[Lint] Add flake8-bugbear (ray-project#19053)

* Add flake8-bugbear

* Add flake8-bugbear

.flake8

@@ -24,4 +24,20 @@ ignore =
   W605
   I
   N
+  B001
+  B002
+  B003
+  B004
+  B005
+  B007
+  B008
+  B009
+  B010
+  B011
+  B012
+  B013
+  B014
+  B015
+  B016
+  B017
 avoid-escape = no

.gitpod/Dockerfile

@@ -15,7 +15,7 @@ RUN set -x; apt update \
     && mv bazel.gpg /etc/apt/trusted.gpg.d/ \
     && echo "deb [arch=amd64] https://storage.googleapis.com/bazel-apt stable jdk1.8" | tee /etc/apt/sources.list.d/bazel.list \
     && apt update && apt install bazel-3.7.2 -y \
-    && pip3 install cython==0.29.0 pytest pandas tree tabulate pexpect sklearn joblib yapf==0.23.0 flake8==3.9.1 mypy==0.782 flake8-quotes setproctitle==1.1.10 psutil \
+    && pip3 install cython==0.29.0 pytest pandas tree tabulate pexpect sklearn joblib yapf==0.23.0 flake8==3.9.1 mypy==0.782 flake8-quotes flake8-bugbear==21.9.2 setproctitle==1.1.10 psutil \
     && python3 -c  'print("startup --output_base=/workspace/ray/.bazel-cache\nstartup --host_jvm_args=-Xmx1800m\nbuild --jobs=6")' > /etc/bazel.bazelrc
 
 RUN update-alternatives --install /usr/local/bin/python python /usr/bin/python3 30 \

ci/travis/format.sh

@@ -83,6 +83,10 @@ if [[ $(flake8 --version) != *"flake8_quotes"* ]]; then
     echo "WARNING: Ray uses flake8 with flake8_quotes. Might error without it. Install with: pip install flake8-quotes"
 fi
 
+if [[ $(flake8 --version) != *"flake8-bugbear"* ]]; then
+    echo "WARNING: Ray uses flake8 with flake8-bugbear. Might error without it. Install with: pip install flake8-bugbear"
+fi
+
 SHELLCHECK_FLAGS=(
   --exclude=1090  # "Can't follow non-constant source. Use a directive to specify location."
   --exclude=1091  # "Not following {file} due to some error"

python/ray/data/datasource/file_based_datasource.py

@@ -275,9 +275,10 @@ def _expand_paths(paths: Union[str, List[str]],
     return expanded_paths, file_infos
 
 
-def _expand_directory(path: str,
-                      filesystem: "pyarrow.fs.FileSystem",
-                      exclude_prefixes: List[str] = [".", "_"]) -> List[str]:
+def _expand_directory(
+        path: str,
+        filesystem: "pyarrow.fs.FileSystem",
+        exclude_prefixes: Optional[List[str]] = None) -> List[str]:
     """
     Expand the provided directory path to a list of file paths.
 
@@ -292,6 +293,9 @@ def _expand_directory(path: str,
     Returns:
         A list of file paths contained in the provided directory.
     """
+    if exclude_prefixes is None:
+        exclude_prefixes = [".", "_"]
+
     from pyarrow.fs import FileSelector
     selector = FileSelector(path, recursive=True)
     files = filesystem.get_file_info(selector)
@@ -304,7 +308,7 @@ def _expand_directory(path: str,
         if not file_path.startswith(base_path):
             continue
         relative = file_path[len(base_path):]
-        if any(relative.startswith(prefix) for prefix in [".", "_"]):
+        if any(relative.startswith(prefix) for prefix in exclude_prefixes):
             continue
         filtered_paths.append((file_path, file_))
     # We sort the paths to guarantee a stable order.

python/ray/experimental/array/remote/core.py

@@ -68,8 +68,8 @@ def diag(v, k=0):
 
