg_entity)})"
+                for kg_entity in kg_entities
+            ]
             combined_filter_parts.append(f"({' or '.join(filter_parts)})")
 
         # TODO: handle complex nested relationship logic (e.g., A participated, and B or C participated)

backend/onyx/evals/eval.py

@@ -292,19 +292,18 @@ def _get_multi_turn_answer_with_tools(
         raise ValueError("Multi-turn eval requires 'messages' array in input")
 
     # Parse messages into EvalMessage objects
-    messages: list[EvalMessage] = []
-    for msg_data in messages_data:
-        messages.append(
-            EvalMessage(
-                message=msg_data["message"],
-                expected_tools=msg_data.get("expected_tools", []),
-                require_all_tools=msg_data.get("require_all_tools", False),
-                model=msg_data.get("model"),
-                model_provider=msg_data.get("model_provider"),
-                temperature=msg_data.get("temperature"),
-                force_tools=msg_data.get("force_tools", []),
-            )
+    messages: list[EvalMessage] = [
+        EvalMessage(
+            message=msg_data["message"],
+            expected_tools=msg_data.get("expected_tools", []),
+            require_all_tools=msg_data.get("require_all_tools", False),
+            model=msg_data.get("model"),
+            model_provider=msg_data.get("model_provider"),
+            temperature=msg_data.get("temperature"),
+            force_tools=msg_data.get("force_tools", []),
         )
+        for msg_data in messages_data
+    ]
 
     turn_results: list[EvalToolResult] = []
 

backend/onyx/external_apps/matching/graphql_parsing.py

@@ -67,8 +67,10 @@ def _operation_invocations(query: str) -> list[tuple[str, str]]:
         if not isinstance(definition, OperationDefinitionNode):
             continue
         operation_type = definition.operation.value  # "query" | "mutation" | ...
-        for field in _root_field_names(definition.selection_set, fragments, set()):
-            invocations.append((operation_type, field))
+        invocations.extend(
+            (operation_type, field)
+            for field in _root_field_names(definition.selection_set, fragments, set())
+        )
     return invocations
 
 

backend/onyx/indexing/chunking/tabular_section_chunker/tabular_section_chunker.py

@@ -113,13 +113,13 @@ def _split_row_by_pairs(
             current_tokens = 0
 
         if pair_tokens > max_tokens:
-            for split_text in split_text_by_tokens(pair_str, tokenizer, max_tokens):
-                pieces.append(
-                    _TokenizedText(
-                        text=split_text,
-                        token_count=count_tokens(split_text, tokenizer),
-                    )
+            pieces.extend(
+                _TokenizedText(
+                    text=split_text,
+                    token_count=count_tokens(split_text, tokenizer),
                 )
+                for split_text in split_text_by_tokens(pair_str, tokenizer, max_tokens)
+            )
         else:
             current_parts = [pair_str]
             current_tokens = pair_tokens

backend/onyx/indexing/chunking/tabular_section_chunker/total_descriptor.py

@@ -24,9 +24,10 @@ def build_total_descriptor_chunks(
     if analysis.row_count == 0:
         return []
 
-    lines: list[str] = []
-    for idx in analysis.numeric_cols:
-        lines.append(_numeric_totals_line(headers[idx], analysis.numeric_stats[idx]))
+    lines: list[str] = [
+        _numeric_totals_line(headers[idx], analysis.numeric_stats[idx])
+        for idx in analysis.numeric_cols
+    ]
     for idx in analysis.categorical_cols:
         line = _categorical_top_line(headers[idx], analysis.categorical_counts[idx])
         if line:

backend/onyx/llm/model_response.py

@@ -105,16 +105,15 @@ def _parse_delta_tool_calls(
     if not tool_calls:
         return []
 
-    parsed_tool_calls: list[ChatCompletionDeltaToolCall] = []
-    for tool_call in tool_calls:
-        parsed_tool_calls.append(
-            ChatCompletionDeltaToolCall(
-                id=tool_call.get("id"),
-                index=tool_call.get("index", 0),
-                type=tool_call.get("type", "function"),
-                function=_parse_function_call(tool_call.get("function")),
-            )
+    parsed_tool_calls: list[ChatCompletionDeltaToolCall] = [
+        ChatCompletionDeltaToolCall(
+            id=tool_call.get("id"),
+            index=tool_call.get("index", 0),
+            type=tool_call.get("type", "function"),
+            function=_parse_function_call(tool_call.get("function")),
         )
+        for tool_call in tool_calls
+    ]
     return parsed_tool_calls
 
 

backend/onyx/onyxbot/discord/handle_message.py

@@ -327,12 +327,15 @@ async def _build_thread_context(
 
     try:
         thread = message.channel
-        messages: list[discord.Message] = []
 
         # Fetch recent messages (excluding current)
-        async for msg in thread.history(limit=MAX_CONTEXT_MESSAGES, oldest_first=False):
-            if msg.id != message.id:
-                messages.append(msg)
+        messages: list[discord.Message] = [
+            msg
+            async for msg in thread.history(
+                limit=MAX_CONTEXT_MESSAGES, oldest_first=False
+            )
+            if msg.id != message.id
+        ]
 
         # Include thread starter message and its reply chain if not already present
         if thread.parent and not isinstance(thread.parent, discord.ForumChannel):

backend/onyx/onyxbot/slack/utils.py

@@ -401,7 +401,7 @@ def get_view_values(state_values: dict[str, Any]) -> dict[str, Any]:
         dict: keys/values of the view state content
     """
     view_values = {}
-    for _, view_data in state_values.items():
+    for view_data in state_values.values():
         for k, v in view_data.items():
             if (
                 "selected_option" in v

backend/onyx/server/documents/connector.py

@@ -1183,14 +1183,14 @@ def get_connector_indexing_status(
     # route is global, so a scoped manager would just 403 on click.
     federated_statuses: list[FederatedConnectorStatus] = []
     if is_connectors_admin:
-        for federated_connector in federated_connectors:
-            federated_statuses.append(
-                FederatedConnectorStatus(
-                    id=federated_connector.id,
-                    source=federated_connector.source,
-                    name=f"{federated_connector.source.replace('_', ' ').title()}",
-                )
+        federated_statuses.extend(
+            FederatedConnectorStatus(
+                id=federated_connector.id,
+                source=federated_connector.source,
+                name=f"{federated_connector.source.replace('_', ' ').title()}",
             )
+            for federated_connector in federated_connectors
+        )
 
     source_to_summary: dict[DocumentSource, SourceSummary] = {}
 

backend/onyx/server/features/build/sandbox/kubernetes/kubernetes_sandbox_manager.py

@@ -603,14 +603,14 @@ class KubernetesSandboxManager(SandboxManager):
         ]
 
