钉钉录播抓取视频

爬取钉钉视频

免责声明

此脚本仅供学习参考，切勿违法使用下载他人资源进行售卖，本人不但任何责任!

仓库地址:

• GItee 源码仓库

执行顺序

• poxyM3u8开启代理
• getM3u8url用于获取m3u8文件
• userAgent随机请求头
• downVideo|downVideoThreadTqdm单线程下载和多线程下载，二选一即可

启动顺序：poxyM3u8开启代理 -> getM3u8url获取文件->downVideo遍历文件进行下载

像这样别人给的钉钉链接我想要它的视频, 但是又没有下载按钮,我该怎么办呢?

我想到了用爬虫爬取

方案一

检查了一下网络请求发现它是采用m3u8文件格式保存的，所以找m3u8的文件。找到了

• 对它写代码进行保存:

  with open("4f8122f4-f8fb-43d5-b8c8-7c1c9a4a70f7_normal.m3u8", "r", encoding="utf-8") as f:centen = f.read()
  print(centen)
  pattern = r'([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}\/[\d]+\.ts\?auth_key=[\d\w-]+)'
  matches = re.findall(pattern, centen)
  print(matches)
  #
  urls = []
  for match in matches:url = "https://dtliving-bj.dingtalk.com/live_hp/" + matchurls.append(url)
  #
  print(len(urls))
  # for i in urls:
  #     print(i)for item in tqdm(urls,disable="下载"):response = requests.get(item)with open("E:/a.mp4", "ab", ) as f:f.write(response.content)
  

  下载是下载下来了, 可是我有很多很多集，我自己下载是不是太麻烦了，也累。所以我就分析了一下这个地址

  发现:

  m3u8:https://dtliving-sz.dingtalk.com/live_hp/8618428f-dc2e-419e-bc6b-b93a6ee6b28c_normal.m3u8?auth_key=1730544823-fb9347e4a68a456b8b265afa36700f15-0-f24f0b45c72dd6547dadf77466f68ce4url:https://n.dingtalk.com/dingding/live-room/index.html?roomId=ZxaInSr3io8j9iZf&liveUuid=8618428f-dc2e-419e-bc6b-b93a6ee6b28c8618428f-dc2e-419e-bc6b-b93a6ee6b28c_normal.m3u8,其中8618428f-dc2e-419e-bc6b-b93a6ee6b28c是Uuid

  既然:8618428f-dc2e-419e-bc6b-b93a6ee6b28c是房间号的话那我把好多集的房间号爬下来然后拼接到dtliving-sz.dingtalk.com/live_hp/房间号_normal.m3u8这样不就行了?

  然后拼接好我就发了一个请求发现并不能下载下来

  

  		原因是`auth_key`的原因, 然后我尝试寻找`auth_key`
  

emmm, 找了许久，打扰了。还是能力不够, 所以打算换一个方式。

方案二

我发现浏览器是可以获取到auth_key的那我不如我去拿浏览器的响应值。

相当于做了一件中间人的方式把我想要的东西抓取出来。

我使用了mitmproxy当我的代理

pip install mitmproxy

然后写一段代码来捕捉我想要抓取的url的响应

from mitmproxy import ctx,http
# http://mitm.it/ 证书
# mitmdump.exe -s .\test5.py
# mitmweb
import re
import requestsdef request(flow):# 获取请求对象request = flow.request# 实例化输出类math = re.match("^(.*?)_normal.m3u8", request.url)if math:info = ctx.log.info# 打印请求的urlinfo("请求地址: " + request.url + "\n")string = request.urlstart_index = string.find("auth_key=") + len("auth_key=")end_index = len(string)result = string[start_index:end_index]print(result)info("请求体: " + request.text + "\n")# # 打印请求方法info("请求方法: " + request.method)def response(flow):m3u8math = re.match("^(.*?)_normal.m3u8", flow.request.url)if m3u8math:print("===============这是m3u8格式的文件响应============================")centen = flow.response.get_text()with open("./m3u8s/{0}.m3u8".format(title), "w") as f:f.write(centen)print("===============结束============================")

代码写好了，然后打开本机代理改成mitmproxy的代理然后安装证书，之后就可以愉快的抓请求了

1、代码启动

2、代理设置:

3、证书安装:

