In NestedStructureCoder: fixed bug where a TensorSpec with name=None would get...

In NestedStructureCoder: fixed bug where a TensorSpec with name=None would get deserialized with name=''.

PiperOrigin-RevId: 257610414

tensorflow/python/saved_model/nested_structure_coder.py

@@ -423,6 +423,7 @@ class _TensorSpecCodec(object):
     return value.HasField("tensor_spec_value")
 
   def do_decode(self, value, decode_fn):
+    name = value.tensor_spec_value.name
     return tensor_spec.TensorSpec(
         shape=decode_fn(
             struct_pb2.StructuredValue(
@@ -430,7 +431,7 @@ class _TensorSpecCodec(object):
         dtype=decode_fn(
             struct_pb2.StructuredValue(
                 tensor_dtype_value=value.tensor_spec_value.dtype)),
-        name=value.tensor_spec_value.name)
+        name=(name if name else None))
 
 
 StructureCoder.register_codec(_TensorSpecCodec())

tensorflow/python/saved_model/nested_structure_coder_test.py

@@ -171,6 +171,22 @@ class NestedStructureTest(test.TestCase):
     decoded = self._coder.decode_proto(encoded)
     self.assertEqual(structure, decoded)
 
+  def testEncodeDecodeTensorSpecWithNoName(self):
+    structure = [tensor_spec.TensorSpec([1, 2, 3], dtypes.int64)]
+    self.assertTrue(self._coder.can_encode(structure))
+    encoded = self._coder.encode_structure(structure)
+    expected = struct_pb2.StructuredValue()
+    expected_list = expected.list_value
+    expected_tensor_spec = expected_list.values.add().tensor_spec_value
+    expected_tensor_spec.shape.dim.add().size = 1
+    expected_tensor_spec.shape.dim.add().size = 2
+    expected_tensor_spec.shape.dim.add().size = 3
+    expected_tensor_spec.name = ""
+    expected_tensor_spec.dtype = dtypes.int64.as_datatype_enum
+    self.assertEqual(expected, encoded)
+    decoded = self._coder.decode_proto(encoded)
+    self.assertEqual(structure, decoded)
+
   def testNotEncodable(self):
 
     class NotEncodable(object):