Fix bug where opening files in GCS forgets to define a name field (#506)

* Fix bug where opening files in GCS forgets to define a name field * update unit tests to check .name * add round trip test * Update CHANGELOG.md Co-authored-by: Michael Penkov <[email protected]> Co-authored-by: Michael Penkov <[email protected]>
piskvorky · Jun 17, 2020 · bac9ab0 · bac9ab0
1 parent 0a36191
commit bac9ab0
Show file tree

Hide file tree

Showing 3 changed files with 19 additions and 2 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -2,6 +2,7 @@
 
   - Correctly pass `newline` parameter to built-in `open` function (PR [#478](https://github.com/RaRe-Technologies/smart_open/pull/478), [@burkovae](https://github.com/burkovae))
   - Azure storage blob support ([@nclsmitchell](https://github.com/nclsmitchell) and [@petedannemann](https://github.com/petedannemann))
+  - Ensure GCS objects always have a .name attribute (PR [#506](https://github.com/RaRe-Technologies/smart_open/pull/506), [@todor-markov](https://github.com/todor-markov))
 
 # 2.0.0, 27 April 2020, "Python 3"
 

diff --git a/smart_open/gcs.py b/smart_open/gcs.py
@@ -129,15 +129,15 @@ def open(
 
     """
     if mode == constants.READ_BINARY:
-        return Reader(
+        fileobj = Reader(
             bucket_id,
             blob_id,
             buffer_size=buffer_size,
             line_terminator=constants.BINARY_NEWLINE,
             client=client,
         )
     elif mode == constants.WRITE_BINARY:
-        return Writer(
+        fileobj = Writer(
             bucket_id,
             blob_id,
             min_part_size=min_part_size,
@@ -146,6 +146,9 @@ def open(
     else:
         raise NotImplementedError('GCS support for mode %r not implemented' % mode)
 
+    fileobj.name = blob_id
+    return fileobj
+
 
 class _RawReader(object):
     """Read an GCS object."""

diff --git a/smart_open/tests/test_gcs.py b/smart_open/tests/test_gcs.py
@@ -888,13 +888,26 @@ def test_read_never_returns_none(self):
         """read should never return None."""
         test_string = u"ветер по морю гуляет..."
         with smart_open.gcs.open(BUCKET_NAME, BLOB_NAME, "wb") as fout:
+            self.assertEqual(fout.name, BLOB_NAME)
             fout.write(test_string.encode('utf8'))
 
         r = smart_open.gcs.open(BUCKET_NAME, BLOB_NAME, "rb")
+        self.assertEqual(r.name, BLOB_NAME)
         self.assertEqual(r.read(), test_string.encode("utf-8"))
         self.assertEqual(r.read(), b"")
         self.assertEqual(r.read(), b"")
 
+    def test_round_trip(self):
+        test_string = u"ветер по морю гуляет..."
+        url = 'gs://%s/%s' % (BUCKET_NAME, BLOB_NAME)
+        with smart_open.open(url, "w") as fout:
+            fout.write(test_string)
+
+        with smart_open.open(url) as fin:
+            actual = fin.read()
+
+        self.assertEqual(test_string, actual)
+
 
 class MakeRangeStringTest(unittest.TestCase):
     def test_no_stop(self):