feat: Fix for UUID (#70)

This corrects the Marshalling for the UUID Scalar.
Referenced in: cloudquery/cloudquery#14825

It assumes Network-Order Bytes as per https://www.rfc-editor.org/rfc/rfc4122.html#section-4.1.1

This should work as it does pass the tests, but our testing framework is a little incomplete.
I'm not going to alter the test framework in this commit - But we should have a Marshalling test that confirms the expected bytes are created for each custom object type.

Author: hydratim

cloudquery/sdk/scalar/uuid.py

@@ -4,6 +4,8 @@
 
 class UUID(Scalar):
     def __init__(self, valid: bool = False, value: uuid.UUID = None):
+        if isinstance(value, uuid.UUID):
+            value = value.bytes
         super().__init__(valid, value)
 
     def __eq__(self, scalar: Scalar) -> bool:
@@ -27,11 +29,13 @@ def set(self, value: any):
             self._value = value.value
             return
 
-        if isinstance(value, uuid.UUID):
+        if isinstance(value, bytes):
             self._value = value
+        elif isinstance(value, uuid.UUID):
+            self._value = value.bytes
         elif isinstance(value, str):
             try:
-                self._value = uuid.UUID(value)
+                self._value = uuid.UUID(value).bytes
             except ValueError as e:
                 raise ScalarInvalidTypeError("Invalid type for UUID scalar") from e
         else: