unify docformatter config (#1642)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

additional_dependencies: [tomli]

.pre-commit-config.yaml

@@ -51,25 +51,15 @@ repos:
     rev: v1.7.3
     hooks:
       - id: docformatter
-        args:
-          - "--in-place"
-          - "--wrap-summaries=120"
-          - "--wrap-descriptions=120"
+        additional_dependencies: [tomli]
+        args: ["--in-place"]
 
   - repo: https://github.com/psf/black
     rev: 23.3.0
     hooks:
       - id: black
         name: Format code
 
-  - repo: https://github.com/asottile/blacken-docs
-    rev: 1.14.0
-    hooks:
-      - id: blacken-docs
-        args:
-          - "--line-length=120"
-          - "--skip-errors"
-
   - repo: https://github.com/astral-sh/ruff-pre-commit
     rev: v0.0.276
     hooks:

pyproject.toml

@@ -48,15 +48,11 @@ exclude_lines = [
 line-length = 120
 exclude = "(.eggs|.git|.hg|.mypy_cache|.venv|_build|buck-out|build|dist)"
 
-[tool.isort]
-known_first_party = [
-    "flash",
-    "examples",
-    "tests",
-]
-skip_glob = []
-profile = "black"
-line_length = 120
+[tool.docformatter]
+recursive = true
+wrap-summaries = 120
+wrap-descriptions = 120
+blank = true
 
 
 [tool.ruff]

setup.py

@@ -33,6 +33,7 @@ def _load_readme_description(path_dir: str, homepage: str, ver: str) -> str:
 
     >>> _load_readme_description(_PATH_ROOT, "", "")  # doctest: +ELLIPSIS +NORMALIZE_WHITESPACE
     '<div align="center">...'
+
     """
     path_readme = os.path.join(path_dir, "README.md")
     text = open(path_readme, encoding="utf-8").read()
@@ -65,6 +66,7 @@ def _augment_requirement(ln: str, comment_char: str = "#", unfreeze: bool = True
     'arrow>=1.2.0, <=1.2.2  # strict'
     >>> _augment_requirement("arrow", unfreeze=True)
     'arrow'
+
     """
     # filer all comments
     if comment_char in ln:
@@ -95,6 +97,7 @@ def _load_requirements(path_dir: str, file_name: str = "base.txt", unfreeze: boo
     >>> path_req = os.path.join(_PATH_ROOT, "requirements")
     >>> _load_requirements(path_req, "docs.txt")  # doctest: +ELLIPSIS +NORMALIZE_WHITESPACE
     ['sphinx>=4.0', ...]
+
     """
     with open(os.path.join(path_dir, file_name)) as file:
         lines = [ln.strip() for ln in file.readlines()]

src/flash/audio/speech_recognition/collate.py

@@ -50,6 +50,7 @@ class DataCollatorCTCWithPadding:
             If set will pad the sequence to a multiple of the provided value.
             This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability >=
             7.5 (Volta).
+
     """
 
     processor: AutoProcessor

src/flash/core/classification.py

@@ -127,6 +127,7 @@ class ClassificationOutput(Output):
 
     Args:
         multi_label: If true, treats outputs as multi label logits.
+
     """
 
     def __init__(self, multi_label: bool = False):

src/flash/core/data/batch.py

@@ -58,6 +58,7 @@ def default_uncollate(batch: Any) -> List[Any]:
         ValueError: If the input is a ``dict`` whose values are not all list-like.
         ValueError: If the input is a ``dict`` whose values are not all the same length.
         ValueError: If the input is not a ``dict`` or list-like.
+
     """
     if isinstance(batch, dict):
         if any(not _is_list_like_excluding_str(sub_batch) for sub_batch in batch.values()):

src/flash/core/data/data_module.py

@@ -566,6 +566,7 @@ def _split_train_val(
 
         Returns:
             A tuple containing the training and validation datasets
+
         """
 
         if not isinstance(val_split, float) or (isinstance(val_split, float) and val_split > 1 or val_split < 0):

src/flash/core/data/io/classification_input.py

@@ -64,5 +64,6 @@ def format_target(self, target: Any) -> Any:
 
         Returns:
             The formatted target.
+
         """
         return getattr(self, "target_formatter", lambda x: x)(target)

src/flash/core/data/io/input.py

@@ -89,6 +89,7 @@ def _has_len(data: Union[Sequence, Iterable]) -> bool:
 
     Args:
         data: The object to check for length support.
+
     """
     try:
         len(data)

src/flash/core/data/io/input_transform.py

@@ -84,6 +84,7 @@ def per_sample_transform(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_sample_transform(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         pass
 
@@ -97,6 +98,7 @@ def train_per_sample_transform(self) -> Callable:
                 DataKeys.TARGET: ...,
                 DataKeys.METADATA: ...,
             }
+
         """
         return self.per_sample_transform()
 
@@ -121,6 +123,7 @@ def val_per_sample_transform(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_sample_transform(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_sample_transform()
 
@@ -134,6 +137,7 @@ def test_per_sample_transform(self) -> Callable:
                 DataKeys.TARGET: ...,
                 DataKeys.METADATA: ...,
             }
+
         """
         return self.per_sample_transform()
 
@@ -158,6 +162,7 @@ def predict_per_sample_transform(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_sample_transform(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_sample_transform()
 
@@ -182,6 +187,7 @@ def serve_per_sample_transform(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_sample_transform(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_sample_transform()
 
@@ -210,6 +216,7 @@ def per_sample_transform_on_device(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_sample_transform_on_device(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         pass
 
@@ -223,6 +230,7 @@ def train_per_sample_transform_on_device(self) -> Callable:
                 DataKeys.TARGET: ...,
                 DataKeys.METADATA: ...,
             }
+
         """
         return self.per_sample_transform_on_device()
 
@@ -247,6 +255,7 @@ def val_per_sample_transform_on_device(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_sample_transform_on_device(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_sample_transform_on_device()
 
@@ -260,6 +269,7 @@ def test_per_sample_transform_on_device(self) -> Callable:
                 DataKeys.TARGET: ...,
                 DataKeys.METADATA: ...,
             }
+
         """
         return self.per_sample_transform_on_device()
 
@@ -284,6 +294,7 @@ def predict_per_sample_transform_on_device(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_sample_transform_on_device(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_sample_transform_on_device()
 
@@ -308,6 +319,7 @@ def serve_per_sample_transform_on_device(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def serve_per_sample_transform_on_device(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_sample_transform_on_device()
 
@@ -336,6 +348,7 @@ def per_batch_transform(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_batch_transform(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         pass
 
@@ -349,6 +362,7 @@ def train_per_batch_transform(self) -> Callable:
                 DataKeys.TARGET: ...,
                 DataKeys.METADATA: ...,
             }
+
         """
         return self.per_batch_transform()
 
@@ -373,6 +387,7 @@ def val_per_batch_transform(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_batch_transform(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_batch_transform()
 
@@ -386,6 +401,7 @@ def test_per_batch_transform(self) -> Callable:
                 DataKeys.TARGET: ...,
                 DataKeys.METADATA: ...,
             }
+
         """
         return self.per_batch_transform()
 
@@ -410,6 +426,7 @@ def predict_per_batch_transform(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_batch_transform(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_batch_transform()
 
@@ -434,6 +451,7 @@ def serve_per_batch_transform(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_batch_transform(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_batch_transform()
 
@@ -462,6 +480,7 @@ def per_batch_transform_on_device(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_batch_transform_on_device(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         pass
 
@@ -475,6 +494,7 @@ def train_per_batch_transform_on_device(self) -> Callable:
                 DataKeys.TARGET: ...,
                 DataKeys.METADATA: ...,
             }
+
         """
         return self.per_batch_transform_on_device()
 
@@ -499,6 +519,7 @@ def val_per_batch_transform_on_device(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_batch_transform_on_device(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_batch_transform_on_device()
 
