def _upload_to_gcs(self, local_file_path, gcs_location):
gcs_bucket, gcs_object = self._get_gcs_bucket_and_name(gcs_location)
request = storage.StorageObjectsInsertRequest(bucket=gcs_bucket,
name=gcs_object)
_LOGGER.info('Starting GCS upload to %s...', gcs_location)
total_size = os.path.getsize(local_file_path)
from apitools.base.py import exceptions
try:
with open(local_file_path, 'rb') as stream:
upload = storage.Upload(stream, 'application/octet-stream',
total_size)
self._storage_client.objects.Insert(request, upload=upload)
except exceptions.HttpError as e:
reportable_errors = {
403: 'access denied',
404: 'bucket not found',
}
if e.status_code in reportable_errors:
raise IOError(
('Could not upload to GCS path %s: %s. Please verify '
'that credentials are valid and that you have write '
'access to the specified path.') %
(gcs_location, reportable_errors[e.status_code]))
raise
_LOGGER.info('Completed GCS upload to %s.', gcs_location)
def __init__(self, client, path, mime_type):
self._client = client
self._path = path
self._bucket, self._name = parse_gcs_path(path)
self._mime_type = mime_type
# Set up communication with child thread.
parent_conn, child_conn = multiprocessing.Pipe()
self._child_conn = child_conn
self._conn = parent_conn
# Set up uploader.
self._insert_request = (storage.StorageObjectsInsertRequest(
bucket=self._bucket, name=self._name))
self._upload = transfer.Upload(
PipeStream(self._child_conn),
self._mime_type,
chunksize=WRITE_CHUNK_SIZE)
self._upload.strategy = transfer.RESUMABLE_UPLOAD
# Start uploading thread.
self._upload_thread = threading.Thread(target=self._start_upload)
self._upload_thread.daemon = True
self._upload_thread.last_error = None
self._upload_thread.start()
def stage_file(self,
gcs_or_local_path,
file_name,
stream,
mime_type='application/octet-stream'):
"""Stages a file at a GCS or local path with stream-supplied contents."""
if not gcs_or_local_path.startswith('gs://'):
local_path = FileSystems.join(gcs_or_local_path, file_name)
logging.info('Staging file locally to %s', local_path)
with open(local_path, 'wb') as f:
f.write(stream.read())
return
gcs_location = FileSystems.join(gcs_or_local_path, file_name)
bucket, name = gcs_location[5:].split('/', 1)
request = storage.StorageObjectsInsertRequest(bucket=bucket, name=name)
logging.info('Starting GCS upload to %s...', gcs_location)
upload = storage.Upload(stream, mime_type)
try:
response = self._storage_client.objects.Insert(request,
upload=upload)
except exceptions.HttpError as e:
reportable_errors = {
403: 'access denied',
404: 'bucket not found',
}
if e.status_code in reportable_errors:
raise IOError(
('Could not upload to GCS path %s: %s. Please verify '
'that credentials are valid and that you have write '
'access to the specified path.') %
(gcs_or_local_path, reportable_errors[e.status_code]))
raise
logging.info('Completed GCS upload to %s', gcs_location)
return response
def __init__(self,
client,
path,
mode='w',
mime_type='application/octet-stream'):
self.client = client
self.path = path
self.mode = mode
self.bucket, self.name = parse_gcs_path(path)
self.closed = False
self.position = 0
# A small buffer to avoid CPU-heavy per-write pipe calls.
self.write_buffer = bytearray()
self.write_buffer_size = 128 * 1024
# Set up communication with uploading thread.
parent_conn, child_conn = multiprocessing.Pipe()
self.child_conn = child_conn
self.conn = parent_conn
# Set up uploader.
self.insert_request = (storage.StorageObjectsInsertRequest(
bucket=self.bucket, name=self.name))
self.upload = transfer.Upload(GcsBufferedWriter.PipeStream(child_conn),
mime_type,
chunksize=WRITE_CHUNK_SIZE)
self.upload.strategy = transfer.RESUMABLE_UPLOAD
# Start uploading thread.
self.upload_thread = threading.Thread(target=self._start_upload)
self.upload_thread.daemon = True
self.upload_thread.last_error = None
self.upload_thread.start()