一、介绍
本例子用scrapy-splash通过搜狗搜索引擎，输入给定关键字抓取资讯信息。
给定关键字：数字；融合；电视
抓取信息内如下：
1、资讯标题
2、资讯链接
3、资讯时间
4、资讯来源
二、网站信息
三、数据抓取
针对上面的网站信息，来进行抓取
1、首先抓取信息列表
抓取代码：sels = site.xpath('//div[@class="vrwrap"]')
2、抓取标题
抓取代码：titles = sel.xpath('.//h3[@class="vrTitle"]/a/text()|.//h3[@class="vrTitle"]/a/em/text()')
3、抓取链接
抓取代码：it['url'] = sel.xpath('.//h3[@class="vrTitle"]/a/@href')[0].extract()
4、抓取日期
抓取代码：strdate = sel.xpath('.//p[@class="news-from"]/text()')
5、抓取来源
抓取代码：sources = sel.xpath('.//p[@class="news-from"]/text()')
四、完整代码
# -*- coding: utf-8 -*-
import scrapy
from scrapy import Request
from scrapy.spiders import Spider
from scrapy_splash import SplashRequest
from scrapy_splash import SplashMiddleware
from scrapy.http import Request, HtmlResponse
from scrapy.selector import Selector
from scrapy_splash import SplashRequest
from scrapy_ott.items import SplashTestItem
from scrapy_ott.mongoDB import mongoDbBase
import scrapy_ott.IniFile
import sys
import os
import re
import time reload(sys)
sys.setdefaultencoding('utf-8') # sys.stdout = open('output.txt', 'w') class sougouSpider(Spider):
name = 'sougou' db = mongoDbBase() configfile = os.path.join(os.getcwd(), 'scrapy_ott\setting.conf')
cf = scrapy_ott.IniFile.ConfigFile(configfile)
information_keywords = cf.GetValue("section", "information_keywords")
information_wordlist = information_keywords.split(';')
# websearchurl = cf.GetValue("sougou", "websearchurl")
start_urls = []
for word in information_wordlist:
start_urls.append('http://news.sogou.com/news?page=1&query=' + word)
start_urls.append('http://news.sogou.com/news?page=2&query=' + word)
start_urls.append('http://news.sogou.com/news?page=3&query=' + word) # request需要封装成SplashRequest
def start_requests(self): for url in self.start_urls:
index = url.rfind('=')
yield SplashRequest(url
, self.parse
, args={'wait': ''},
meta={'keyword': url[index + 1:]}
) def Comapre_to_days(self,leftdate, rightdate):
'''
比较连个字符串日期，左边日期大于右边日期多少天
:param leftdate: 格式：2017-04-15
:param rightdate: 格式：2017-04-15
:return: 天数
'''
l_time = time.mktime(time.strptime(leftdate, '%Y-%m-%d'))
r_time = time.mktime(time.strptime(rightdate, '%Y-%m-%d'))
result = int(l_time - r_time) / 86400
return result def date_isValid(self, strDateText):
'''
判断日期时间字符串是否合法：如果给定时间大于当前时间是合法，或者说当前时间给定的范围内
:param strDateText: 四种格式 '慧聪网 7小时前'; '新浪游戏 29分钟前' ; '中国行业研究网 2017-6-13'
:return: True:合法；False:不合法
'''
currentDate = time.strftime('%Y-%m-%d')
source = strDateText.split(' ')[0]
if strDateText.find('分钟前') > 0 :
return True,source,currentDate
elif strDateText.find('小时前') > 0:
datePattern = re.compile(r'\d{1,2}')
ch = int(time.strftime('%H')) # 当前小时数
strDate = re.findall(datePattern, strDateText)
if len(strDate) == 1:
if int(strDate[0]) <= ch: # 只有小于当前小时数，才认为是今天
return True,source,currentDate
else:
datePattern = re.compile(r'\d{4}-\d{1,2}-\d{1,2}')
strDate = re.findall(datePattern, strDateText)
if len(strDate) == 1:
if self.Comapre_to_days(currentDate, strDate[0]) == 0:
return True,source,currentDate
return False, '','' def parse(self, response):
keyword = response.meta['keyword']
site = Selector(response)
sels = site.xpath('//div[@class="vrwrap"]')
item_list = []
for sel in sels:
strdate = sel.xpath('.//p[@class="news-from"]/text()')
if len(strdate)>0:
flag, source, date = self.date_isValid(strdate[0].extract())
if flag:
titles = sel.xpath('.//h3[@class="vrTitle"]/a/text()|.//h3[@class="vrTitle"]/a/em/text()')
title = ''
for t in titles:
title += str(t.extract())
if title.find(keyword) > -1:
it = SplashTestItem()
it['source'] = source
it['url'] = sel.xpath('.//h3[@class="vrTitle"]/a/@href')[0].extract()
it['date'] = date
it['keyword'] = keyword
it['title'] = title item_list.append(it) if len(item_list)>0:
self.db.SaveInformations(item_list)
return item_list