带有进度条显示的 多线程快读快写 sql,快速迁移表的 python 包
项目地址:https://github.com/tosmart01/fastsql
一、安装
- pip install fast_sql
二、依赖环境
- python3.6+
- mysql | oracle
二、示例
- 快读读取 sql 生成 DataFrame
import fast_sql
from sqlalchemy import create_engine
sql = "select * from test where id <1000000"
con = create_engine("oracle+cx_oracle://wuwukai:test@localhost:1521/helowin")
res = fast_sql.read_sql(sql,con,show_progress=True,thread_num=8)
res.head()
image:

-
快速读 sql 生成 csv
sql = "select * from test where id <1000000" path = '/home/test.csv' con = create_engine("oracle+cx_oracle://wuwukai:wuwukai@localhost:1521/helowin") res = fast_sql.to_csv(sql,con,path_or_buf=path, show_progress=True,thread_num=8,index=None) -
快速表迁移
#1.在线迁移 sql = "select * from student where id <1000000" con = create_engine("oracle+cx_oracle://wuwukai:test@localhost:1521/helowin") to_db = create_engine("mysql+pymysql://root:123456@localhost:3306/aps_2") res = fast_sql.to_sql(sql,from_db=con,to_db=to_db,if_exists='delete', mode='rw',to_table='stu',delete_cache=True,show_progress=True,)
# 2.离线迁移,先序列化到本地,在迁移 sql = "select * from student where id <1000000" con = create_engine("oracle+cx_oracle://wuwukai:test@localhost:1521/helowin") to_db = create_engine("mysql+pymysql://root:123456@localhost:3306/aps_2") # 写入本地,保存在当前工作目录 res = fast_sql.to_sql(sql,from_db=con, mode='r',to_table='stu',show_progress=True,) # 入库,file_path 为保存的目录 res = fast_sql.to_sql(sql,to_db=con,file_path='/home/test' mode='w',to_table='stu',show_progress=True,)
三、参数说明
其他参数兼容 pandas read_sql,to_csv
sql: sql 语句
con: 数据库连接 sqlalchemy 连接对象 | sqlalchemy 连接字符串 | pymysql 连接对象
thread_num: 开启的线程数量
encoding: 编码
show_progress: 是否显示进度条
from_db: 数据源库
to_db: 目标库
if_exists: 目标库相同数据是否删除 delete append orthe(不做处理)
to_table: 目标库表名,默认与原始 sql 表名相同
mode: r > 读取 sql 序列化到本地
w > 将序列化的文件入库
rw > 从源库读取写入到目标库
file_path: 读取数据序列化路径
delete_cache: 是否删除迁移过程中缓存的序列化文件
to_columns: 指定目标库的列名,默认与原列名相同