def test_init(self):
sensor = WasbPrefixSensor(task_id='wasb_sensor',
dag=self.dag,
**self._config)
self.assertEqual(sensor.container_name, self._config['container_name'])
self.assertEqual(sensor.prefix, self._config['prefix'])
self.assertEqual(sensor.wasb_conn_id, self._config['wasb_conn_id'])
self.assertEqual(sensor.check_options, {})
self.assertEqual(sensor.timeout, self._config['timeout'])
sensor = WasbPrefixSensor(task_id='wasb_sensor',
dag=self.dag,
check_options={'timeout': 2},
**self._config)
self.assertEqual(sensor.check_options, {'timeout': 2})
def test_poke(self, mock_hook):
mock_instance = mock_hook.return_value
sensor = WasbPrefixSensor(task_id='wasb_sensor',
dag=self.dag,
check_options={'timeout': 2},
**self._config)
sensor.poke(None)
mock_instance.check_for_prefix.assert_called_once_with('container',
'prefix',
timeout=2)
def transform(self, src_operator: BaseOperator,
parent_fragment: DAGFragment,
upstream_fragments: List[DAGFragment]) -> DAGFragment:
"""
You need to add the ``wasb_conn_id`` to the source operator (or preferably DAG) for this to work.
The ``container_name`` and ``prefix`` for the blob based sensors are coped from the ``bucket_name``
and ``bucket_key`` of the s3 sensor, so make sure they are templatized for changing between
`s3://` and `wasb://` paths, etc. using config
"""
s3_key_sensor: S3KeySensor = src_operator
wasb_conn_id = s3_key_sensor.params.get('wasb_conn_id', None)
if not wasb_conn_id:
wasb_conn_id = self.dag.params.get('wasb_conn_id', None)
if not wasb_conn_id:
raise TransformerException(
"Could not find wasb_conn_id in operator or DAG params")
if s3_key_sensor.wildcard_match:
wasb_sensor_op = WasbWildcardPrefixSensor(
task_id=src_operator.task_id,
wasb_conn_id=wasb_conn_id,
container_name=s3_key_sensor.bucket_name,
wildcard_prefix=s3_key_sensor.bucket_key,
dag=self.dag)
else:
wasb_sensor_op = WasbPrefixSensor(
task_id=src_operator.task_id,
wasb_conn_id=wasb_conn_id,
container_name=s3_key_sensor.bucket_name,
prefix=s3_key_sensor.bucket_key,
dag=self.dag)
self.copy_op_attrs(wasb_sensor_op, src_operator)
self.sign_op(wasb_sensor_op)
return DAGFragment([wasb_sensor_op])
output_container = '222'
processing_file_prefix = ''
blob_service = WasbHook(wasb_conn_id=wasb_connection_id)
dag = DAG(
dag_id='azure_blob_reader',
default_args=default_args,
description='A dag to pull new images from blob and process them',
schedule_interval=timedelta(days=1),
)
new_files = WasbPrefixSensor(
task_id='new_files_sensor',
container_name=input_container,
prefix=processing_file_prefix,
wasb_conn_id=wasb_connection_id,
dag=dag,
)
def move_blobs_to_processing(**context):
results = blob_service.connection.list_blobs(
input_container, processing_file_prefix)
blobs_moved = 0
blob_urls = []
for blob in results:
print("\t Blob name: " + blob.name)
# Generate a SAS token for blob access
blob_input_url = blob_service.connection.make_blob_url(
input_container,
import airflow.utils.dates
from airflow.models import DAG
from airflow.operators.bash_operator import BashOperator
from airflow.contrib.sensors.wasb_sensor import WasbBlobSensor, WasbPrefixSensor
dag = DAG(dag_id="azure_blob_sensor",
start_date=airflow.utils.dates.days_ago(3),
schedule_interval="@once")
data_arrival_sensor = WasbBlobSensor(task_id="data_arrival_sensor",
container_name="landing",
blob_name="raw_data.csv",
wasb_conn_id="blob_default",
poke_interval=60,
timeout=60 * 60 * 24)
data_file_prefix_sensor = WasbPrefixSensor(task_id="data_file_prefix_sensor",
container_name="landing",
prefix="raw_",
wasb_conn_id="blob_default",
poke_interval=60,
timeout=60 * 60 * 24)
data_has_arrived = BashOperator(task_id="data_has_arrived",
bash_command="echo 'The data has arrived!'",
dag=dag)
[data_arrival_sensor, data_file_prefix_sensor] >> data_has_arrived