1心得体会
之前还没有听说过selenium模板,现在发现这个模板挺牛逼的。能自动模拟用户操作浏览器,不过缺点就是慢了写,但是不需要自己写cookie、headers这些了,毕竟是用真实的浏览器去模板。主要是好不是很熟练,还需要多敲代码。
2、selenium介绍
selenium是一款自动化测试工具,支持很多主流的浏览器。只要浏览器安装了依赖驱动就行。
下面代码是模拟访问百度操作
1 from selenium import webdriver 2 driver = webdriver.Chrome() 3 driver.get("https://www.baidu.com")
只需简单的代码就可以让浏览器自己访问百度了。当然,这只是简单的操作。
对于用来操作来说,还有输入内容和点击这两个常用的功能
test = driver.find_element_by_xpath(xpath)这个语句可以根据xpath来找到相应的位置.
如果可以点击的话,test.click() 就会自己点击。
如果是需要输入内容的话,test.send_keys()
3.模拟登陆 https://dig.chouti.com并点赞
1、先配置驱动
1 # 配置驱动 2 option = webdriver.ChromeOptions() 3 # driver = webdriver.Chrome('/Users/wupeiqi/drivers/chromedriver', chrome_options=option) 4 driver = webdriver.Chrome( chrome_options=option)
2、找到登陆按钮并点击
1 # 1. 控制浏览器打开指定页面 2 driver.get("https://dig.chouti.com") 3 4 5 # 2. 找到登录按钮 6 btn_login = driver.find_element_by_xpath('//*[@id="login-link-a"]') 7 # 3. 点击按钮 8 btn_login.click()
3、输入账号和密码
1 # 4. 找到手机标签 2 input_user = driver.find_element_by_xpath('//*[@id="mobile"]') 3 # 5. 找到密码标签 4 input_pwd = driver.find_element_by_xpath('//*[@id="mbpwd"]') 5 6 # 6. 输入用户名 7 input_user.send_keys('18879970712') 8 # 7. 输入密码 9 input_pwd.send_keys('lg970712')
4、点击登陆按钮
1 # 8. 点击登录按钮 2 input_submit = driver.find_element_by_xpath( 3 '//*[@id="footer-band"]/div[5]/div/div/div[1]/div[2]/div[4]/div[2]/div/span[1]') 4 input_submit.click()
5、找到要点赞的新闻并点赞
1 # 9. 点击跳转 2 news = driver.find_element_by_xpath('//*[@id="newsContent20696066"]/div[2]/a[1]') 3 news.click() 4 driver.execute_script("arguments[0].click();", news)
6、最后关闭驱动
1 # 10.关闭浏览器 2 driver.close()
4、滑动验证
实现滑动验证之前的步骤和前面的一样。登陆网站,找到输入账号和密码的位置输入账号和密码。不过验证的时候就需要滑动验证的,这就需要处理图片了。滑动验证一开始有一张完整的图片,将完整的图片下载下来,然后还有一张不完整的图片,也下载下来。我们需要做的就是对比着两张图片哪里的像素不同,根据不同像素的位置来计算需要滑动多少像素,然后去模拟浏览器去滑动。
完整代码如下:
1 from selenium import webdriver 2 from selenium.webdriver import ActionChains 3 from selenium.webdriver.common.by import By 4 from selenium.webdriver.common.keys import Keys 5 from selenium.webdriver.support import expected_conditions as EC 6 from selenium.webdriver.support.wait import WebDriverWait 7 import os 8 import shutil 9 from PIL import Image 10 import time 11 12 13 def get_snap(driver): 14 driver.save_screenshot('full_snap.png') 15 page_snap_obj = Image.open('full_snap.png') 16 17 return page_snap_obj 18 19 20 def get_image(driver): 21 img = driver.find_element_by_class_name('geetest_canvas_img') 22 time.sleep(2) 23 location = img.location 24 size = img.size 25 26 left = location['x'] 27 top = location['y'] 28 right = left + size['width'] 29 bottom = top + size['height'] 30 31 page_snap_obj = get_snap(driver) 32 33 image_obj = page_snap_obj.crop((left, top, right, bottom)) 34 # image_obj.show() 35 with open('code.png', 'wb') as f: 36 image_obj.save(f, format='png') 37 return image_obj 38 39 40 def get_distance(image1, image2): 41 # start = 0 42 # threhold = 70 43 # for i in range(start, image1.size[0]): 44 # for j in range(0, image1.size[1]): 45 # rgb1 = image1.load()[i, j] 46 # rgb2 = image2.load()[i, j] 47 # res1 = abs(rgb1[0] - rgb2[0]) 48 # res2 = abs(rgb1[1] - rgb2[1]) 49 # res3 = abs(rgb1[2] - rgb2[2]) 50 # # print(res1,res2,res3) 51 # if not (res1 < threhold and res2 < threhold and res3 < threhold): 52 # print(111111, i, j) 53 # return i - 13 54 # print(2222, i, j) 55 # return i - 13 56 start = 0 57 threhold = 70 58 v = [] 59 for i in range(start, image1.size[0]): 60 for j in range(0, image1.size[1]): 61 rgb1 = image1.load()[i, j] 62 rgb2 = image2.load()[i, j] 63 res1 = abs(rgb1[0] - rgb2[0]) 64 res2 = abs(rgb1[1] - rgb2[1]) 65 res3 = abs(rgb1[2] - rgb2[2]) 66 67 if not (res1 < threhold and res2 < threhold and res3 < threhold): 68 print(i) 69 if i not in v: 70 v.append(i) 71 72 stop = 0 73 for i in range(0, len(v)): 74 val = i + v[0] 75 if v[i] != val: 76 stop = v[i] 77 break 78 79 width = stop - v[0] 80 print(stop, v[0], width) 81 return width 82 83 84 def get_tracks(distance): 85 import random 86 exceed_distance = random.randint(0, 5) 87 distance += exceed_distance # 先滑过一点,最后再反着滑动回来 88 v = 0 89 t = 0.2 90 forward_tracks = [] 91 92 current = 0 93 mid = distance * 3 / 5 94 while current < distance: 95 if current < mid: 96 a = random.randint(1, 3) 97 else: 98 a = random.randint(1, 3) 99 a = -a 100 s = v * t + 0.5 * a * (t ** 2) 101 v = v + a * t 102 current += s 103 forward_tracks.append(round(s)) 104 105 # 反着滑动到准确位置 106 v = 0 107 t = 0.2 108 back_tracks = [] 109 110 current = 0 111 mid = distance * 4 / 5 112 while abs(current) < exceed_distance: 113 if current < mid: 114 a = random.randint(1, 3) 115 else: 116 a = random.randint(-3, -5) 117 a = -a 118 s = -v * t - 0.5 * a * (t ** 2) 119 v = v + a * t 120 current += s 121 back_tracks.append(round(s)) 122 return {'forward_tracks': forward_tracks, 'back_tracks': list(reversed(back_tracks))} 123 124 125 def crack(driver): # 破解滑动认证 126 # 1、点击按钮,得到没有缺口的图片 127 button = driver.find_element_by_xpath('//*[@id="embed-captcha"]/div/div[2]/div[1]/div[3]') 128 button.click() 129 130 # 2、获取没有缺口的图片 131 image1 = get_image(driver) 132 133 # 3、点击滑动按钮,得到有缺口的图片 134 button = driver.find_element_by_class_name('geetest_slider_button') 135 button.click() 136 137 # 4、获取有缺口的图片 138 image2 = get_image(driver) 139 140 # 5、对比两种图片的像素点,找出位移 141 distance = get_distance(image1, image2) 142 print(distance) 143 # 144 # 6、模拟人的行为习惯,根据总位移得到行为轨迹 145 tracks = get_tracks(int(distance / 2)) 146 147 # 7、按照行动轨迹先正向滑动,后反滑动 148 button = driver.find_element_by_class_name('geetest_slider_button') 149 ActionChains(driver).click_and_hold(button).perform() 150 151 # 正常人类总是自信满满地开始正向滑动,自信地表现是疯狂加速 152 for track in tracks['forward_tracks']: 153 ActionChains(driver).move_by_offset(xoffset=track, yoffset=0).perform() 154 155 # 结果傻逼了,正常的人类停顿了一下,回过神来发现,卧槽,滑过了,然后开始反向滑动 156 time.sleep(0.5) 157 for back_track in tracks['back_tracks']: 158 ActionChains(driver).move_by_offset(xoffset=back_track, yoffset=0).perform() 159 # 160 # # 小范围震荡一下,进一步迷惑极验后台,这一步可以极大地提高成功率 161 ActionChains(driver).move_by_offset(xoffset=3, yoffset=0).perform() 162 ActionChains(driver).move_by_offset(xoffset=-3, yoffset=0).perform() 163 164 # # 成功后,骚包人类总喜欢默默地欣赏一下自己拼图的成果,然后恋恋不舍地松开那只脏手 165 time.sleep(0.5) 166 ActionChains(driver).release().perform() 167 168 169 def login_luffy(username, password): 170 driver = webdriver.Chrome() 171 driver.set_window_size(960, 800) 172 try: 173 # 1、输入账号密码回车 174 driver.implicitly_wait(3) 175 driver.get('https://www.luffycity.com/login') 176 input_username = driver.find_element_by_xpath('//*[@id="router-view"]/div/div/div[2]/div[2]/input[1]') 177 input_pwd = driver.find_element_by_xpath('//*[@id="router-view"]/div/div/div[2]/div[2]/input[2]') 178 179 input_username.send_keys(username) 180 input_pwd.send_keys(password) 181 182 # 2、破解滑动认证 183 crack(driver) 184 185 time.sleep(10) # 睡时间长一点,确定登录成功 186 finally: 187 pass 188 # driver.close() 189 190 191 if __name__ == '__main__': 192 login_luffy(username='wupeiqi', password='123123123')