极客验证码破解
很多網(wǎng)站的登陸都有驗證碼一項,而極客的方案就是應(yīng)用的非常普遍。更多的場景是反反爬蟲的對抗中,極客驗證碼更是首選。
圖片來看一下
點擊后就出現(xiàn)上面的滑動圖片的窗口。本文親自嘗試代碼,來分享其中的坑。
本文是使用selenium自動化測試工具來驅(qū)動代碼完成驗證,因此要有谷歌的driver和谷歌瀏覽器。以http://www.epicc.com.cn/idprovider/views/login.jsp為例子。
首先
def get_geetest_button(self):"""獲取初始驗證按鈕:return:"""button = self.wait.until(EC.element_to_be_clickable((By.CLASS_NAME, 'slide-tip')))return buttondef get_position(self):"""獲取驗證碼位置:return: 驗證碼位置元組"""img=self.wait.until(EC.presence_of_element_located((By.CLASS_NAME ,'captcha-box-content')))time.sleep(2)location = img.locationprint(location)size = img.sizetop, bottom, left, right = location['y'], location['y'] + size['height'], location['x'], location['x'] + size['width']print((top, bottom, left, right))# return (top, bottom, left, right)return (238, 400, 1103, 1473)def get_screenshot(self):"""獲取網(wǎng)頁截圖:return: 截圖對象"""name = int(time.time())screenshot = self.browser.get_screenshot_as_png()with open("screenshot"+ str(name) + ".png","wb")as f:f.write(screenshot)screenshot = Image.open(BytesIO(screenshot))return screenshotdef get_slider(self):"""獲取滑塊:return: 滑塊對象"""slider = self.wait.until(EC.element_to_be_clickable((By.CLASS_NAME, 'slide-bar')))return sliderdef get_geetest_image(self, name='captcha.png'):"""獲取驗證碼圖片:return: 圖片對象"""top, bottom, left, right = self.get_position()print('驗證碼位置', top, bottom, left, right)time.sleep(3)screenshot = self.get_screenshot()# crop函數(shù)帶的參數(shù)為(起始點的橫坐標(biāo),起始點的縱坐標(biāo),寬度,高度)captcha = screenshot.crop((left, top, right, bottom))captcha.save(name)return captchadef open(self):"""打開網(wǎng)頁輸入用戶名密碼:return: None"""self.browser.get(self.url)time.sleep(1)passLoginButton = self.browser.find_element_by_xpath("//ul[@class='login-tab-list']/li[2]")passLoginButton.click()email = self.wait.until(EC.presence_of_element_located((By.ID, 'entryId')))password = self.wait.until(EC.presence_of_element_located((By.ID, 'password')))email.send_keys(self.email)password.send_keys(self.password)def get_gap(self, image1, image2):"""獲取缺口偏移量:param image1: 不帶缺口圖片:param image2: 帶缺口圖片:return:"""for i in range(INIT_LEFT, image1.size[0]):for j in range(image1.size[1]):if not self.is_pixel_equal(image1, image2, i, j):left = ireturn leftdef is_pixel_equal(self, image1, image2, x, y):"""判斷兩個像素是否相同:param image1: 圖片1:param image2: 圖片2:param x: 位置x:param y: 位置y:return: 像素是否相同"""# 取兩個圖片的像素點pixel1 = image1.load()[x, y]pixel2 = image2.load()[x, y]threshold = 60if abs(pixel1[0] - pixel2[0]) < threshold and abs(pixel1[1] - pixel2[1]) < threshold and abs(pixel1[2] - pixel2[2]) < threshold:return Trueelse:return Falsedef get_track(self, distance):"""根據(jù)偏移量獲取移動軌跡:param distance: 偏移量:return: 移動軌跡"""# 移動軌跡track = []# 當(dāng)前位移current = 0# 減速閾值mid = distance * 4 / 5# 計算間隔t = 0.2# 初速度v = 0while current < distance:if current < mid:# 加速度為正2a = 2else:# 加速度為負(fù)3a = -3# 初速度v0v0 = v# 當(dāng)前速度v = v0 + atv = v0 + a * t# 移動距離x = v0t + 1/2 * a * t^2move = v0 * t + 1 / 2 * a * t * t# 當(dāng)前位移current += move# 加入軌跡 track.append(round(move))return trackdef move_to_gap(self, slider, track):"""拖動滑塊到缺口處:param slider: 滑塊:param track: 軌跡:return:"""ActionChains(self.browser).click_and_hold(slider).perform()for x in track:ActionChains(self.browser).move_by_offset(xoffset=x, yoffset=0).perform()time.sleep(0.5)ActionChains(self.browser).release().perform()def login(self):"""登錄:return: None"""submit = self.wait.until(EC.element_to_be_clickable((By.CLASS_NAME, 'epiccLogin')))submit.click()time.sleep(10)print('登錄成功')def crack(self):# 輸入用戶名密碼 self.open()# 點擊驗證按鈕button = self.get_geetest_button()ActionChains(self.browser).move_to_element(button).perform()# button.click()time.sleep(5)# 獲取驗證碼圖片image1 = self.get_geetest_image('captcha1.png')# 點按呼出缺口print("點按呼出缺口")slider = self.get_slider()slider.click()# 獲取帶缺口的驗證碼圖片image2 = self.get_geetest_image('captcha2.png')# 獲取缺口位置gap = self.get_gap(image1, image2)print('缺口位置', gap)# 減去缺口位移 BORDER指的是待拼合的滑塊距離驗證碼圖片左邊的距離(經(jīng)過驗證大多數(shù)網(wǎng)站滑塊驗證碼這個值是是固定的)gap -= BORDER #滑塊要滑動的距離print("需要滑動的距離:{}".format(gap))gap = gap * 0.8# 獲取移動軌跡track = self.get_track(gap)print('滑動軌跡', track)# 拖動滑塊 self.move_to_gap(slider, track)success = self.wait.until(EC.text_to_be_present_in_element((By.CLASS_NAME, 'geetest_success_radar_tip_content'), '驗證成功'))print(success)# 失敗后重試if not success:self.crack()else:self.login()if __name__ == '__main__':crack = CrackGeetest()crack.crack()?
?
轉(zhuǎn)載于:https://www.cnblogs.com/hd-zg/p/9010919.html
總結(jié)
- 上一篇: 最大并发连接数和最大会话数的区别
- 下一篇: linux常用命令:wget 命令