         # Add ports for session Next.js servers (one port per potential session)
-        for port in range(SANDBOX_NEXTJS_PORT_START, SANDBOX_NEXTJS_PORT_END):
-            ports.append(
-                client.V1ServicePort(
-                    name=f"nextjs-{port}",
-                    port=port,
-                    target_port=port,
-                )
+        ports.extend(
+            client.V1ServicePort(
+                name=f"nextjs-{port}",
+                port=port,
+                target_port=port,
             )
+            for port in range(SANDBOX_NEXTJS_PORT_START, SANDBOX_NEXTJS_PORT_END)
+        )
 
         return client.V1Service(
             api_version="v1",

backend/onyx/server/features/build/session/md_to_docx.py

@@ -590,8 +590,9 @@ def _render_table(
         if section_type == "table_head":
             header_cells = section.get("children", [])
         elif section_type == "table_body":
-            for row in section.get("children", []):
-                body_rows.append(row.get("children", []))
+            body_rows.extend(
+                row.get("children", []) for row in section.get("children", [])
+            )
 
     num_cols = len(header_cells) or (len(body_rows[0]) if body_rows else 0)
     if num_cols == 0:

backend/onyx/server/features/build/session/streaming.py

@@ -235,10 +235,11 @@ def _extract_text_from_content(content: Any) -> str:
     if hasattr(content, "type") and content.type == "text":
         return getattr(content, "text", "") or ""
     if isinstance(content, list):
-        texts = []
-        for block in content:
-            if hasattr(block, "type") and block.type == "text":
-                texts.append(getattr(block, "text", "") or "")
+        texts = [
+            getattr(block, "text", "") or ""
+            for block in content
+            if hasattr(block, "type") and block.type == "text"
+        ]
         return "".join(texts)
     return ""
 

backend/onyx/server/features/web_search/api.py

@@ -176,16 +176,16 @@ def _run_web_search(
             list(search_results)
         )
         trimmed_results = list(filtered_results)[: request.max_results]
-        for search_result in trimmed_results:
-            results.append(
-                LlmWebSearchResult(
-                    document_citation_number=DOCUMENT_CITATION_NUMBER_EMPTY_VALUE,
-                    url=search_result.link,
-                    title=search_result.title,
-                    snippet=search_result.snippet or "",
-                    unique_identifier_to_strip_away=search_result.link,
-                )
+        results.extend(
+            LlmWebSearchResult(
+                document_citation_number=DOCUMENT_CITATION_NUMBER_EMPTY_VALUE,
+                url=search_result.link,
+                title=search_result.title,
+                snippet=search_result.snippet or "",
+                unique_identifier_to_strip_away=search_result.link,
             )
+            for search_result in trimmed_results
+        )
     return provider_view.provider_type, results
 
 
@@ -211,15 +211,14 @@ def _open_urls(
             "Web content provider failed to fetch URLs.",
         ) from exc
 
-    results: list[LlmOpenUrlResult] = []
-    for doc in docs:
-        results.append(
-            LlmOpenUrlResult(
-                document_citation_number=DOCUMENT_CITATION_NUMBER_EMPTY_VALUE,
-                content=truncate_search_result_content(doc.full_content),
-                unique_identifier_to_strip_away=doc.link,
-            )
+    results: list[LlmOpenUrlResult] = [
+        LlmOpenUrlResult(
+            document_citation_number=DOCUMENT_CITATION_NUMBER_EMPTY_VALUE,
+            content=truncate_search_result_content(doc.full_content),
+            unique_identifier_to_strip_away=doc.link,
         )
+        for doc in docs
+    ]
     provider_type = (
         provider_view.provider_type
         if provider_view

backend/onyx/server/federated/api.py

@@ -541,17 +541,14 @@ def get_federated_connector_detail(
             break
 
     # Get document set mappings
-    document_sets = []
-    for mapping in federated_connector.document_sets:
-        document_sets.append(
-            {
-                "id": mapping.document_set_id,
-                "name": (
-                    mapping.document_set.name if mapping.document_set else "Unknown"
-                ),
-                "entities": mapping.entities,
-            }
-        )
+    document_sets = [
+        {
+            "id": mapping.document_set_id,
+            "name": (mapping.document_set.name if mapping.document_set else "Unknown"),
+            "entities": mapping.entities,
+        }
+        for mapping in federated_connector.document_sets
+    ]
 
     return FederatedConnectorDetail(
         id=federated_connector.id,

backend/onyx/server/manage/llm/api.py

@@ -882,19 +882,19 @@ def list_llm_provider_basics(
 
     all_providers = fetch_existing_llm_providers(db_session, [])
 
-    accessible_providers = []
-
-    for provider in all_providers:
-        # Use centralized access control logic with persona=None since we're
-        # listing providers without a specific persona context. This correctly:
-        # - Includes public providers WITHOUT persona restrictions
-        # - Includes providers user can access via group membership
-        # - Excludes providers with persona restrictions (requires specific persona)
-        # - Excludes non-public providers with no restrictions (admin-only)
+    # Use centralized access control logic with persona=None since we're
+    # listing providers without a specific persona context. This correctly:
+    # - Includes public providers WITHOUT persona restrictions
+    # - Includes providers user can access via group membership
+    # - Excludes providers with persona restrictions (requires specific persona)
+    # - Excludes non-public providers with no restrictions (admin-only)
+    accessible_providers = [
+        LLMProviderDescriptor.from_model(provider)
+        for provider in all_providers
         if can_user_access_llm_provider(
             provider, user_group_ids, persona=None, can_manage_llms=can_manage_llms
-        ):
-            accessible_providers.append(LLMProviderDescriptor.from_model(provider))
+        )
+    ]
 
     end_time = datetime.now(timezone.utc)
     duration = (end_time - start_time).total_seconds()
@@ -956,9 +956,11 @@ def get_valid_model_names_for_persona(
             llm_provider_model, user_group_ids, persona, can_manage_llms=can_manage_llms
         ):
             # Collect all model names from this provider
-            for model_config in llm_provider_model.model_configurations:
-                if model_config.is_visible:
-                    valid_models.append(model_config.name)
+            valid_models.extend(
+                model_config.name
+                for model_config in llm_provider_model.model_configurations
+                if model_config.is_visible
+            )
 
     return valid_models
 
@@ -1036,16 +1038,14 @@ def list_llm_providers_for_persona(
         db_session, [LLMModelFlowType.CHAT, LLMModelFlowType.VISION]
     )
 
