首页 > > 网络编程 > 其它 >

python 斗图图片爬虫

2018-06-18 03:29:25来源：未知阅读 ()

捣鼓了三小时，有一些小Bug，望大佬指导

废话不说，直接上代码：

#!/usr/bin/python3
# -*- coding:UTF-8 -*-
import os,re,requests
from urllib import request,parse

class Doutu_api(object):
    def __init__(self):
        self.api_html = r'http://www.doutula.com/search?keyword=%s'
        self.headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 '
                                      '(KHTML, like Gecko) Chrome/66.0.3359.181 Safari/537.36'}
        self.path = os.path.dirname(os.path.realpath(__file__))+'\\temp'

    def make_path(self,path=''):#返回假为已创建，否则创建新文件夹
        self.path = self.path+'\\'+path
        if os.path.exists(self.path):  # 判断文件夹是否存在
            return False
        else:
            os.mkdir(self.path)  # 创建文件夹
            return True

    def get_img_html(self,html):
        self.make_path(path=html)
        html = self.api_html%parse.quote(html)
        pattern = re.compile(u'<a.*?class="col-xs-6 col-md-2".*?href="(.*?)".*?style="padding:5px;">.*?</a>',re.S)
        pattern_img = re.compile(u'<td>.*?<img.*?src="(.*?)".*?alt="(.*?)".*?onerror=".*?">.*?</td>',re.S)
        try:
            req = request.Request(html, headers=self.headers)
            imgs = request.urlopen(req)
            imgs = imgs.read().decode('utf-8')
            imgs = re.findall(pattern, imgs)
            for img in imgs:
                req = request.Request(img, headers=self.headers)
                imgurl = request.urlopen(req).read().decode('utf-8')
                imgurl =re.findall(pattern_img, imgurl)
                with open(self.path+'\\{}.png'.format(imgurl[0][1].replace('/','-')), 'wb') as file:
                    response = requests.get(imgurl[0][0]).content  # 下载图片
                    file.write(response)  # 读取图片
            print('已完成下载,图片地址:',self.path)
        except Exception as e:
            print(e)
        return None

doutu = Doutu_api()
doutu.get_img_html(input('斗图内容关键字：'))