PR Review: fix pre-commit errors

src/mdio/schemas/v1/dataset_builder.py

@@ -25,16 +25,13 @@
 from mdio.schemas.v1.variable import Coordinate
 from mdio.schemas.v1.variable import Variable
 
-AnyMetadataList: TypeAlias = list[AllUnits |
-                                  UserAttributes |
-                                  ChunkGridMetadata |
-                                  StatisticsMetadata |
-                                  DatasetInfo]
+AnyMetadataList: TypeAlias = list[
+    AllUnits | UserAttributes | ChunkGridMetadata | StatisticsMetadata | DatasetInfo
+]
 CoordinateMetadataList: TypeAlias = list[AllUnits | UserAttributes]
-VariableMetadataList: TypeAlias = list[AllUnits |
-                                       UserAttributes |
-                                       ChunkGridMetadata |
-                                       StatisticsMetadata]
+VariableMetadataList: TypeAlias = list[
+    AllUnits | UserAttributes | ChunkGridMetadata | StatisticsMetadata
+]
 DatasetMetadataList: TypeAlias = list[DatasetInfo | UserAttributes]
 
 
@@ -52,7 +49,7 @@
 ) -> NamedDimension | None:
     """Get a dimension by name and optional size from the list[NamedDimension]."""
     if dimensions is None:
         return False
     if not isinstance(name, str):
         msg = f"Expected str, got {type(name).__name__}"
         raise TypeError(msg)
@@ -97,17 +94,12 @@
     """
 
     def __init__(self, name: str, attributes: UserAttributes | None = None):
-
         try:
             api_version = metadata.version("multidimio")
         except metadata.PackageNotFoundError:
             api_version = "unknown"
 
-        self._info = DatasetInfo(
-            name=name,
-            api_version=api_version,
-            created_on=datetime.now(UTC)
-        )
+        self._info = DatasetInfo(name=name, api_version=api_version, created_on=datetime.now(UTC))
         self._attributes = attributes
         self._dimensions: list[NamedDimension] = []
         self._coordinates: list[Coordinate] = []
@@ -116,9 +108,7 @@
         self._unnamed_variable_counter = 0
 
     def add_dimension(  # noqa: PLR0913
-        self,
-        name: str,
-        size: int
+        self, name: str, size: int
     ) -> "MDIODatasetBuilder":
         """Add a dimension.
 
@@ -128,6 +118,10 @@
             name: Name of the dimension
             size: Size of the dimension
 
+        Raises:
+            ValueError: If 'name' is not a non-empty string.
+                        if the dimension is already defined.
+
         Returns:
             self: Returns self for method chaining
         """
@@ -146,7 +140,6 @@
         self._state = _BuilderState.HAS_DIMENSIONS
         return self
 
-
     def add_coordinate(  # noqa: PLR0913
         self,
         name: str,
@@ -170,6 +163,13 @@
             compressor: Compressor used for the variable (defaults to None)
             metadata_info: Optional metadata information for the coordinate
 
+        Raises:
+            ValueError: If no dimensions have been added yet.
+                        If 'name' is not a non-empty string.
+                        If 'dimensions' is not a non-empty list.
+                        If the coordinate is already defined.
+                        If any referenced dimension is not already defined.
+
         Returns:
             self: Returns self for method chaining
         """
@@ -195,7 +195,7 @@
             if nd is None:
                 msg = f"Pre-existing dimension named {dim_name!r} is not found"
                 raise ValueError(msg)
-            named_dimensions.append(nd)    
+            named_dimensions.append(nd)
 
         meta_dict = _to_dictionary(metadata_info)
         coord = Coordinate(
@@ -204,7 +204,7 @@
             dimensions=named_dimensions,
             compressor=compressor,
             dataType=data_type,
-            metadata=meta_dict
+            metadata=meta_dict,
         )
         self._coordinates.append(coord)
 
@@ -216,7 +216,7 @@
             data_type=coord.data_type,
             compressor=compressor,
             coordinates=[name],  # Use the coordinate name as a reference
-            metadata_info=coord.metadata
+            metadata_info=coord.metadata,
         )
 
         self._state = _BuilderState.HAS_COORDINATES
@@ -252,6 +252,14 @@
                          (defaults to None, meaning no coordinates)
             metadata_info: Optional metadata information for the variable
 
+        Raises:
+            ValueError: If no dimensions have been added yet.
+                        If 'name' is not a non-empty string.
+                        If 'dimensions' is not a non-empty list.
+                        If the variable is already defined.
+                        If any referenced dimension is not already defined.
+                        If any referenced coordinate is not already defined.
+
         Returns:
             self: Returns self for method chaining.
         """
@@ -293,7 +301,7 @@
 
         # If this is a dimension coordinate variable, embed the Coordinate into it
         if coordinates is not None and len(coordinates) == 1 and coordinates[0] == name:
-                coordinates = coordinate_objs
+            coordinates = coordinate_objs
 
         meta_dict = _to_dictionary(metadata_info)
         var = Variable(
@@ -303,7 +311,8 @@
             data_type=data_type,
             compressor=compressor,
             coordinates=coordinates,
-            metadata=meta_dict)
+            metadata=meta_dict,
+        )
         self._variables.append(var)
 
         self._state = _BuilderState.HAS_VARIABLES
@@ -315,12 +324,15 @@
         This function must be called after at least one dimension is added via add_dimension().
         It will create a Dataset object with all added dimensions, coordinates, and variables.
 
+        Raises:
+            ValueError: If no dimensions have been added yet.
+
         Returns:
             Dataset: The built dataset with all added dimensions, coordinates, and variables.
         """
         if self._state == _BuilderState.INITIAL:
             msg = "Must add at least one dimension before building"
             raise ValueError(msg)
 
         var_meta_dict = _to_dictionary([self._info, self._attributes])
         return Dataset(variables=self._variables, metadata=var_meta_dict)

tests/unit/v1/helpers.py

@@ -1,4 +1,5 @@
 """Helper methods used in unit tests."""
+
 from mdio.schemas.chunk_grid import RegularChunkGrid
 from mdio.schemas.chunk_grid import RegularChunkShape
 from mdio.schemas.compressors import Blosc
@@ -23,22 +24,19 @@
 from mdio.schemas.v1.variable import Variable
 
 
-def validate_builder(builder: MDIODatasetBuilder,
-                     state: _BuilderState,
-                     n_dims: int,
-                     n_coords: int,
-                     n_var: int) -> None:
+def validate_builder(
+    builder: MDIODatasetBuilder, state: _BuilderState, n_dims: int, n_coords: int, n_var: int
+) -> None:
     """Validate the state of the builder, the number of dimensions, coordinates, and variables."""
     assert builder._state == state
     assert len(builder._dimensions) == n_dims
     assert len(builder._coordinates) == n_coords
     assert len(builder._variables) == n_var
 
 
-def validate_coordinate(builder: MDIODatasetBuilder,
-                        name: str,
-                        dims: list[tuple[str, int]],
-                        dtype: ScalarType) -> Coordinate:
+def validate_coordinate(
+    builder: MDIODatasetBuilder, name: str, dims: list[tuple[str, int]], dtype: ScalarType
+) -> Coordinate:
     """Validate existence and the structure of the created coordinate."""
     # Validate that coordinate exists
     c = next((c for c in builder._coordinates if c.name == name), None)
@@ -55,11 +53,13 @@
     return c
 
 
-def validate_variable(container: MDIODatasetBuilder | Dataset,
-                      name: str,
-                      dims: list[tuple[str, int]],
-                      coords: list[str],
-                      dtype: ScalarType) -> Variable:
+def validate_variable(
+    container: MDIODatasetBuilder | Dataset,
+    name: str,
+    dims: list[tuple[str, int]],
+    coords: list[str],
+    dtype: ScalarType,
+) -> Variable:
     """Validate existence and the structure of the created variable."""
     if isinstance(container, MDIODatasetBuilder):
         var_list = container._variables
@@ -68,8 +68,8 @@
         var_list = container.variables
         global_coord_list = _get_all_coordinates(container)
     else:
         err_msg = f"Expected MDIODatasetBuilder or Dataset, got {type(container)}"
         raise TypeError(err_msg)
 
     # Validate that the variable exists
     v = next((e for e in var_list if e.name == name), None)
@@ -90,23 +90,24 @@
         assert len(v.coordinates) == len(coords)
         for coord_name in coords:
             assert _get_coordinate(global_coord_list, v.coordinates, coord_name) is not None
-            
+
     assert v.data_type == dtype
     return v
 
 
 def _get_coordinate(
-        global_coord_list: list[Coordinate],
-        coordinates_or_references: list[Coordinate] | list[str],
-        name: str) -> Coordinate | None:     
+    global_coord_list: list[Coordinate],
+    coordinates_or_references: list[Coordinate] | list[str],
+    name: str,
+) -> Coordinate | None:
     """Get a coordinate by name from the list[Coordinate] | list[str].
 
-    The function validates that the coordinate referenced by the name can be found 
-    in the global coordinate list.   
+    The function validates that the coordinate referenced by the name can be found
+    in the global coordinate list.
     If the coordinate is stored as a Coordinate object, it is returned directly.
     """
     if coordinates_or_references is None:
         return None
 
     for c in coordinates_or_references:
         if isinstance(c, str) and c == name:
@@ -114,18 +115,17 @@
             cc = None
             # Find the Coordinate in the global list and return it.
             if global_coord_list is not None:
-                cc = next(
-                    (cc for cc in global_coord_list if cc.name == name), None)
+                cc = next((cc for cc in global_coord_list if cc.name == name), None)
             if cc is None:
                 msg = f"Pre-existing coordinate named {name!r} is not found"
                 raise ValueError(msg)
             return cc
         if isinstance(c, Coordinate) and c.name == name:
             # The coordinate is stored as an embedded Coordinate object.
             # Return it.
             return c
 
     return None
 
 
 def _get_all_coordinates(dataset: Dataset) -> list[Coordinate]:
@@ -142,39 +142,42 @@
     """Create in-memory campos_3d dataset."""
     ds = MDIODatasetBuilder(
         "campos_3d",
-        attributes=UserAttributes(attributes={
-            "textHeader": [
-                "C01 .......................... ",
-                "C02 .......................... ",
-                "C03 .......................... ",
-            ],
-            "foo": "bar"
-        }))
+        attributes=UserAttributes(
+            attributes={
+                "textHeader": [
+                    "C01 .......................... ",
+                    "C02 .......................... ",
+                    "C03 .......................... ",
+                ],
+                "foo": "bar",
+            }
+        ),
+    )
 