• 设置好系统代理后,浏览器输入http://mitm.it/, 然后选择对应系统的证书安装就行。

4、抓取

• 当我使用浏览器打开https://n.dingtalk.com/dingding/live-room/index.html?roomId=AAToXdFAVGArvaQx&liveUuid=9aac3549-698f-46b9-9bb0-f2f44d4faaca的时候它就会帮我把特定m3u8的请求响应做文件保存

from mitmproxy import ctx,http
# http://mitm.it/ 证书
# mitmdump.exe -s .\xiaoyuan.py
# mitmweb
import re
import requests
def response(flow):titlesearch = re.search(r"roomId=(.*?)&liveUuid=(.*)", flow.request.url)if titlesearch:global roomIdAndUidroomIdAndUid = titlesearchcentent = flow.response.get_content().decode('utf-8')titleRe = re.search(r'<meta property="og:title" content="(.*?)">',centent)global titletitle = titleRe.group(1)print(title)else:m3u8math = re.match(r"^(.*)/(.*?)_normal.m3u8", flow.request.url)if m3u8math:print("===============这是m3u8格式的文件响应============================")print("房间号：", roomIdAndUid.group(2), "========", roomIdAndUid.group(1))centen = flow.response.get_text()try:with open("./杰哥数学m3u8/{0}.m3u8".format(title), "w") as f:f.write(centen)except OSError:with open("./log.txt".format(title), "a") as f:f.write("标题: {0}, roomId:{1}, UuId: {2}, url:https://n.dingtalk.com/dingding/live-room/index.html?roomId={3}&liveUuid={4}\n".format(title,roomIdAndUid.group(1),roomIdAndUid.group(2),roomIdAndUid.group(1),roomIdAndUid.group(2),))print("===============结束============================")

可是我有很多个链接

所以我打算使用webdriver帮我做批量的链接请求, 而且这个必须要登录才能播放而webdriver会打断我的登录状态，为了保存我的登录状态所以我直接调试本机的chrome。

1、关闭chrome浏览器

2、终端输入

chrome.exe --remote-debugging-port=9222

3、确认是登录状态后，执行代码

import time
from selenium import webdriver
from selenium import webdriveroptions = webdriver.ChromeOptions()
options.set_headless()
options.add_experimental_option("debuggerAddress", "127.0.0.1:9222")
driver = webdriver.Chrome("chromedriver-win64/chromedriver-win64/chromedriver.exe",chrome_options=options)
driver.get('https://n.dingtalk.com/dingding/live-room/index.html?roomId=AAToXdFAVGArvaQx&liveUuid=9aac3549-698f-46b9-9bb0-f2f44d4faaca')

这段代码一执行马上就把这个m3u8文件下载下来了

接下来执行多个url把他们m3u8都下载下来，我只需要把它们都打开然后进行代理检测到就会帮我们下载m3u8文件

import timefrom selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.common.by import By
from selenium import webdriver
from selenium.webdriver.support import expected_conditions as ECoptions = webdriver.ChromeOptions()
options.set_headless()
options.add_experimental_option("debuggerAddress", "127.0.0.1:9222")
driver = webdriver.Chrome("chromedriver-win64/chromedriver-win64/chromedriver.exe",chrome_options=options)
driver.implicitly_wait(10)def newTable(urls, i):if len(urls) > i:window_handles = driver.window_handles# 切换到新标签页print(window_handles)new_tab = window_handles[-1]driver.switch_to.window(new_tab)driver.get(urls[i])login_btn = WebDriverWait(driver, 10, 0.5).until(EC.visibility_of_element_located((By.ID, "live-room")))if login_btn:time.sleep(5)i += 1print(i)newTable(urls, i)with open("钉钉1.txt", "r", encoding="utf-8") as f:urls = f.readlines()driver.get(urls[0])time.sleep(5)newTable(urls, 1)

然后再对m3u8文件进行遍历下载

import re
import requests
import os
import tqdm
requests.packages.urllib3.disable_warnings()with open("m3u8/af941a57-92ad-487f-a2a1-a4682f07afc4_normal.m3u8", "r", encoding="utf-8") as file:content = file.read()# fileName = os.path.basename(file_path).split(".")[0]
# print(f"文件 {os.path.basename(file_path)} 的内容为：{content}")
pattern = r'([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}\/[\d]+\.ts\?auth_key=[\d\w-]+)'
matches = re.findall(pattern, content)m3u8Url = ["https://dtliving-sz.dingtalk.com/live_hp/", "https://dtliving-sh.dingtalk.com/live_hp/"]def getStatusUrl():for status in m3u8Url:url = status + matches[0]responseStatus = requests.get(url, verify=False)print(status, responseStatus.status_code)if responseStatus.status_code == 200:return status
def getMp4Url():urls = []status = getStatusUrl()for match in matches:url = status+matchurls.append(url)return urlsdef run():urls = getMp4Url()for item in tqdm.tqdm(urls):response = requests.get(item, verify=False)if response.status_code == 200:# with open("/disk/data/杰哥数学/{0}.mp4".format("习题课1"), "ab", ) as f:with open(r"E:\杰哥数学\{0}.mp4".format("习题课1"), "ab", ) as f:f.write(response.content)run()

