使用 Scrapy 的 ImagesPipeline 下载图片

下载 百度贴吧-动漫壁纸吧 所有图片

定义item

spider

spider 只需要得到图片的url，必须以列表的形式给管道处理

class picturespiderspider(scrapy.spider):
    name = 'picture_spider'
    allowed_domains = ['tieba.baidu.com']
    start_urls = ['https://tieba.baidu.com/f?kw=%e5%8a%a8%e6%bc%ab%e5%a3%81%e7%ba%b8']

    def parse(self, response):

        #   贴吧中一页帖子的id和标题
        theme_urls = re.findall(r'<a rel="noreferrer" href="/p/(\d+)" title="(.*?)" target="_blank" class="j_th_tit ">',
                                response.text, re.s)

        for theme in theme_urls:
            #   帖子的url
            theme_url = 'https://tieba.baidu.com/p/' + theme[0]
            #   进入各个帖子
            yield scrapy.request(url=theme_url, callback=self.parse_theme)

        #   贴吧下一页的url
        next_url = re.findall(
            r'<a href="//tieba.baidu.com/f\?kw=%e5%8a%a8%e6%bc%ab%e5%a3%81%e7%ba%b8&ie=utf-8&pn=(\d+)" class="next pagination-item " >下一页&gt;</a>',
            response.text, re.s)
        if next_url:
            next_url = self.start_urls[0] + '&pn=' + next_url[0]
            yield scrapy.request(url=next_url)

    # 下载每个帖子里的所有图片
    def parse_theme(self, response):
        item = postbaritem()
        #   每个贴子一页图片的缩略图的url
        pic_ids = response.xpath('//img[@class="bde_image"]/@src').extract()
        #   用列表来装图片的url
        item['pic_urls'] = []

        for pic_url in pic_ids:
            #   取出每张图片的名称
            item['pic_name'] = pic_url.split('/')[-1]
            #   图片url
            url = 'http://imgsrc.baidu.com/forum/pic/item/' + item['pic_name']
            #   将url添加进列表
            item['pic_urls'].append(url)

        #   将item交给pipelines下载
        yield item

        #   下完一页图片后继续下一页
        next_url = response.xpath('//a[contains(text(),"下一页")]/@href').extract_first()
        if next_url:
            yield scrapy.request('https://tieba.baidu.com' + next_url, callback=self.parse_theme)

imagespipeline

• from scrapy.pipelines.images import imagespipeline
• 继承imagespipeline，重写get_media_requests()和file_path()方法

from scrapy.pipelines.images import imagespipeline
import scrapy


class postbarpipeline(imagespipeline):
    #   需要headers的网站，再使用
    headers = {
        'user-agent': '',
        'referer': '',
    }

    def get_media_requests(self, item, info):
        for pic_url in item['pic_urls']:
            # 为每个url生成一个request
            yield scrapy.request(pic_url)
            # 需要请求头的时候，添加headers参数
            # yield scrapy.request(pic_url, headers=self.headers)

    def file_path(self, request, response=none, info=none):
        # 重命名(包含后缀名)，若不重写这函数，图片名为哈希
        pic_path = request.url.split('/')[-1]
        return pic_path

settings文件

• 激活管道
  

• 设置图片保存地址
  
  
  

运行结果