python 操作原生sql

介绍

python多线程并发操作数据库，会存在链接数据库超时、数据库连接丢失、数据库操作超时等问题。

解决方法：使用数据库连接池，并且每次操作都从数据库连接池获取数据库操作句柄，操作完关闭连接返回数据库连接池。

*连接数据库需要设置charset = ‘utf8’, use_unicode = True，不然会报中文乱码问题

*网上说解决python多线程并发操作数据库问题，连接时使用self.conn.ping(True)（检查并保持长连接），但是我这边亲测无法解决，建议还是使用数据库连接池

python多线程代码

import threading

class MyThread(threading.Thread):

def __init__(self, name, count, exec_object):
	threading.Thread.__init__(self)
	self.name = name
	self.count = count
	self.exec_object = exec_object

def run(self):
	while self.count >= 0:
	count = count - 1
	self.exec_object.execFunc(count)

thread1 = MyThread('MyThread1', 3, ExecObject())
thread2 = MyThread('MyThread2', 5, ExecObject())
thread1.start()
thread2.start()
thread1.join() # join方法 执行完thread1的方法才继续主线程
thread2.join() # join方法 执行完thread2的方法才继续主线程
# 执行顺序 并发执行thread1 thread2，thread1和thread2执行完成才继续执行主线程

# ExecObject类是自定义数据库操作的业务逻辑类
# 

########join方法详解########
thread1 = MyThread('MyThread1', 3, ExecObject())
thread2 = MyThread('MyThread2', 5, ExecObject())
thread1.start()
thread1.join() # join方法 执行完thread1的方法才继续主线程
thread2.start()
thread2.join() # join方法 执行完thread2的方法才继续主线程
# 执行顺序 先执行thread1，执行完thread1再执行thread2，执行完thread2才继续执行主线程

mysql数据库连接池代码：

import MySQLdb
from DBUtils.PooledDB import PooledDB

class MySQL:

host = 'localhost'
user = 'root'
port = 3306
pasword = ''
db = 'testDB'
charset = 'utf8'

pool = None
limit_count = 3 # 最低预启动数据库连接数量

def __init__(self):
	self.pool = PooledDB(MySQLdb, self.limit_count, host = self.host, user = self.user, passwd = self.pasword, db = self.db,
	port = self.port, charset = self.charset, use_unicode = True)

def select(self, sql):
	conn = self.pool.connection()
	cursor = conn.cursor()
	cursor.execute(sql)
	result = cursor.fetchall()
	cursor.close()
	conn.close()
	return result

def insert(self, table, sql):
	conn = self.pool.connection()
	cursor = conn.cursor()
	try:
		cursor.execute(sql)
		conn.commit()
		return {'result':True, 'id':int(cursor.lastrowid)}
	except Exception as err:
		conn.rollback()
		return {'result':False, 'err':err}
	finally:
		cursor.close()
		conn.close()