这样就可以下载文件了

总结

流程分析

由于解密困难,所以采用mitmproxy进行代理实现直接抓取视频需要请求的m3u8格式的文件，然后进行保存

• 启动代理
• 模拟浏览器访问视频地址
• 下载所有m3u8的文件
• 对m3u8文件进行清洗
• 拼装ts片段视频的地址
• 保存视频

完整代码

1、启动代理

poxyM3u8.py

from mitmproxy import ctx,http
# http://mitm.it/ 证书
# mitmdump.exe -s .\xiaoyuan.py
# mitmweb
import re
import requestsdef setM3u8Status():"""1 表示下载好了 """with open("m3u8Status.txt", "w") as f:f.write("0")def response(flow):titlesearch = re.search(r"roomId=(.*?)&liveUuid=(.*)", flow.request.url)if titlesearch:global roomIdAndUidroomIdAndUid = titlesearchcentent = flow.response.get_content().decode('utf-8')titleRe = re.search(r'<meta property="og:title" content="(.*?)">',centent)global titletitle = titleRe.group(1)print(title)else:m3u8math = re.match(r"^(.*)/(.*?)_normal.m3u8", flow.request.url)if m3u8math:print("===============这是m3u8格式的文件响应============================")print("房间号：", roomIdAndUid.group(2), "========", roomIdAndUid.group(1))centen = flow.response.get_text()try:with open(r"./m3u8/{0}.m3u8".format(title.replace("/", "-")).replace("\t", " "), "w") as f:f.write(centen)setM3u8Status()except OSError as e:print("==================================错误====================================")print(e)print("==================================错误====================================")with open("./log.txt".format(title), "a") as f:f.write("标题: {0}, roomId:{1}, UuId: {2}, url:https://n.dingtalk.com/dingding/live-room/index.html?roomId={3}&liveUuid={4}\n".format(title,roomIdAndUid.group(1),roomIdAndUid.group(2),roomIdAndUid.group(1),roomIdAndUid.group(2),))print("===============结束============================")

2、模拟浏览器进行请求

getM3u8.py

import time
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.common.by import By
from selenium import webdriver
from selenium.webdriver.support import expected_conditions as ECoptions = webdriver.ChromeOptions()
options.set_headless()
options.add_experimental_option("debuggerAddress", "127.0.0.1:9222")
driver = webdriver.Chrome("../chromedriver-win64/chromedriver-win64/chromedriver.exe", chrome_options=options)
driver.implicitly_wait(10)def newTable(urls, i):if len(urls) > i:window_handles = driver.window_handles# 切换到新标签页print(window_handles)new_tab = window_handles[-1]driver.switch_to.window(new_tab)setM3u8Status()driver.get(urls[i])while getM3u8FileStatus():time.sleep(5)i += 1print(i)newTable(urls, i)# login_btn = WebDriverWait(driver, 10, 0.5).until(EC.visibility_of_element_located((By.ID, "live-room")))# if login_btn:#     time.sleep(5)#     i += 1#     print(i)#     newTable(urls, i)def getM3u8FileStatus():with open("m3u8Status.txt", "r", encoding="utf-8") as f:status = f.read()time.sleep(2)return "1" == statusdef setM3u8Status():"""0 表示新的请求等待 """with open("m3u8Status.txt", "w") as f:f.write("1")time.sleep(2)with open("钉钉1.txt", "r", encoding="utf-8") as f:urls = f.readlines()driver.get(urls[0])while getM3u8FileStatus():time.sleep(5)newTable(urls, 1)

3、最后下载文件

我发现m3u8里面的ts请求不止一个域名

有两个，用错了域名会报404状态码
m3u8Url = ["https://dtliving-sz.dingtalk.com/live_hp/", "https://dtliving-sh.dingtalk.com/live_hp/"]

userAgent 随机请求头

import random
import stringbrowsers = ["Chrome", "Firefox", "Safari", "Edge", "Opera"]
operating_systems = ["Windows NT", "Macintosh", "Linux", "iPhone", "iPad", "Android"]
versions = [str(i) for i in range(80, 130)]def generate_random_string(length):return ''.join(random.choices(string.ascii_letters + string.digits, k=length))def generate_user_agents(num):user_agents = []for _ in range(num):browser = random.choice(browsers)os = random.choice(operating_systems)version = random.choice(versions)if os == "Windows NT":ua = f"Mozilla/5.0 ({os}; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) {browser}/{version} Safari/537.36"elif os == "Macintosh":ua = f"Mozilla/5.0 ({os}; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) {browser}/{version} Safari/537.36"elif os == "Linux":ua = f"Mozilla/5.0 ({os}; x86_64) AppleWebKit/537.36 (KHTML, like Gecko) {browser}/{version} Safari/537.36"elif os == "iPhone":ua = f"Mozilla/5.0 (iPhone; CPU iPhone OS {generate_random_string(2)}_0 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/{version} Mobile/15E148 Safari/604.1"elif os == "iPad":ua = f"Mozilla/5.0 (iPad; CPU OS {generate_random_string(2)}_0 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/{version} Mobile/15E148 Safari/604.1"elif os == "Android":ua = f"Mozilla/5.0 (Linux; Android {generate_random_string(2)}; {generate_random_string(10)}) AppleWebKit/537.36 (KHTML, like Gecko) {browser}/{version} Mobile Safari/537.36"user_agents.append(ua)return user_agents# print(random.choice(generate_user_agents(100)))

