爬虫+百度图片

#coding:utf-8
import requests
import re,time
import os
from multiprocessing import Pool

#获取网站内容
def get_page_text(url,prama):
    response = requests.get(url,params=prama)
    #utf-8会报错，超出utf-8的编码范围
    response = response.content.decode('ISO-8859-1')
    return response
#利用正则，获取图片地址
def get_img_url(html):
    r = re.compile('"thumbURL":"(.*?\.jpg)"')
    urlList = re.findall(r,html)
    return urlList

def run(keyword,num,path):
    x = 1
    url = 'https://image.baidu.com/search/acjson'
    #第一重循环，动态刷新图片数量，利用pn每次加30，
    for i in range(30,30*num+30,30):
        prama = {"ipn": "rj", "tn": "resultjson_com", "word": keyword, "pn": str(i)}
        html = get_page_text(url,prama)
        img_url = get_img_url(html)
        #第二重循环遍历每次刷新的30张图片
        for j in img_url:
            try:
                img = requests.get(j,timeout=5).content
                with open(path+'/'+str(x)+'.jpg','wb')as f:
                    f.write(img)
                    print('第%d张下载成功...'%x)
                x += 1
            except:
                print('下载失败...')
#判断目录是否存在，不存在则创建目录
def mkdir(keyword):
    path = './图片/'+keyword
    dir = os.path.exists(path)
    if not dir:
        os.makedirs(path)
        return path
    else:
        print('目录已存在...')
        return path
def main():
    keyword = input('输入下载图片的关键字：')
    number = int(input('输入下载的数量(单位：页,每页30张)：'))
    path = mkdir(keyword)
    run(keyword,number,path)
    
if __name__ == '__main__':
    main()

原理就是想方设法获取图片的地址，然后以二进制的格式写入文件。