selenium 实现点触验证码识别

YY点触验证码识别案例

这个网站在第一次请求网站的时候，用selenium定位不到验证码，最后发现是iframe标签的问题

我们需要再次请求iframe里面的链接，打开新的网页之后就可以定位操作了。

 

  1 # !/usr/bin/env python
  2 # -*- coding:utf-8 -*-
  3 
  4 from chaojiying import Chaojiying
  5 import random
  6 import time
  7 from PIL import Image
  8 from io import BytesIO
  9 from selenium import webdriver
 10 from selenium.webdriver.support.ui import WebDriverWait
 11 from selenium.webdriver import ActionChains
 12 from selenium.webdriver.common.by import By
 13 from selenium.webdriver.support import expected_conditions as ec
 14 
 15 CHAOJIYING_NAME = '******' # 超级鹰账号
 16 CHAOJIYING_PWD = '********'  # 超级鹰密码
 17 CHAOJIYING_ID = 894613  # 软件ID
 18 CHAOJIYING_KIND = 9004  # 验证码类型
 19 
 20 
 21 class YYVerification(object):
 22     """
 23     此类用于YY的验证码识别，可以应用到类似的验证码识别上，这种验证码类型是
 24     点击类验证码.这里我们是对接超级鹰平台。
 25     """
 26     def __init__(self):
 27         self.url = 'https://aq.yy.com/p/reg/account.do?appid=&url=&fromadv=udbclsd_r'
 28         self.driver = webdriver.Chrome()
 29         self.chaojiying = Chaojiying(CHAOJIYING_NAME, CHAOJIYING_PWD, CHAOJIYING_ID)
 30 
 31     def __del__(self):
 32         self.driver.close()
 33 
 34     def screen_shot(self):
 35         """
 36         网页截屏
 37         :return: bool值
 38         """
 39         # self.driver.maximize_window()
 40         time.sleep(2)
 41         # # 用于网页的向下滚动
 42         # js = 'var q=document.documentElement.scrollTop=300'
 43         # self.driver.execute_script(js)
 44         # time.sleep(1)
 45         self.driver.save_screenshot('yy.png')
 46         return True
 47 
 48     def i_url(self):
 49         """
 50         实现对iframe标签的url请求，并定位新网页
 51         :return: 定位对象
 52         """
 53         i = self.driver.find_element_by_xpath('/html/body/div[2]/div[2]/div/iframe')
 54         url_1 = i.get_attribute('src')
 55         print(url_1)
 56         self.driver.get(url_1)
 57         wait = WebDriverWait(self.driver, 20)
 58         element = wait.until(ec.presence_of_element_located((By.XPATH, '//*[@id="mPickWords"]/div[1]')))
 59         return element
 60 
 61     def shear_location(self):
 62         """
 63         计算对截屏剪切的坐标，注意这里不同电脑可能会修改start_x到end_y
 64         后边添加的数是根据作者的电脑优化的
 65         :return: 返回坐标以及网页剪切定位的对象
 66         """
 67         time.sleep(random.random() + 1)
 68         print('正在获取div')
 69         div = self.i_url()
 70         start_x = div.location['x'] + 25
 71         start_y = div.location['y'] + 65
 72         end_x = div.location['x'] + 350 + 25
 73         end_y = div.location['y'] + 218 + 65
 74         result = (start_x, start_y, end_x, end_y)
 75         print(result)
 76         return result, div
 77 
 78     @staticmethod
 79     def shear_image(axis):
 80         im = Image.open('yy.png')
 81         new_image = im.crop(axis)
 82         new_image.save('new_img.png')
 83         return new_image
 84 
 85     def upload_picture(self, img):
 86         """
 87         上传图片(Byte)，返回点击的坐标。这里对返回的坐标进行了处理，
 88         处理过程可以看成 shear_location 函数的坐标数据逆处理过程
 89         :param img: 上传的图片
 90         :return: 点击坐标
 91         """
 92         image = img
 93         byte_array = BytesIO()
 94         image.save(byte_array, format('PNG'))
 95         # 提交图片
 96         result = self.chaojiying.post_pic(byte_array.getvalue(), CHAOJIYING_KIND)
 97         print(result)
 98         if '无可用题分' in result['err_str']:
 99             print('题分已经不足请充值')
100             raise ValueError
101         pic_str = result['pic_str']
102         pic_list = [[i for i in x.split(',')] for x in pic_str.split('|')]
103         for i in pic_list:
104             i[0] = int(int(i[0])*(282/354))
105             i[1] = int(int(i[1])*(219/274))
106         print(pic_list)
107         return pic_list
108 
109     def click_now(self, coordinates, axis, element):
110         """
111         这里如果验证失败会进行回调，实现验证码的自动验证。
112         :param coordinates: 点击坐标
113         :param axis: 剪切坐标
114         :param element: 剪切位置的定位
115         :return: 返回点击的成功信息
116         """
117         print('点击开始')
118         print(coordinates)
119         for location in coordinates:
120             ActionChains(self.driver).move_to_element_with_offset(element, location[0], location[1]).click().perform()
121             time.sleep(random.random() + 1.8)
122         submission = self.driver.find_element_by_xpath('//*[@id="mPickWords"]/div[2]/span[4]')
123         submission.click()
124         time.sleep(1)
125         if 'pw_submit_disabled' in submission.get_attribute('class'):
126             return '点击成功'
127         else:
128             time.sleep(2)
129             print('正在重新识别新的验证码')
130             self.revalidation(axis, element)
131             return '点击成功'
132 
133     def revalidation(self, axis, element):
134         """
135         网页跳转之后，执行验证的主程序
136         :param axis: 剪切坐标
137         :param element: 剪切位置的定位对象
138         :return: 验证成果
139         """
140         self.screen_shot()
141         if not self.screen_shot():
142             return '截图失败'
143         time.sleep(1)
144         # 剪切图片
145         new_image = self.shear_image(axis)
146         new_image.show()
147         # 上传图片并返回点击坐标
148         click_coordinates = self.upload_picture(new_image)
149         # 点击验证码
150         print(self.click_now(click_coordinates, axis, element))
151         return '点击验证结束'
152 
153     def main(self):
154         self.driver.get(self.url)
155         wait = WebDriverWait(self.driver, 20)
156         wait.until(ec.presence_of_element_located((By.XPATH, '/html/body/div[2]/div[2]/div/iframe')))
157         axis, element = self.shear_location()
158         self.revalidation(axis, element)
159 
160 
161 if __name__ == '__main__':
162     yy = YYVerification()
163     yy.main()

 

 

转载于:https://www.cnblogs.com/dan-baishucaizi/p/8903758.html