Replaces convert_inputs_to_tensors decorator with a better decorator

- Replaces `convert_inputs_to_tensors` decorator with `convert_to_tensors`,
     which will convert any `TensorLike` arguments by default and sync data types
     according to any specified data type constraints. This means that information
     does not need to be duplicated in the decorator like it used to be.

- Disallows unsafe downcasting in the new `convert_to_tensors` decorator.
     Previously, you could have:
     ```py
     x = tp.Tensor([1], dtype=tp.int32)
     x * 3.99
     ```
     which would yield a result equal to:
     ```py
     tp.Tensor([3])
     ```
     due to truncation when casting.

- Fixes several type annotations for binary elementwise ops.

tripy/docs/post0_developer_guides/how-to-add-new-ops.md

@@ -201,7 +201,7 @@ import tripy.frontend.utils as frontend_utils
 @export.public_api(document_under="tensor_operations")
 
 # The `convert_shape_inputs` decorator converts the specified function arguments into `tripy.Shape`s,
-# which would allow for using Python numbers and sequences. The `convert_inputs_to_tensors` decorator more generally converts
+# which would allow for using Python numbers and sequences. The `convert_to_tensors` decorator more generally converts
 # function arguments into Tripy tensors and is also commonly used in the codebase.
 @frontend_utils.convert_shape_inputs(["shape"])
 def theta(shape: Tuple[int], dim: int = 0, dtype: datatype.dtype = datatype.float32) -> "tripy.Tensor":

tripy/tests/constraints/object_builders.py

@@ -20,6 +20,7 @@
 
 import tripy as tp
 from tripy.common import datatype
+from tripy.types import TensorLike
 
 
 def tensor_builder(init, dtype, namespace):
@@ -64,6 +65,7 @@ def default_builder(init, dtype, namespace):
 find_func = {
     "tripy.Tensor": tensor_builder,
     "tripy.types.TensorLike": tensor_builder,
+    TensorLike: tensor_builder,
     "tripy.Shape": tensor_builder,
     "tripy.dtype": dtype_builder,
     datatype.dtype: dtype_builder,

tripy/tests/constraints/test_dtypes.py

@@ -121,7 +121,7 @@ def test_all_public_apis_verified(api):
                         exception = True
                         positive_case = False
 
-                ids = [f"{dtype_name}={dtype}" for dtype_name, dtype in namespace.items()]
+                ids = [f"{dtype_name}-{dtype}" for dtype_name, dtype in namespace.items()]
 
                 DTYPE_CONSTRAINT_CASES.append(
                     pytest.param(
@@ -160,11 +160,18 @@ def _run_dtype_constraints_subtest(test_data):
         args = [kwargs["self"]]
         del kwargs["self"]
 
+    def cast_to_bool(arg0, arg1):
+        if arg1.dtype == tp.bool:
+            return bool(arg0)
+        return arg0
+
     SPECIAL_FUNCS = {
-        "__radd__": (lambda self, other: self + other),
-        "__rsub__": (lambda self, other: self - other),
-        "__rpow__": (lambda self, other: self**other),
-        "__rmul__": (lambda self, other: self * other),
+        "__add__": (lambda self, other: self + cast_to_bool(other, self)),
+        "__mul__": (lambda self, other: self * cast_to_bool(other, self)),
+        "__radd__": (lambda self, other: cast_to_bool(self, other) + other),
+        "__rsub__": (lambda self, other: cast_to_bool(self, other) - other),
+        "__rpow__": (lambda self, other: cast_to_bool(self, other) ** other),
+        "__rmul__": (lambda self, other: cast_to_bool(self, other) * other),
         "__rtruediv__": (lambda self, other: self / other),
         "shape": (lambda self: self.shape),
     }
@@ -200,6 +207,7 @@ def _run_dtype_constraints_subtest(test_data):
 
         all_locals = locals()
         exec(dedent(code), globals(), all_locals)
+
         ret_val = all_locals["ret_val"]
 
     # If output does not have dtype skip .eval().

tripy/tests/frontend/test_shape.py

@@ -512,7 +512,7 @@ def test_quantize_rejected(self, values):
     def test_binary_elementwise_broadcast_rejected(self, values):
         with raises(
             tp.TripyException,
-            match=r"\_\_mul\_\_ expects tensor arguments to have matching class types, but got mixed `tp\.Tensor` and `tp\.Shape` arguments\.",
+            match=r"Error processing shape inputs in operator BinaryElementwise Further information: Binary elementwise operators do not accept combinations of Shape and Tensor arguments.",
         ):
             tp.Shape(values).multiply(tp.Tensor([values, values]))