Expose is_supported_dtype to the public interface (#150)

Also take this opportunity to clean up a naming inconsistency;
NumPy types are "dtypes", core types are "types".

Author: manopapad

cunumeric/__init__.py

@@ -31,6 +31,7 @@
 from ._array.util import maybe_convert_to_np_ndarray
 from ._module import *
 from ._ufunc import *
+from ._utils.array import is_supported_dtype
 from ._utils.coverage import clone_module
 
 clone_module(_np, globals(), maybe_convert_to_np_ndarray)

cunumeric/_array/array.py

@@ -30,7 +30,7 @@
 )
 
 from .. import _ufunc
-from .._utils.array import calculate_volume, to_core_dtype
+from .._utils.array import calculate_volume, to_core_type
 from .._utils.coverage import FALLBACK_WARNING, clone_class, is_implemented
 from .._utils.linalg import dot_modes
 from .._utils.structure import deep_apply
@@ -128,7 +128,7 @@ def __init__(
                         for inp in inputs
                         if isinstance(inp, ndarray)
                     ]
-                core_dtype = to_core_dtype(dtype)
+                core_dtype = to_core_type(dtype)
                 self._thunk = runtime.create_empty_thunk(
                     sanitized_shape, core_dtype, inputs
                 )
@@ -660,7 +660,7 @@ def __contains__(self, item: Any) -> ndarray:
             args = (np.array(item, dtype=self.dtype),)
         if args[0].size != 1:
             raise ValueError("contains needs scalar item")
-        core_dtype = to_core_dtype(self.dtype)
+        core_dtype = to_core_type(self.dtype)
         return perform_unary_reduction(
             UnaryRedCode.CONTAINS,
             self,
@@ -1975,7 +1975,7 @@ def clip(
                 return convert_to_cunumeric_ndarray(
                     self.__array__().clip(args[0], args[1])
                 )
-        core_dtype = to_core_dtype(self.dtype)
+        core_dtype = to_core_type(self.dtype)
         extra_args = (Scalar(min, core_dtype), Scalar(max, core_dtype))
         return perform_unary_op(
             UnaryOpCode.CLIP, self, out=out, extra_args=extra_args
@@ -2971,7 +2971,7 @@ def var(
                 # FIXME(wonchanl): the following code blocks on mu to convert
                 # it to a Scalar object. We need to get rid of this blocking by
                 # allowing the extra arguments to be Legate stores
-                args=(Scalar(mu.__array__(), to_core_dtype(self.dtype)),),
+                args=(Scalar(mu.__array__(), to_core_type(self.dtype)),),
             )
         else:
             # TODO(https://github.com/nv-legate/cunumeric/issues/591)

cunumeric/_thunk/deferred.py

@@ -49,7 +49,7 @@
     normalize_axis_tuple,
 )
 
-from .._utils.array import is_advanced_indexing, to_core_dtype
+from .._utils.array import is_advanced_indexing, to_core_type
 from ..config import (
     BinaryOpCode,
     BitGeneratorDistribution,
@@ -1701,7 +1701,7 @@ def select(
             c_arr = c._broadcast(self.shape)
             task.add_input(c_arr)
             task.add_alignment(c_arr, out_arr)
-        task.add_scalar_arg(default, to_core_dtype(default.dtype))
+        task.add_scalar_arg(default, to_core_type(default.dtype))
         task.execute()
 
     # Create or extract a diagonal from a matrix

cunumeric/_utils/array.py

@@ -40,11 +40,24 @@
 }
 
 
-def is_supported_type(dtype: str | np.dtype[Any]) -> bool:
+def is_supported_dtype(dtype: str | np.dtype[Any]) -> bool:
+    """
+    Whether a NumPy dtype is supported by cuNumeric
+
+    Parameters
+    ----------
+    dtype : data-type
+        The dtype to query
+
+    Returns
+    -------
+    res : bool
+        True if `dtype` is a supported dtype
+    """
     return np.dtype(dtype) in SUPPORTED_DTYPES
 
 
-def to_core_dtype(dtype: str | np.dtype[Any]) -> ty.Type:
+def to_core_type(dtype: str | np.dtype[Any]) -> ty.Type:
     core_dtype = SUPPORTED_DTYPES.get(np.dtype(dtype))
     if core_dtype is None:
         raise TypeError(f"cuNumeric does not support dtype={dtype}")

cunumeric/runtime.py

@@ -23,7 +23,7 @@
 from legate.core import LEGATE_MAX_DIM, Scalar, TaskTarget, get_legate_runtime
 from legate.settings import settings as legate_settings
 
-from ._utils.array import calculate_volume, is_supported_type, to_core_dtype
+from ._utils.array import calculate_volume, is_supported_dtype, to_core_type
 from ._utils.stack import find_last_user_stacklevel
 from .config import (
     BitGeneratorOperation,
@@ -60,7 +60,7 @@ def thunk_from_scalar(
     from ._thunk.deferred import DeferredArray
 
     store = legate_runtime.create_store_from_scalar(
-        Scalar(bytes, to_core_dtype(dtype)),
+        Scalar(bytes, to_core_type(dtype)),
         shape=shape,
     )
     return DeferredArray(store)
@@ -377,7 +377,7 @@ def find_or_create_array_thunk(
         from ._thunk.deferred import DeferredArray
 
         assert isinstance(array, np.ndarray)
-        if not is_supported_type(array.dtype):
+        if not is_supported_dtype(array.dtype):
             raise TypeError(f"cuNumeric does not support dtype={array.dtype}")
 
         # We have to be really careful here to handle the case of
@@ -429,7 +429,7 @@ def find_or_create_array_thunk(
             # This is not a scalar so make a field.
             # We won't try to cache these bigger arrays.
             store = legate_runtime.create_store_from_buffer(
-                to_core_dtype(array.dtype),
+                to_core_type(array.dtype),
                 array.shape,
                 array.copy() if transfer == TransferType.MAKE_COPY else array,
                 # This argument should really be called "donate"

tests/integration/test_argsort.py

@@ -98,7 +98,7 @@ def test_structured_array_order(self):
         #     if self.deferred is None:
         #         if self.parent is None:
         #
-        # > assert self.runtime.is_supported_type(self.array.dtype)
+        # > assert self.runtime.is_supported_dtype(self.array.dtype)
         # E
         # AssertionError
         #

tests/integration/test_prod.py

@@ -222,7 +222,7 @@ def test_dtype_complex(self, dtype):
         # allclose hits assertion error:
         # File "/legate/cunumeric/cunumeric/eager.py", line 293,
         # in to_deferred_array
-        #   assert self.runtime.is_supported_type(self.array.dtype)
+        #   assert self.runtime.is_supported_dtype(self.array.dtype)
         #   AssertionError
         assert allclose(out_np, out_num)
 

tests/integration/test_searchsorted.py

@@ -83,7 +83,7 @@ def test_val_none(self):
             # cuNumeric raises AssertionError
             #       if self.deferred is None:
             #           if self.parent is None:
-            #    >          assert self.runtime.is_supported_type
+            #    >          assert self.runtime.is_supported_dtype
             #                    (self.array.dtype)
             #    E               AssertionError
             # cunumeric/cunumeric/eager.py:to_deferred_array()

tests/unit/cunumeric/test_utils_array.py

@@ -75,17 +75,17 @@ class Test_is_supported_dtype:
     @pytest.mark.parametrize("value", ["foo", 10, 10.2, (), set()])
     def test_type_bad(self, value) -> None:
         with pytest.raises(TypeError):
-            m.to_core_dtype(value)
+            m.to_core_type(value)
 
     @pytest.mark.parametrize("value", EXPECTED_SUPPORTED_DTYPES)
     def test_supported(self, value) -> None:
-        m.to_core_dtype(value)
+        m.to_core_type(value)
 
     # This is just a representative sample, not exhasutive
     @pytest.mark.parametrize("value", [np.float128, np.datetime64, [], {}])
     def test_unsupported(self, value) -> None:
         with pytest.raises(TypeError):
-            m.to_core_dtype(value)
+            m.to_core_type(value)
 
 
 @pytest.mark.parametrize(