refactor: replace dataclasses to attrs

Signed-off-by: ktro2828 <[email protected]>

t4_devkit/common/converter.py

@@ -0,0 +1,27 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+import numpy as np
+from pyquaternion import Quaternion
+
+if TYPE_CHECKING:
+    from t4_devkit.typing import ArrayLike, NDArray
+
+__all__ = ["as_quaternion"]
+
+
+def as_quaternion(value: ArrayLike | NDArray) -> Quaternion:
+    """Convert input rotation like array to `Quaternion`.
+
+    Args:
+        value (ArrayLike | NDArray): Rotation matrix or quaternion.
+
+    Returns:
+        Quaternion: Converted instance.
+    """
+    return (
+        Quaternion(matrix=value)
+        if isinstance(value, np.ndarray) and value.ndim == 2
+        else Quaternion(value)
+    )

t4_devkit/common/io.py

@@ -2,7 +2,7 @@
 import json
 from typing import Any
 
-__all__ = ("load_json",)
+__all__ = ("load_json", "save_json")
 
 
 def load_json(filename: str) -> Any:

t4_devkit/dataclass/box.py

@@ -1,13 +1,15 @@
 from __future__ import annotations
 
-from dataclasses import dataclass, field
 from typing import TYPE_CHECKING, TypeVar
 
 import numpy as np
-from pyquaternion import Quaternion
+from attrs import define, field
+from attrs.converters import optional
 from shapely.geometry import Polygon
 from typing_extensions import Self
 
+from t4_devkit.common.converter import as_quaternion
+
 from .roi import Roi
 from .trajectory import to_trajectories
 
@@ -57,7 +59,7 @@ def distance_box(box: BoxType, tf_matrix: HomogeneousMatrix) -> float | None:
     return np.linalg.norm(position)
 
 
-@dataclass(eq=False)
+@define(eq=False)
 class BaseBox:
     """Abstract base class for box objects."""
 
@@ -72,7 +74,7 @@ class BaseBox:
 # >>> e.g.) box.as_state() -> BoxState
 
 
-@dataclass(eq=False)
+@define(eq=False)
 class Box3D(BaseBox):
     """A class to represent 3D box.
 
@@ -109,25 +111,15 @@ class Box3D(BaseBox):
         ... )
     """
 
-    position: TranslationType
-    rotation: RotationType
+    position: TranslationType = field(converter=np.asarray)
+    rotation: RotationType = field(converter=as_quaternion)
     shape: Shape
-    velocity: VelocityType | None = field(default=None)
+    velocity: VelocityType | None = field(default=None, converter=optional(np.asarray))
     num_points: int | None = field(default=None)
 
     # additional attributes: set by `with_**`
     future: list[Trajectory] | None = field(default=None, init=False)
 
-    def __post_init__(self) -> None:
-        if not isinstance(self.position, np.ndarray):
-            self.position = np.array(self.position)
-
-        if not isinstance(self.rotation, Quaternion):
-            self.rotation = Quaternion(self.rotation)
-
-        if self.velocity is not None and not isinstance(self.velocity, np.ndarray):
-            self.velocity = np.array(self.velocity)
-
     def with_future(
         self,
         waypoints: list[TrajectoryType],
@@ -195,7 +187,7 @@ def corners(self, box_scale: float = 1.0) -> NDArrayF64:
         return np.dot(self.rotation.rotation_matrix, corners).T + self.position
 
 
-@dataclass(eq=False)
+@define(eq=False)
 class Box2D(BaseBox):
     """A class to represent 2D box.
 
@@ -222,15 +214,11 @@ class Box2D(BaseBox):
         >>> box2d = box2d.with_position(position=(1.0, 1.0, 1.0))
     """
 
-    roi: Roi | None = field(default=None)
+    roi: Roi | None = field(default=None, converter=lambda x: None if x is None else Roi(x))
 
     # additional attributes: set by `with_**`
     position: TranslationType | None = field(default=None, init=False)
 
-    def __post_init__(self) -> None:
-        if self.roi is not None and not isinstance(self.roi, Roi):
-            self.roi = Roi(self.roi)
-
     def with_position(self, position: TranslationType) -> Self:
         """Return a self instance setting `position` attribute.
 
@@ -240,7 +228,7 @@ def with_position(self, position: TranslationType) -> Self:
         Returns:
             Self instance after setting `position`.
         """
-        self.position = np.array(position) if not isinstance(position, np.ndarray) else position
+        self.position = np.asarray(position)
         return self
 
     def __eq__(self, other: Box2D | None) -> bool:

t4_devkit/dataclass/label.py

@@ -1,9 +1,9 @@
 from __future__ import annotations
 
 import warnings
-from dataclasses import dataclass, field
 from enum import Enum, auto, unique
 
+from attrs import define, field
 from typing_extensions import Self
 
 __all__ = ["LabelID", "SemanticLabel", "convert_label"]
@@ -58,7 +58,7 @@ def __eq__(self, other: str | LabelID) -> bool:
         return self.name == other.upper() if isinstance(other, str) else self.name == other.name
 
 
-@dataclass(frozen=True, eq=False)
+@define(frozen=True, eq=False)
 class SemanticLabel:
     """A dataclass to represent semantic labels.
 
