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>

llama_stack/providers/utils/inference/openai_compat.py

@@ -605,7 +605,7 @@ def convert_tool_call(
             tool_name=tool_call.function.name,
             arguments=json.loads(tool_call.function.arguments),
         )
-    except Exception as e:
+    except Exception:
         return UnparseableToolCall(
             call_id=tool_call.id or "",
             tool_name=tool_call.function.name or "",
@@ -876,7 +876,9 @@ async def convert_openai_chat_completion_stream(
             # it is possible to have parallel tool calls in stream, but
             # ChatCompletionResponseEvent only supports one per stream
             if len(choice.delta.tool_calls) > 1:
-                warnings.warn("multiple tool calls found in a single delta, using the first, ignoring the rest")
+                warnings.warn(
+                    "multiple tool calls found in a single delta, using the first, ignoring the rest", stacklevel=2
+                )
 
             if not enable_incremental_tool_calls:
                 yield ChatCompletionResponseStreamChunk(

llama_stack/providers/utils/kvstore/redis/redis.py

@@ -36,7 +36,7 @@ class RedisKVStoreImpl(KVStore):
         value = await self.redis.get(key)
         if value is None:
             return None
-        ttl = await self.redis.ttl(key)
+        await self.redis.ttl(key)
         return value
 
     async def delete(self, key: str) -> None:

llama_stack/providers/utils/scoring/aggregation_utils.py

@@ -32,7 +32,7 @@ def aggregate_categorical_count(
     scoring_results: List[ScoringResultRow],
 ) -> Dict[str, Any]:
     scores = [str(r["score"]) for r in scoring_results]
-    unique_scores = sorted(list(set(scores)))
+    unique_scores = sorted(set(scores))
     return {"categorical_count": {s: scores.count(s) for s in unique_scores}}
 
 

llama_stack/providers/utils/scoring/base_scoring_fn.py

@@ -66,7 +66,7 @@ class RegisteredBaseScoringFn(BaseScoringFn):
         return self.__class__.__name__
 
     def get_supported_scoring_fn_defs(self) -> List[ScoringFn]:
-        return [x for x in self.supported_fn_defs_registry.values()]
+        return list(self.supported_fn_defs_registry.values())
 
     def register_scoring_fn_def(self, scoring_fn: ScoringFn) -> None:
         if scoring_fn.identifier in self.supported_fn_defs_registry: