[Python爬虫] 之九：Selenium +phantomjs抓取活动行中会议活动（单线程抓取）

　　思路是这样的，给一系列关键字：互联网电视；智能电视；数字；影音；家庭娱乐；节目；视听；版权；数据等。在活动行网站搜索页（http://www.huodongxing.com/search?city=%E5%85%A8%E5%9B%BD&pi=1）的文本输入框中分别输入每个关键字，在搜索结果中抓取需要的数据。

　　首先通过Selenium+IE驱动得到每个关键字搜索结果的url（首页，因为以后各个页的url就是索引不一样）和总页数，保存的列表里面。然后再循环列表，用Selenium +phantomjs抓取每个关键字对应的url抓取对应的数据。

　　

　　

　　

　　具体代码如下：

　　

# coding=utf-8
import os
import re
from selenium import webdriver
import selenium.webdriver.support.ui as ui
from selenium.webdriver.common.keys import Keys
import time
from selenium.webdriver.common.action_chains import ActionChains
from selenium.webdriver.support.select import Select
import IniFile
from selenium.webdriver.common.keys import Keys
import LogFile

class huodongxing:

    def __init__(self):
        #通过配置文件获取IEDriverServer.exe路径
        configfile = os.path.join(os.getcwd(),'MeetingConfig.conf')
        self.cf = IniFile.ConfigFile(configfile)
        IEDriverServer = self.cf.GetValue("section", "IEDriverServer")
        #每抓取一页数据延迟的时间，单位为秒，默认为5秒
        self.pageDelay = 5
        pageInteralDelay = self.cf.GetValue("section", "pageInteralDelay")
        if pageInteralDelay:
            self.pageDelay = int(pageInteralDelay)

        os.environ["webdriver.ie.driver"] = IEDriverServer
        self.urldriver = webdriver.Ie(IEDriverServer)
        # self.driver = webdriver.PhantomJS()
        self.wait = ui.WebDriverWait(self.urldriver, 20)
        self.urldriver.maximize_window()

    def compareDate(self,dateLeft, dateRight):
        '''
        比较俩个日期的大小
        :param dateLeft: 日期 格式2017-03-04
        :param dateRight:日期 格式2017-03-04
        :return: 1：左大于右，0：相等，-1：左小于右
        '''
        dls = dateLeft.split('-')
        drs = dateRight.split('-')
        if len(dls) > len(drs):
            return 1
        if int(dls[0]) == int(drs[0]) and int(dls[1]) == int(drs[1]) and int(dls[2]) == int(drs[2]):
            return 0

        if int(dls[0]) > int(drs[0]):
            return 1
        elif int(dls[0]) == int(drs[0]) and int(dls[1]) > int(drs[1]):
            return 1
        elif int(dls[0]) == int(drs[0]) and int(dls[1]) == int(drs[1]) and int(dls[2]) > int(drs[2]):
            return 1
        return -1

    def scroll_top(self):
        '''
        滚动条拉到顶部
        :return:
        '''
        if self.urldriver.name == "chrome":
            js = "var q=document.body.scrollTop=0"

        else:
            js = "var q=document.documentElement.scrollTop=0"
        return self.urldriver.execute_script(js)

    def scroll_foot(self):
        '''
        滚动条拉到底部
        :return:
        '''

        if self.urldriver.name == "chrome":
            js = "var q=document.body.scrollTop=10000"

        else:
            js = "var q=document.documentElement.scrollTop=10000"
        return self.urldriver.execute_script(js)

    def get_UrlPageCountList(self,webSearchUrl,keywordList):
        '''
        根据关键字列表获取每个关键字搜索出内容url及对应的页数
        :param webSearchUrl: 给定网址的搜索首页
        :param keywordList: 关键字列表
        :return: url及对应的页数的列表
        '''
        search_url_pageCount_list = []
        # firstUrl = self.cf.GetValue("section", "webSearchUrl")
        self.urldriver.get(webSearchUrl)
        # self.urldriver.implicitly_wait(3)
        time.sleep(3)
        pageCountLable = self.cf.GetValue("section", "pageCountLable")
        for keyword in keywordList:
            if len(keyword) > 0:
                js = "var obj = document.getElementById('mainSearchTextbox');obj.value='" + keyword + "';"
                print '关键字：%s' % keyword
                self.urldriver.execute_script(js)
                # 点击搜索链接
                ss_elements = self.urldriver.find_element_by_id("mainSearchTextbox")
                ss_elements.send_keys(Keys.RETURN)
                time.sleep(5)
                current_url = self.urldriver.current_url.replace('pi=1', 'pi=')
                try:
                    elements = self.urldriver.find_elements_by_xpath(pageCountLable)
                    # 要爬虫的页数
                    strCount = elements[0].text.encode('utf8')
                    pageCount = int(strCount) / 10
                    if int(strCount) % 10 > 0:
                        pageCount = pageCount + 1
                    search_url_pageCount_list.append(current_url + '_' + str(pageCount))
                except Exception, e:
                    print e.message

