day6-多线程和多进程

day6-多线程和多进程

1.多线程

• 程序默认都是单线程(这个默认线程又叫主线程，其他的线程都叫子线程)
• Thread类的对象就是线程对象，程序需要多少个子线程，就要创建多少个Thread的对象

import time
from datetime import datetime
from threading import Thread, current_thread

def download(name):
    print(f'{name}开始下载:{datatime.now()}')
    print('当前线程:', current_thread())
    time.sleep(2)
    print({f'{name}下载结束:{datetime.now()}'})
    
# 1. 单线程下载三部电影
# download('大话西游')
# download('忍者神龟')
# download('狐妖小红娘')
# 1) 创建线程对象
t1 = Thread(target=download, args=('大话西游'))
t2 = Thread(target=download, args=('忍者神龟'))
t3 = Thread(target=download, args=('狐妖小红娘'))

# 2) 启动线程
# 线程对象.start()
t1.start()
t2.start()
t3.start()

(2) 多线程下载三部电影

• 创建子线程: Thread(target=函数, args=元组)
• 函数 - 需要在子线程中调用的函数
• 元组 - 调用 target 对应的函数的时候传递实参列表

2.图片下载

import requests
from lxml import etree
from threading import Thread

ts1 = []
ts2 = []


def get_one_page(page: int):
    print(f'获取第{page}页数据')
    if page == 1:
        url = 'https://pic.netbian.com/index.html'
    else:
        url = f'https://pic.netbian.com/index_{page}.html'
    response = requests.get(url)
    response.encoding = 'gbk'
    html = etree.HTML(response.text)
    result = html.xpath('//div[@class="slist"]/ul/li/a/img/@src|//div[@class="slist"]/ul/li/a/span/img/@src')
    # return ['https://pic.netbian.com'+x for x in result]

    #  一页图片用一个线程来下载
    t = Thread(target=download_page_image, args=(['https://pic.netbian.com'+x for x in result],))
    t.start()
    ts1.append(t)

    # 一张图片一个线程
    # for url in ['https://pic.netbian.com'+x for x in result]:
    #     t = Thread(target=download_image, args=(url,))
    #     t.start()


def get_all_data():
    for page in range(1, 11):
        t = Thread(target=get_one_page, args=(page, ))
        t.start()
        ts2.append(t)

        for t in ts2:
            t.join()

        for t in ts1:
            t.join()
        print('图片全部下载完成！')


def download_page_image(urls: list):
    print(f'下载一页数据:{len(urls)}')
    for url in urls:
        download_image(url)


def download_image(url: str):
    response = requests.get(url)
    f = open(f'files/{url.split("/")[-1]}', 'wb')
    f.write(response.content)
    # print('下载完成！')


get_all_data()

3.线程类的子类对象

from threading import Thread, current_thread
import time
from datetime import datetime


class DownloadThread(Thread):
    def __init__(self, name):
        super().__init__()
        self.name = name

    # 在子线程中添加任务
    def run(self) -> None:
        print(f'{self.name}开始下载:{datetime.now()}')
        print('当前线程:', current_thread())
        time.sleep(2)
        print(f'{self.name}下载结束:{datetime.now()}')


t1 = DownloadThread('花园宝宝')
t2 = DownloadThread('天线宝宝')
t3 = DownloadThread('灰蛇传')

# 通过start调用run，run方法会在相应的子线程中执行
t1.start()
t2.start()
t3.start()

# 直接调用run，run方法会在主线程中执行
# t1.run()
# t2.run()
# t3.run()

4.jion 操作

(1) jion 的用法

• 线程对象.jion() - 等待当前线程的任务结束后才执行后面的代码

import time
from datetime import datetime
from threading import Thread, current_thread
from random import randint

def download(name):
    print(f'{name}开始下载:{datetime.now()}')
    print('当前线程:', current_thread())
    time.sleep(randint(3, 7))
    print({f'{name}下载结束:{datetime.now()}'})
    

t1 = Thread(target=download, args=('大话西游'))
t2 = Thread(target=download, args=('忍者神龟'))
t3 = Thread(target=download, args=('狐妖小红娘'))

t1.start()
t2.start()
t3.start()

t1.join()
t2.join()
t3.join()

5.多进程

from multiprocessing import Process, current_process
from datetime import datetime
import time
from dandom import randint

def download(name):
    print(f'{name}开始下载:{datetime.now()}')
    print('当前线程:', current_process())
    time.sleep(randint(3, 7))
    print({f'{name}下载结束:{datetime.now()}'})
    
if __name__ == '__main__':
    # 1. 创建进程对象
    p1 = Process(target=download, args=('肖申克的救赎',))
    p2 = Process(target=download, args=('触不可及',))
    p3 = Process(target=download, args=('雪国列车',))
    
    # 2. 启动进程
    p1.start()
    p2.start()
    p3.start()
    
    p1.join()
    p2.join()
    p3.join()
    print('电影下载完成')

6.线程间通信(了解)

(1) 在同一个进程中的多个线程间的数据可以直接相互使用

a = 100
list1 = []

def func1():
    time.sleep(1)
    global a
    print(a)
    a = 300
    list1.append(200)
    
t = Thread(target=func1)
t.start()

t.join()
print(a, list1)

(2) 线程间通信的数据安全

• 加锁: 一个数据一把锁(数据和锁需要对应)

• Lock对象: 锁对象.acquire() - 加锁； 锁对象.release() - 释放锁

• Rlock 对象:

  with 锁对象

  ​ 操作数据的代码

from threading import Lock, Rlock

balance = 10000
# 1) 创建锁对象，保证一个数据一个锁
lock = Lock()
lock2 = Rlock()

def save_money(num: int):
    """存钱"""
    print('开始存钱......')
    
    # 2) 在使用需要锁的数据之前加锁
    # lock.acquire()
    with lock2:
        global balance
        b1 = balance
        time.sleep(1)
        balance = b1 + num
        
    # 3) 在数据用完以后释放锁
    # lock.release()


def draw_money(num: int):
    print('开始取钱......')
    # lock.acquire()
    with lock2:
        global balance
        b1 = balance
        if b1 >= num:
            time.sleep(1)
            balance = b1 - num
        else:
            print('余额不足!')
    # lock.release()


t1 = Thread(target=save_money, args=(2000,))
t2 = Thread(target=draw_money, args=(3000,))

t2.start()
t1.start()


t1.join()
t2.join()
print('余额:', balance)

7.多线程数据返回问题

• 结论: 在子线程中调用的函数如果有返回值，这个返回值是无法在任何地方获取的

from threading import Thread, current_thread
from multiprocessing import Process

data = []

def download(name):
    """收集数据"""
    print('---------')
    # return f'{name}数据'
    data.append(f'{name}数据')
    
def use_data():
    for x in data:
        print(f'使用{current_thread()}:', x)
        
t1 = Thread(target=download, args=('肖申克的救赎'))
t2 = Thread(target=download, args=('触不可及'))
t1.start()
t2.start()

t1.join()
t2.join()
# print(data)

t3 = Thread(target=use_data)
t3.start()