-    llm_provider_list: list[LLMProviderDescriptor] = []
-
-    for llm_provider_model in all_providers:
-        # Check access with persona context — respects persona restrictions
+    # Check access with persona context — respects persona restrictions
+    llm_provider_list: list[LLMProviderDescriptor] = [
+        LLMProviderDescriptor.from_model(llm_provider_model)
+        for llm_provider_model in all_providers
         if can_user_access_llm_provider(
             llm_provider_model, user_group_ids, persona, can_manage_llms=can_manage_llms
-        ):
-            llm_provider_list.append(
-                LLMProviderDescriptor.from_model(llm_provider_model)
-            )
+        )
+    ]
 
     end_time = datetime.now(timezone.utc)
     duration = (end_time - start_time).total_seconds()

backend/onyx/server/query_and_chat/chat_backend.py

@@ -432,15 +432,11 @@ def get_chat_session(
 
     # Every assistant message might have a set of tool calls associated with it, these need to be replayed back for the frontend
     # Each list is the set of tool calls for the given assistant message.
-    replay_packet_lists: list[list[Packet]] = []
-    for msg in session_messages:
-        if msg.message_type == MessageType.ASSISTANT:
-            replay_packet_lists.append(
-                translate_assistant_message_to_packets(
-                    chat_message=msg, db_session=db_session
-                )
-            )
-            # msg_packet_list.append(Packet(ind=end_step_nr, obj=OverallStop()))
+    replay_packet_lists: list[list[Packet]] = [
+        translate_assistant_message_to_packets(chat_message=msg, db_session=db_session)
+        for msg in session_messages
+        if msg.message_type == MessageType.ASSISTANT
+    ]
 
     return ChatSessionDetailResponse(
         chat_session_id=session_id,

backend/onyx/server/query_and_chat/session_loading.py

@@ -120,16 +120,14 @@ def create_message_packets(
 def create_citation_packets(
     citation_info_list: list[CitationInfo], turn_index: int
 ) -> list[Packet]:
-    packets: list[Packet] = []
-
     # Emit each citation as a separate CitationInfo packet
-    for citation_info in citation_info_list:
-        packets.append(
-            Packet(
-                placement=Placement(turn_index=turn_index),
-                obj=citation_info,
-            )
+    packets: list[Packet] = [
+        Packet(
+            placement=Placement(turn_index=turn_index),
+            obj=citation_info,
         )
+        for citation_info in citation_info_list
+    ]
 
     packets.append(Packet(placement=Placement(turn_index=turn_index), obj=SectionEnd()))
 

backend/onyx/skills/builtin/google-drive/gslides_api.py

@@ -93,11 +93,15 @@ def _element_text(element: dict[str, Any]) -> str:
     if shape_text:
         parts.append(_text_runs(shape_text))
     for row in element.get("table", {}).get("tableRows", []):
-        for cell in row.get("tableCells", []):
-            if cell.get("text"):
-                parts.append(_text_runs(cell["text"]))
-    for child in element.get("elementGroup", {}).get("children", []):
-        parts.append(_element_text(child))
+        parts.extend(
+            _text_runs(cell["text"])
+            for cell in row.get("tableCells", [])
+            if cell.get("text")
+        )
+    parts.extend(
+        _element_text(child)
+        for child in element.get("elementGroup", {}).get("children", [])
+    )
     return "".join(parts)
 
 

backend/onyx/skills/builtin/pptx/scripts/office/validators/base.py

@@ -301,14 +301,13 @@ class BaseSchemaValidator:
                 print("PASSED - No .rels files found")
             return True
 
-        all_files = []
-        for file_path in self.unpacked_dir.rglob("*"):
-            if (
-                file_path.is_file()
-                and file_path.name != "[Content_Types].xml"
-                and not file_path.name.endswith(".rels")
-            ):
-                all_files.append(file_path.resolve())
+        all_files = [
+            file_path.resolve()
+            for file_path in self.unpacked_dir.rglob("*")
+            if file_path.is_file()
+            and file_path.name != "[Content_Types].xml"
+            and not file_path.name.endswith(".rels")
+        ]
 
         all_referenced_files = set()
 
@@ -665,10 +664,10 @@ class BaseSchemaValidator:
                 continue
 
             new_errors.append(f"  {relative_path}: {len(new_file_errors)} new error(s)")
-            for error in list(new_file_errors)[:3]:
-                new_errors.append(
-                    f"    - {error[:250]}..." if len(error) > 250 else f"    - {error}"
-                )
+            new_errors.extend(
+                f"    - {error[:250]}..." if len(error) > 250 else f"    - {error}"
+                for error in list(new_file_errors)[:3]
+            )
 
         if self.verbose:
             print(f"Validated {len(self.xml_files)} files:")
@@ -831,10 +830,10 @@ class BaseSchemaValidator:
                 return text
             matches = list(template_pattern.finditer(text))
             if matches:
-                for match in matches:
-                    warnings.append(
-                        f"Found template tag in {content_type}: {match.group()}"
-                    )
+                warnings.extend(
+                    f"Found template tag in {content_type}: {match.group()}"
+                    for match in matches
+                )
                 return template_pattern.sub("", text)
             return text
 

backend/onyx/skills/builtin/pptx/scripts/office/validators/docx.py

@@ -332,20 +332,20 @@ class DOCXSchemaValidator(BaseSchemaValidator):
             }
 
             orphaned_ends = range_ends - range_starts
-            for comment_id in sorted(
-                orphaned_ends, key=lambda x: int(x) if x and x.isdigit() else 0
-            ):
-                errors.append(
-                    f'  document.xml: commentRangeEnd id="{comment_id}" has no matching commentRangeStart'
+            errors.extend(
+                f'  document.xml: commentRangeEnd id="{comment_id}" has no matching commentRangeStart'
+                for comment_id in sorted(
+                    orphaned_ends, key=lambda x: int(x) if x and x.isdigit() else 0
                 )
+            )
 
             orphaned_starts = range_starts - range_ends
-            for comment_id in sorted(
-                orphaned_starts, key=lambda x: int(x) if x and x.isdigit() else 0
-            ):
-                errors.append(
-                    f'  document.xml: commentRangeStart id="{comment_id}" has no matching commentRangeEnd'
+            errors.extend(
+                f'  document.xml: commentRangeStart id="{comment_id}" has no matching commentRangeEnd'
+                for comment_id in sorted(
+                    orphaned_starts, key=lambda x: int(x) if x and x.isdigit() else 0
                 )
+            )
 
             comment_ids = set()
             if comments_xml and comments_xml.exists():
@@ -359,13 +359,13 @@ class DOCXSchemaValidator(BaseSchemaValidator):
 
                 marker_ids = range_starts | range_ends | references
                 invalid_refs = marker_ids - comment_ids
-                for comment_id in sorted(
-                    invalid_refs, key=lambda x: int(x) if x and x.isdigit() else 0
-                ):
-                    if comment_id:
-                        errors.append(
-                            f'  document.xml: marker id="{comment_id}" references non-existent comment'
-                        )
+                errors.extend(
+                    f'  document.xml: marker id="{comment_id}" references non-existent comment'
+                    for comment_id in sorted(
+                        invalid_refs, key=lambda x: int(x) if x and x.isdigit() else 0
+                    )
+                    if comment_id
+                )
 
         except (lxml.etree.XMLSyntaxError, Exception) as e:
             errors.append(f"  Error parsing XML: {e}")

backend/onyx/skills/builtin/pptx/scripts/office/validators/redlining.py

@@ -204,10 +204,11 @@ class RedliningValidator:
         author_attr = f"{{{self.namespaces['w']}}}author"
 
         for parent in root.iter():
-            to_remove = []
-            for child in parent:
-                if child.tag == ins_tag and child.get(author_attr) == self.author:
-                    to_remove.append(child)
+            to_remove = [
+                child
+                for child in parent
+                if child.tag == ins_tag and child.get(author_attr) == self.author
+            ]
             for elem in to_remove:
                 parent.remove(elem)
 
