def get_check_date(context):
# logging.info(context)
ds = context.get('ds')
dt = datetime.strptime(ds, '%Y-%m-%d')
(y, w, wd) = datetime.strptime(ds, '%Y-%m-%d').isocalendar()
params = context['params']
schedule = params.get('scheduling', 0)
start = params.get('start', 0)
if schedule > 0:
sub_n = (7 + (wd - schedule)) % 7
else:
sub_n = 0
sdt = timedelta(days=sub_n)
chk_d = datetime.strftime(dt - sdt, '%Y-%m-%d')
chk_d2 = datetime.strftime(dt - sdt, '%Y%m%d')
# logging.info(context["ti"].task_id)
logging.info(dt)
logging.info(chk_d)
logging.info(start)
partition = "dt='{}'".format(
chk_d if int(chk_d2) > int(start) else datetime.
strftime(datetime.strptime(start, '%Y%m%d'), '%Y-%m-%d'))
logging.info('Poking for table %s.%s, partition %s',
params.get('hive_db'), params.get('hive_table'),
partition)
from airflow.hooks.hive_hooks import HiveMetastoreHook
hook = HiveMetastoreHook(metastore_conn_id="metastore_default")
return hook.check_for_partition(params.get('hive_db'),
params.get('hive_table'), partition)
class HivePartitionSensor(BaseSensorOperator):
"""
Waits for a partition to show up in Hive.
Note: Because ``partition`` supports general logical operators, it
can be inefficient. Consider using NamedHivePartitionSensor instead if
you don't need the full flexibility of HivePartitionSensor.
:param table: The name of the table to wait for, supports the dot
notation (my_database.my_table)
:type table: string
:param partition: The partition clause to wait for. This is passed as
is to the metastore Thrift client ``get_partitions_by_filter`` method,
and apparently supports SQL like notation as in ``ds='2015-01-01'
AND type='value'`` and comparison operators as in ``"ds>=2015-01-01"``
:type partition: string
:param metastore_conn_id: reference to the metastore thrift service
connection id
:type metastore_conn_id: str
"""
template_fields = (
'schema',
'table',
'partition',
)
ui_color = '#C5CAE9'
@apply_defaults
def __init__(self,
table,
partition="ds='{{ ds }}'",
metastore_conn_id='metastore_default',
schema='default',
poke_interval=60 * 3,
*args,
**kwargs):
super(HivePartitionSensor, self).__init__(poke_interval=poke_interval,
*args,
**kwargs)
if not partition:
partition = "ds='{{ ds }}'"
self.metastore_conn_id = metastore_conn_id
self.table = table
self.partition = partition
self.schema = schema
def poke(self, context):
if '.' in self.table:
self.schema, self.table = self.table.split('.')
self.log.info('Poking for table {self.schema}.{self.table}, '
'partition {self.partition}'.format(**locals()))
if not hasattr(self, 'hook'):
from airflow.hooks.hive_hooks import HiveMetastoreHook
self.hook = HiveMetastoreHook(
metastore_conn_id=self.metastore_conn_id)
return self.hook.check_for_partition(self.schema, self.table,
self.partition)
class HivePartitionSensor(BaseSensorOperator):
"""
Waits for a partition to show up in Hive.
Note: Because ``partition`` supports general logical operators, it
can be inefficient. Consider using NamedHivePartitionSensor instead if
you don't need the full flexibility of HivePartitionSensor.
:param table: The name of the table to wait for, supports the dot
notation (my_database.my_table)
:type table: string
:param partition: The partition clause to wait for. This is passed as
is to the metastore Thrift client ``get_partitions_by_filter`` method,
and apparently supports SQL like notation as in ``ds='2015-01-01'
AND type='value'`` and comparison operators as in ``"ds>=2015-01-01"``
:type partition: string
:param metastore_conn_id: reference to the metastore thrift service
connection id
:type metastore_conn_id: str
"""
template_fields = ('schema', 'table', 'partition',)
ui_color = '#C5CAE9'
@apply_defaults
def __init__(
self,
table, partition="ds='{{ ds }}'",
metastore_conn_id='metastore_default',
schema='default',
poke_interval=60*3,
*args, **kwargs):
super(HivePartitionSensor, self).__init__(
poke_interval=poke_interval, *args, **kwargs)
if not partition:
partition = "ds='{{ ds }}'"
self.metastore_conn_id = metastore_conn_id
self.table = table
self.partition = partition
self.schema = schema
def poke(self, context):
if '.' in self.table:
self.schema, self.table = self.table.split('.')
self.log.info(
'Poking for table {self.schema}.{self.table}, '
'partition {self.partition}'.format(**locals()))
if not hasattr(self, 'hook'):
from airflow.hooks.hive_hooks import HiveMetastoreHook
self.hook = HiveMetastoreHook(
metastore_conn_id=self.metastore_conn_id)
return self.hook.check_for_partition(
self.schema, self.table, self.partition)
