Python selenium PIL 全网页滚动截屏 && headless全网页截屏

1. 思路

   ​ 先截取当前屏幕的图片，获取其高度作为base高度 h，再获取全网页body到尾部的高度 H ，循环截取图片，再通过PIL进行拼接。

2. 代码

   # -*- coding:utf-8 -*-
   # author: [email protected]
   # software: PyCharm
   import os
   
   from PIL import Image
   from time import sleep
   
   
   class ScreenShot:
       __JS__ = {
         
         
           'scroll_to_bottom': "window.scroll({top:document.body.clientHeight,left:0,behavior:'auto'});",
           'scroll_to_y': "window.scroll({top:%d,left:0,behavior:'auto'});",
       }
       __base_end__ = 'tmp_end.png'
       __scroll_bottom__ = 'scroll_to_bottom'
       __scroll_y__ = 'scroll_to_y'
       __body__ = '//body'
       __height__ = 'height'
       __clear_shell__ = 'rm -rf *.png'
       __RGB__ = 'RGB'
   
       @classmethod
       def screen_shot(cls, driver, title, uploader_url='', delete=False):
           """
           全网页滚动截屏
           :param driver: webdriver 示例
           :param title: 标题（最终图片命名）
           :param uploader_url: 上传url
           :param delete: 是否清除所有图片
           :return:
           """
           base_image = '{}.png'.format(title)
           driver.save_screenshot(base_image)
           body_h = int(driver.find_element_by_xpath(cls.__body__).size.get(cls.__height__))
           current_h = Image.open(base_image).size[1] / 2
           for i in range(1, int(body_h / current_h)):
               driver.execute_script(cls.__JS__[cls.__scroll_y__] % (current_h * i))
               sleep(.5)
               driver.save_screenshot(f'tmp_{
           
           i}.png')
               cls.__join_images__(base_image, f'tmp_{
           
           i}.png', 0, base_image)
           driver.execute_script(cls.__JS__[cls.__scroll_bottom__])
           driver.save_screenshot(cls.__base_end__)
           cls.__join_images__(base_image, cls.__base_end__, int(current_h - int(body_h % current_h)), base_image)
           # TODO 上传图片
           url = ''
           # 移除图片
           if delete:
               os.system(cls.__clear_shell__)
           return url
   
       @classmethod
       def __join_images__(cls, png1, png2, size=0, output='result.png'):
           """
           图片拼接
           :param png1: 图片1
           :param png2: 图片2
           :param size: 两个图片重叠的距离
           :param output: 输出的图片文件
           :return:
           """
           size = size * 2
           img1, img2 = Image.open(png1), Image.open(png2)
           size1, size2 = img1.size, img2.size
           joint = Image.new(cls.__RGB__, (size1[0], size1[1] + size2[1] - size))
           loc1, loc2 = (0, 0), (0, size1[1] - size)
           joint.paste(img1, loc1)
           joint.paste(img2, loc2)
           joint.save(output)
   
   
   if __name__ == '__main__':
       from selenium import webdriver
       driver = webdriver.Chrome()
       driver.get("https://www.cnblogs.com/worldline/")
       ScreenShot.screen_shot(driver, 'worldline')
       driver.quit()
   
   

3. 其他

   如果是在headless模式,可以使用

   
   def get_image(url, pic_name):
       """
       适用于无头全屏截图
       :param url: url访问路径
       :param pic_name: 图片名称
       :return:
       """
       chrome_options = Options()
       chrome_options.add_argument('headless')
       driver = webdriver.Chrome(options=chrome_options)
       driver.get(url)
       time.sleep(.5)
       width = driver.execute_script("return document.documentElement.scrollWidth")
       height = driver.execute_script("return document.documentElement.scrollHeight")
       print(width, height)
       driver.set_window_size(width, height)
       time.sleep(.5)
       driver.save_screenshot(pic_name)
       driver.close()