def __init__(self, keyspace, table, ctx, row_format=None, read_conf=None):
if not keyspace:
raise ValueError("keyspace not set")
if not table:
raise ValueError("table not set")
if row_format is None:
row_format = RowFormat.ROW
elif row_format < 0 or row_format >= len(RowFormat.values):
raise ValueError("invalid row_format %s" % row_format)
self.keyspace = keyspace
self.table = table
self.read_conf = read_conf
self.row_format = row_format
jread_conf = as_java_object(ctx._gateway,
read_conf.__dict__) if read_conf else None
self._helper = ctx._jvm.java.lang.Thread.currentThread().getContextClassLoader() \
.loadClass("pyspark_cassandra.PythonHelper").newInstance()
self._cjrdd = self._helper \
.cassandraTable(
keyspace,
table,
ctx._jsc,
jread_conf,
)
jrdd = self._helper.parseRows(self._cjrdd, row_format)
super(CassandraRDD, self).__init__(jrdd, ctx)
def __init__(self, keyspace, table, ctx, row_format=None, read_conf=None):
if not keyspace:
raise ValueError("keyspace not set")
if not table:
raise ValueError("table not set")
if not row_format:
row_format = RowFormat.ROW
elif row_format < 0 or row_format >= len(RowFormat.values):
raise ValueError("invalid row_format %s" % row_format)
self.keyspace = keyspace
self.table = table
self.read_conf = read_conf
self.row_format = row_format
jread_conf = as_java_object(ctx._gateway, read_conf.__dict__) if read_conf else None
self._helper = ctx._jvm.java.lang.Thread.currentThread().getContextClassLoader() \
.loadClass("pyspark_cassandra.PythonHelper").newInstance()
self._cjrdd = self._helper \
.cassandraTable(
keyspace,
table,
ctx._jsc,
jread_conf,
)
jrdd = self._helper.parseRows(self._cjrdd, row_format)
super(CassandraRDD, self).__init__(jrdd, ctx)
def saveToCassandra(rdd,
keyspace=None,
table=None,
columns=None,
write_conf=None,
row_format=None):
'''
Saves an RDD to Cassandra. The RDD is expected to contain dicts with keys mapping to CQL columns.
Arguments:
@param rdd(RDD):
The RDD to save. Equals to self when invoking saveToCassandra on a monkey patched RDD.
@param keyspace(string):in
The keyspace to save the RDD in. If not given and the rdd is a CassandraRDD the same keyspace is used.
@param table(string):
The CQL table to save the RDD in. If not given and the rdd is a CassandraRDD the same table is used.
Keyword arguments:
@param columns(iterable):
The columns to save, i.e. which keys to take from the dicts in the RDD.
If None given all columns are be stored.
@param write_conf(WriteConf):
A WriteConf object to use when saving to Cassandra
@param row_format(RowFormat):
Make explicit how to map the RDD elements into Cassandra rows.
If None given the mapping is auto-detected as far as possible.
'''
keyspace = keyspace or rdd.keyspace
if not keyspace:
raise ValueError("keyspace not set")
table = table or rdd.table
if not table:
raise ValueError("table not set")
# create write config as map and convert the columns to a string array
write_conf = as_java_object(rdd.ctx._gateway,
write_conf.__dict__) if write_conf else None
columns = as_java_array(rdd.ctx._gateway, "String",
columns) if columns else None
# create a helper object
helper = rdd.ctx._jvm.java.lang.Thread.currentThread().getContextClassLoader() \
.loadClass("pyspark_cassandra.PythonHelper").newInstance()
# delegate to helper
helper \
.saveToCassandra(
rdd._jrdd,
keyspace,
table,
columns,
write_conf,
row_format
)
def saveToCassandra(rdd, keyspace=None, table=None, columns=None, row_format=None, keyed=None, write_conf=None,
**write_conf_kwargs):
'''
Saves an RDD to Cassandra. The RDD is expected to contain dicts with keys mapping to CQL columns.
Arguments:
@param rdd(RDD):
The RDD to save. Equals to self when invoking saveToCassandra on a monkey patched RDD.
@param keyspace(string):in
The keyspace to save the RDD in. If not given and the rdd is a CassandraRDD the same keyspace is used.
@param table(string):
The CQL table to save the RDD in. If not given and the rdd is a CassandraRDD the same table is used.
Keyword arguments:
@param columns(iterable):
The columns to save, i.e. which keys to take from the dicts in the RDD.
If None given all columns are be stored.
@param row_format(RowFormat):
Make explicit how to map the RDD elements into Cassandra rows.
If None given the mapping is auto-detected as far as possible.
@param keyed(bool):
Make explicit that the RDD consists of key, value tuples (and not arrays of length two).
@param write_conf(WriteConf):
A WriteConf object to use when saving to Cassandra
@param **write_conf_kwargs:
WriteConf parameters to use when saving to Cassandra
'''
keyspace = keyspace or getattr(rdd, 'keyspace', None)
if not keyspace:
raise ValueError("keyspace not set")
table = table or getattr(rdd, 'table', None)
if not table:
raise ValueError("table not set")
# create write config as map
write_conf = WriteConf.build(write_conf, **write_conf_kwargs)
write_conf = as_java_object(rdd.ctx._gateway, write_conf.settings())
# convert the columns to a string array
columns = as_java_array(rdd.ctx._gateway, "String", columns) if columns else None
helper(rdd.ctx) \
.saveToCassandra(
rdd._jrdd,
keyspace,
table,
columns,
row_format,
keyed,
write_conf,
)
def saveToCassandra(rdd, keyspace=None, table=None, columns=None, write_conf=None, row_format=None):
'''
Saves an RDD to Cassandra. The RDD is expected to contain dicts with keys mapping to CQL columns.
Arguments:
@param rdd(RDD):
The RDD to save. Equals to self when invoking saveToCassandra on a monkey patched RDD.
@param keyspace(string):in
The keyspace to save the RDD in. If not given and the rdd is a CassandraRDD the same keyspace is used.
@param table(string):
The CQL table to save the RDD in. If not given and the rdd is a CassandraRDD the same table is used.
Keyword arguments:
@param columns(iterable):
The columns to save, i.e. which keys to take from the dicts in the RDD.
If None given all columns are be stored.
@param write_conf(WriteConf):
A WriteConf object to use when saving to Cassandra
@param row_format(RowFormat):
Make explicit how to map the RDD elements into Cassandra rows.
If None given the mapping is auto-detected as far as possible.
'''
keyspace = keyspace or rdd.keyspace
if not keyspace:
raise ValueError("keyspace not set")
table = table or rdd.table
if not table:
raise ValueError("table not set")
# create write config as map and convert the columns to a string array
write_conf = as_java_object(rdd.ctx._gateway, write_conf.__dict__) if write_conf else None
columns = as_java_array(rdd.ctx._gateway, "String", columns) if columns else None
# create a helper object
helper = rdd.ctx._jvm.java.lang.Thread.currentThread().getContextClassLoader() \
.loadClass("pyspark_cassandra.PythonHelper").newInstance()
# delegate to helper
helper \
.saveToCassandra(
rdd._jrdd,
keyspace,
table,
columns,
write_conf,
row_format
)
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,
)
