python爬虫——破解滑动验证码

版权声明：本文为博主原创文章，未经博主允许不得转载。 https://blog.csdn.net/wzyaiwl/article/details/89603059

滑动验证码的破解是验证码类中唯一一个不需要涉及深度学习的一类反爬措施。它完全可以依靠python的一些包来进行破解。下面例举下要用的模块。

1. pillow：处理图片
2. request：获取图片
3. selenium：模拟浏览器的行为
4. random：取随机数
5. io：模拟文件存储
6. re：正则匹配
7. time：时间模块，用到时间等待功能

这里以虎嗅网为例，虎嗅网用的滑动验证码是由极验提供的。下面叙述下破解滑块验证码的难点：

1. 得到完整的验证码图片：难点存在于网页中直接得到的图片是乱的，需要我们自己重新进行裁剪和拼接。
2. 获取缺口的距离：之所以说这个是难点，是考虑到有些人对图片处理不熟悉，如果有点图片处理的基础，这个并不难。
3. 模拟人为的滑块移动轨迹：难点在于不合理的移动轨迹会被极验判断为非人类操作。

源代码展示：

from selenium import webdriver
from selenium.webdriver.chrome.options import Options
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver import ActionChains

import time
import re
import requests
from io import BytesIO
from PIL import Image
import random

class SlipCaptcha(object):
    def __init__(self):
        """
        初始化界面,初始化了driver对象和wait对象,以及传入了一个url地址
        """
        self.url = 'https://www.huxiu.com/'
        option = Options()
        option.add_argument('--window-size=1332,700')
        option.add_argument('--disable-infobars')
        # option.add_argument('headless')
        option.add_argument('incognito')
        self.driver = webdriver.Chrome(chrome_options=option)
        self.wait = WebDriverWait(self.driver,10)
    def home_page(self):
        """
        主要流程实施的函数
        1:driver.get函数先到达虎嗅首页面
        2:点击登陆按钮
        3:获取验证码的图片,图片有两份,一份是有缺口的,一份没有缺口
        4:获得的两份图片是被打乱,需要我们根据坐标信息,重新裁剪拼接
        5:拼接后,比较两份图片的区别,得到缺口的x方向的距离
        6:依据得到的距离,滑动滑块,由于存在对滑块轨迹的限制.随意我们还要设置如何活动,即以什么样的速度,加速度来滑动.
        :return:
        """
        self.driver.get(self.url)
        login_button = self.driver.find_element_by_xpath('//a[@class="js-login"]')
        login_button.click()
        gap_image_position,nogap_image_position, gap_image_url,nogap_image_url = self.get_image_info()
        new_gapimage,new_nogapimage = self.get_image_complete(gap_image_position,nogap_image_position,gap_image_url,nogap_image_url)
        distance = self.get_move_distance(new_gapimage, new_nogapimage)
        self.slid_button(distance)

    def get_image_info(self):
        """
        获得图片的信息，如图片的url，图片的坐标信息。
        共获得两份图片，一份是有缺口的，一份没有缺口。
        :return:
        """
        gap_image_list = self.wait.until(EC.presence_of_all_elements_located((By.XPATH,'//div[@class="user-login-box"]//div[@class="gt_cut_bg gt_show"]/div[@class="gt_cut_bg_slice"]')))
        nogap_image_list = self.wait.until(EC.presence_of_all_elements_located((By.XPATH,'//div[@class="user-login-box"]//div[@class="gt_cut_fullbg gt_show"]/div[@class="gt_cut_fullbg_slice"]')))
        gap_image_url = re.findall(r'url\("(.*?)"\)',gap_image_list[0].get_attribute('style'))[0]
        nogap_image_url = re.findall(r'url\("(.*?)"\)',nogap_image_list[0].get_attribute('style'))[0]
        gap_image_position = [re.findall(r'background-position: -(.*?)px -?(.*?)px;',i.get_attribute('style'))[0] for i in gap_image_list]
        nogap_image_position = [re.findall(r'background-position: -(.*?)px -?(.*?)px;',i.get_attribute('style'))[0] for i in nogap_image_list]
        return gap_image_position,nogap_image_position,gap_image_url,nogap_image_url

