Python连接SQL与hive

--Zstarling

• SQL连接
• hive连接
• 补充写入方法
• • • • write与writelines对比
      • 表格Dataframe形式的写入
      • for、with 位置对比

SQL连接

import pymysql
def sql(path):
    conn = pymysql.Connect(host='36.104.34.123', user='用户名', passwd='密码', db='库名')
    # 获取游标
    cursor = conn.cursor()
    # 1、从数据库中查询
    sql = """
    SELECT 
        x.province,
        x.city,
        count(DISTINCT x.pid_new)
    FROM database.table AS x
    WHERE x.province IN ("江苏")
    group by x.city
    """
    try:
        result = cursor.execute(sql)  # cursor执行sql语句
        print(result)  # 会输出 *** 作的行数
        # 使用fetch方法进行遍历结果  总共有三条数据
        rr = cursor.fetchall()  # 会输出查询的所有结果
        # print(cursor.fetchone())  # 会输出一条查询的结果
        # print(cursor.fetchmany(2))  # 传入参数会输出对应条数的查询结果
        conn.commit()  # 提交 *** 作，只要涉及增删改就必须有commit，否则写不进数据库
    except Exception as e:
        print(e)
        conn.rollback()

    for row in rr:
        with open(path, 'a+') as f:
            print(row)
            f.writelines(str(row) + '\n')
    f.close()
    cursor.close()
    conn.close()

hive连接

from impala.dbapi import connect
class Hiveclient():
    # Hive登录
    def __init__(self, user, password):
        self.user = user
        self.conn = connect(host='master1', port=99999, auth_mechanism='GSSAPI', user=user, password=password,
                            kerberos_service_name='hive')

    # 数据查询与返回
    def query(self):
        print("开始取数啦！！！")
        with self.conn.cursor() as curs:
            qu_sql = """SELECT
                        province,
                        city
                        FROM database.table
                        limit 10
                    """
            print(qu_sql)
            try:
                curs.execute(qu_sql)
                rr = curs.fetchall()
                print(type(rr), rr)

                # 对结果进行处理
                for i in rr:
                    with open(path, 'a+', encoding='GB2312') as f:
                        print(i, type(i))
                        f.writelines(str(i) + '\n')
                print('=============全部完成===============')
            except Exception as e:
                print(e)
            return

    # 关闭客户端
    def close(self):
        self.conn.close()

补充写入方法 write与writelines对比

write():只能写入字符串。会把
writelines():可以写入字符串，列表，元组等。

 with open(path, 'a+') as f:
        print(rr)
        f.write(str(rr)+'\n')
        f.writelines(str(rr)+'\n')

都是字符串形式时两者无区别。结果如下:

注意：如果此时需要换行的话需要转化成字符串的形式与\n连接。否则则会在尾部直接追加。示例如下：

表格Dataframe形式的写入

另外，如果是Dataframe或者series的数据需要写入的话，需要采用to_csv或者to_excel的方式。

import os
import pandas as pd

path = os.getcwd() + '/test.csv'
data = pd.read_excel(r"C:\Users\EDZ\Desktop\test2.xlsx")
print(type(data))
data.to_csv(path, index=False, encoding='GB2312')

for、with 位置对比

    with open(path, 'a+') as f:
        for row in rr:
            print(row)
            f.writelines(str(row)+'\n')

    for row in rr:
        with open(path, 'a+') as f:
            print(row)
            f.writelines(str(row)+'\n')

  for row in rr:
        with open(path, 'a+') as f:
            print(row)
            f.write(str(row)+'\n')

• 用了for循坏之后，write与writelines结果相同。
• with 和for 调换位置后并不影响循坏，即输出结果相同。