@@ -215,10 +216,11 @@ class RedliningValidator:
         t_tag = f"{{{self.namespaces['w']}}}t"
 
         for parent in root.iter():
-            to_process = []
-            for child in parent:
-                if child.tag == del_tag and child.get(author_attr) == self.author:
-                    to_process.append((child, list(parent).index(child)))
+            to_process = [
+                (child, list(parent).index(child))
+                for child in parent
+                if child.tag == del_tag and child.get(author_attr) == self.author
+            ]
 
             for del_elem, del_index in reversed(to_process):
                 for elem in del_elem.iter():
@@ -235,10 +237,9 @@ class RedliningValidator:
 
         paragraphs = []
         for p_elem in root.findall(f".//{p_tag}"):
-            text_parts = []
-            for t_elem in p_elem.findall(f".//{t_tag}"):
-                if t_elem.text:
-                    text_parts.append(t_elem.text)
+            text_parts = [
+                t_elem.text for t_elem in p_elem.findall(f".//{t_tag}") if t_elem.text
+            ]
             paragraph_text = "".join(text_parts)
             if paragraph_text:
                 paragraphs.append(paragraph_text)

backend/onyx/tools/tool_implementations/open_url/open_url_tool.py

@@ -305,7 +305,7 @@ def _resolve_urls_to_document_ids(
 def _estimate_result_chars(result: dict[str, Any]) -> int:
     """Estimate character count from document fields in a result dict."""
     total = 0
-    for _key, value in result.items():
+    for value in result.values():
         if value is not None:
             total += len(str(value))
     return total

backend/onyx/tools/tool_implementations/open_url/snippet_matcher.py

@@ -114,8 +114,7 @@ def _normalize_text_with_mapping(text: str) -> tuple[str, list[int]]:
     nfd_to_orig: list[int] = []
     for orig_idx, orig_char in enumerate(original_text):
         nfd_of_char = unicodedata.normalize("NFD", orig_char)
-        for _ in nfd_of_char:
-            nfd_to_orig.append(orig_idx)
+        nfd_to_orig.extend([orig_idx] * len(nfd_of_char))
 
     # Map NFC positions → NFD positions.
     # Each NFC char, when decomposed, tells us exactly how many NFD

backend/onyx/tools/tool_implementations/open_url/utils.py

@@ -10,8 +10,9 @@ def filter_web_contents_with_no_title_or_content(
     Downstream uses these fields for display + prompting; drop empty ones centrally
     rather than duplicating checks across provider clients.
     """
-    filtered: list[WebContent] = []
-    for content in contents:
-        if content.title.strip() or content.full_content.strip():
-            filtered.append(content)
+    filtered: list[WebContent] = [
+        content
+        for content in contents
+        if content.title.strip() or content.full_content.strip()
+    ]
     return filtered

backend/onyx/tools/tool_implementations/search/search_tool.py

@@ -923,12 +923,12 @@ class SearchTool(Tool[SearchToolOverrideKwargs]):
             if semantic_query
             else []
         )
-        for llm_query in llm_queries:
-            # In rare cases, the LLM may fail to provide real queries
-            if llm_query:
-                semantic_queries_with_weights.append(
-                    (llm_query, LLM_NON_CUSTOM_QUERY_WEIGHT)
-                )
+        # In rare cases, the LLM may fail to provide real queries
+        semantic_queries_with_weights.extend(
+            (llm_query, LLM_NON_CUSTOM_QUERY_WEIGHT)
+            for llm_query in llm_queries
+            if llm_query
+        )
         if override_kwargs.original_query:
             semantic_queries_with_weights.append(
                 (override_kwargs.original_query, ORIGINAL_QUERY_WEIGHT)

backend/onyx/tools/tool_implementations/search/search_utils.py

@@ -243,7 +243,7 @@ def merge_overlapping_sections(
     merged_sections: dict[tuple[str, int], InferenceSection] = {}
 
     # Process each document's sections
-    for _doc_id, doc_section_list in doc_sections.items():
+    for doc_section_list in doc_sections.values():
         if not doc_section_list:
             continue
 

backend/onyx/tools/tool_implementations/web_search/utils.py

@@ -24,10 +24,9 @@ def filter_web_search_results_with_no_title_or_snippet(
     titles/snippets for display and prompting, so we drop those empty entries
     centrally (rather than duplicating the check in each client).
     """
-    filtered: list[WebSearchResult] = []
-    for result in results:
-        if result.title.strip() or result.snippet.strip():
-            filtered.append(result)
+    filtered: list[WebSearchResult] = [
+        result for result in results if result.title.strip() or result.snippet.strip()
+    ]
     return filtered
 
 

backend/onyx/utils/middleware.py

@@ -88,10 +88,11 @@ def _build_route_map(app: FastAPI) -> list[tuple[re.Pattern[str], str]]:
     Used by endpoint context middleware to resolve request paths to route
     templates, avoiding high-cardinality raw paths in metrics labels.
     """
-    route_map: list[tuple[re.Pattern[str], str]] = []
-    for route in app.routes:
-        if isinstance(route, APIRoute):
-            route_map.append((route.path_regex, route.path))
+    route_map: list[tuple[re.Pattern[str], str]] = [
+        (route.path_regex, route.path)
+        for route in app.routes
+        if isinstance(route, APIRoute)
+    ]
     return route_map
 
 

backend/scripts/query_time_check/seed_dummy_docs.py

@@ -76,9 +76,7 @@ def generate_dummy_chunk(
         image_file_id=None,
     )
 