 
 @ray.remote
-def transpose(a, axes=[]):
-    axes = None if axes == [] else axes
+def transpose(a, axes=None):
+    axes = None if (axes == [] or axes is None) else axes
     return np.transpose(a, axes=axes)
 
 

python/ray/serve/http_proxy.py

@@ -259,8 +259,11 @@ def __init__(self,
                  port: int,
                  controller_name: str,
                  controller_namespace: str,
-                 http_middlewares: List[
-                     "starlette.middleware.Middleware"] = []):  # noqa: F821
+                 http_middlewares: Optional[List[
+                     "starlette.middleware.Middleware"]] = None):  # noqa: F821
+        if http_middlewares is None:
+            http_middlewares = []
+
         self.host = host
         self.port = port
 

python/ray/tests/test_autoscaler.py

@@ -464,7 +464,10 @@ def waitFor(self, condition, num_retries=50, fail_msg=None):
         fail_msg = fail_msg or "Timed out waiting for {}".format(condition)
         raise RayTestTimeoutException(fail_msg)
 
-    def waitForNodes(self, expected, comparison=None, tag_filters={}):
+    def waitForNodes(self, expected, comparison=None, tag_filters=None):
+        if tag_filters is None:
+            tag_filters = {}
+
         MAX_ITER = 50
         for i in range(MAX_ITER):
             n = len(self.provider.non_terminated_nodes(tag_filters))

python/ray/tests/test_resource_demand_scheduler.py

@@ -1322,7 +1322,10 @@ def tearDown(self):
         shutil.rmtree(self.tmpdir)
         ray.shutdown()
 
-    def waitForNodes(self, expected, comparison=None, tag_filters={}):
+    def waitForNodes(self, expected, comparison=None, tag_filters=None):
+        if tag_filters is None:
+            tag_filters = {}
+
         MAX_ITER = 50
         for i in range(MAX_ITER):
             n = len(self.provider.non_terminated_nodes(tag_filters))

python/ray/tune/tests/test_trial_scheduler_pbt.py

@@ -201,7 +201,10 @@ def MockTrainingFuncSync(config, checkpoint_dir=None):
     def tearDown(self):
         ray.shutdown()
 
-    def synchSetup(self, synch, param=[10, 20, 30]):
+    def synchSetup(self, synch, param=None):
+        if param is None:
+            param = [10, 20, 30]
+
         scheduler = PopulationBasedTraining(
             time_attr="training_iteration",
             metric="mean_accuracy",

python/ray/util/dask/scheduler_utils.py

@@ -371,8 +371,11 @@ def fire_task():
     return nested_get(result, state["cache"])
 
 
-def apply_sync(func, args=(), kwds={}, callback=None):
+def apply_sync(func, args=(), kwds=None, callback=None):
     """ A naive synchronous version of apply_async """
+    if kwds is None:
+        kwds = {}
+
     res = func(*args, **kwds)
     if callback is not None:
         callback(res)

python/ray/workflow/step_executor.py

@@ -401,7 +401,10 @@ def __reduce__(self):
 
 def _record_step_status(step_id: "StepID",
                         status: "WorkflowStatus",
-                        outputs: List["ObjectRef"] = []) -> None:
+                        outputs: Optional[List["ObjectRef"]] = None) -> None:
+    if outputs is None:
+        outputs = []
+
     workflow_id = workflow_context.get_current_workflow_id()
     workflow_manager = get_management_actor()
     ray.get(

python/requirements_linters.txt

@@ -1,5 +1,6 @@
 flake8==3.9.1
 flake8-comprehensions
 flake8-quotes==2.0.0
+flake8-bugbear==21.9.2
 mypy==0.782
 yapf==0.23.0

rllib/agents/ddpg/ddpg_tf_model.py

@@ -1,6 +1,6 @@
 import numpy as np
 import gym
-from typing import List
+from typing import List, Optional
 
