python MySQL db 和 Cursor 无法执行 INSERT INTO

Question

In Python version 2.7.6

Pandas version 0.18.1

MySQL 5.7

import MySQLdb as dbapi
import sys
import csv
import os
import sys, getopt
import pandas as pd

df = pd.read_csv('test.csv')
rows = df.apply(tuple, 1).unique().tolist()

db=dbapi.connect(host=dbServer,user=dbUser,passwd=dbPass)

cur=db.cursor()

for (CLIENT_ID,PROPERTY_ID,YEAR) in rows:

                INSERT_QUERY=("INSERT INTO {DATABASE}.TEST SELECT * FROM {DATABASE}_{CLIENT_ID}.TEST WHERE PROPERTY_ID = {PROPERTY_ID} AND YEAR = {YEAR};".format(
                CLIENT_ID=CLIENT_ID,
                PROPERTY_ID=PROPERTY_ID,
                YEAR=YEAR,
                DATABASE=DATABASE
             ))

                print INSERT_QUERY
                cur.execute(INSERT_QUERY)
                db.query(INSERT_QUERY)

This will print out the query I am looking for, however, without successfully returning the results of INSERT INTO when I checked the results in MySQL

INSERT INTO test.TEST SELECT * FROM test_1.TEST WHERE PROPERTY_ID = 1 AND YEAR = 2015;

However, if I just copy and paste this MySQL query into MySQL GUI, it will execute without any problem. Could any guru enlighten?

I also tried the following

cur.execute(INSERT_QUERY, multi=True)

Returns an error

TypeError: execute() got an unexpected keyword argument 'multi'

Answer 1

The answer here is we need to use "from mysql.connector" and a db.commit(). Here is a good example

http://www.mysqltutorial.org/python-mysql-insert/

import MySQLdb as dbapi
import mysql.connector
import sys
import csv
import os
import sys, getopt
import pandas as pd

df = pd.read_csv('test.csv')
rows = df.apply(tuple, 1).unique().tolist()

db=dbapi.connect(host=dbServer,user=dbUser,passwd=dbPass)

cur=db.cursor()

conn = mysql.connector.connect(host=dbServer,user=dbUser,port=dbPort,password=dbPass)

cursor=conn.cursor()

for (CLIENT_ID,PROPERTY_ID,YEAR) in rows:

                INSERT_QUERY=("INSERT INTO {DATABASE}.TEST SELECT * FROM {DATABASE}_{CLIENT_ID}.TEST WHERE PROPERTY_ID = {PROPERTY_ID} AND YEAR = {YEAR};".format(
                CLIENT_ID=CLIENT_ID,
                PROPERTY_ID=PROPERTY_ID,
                YEAR=YEAR,
                DATABASE=DATABASE
             ))

                print INSERT_QUERY
                cursor.execute(INSERT_QUERY)
                conn.commit()

Only by having the commit, the database/ table changes will be accepted

Answer 2

I was using mysql-connector pool, trying to insert a new row into a table, and got the same problem. The version info: mysql-8, python3.7. The solution is to add connection.commit at last even you didn't start transaction.