Skip to content

Instantly share code, notes, and snippets.

@2shou
Last active January 17, 2023 06:34
Show Gist options
  • Star 3 You must be signed in to star a gist
  • Fork 2 You must be signed in to fork a gist
  • Save 2shou/c4329fa17f1b3f287341 to your computer and use it in GitHub Desktop.
Save 2shou/c4329fa17f1b3f287341 to your computer and use it in GitHub Desktop.
async client for hiveserver2
# coding: utf-8
from TCLIService import TCLIService
from TCLIService.ttypes import TOpenSessionReq, TGetTablesReq, TFetchResultsReq, \
TStatusCode, TGetResultSetMetadataReq, TGetColumnsReq, TType, TTypeId, \
TExecuteStatementReq, TGetOperationStatusReq, TFetchOrientation, TCloseOperationReq, \
TCloseSessionReq, TGetSchemasReq, TCancelOperationReq, THandleIdentifier, \
TOperationHandle, TOperationState
from thrift import Thrift
from thrift.transport import TSocket
from thrift.transport import TTransport
from thrift.protocol import TBinaryProtocol
class HiveServerTColumnValue:
def __init__(self, tcolumn_value):
self.column_value = tcolumn_value
@property
def val(self):
if self.column_value.boolVal is not None:
return self.column_value.boolVal.value
elif self.column_value.byteVal is not None:
return self.column_value.byteVal.value
elif self.column_value.i16Val is not None:
return self.column_value.i16Val.value
elif self.column_value.i32Val is not None:
return self.column_value.i32Val.value
elif self.column_value.i64Val is not None:
return self.column_value.i64Val.value
elif self.column_value.doubleVal is not None:
return self.column_value.doubleVal.value
elif self.column_value.stringVal is not None:
return self.column_value.stringVal.value
class HiveServer2Connection(object):
def __init__(self, host, port, user):
self.host = host
self.port = port
self.user = user
self.session_handle = None
self._client = None
def connect(self):
transport = TSocket.TSocket(self.host, self.port)
transport = TTransport.TBufferedTransport(transport)
protocol = TBinaryProtocol.TBinaryProtocol(transport)
client = TCLIService.Client(protocol)
transport.open()
self._client = client
def open_session(self, username):
req = TOpenSessionReq(username=username, configuration={})
res = self._client.OpenSession(req)
session_handle = res.sessionHandle
return session_handle
def call(self, fn, req, status=TStatusCode.SUCCESS_STATUS):
if self.session_handle is None:
self.session_handle = self.open_session(self.user)
if hasattr(req, 'sessionHandle') and req.sessionHandle is None:
req.sessionHandle = self.session_handle
res = fn(req)
return res
def execute_statement(self, statement):
req = TExecuteStatementReq(statement=statement, confOverlay={}, runAsync=True)
res = self.call(self._client.ExecuteStatement, req)
return res.operationHandle
def fetch_result(self, operation_handle, orientation=TFetchOrientation.FETCH_NEXT, max_rows=1000):
fetch_req = TFetchResultsReq(operationHandle=operation_handle, orientation=orientation, maxRows=max_rows)
res = self.call(self._client.FetchResults, fetch_req)
if operation_handle.hasResultSet:
meta_req = TGetResultSetMetadataReq(operationHandle=operation_handle)
schema = self.call(self._client.GetResultSetMetadata, meta_req)
else:
schema = None
return res, schema
def query_state(self, operation_handle):
req = TGetOperationStatusReq(operation_handle)
res = self._client.GetOperationStatus(req)
return res
def cancel_operation(self, operation_handle):
req = TCancelOperationReq(operation_handle)
return self._client.CancelOperation(req)
def close_session(self):
req = TCloseSessionReq(sessionHandle=self.session_handle)
return self._client.CloseSession(req)
def close(self):
self._client._iprot.trans.close()
def cursor(self, secret=None, guid=None):
if self._client is None:
self.connect()
return HiveServer2Cursor(self, secret, guid)
class HiveServer2Cursor(object):
def __init__(self, conn, secret=None, guid=None):
self.conn = conn
self.secret = secret
self.guid = guid
self.operation_handle = self._make_operation_handle(self.secret, self.guid)
@staticmethod
def _make_operation_handle(secret, guid):
if secret is None or guid is None:
return None
operation_id = THandleIdentifier(secret=secret, guid=guid)
return TOperationHandle(hasResultSet=True, modifiedRowCount=None, operationType=0, operationId=operation_id)
def execute(self, statement):
self.operation_handle = self.conn.execute_statement(statement)
if self.operation_handle is None:
return False
self.secret = self.operation_handle.operationId.secret
self.guid = self.operation_handle.operationId.guid
return True
def get_state(self):
res = self.conn.query_state(self.operation_handle)
if res.operationState == TOperationState.FINISHED_STATE:
return 1
elif res.operationState == TOperationState.CLOSED_STATE or res.operationState == TOperationState.ERROR_STATE \
or res.operationState == TOperationState.UKNOWN_STATE:
return 2
elif res.operationState == TOperationState.CANCELED_STATE:
return 3
return 0
def fetch_result(self, max_rows=1000):
def append_rows(_rlst, results):
for row in results.rows:
_rlst.append([HiveServerTColumnValue(column).val for column in row.colVals])
return _rlst
rlst = []
while True:
res, schema = self.conn.fetch_result(self.operation_handle, max_rows=max_rows)
# hasMoreRows属性失效了
if len(res.results.rows) == 0:
break
append_rows(rlst, res.results)
return rlst
def cancel_operation(self):
return self.conn.cancel_operation(self.operation_handle)
def close(self):
return self.conn.close_session()
def main():
from time import sleep
conn1 = HiveServer2Connection(host='yourhost', port=10000, user='root')
try:
cur = conn1.cursor()
cur.execute('select count(*) from test')
secret = cur.secret
guid = cur.guid
# 原有的session不能断,否则查询会中断
# cur.close()
finally:
conn1.close()
conn2 = HiveServer2Connection(host='yourhost', port=10000, user='root')
try:
cur = conn2.cursor(secret=secret, guid=guid)
while cur.get_state() == 0:
print 'running'
sleep(1)
if cur.get_state() == 1:
print 'success'
for r in cur.fetch_result():
print r[0]
elif cur.get_state() == 2:
print 'failed'
cur.close()
finally:
conn2.close()
if __name__ == '__main__':
main()
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment