This is an automated email from the ASF dual-hosted git repository.
shunping pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/beam.git
The following commit(s) were added to refs/heads/master by this push:
new 980c11432a1 Clean up legacy references to apitools in GCS I/O (#39433)
980c11432a1 is described below
commit 980c11432a1f5b068f8732654a7a7af63fbd295e
Author: Jack McCluskey <[email protected]>
AuthorDate: Thu Jul 30 21:20:49 2026 -0400
Clean up legacy references to apitools in GCS I/O (#39433)
---
sdks/python/apache_beam/io/filesystemio.py | 5 ++---
sdks/python/apache_beam/io/gcp/__init__.py | 20 --------------------
sdks/python/apache_beam/io/gcp/gcsfilesystem_test.py | 4 +++-
3 files changed, 5 insertions(+), 24 deletions(-)
diff --git a/sdks/python/apache_beam/io/filesystemio.py
b/sdks/python/apache_beam/io/filesystemio.py
index 571d1f2d269..daa02e58601 100644
--- a/sdks/python/apache_beam/io/filesystemio.py
+++ b/sdks/python/apache_beam/io/filesystemio.py
@@ -284,9 +284,8 @@ class PipeStream(object):
return self.position
def seek(self, offset, whence=os.SEEK_SET):
- # The apitools library used by the gcsio.Uploader class insists on seeking
- # to the end of a stream to do a check before completing an upload, so we
- # must have this no-op method here in that case.
+ # Certain upload stream implementations seek to the end of a stream to
check
+ # length before completing an upload, so we support a no-op seek(0,
SEEK_END).
if whence == os.SEEK_END and offset == 0:
return
elif whence == os.SEEK_SET:
diff --git a/sdks/python/apache_beam/io/gcp/__init__.py
b/sdks/python/apache_beam/io/gcp/__init__.py
index 861a39f5c75..cce3acad34a 100644
--- a/sdks/python/apache_beam/io/gcp/__init__.py
+++ b/sdks/python/apache_beam/io/gcp/__init__.py
@@ -14,23 +14,3 @@
# See the License for the specific language governing permissions and
# limitations under the License.
#
-
-# Important: the MIME library in the Python 3.x standard library used by
-# apitools causes uploads containing '\r\n' to be corrupted, unless we
-# patch the BytesGenerator class to write contents verbatim.
-try:
- # pylint: disable=wrong-import-order, wrong-import-position
- # pylint: disable=ungrouped-imports
- import email.generator as email_generator
-
- from apitools.base.py import transfer
-
- class _WrapperNamespace(object):
- class BytesGenerator(email_generator.BytesGenerator):
- def _write_lines(self, lines):
- self.write(lines)
-
- transfer.email_generator = _WrapperNamespace
-except ImportError:
- # We may not have the GCP dependencies installed, so we pass in this case.
- pass
diff --git a/sdks/python/apache_beam/io/gcp/gcsfilesystem_test.py
b/sdks/python/apache_beam/io/gcp/gcsfilesystem_test.py
index 08fdd630288..0ab8c4c48f8 100644
--- a/sdks/python/apache_beam/io/gcp/gcsfilesystem_test.py
+++ b/sdks/python/apache_beam/io/gcp/gcsfilesystem_test.py
@@ -29,9 +29,11 @@ from apache_beam.io.filesystem import BeamIOError
from apache_beam.io.filesystem import FileMetadata
from apache_beam.options.pipeline_options import PipelineOptions
-# Protect against environments where apitools library is not available.
+# Protect against environments where GCP storage library is not available.
# pylint: disable=wrong-import-order, wrong-import-position
try:
+ from google.cloud import storage # pylint: disable=unused-import
+
from apache_beam.io.gcp import gcsfilesystem
except ImportError:
gcsfilesystem = None # type: ignore