     # Add dimensions
     ds.add_dimension("inline", 256)
     ds.add_dimension("crossline", 512)
     ds.add_dimension("depth", 384)
-    ds.add_coordinate("inline", dimensions=["inline"], data_type=ScalarType.UINT32) 
-    ds.add_coordinate("crossline", dimensions=["crossline"], data_type=ScalarType.UINT32) 
-    ds.add_coordinate("depth", dimensions=["depth"], data_type=ScalarType.FLOAT64, 
-                      metadata_info=[
-                          AllUnits(units_v1=LengthUnitModel(length=LengthUnitEnum.METER))
-                      ])
+    ds.add_coordinate("inline", dimensions=["inline"], data_type=ScalarType.UINT32)
+    ds.add_coordinate("crossline", dimensions=["crossline"], data_type=ScalarType.UINT32)
+    ds.add_coordinate(
+        "depth",
+        dimensions=["depth"],
+        data_type=ScalarType.FLOAT64,
+        metadata_info=[AllUnits(units_v1=LengthUnitModel(length=LengthUnitEnum.METER))],
+    )
     # Add coordinates
     ds.add_coordinate(
         "cdp-x",
         dimensions=["inline", "crossline"],
         data_type=ScalarType.FLOAT32,
-        metadata_info=[
-            AllUnits(units_v1=LengthUnitModel(length=LengthUnitEnum.METER))]
+        metadata_info=[AllUnits(units_v1=LengthUnitModel(length=LengthUnitEnum.METER))],
     )
     ds.add_coordinate(
         "cdp-y",
         dimensions=["inline", "crossline"],
         data_type=ScalarType.FLOAT32,
-        metadata_info=[
-            AllUnits(units_v1=LengthUnitModel(length=LengthUnitEnum.METER))]
+        metadata_info=[AllUnits(units_v1=LengthUnitModel(length=LengthUnitEnum.METER))],
     )
 
     # Add image variable
@@ -187,7 +190,8 @@
         metadata_info=[
             ChunkGridMetadata(
                 chunk_grid=RegularChunkGrid(
-                    configuration=RegularChunkShape(chunk_shape=[128, 128, 128]))
+                    configuration=RegularChunkShape(chunk_shape=[128, 128, 128])
+                )
             ),
             StatisticsMetadata(
                 stats_v1=SummaryStatistics(
@@ -196,13 +200,12 @@
                     sumSquares=125.12,
                     min=5.61,
                     max=10.84,
-                    histogram=CenteredBinHistogram(
-                        binCenters=[1, 2], counts=[10, 15]),
+                    histogram=CenteredBinHistogram(binCenters=[1, 2], counts=[10, 15]),
                 )
             ),
-            UserAttributes(
-                attributes={"fizz": "buzz", "UnitSystem": "Canonical"}),
-        ])
+            UserAttributes(attributes={"fizz": "buzz", "UnitSystem": "Canonical"}),
+        ],
+    )
     # Add velocity variable
     ds.add_variable(
         name="velocity",
@@ -212,10 +215,10 @@
         metadata_info=[
             ChunkGridMetadata(
                 chunk_grid=RegularChunkGrid(
-                    configuration=RegularChunkShape(chunk_shape=[128, 128, 128]))
+                    configuration=RegularChunkShape(chunk_shape=[128, 128, 128])
+                )
             ),
-            AllUnits(units_v1=SpeedUnitModel(
-                speed=SpeedUnitEnum.METER_PER_SECOND)),
+            AllUnits(units_v1=SpeedUnitModel(speed=SpeedUnitEnum.METER_PER_SECOND)),
         ],
     )
     # Add inline-optimized image variable
@@ -229,8 +232,10 @@
         metadata_info=[
             ChunkGridMetadata(
                 chunk_grid=RegularChunkGrid(
-                    configuration=RegularChunkShape(chunk_shape=[4, 512, 512]))
-            )]
+                    configuration=RegularChunkShape(chunk_shape=[4, 512, 512])
+                )
+            )
+        ],
     )
     # Add headers variable with structured dtype
     ds.add_variable(
@@ -240,7 +245,7 @@
             fields=[
                 StructuredField(name="cdp-x", format=ScalarType.FLOAT32),
                 StructuredField(name="cdp-y", format=ScalarType.FLOAT32),
-                StructuredField(name="inline", format=ScalarType.UINT32),       
+                StructuredField(name="inline", format=ScalarType.UINT32),
                 StructuredField(name="crossline", format=ScalarType.UINT32),
             ]
         ),