利用Python将Oracle表中的数据抽到PostgreSQL

Posted 2022-04-09 robinson1988

import cx_Oracle
import psycopg2
import os
import time
from io import StringIO
import pandas as pd

#说明：本脚本用于将Oracle数据迁移到PG
#注意：源表与目标表字段数量必须一致

os.environ['NLS_LANG'] = 'SIMPLIFIED CHINESE_CHINA.UTF8'       #设置字符集为UTF8，防止中文乱码
source_db = cx_Oracle.connect('scott/tiger@192.168.3.10/orcl') #源库
target_db = psycopg2.connect(database="orcl", user="scott", password="tiger", host="192.168.56.10", port="5432") #目标库
cur_select = source_db.cursor()  #源库查询对象
cur_insert = target_db.cursor()  #目标库插入对象
cur_select.arraysize = 5000
cur_insert.arraysize = 5000
source_table = input("请输入源表名称:")   #从键盘获取源表名称
target_table = input("请输入目标表名称:") #从键盘获取目标表名称
select_sql = 'select * from ' + source_table  #源查询SQL，如果有where过滤条件，在这里拼接
cur_select.execute(select_sql) #执行
print('开始执行:',time.strftime("%Y-%m-%d %H:%M:%S", time.localtime()))
while True:
 rows = cur_select.fetchmany(5000)     #每次获取5000行，由cur_select.arraysize值决定
 data = pd.DataFrame(rows)
 output = StringIO()
 data.to_csv(output, sep='\\t', index=False, header=False)
 output1 = output.getvalue()
 cur_insert.copy_from(StringIO(output1),target_table,sep='\\t',null='')
 target_db.commit()
 if not rows:
     break                    #中断循环
cur_select.close()
cur_insert.close()
source_db.close()
target_db.close()
print('执行成功:',time.strftime("%Y-%m-%d %H:%M:%S", time.localtime()))