使用pyppteer+ddddocr通过极验滑块验证码附源码

前言：

        验证码是大多数爬虫都需要克服的难题。pyppteer是目前市场主流的自动化工具之一，它的优势在于不易被浏览器检测到，ddddocr也是目前主流的验证码识别的第三方模块。接下来就通过python自动化工具和ddddocr等第三方模块完成极验滑块。

一：使用自动化工具打开网站

        目标网站：行为验证4.0-适应型验证码-滑动验证,点选验证,图片验证-极验GeeTest

1.  
   # 浏览器 启动参数
2.  
   start_parm = {
3.  
   # 关闭无头浏览器
4.  
   "headless": False,
5.  
   "args": [
6.  
   '--disable-infobars', # 关闭自动化提示框
7.  
   '--no-sandbox', # 关闭沙盒模式
8.  
   '--start-maximized', # 窗口最大化模式
9.  
    
10.  
    ],
11.  
    }
12.  
    browser = await launch(**start_parm)
13.  
    page = await browser.newPage()
14.  
     
15.  
    # 设置网页 视图大小
16.  
    await page.setViewport(viewport={'width': 1920, 'height': 1080})
17.  
    await page.goto('https://www.geetest.com/adaptive-captcha-demo')

二：控制鼠标定位到指定元素

1.  
   await page.waitForXPath('//div[@class="type-config"]') # 等待元素加载
2.  
   botton1 = await page.xpath('//div[@class="tab-item tab-item-1"]') # 滑块拼图验证按钮
3.  
   await botton1[0].click()
4.  
   await page.click('#captcha', options={
5.  
   'button': 'left',
6.  
   'clickCount': 2,
7.  
   'delay': 300, # 延迟点击(ms)
8.  
   })
9.  
   botton2 = await page.xpath('//*[@aria-label="点击按钮开始验证"]') # 开始验证按钮
10.  
    await botton2[0].click()

三：提取滑块拼图照片url

1.  
   elements_1 = await page.xpath(
2.  
   '//*[@id="captcha"]/div[2]/div[1]/div[4]/div[1]/div[2]/div/div/div[1]/div[1]/div[1]/@style') # 滑块图片链接
3.  
   elements_2 = await page.xpath(
4.  
   '//*[@id="captcha"]/div[2]/div[1]/div[4]/div[1]/div[2]/div/div/div[1]/div[2]/@style') # 背景图片链接
5.  
   for element in elements_1:
6.  
   sc = await page.evaluate('(element) => element.textContent', element)
7.  
   sc_url = sc.split('"')[1].split('"')[0] # 提取滑块图片链接
8.  
   with open('slice.png', 'wb')as f1:
9.  
   f1.write(requests.get(sc_url).content)
10.  
    for element in elements_2:
11.  
    bg = await page.evaluate('(element) => element.textContent', element)
12.  
    bg_url = bg.split('"')[1].split('"')[0] # 提取背景图片链接
13.  
    with open('bg.png', 'wb') as f2:
14.  
    f2.write(requests.get(bg_url).content)

四：获取拼图坐标偏移量

1.  
   async def get_xy():
2.  
   det = ddddocr.DdddOcr(det=False, ocr=False)
3.  
    
4.  
   with open('slice.png', 'rb') as f:
5.  
   target_bytes = f.read()
6.  
    
7.  
   with open('bg.png', 'rb') as f:
8.  
   background_bytes = f.read()
9.  
   try:
10.  
    res = det.slide_match(target_bytes, background_bytes)
11.  
    print(res)
12.  
    return res.get('target')[0]
13.  
    except:
14.  
    return False

五：操作鼠标移动滑块

        其中x，y需要根据自己电脑进行微调。

1.  
   if target:
2.  
   # print(target)
3.  
   botton3 = await page.xpath(
4.  
   '//*[@id="captcha"]/div[2]/div[1]/div[4]/div[1]/div[2]/div/div/div[2]/div/div[3]')
5.  
   await botton3[0].hover() # 鼠标悬停元素上
6.  
   await page.mouse.down() # 鼠标落下
7.  
   await page.waitFor(500)
8.  
   x = 1116 target
9.  
   y = 641
10.  
    await page.mouse.move(x, y, {'steps': 2}) # 鼠标移动
11.  
    await page.waitFor(500)
12.  
    await page.mouse.up() # 鼠标松开
13.  
    time.sleep(2)
14.  
    elements_3 = await page.xpath('//*[@id="captcha"]/div[2]/div[1]/div[3]/div[2]/div/div[2]/text()')
15.  
    msg = ''
16.  
    for element in elements_3:
17.  
    msg = await page.evaluate('(element) => element.textContent', element)
18.  
    if msg == '验证通过':
19.  
    break
20.  
    else:
21.  
    print(msg)
22.  
    else: # 获取坐标失败时刷新验证
23.  
    botton4 = await page.xpath('//*[@aria-label="刷新验证"]')
24.  
    await botton4[0].click()

验证通过，欧耶~

六：完整代码

1.  
   # coding:utf-8
2.  
   import ddddocr
3.  
   # coding:utf-8
4.  
   import asyncio
5.  
   import time
6.  
    
