python自动登陆网页+拼图自动验证

---

前言

因自动化测试需要，自动登陆网页并上传下载应用，增加python自动登陆网页的程序，主要难点是滑块的自动验证，这部分采用OpenCV算法实现。

---

一、使用的库

1. selenium（Web应用程序测试的工具。Selenium测试直接运行在浏览器中，就像真正的用户在操作一样。）
2. opencv(CV：Computer Vision，计算机视觉，包含超级多图形相关的算法)

二、步骤

示例使用谷歌浏览器，需要下载对应版本的driver，相关流程百度解决。

1.引入库

代码如下（示例）：

import base64
import time
import os
import random
from selenium import webdriver
import cv2 as cv
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.support.ui import WebDriverWait as WAIT
from selenium.webdriver.common.action_chains import ActionChains

# 新建selenium浏览器对象，后面是geckodriver.exe下载后本地路径
from selenium.webdriver.common.by import By

2.步骤和流程

1、打开浏览器，进入界面
2、输入账号和密码
3、保存拼图的验证背景、拼图块两张图片
4、计算两张图片的距离
5、移动拼图到指定位置
6、校验成功
7、登陆网页

重点说明：
计算两张图片的距离，使用到cv2中的算法，高斯模糊去噪、边缘检测，再计算出两张图的距离。

3.完整的测试代码

代码如下（示例）：

​
# This is a sample Python script.

# Press Shift+F10 to execute it or replace it with your code.
# Press Double Shift to search everywhere for classes, files, tool windows, actions, and settings.
import base64
import time
import os
import random

from selenium import webdriver

import cv2 as cv
from selenium.webdriver.support import expected_conditions as EC
#from selenium.webdriver import ActionChains
from selenium.webdriver.support.ui import WebDriverWait as WAIT
from selenium.webdriver.common.action_chains import ActionChains

# 新建selenium浏览器对象，后面是geckodriver.exe下载后本地路径
from selenium.webdriver.common.by import By

browser = webdriver.ChromeOptions()

# 网站登陆页面
# 浏览器访问登录页面
class pagesign:
    driver = None

    def __init__(self, userName, password, pwd):
        if len(pwd) > 0:
            self.downloadpath = pwd

        download_dir = "f:"
        chrome_options = webdriver.ChromeOptions()
        # chrome_options.add_argument('--headless')
        chrome_options.add_argument('--no-sandbox')
        chrome_options.add_argument('--safebrowsing-disable-download-protection') #关闭这种类型的文件可能会损害计算机…
        chrome_options.add_argument("--safebrowsing-disable-extension-blacklist")
        preferences = {"download.default_directory": download_dir,
                       "directory_upgrade": True,
                       "safebrowsing.enabled": True,
                       "download.prompt_for_download": False}
        chrome_options.add_experimental_option("prefs", preferences)

        self.driver = webdriver.Chrome(options=chrome_options, executable_path=r'chromedriver')

        self.driver.command_executor._commands["send_command"] = ("POST", '/session/$sessionId/chromium/send_command')
        params = {'cmd': 'Page.setDownloadBehavior', 'params': {'behavior': 'allow', 'downloadPath': self.downloadpath}}
        command_result = self.driver.execute("send_command", params)

        self.driver.implicitly_wait(10)
        self.driver.get(self.paxurl)

        self.driver.find_element(By.ID, "userName").send_keys(userName)
        self.driver.find_element(By.XPATH, value="/html/body/div[2]/div/div[2]/div/form/div[3]/div/div[1]/input").send_keys(
            password)

        self.wait = WAIT(self.driver, 5)

        if self.slide_verify():
            print('登陆成功')
            #self.save_cookie()
            #self.driver.close()
            self.driver.find_element(By.XPATH, value="/html/body/div[2]/div/div[2]/div/form/div[5]/div/button").click()
        else:
            print('第1次登陆失败')
            for i in range(4):
                print('正在尝试第%d次登陆' % (i + 2))
                if self.slide_verify():
                    print('第%d次登陆成功' % (i + 2))
                    #self.save_cookie()
                    #self.driver.close()
                    self.driver.find_element(By.XPATH, value="/html/body/div[2]/div/div[2]/div/form/div[5]/div/button").click()
                    return
                print('第%d次登陆失败' % (i + 2))
            print('登陆失败5次，停止登陆')
           # self.driver.close()

    def slide_verify(self):
        '''滑动验证'''
        js = f'''return document.getElementsByTagName("canvas")[{0}].toDataURL("image/png");'''
        base64str = self.driver.execute_script(js)
        page = base64str.split(',')[1]
        imagedata = base64.b64decode(page)

        with open(self.downloadpath + '\\' + "slider.png", "wb") as f:
            f.write(imagedata)

        js = f'''return document.getElementsByTagName("canvas")[{1}].toDataURL("image/png");'''
        base64str = self.driver.execute_script(js)
        page = base64str.split(',')[1]
        imagedata = base64.b64decode(page)

        with open(self.downloadpath + '\\' + "bg.png", "wb") as f:
            f.write(imagedata)

        slider_button = self.wait.until(EC.element_to_be_clickable((By.XPATH, '//div[@class="sliderMask"]')))

        distance = self.get_distance(self.downloadpath + '\\' + "slider.png", self.downloadpath + '\\' + "bg.png")
        print("distance=" + str(distance))
        tracks = self.get_tracks(distance)
        self.mouse_move(slider_button, tracks)
        try:
            element = self.wait.until(EC.visibility_of_element_located((By.XPATH, '//div[@class="sliderContainer sliderContainer_success"]')))
        except:
            return False
        else:
            return True


    def mouse_move(self, slide, tracks):
        '''鼠标滑动'''

        # 鼠标点击滑块并按照不放
        ActionChains(self.driver).click_and_hold(slide).perform()
        # 按照轨迹进行滑动，
        for track in tracks:
            ActionChains(self.driver).move_by_offset(track, 0).perform()
        print("time sleep 1")
        time.sleep(0.5)
        print("after time sleep 1")
        ActionChains(self.driver).release().perform()
        print("mouse release")
        time.sleep(3)
        print("after sleep")
        return


    def get_distance(self, slider_img_path, bg_img_path):
        """获取滑块移动距离"""

        # 背景图片处理
        bg_img = cv.imread(bg_img_path, 0)  # 读入灰度图片
        bg_img = cv.GaussianBlur(bg_img, (5, 5), 0)  # 高斯模糊去噪
        bg_img = cv.Canny(bg_img, 100, 200)  # Canny算法进行边缘检测
        #cv.imshow('canny', bg_img)
        # 滑块做同样处理
        slider_img = cv.imread(slider_img_path, 0)
        slider_img = cv.GaussianBlur(slider_img, (5, 5), 0)
        slider_img = cv.Canny(slider_img, 100, 200)
        #cv.imshow('slider', slider_img)
        # 寻找最佳匹配
        res = cv.matchTemplate(bg_img, slider_img, cv.TM_CCOEFF_NORMED)

        # 最小值，最大值，并得到最小值, 最大值的索引
        min_val, max_val, min_loc, max_loc = cv.minMaxLoc(res)
        # 例如：(-0.05772797390818596, 0.30968162417411804, (0, 0), (196, 1))
        top_left = max_loc[0]  # 横坐标
        print("min_loc=", min_loc[0])
        print("max_loc=", max_loc[0])

        len = min_loc[0] + (max_loc[0]-min_loc[0])/2
        return len


    def get_tracks(self, distance):
        '''滑动轨迹 '''
        tracks = []
        v = 0
        t = 0.2  # 单位时间
        current = 0  # 滑块当前位移
        while current < distance:
            a = random.randint(1, 3)
            v0 = v  # 初速度
            track = v0 * t + 0.5 * a * (t ** 2)  # 单位时间（0.2s）的滑动距离
            tracks.append(round(track))  # 加入轨迹
            current += round(track)
            v = v0 + a * t
        return tracks


# Press the green button in the gutter to run the script.
if __name__ == '__main__':
    path = os.getcwd()
    sign = pagesign("1234", "gewrwfas", path)

# See PyCharm help at https://www.jetbrains.com/help/pycharm/

​

获取距离需要按照实际情况调试，示例中的验证码相对比较简单，另外鼠标移动后，需要判断界面是否登陆成功，再点击登陆按钮

总结

以上是对自动登陆网页简单说明，花费了不到两天时间，对opencv、selenium有了初步的了解。