-    document_set_names = []
-    for i in range(number_of_document_sets):
-        document_set_names.append(f"Document Set {i}")
+    document_set_names = [f"Document Set {i}" for i in range(number_of_document_sets)]
 
     user_emails: list[str | None] = []
     user_groups: list[str] = []

backend/scripts/tenant_cleanup/analyze_current_tenants.py

@@ -132,8 +132,7 @@ def collect_control_plane_data() -> list[dict[str, Any]]:
     control_plane_data = []
     with open("control_plane_data.csv", "r", newline="", encoding="utf-8") as csvfile:
         reader = csv.DictReader(csvfile)
-        for row in reader:
-            control_plane_data.append(row)
+        control_plane_data.extend(reader)
 
     return control_plane_data
 

backend/tests/daily/connectors/airtable/test_airtable_basic.py

@@ -345,9 +345,7 @@ def test_airtable_connector_index_all(
 
     all_docs: list[Document] = []
     for batch in connector.load_from_state():
-        for item in batch:
-            if isinstance(item, Document):
-                all_docs.append(item)
+        all_docs.extend(item for item in batch if isinstance(item, Document))
 
     # 2 from Tickets + 4 from Support Categories + 1 from Table 3 = 7
     assert len(all_docs) == 7

backend/tests/daily/connectors/coda/test_coda_connector.py

@@ -258,9 +258,7 @@ class TestLoadFromState:
         reference_data: dict[str, Any],  # noqa: ARG002
     ) -> None:
         """Test that no documents are yielded twice."""
-        document_ids = []
-        for doc in connector_doc_generator(connector):
-            document_ids.append(doc.id)
+        document_ids = [doc.id for doc in connector_doc_generator(connector)]
 
         unique_ids = set(document_ids)
         assert len(document_ids) == len(unique_ids), (

backend/tests/daily/connectors/discord/test_discord_connector.py

@@ -28,10 +28,9 @@ def test_discord_connector_basic(discord_connector: DiscordConnector) -> None:
 
     doc_batch = next(doc_batch_generator)
 
-    docs: list[Document] = []
-    for doc in doc_batch:
-        if not isinstance(doc, HierarchyNode):
-            docs.append(doc)
+    docs: list[Document] = [
+        doc for doc in doc_batch if not isinstance(doc, HierarchyNode)
+    ]
 
     assert len(docs) > 0, "No documents were retrieved from the connector"
 

backend/tests/external_dependency_unit/answer/test_answer_without_openai.py

@@ -77,12 +77,12 @@ def test_answer_with_only_anthropic_provider(
             chat_session_id=chat_session.id,
         )
 
-        response_stream: list[AnswerStreamPart] = []
-        for packet in handle_stream_message_objects(
-            new_msg_req=chat_request,
-            user=test_user,
-        ):
-            response_stream.append(packet)
+        response_stream: list[AnswerStreamPart] = list(
+            handle_stream_message_objects(
+                new_msg_req=chat_request,
+                user=test_user,
+            )
+        )
 
         assert response_stream, "Should receive streamed packets"
         assert not any(

backend/tests/external_dependency_unit/background/test_periodic_task_claim.py

@@ -193,8 +193,7 @@ class TestClaimConcurrency:
         results: list[bool] = []
         with ThreadPoolExecutor(max_workers=num_threads) as executor:
             futures = [executor.submit(claim) for _ in range(num_threads)]
-            for future in as_completed(futures):
-                results.append(future.result())
+            results.extend(future.result() for future in as_completed(futures))
 
         winners = sum(1 for r in results if r)
         assert winners == 1, f"Expected 1 winner, got {winners}"

backend/tests/external_dependency_unit/connectors/google_drive/test_google_drive_group_sync.py

@@ -435,9 +435,8 @@ class TestPerformExternalGroupSync:
     def test_batch_processing(self, db_session: Session) -> None:
         """Test that large numbers of groups are processed in batches"""
         # Create many test users
-        users = []
-        for i in range(150):  # More than the batch size of 100
-            users.append(_create_ext_perm_user(db_session, f"user{i}"))
+        # More than the batch size of 100
+        users = [_create_ext_perm_user(db_session, f"user{i}") for i in range(150)]
 
         cc_pair = _create_test_connector_credential_pair(db_session)
 

backend/tests/external_dependency_unit/db/test_tag_race_condition.py

@@ -249,11 +249,12 @@ class TestTagRaceCondition:
 
         # Run both types of operations concurrently
         with ThreadPoolExecutor(max_workers=num_documents * 2) as executor:
-            futures: list[Future[Union[Tag | None] | list[Tag]]] = []
-            for doc_id in doc_ids_single:
-                futures.append(executor.submit(create_single_tag, doc_id))
-            for doc_id in doc_ids_list:
-                futures.append(executor.submit(create_list_tag, doc_id))
+            futures: list[Future[Union[Tag | None] | list[Tag]]] = [
+                executor.submit(create_single_tag, doc_id) for doc_id in doc_ids_single
+            ]
+            futures.extend(
+                executor.submit(create_list_tag, doc_id) for doc_id in doc_ids_list
+            )
 
             for future in as_completed(futures):
                 try:

backend/tests/external_dependency_unit/hierarchy/test_upsert_parents_deep_chain.py

@@ -48,16 +48,15 @@ def test_upsert_parents_handles_deep_chain(
     """A 1500-deep parent chain must upsert without RecursionError."""
     tag = uuid4().hex[:8]
     # Build chain: root -> n1 -> n2 -> ... -> n{DEPTH-1}
-    nodes: list[PydanticHierarchyNode] = []
-    for i in range(DEEP_CHAIN_DEPTH):
-        nodes.append(
-            PydanticHierarchyNode(
-                raw_node_id=f"deep_{tag}_{i}",
-                raw_parent_id=f"deep_{tag}_{i - 1}" if i > 0 else None,
-                display_name=f"Deep {i}",
-                node_type=HierarchyNodeType.PAGE,
-            )
+    nodes: list[PydanticHierarchyNode] = [
+        PydanticHierarchyNode(
+            raw_node_id=f"deep_{tag}_{i}",
+            raw_parent_id=f"deep_{tag}_{i - 1}" if i > 0 else None,
+            display_name=f"Deep {i}",
+            node_type=HierarchyNodeType.PAGE,
         )
+        for i in range(DEEP_CHAIN_DEPTH)
+    ]
 
     # Pass the deepest child first so upsert_parents has to walk the entire
     # chain in one call (this is the worst-case shape that originally blew the

backend/tests/external_dependency_unit/llm/test_llm_provider.py

@@ -527,13 +527,9 @@ class TestDefaultProviderEndpoint:
             existing_providers = fetch_existing_llm_providers(
                 db_session, flow_type_filter=[LLMModelFlowType.CHAT]
             )
-            provider_names_to_restore: list[str] = []
 
-            for provider in existing_providers:
-                if provider.name is not None:
-                    provider_names_to_restore.append(provider.name)
-
-            # Remove all providers temporarily
+            # Remove all providers temporarily. The `finally` rollback restores
+            # them, since none of these deletes are committed.
             for provider in existing_providers:
                 remove_llm_provider(db_session, provider.id)
 

backend/tests/external_dependency_unit/server/security/test_security_settings_store.py