@@ -512,6 +533,7 @@ def test_per_batch_transform_on_device(self) -> Callable:
                 DataKeys.TARGET: ...,
                 DataKeys.METADATA: ...,
             }
+
         """
         return self.per_batch_transform_on_device()
 
@@ -536,6 +558,7 @@ def predict_per_batch_transform_on_device(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def per_batch_transform_on_device(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_batch_transform_on_device()
 
@@ -560,6 +583,7 @@ def serve_per_batch_transform_on_device(self) -> Callable:
             class MyInputTransform(InputTransform):
                 def serve_per_batch_transform_on_device(self) -> Callable:
                     return ApplyToKeys("input", my_func)
+
         """
         return self.per_batch_transform_on_device()
 
@@ -606,6 +630,7 @@ def _per_batch_transform(self, batch: Any, stage: RunningStage) -> Any:
 
         .. note:: This option is mutually exclusive with :meth:`per_sample_transform_on_device`, since if both are
         specified, uncollation has to be applied.
+
         """
         return self.current_transform(stage=stage, current_fn="per_batch_transform")(batch)
 
@@ -620,6 +645,7 @@ def _per_sample_transform_on_device(self, sample: Any, stage: RunningStage) -> A
         specified, uncollation has to be applied. .. note::     This function won't be called within the dataloader
         workers, since to make that happen     each of the workers would have to create it's own CUDA-context which
         would pollute GPU memory (if on GPU).
+
         """
         fn = self.current_transform(stage=stage, current_fn="per_sample_transform_on_device")
         if isinstance(sample, list):
@@ -631,6 +657,7 @@ def _per_batch_transform_on_device(self, batch: Any, stage: RunningStage) -> Any
 
         .. note::     This function won't be called within the dataloader workers, since to make that happen     each of
         the workers would have to create it's own CUDA-context which would pollute GPU memory (if on GPU).
+
         """
         return self.current_transform(stage=stage, current_fn="per_batch_transform_on_device")(batch)
 

src/flash/core/data/io/output.py

@@ -37,6 +37,7 @@ def transform(sample: Any) -> Any:
 
         Returns:
             The converted output.
+
         """
         return sample
 

src/flash/core/data/io/output_transform.py

@@ -25,6 +25,7 @@ def per_batch_transform(batch: Any) -> Any:
         """Transforms to apply on a whole batch before uncollation to individual samples.
 
         Can involve both CPU and Device transforms as this is not applied in separate workers.
+
         """
         return batch
 
@@ -33,6 +34,7 @@ def per_sample_transform(sample: Any) -> Any:
         """Transforms to apply to a single sample after splitting up the batch.
 
         Can involve both CPU and Device transforms as this is not applied in separate workers.
+
         """
         return sample
 
@@ -41,6 +43,7 @@ def uncollate(batch: Any) -> Any:
         """Uncollates a batch into single samples.
 
         Tries to preserve the type wherever possible.
+
         """
         return default_uncollate(batch)
 

src/flash/core/data/splits.py

@@ -21,6 +21,7 @@ class SplitDataset(Properties, Dataset):
         split_ds = SplitDataset(dataset, indices=[10, 14, 25])
 
         split_ds = SplitDataset(dataset, indices=[10, 10, 10, 14, 25], use_duplicated_indices=True)
+
     """
 
     def __init__(

src/flash/core/data/utilities/data_frame.py

@@ -30,6 +30,7 @@ def resolve_targets(data_frame: pd.DataFrame, target_keys: Union[str, List[str]]
     Args:
         data_frame: The ``pd.DataFrame`` containing the target column / columns.
         target_keys: The column in the data frame (or a list of columns) from which to resolve the target.
+
     """
     if not isinstance(target_keys, List):
         return data_frame[target_keys].tolist()
@@ -63,6 +64,7 @@ def resolve_files(
         root: The root path to use when resolving files.
         resolver: The resolver function to use. This function should receive the root and a file ID as input and return
             the path to an existing file.
+
     """
     if resolver is None:
         resolver = default_resolver