爬取知名社区技术文章_article_3

Posted 2020-09-29 北门吹雪

爬虫主逻辑处理，获取字段，获取主url和子url

#!/usr/bin/python3
# -*- coding: utf-8 -*-

import scrapy
from scrapy.http import Request
from urllib import parse
from JobBole.items import JobboleItem, ArticleItemLoader


class ExampleSpider(scrapy.Spider):
    name = ‘jobbole‘
    # allowed_domains = [‘example.com‘]
    # 起始url
    start_urls = [‘http://blog.jobbole.com/all-posts/‘]

    def parse(self, response):
        # 获取主页面所以有效详情页面url，图片url，下一页url
        home_all_node = response.css(‘.post.floated-thumb .post-thumb a‘)
        # 获取节点，循环出每个单张图片和单个url交个下载器和子url处理
        for home_node in home_all_node:
            img_url = home_node.css(‘img::attr(src)‘).extract_first(‘‘)
            cont_url = home_node.css(‘::attr(href)‘).extract_first(‘‘)
            # yield 提供异步方法，parser 进行域名拼接，meta进行传值
            yield Request(url=parse.urljoin(response.url, cont_url),
                          meta={‘img_url‘: img_url},
                          callback=self.analysie_go)
        # 提取下一页并下载
        next_page_url = response.css(‘.next.page-numbers::attr(href)‘).extract_first(‘‘)
        if next_page_url:
            yield Request(url=parse.urljoin(response.url, next_page_url), callback=self.parse)
            
    def analysie_go(self, response):
        # 解析详情页面获取所需其他字段的值
        img_url = response.meta.get(‘img_url‘, ‘0‘)
        load_item = ArticleItemLoader(item=JobboleItem(), response=response)
        load_item.add_value(‘img_url‘, img_url)
        load_item.add_value(‘cont_url‘, response.url)
        load_item.add_value(‘cont_id‘, response.url)
        load_item.add_css(‘title‘, ‘.entry-header h1::text‘)
        load_item.add_css(‘publish_time‘, ‘.entry-meta-hide-on-mobile::text‘)
        load_item.add_xpath(‘cont‘, ‘//div[@class="entry"]//text()‘)
        load_item.add_css(‘link_num‘, ‘.vote-post-up h10::text‘)
        load_item.add_css(‘collection_num‘, ‘.bookmark-btn::text‘)
        load_item.add_css(‘comment_num‘, ‘.post-adds a span::text‘)
        article_items = load_item.load_item()
        # 把获取的字段交给items
        yield article_items