    def get_image_complete(self,gap_image_position, nogap_image_position, gap_image_url,nogap_image_url):
        """
        将获得混乱的图片，用获得的信息，拼接成正常的图片。
        :param gap_image_position:
        :param nogap_image_position:
        :param gap_image_url:
        :param nogap_image_url:
        :return:
        """
        gap_image_file = BytesIO(requests.get(gap_image_url).content)
        nogap_image_file = BytesIO(requests.get(nogap_image_url).content)
        old_gapimage = Image.open(gap_image_file)
        new_gapimage = Image.new('RGB', (260, 116))
        old_nogapimage = Image.open(nogap_image_file)
        new_nogapimage = Image.new('RGB', (260, 116))
        up_count = 0
        down_count = 0
        # 拼接缺口图片
        for i in gap_image_position[:26]:
            cut_image = old_gapimage.crop((int(i[0]), 58, int(i[0]) + 10, 116))  # 左上顶点，右下顶点
            new_gapimage.paste(cut_image, (up_count, 0))
            up_count = up_count + 10
        for i in gap_image_position[26:]:
            cut_image = old_gapimage.crop((int(i[0]), 0, int(i[0]) + 10, 58))  # 左上顶点，右下顶点
            new_gapimage.paste(cut_image, (down_count, 58))
            down_count = down_count + 10
        # 拼接无缺口图片
        up_count = 0
        down_count = 0
        for i in nogap_image_position[:26]:
            cut_image = old_nogapimage.crop((int(i[0]), 58, int(i[0]) + 10, 116))  # 左上顶点，右下顶点
            new_nogapimage.paste(cut_image, (up_count, 0))
            up_count = up_count + 10
        for i in gap_image_position[26:]:
            cut_image = old_nogapimage.crop((int(i[0]), 0, int(i[0]) + 10, 58))  # 左上顶点，右下顶点
            new_nogapimage.paste(cut_image, (down_count, 58))
            down_count = down_count + 10
        return new_gapimage,new_nogapimage

    def get_move_distance(self,new_gapimage, new_nogapimage):
        def compare_image(p1,p2):
            """
            比较图片的像素
            由于RGB图片一个像素点是三维的，所以循环三次
            :return:
            """
            for i in range(3):
                if abs(p1[i]-p2[i]) >= 50:
                    return False
                return True
        for i in range(260):
            for j in range(116):
                gap_pixel = new_gapimage.getpixel((i,j))
                nogap_pixel = new_nogapimage.getpixel((i,j))
                if not compare_image(gap_pixel,nogap_pixel):
                    return i

    def slid_button(self,distance):
        """
        根据缺口位置，移动滑块特定的距离distance
        :param diatance:
        :return:
        """
        #获取滑块元素
        button = self.driver.find_element_by_xpath('//div[@class="user-login-box"]//div[@class="gt_slider_knob gt_show"]')
        ActionChains(self.driver).click_and_hold(button).perform()
        time.sleep(0.5)
        track_list = self.track(distance-3)
        # print(track_list)
        for i in track_list:
            ActionChains(self.driver).move_by_offset(i,0).perform()
        time.sleep(1)
        ActionChains(self.driver).release().perform()
    def track(self,distance):
        """
        规划移动的轨迹
        加速度用到random模块,随机选择给定的加速度
        :param distance:
        :return:
        """
        #匀速移动
        # for i in range(distance):
        #     ActionChains(self.driver).move_by_offset(1, 0).perform()
        # ActionChains(self.driver).move_by_offset(distance-5, 0).perform()
        t = 0.1
        speed = 0
        current = 0
        mid = 3 / 5 * distance
        track_list = []
        while current < distance:
            if current < mid:
                a = random.choice([1,2,3])
                # a = 3
            else:
                a = random.choice([-1,-2,-3])
                # a = -4
            move_track = speed * t + 0.5 * a * t**2
            track_list.append(round(move_track))
            speed = speed + a*t
            current += move_track
        #模拟人类来回移动了一小段
        end_track = [1,0]*10 +[0]*10+[-1,0]*10
        track_list.extend(end_track)
        offset = sum(track_list) - distance
        #由于四舍五入带来的误差,这里需要补回来
        if offset > 0:
            track_list.extend(offset*[-1,0])
        elif offset < 0:
            track_list.extend(offset*[1,0])
        return track_list
    def run(self):
        """
        运行函数,在主函数中执行该函数即可
        :return:
        """
        try:
            self.home_page()
        finally:
            time.sleep(0.5)
            self.driver.quit()
if __name__ == '__main__':
    S = SlipCaptcha()
    S.run()