@@ -165,8 +165,9 @@ def test_thread_safe_concurrent_reads() -> None:
 
     def worker() -> None:
         try:
-            for _ in range(50):
-                results.append(get_security_settings().user_directory_admin_only)
+            results.extend(
+                get_security_settings().user_directory_admin_only for _ in range(50)
+            )
         except BaseException as e:  # noqa: BLE001
             errors.append(e)
 

backend/tests/integration/common_utils/managers/cc_pair.py

@@ -199,8 +199,9 @@ class CCPairManager:
         indexing_statuses = []
         for connectors_by_source in indexing_status_response:
             connectors = connectors_by_source["indexing_statuses"]
-            for connector in connectors:
-                indexing_statuses.append(ConnectorIndexingStatusLite(**connector))
+            indexing_statuses.extend(
+                ConnectorIndexingStatusLite(**connector) for connector in connectors
+            )
         return indexing_statuses
 
     @staticmethod

backend/tests/integration/common_utils/managers/file.py

@@ -43,16 +43,15 @@ class FileManager:
 
         response_json = response.json()
         # Convert UserFileSnapshot to FileDescriptor format
-        file_descriptors: List[FileDescriptor] = []
-        for user_file in response_json.get("user_files", []):
-            file_descriptors.append(
-                {
-                    "id": user_file["file_id"],
-                    "type": user_file["chat_file_type"],
-                    "name": user_file["name"],
-                    "user_file_id": str(user_file["id"]),
-                }
-            )
+        file_descriptors: List[FileDescriptor] = [
+            {
+                "id": user_file["file_id"],
+                "type": user_file["chat_file_type"],
+                "name": user_file["name"],
+                "user_file_id": str(user_file["id"]),
+            }
+            for user_file in response_json.get("user_files", [])
+        ]
         return file_descriptors, ""
 
     @staticmethod

backend/tests/integration/tests/index_attempt/test_index_attempt_pagination.py

@@ -100,8 +100,7 @@ def test_index_attempt_pagination(reset: None) -> None:  # noqa: ARG001
         base_time=base_time,
     )
 
-    for attempt in generated_attempts:
-        all_attempt_ids.append(attempt.id)
+    all_attempt_ids.extend(attempt.id for attempt in generated_attempts)
 
     # Verify basic pagination with different page sizes
     print("Verifying basic pagination with page size 5")

backend/tests/integration/tests/query_history/test_query_history_pagination.py

@@ -60,7 +60,7 @@ def test_query_history_pagination(reset: None) -> None:  # noqa: ARG001
     ) = setup_chat_sessions_with_different_feedback()
 
     all_chat_sessions = []
-    for _, chat_sessions in chat_sessions_by_feedback_type.items():
+    for chat_sessions in chat_sessions_by_feedback_type.values():
         all_chat_sessions.extend(chat_sessions)
 
     # Verify basic pagination with different page sizes

backend/tests/regression/answer_quality/cli_utils.py

@@ -248,11 +248,12 @@ def get_api_server_host_port(env_name: str) -> str:
 
     stdout, _ = _run_command("docker ps -a --format '{{json .}}'")
     containers = [json.loads(line) for line in stdout.splitlines()]
-    server_jsons = []
 
-    for container in containers:
-        if container_name in container["Names"] and env_name in container["Names"]:
-            server_jsons.append(container)
+    server_jsons = [
+        container
+        for container in containers
+        if container_name in container["Names"] and env_name in container["Names"]
+    ]
 
     if not server_jsons:
         raise RuntimeError(

backend/tests/regression/answer_quality/file_uploader.py

@@ -25,8 +25,7 @@ def unzip_and_get_file_paths(zip_file_path: str) -> list[str]:
 
     file_paths = []
     for root, _, files in os.walk(persistent_dir):
-        for file in sorted(files):
-            file_paths.append(os.path.join(root, file))
+        file_paths.extend(os.path.join(root, file) for file in sorted(files))
 
     return file_paths
 

backend/tests/unit/onyx/chat/test_argument_delta_streaming.py

@@ -68,8 +68,7 @@ def _stream_fragments(
         for packet in maybe_emit_argument_delta(tc_map, delta, pl, parsers=parsers):
             obj = packet.obj
             assert isinstance(obj, ToolCallArgumentDelta)
-            for value in obj.argument_deltas.values():
-                emitted.append(value)
+            emitted.extend(obj.argument_deltas.values())
     return emitted
 
 

backend/tests/unit/onyx/chat/test_citation_processor.py

@@ -358,8 +358,7 @@ def test_formatted_citation_yielded_separately(
 
     results = []
     for token in ["Text [", "1", "] here."]:
-        for result in processor.process_token(token):
-            results.append(result)
+        results.extend(processor.process_token(token))
 
     # Should have text chunks and formatted citation
     text_results = [r for r in results if isinstance(r, str)]
@@ -887,12 +886,10 @@ def test_stop_token_detection_stops_processing() -> None:
 
     results = []
     for token in ["Text ", "ST", "OP"]:
-        for result in processor.process_token(token):
-            results.append(result)
+        results.extend(processor.process_token(token))
 
     # Try to add more text after stop token
-    for result in processor.process_token(" more text"):
-        results.append(result)
+    results.extend(processor.process_token(" more text"))
 
     # Processing should stop at STOP token - no results after STOP
     output = "".join(r for r in results if isinstance(r, str))
@@ -907,8 +904,7 @@ def test_partial_stop_token_held_back() -> None:
 
     results = []
     for token in ["Text ", "ST"]:
-        for result in processor.process_token(token):
-            results.append(result)
+        results.extend(processor.process_token(token))
 
     # Partial stop token should be held back
     output = "".join(r for r in results if isinstance(r, str))
@@ -924,8 +920,7 @@ def test_stop_token_at_different_positions() -> None:
     processor1 = DynamicCitationProcessor(stop_stream=stop_stream)
     results1 = []
     for token in ["END"]:
-        for result in processor1.process_token(token):
-            results1.append(result)
+        results1.extend(processor1.process_token(token))
     # Stop token detection returns early, so no results
     output1 = "".join(r for r in results1 if isinstance(r, str))
     assert output1 == ""  # Stop token detected, no output
@@ -934,8 +929,7 @@ def test_stop_token_at_different_positions() -> None:
     processor2 = DynamicCitationProcessor(stop_stream=stop_stream)
     results2 = []
     for token in ["Start ", "EN", "D"]:
-        for result in processor2.process_token(token):
-            results2.append(result)
+        results2.extend(processor2.process_token(token))
     output2 = "".join(r for r in results2 if isinstance(r, str))
     # "Start " should be processed before stop token is detected
     assert "Start " in output2
@@ -967,12 +961,10 @@ def test_none_token_flushes_remaining_segment(
 
     results = []
     for token in ["Remaining ", "text"]:
-        for result in processor.process_token(token):
-            results.append(result)
+        results.extend(processor.process_token(token))
 
