python calls a simple instance of HBase

  • 2020-05-19 05:05:30
  • OfStack

The new engineer does not know HBase, java is not familiar, python is ok, I suggest that he can consider using thrift of HBase to complete the current work.

First, install thrift

Download thrift, in this case, I'm using thrift-0.7.0-dev.tar.gz


tar xzf thrift-0.7.0-dev.tar.gz
cd thrift-0.7.0-dev
sudo ./configure --with-cpp=no --with-ruby=no
sudo make
sudo make install

Then, go to the HBase source package and find it

src/main/resources/org/apache/hadoop/hbase/thrift/

perform

thrift --gen py Hbase.thrift
mv gen - py/hbase / / usr/lib/python2. 4 / site packages/(according to python versions may have different)

I wrote some call scripts here for your reference


from unittest import TestCase, main
from thrift import Thrift
from thrift.transport import TSocket
from thrift.transport import TTransport
from thrift.protocol import TBinaryProtocol

from hbase import Hbase
from hbase.ttypes import ColumnDescriptor, Mutation, BatchMutation
class HBaseTester:

  def __init__(self, netloc, port, table="staftesttable"):
    self.tableName = table

    self.transport = TTransport.TBufferedTransport(
      TSocket.TSocket(netloc, port))
    self.protocol = TBinaryProtocol.TBinaryProtocol(self.transport)
    self.client = Hbase.Client(self.protocol)
    self.transport.open()

    tables = self.client.getTableNames()
    if self.tableName not in tables:
      self.__createTable()

  def __del__(self):
    self.transport.close()

  def __createTable(self):
    name = ColumnDescriptor(name='name')
    foo = ColumnDescriptor(name='foo')

    self.client.createTable(self.tableName,
                [name,foo])
  
  def put(self,key,name,foo):
    name = Mutation(column="name:v", value=name)
    foo = Mutation(column="foo:v",value=foo)
    
    self.client.mutateRow(self.tablename,key,[name,foo])
  
  def scanner(self,column):
    scanner = client.scannerOpen(self.tablename,"",[column])
				r = client.scannerGet(scanner)
				result= []
    while r:
     print r[0]
     result.append(r[0])
     r = client.scannerGet(scanner)
    print "Scanner finished"
    return result
   
  
    
    
    
     
class TestHBaseTester(TestCase):
  
  def setUp(self):
    self.writer = HBaseTester("localhost", 9090)

  def tearDown(self):
    name = self.writer.tableName
    client = self.writer.client
    client.disableTable(name)
    client.deleteTable(name)


  def testCreate(self):
    tableName = self.writer.tableName
    client = self.writer.client
    self.assertTrue(self.writer.tableName in client.getTableNames())
    columns =['name:','foo:']
    for i in client.getColumnDescriptors(tableName):
      self.assertTrue(i in columns)
      
  def testPut(self):
    self.writer.put("r1","n1","f1")
    self.writer.put("r2","n2","f2")
    self.writer.put("r3","n3","")
    self.writer.scanner("name:") 
    
if __name__ == "__main__":
  main()

Related articles: