python并发模块之concurrent.futures(二)

python并发模块之concurrent.futures(二)

上次我们简单的了解下，模块的一些基本方法和用法，这里我们进一步对concurrent.futures做一个了解和拓展.
上次的内容点这。
python并发模块之concurrent.futures(二)
以下载图片为例子，下面的程序是顺序下载http://www.58pic.com/newpic/28660111.html网站的24个表情 。

from requests_html import HTMLSession
import os
import time
BASE_PATH="downloads"
class Get_Image():
    def __init__(self):
        self.timeout=20
        self.session=HTMLSession()
    def getiamge(self,url):
        req=self.session.get(url,timeout=self.timeout)
        if req.status_code==200:
            imgurllist=req.html.xpath("//ul[@class='emoticon-model']/li/img/@data-big")
            for index,url in enumerate(imgurllist):
                print(f"开始下载第{index+1}张图片")
                self.save_image(url,index+1)
        else:
            print("下载失败")
    def save_image(self,imgurl,index):
        print(f"当前下载链接：{imgurl}")
        buff=self.session.get(imgurl,timeout=self.timeout).content
        file_path=os.path.join(os.path.dirname(os.path.abspath(__file__)),BASE_PATH)
        if not os.path.exists(file_path):
            os.makedirs(file_path)
        with open(os.path.join(file_path,f"{index}.png"),"wb") as fs:
            fs.write(buff)
if __name__ == '__main__':
    start_url="http://www.58pic.com/newpic/28660111.html"
    start=time.time()
    Get_Image().getiamge(start_url)
    end=time.time()
    print(f"顺序下载24张图片用时:{end-start}")
#运行了两次结果分别为
#顺序下载24张图片用时:14.926000356674194
#顺序下载24张图片用时:14.07800030708313

使用concurrent.futures修改成并发之后

from requests_html import HTMLSession
import os
import time
from concurrent.futures import ThreadPoolExecutor
BASE_PATH="downloads"
MAX_WORKERS = 10 #最多使用10个线程
class Get_Image():
    def __init__(self):
        self.timeout=20
        self.session=HTMLSession()
    def getiamge(self,url):
        req=self.session.get(url,timeout=self.timeout)
        if req.status_code==200:
            imgurllist=req.html.xpath("//ul[@class='emoticon-model']/li/img/@data-big")
            works=min(len(imgurllist),MAX_WORKERS)
            with ThreadPoolExecutor(works) as excutor:
                res=excutor.map(self.save_image,imgurllist,range(1,25))
            return len(list(res))
        else:
            print("下载失败")
    def save_image(self,imgurl,index):
        print(f"当前下载链接：{imgurl}")
        buff=self.session.get(imgurl,timeout=self.timeout).content
        file_path=os.path.join(os.path.dirname(os.path.abspath(__file__)),BASE_PATH)
        if not os.path.exists(file_path):
            os.makedirs(file_path)
        with open(os.path.join(file_path,f"{index}.png"),"wb") as fs:
            fs.write(buff)
if __name__ == '__main__':
    start_url="http://www.58pic.com/newpic/28660111.html"
    start=time.time()
    Get_Image().getiamge(start_url)
    end=time.time()
    print(f"并发下载24张图片用时:{end-start}")
#运行了两次结果分别为
#并发下载24张图片用时:7.737000226974487
#并发下载24张图片用时:7.083999872207642

通过观察发现速度并发之后效率大大提高了。