     # Flush with None
-    for result in processor.process_token(None):
-        results.append(result)
+    results.extend(processor.process_token(None))
 
     output = "".join(r for r in results if isinstance(r, str))
     assert "Remaining text" in output

backend/tests/unit/onyx/connectors/airtable/test_airtable_index_all.py

@@ -179,9 +179,7 @@ SAMPLE_BASES = [
 def _collect_docs(connector: AirtableConnector) -> list[Document]:
     docs: list[Document] = []
     for batch in connector.load_from_state():
-        for item in batch:
-            if isinstance(item, Document):
-                docs.append(item)
+        docs.extend(item for item in batch if isinstance(item, Document))
     return docs
 
 

backend/tests/unit/onyx/connectors/confluence/test_include_attachments_skip.py

@@ -81,9 +81,7 @@ def _collect_slim_doc_ids(
     ):
         ids: list[str] = []
         for batch in connector.retrieve_all_slim_docs():
-            for item in batch:
-                if isinstance(item, SlimDocument):
-                    ids.append(item.id)
+            ids.extend(item.id for item in batch if isinstance(item, SlimDocument))
     return ids
 
 

backend/tests/unit/onyx/connectors/confluence/test_slim_doc_image_skip.py

@@ -98,9 +98,7 @@ def _collect_slim_doc_ids(connector: ConfluenceConnector) -> list[str]:
     ):
         ids: list[str] = []
         for batch in connector.retrieve_all_slim_docs():
-            for item in batch:
-                if isinstance(item, SlimDocument):
-                    ids.append(item.id)
+            ids.extend(item.id for item in batch if isinstance(item, SlimDocument))
     return ids
 
 

backend/tests/unit/onyx/connectors/source_operation_harnesses.py

@@ -101,15 +101,15 @@ def compute_uncovered_units(
         if spec.untested is not None:
             continue
         for variant in spec.variants or (None,):
-            for capability in spec.capabilities:
-                if (operation, variant, capability) not in exercised:
-                    uncovered.append(
-                        UncoveredUnit(
-                            operation=operation,
-                            variant=variant,
-                            capability=capability,
-                        )
-                    )
+            uncovered.extend(
+                UncoveredUnit(
+                    operation=operation,
+                    variant=variant,
+                    capability=capability,
+                )
+                for capability in spec.capabilities
+                if (operation, variant, capability) not in exercised
+            )
     return uncovered
 
 
@@ -171,13 +171,13 @@ def find_import_fence_violations(
                     imported = [node.module]
                 else:
                     continue
-                for module in imported:
-                    if module.split(".")[0] in roots:
-                        violations.append(
-                            FenceViolation(
-                                file=str(path),
-                                line=node.lineno,
-                                module=module,
-                            )
-                        )
+                violations.extend(
+                    FenceViolation(
+                        file=str(path),
+                        line=node.lineno,
+                        module=module,
+                    )
+                    for module in imported
+                    if module.split(".")[0] in roots
+                )
     return violations

backend/tests/unit/onyx/document_index/opensearch/test_opensearch_batch_flush.py

@@ -183,8 +183,7 @@ def test_multiple_docs_each_under_limit_flush_per_doc() -> None:
     chunks = []
     for doc_idx in range(3):
         doc_id = f"doc_{doc_idx}"
-        for chunk_idx in range(50):
-            chunks.append(_make_chunk(doc_id, chunk_idx))
+        chunks.extend(_make_chunk(doc_id, chunk_idx) for chunk_idx in range(50))
 
     metadata = IndexingMetadata(
         doc_id_to_chunk_cnt_diff={

backend/tests/unit/onyx/indexing/test_embed_chunks_in_batches.py

@@ -262,8 +262,7 @@ class TestEmbedChunksInBatches:
         def _embed(
             chunks: list[DocAwareChunk], **_kwargs: object
         ) -> tuple[list[IndexChunk], list[ConnectorFailure]]:
-            for c in chunks:
-                embedded_doc_ids.append(c.source_document.id)
+            embedded_doc_ids.extend(c.source_document.id for c in chunks)
             return _mock_embed_fail_doc("docA")(chunks)
 
         mock_embed.side_effect = _embed

backend/tests/unit/onyx/server/features/craft/sandbox/test_send_message_with_bus.py

@@ -128,17 +128,18 @@ def _run_send_message(
     events: list[Any] = []
 
     def runner() -> None:
-        for evt in client.send_message(
-            _SESSION,
-            "hello",
-            directory=_DIRECTORY,
-            model_provider=model_provider,
-            model_id=model_id,
-            attachments=attachments,
-            timeout=timeout,
-            absolute_timeout=absolute_timeout,
-        ):
-            events.append(evt)
+        events.extend(
+            client.send_message(
+                _SESSION,
+                "hello",
+                directory=_DIRECTORY,
+                model_provider=model_provider,
+                model_id=model_id,
+                attachments=attachments,
+                timeout=timeout,
+                absolute_timeout=absolute_timeout,
+            )
+        )
 
     t = threading.Thread(target=runner, daemon=True)
     t.start()

backend/tests/unit/onyx/server/features/craft/session/test_md_to_docx.py

@@ -263,16 +263,15 @@ def test_table_cell_alignment_follows_column_spec() -> None:
     # The Markdown separator row (:--, :--:, --:) sets per-column alignment, the
     # way pandoc renders it.
     doc = _render("| L | C | R |\n|:--|:--:|--:|\n| a | b | c |\n")
-    row_alignments = []
-    for row in doc.tables[0].rows:
-        row_alignments.append(
-            [
-                None
-                if cell.paragraphs[0].alignment is None
-                else cell.paragraphs[0].alignment
-                for cell in row.cells
-            ]
-        )
+    row_alignments = [
+        [
+            None
+            if cell.paragraphs[0].alignment is None
+            else cell.paragraphs[0].alignment
+            for cell in row.cells
+        ]
+        for row in doc.tables[0].rows
+    ]
     # WD_ALIGN_PARAGRAPH: left default (None), CENTER == 1, RIGHT == 2.
     assert all(
         alignments[1] == WD_ALIGN_PARAGRAPH.CENTER for alignments in row_alignments

backend/tests/unit/onyx/server/features/craft/test_session_manager_merger.py

@@ -32,8 +32,7 @@ def _collect_with_timeout(
 
     def runner() -> None:
         try:
-            for item in gen:
-                items.append(item)
+            items.extend(gen)
         except BaseException as e:  # noqa: BLE001
             error.append(e)
 

backend/tests/unit/onyx/utils/test_audit.py

@@ -29,10 +29,11 @@ from onyx.utils.audit import (
 
 def _capture(caplog: pytest.LogCaptureFixture) -> list[dict[str, Any]]:
     """Parse every captured ``onyx.audit`` JSON message into a dict."""
-    events: list[dict[str, Any]] = []
-    for record in caplog.records:
-        if record.name.startswith("onyx.audit"):
-            events.append(json.loads(record.getMessage()))
+    events: list[dict[str, Any]] = [
+        json.loads(record.getMessage())
+        for record in caplog.records
+        if record.name.startswith("onyx.audit")
+    ]
     return events
 
