一、带有返回值的多线程

1.1 实现代码

# -*- coding:utf-8 -*-
"""
作者：wyt
日期：2022年04月21日
"""
import threading
import requests
import time
urls = [
    f'https://www.cnblogs.com/#p{page}' # 待爬地址
    for page in range(1, 10)  # 爬取1-10页
]
def craw(url):
    r = requests.get(url)
    num = len(r.text)  # 爬取博客园当页的文字数
    return num  # 返回当页文字数
 
def sigle():  # 单线程
    res = []
    for i in urls:
        res.append(craw(i))
    return res
class MyThread(threading.Thread):  # 重写threading.Thread类，加入获取返回值的函数
    def __init__(self, url):
        threading.Thread.__init__(self)
        self.url = url                # 初始化传入的url
    def run(self):                    # 新加入的函数，该函数目的：
        self.result = craw(self.url)  # ①。调craw(arg)函数，并将初试化的url以参数传递——实现爬虫功能
                                      # ②。并获取craw(arg)函数的返回值存入本类的定义的值result中
    def get_result(self):  #新加入函数，该函数目的：返回run()函数得到的result
        return self.result
def multi_thread():
    print("start")
    threads = []           # 定义一个线程组
    for url in urls:
        threads.append(    # 线程组中加入赋值后的MyThread类
            MyThread(url)  # 将每一个url传到重写的MyThread类中
        )
    for thread in threads: # 每个线程组start
        thread.start()
    for thread in threads: # 每个线程组join
        thread.join()
    list = []
    for thread in threads:
        list.append(thread.get_result())  # 每个线程返回结果(result)加入列表中
    print("end")
    return list  # 返回多线程返回的结果组成的列表
if __name__ == '__main__':
    start_time = time.time()
    result_multi = multi_thread()
    print(result_multi)  # 输出返回值-列表
    # result_sig = sigle()
    # print(result_sig)
    end_time = time.time()
    print('用时：', end_time - start_time)

1.2 结果

单线程：

多线程：

加速效果明显。

二、实现过程

2.1 一个普通的爬虫函数

import threading
import requests
import time
urls = [
    f'https://www.cnblogs.com/#p{page}' # 待爬地址
    for page in range(1, 10)  # 爬取1-10页
]
def craw(url):
    r = requests.get(url)
    num = len(r.text)  # 爬取博客园当页的文字数
    print(num)
def sigle():  # 单线程
    res = []
    for i in urls:
        res.append(craw(i))
    return res
def multi_thread():
    print("start")
    threads = []           # 定义一个线程组
    for url in urls:
        threads.append(
            threading.Thread(target=craw,args=(url,))  # 注意args=(url,)，元组
        )
    for thread in threads: # 每个线程组start
        thread.start()
    for thread in threads: # 每个线程组join
        thread.join()
    print("end")
if __name__ == '__main__':
    start_time = time.time()
    result_multi = multi_thread()
    # result_sig = sigle()
    # print(result_sig)
    end_time = time.time()
    print('用时：', end_time - start_time)

返回：

start
69915
69915
69915
69915
69915
69915
69915
69915
69915
end
用时： 0.316709041595459

2.2 一个简单的多线程传值实例

import time
from threading import Thread
def foo(number):
    time.sleep(1)
    return number
class MyThread(Thread):
    def __init__(self, number):
        Thread.__init__(self)
        self.number = number
    def run(self):
        self.result = foo(self.number)
    def get_result(self):
        return self.result
if __name__ == '__main__':
    thd1 = MyThread(3)
    thd2 = MyThread(5)
    thd1.start()
    thd2.start()
    thd1.join()
    thd2.join()
    print(thd1.get_result())
    print(thd2.get_result())

返回：

3
5

2.3 实现重点

多线程入口

threading.Thread(target=craw,args=(url,))  # 注意args=(url,)，元组

多线程传参

需要重写一下threading.Thread类，加一个接收返回值的函数。

三、代码实战

使用这种带返回值的多线程技术重写了一下之前发布过的一个爬取子域名的代码，原始代码在这里：https://www.jb51.net/article/254460.htm

import threading
import requests
from bs4 import BeautifulSoup
from static.plugs.headers import get_ua
#https://cn.bing.com/search?q=site%3Abaidu.com&go=Search&qs=ds&first=20&FORM=PERE
def search_1(url):
    Subdomain = []
    html = requests.get(url, stream=True, headers=get_ua())
    soup = BeautifulSoup(html.content, 'html.parser')
    job_bt = soup.findAll('h2')
    for i in job_bt:
        link = i.a.get('href')
        # print(link)
        if link not in Subdomain:
            Subdomain.append(link)
    return Subdomain
class MyThread(threading.Thread):
    def __init__(self, url):
        threading.Thread.__init__(self)
        self.url = url
    def run(self):
        self.result = search_1(self.url)
    def get_result(self):
        return self.result
def Bing_multi_thread(site):
    print("start")
    threads = []
    for i in range(1, 30):
        url = "https://cn.bing.com/search?q=site%3A" + site + "&go=Search&qs=ds&first=" + str(
            (int(i) - 1) * 10) + "&FORM=PERE"
        threads.append(
            MyThread(url)
        )
    for thread in threads:
        thread.start()
    for thread in threads:
        thread.join()
    res_list = []
    for thread in threads:
        res_list.extend(thread.get_result())
    res_list = list(set(res_list)) #列表去重
    number = 1
    for i in res_list:
        number += 1
    number_list = list(range(1, number + 1))
    dict_res = dict(zip(number_list, res_list))
    print("end")
    return dict_res
if __name__ == '__main__':
    print(Bing_multi_thread("qq.com"))

返回：

{
1:'https://transmart.qq.com/index',
2:'https://wpa.qq.com/msgrd?v=3&uin=448388692&site=qq&menu=yes',
3:'https://en.exmail.qq.com/',
4:'https://jiazhang.qq.com/wap/com/v1/dist/unbind_login_qq.shtml?source=h5_wx',
5:'http://imgcache.qq.com/',
6:'https://new.qq.com/rain/a/20220109A040B600',
7:'http://cp.music.qq.com/index.html',
8:'http://s.syzs.qq.com/',
9:'https://new.qq.com/rain/a/20220321A0CF1X00',
10:'https://join.qq.com/about.html',
11:'https://live.qq.com/10016675',
12:'http://uni.mp.qq.com/',
13:'https://new.qq.com/omn/TWF20220/TWF2022042400147500.html',
14:'https://wj.qq.com/?from=exur#!',
15:'https://wj.qq.com/answer_group.html',
16:'https://view.inews.qq.com/a/20220330A00HTS00',
17:'https://browser.qq.com/mac/en/index.html',
18:'https://windows.weixin.qq.com/?lang=en_US',
19:'https://cc.v.qq.com/upload',
20:'https://xiaowei.weixin.qq.com/skill',
21:'http://wpa.qq.com/msgrd?v=3&uin=286771835&site=qq&menu=yes',
22:'http://huifu.qq.com/',
23:'https://uni.weixiao.qq.com/',
24:'http://join.qq.com/',
25:'https://cqtx.qq.com/',
26:'http://id.qq.com/',
27:'http://m.qq.com/',
28:'https://jq.qq.com/?_wv=1027&k=pevCjRtJ',
29:'https://v.qq.com/x/page/z0678c3ys6i.html',
30:'https://live.qq.com/10018921',
31:'https://m.campus.qq.com/manage/manage.html',
32:'https://101.qq.com/',
33:'https://new.qq.com/rain/a/20211012A0A3L000',
34:'https://live.qq.com/10021593',
35:'https://pc.weixin.qq.com/?t=win_weixin&lang=en',
36:'https://sports.qq.com/lottery/09fucai/cqssc.htm'
}

非常非常非常能感受到速度快了超级多，用这种方式爆破子域名也比较爽。没有多线程，我的项目里可能缺少了好几个功能：因为之前写过的一些程序都因执行时间过长被我砍掉。这个功能还是很实用的。

四、学习