Python脚本：不同Oracle库的表进行读写

准备工作：

①下载cx_Oracle、pandas、numpy包；

python 复制代码

pip install cx_Oracle
pip install pandas
pip install numpy

②已知两库的连接信息；

库1的主要为读数据表，并将表内容进行条件转换

python 复制代码

sql = ('SELECT NAME, SEX, AGE, PROVINCE, RANK, INCOME, BELONG_PROJECT, REPORT_TIME, '
       'APPROVER, APPROVAL_STATUS, SERVICE_RATING FROM REIMBURSEMENT_FIFTY_THOUSAND WHERE ROWNUM <= 10')

conn_govern_dev = cx_Oracle.connect(数据库1的连接信息)
cursor_govern_dev = conn_govern_dev.cursor()

# 读取govern_dev库的表
df = pd.read_sql(sql, conn_govern_dev)

# 性别为男切换为boy，女切换为girl
def myfunc_sex(x):
    if x == '男':
        return 'boy'
    elif x == '女':
        return 'girl'
    else:
        return 'maybe'

df['SEX'] = df['SEX'].map(myfunc_sex)

# 对工资列，低于1万的乘以3，加一个值，高于1万的除以2，加另外一个值
def myfunc_income(x, *args, **args_dict):
    if x < 10000:
        return x * 3 + args_dict['high']
    else:
        return x / 2 + args_dict['low']

df['INCOME'] = df['INCOME'].apply(myfunc_income, low=0.2, high=0.1)

cursor_govern_dev.close()
conn_govern_dev.close()

库2的操作就是将库1的表进行写入

python 复制代码

conn_test = cx_Oracle.connect(数据库2的连接信息)

df_array = np.array(df)  # dataframe转换array
df_list = df_array.tolist()  # array转换list
# print(df_list)
sql = ('insert into REIMBURSEMENT_FIFTY_THOUSAND(NAME, SEX, AGE, PROVINCE, RANK, '
       'INCOME, BELONG_PROJECT, REPORT_TIME, APPROVER, APPROVAL_STATUS, SERVICE_RATING) '
       'values(:1,:2,:3,:4,:5,:6,:7,:8,:9,:10,:11)')
cur = conn_test.cursor()
# 写入test库的表
cur.executemany(sql, df_list)
cur.close()
conn_test.commit()
conn_test.close()

完整代码

python 复制代码

# -*- coding: utf8 -*-
import cx_Oracle
import pandas as pd
import numpy as np

sql = ('SELECT NAME, SEX, AGE, PROVINCE, RANK, INCOME, BELONG_PROJECT, REPORT_TIME, '
       'APPROVER, APPROVAL_STATUS, SERVICE_RATING FROM REIMBURSEMENT_FIFTY_THOUSAND WHERE ROWNUM <= 10')

conn_govern_dev = cx_Oracle.connect(库1的链接信息)
cursor_govern_dev = conn_govern_dev.cursor()

# 读取govern_dev库的表
df = pd.read_sql(sql, conn_govern_dev)

# 性别为男切换为boy，女切换为girl
def myfunc_sex(x):
    if x == '男':
        return 'boy'
    elif x == '女':
        return 'girl'
    else:
        return 'maybe'

df['SEX'] = df['SEX'].map(myfunc_sex)

# 对工资列，低于1万的乘以3，加一个值，高于1万的除以2，加另外一个值
def myfunc_income(x, *args, **args_dict):
    if x < 10000:
        return x * 3 + args_dict['high']
    else:
        return x / 2 + args_dict['low']

df['INCOME'] = df['INCOME'].apply(myfunc_income, low=0.2, high=0.1)

cursor_govern_dev.close()
conn_govern_dev.close()

conn_test = cx_Oracle.connect(库2的链接信息)

df_array = np.array(df)  # dataframe转换array
df_list = df_array.tolist()  # array转换list
# print(df_list)
sql = ('insert into REIMBURSEMENT_FIFTY_THOUSAND(NAME, SEX, AGE, PROVINCE, RANK, '
       'INCOME, BELONG_PROJECT, REPORT_TIME, APPROVER, APPROVAL_STATUS, SERVICE_RATING) '
       'values(:1,:2,:3,:4,:5,:6,:7,:8,:9,:10,:11)')
cur = conn_test.cursor()
# 写入test库的表
cur.executemany(sql, df_list)
cur.close()
conn_test.commit()
conn_test.close()