selenium 之百度搜索，结果列表翻页查询

selenium之百度搜索，结果列表翻页查询

by:授客 QQ：1033553122

实例：百度搜索，结果列表翻页查询

解决问题：解决selenium driver获取web页面元素时,元素过期问题

 

思路1：获取所有“页面翻页链接”元素，然后遍历元素并点击

# -*- coding: utf-8 -*-

from selenium import webdriver

import time

 

if __name__ == "__main__":

    driver = webdriver.Firefox()

    driver.maximize_window()   

    driver.get('http://www.baidu.com')   

    driver.implicitly_wait(5)

   

    driver.find_element_by_id('kw1').send_keys('selenium')

    driver.find_element_by_id('su1').click()   

   

    page = driver.find_element_by_id('page')      

    pages = page.find_elements_by_tag_name('a')  #查找所有翻页跳转链接

   

#设置滚动条位置为底部

    js = 'document.documentElement.scrollTop=10000'

    for each in pages:

        driver.execute_script(js)   #拖动滚动条到底部

        each.click()

        driver.execute_script(js)

           

    time.sleep(3)     

driver.quit()

 

结果：点击第3页时，程序出错

selenium.common.exceptions.StaleElementReferenceException: Message: u'Element not found in the cache - perhaps the page has changed since it was looked up' ; Stacktrace:

 

即在cache中找不到元素，可能是在元素被找到之后页面变换了。 这就说明，当前页面发生跳转之后，存在cache中的与这个页面相关的元素也被清空了

 

思路2：基于思路1的错误结果分析>先获取每个页面数，然后每次点击某个页面，跳转后重新获取下一个页面翻页链接，然后点击，循环。。

# -*- coding: utf-8 -*-

 

from selenium import webdriver

import time

 

if __name__ == "__main__":

    driver = webdriver.Firefox()

    driver.maximize_window()   

    driver.get('http://www.baidu.com')   

    driver.implicitly_wait(5)

   

    driver.find_element_by_id('kw1').send_keys('selenium')

    driver.find_element_by_id('su1').click()   

   

    page = driver.find_element_by_id('page')      

    pages = page.find_elements_by_tag_name('a')

   

    js = 'document.documentElement.scrollTop=10000'

    total = len(pages)

    has_pre_page = False

    page_num = 0

    for i in range(total):

        driver.execute_script(js)

        pn=10

        page_num = page_num + 1 #设置页面号

        one_page = driver.find_element_by_css_selector('p[id="page"]>a:nth-of-type('+str(page_num)+')')

        one_page.click()

    

        #备注以下小段代码描述了页面变化规律，这个得自己去研究

  if not has_pre_page:   #点击第2页时会出现上一页，页号加1

            has_pre_page = True

            page_num = page_num + 1

        if page_num % 7 == 0: #page_num等于7时，页号减1

            page_num = page_num - 1

     

  time.sleep(2)

        driver.execute_script(js)

        time.sleep(2)

           

    time.sleep(3)     

driver.quit()

 

 

改进版(自动翻页，前翻页后翻页)

# -*- coding: utf-8 -*-

 

from selenium import webdriver

import time

 

if __name__ == "__main__":

    driver = webdriver.Firefox()

    driver.maximize_window()   

    driver.get('http://www.baidu.com')   

    driver.implicitly_wait(5)

   

    driver.find_element_by_id('kw1').send_keys('selenium selenium')#测试数据 selenium zhidashso dld#selenium zhidashso dldld

    driver.find_element_by_id('su1').click()   

   

    js = 'document.documentElement.scrollTop=10000'

               

    total = 0  #页面数

    is_next_page = True  #存在下一页

    page_num = 0   #要点击的页面号

 

    #往后翻页

    while is_next_page:#'sv_page\=1' in one_page.get_attribute('href')

        driver.execute_script(js)

        page_num = page_num + 1    #设置页号为下一页

        total = page_num   #记录页面数

        value=str(page_num)      

        try:       

            #查找指定页面

            one_page = driver.find_element_by_css_selector('p[id="page"]>a[href*=pn\='+value+']')  

            one_page.click()

            time.sleep(1)

            driver.execute_script(js)

            time.sleep(1)

        except:

            print('no next page')

            is_next_page = False

            total = total - 1

            break      

 

    #往前翻页

    while total >= 0:

        driver.execute_script(js) 

        try:

            total = total -1

            value = str(total)

            one_page = driver.find_element_by_css_selector('p[id="page"]>a[href*=pn\='+value+']')

            one_page.click()

            time.sleep(1)

            driver.execute_script(js)

            time.sleep(1)

        except:

            print('no pre page')

            break;  

   

    time.sleep(3)     

    driver.quit()