 

backend/tests/unit/onyx/utils/test_json_river.py

@@ -123,58 +123,65 @@ class TestEscapeSequences:
 
     def test_newline_escape(self) -> None:
         deltas = _all_deltas(['{"text": "line1\\nline2"}'])
-        text_parts = []
-        for d in deltas:
-            if isinstance(d, dict) and "text" in d and isinstance(d["text"], str):
-                text_parts.append(d["text"])
+        text_parts = [
+            d["text"]
+            for d in deltas
+            if isinstance(d, dict) and "text" in d and isinstance(d["text"], str)
+        ]
         assert "".join(text_parts) == "line1\nline2"
 
     def test_tab_escape(self) -> None:
         deltas = _all_deltas(['{"t": "a\\tb"}'])
-        parts = []
-        for d in deltas:
-            if isinstance(d, dict) and "t" in d and isinstance(d["t"], str):
-                parts.append(d["t"])
+        parts = [
+            d["t"]
+            for d in deltas
+            if isinstance(d, dict) and "t" in d and isinstance(d["t"], str)
+        ]
         assert "".join(parts) == "a\tb"
 
     def test_escaped_quote(self) -> None:
         deltas = _all_deltas(['{"q": "say \\"hi\\""}'])
-        parts = []
-        for d in deltas:
-            if isinstance(d, dict) and "q" in d and isinstance(d["q"], str):
-                parts.append(d["q"])
+        parts = [
+            d["q"]
+            for d in deltas
+            if isinstance(d, dict) and "q" in d and isinstance(d["q"], str)
+        ]
         assert "".join(parts) == 'say "hi"'
 
     def test_unicode_escape(self) -> None:
         deltas = _all_deltas(['{"u": "\\u0041\\u0042"}'])
-        parts = []
-        for d in deltas:
-            if isinstance(d, dict) and "u" in d and isinstance(d["u"], str):
-                parts.append(d["u"])
+        parts = [
+            d["u"]
+            for d in deltas
+            if isinstance(d, dict) and "u" in d and isinstance(d["u"], str)
+        ]
         assert "".join(parts) == "AB"
 
     def test_escape_split_across_chunks(self) -> None:
         deltas = _all_deltas(['{"x": "a\\', 'nb"}'])
-        parts = []
-        for d in deltas:
-            if isinstance(d, dict) and "x" in d and isinstance(d["x"], str):
-                parts.append(d["x"])
+        parts = [
+            d["x"]
+            for d in deltas
+            if isinstance(d, dict) and "x" in d and isinstance(d["x"], str)
+        ]
         assert "".join(parts) == "a\nb"
 
     def test_unicode_escape_split_across_chunks(self) -> None:
         deltas = _all_deltas(['{"u": "\\u00', '41"}'])
-        parts = []
-        for d in deltas:
-            if isinstance(d, dict) and "u" in d and isinstance(d["u"], str):
-                parts.append(d["u"])
+        parts = [
+            d["u"]
+            for d in deltas
+            if isinstance(d, dict) and "u" in d and isinstance(d["u"], str)
+        ]
         assert "".join(parts) == "A"
 
     def test_backslash_escape(self) -> None:
         deltas = _all_deltas(['{"p": "c:\\\\dir"}'])
-        parts = []
-        for d in deltas:
-            if isinstance(d, dict) and "p" in d and isinstance(d["p"], str):
-                parts.append(d["p"])
+        parts = [
+            d["p"]
+            for d in deltas
+            if isinstance(d, dict) and "p" in d and isinstance(d["p"], str)
+        ]
         assert "".join(parts) == "c:\\dir"
 
 
@@ -196,9 +203,7 @@ class TestNestedStructures:
         all_items: list[str] = []
         for d in deltas:
             if isinstance(d, list):
-                for item in d:
-                    if isinstance(item, str):
-                        all_items.append(item)
+                all_items.extend(item for item in d if isinstance(item, str))
             elif isinstance(d, str):
                 all_items.append(d)
         joined = "".join(all_items)

backend/tests/unit/onyx/utils/test_threadpool_concurrency.py

@@ -304,11 +304,12 @@ def test_parallel_yield_basic() -> None:
     gen3 = make_gen([3, 6, 9], 0.15)  # Slowest generator
 
     # Collect results with timestamps
-    results: list[tuple[float, int]] = []
     start_time = time.time()
 
-    for value in parallel_yield([gen1, gen2, gen3]):
-        results.append((time.time() - start_time, value))
+    results: list[tuple[float, int]] = [
+        (time.time() - start_time, value)
+        for value in parallel_yield([gen1, gen2, gen3])
+    ]
 
     # Verify all values were yielded
     assert sorted(v for _, v in results) == list(range(1, 10))

backend/tests/unit/onyx/utils/test_threadpool_contextvars.py

@@ -46,7 +46,7 @@ def test_run_functions_in_parallel_preserves_contextvar() -> None:
     # Run in parallel and verify all results have the correct value
     results = run_functions_in_parallel(function_calls)
 
-    for _result_id, value in results.items():
+    for value in results.values():
         assert value == "parallel_test"
 
 

loadtest/tests/test_mock_llm.py

@@ -131,9 +131,11 @@ def stream_chunks(
     chunks = []
     with client.stream("POST", "/v1/chat/completions", json=body) as response:
         assert response.status_code == 200
-        for line in response.iter_lines():
-            if line.startswith("data: ") and line != "data: [DONE]":
-                chunks.append(json.loads(line[len("data: ") :]))
+        chunks.extend(
+            json.loads(line[len("data: ") :])
+            for line in response.iter_lines()
+            if line.startswith("data: ") and line != "data: [DONE]"
+        )
     return chunks
 
 

pyproject.toml

@@ -292,7 +292,7 @@ ignore = [
     "B904", # raise-without-from-inside-except (~465 existing violations).
 ]
 # G004: f-strings in logging break Sentry's message-pattern grouping.
-select = ["ARG", "B", "E", "F", "G004", "I", "S", "W"]
+select = ["ARG", "B", "E", "F", "G004", "I", "PERF", "S", "W"]
 
 [tool.ruff.lint.flake8-bugbear]
 # FastAPI dependency-injection markers are the intended way to write these