fix 816 (#817)

README.rst

@@ -411,6 +411,8 @@ GCS Advanced Usage
 
 Additional keyword arguments can be propagated to the GCS open method (`docs <https://cloud.google.com/python/docs/reference/storage/latest/google.cloud.storage.blob.Blob#google_cloud_storage_blob_Blob_open>`__), which is used by ``smart_open`` under the hood, using the ``blob_open_kwargs`` transport parameter.
 
+Additionally keyword arguments can be propagated to the GCS ``get_blob`` method (`docs <https://cloud.google.com/python/docs/reference/storage/latest/google.cloud.storage.bucket.Bucket#google_cloud_storage_bucket_Bucket_get_blob>`__) when in a read-mode, using the ``get_blob_kwargs`` transport parameter.
+
 Additional blob properties (`docs <https://cloud.google.com/python/docs/reference/storage/latest/google.cloud.storage.blob.Blob#properties>`__) can be set before an upload, as long as they are not read-only, using the ``blob_properties`` transport parameter.
 
 .. code-block:: python
@@ -507,4 +509,3 @@ issues or pull requests there. Suggestions, pull requests and improvements welco
 
 ``smart_open`` is open source software released under the `MIT license <https://github.com/piskvorky/smart_open/blob/master/LICENSE>`_.
 Copyright (c) 2015-now `Radim Řehůřek <https://radimrehurek.com>`_.
-

smart_open/gcs.py

@@ -59,6 +59,7 @@ def open(
     buffer_size=None,
     min_part_size=_DEFAULT_MIN_PART_SIZE,
     client=None,  # type: google.cloud.storage.Client
+    get_blob_kwargs=None,
     blob_properties=None,
     blob_open_kwargs=None,
 ):
@@ -78,6 +79,9 @@ def open(
         The minimum part size for multipart uploads. For writing only.
     client: google.cloud.storage.Client, optional
         The GCS client to use when working with google-cloud-storage.
+    get_blob_kwargs: dict, optional
+        Additional keyword arguments to propagate to the bucket.get_blob
+        method of the google-cloud-storage library. For reading only.
     blob_properties: dict, optional
         Set properties on blob before writing. For writing only.
     blob_open_kwargs: dict, optional
@@ -95,6 +99,7 @@ def open(
         _blob = Reader(bucket=bucket_id,
                        key=blob_id,
                        client=client,
+                       get_blob_kwargs=get_blob_kwargs,
                        blob_open_kwargs=blob_open_kwargs)
 
     elif mode in (constants.WRITE_BINARY, 'w', 'wt'):
@@ -116,8 +121,11 @@ def Reader(bucket,
            buffer_size=None,
            line_terminator=None,
            client=None,
+           get_blob_kwargs=None,
            blob_open_kwargs=None):
 
+    if get_blob_kwargs is None:
+        get_blob_kwargs = {}
     if blob_open_kwargs is None:
         blob_open_kwargs = {}
     if client is None:
@@ -128,7 +136,7 @@ def Reader(bucket,
         warn_deprecated('line_terminator')
 
     bkt = client.bucket(bucket)
-    blob = bkt.get_blob(key)
+    blob = bkt.get_blob(key, **get_blob_kwargs)
 
     if blob is None:
         raise google.cloud.exceptions.NotFound(f'blob {key} not found in {bucket}')

smart_open/tests/test_gcs.py

@@ -45,6 +45,8 @@ def __init__(self, client, name=None):
         #
         self.client.register_bucket(self)
 
+        self.get_blob = mock.Mock(side_effect=self._get_blob)
+
     def blob(self, blob_id, **kwargs):
         return self.blobs.get(blob_id, FakeBlob(blob_id, self, **kwargs))
 
@@ -57,7 +59,7 @@ def delete(self):
     def exists(self):
         return self._exists
 
-    def get_blob(self, blob_id):
+    def _get_blob(self, blob_id, **kwargs):
         try:
             return self.blobs[blob_id]
         except KeyError as e:
@@ -300,6 +302,15 @@ def test_property_passthrough(self):
         for k, v in blob_properties.items():
             self.assertEqual(getattr(b, k), v)
 
+    def test_get_blob_kwargs_passthrough(self):
+        get_blob_kwargs = {'generation': '1111111111111111'}
+
+        with self.assertRaises(google.cloud.exceptions.NotFound):
+            smart_open.gcs.Reader(BUCKET_NAME, BLOB_NAME, get_blob_kwargs=get_blob_kwargs)
+
+        self.client.bucket(BUCKET_NAME) \
+            .get_blob.assert_called_once_with(BLOB_NAME, **get_blob_kwargs)
+
     def test_default_open_kwargs(self):
         smart_open.gcs.Writer(BUCKET_NAME, BLOB_NAME)