 from ray.rllib.models.tf.tf_modelv2 import TFModelV2
 from ray.rllib.utils.framework import try_import_tf
@@ -29,9 +29,9 @@ def __init__(
             model_config: ModelConfigDict,
             name: str,
             # Extra DDPGActionModel args:
-            actor_hiddens: List[int] = [256, 256],
+            actor_hiddens: Optional[List[int]] = None,
             actor_hidden_activation: str = "relu",
-            critic_hiddens: List[int] = [256, 256],
+            critic_hiddens: Optional[List[int]] = None,
             critic_hidden_activation: str = "relu",
             twin_q: bool = False,
             add_layer_norm: bool = False):
@@ -48,6 +48,12 @@ def __init__(
         should be defined in subclasses of DDPGActionModel.
         """
 
+        if actor_hiddens is None:
+            actor_hiddens = [256, 256]
+
+        if critic_hiddens is None:
+            critic_hiddens = [256, 256]
+
         super(DDPGTFModel, self).__init__(obs_space, action_space, num_outputs,
                                           model_config, name)
 

rllib/agents/ddpg/ddpg_torch_model.py

@@ -1,6 +1,6 @@
 import numpy as np
 import gym
-from typing import List, Dict, Union
+from typing import List, Dict, Union, Optional
 
 from ray.rllib.models.torch.misc import SlimFC
 from ray.rllib.models.torch.torch_modelv2 import TorchModelV2
@@ -31,9 +31,9 @@ def __init__(
             model_config: ModelConfigDict,
             name: str,
             # Extra DDPGActionModel args:
-            actor_hiddens: List[int] = [256, 256],
+            actor_hiddens: Optional[List[int]] = None,
             actor_hidden_activation: str = "relu",
-            critic_hiddens: List[int] = [256, 256],
+            critic_hiddens: Optional[List[int]] = None,
             critic_hidden_activation: str = "relu",
             twin_q: bool = False,
             add_layer_norm: bool = False):
@@ -51,6 +51,12 @@ def __init__(
         only defines the layers for the output heads. Those layers for
         forward() should be defined in subclasses of DDPGTorchModel.
         """
+        if actor_hiddens is None:
+            actor_hiddens = [256, 256]
+
+        if critic_hiddens is None:
+            critic_hiddens = [256, 256]
+
         nn.Module.__init__(self)
         super(DDPGTorchModel, self).__init__(obs_space, action_space,
                                              num_outputs, model_config, name)

rllib/contrib/sumo/connector.py

@@ -162,7 +162,7 @@ def _stopping_condition(self, current_step_counter, until_end):
             return True
         return False
 
-    def step(self, until_end=False, agents=set()):
+    def step(self, until_end=False, agents=None):
         """
             Runs a "learning" step and returns if the simulation has finished.
             This function in meant to be called by the RLLIB Environment.
@@ -176,6 +176,9 @@ def step(self, until_end=False, agents=set()):
             Return:
                 Bool. True iff the simulation is still ongoing.
         """
+        if agents is None:
+            agents = set()
+
         # Execute SUMO steps until the learning needs to happen
         current_step_counter = 0
         logger.debug(

rllib/evaluation/metrics.py

@@ -56,10 +56,15 @@ def get_learner_stats(grad_info: GradInfoDict) -> LearnerStatsDict:
 
 @DeveloperAPI
 def collect_metrics(local_worker: Optional["RolloutWorker"] = None,
-                    remote_workers: List[ActorHandle] = [],
-                    to_be_collected: List[ObjectRef] = [],
+                    remote_workers: Optional[List[ActorHandle]] = None,
+                    to_be_collected: Optional[List[ObjectRef]] = None,
                     timeout_seconds: int = 180) -> ResultDict:
     """Gathers episode metrics from RolloutWorker instances."""
+    if remote_workers is None:
+        remote_workers = []
+
+    if to_be_collected is None:
+        to_be_collected = []
 
