!1195 add validate for geswitch and merge

Merge pull request !1195 from jiangjinsheng/issue_geswitch

mindspore/ops/operations/control_ops.py

@@ -128,6 +128,7 @@ class GeSwitch(PrimitiveWithInfer):
         return (data, data)
 
     def infer_dtype(self, data_type, pred_type):
+        validator.check_subclass("data", data_type, (mstype.tensor,) + mstype.number_type, self.name)
         validator.check_tensor_type_same({"pred": pred_type}, [mstype.bool_], self.name)
         return (data_type, data_type)
 
@@ -153,7 +154,20 @@ class Merge(PrimitiveWithInfer):
         raise NotImplementedError
 
     def infer_shape(self, inputs):
+        validator.check_integer('inputs len', len(inputs), 0, Rel.GT, self.name)
+        input_0 = inputs[0]
+
+        for i in range(1, len(inputs)):
+            if inputs[i] != input_0:
+                raise ValueError(f"For \'{self.name}\', the shape of {i}th input should be same as "
+                                 f"first input {input_0}, but got {inputs[i]}.")
+
         return (inputs[0], [1])
 
     def infer_dtype(self, inputs):
+        args = {}
+        for i, item in enumerate(inputs):
+            args['inputs[%d]' % i] = item
+
+        validator.check_tensor_type_same(args, (mstype.bool_,) + mstype.number_type, self.name)
         return (inputs[0], mstype.int32)

mindspore/ops/operations/nn_ops.py

@@ -2084,7 +2084,7 @@ class GetNext(PrimitiveWithInfer):
     Note:
         GetNext op needs to be associated with network and also depends on the init_dataset interface,
         it can't be used directly as a single op.
-        For details, please refer to `nn.cell_wrapper.DataWrapper` source code.
+        For details, please refer to `nn.DataWrapper` source code.
 
     Args:
         types (list[:class:`mindspore.dtype`]): The type of the outputs.

tests/ut/python/ops/test_control_ops.py

@@ -33,7 +33,7 @@ def cond_data_test(x_init, y_init):
             super(Net, self).__init__()
             self.square = P.Square()
             self.add = P.TensorAdd()
-            self.value = Tensor(np.full((1), 3, dtype=np.float32))
+            self.value = Tensor(3, dtype=ms.float32)
             self.switch = P.GeSwitch()
             self.merge = P.Merge()
             self.less = P.Less()