def key_by(self, primary_key=True, partition_key=False, *columns):
# TODO implement keying by arbitrary columns
if columns:
raise NotImplementedError('keying by arbitrary columns is not (yet) supported')
if partition_key:
raise NotImplementedError('keying by partition key is not (yet) supported')
new = copy(self)
new._key_by = ColumnSelector(partition_key, primary_key, *columns)
new.crdd = self.crdd
return new
def __init__(self, ctx, keyspace, table, row_format=None, read_conf=None, **read_conf_kwargs):
super(CassandraTableScanRDD, self).__init__(ctx, keyspace, table, row_format, read_conf, **read_conf_kwargs)
self._key_by = ColumnSelector.none()
read_conf = as_java_object(ctx._gateway, self.read_conf.settings())
self.crdd = self._helper \
.cassandraTable(
ctx._jsc,
keyspace,
table,
read_conf,
)
def __init__(self, ctx, keyspace, table, row_format=None, read_conf=None, **read_conf_kwargs):
super(CassandraTableScanRDD, self).__init__(ctx, keyspace, table, row_format, read_conf,
**read_conf_kwargs)
self._key_by = ColumnSelector.none()
read_conf = as_java_object(ctx._gateway, self.read_conf.settings())
self.crdd = self._helper \
.cassandraTable(
ctx._jsc,
keyspace,
table,
read_conf,
)