新浪网分类资讯爬虫

      从GitHub得到完整项目（https://github.com/daleyzou/sinainfo.git）

1、简介

爬取新浪网导航页所有下所有大类、小类、小类里的子链接，以及子链接页面的新闻内容。

效果演示图：

2、代码

items.py

import scrapy


class SinainfoItem(scrapy.Item):
    # 大类的标题和url
    parentTitle = scrapy.Field()
    parentUrls = scrapy.Field()

    # 小类的标题和子url
    subTitle = scrapy.Field()
    subUrls = scrapy.Field()

    # 小类目录存储路径
    subFilename = scrapy.Field()

    # 小类下的子链接
    sonUrls = scrapy.Field()

    # 大文章标题和内容
    head = scrapy.Field()
    content = scrapy.Field()

spiders/sina.py(爬虫)

# -*- coding: utf-8 -*-
import scrapy
import sys
import os

# noinspection PyUnresolvedReferences
from sinainfo.items import SinainfoItem

reload(sys)
sys.setdefaultencoding('utf-8')


class SinaSpider(scrapy.Spider):
    name = 'sina'
    allowed_domains = ['sina.com.cn']
    start_urls = ['http://news.sina.com.cn/guide/']

    def parse(self, response):
        items = []
        # 所有大类的标题和url
        parentTitle = response.xpath("//div[@id='tab01']/div/h3/a/text()").extract()
        parentUrls = response.xpath('//div[@id="tab01"]/div/h3/a/@href').extract()

        # 所有小类的ur 和 标题
        subUrls = response.xpath('//div[@id="tab01"]/div/ul/li/a/@href').extract()
        subTitle = response.xpath('//div[@id="tab01"]/div/ul/li/a/text()').extract()

        # 爬取所有大类
        for i in range(0, len(parentTitle)):
            # 指定大类目录的路径和目录名
            parentFilename = "./Data/" + parentTitle[i]
            # 如果目录不存在，则创建目录
            if (not os.path.exists(parentFilename)):
                os.makedirs(parentFilename)

            # 爬取所有小类
            for j in range(0, len(subUrls)):
                item = SinainfoItem()
                # 保存大类的title和urls
                item['parentTitle'] = parentTitle[i]
                item['parentUrls'] = parentUrls[i]
                # 检查小类的url是否以同类别大类url开头，如果是返回Ture
                if_belong = subUrls[j].startswith(item['parentUrls'])
                # 如果属于本大类，将存储目录放在本大类下
                if (if_belong):
                    subFilename = parentFilename + '/' + subTitle[j]
                    # 如果目录不存在，则创建目录
                    if (not os.path.exists(subFilename)):
                        os.makedirs(subFilename)
                    # 存储 小类url、title、和filename字段数据
                    item['subUrls'] = subUrls[j]
                    item['subTitle'] = subTitle[j]
                    item['subFilename'] = subFilename
                    items.append(item)

        # 发送每个小类url的Request请求，得到Response连同包含meta数据
                    # 一同交给回调函数second_parse()处理
        for item in items:
            yield scrapy.Request(url = item['subUrls'],
                                 meta={'meta_1':item}, callback=self.second_parse)

    # 对于返回的小类url，在进行递归请求
    def second_parse(self, response):
        # 提取每次Response的meta数据
        meta_1 = response.meta['meta_1']
        # 取出小类里所有字链接
        sonUrls = response.xpath('//a/@href').extract()

        items = []
        for i in range(0, len(sonUrls)):
            # 检查每个链接是否以大类url开头、以.shtml结尾，如果是返回True
            if_belong = sonUrls[i].endswith('.shtml') and sonUrls[i].startswith(
                meta_1['parentUrls'])
            # 如果属于本大类，获取字段值放在同一个item下便于传输
            if (if_belong):
                item = SinainfoItem()
                item['parentTitle'] = meta_1['parentTitle']
                item['parentUrls'] = meta_1['parentUrls']
                item['subTitle'] = meta_1['subTitle']
                item['subUrls'] = meta_1['subUrls']
                item['subFilename'] = meta_1['subFilename']
                item['sonUrls'] = sonUrls[i]
                items.append(item)

        for item in items:
            yield scrapy.Request(url = item['sonUrls'],
                                 meta = {'meta_2':item}, callback=self.detail_parse)

    # 数据解析方法，获取文章标题和内容
    def detail_parse(self, response):
        item = response.meta['meta_2']
        content = ""
        head = response.xpath('//h1[@id="main_title"]/text()')
        content_list = response.xpath('//div[@id="artibody"]/p/text()').extract()
        # 将p标签里的文本内容合并到一起
        for content_one in content_list:
            content += content_one
        item['head'] = head
        item['content'] = content

pipelines.py

class SinainfoPipeline(object):
    def process_item(self, item, spider):
        sonUrls = item['sonUrls']

        # 文件名为子链接url中间部分，并将/替换为_,保存为.txt
        filename = sonUrls[7:-6].replace('/', '_')
        filename += ".txt"

        fp = open(item['subFilename']+'/'+filename, 'w')
        fp.write(item['content'])
        fp.close()
        return item

settings.py

BOT_NAME = 'sinainfo'

SPIDER_MODULES = ['sinainfo.spiders']
NEWSPIDER_MODULE = 'sinainfo.spiders'

LOG_LEVEL = 'DEBUG'
# Crawl responsibly by identifying yourself (and your website) on the user-agent
USER_AGENT = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_3) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.116 Safari/537.36'
DOWNLOAD_DELAY = 3
COOKIES_ENABLED = False

ITEM_PIPELINES = {
   'sinainfo.pipelines.SinainfoPipeline': 300,
}

3、运行

方法一：

（1）在项目根目录下新建main.py文件,用于调试

from scrapy import cmdline
cmdline.execute('scrapy crawl sina'.split())

（2）执行程序

py2 main.py

方法二：

在命令行下：

(1)切换到项目/sinainfo/sinainfo/spiders

(2)执行 scrapy crawl sina

作者：Daley Zou

出处：http://www.cnblogs.com/daleyzou/p/8325990.html