@@ -70,7 +70,7 @@ class SemanticLabel:
 
     label: LabelID
     original: str | None = field(default=None)
-    attributes: list[str] = field(default_factory=list)
+    attributes: list[str] = field(factory=list)
 
     def __eq__(self, other: str | SemanticLabel) -> bool:
         return self.label == other if isinstance(other, str) else self.label == other.label

t4_devkit/dataclass/pointcloud.py

@@ -2,10 +2,10 @@
 
 import struct
 from abc import abstractmethod
-from dataclasses import dataclass
 from typing import TYPE_CHECKING, ClassVar, TypeVar
 
 import numpy as np
+from attrs import define, field
 
 if TYPE_CHECKING:
     from typing_extensions import Self
@@ -21,14 +21,18 @@
 ]
 
 
-@dataclass
+@define
 class PointCloud:
     """Abstract base dataclass for pointcloud data."""
 
-    points: NDArrayFloat
+    points: NDArrayFloat = field(converter=np.asarray)
 
-    def __post_init__(self) -> None:
-        assert self.points.shape[0] == self.num_dims()
+    @points.validator
+    def check_dims(self, attribute, value) -> None:
+        if value.shape[0] != self.num_dims():
+            raise ValueError(
+                f"Expected point dimension is {self.num_dims()}, but got {value.shape[0]}"
+            )
 
     @staticmethod
     @abstractmethod
@@ -74,7 +78,7 @@ def transform(self, matrix: NDArrayFloat) -> None:
         )[:3, :]
 
 
-@dataclass
+@define
 class LidarPointCloud(PointCloud):
     """A dataclass to represent lidar pointcloud."""
 
@@ -91,7 +95,7 @@ def from_file(cls, filepath: str) -> Self:
         return cls(points.T)
 
 
-@dataclass
+@define
 class RadarPointCloud(PointCloud):
     # class variables
     invalid_states: ClassVar[list[int]] = [0]
@@ -188,9 +192,9 @@ def from_file(
         return cls(points)
 
 
-@dataclass
+@define
 class SegmentationPointCloud(PointCloud):
-    labels: NDArrayU8
+    labels: NDArrayU8 = field(converter=lambda x: np.asarray(x, dtype=np.uint8))
 
     @staticmethod
     def num_dims() -> int:

t4_devkit/dataclass/roi.py

@@ -1,26 +1,24 @@
 from __future__ import annotations
 
-from dataclasses import dataclass
 from typing import TYPE_CHECKING
 
+from attrs import define, field
+
 if TYPE_CHECKING:
     from t4_devkit.typing import RoiType
 
 __all__ = ["Roi"]
 
 
-@dataclass
+@define
 class Roi:
-    roi: RoiType
+    roi: RoiType = field(converter=tuple)
 
     def __post_init__(self) -> None:
         assert len(self.roi) == 4, (
             "Expected roi is (x, y, width, height), " f"but got length with {len(self.roi)}."
         )
 
-        if not isinstance(self.roi, tuple):
-            self.roi = tuple(self.roi)
-
     @property
     def offset(self) -> tuple[int, int]:
         return self.roi[:2]

t4_devkit/dataclass/shape.py

@@ -1,10 +1,10 @@
 from __future__ import annotations
 
-from dataclasses import dataclass, field
 from enum import Enum, auto, unique
 from typing import TYPE_CHECKING
 
 import numpy as np
+from attrs import define, field
 from shapely.geometry import Polygon
 from typing_extensions import Self
 
@@ -35,7 +35,7 @@ def from_name(cls, name: str) -> Self:
         return cls.__members__[name]
 
 
-@dataclass
+@define
 class Shape:
     """A dataclass to represent the 3D box shape.
 
@@ -47,13 +47,10 @@ class Shape:
     """
 
     shape_type: ShapeType
-    size: SizeType
+    size: SizeType = field(converter=np.asarray)
     footprint: Polygon = field(default=None)
 
-    def __post_init__(self) -> None:
-        if not isinstance(self.size, np.ndarray):
-            self.size = np.array(self.size)
-
+    def __attrs_post_init__(self) -> None:
         if self.shape_type == ShapeType.POLYGON and self.footprint is None:
             raise ValueError("`footprint` must be specified for `POLYGON`.")
 

t4_devkit/dataclass/trajectory.py

@@ -1,17 +1,17 @@
 from __future__ import annotations
 
-from dataclasses import dataclass, field
 from typing import TYPE_CHECKING, Generator
 
 import numpy as np
+from attrs import define, field
 
 if TYPE_CHECKING:
     from t4_devkit.typing import TrajectoryType, TranslationType
 
 __all__ = ["Trajectory", "to_trajectories"]
 
 
-@dataclass
+@define
 class Trajectory:
     """A dataclass to represent trajectory.
 
@@ -41,14 +41,12 @@ class Trajectory:
         [2. 2. 2.]
     """
 
-    waypoints: TrajectoryType
+    waypoints: TrajectoryType = field(converter=np.asarray)
     confidence: float = field(default=1.0)
 
-    def __post_init__(self) -> None:
-        if not isinstance(self.waypoints, np.ndarray):
-            self.waypoints = np.array(self.waypoints)
-
-        assert self.waypoints.shape[1] == 3
+    def __attrs_post_init__(self) -> None:
+        if self.waypoints.shape[1] != 3:
+            raise ValueError("Trajectory dimension must be 3.")
 
     def __len__(self) -> int:
         return len(self.waypoints)