def test_execute_without_replace(self, mock_hook, mock_hook2):
mock_hook.return_value.list.return_value = MOCK_FILES
mock_hook.return_value.download.return_value = b"testing"
mock_hook2.return_value.list.return_value = MOCK_FILES
operator = GCSToS3Operator(
task_id=TASK_ID,
bucket=GCS_BUCKET,
prefix=PREFIX,
delimiter=DELIMITER,
dest_aws_conn_id="aws_default",
dest_s3_key=S3_BUCKET,
replace=False,
)
# create dest bucket with all the files
hook = S3Hook(aws_conn_id='airflow_gcs_test')
bucket = hook.get_bucket('bucket')
bucket.create()
for mock_file in MOCK_FILES:
bucket.put_object(Key=mock_file, Body=b'testing')
# we expect nothing to be uploaded
# and all the MOCK_FILES to be present at the S3 bucket
uploaded_files = operator.execute(None)
self.assertEqual([], uploaded_files)
self.assertEqual(sorted(MOCK_FILES),
sorted(hook.list_keys('bucket', delimiter='/')))
def test_execute_with_acl_policy(self, mock_load_bytes, mock_gcs_hook,
mock_gcs_hook2):
mock_gcs_hook.return_value.list.return_value = MOCK_FILES
mock_gcs_hook.return_value.download.return_value = b"testing"
mock_gcs_hook2.return_value.list.return_value = MOCK_FILES
operator = GCSToS3Operator(
task_id=TASK_ID,
bucket=GCS_BUCKET,
prefix=PREFIX,
delimiter=DELIMITER,
dest_aws_conn_id="aws_default",
dest_s3_key=S3_BUCKET,
replace=False,
acl_policy=ACL_POLICY,
)
# Create dest bucket without files
hook = S3Hook(aws_conn_id='airflow_gcs_test')
bucket = hook.get_bucket('bucket')
bucket.create()
operator.execute(None)
# Make sure the acl_policy parameter is passed to the upload method
self.assertEqual(mock_load_bytes.call_args.kwargs['acl_policy'],
ACL_POLICY)
def test_execute_should_pass_dest_s3_extra_args_to_s3_hook(
self, s3_mock_hook, mock_hook, mock_hook2):
mock_hook.return_value.list.return_value = MOCK_FILES
mock_hook.return_value.download.return_value = b"testing"
mock_hook2.return_value.list.return_value = MOCK_FILES
s3_mock_hook.return_value = mock.Mock()
s3_mock_hook.parse_s3_url.return_value = mock.Mock()
operator = GCSToS3Operator(
task_id=TASK_ID,
bucket=GCS_BUCKET,
prefix=PREFIX,
delimiter=DELIMITER,
dest_aws_conn_id="aws_default",
dest_s3_key=S3_BUCKET,
replace=True,
dest_s3_extra_args={
"ContentLanguage": "value",
},
)
operator.execute(None)
s3_mock_hook.assert_called_once_with(
aws_conn_id='aws_default',
extra_args={'ContentLanguage': 'value'},
verify=None)
def test_execute_incremental(self, mock_hook, mock_hook2):
mock_hook.return_value.list.return_value = MOCK_FILES
mock_hook.return_value.download.return_value = b"testing"
mock_hook2.return_value.list.return_value = MOCK_FILES
operator = GCSToS3Operator(
task_id=TASK_ID,
bucket=GCS_BUCKET,
prefix=PREFIX,
delimiter=DELIMITER,
dest_aws_conn_id="aws_default",
dest_s3_key=S3_BUCKET,
replace=False,
)
# create dest bucket
hook = S3Hook(aws_conn_id='airflow_gcs_test')
bucket = hook.get_bucket('bucket')
bucket.create()
bucket.put_object(Key=MOCK_FILES[0], Body=b'testing')
# we expect all except first file in MOCK_FILES to be uploaded
# and all the MOCK_FILES to be present at the S3 bucket
uploaded_files = operator.execute(None)
assert sorted(MOCK_FILES[1:]) == sorted(uploaded_files)
assert sorted(MOCK_FILES) == sorted(
hook.list_keys('bucket', delimiter='/'))
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
import os
from datetime import datetime
from airflow import DAG
from airflow.providers.amazon.aws.transfers.gcs_to_s3 import GCSToS3Operator
BUCKET = os.getenv("BUCKET", "bucket")
S3_KEY = os.getenv("S3_KEY", "s3://<bucket>/<prefix>")
with DAG(
dag_id="example_gcs_to_s3",
start_date=datetime(2021, 1, 1),
tags=["example"],
catchup=False,
) as dag:
# [START howto_transfer_gcs_to_s3]
gcs_to_s3 = GCSToS3Operator(
task_id="gcs_to_s3",
bucket=BUCKET,
dest_s3_key=S3_KEY,
replace=True,
)
# [END howto_transfer_gcs_to_s3]