🧾 Update all copyright headers (and added back tests).

Doing this as 2025 is near (2024 is the first year I started putting in headers).

koila/__init__.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from . import constants, gpus
 from .eager import EagerTensor
 from .errors import UnsupportedError

koila/constants.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from typing import Dict
 
 import torch

koila/eager.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from __future__ import annotations
 
 import logging

koila/errors.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from typing import NoReturn
 
 

koila/gpus.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from __future__ import annotations
 
 import math

koila/interfaces.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from __future__ import annotations
 
 import functools

koila/lazy.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from __future__ import annotations
 
 import builtins

koila/prepasses.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from __future__ import annotations
 
 import functools

koila/shapes.py

@@ -1,3 +1,5 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
 from __future__ import annotations
 
 import functools

tests/__init__.py

@@ -0,0 +1 @@
+# Copyright (c) RenChu Wang - All Rights Reserved

tests/common.py

@@ -0,0 +1,96 @@
+# Copyright (c) RenChu Wang - All Rights Reserved
+
+from __future__ import annotations
+
+import dataclasses as dcls
+import math
+import typing
+from dataclasses import dataclass
+from typing import Any, Callable, Dict, Sequence
+
+import numpy as np
+import torch
+from numpy import ndarray
+from torch import Tensor
+from torch.types import Number
+
+
+@dataclass(init=False)
+class ArgsKwargs:
+    def __init__(self, *args: Any, **kwargs: Any) -> None:
+        self.args = args
+        self.kwargs = kwargs
+
+    args: Sequence[Any] = dcls.field(default_factory=tuple)
+    kwargs: Dict[str, Any] = dcls.field(default_factory=dict)
+
+
+@dataclass(init=False)
+class Caller:
+    func: Callable[..., Any]
+    arguments: Sequence[ArgsKwargs] = dcls.field(default_factory=list)
+
+    def __init__(
+        self,
+        func: Callable[..., Any],
+        arguments: Sequence[ArgsKwargs | Sequence[Any] | Dict[str, Any]],
+    ) -> None:
+        self.func = func
+        self.arguments = []
+
+        for argument in arguments:
+            if isinstance(argument, Sequence):
+                argument = ArgsKwargs(*argument)
+
+            if isinstance(argument, dict):
+                assert all(isinstance(key, str) for key in argument.keys())
+                argument = ArgsKwargs(**argument)
+
+            self.arguments.append(argument)
+
+    def call(self) -> None:
+        for argument in self.arguments:
+            self.func(*argument.args, **argument.kwargs)
+
+
+def call(
+    func: Callable[..., Any],
+    arguments: Sequence[ArgsKwargs | Sequence[Any] | Dict[str, Any]],
+) -> None:
+    Caller(func, arguments=arguments).call()
+
+
+def assert_equal(
+    input: Tensor | ndarray | Number, other: Tensor | ndarray | Number
+) -> None:
+    if isinstance(input, ndarray) or isinstance(other, ndarray):
+        assert np.all(input == other), input != other
+        return
+
+    if isinstance(input, Tensor) or isinstance(other, Tensor):
+        assert typing.cast(Tensor, input == other).all(), input != other
+        return
+
+    assert input == other, [input, other]
+
+
+def assert_isclose(
+    input: Tensor | ndarray | Number, other: Tensor | ndarray | Number
+) -> None:
+    if isinstance(input, ndarray) or isinstance(other, ndarray):
+        assert np.allclose(input, other, atol=1e-5), [input, other]
+        return
+
+    if isinstance(input, Tensor) and isinstance(other, Tensor):
+        assert torch.allclose(input, other, atol=1e-5), [input, other]
+        return
+
+    assert math.isclose(input, other, abs_tol=1e-5), [input, other]
+
+
+def is_notimplemented(func: Callable[[], Any]) -> bool:
+    try:
+        func()
+        return False
+    except:
+        return True