7.  
   import requests
8.  
   from pyppeteer.launcher import DEFAULT_ARGS
9.  
    
10.  
    DEFAULT_ARGS.remove("--enable-automation")
11.  
    from pyppeteer import launch
12.  
     
13.  
     
14.  
    async def main():
15.  
    # 浏览器 启动参数
16.  
    start_parm = {
17.  
    # 关闭无头浏览器
18.  
    "headless": False,
19.  
    "args": [
20.  
    '--disable-infobars', # 关闭自动化提示框
21.  
    '--no-sandbox', # 关闭沙盒模式
22.  
    '--start-maximized', # 窗口最大化模式
23.  
     
24.  
    ],
25.  
    }
26.  
    browser = await launch(**start_parm)
27.  
    page = await browser.newPage()
28.  
     
29.  
    # 设置网页 视图大小
30.  
    await page.setViewport(viewport={'width': 1920, 'height': 1080})
31.  
    await page.goto('https://www.geetest.com/adaptive-captcha-demo')
32.  
    time.sleep(2)
33.  
    await page.waitForXPath('//div[@class="type-config"]') # 等待元素加载
34.  
    botton1 = await page.xpath('//div[@class="tab-item tab-item-1"]') # 滑块拼图验证按钮
35.  
    await botton1[0].click()
36.  
    await page.click('#captcha', options={
37.  
    'button': 'left',
38.  
    'clickCount': 2,
39.  
    'delay': 300, # 延迟点击(ms)
40.  
    })
41.  
    botton2 = await page.xpath('//*[@aria-label="点击按钮开始验证"]') # 开始验证按钮
42.  
    await botton2[0].click()
43.  
    while True:
44.  
    time.sleep(5)
45.  
    elements_1 = await page.xpath(
46.  
    '//*[@id="captcha"]/div[2]/div[1]/div[4]/div[1]/div[2]/div/div/div[1]/div[1]/div[1]/@style') # 滑块图片链接
47.  
    elements_2 = await page.xpath(
48.  
    '//*[@id="captcha"]/div[2]/div[1]/div[4]/div[1]/div[2]/div/div/div[1]/div[2]/@style') # 背景图片链接
49.  
    for element in elements_1:
50.  
    sc = await page.evaluate('(element) => element.textContent', element)
51.  
    sc_url = sc.split('"')[1].split('"')[0] # 提取滑块图片链接
52.  
    with open('slice.png', 'wb')as f1:
53.  
    f1.write(requests.get(sc_url).content)
54.  
    for element in elements_2:
55.  
    bg = await page.evaluate('(element) => element.textContent', element)
56.  
    bg_url = bg.split('"')[1].split('"')[0] # 提取背景图片链接
57.  
    with open('bg.png', 'wb') as f2:
58.  
    f2.write(requests.get(bg_url).content)
59.  
    target = await get_xy() # 得到滑块x坐标偏移量
60.  
    if target:
61.  
    # print(target)
62.  
    botton3 = await page.xpath(
63.  
    '//*[@id="captcha"]/div[2]/div[1]/div[4]/div[1]/div[2]/div/div/div[2]/div/div[3]')
64.  
    await botton3[0].hover() # 鼠标悬停元素上
65.  
    await page.mouse.down() # 鼠标落下
66.  
    await page.waitFor(500)
67.  
    x = 1116 target
68.  
    y = 641
69.  
    await page.mouse.move(x, y, {'steps': 2}) # 鼠标移动
70.  
    await page.waitFor(500)
71.  
    await page.mouse.up() # 鼠标松开
72.  
    time.sleep(2)
73.  
    elements_3 = await page.xpath('//*[@id="captcha"]/div[2]/div[1]/div[3]/div[2]/div/div[2]/text()')
74.  
    msg = ''
75.  
    for element in elements_3:
76.  
    msg = await page.evaluate('(element) => element.textContent', element)
77.  
    if msg == '验证通过':
78.  
    break
79.  
    else:
80.  
    print(msg)
81.  
    else: # 获取坐标失败时刷新验证
82.  
    botton4 = await page.xpath('//*[@aria-label="刷新验证"]')
83.  
    await botton4[0].click()
84.  
    input('---验证通过---')
85.  
    await browser.close()
86.  
     
87.  
     
88.  
    async def get_xy():
89.  
    det = ddddocr.DdddOcr(det=False, ocr=False)
90.  
     
91.  
    with open('slice.png', 'rb') as f:
92.  
    target_bytes = f.read()
93.  
     
94.  
    with open('bg.png', 'rb') as f:
95.  
    background_bytes = f.read()
96.  
    try:
97.  
    res = det.slide_match(target_bytes, background_bytes)
98.  
    print(res)
99.  
    return res.get('target')[0]
100.  
     except:
101.  
     return False
102.  
      
103.  
      
104.  
     if __name__ == '__main__':
105.  
     main()
106.  
     asyncio.get_event_loop().run_until_complete(main())

这篇好文章是转载于：学新通技术网