chore(lint): update Ruff ignores for project conventions and maintainability (#1184)

- Added new ignores from flake8-bugbear (`B007`, `B008`)
- Ignored `C901` (high function complexity) for now, pending review
- Maintained PyTorch conventions (`N812`, `N817`)
- Allowed `E731` (lambda assignments) for flexibility
- Consolidated existing ignores (`E402`, `E501`, `F405`, `C408`, `N812`)
- Documented rationale for each ignored rule

This keeps our linting aligned with project needs while tracking
potential fixes.

Signed-off-by: Sébastien Han <seb@redhat.com>

Signed-off-by: Sébastien Han <seb@redhat.com>

Author: leseb

llama_stack/cli/stack/_build.py

@@ -141,7 +141,7 @@ def run_stack_build_command(args: argparse.Namespace) -> None:
                 completer=WordCompleter(available_providers),
                 complete_while_typing=True,
                 validator=Validator.from_callable(
-                    lambda x: x in available_providers,
+                    lambda x: x in available_providers,  # noqa: B023 - see https://github.com/astral-sh/ruff/issues/7847
                     error_message="Invalid provider, use <TAB> to see options",
                 ),
             )

llama_stack/cli/tests/test_stack_config.py

@@ -112,7 +112,7 @@ def test_parse_and_maybe_upgrade_config_old_format(old_config):
 
     inference_providers = result.providers["inference"]
     assert len(inference_providers) == 2
-    assert set(x.provider_id for x in inference_providers) == {
+    assert {x.provider_id for x in inference_providers} == {
         "remote::ollama-00",
         "meta-reference-01",
     }

llama_stack/distribution/distribution.py

@@ -13,7 +13,7 @@
 
 
 def stack_apis() -> List[Api]:
-    return [v for v in Api]
+    return list(Api)
 
 
 class AutoRoutedApiInfo(BaseModel):
@@ -55,7 +55,7 @@ def builtin_automatically_routed_apis() -> List[AutoRoutedApiInfo]:
 
 
 def providable_apis() -> List[Api]:
-    routing_table_apis = set(x.routing_table_api for x in builtin_automatically_routed_apis())
+    routing_table_apis = {x.routing_table_api for x in builtin_automatically_routed_apis()}
     return [api for api in Api if api not in routing_table_apis and api != Api.inspect]
 
 

llama_stack/distribution/resolver.py

@@ -115,8 +115,8 @@ async def resolve_impls(
     - flatmaps, sorts and resolves the providers in dependency order
     - for each API, produces either a (local, passthrough or router) implementation
     """
-    routing_table_apis = set(x.routing_table_api for x in builtin_automatically_routed_apis())
-    router_apis = set(x.router_api for x in builtin_automatically_routed_apis())
+    routing_table_apis = {x.routing_table_api for x in builtin_automatically_routed_apis()}
+    router_apis = {x.router_api for x in builtin_automatically_routed_apis()}
 
     providers_with_specs = {}
 

llama_stack/distribution/ui/page/playground/rag.py

@@ -134,7 +134,7 @@ def rag_chat_page():
             dict(
                 name="builtin::rag/knowledge_search",
                 args={
-                    "vector_db_ids": [vector_db_id for vector_db_id in selected_vector_dbs],
+                    "vector_db_ids": list(selected_vector_dbs),
                 },
             )
         ],

llama_stack/providers/inline/agents/meta_reference/agent_instance.py

@@ -797,10 +797,10 @@ async def _get_tool_defs(
         self, toolgroups_for_turn: Optional[List[AgentToolGroup]] = None
     ) -> Tuple[List[ToolDefinition], Dict[str, str]]:
         # Determine which tools to include
-        agent_config_toolgroups = set(
-            (toolgroup.name if isinstance(toolgroup, AgentToolGroupWithArgs) else toolgroup)
+        agent_config_toolgroups = {
+            toolgroup.name if isinstance(toolgroup, AgentToolGroupWithArgs) else toolgroup
             for toolgroup in self.agent_config.toolgroups
-        )
+        }
         toolgroups_for_turn_set = (
             agent_config_toolgroups
             if toolgroups_for_turn is None

llama_stack/providers/inline/eval/meta_reference/eval.py

@@ -86,7 +86,6 @@ async def run_eval(
     ) -> Job:
         task_def = self.benchmarks[benchmark_id]
         dataset_id = task_def.dataset_id
-        candidate = task_config.eval_candidate
         scoring_functions = task_def.scoring_functions
         dataset_def = await self.datasets_api.get_dataset(dataset_id=dataset_id)
         validate_dataset_schema(dataset_def.dataset_schema, get_valid_schemas(Api.eval.value))

llama_stack/providers/inline/inference/meta_reference/inference.py

@@ -208,7 +208,6 @@ def impl():
             logprobs = []
             stop_reason = None
 
-            tokenizer = self.generator.formatter.tokenizer
             for token_result in self.generator.completion(request):
                 tokens.append(token_result.token)
                 if token_result.text == "<|eot_id|>":

llama_stack/providers/inline/inference/meta_reference/parallel_utils.py

@@ -207,7 +207,7 @@ def maybe_parse_message(maybe_json: Optional[str]) -> Optional[ProcessingMessage
         return parse_message(maybe_json)
     except json.JSONDecodeError:
         return None
-    except ValueError as e:
+    except ValueError:
         return None
 
 
@@ -352,7 +352,7 @@ def run_inference(
                 if isinstance(obj, TaskResponse):
                     yield obj.result
 
-        except GeneratorExit as e:
+        except GeneratorExit:
             self.request_socket.send(encode_msg(CancelSentinel()))
             while True:
                 obj_json = self.request_socket.send()

llama_stack/providers/inline/inference/meta_reference/quantization/fp8_txest_disabled.py

@@ -7,6 +7,9 @@
 # Copyright (c) Meta Platforms, Inc. and affiliates.
 # This software may be used and distributed in accordance with the terms of the Llama 3 Community License Agreement.
 
+# The file gets a special treatment for now?
+# ruff: noqa: N803
+
 import unittest
 
 import torch