PYTHON-1648 Binary buffer protocol support

bson/binary.py

@@ -133,22 +133,25 @@ class Binary(bytes):
       directly to :class:`bytes`.
 
     :Parameters:
-      - `data`: the binary data to represent
+      - `data`: the binary data to represent. Can be any bytes-like type
+        that implements the buffer protocol.
       - `subtype` (optional): the `binary subtype
         <http://bsonspec.org/#/specification>`_
         to use
+
+    .. versionchanged:: 3.9
+      Support any bytes-like type that implements the buffer protocol.
     """
 
     _type_marker = 5
 
     def __new__(cls, data, subtype=BINARY_SUBTYPE):
-        if not isinstance(data, bytes):
-            raise TypeError("data must be an instance of bytes")
         if not isinstance(subtype, int):
             raise TypeError("subtype must be an instance of int")
         if subtype >= 256 or subtype < 0:
             raise ValueError("subtype must be contained in [0, 256)")
-        self = bytes.__new__(cls, data)
+        # Support any type that implements the buffer protocol.
+        self = bytes.__new__(cls, memoryview(data).tobytes())
         self.__subtype = subtype
         return self
 

doc/changelog.rst

@@ -60,6 +60,8 @@ Version 3.9 adds support for MongoDB 4.2. Highlights include:
   :meth:`~pymongo.collection.Collection.find_one_and_update`,
   :meth:`~pymongo.operations.UpdateOne`, and
   :meth:`~pymongo.operations.UpdateMany`.
+- :class:`~bson.binary.Binary` now supports any bytes-like type that implements
+  the buffer protocol.
 
 .. _URI options specification: https://github.com/mongodb/specifications/blob/master/source/uri-options/uri-options.rst
 

test/test_binary.py

@@ -14,9 +14,11 @@
 
 """Tests for the Binary wrapper."""
 
+import array
 import base64
 import copy
 import pickle
+import platform
 import sys
 import uuid
 
@@ -26,6 +28,7 @@ import bson
 
 from bson.binary import *
 from bson.codec_options import CodecOptions
+from bson.py3compat import PY3
 from bson.son import SON
 from pymongo.mongo_client import MongoClient
 from test import client_context, unittest
@@ -81,7 +84,6 @@ class TestBinary(unittest.TestCase):
 
     def test_exceptions(self):
         self.assertRaises(TypeError, Binary, None)
-        self.assertRaises(TypeError, Binary, u"hello")
         self.assertRaises(TypeError, Binary, 5)
         self.assertRaises(TypeError, Binary, 10.2)
         self.assertRaises(TypeError, Binary, b"hello", None)
@@ -90,6 +92,10 @@ class TestBinary(unittest.TestCase):
         self.assertRaises(ValueError, Binary, b"hello", 256)
         self.assertTrue(Binary(b"hello", 0))
         self.assertTrue(Binary(b"hello", 255))
+        if platform.python_implementation() != "Jython":
+            # Jython's memoryview accepts unicode strings...
+            # https://bugs.jython.org/issue2784
+            self.assertRaises(TypeError, Binary, u"hello")
 
     def test_subtype(self):
         one = Binary(b"hello")
@@ -356,6 +362,22 @@ class TestBinary(unittest.TestCase):
         for proto in range(pickle.HIGHEST_PROTOCOL + 1):
             self.assertEqual(uul, pickle.loads(pickle.dumps(uul, proto)))
 
+    def test_buffer_protocol(self):
+        b0 = Binary(b'123', 2)
+
+        self.assertEqual(b0, Binary(memoryview(b'123'), 2))
+        self.assertEqual(b0, Binary(bytearray(b'123'), 2))
+        # mmap.mmap and array.array only expose the
+        # buffer interface in python 3.x
+        if PY3:
+            # No mmap module in Jython
+            import mmap
+            with mmap.mmap(-1, len(b'123')) as mm:
+                mm.write(b'123')
+                mm.seek(0)
+                self.assertEqual(b0, Binary(mm, 2))
+            self.assertEqual(b0, Binary(array.array('B', b'123'), 2))
+
 
 if __name__ == "__main__":
     unittest.main()