本文主要是介绍python 将mysql转为csv、json导入到Doris数据库，希望对大家解决编程问题提供一定的参考价值，需要的开发者们随着小编来一起学习吧！
上一篇文章是将mysql导出成csv文件，适合csv不是很大的情况，以下对脚本进行了优化，采用分块读取csv，降低了内存的使用率，提高了传输速度。
from pydoris.doris_client import DorisClient
import requests
import pandas as pd
import numpy as npfe_host = ''
username = ''
passwd = ''
fe_http_port = ""
fe_query_port = ""doris_client = DorisClient(fe_host=fe_host,fe_query_port=fe_query_port,fe_http_port=fe_http_port,username=username,password=passwd,db='zst_cep_model')# 由于导入的csv文件过大，都是五六G以上，所以使用chunksize 分块获取数据进行操作。
for key, chunk in enumerate(pd.read_csv('xxxx.csv', chunksize=10000, dtype=str)):# 这一步也是很迷，需要读csv，然后转为字符串类型csv = chunk.to_csv(header=False, index=False, sep=',').encode('utf-8')# table_name 必须的是database.table的形式success = doris_client.write('xxx.xxx', csv)if success:print("数据写入成功！")else:print("数据写入失败。", key)breakdel csv后面需要进行增量更新，使用csv太麻烦，想着使用json传输可能会好点，并直接从mysql中获取数据直接传输到doris，以下是脚本。
import pandas as pd
from pydoris.doris_client import *from tools import *# 配置
config = {// Mysql的配置'mysql_config': {'host': '','port': ,'user': '','password': '','database': ''},// Doris的配置'doris_db_config': {'fe_host': '','username': '','passwd': '','fe_http_port': "",'fe_query_port': "",'db': ''},# 要传输的mysql的目标表'mysql_table': '',# 要接收的Doris的目标表'doris_table': '',
}# MysqlDataBaseClass 是自己编写的Mysql工具类，返回的是Mysql连接对象
yp_apidb = MysqlDataBaseClass(host=config['mysql_config']['host'], port=config['mysql_config']['port'], user=config['mysql_config']['user'],password=config['mysql_config']['password'], database=config['mysql_config']['database'])doris_client = DorisClient(fe_host=config['doris_db_config']['fe_host'],fe_query_port=config['doris_db_config']['fe_query_port'],fe_http_port=config['doris_db_config']['fe_http_port'],username=config['doris_db_config']['username'],password=config['doris_db_config']['passwd'],db=config['doris_db_config']['db'])def get_data_from_mysql(page=1, end_page=None, total_row_num=None, page_size=10000, limit_date='2024-08-12'):result = {'total_page': 0, 'total': 0, 'now_page': page, 'data': [], 'code': False, 'msg': ''}if total_row_num is None:select_res = yp_apidb.ExecuteSQL_Select(sql=f'''SELECT count(1) as total_num FROM `{config["mysql_table"]}` where collect_batch_date >= '{limit_date}';''')totalRowsNum = int(select_res[0]['total_num'])else:totalRowsNum = total_row_numif (totalRowsNum % page_size) == 0:totalPages = int(totalRowsNum / page_size)else:totalPages = int((totalRowsNum / page_size) + 1)result['total_page'] = totalPagesresult['total'] = totalRowsNumif end_page and page > end_page:result['msg'] = '已经达到设置的最后一页'return resultif page > totalPages:result['msg'] = '已经是最后一页'return resultstart_num = int((page - 1) * page_size)limit = f'{start_num}, {page_size}'sql = f'''SELECT * FROM `{config["mysql_table"]}` where collect_batch_date >= '{limit_date}' limit {limit};'''data_list = yp_apidb.ExecuteSQL_Select(sql)result['data'] = data_listresult['code'] = Trueresult['msg'] = '获取成功'return resultdef insert_to_doris(data_list):if len(data_list) > 0:df = pd.DataFrame(data_list)json_data = df.to_json(orient='records')options = WriteOptions()options.set_json_format()options.set_option("strip_outer_array", "true")success = doris_client.write(f"{config['doris_table']}", json_data, options=options)if success:return Trueelse:print("数据写入失败。")return Falseif __name__ == '__main__':page = 1total_row_num = Nonelimit_date = '2024-08-01'# 循环获取下一页，从而达到自动翻页的功能while True:res = get_data_from_mysql(page=page, total_row_num=total_row_num, limit_date=limit_date)print(res['msg'], res['total_page'], res['total'], res['now_page'])total_row_num = res['total']if res['code']:data_list = res['data']flage = insert_to_doris(data_list)if flage is False:breakpage += 1else:print(res['msg'], page)break以上脚本仅供学习参考，仅为实现临时功能而编写，还有优化的空间。
这篇关于python 将mysql转为csv、json导入到Doris数据库的文章就介绍到这儿，希望我们推荐的文章对编程师们有所帮助！