        self.urldriver.close()
        self.urldriver.quit()
        self.driver = webdriver.PhantomJS()
        self.wait = ui.WebDriverWait(self.driver, 20)
        self.driver.maximize_window()
        return search_url_pageCount_list

    def scrapy_Data(self):
        '''抓取数据'''
        start = time.clock()

        webSearchUrl = self.cf.GetValue("section", "webSearchUrl")
        keyword = self.cf.GetValue("section", "keywords")
        keywordList = keyword.split(';')
        #搜索页及对应的页数
        search_url_pageCount_list = self.get_UrlPageCountList(webSearchUrl,keywordList)

        if len(search_url_pageCount_list) > 0:
            htmlLable = self.cf.GetValue("section", "htmlLable")
            # logfile = os.path.join(os.getcwd(), r'log.txt')
            # log = LogFile.LogFile(logfile)

            OriginalUrlLabel = self.cf.GetValue("section", "OriginalUrlLabel")
            currentDate = time.strftime('%Y-%m-%d')
            datePattern = re.compile(r'\d{4}-\d{2}-\d{2}')
            keyword_index = 0
            for url_pageCount in search_url_pageCount_list:
                try:
                    kword = keywordList[keyword_index]
                    print ''
                    print '关键字：%s ' % kword
                    pageCount = int(url_pageCount.split('_')[1])
                    page_Count = pageCount
                    recordCount = 0
                    if pageCount > 0:
                        current_url = url_pageCount.split('_')[0]
                        pageIndex = 0
                        while pageCount > 0:
                            url = current_url + str(pageIndex)
                            self.driver.get(url)

                            # 延迟3秒
                            time.sleep(3)
                            # self.driver.implicitly_wait(3)
                            pageCount = pageCount - 1
                            self.wait.until(lambda driver: self.driver.find_elements_by_xpath(htmlLable))
                            Elements = self.driver.find_elements_by_xpath(htmlLable)

                            # 查找微博对应的原始url
                            urlList = []
                            self.wait.until(lambda driver: self.driver.find_elements_by_xpath(OriginalUrlLabel))
                            hrefElements = self.driver.find_elements_by_xpath(OriginalUrlLabel)
                            for hrefe in hrefElements:
                                urlList.append(hrefe.get_attribute('href').encode('utf8'))

                            # self.driver.implicitly_wait(2)
                            index = 0
                            strMessage = ' '
                            strsplit = '\n------------------------------------------------------------------------------------\n'
                            index = 0
                            # 每页中有用记录
                            usefulCount = 0
                            for element in Elements:
                                txt = element.text.encode('utf8')
                                txts = txt.split('\n')
                                strDate = re.findall(datePattern, txt)
                                # 日期大于今天并且搜索的关键字在标题中才认为是复合要求的数据
                                if len(strDate) > 0 and self.compareDate(strDate[0], currentDate) == 1 and \
                                                txts[0].find(kword) > -1:
                                    print ' '
                                    print txt
                                    print '活动链接：' + urlList[index]
                                    print strsplit

                                    strMessage = txt + "\n"
                                    strMessage += '活动链接：' + urlList[index] + "\n"
                                    strMessage += strsplit
                                    strMessage = unicode(strMessage, 'utf8')
                                    log.WriteLog(strMessage)
                                    usefulCount = usefulCount + 1
                                    recordCount = recordCount + 1
                                index = index + 1

                            pageIndex = pageIndex + 1
                            if usefulCount == 0:
                                break

                    print "共浏览了: %d 页数据" % page_Count
                    print "共抓取了: %d 个符合条件的活动记录" % recordCount
                except Exception, e:
                    print e.message
                keyword_index = keyword_index + 1

        self.driver.close()
        self.driver.quit()
        end = time.clock()
        print "整个过程用时间: %f 秒" % (end - start)

# #测试抓取数据
obj = huodongxing()
obj.scrapy_Data()

配置文件内容：

[section]
#IE驱动的路径
iedriverserver = C:\Program Files\Internet Explorer\IEDriverServer.exe

pageinteraldelay = 5

#要搜索的标签，如果有多个，中间用分号隔开
htmlLable = //ul[@class='event-horizontal-list-new']/li

#要获取爬虫也是的标签
pageCountLable = //span[@class='text-primary']

#给定网址的搜索首页Url
webSearchUrl = http://www.huodongxing.com/search?city=%E5%85%A8%E5%9B%BD&pi=1

#查找对应的原始url
OriginalUrlLabel = //ul[@class='event-horizontal-list-new']/li/h3/a

#文本输入框要搜索的关键字
keywords = 互联网电视;智能电视;数字;影音;家庭娱乐;节目;视听;版权;数据

抓取的数据结构为：

整个过程用时间: 184.897053 秒