     episodes, to_be_collected = collect_episodes(
         local_worker,
@@ -73,11 +78,16 @@ def collect_metrics(local_worker: Optional["RolloutWorker"] = None,
 @DeveloperAPI
 def collect_episodes(
         local_worker: Optional["RolloutWorker"] = None,
-        remote_workers: List[ActorHandle] = [],
-        to_be_collected: List[ObjectRef] = [],
+        remote_workers: Optional[List[ActorHandle]] = None,
+        to_be_collected: Optional[List[ObjectRef]] = None,
         timeout_seconds: int = 180
 ) -> Tuple[List[Union[RolloutMetrics, OffPolicyEstimate]], List[ObjectRef]]:
     """Gathers new episodes metrics tuples from the given evaluators."""
+    if remote_workers is None:
+        remote_workers = []
+
+    if to_be_collected is None:
+        to_be_collected = []
 
     if remote_workers:
         pending = [

rllib/examples/env/coin_game_non_vectorized_env.py

@@ -13,7 +13,7 @@
 from gym.utils import seeding
 from ray.rllib.env.multi_agent_env import MultiAgentEnv
 from ray.rllib.utils import override
-from typing import Dict
+from typing import Dict, Optional
 
 from ray.rllib.examples.env.utils.interfaces import InfoAccumulationInterface
 
@@ -36,7 +36,9 @@ class CoinGame(InfoAccumulationInterface, MultiAgentEnv, gym.Env):
         np.array([-1, 0]),
     ]
 
-    def __init__(self, config: Dict = {}):
+    def __init__(self, config: Optional[Dict] = None):
+        if config is None:
+            config = {}
 
         self._validate_config(config)
 
@@ -325,7 +327,10 @@ def _init_info(self):
 class AsymCoinGame(CoinGame):
     NAME = "AsymCoinGame"
 
-    def __init__(self, config: dict = {}):
+    def __init__(self, config: Optional[dict] = None):
+        if config is None:
+            config = {}
+
         if "asymmetric" in config:
             assert config["asymmetric"]
         else:

rllib/examples/env/coin_game_vectorized_env.py

@@ -21,7 +21,9 @@ class VectorizedCoinGame(CoinGame):
     Vectorized Coin Game environment.
     """
 
-    def __init__(self, config={}):
+    def __init__(self, config=None):
+        if config is None:
+            config = {}
 
         super().__init__(config)
 
@@ -159,7 +161,10 @@ def _load_env(self, env_state):
 class AsymVectorizedCoinGame(VectorizedCoinGame):
     NAME = "AsymCoinGame"
 
-    def __init__(self, config={}):
+    def __init__(self, config=None):
+        if config is None:
+            config = {}
+
         if "asymmetric" in config:
             assert config["asymmetric"]
         else:

rllib/examples/env/matrix_sequential_social_dilemma.py

@@ -8,7 +8,7 @@
 import logging
 from abc import ABC
 from collections import Iterable
-from typing import Dict
+from typing import Dict, Optional
 
 import numpy as np
 from gym.spaces import Discrete
@@ -39,7 +39,9 @@ class MatrixSequentialSocialDilemma(InfoAccumulationInterface, MultiAgentEnv,
     episode.
     """
 
-    def __init__(self, config: Dict = {}):
+    def __init__(self, config: Optional[Dict] = None):
+        if config is None:
+            config = {}
 
         assert "reward_randomness" not in config.keys()
         assert self.PAYOUT_MATRIX is not None

rllib/utils/tf_run_builder.py

@@ -59,7 +59,10 @@ def get(self, to_fetch):
 _count = 0
 
 
-def run_timeline(sess, ops, debug_name, feed_dict={}, timeline_dir=None):
+def run_timeline(sess, ops, debug_name, feed_dict=None, timeline_dir=None):
+    if feed_dict is None:
+        feed_dict = {}
+
     if timeline_dir:
         from tensorflow.python.client import timeline