自动化测试解决滑块解锁（解决非人工操作提示）

# coding = utf-8

import base64
import time
import cv2
import numpy as np
import random

from selenium import webdriver
from selenium.webdriver import ChromeOptions, ActionChains
from selenium.webdriver.common.by import By


class loginTAMS(object):

    def __init__(self):
        options = ChromeOptions()
        options.add_experimental_option('excludeSwitches', ['enable-automation'])

        # 创建webdriver对象
        self.driver = webdriver.Chrome(options=options)


    def get_distance(self, driver):
        script = """
                return document.getElementsByTagName("canvas")[0].toDataURL("image/png");
            """
        # 执行 JS 代码并拿到图片 base64 数据
        im_info = driver.execute_script(script)  # 执行js文件得到带图片信息的图片数据
        im_base64 = im_info.split(',')[1]  # 拿到base64编码的图片信息
        im_bytes = base64.b64decode(im_base64)  # 转为bytes类型
        with open('target.png', 'wb') as f:  # 保存图片到本地
            f.write(im_bytes)

        script = """
                return document.getElementsByTagName("canvas")[1].toDataURL("image/png");
            """
        # 执行 JS 代码并拿到图片 base64 数据
        im_info = driver.execute_script(script)  # 执行js文件得到带图片信息的图片数据
        im_base64 = im_info.split(',')[1]  # 拿到base64编码的图片信息
        im_bytes = base64.b64decode(im_base64)  # 转为bytes类型
        with open('target2.png', 'wb') as f:  # 保存图片到本地
            f.write(im_bytes)

        # 使用opencv读取两张图片
        simg = cv2.imread('target2.png')
        bimg = cv2.imread('target.png')

        # 灰度处理，降低偏差
        s_img = cv2.cvtColor(simg, cv2.COLOR_BGR2GRAY)
        b_img = cv2.cvtColor(bimg, cv2.COLOR_BGR2GRAY)

        # 保存两张灰度处理的图片
        cv2.imwrite('hui_simg.png', s_img)
        cv2.imwrite('hui_bimg.png', b_img)

        # 处理滑块图片，保存有效部分
        s_img = s_img[s_img.any(1)]

        # opencv的匹配算法，匹配模块寻找两张图片的相似之处
        result = cv2.matchTemplate(b_img, s_img, cv2.TM_CCOEFF_NORMED)

        # 获取坐标
        # 获取最大索引
        index_max = np.argmax(result)
        # 获取到坐标
        y, x = np.unravel_index(index_max, result.shape)
        print("y:", y, "x:", x)
        return x


    def get_track(self, distance):
        '''
        拿到移动轨迹，模仿人的滑动行为，先匀加速后匀减速
        :param self:
        :param distance:需要移动的距离
        :return: 存放每0.2秒移动的距离
        '''
        # 初速度
        v = 0
        # 单位时间为0.2s来统计轨迹，轨迹即0.2内的位移
        t = 0.3
        # 位移/轨迹列表，列表内的一个元素代表0.2s的位移
        tracks = []
        # 当前的位移
        current = 0
        # 到达mid开始减速
        mid = distance * 5/8

        #distance += 10 # 先滑过一点，最后再反着滑动回来
        # a = random.randint(1, 3)
        while current < distance:
            if current < mid:
                # 加速度越小，单位时间的位移越小，模拟的轨迹就越多越详细
                a = random.randint(2, 5)  # 加速运动
            else:
                a = -random.randint(2, 4)  # 减速运动

            # 初速度
            v0 = v
            # 0.2秒时间内的位移
            s = v0*t+0.5*a*(t**2)
            # 当前位置
            current += s
            # 添加到轨迹列表
            tracks.append(round(s))
            # 速度已经达到v，该速度作为下次的初速度
            v = v0+a*t

        # 反着滑动到大概准确位置
        # for i in range(4):
        #     tracks.append(-random.randint(9, 10))
        random.shuffle(tracks)
        return tracks


    def slide(self, tracks, webdriver):
        slider = webdriver.find_element(By.XPATH, "//span[text()='向右滑']/preceding-sibling::div/div")
        # 鼠标点击并按住不松
        action = ActionChains(webdriver)
        action.click_and_hold(slider).perform()
        # 让鼠标随机往下移动一段距离
        action.move_by_offset(xoffset=0, yoffset=100).perform()
        time.sleep(0.15)
        for item in tracks:
            action.move_by_offset(xoffset=item, yoffset=random.randint(-2, 2)).perform()
        # 稳定一秒再松开
        time.sleep(1)
        action.release()
        action.perform()
        time.sleep(1)
        # 随机拿开鼠标
        #action.move_by_offset(xoffset=random.randint(200, 300), yoffset=random.randint(200, 300)).perform()
        time.sleep(0.2)
        info = webdriver.find_element(By.XPATH, "//div[@id='slideVerify']/following-sibling::div")
        if '验证通过' in info.text:
            return 1

        if '验证失败' in info.text:
            return 2

    def enterLoginInfo(self, url, userIdXpath, userid, passwordXpath, password):
        self.driver.get(url)
        self.driver.maximize_window()
        time.sleep(2)

        self.driver.find_element(By.XPATH, userIdXpath).send_keys(userid)
        self.driver.find_element(By.XPATH, passwordXpath).send_keys(password)


    def verifyCanvas(self, submitxpath):
        while True:
            distance = self.get_distance(self.driver)
            tracks = self.get_track(distance)
            returnValue = self.slide(tracks, self.driver)
            if returnValue == 1:
                break

        self.driver.find_element(By.XPATH, submitxpath).click()

由于我用的自动化测试框架为robotframework，上述的方法需要浏览器的驱动操作，所以需要将上述方法放到seleniumlibrary文件的keywords下的element.py文件中，

    def get_distance(self):
        script = """
                return document.getElementsByTagName("canvas")[0].toDataURL("image/png");
            """
        # 执行 JS 代码并拿到图片 base64 数据
        im_info = self.driver.execute_script(script)  # 执行js文件得到带图片信息的图片数据
        im_base64 = im_info.split(',')[1]  # 拿到base64编码的图片信息
        im_bytes = base64.b64decode(im_base64)  # 转为bytes类型
        with open('target.png', 'wb') as f:  # 保存图片到本地
            f.write(im_bytes)

        script = """
                return document.getElementsByTagName("canvas")[1].toDataURL("image/png");
            """
        # 执行 JS 代码并拿到图片 base64 数据
        im_info = self.driver.execute_script(script)  # 执行js文件得到带图片信息的图片数据
        im_base64 = im_info.split(',')[1]  # 拿到base64编码的图片信息
        im_bytes = base64.b64decode(im_base64)  # 转为bytes类型
        with open('target2.png', 'wb') as f:  # 保存图片到本地
            f.write(im_bytes)

        # 使用opencv读取两张图片
        simg = cv2.imread('target2.png')
        bimg = cv2.imread('target.png')

        # 灰度处理，降低偏差
        s_img = cv2.cvtColor(simg, cv2.COLOR_BGR2GRAY)
        b_img = cv2.cvtColor(bimg, cv2.COLOR_BGR2GRAY)

        # 保存两张灰度处理的图片
        cv2.imwrite('hui_simg.png', s_img)
        cv2.imwrite('hui_bimg.png', b_img)

        # 处理滑块图片，保存有效部分
        s_img = s_img[s_img.any(1)]

        # opencv的匹配算法，匹配模块寻找两张图片的相似之处
        result = cv2.matchTemplate(b_img, s_img, cv2.TM_CCOEFF_NORMED)

        # 获取坐标
        # 获取最大索引
        index_max = np.argmax(result)
        # 获取到坐标
        y, x = np.unravel_index(index_max, result.shape)
        print("y:", y, "x:", x)
        return x+10

    def get_track(self, distance):
        '''
        拿到移动轨迹，模仿人的滑动行为，先匀加速后匀减速
        :param self:
        :param distance:需要移动的距离
        :return: 存放每0.2秒移动的距离
        '''
        # 初速度
        v = 0
        # 单位时间为0.2s来统计轨迹，轨迹即0.2内的位移
        t = 0.3
        # 位移/轨迹列表，列表内的一个元素代表0.2s的位移
        tracks = []
        # 当前的位移
        current = 0
        # 到达mid开始减速
        mid = distance * 5 / 8

        # distance += 10 # 先滑过一点，最后再反着滑动回来
        # a = random.randint(1, 3)
        while current < distance:
            if current < mid:
                # 加速度越小，单位时间的位移越小，模拟的轨迹就越多越详细
                a = random.randint(2, 5)  # 加速运动
            else:
                a = -random.randint(2, 4)  # 减速运动

            # 初速度
            v0 = v
            # 0.2秒时间内的位移
            s = v0 * t + 0.5 * a * (t ** 2)
            # 当前位置
            current += s
            # 添加到轨迹列表
            tracks.append(round(s))
            # 速度已经达到v，该速度作为下次的初速度
            v = v0 + a * t

        # 反着滑动到大概准确位置
        # for i in range(4):
        #     tracks.append(-random.randint(9, 10))
        random.shuffle(tracks)
        return tracks

    def slide(self, tracks, sliderXpath, checkXpath):
        slider = self.driver.find_element(By.XPATH, sliderXpath)
        # 鼠标点击并按住不松
        action = ActionChains(self.driver)
        action.click_and_hold(slider).perform()
        # 让鼠标随机往下移动一段距离
        action.move_by_offset(xoffset=0, yoffset=100).perform()
        time.sleep(0.15)
        for item in tracks:
            action.move_by_offset(xoffset=item, yoffset=random.randint(-2, 2)).perform()
        # 稳定一秒再松开
        time.sleep(1)
        action.release()
        action.perform()
        time.sleep(1)
        # 随机拿开鼠标
        # action.move_by_offset(xoffset=random.randint(200, 300), yoffset=random.randint(200, 300)).perform()
        time.sleep(0.2)
        info = self.driver.find_element(By.XPATH, checkXpath)
        if '验证通过' in info.text:
            return 1

        if '验证失败' in info.text:
            return 2

然后在robotframework中直接调用verifyCanvas方法即可。