使用 Selenium 抓取网页时阻止登录覆盖窗口 [英] Blocking login overlay window when scraping web page using Selenium

问题描述

我试图在 10 个网页中抓取一长串书籍.当循环第一次点击 next > 按钮时，网站会显示登录覆盖，因此 selenium 无法找到目标元素.我已经尝试了所有可能的解决方案:

I am trying to scrape a long list of books in 10 web pages. When the loop clicks on next > button for the first time the website displays a login overlay so selenium can not find the target elements. I have tried all the possible solutions:

1. 使用一些镀铬选项.
2. 使用 try-except 点击叠加层上的 X 按钮.但它只出现一次(第一次点击 next > 时).问题是，当我将这个 try-except 块放在 while True: 循环的末尾时，它变得无限，因为我使用 continue in except 因为我不想打破循环.
3. 向 Chrome 添加一些弹出窗口阻止程序扩展程序，但在我运行代码时它们不起作用，尽管我使用 options.add_argument('load-extension=' + ExtensionPath) 添加了扩展程序.

1. Use some chrome options.
2. Use try-except to click X button on the overlay. But it appears only one time (when clicking next > for the first time). The problem is that when I put this try-except block at the end of while True: loop, it became infinite as I use continue in except as I do not want to break the loop.
3. Add some popup blocker extensions to Chrome but they do not work when I run the code although I add the extension using options.add_argument('load-extension=' + ExtensionPath).

这是我的代码:

options = Options()
options.add_argument('start-maximized')
options.add_argument('disable-infobars')
options.add_argument('disable-avfoundation-overlays')
options.add_argument('disable-internal-flash')
options.add_argument('no-proxy-server')
options.add_argument("disable-notifications")
options.add_argument("disable-popup")
Extension = (r'C:\Users\DELL\AppData\Local\Google\Chrome\User Data\Profile 1\Extensions\ifnkdbpmgkdbfklnbfidaackdenlmhgh\1.1.9_0')
options.add_argument('load-extension=' + Extension)
options.add_argument('--disable-overlay-scrollbar')

driver = webdriver.Chrome(options=options)
driver.get('https://www.goodreads.com/list/show/32339._50_?page=')
wait = WebDriverWait(driver, 2)

review_dict = {'title':[], 'author':[],'rating':[]}


html_soup = BeautifulSoup(driver.page_source, 'html.parser')
prod_containers = html_soup.find_all('table', class_ = 'tableList js-dataTooltip')


while True:
   
    table =  driver.find_element_by_xpath('//*[@id="all_votes"]/table')

    for product in table.find_elements_by_xpath(".//tr"):
        
        for td in product.find_elements_by_xpath('.//td[3]/a'):
            title = td.text
            review_dict['title'].append(title)

        for td in product.find_elements_by_xpath('.//td[3]/span[2]'):
            author = td.text
            review_dict['author'].append(author)

        for td in product.find_elements_by_xpath('.//td[3]/div[1]'):
            rating = td.text[0:4]
            review_dict['rating'].append(rating)
            
    try:
        close = wait.until(EC.element_to_be_clickable((By.XPATH, '/html/body/div[3]/div/div/div[1]/button')))
        close.click()
        
    except NoSuchElementException:
        continue
                
    try:
        element = wait.until(EC.element_to_be_clickable((By.CLASS_NAME, 'next_page')))
        element.click()
        
    except TimeoutException:    
        break
    
    
df = pd.DataFrame.from_dict(review_dict) 
df

任何帮助，例如我是否可以将循环更改为 for 循环点击 next > 按钮直到结束而不是 while 循环，或者我应该在哪里放置 try-except 块以关闭叠加层，或者如果有Chromeoption 可以禁用叠加层.提前致谢

Any help like if I can change the loop to for loop clicks next > button until the end rather than while loop or where should I put try-except block to close the overlay or if there is Chromeoption can disable overlay. Thanks in advance

推荐答案

感谢您分享您的代码和遇到问题的网站.我能够使用 xpath 关闭登录模式.我接受了这个挑战并使用类对象分解了代码.

1 个对象用于 selenium.webdriver.chrome.webdriver，另一个对象用于您想要抓取数据的页面 ( https://www.goodreads.com/list/show/32339).

在以下方法中，我使用了 Javascript return arguments[0].scrollIntoView(); 方法，并且能够滚动到页面上显示的最后一本书.完成后，我可以点击下一步按钮

Thank you for sharing your code and the website that you are having trouble with. I was able to close the Login Modal by using xpath. I took this challenge and broke up the code using class objects.

1 object is for the selenium.webdriver.chrome.webdriver and the other object is for the page that you wanted to scrape the data against ( https://www.goodreads.com/list/show/32339 ).

In the following methods, I used the Javascript return arguments[0].scrollIntoView(); method and was able to scroll to the last book that displayed on the page. After I did that, I was able to click the next button

def scroll_to_element(self, xpath : str):
        element = self.chrome_driver.find_element(By.XPATH, xpath)
        self.chrome_driver.execute_script("return arguments[0].scrollIntoView();", element)

def get_book_count(self):
        return self.chrome_driver.find_elements(By.XPATH, "//div[@id='all_votes']//table[contains(@class, 'tableList')]//tbody//tr").__len__()

def click_next_page(self):
        # Scroll to last record and click "next page"
        xpath = "//div[@id='all_votes']//table[contains(@class, 'tableList')]//tbody//tr[{0}]".format(self.get_book_count())
        self.scroll_to_element(xpath)
        self.chrome_driver.find_element(By.XPATH, "//div[@id='all_votes']//div[@class='pagination']//a[@class='next_page']").click()

一旦我点击了下一步"按钮，我看到了模态显示.我能够找到模态的 xpath 并能够关闭模态.

Once I clicked on the "Next" button, I saw the modal display. I was able to find the xpath for the modal and was able to close the modal.

def is_displayed(self, xpath: str, int = 5):
        try:
            webElement = DriverWait(self.chrome_driver, int).until(
                DriverConditions.presence_of_element_located(locator = (By.XPATH, xpath))
            )
            
            return True if webElement != None else False
        except:
            return False

def is_modal_displayed(self):
        return self.is_displayed("//body[@class='modalOpened']")

def close_modal(self):
        self.chrome_driver.find_element(By.XPATH, "//div[@class='modal__content']//div[@class='modal__close']").click()
        if(self.is_modal_displayed()):
            raise Exception("Modal Failed To Close")

希望这能帮助您解决问题.

I hope this helps you to solve your problem.

这篇关于使用 Selenium 抓取网页时阻止登录覆盖窗口的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！