downVideo.py

import os
import random
import time
import requests
import re
from tqdm import tqdm
requests.packages.urllib3.disable_warnings()
import glob
from userAgent import generate_user_agentsheaders = {'User-Agent': random.choice(generate_user_agents(100))}
folder_path = 'm3u8'
# 获取文件列表并添加进度条
file_paths = list(tqdm(glob.glob(folder_path + '/**/*', recursive=True), desc="获取文件列表进度"))m3u8Url = ["https://dtliving-sz.dingtalk.com/live_hp/", "https://dtliving-sh.dingtalk.com/live_hp/", "https://dtliving-bj.dingtalk.com/live_hp/"]def getStatusUrl(fileName, m3u8, i):url = m3u8Url[i] + m3u8print(f"第{i}次,正在检测{url}....")response = requests.get(url, headers=headers, verify=False)if response.status_code == 200:print(f"域名检测成功{m3u8}的域名是{m3u8Url[i]}")return m3u8Url[i]else:i+=1if len(m3u8Url) > i:print(f"第{i},次检测{url}....")getStatusUrl(fileName, m3u8, i)else:print(f"错误")with open("./errorUrl.txt", "a", encoding="utf-8") as f:f.write(f"{m3u8}没有找到合适的域名, 文件名称是: {fileName}")def getMp4Url(matches,fileName, m3u8, i):urls = []status = getStatusUrl(fileName, m3u8, i)for match in matches:url = status+matchurls.append(url)return urlsfor file_path in file_paths:if os.path.isfile(file_path):with open(file_path, 'r', encoding="utf-8") as file:content = file.read()fileName = os.path.basename(file_path).split(".")[0]# print(f"文件 {os.path.basename(file_path)} 的内容为：{content}")pattern = r'([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}\/[\d]+\.ts\?auth_key=[\d\w-]+)'matches = re.findall(pattern, content)urls = getMp4Url(matches, fileName, matches[0], 0)# 处理每个文件中的链接列表并添加进度条for item in tqdm(urls, desc=f"处理 {fileName} 文件内链接进度"):response = requests.get(item, verify=False)time.sleep(2)# with open("/disk/data/杰哥数学/{0}.mp4".format(fileName), "ab", ) as f:with open(r"E:\杰哥数学\{0}.mp4".format(fileName), "ab", ) as f:f.write(response.content)

也可以是多线程这样下载更快

import os
import random
import time
import requests
import re
from tqdm import tqdm
import glob
from userAgent import generate_user_agents
import threadingrequests.packages.urllib3.disable_warnings()# 设置请求头
headers = {'User-Agent': random.choice(generate_user_agents(100))}# 设置文件夹路径
folder_path = 'm3u8'# 获取文件列表并添加进度条
file_paths = list(tqdm(glob.glob(folder_path + '/**/*', recursive=True), desc="获取文件列表进度"))# 定义m3u8的URL列表
m3u8Url = ["https://dtliving-sz.dingtalk.com/live_hp/", "https://dtliving-sh.dingtalk.com/live_hp/", "https://dtliving-bj.dingtalk.com/live_hp/"]def getStatusUrl(fileName, m3u8, i):url = m3u8Url[i] + m3u8print(f"第{i}次,正在检测{url}....")log.write(f"第{i}次,正在检测{url}....")response = requests.get(url, headers=headers, verify=False)if response.status_code == 200:print(f"域名检测成功{m3u8}的域名是{m3u8Url[i]}")log.write(f"域名检测成功{m3u8}的域名是{m3u8Url[i]}")return m3u8Url[i]else:i += 1if len(m3u8Url) > i:print(f"第{i},次检测{url}....")log.write(f"第{i},次检测{url}....")return getStatusUrl(fileName, m3u8, i)else:log.write(f"第{i},{url}域名匹配失败....")with open("./errorUrl.txt", "a", encoding="utf-8") as f:f.write(f"{m3u8}没有找到合适的域名, 文件名称是: {fileName}")def getMp4Url(matches, fileName, m3u8, i):urls = []status = getStatusUrl(fileName, m3u8, i)for match in matches:url = status+matchurls.append(url)return urlsdef process_file(file_path):global loglog = open("log.txt", "a", encoding="utf-8")if os.path.isfile(file_path):with open(file_path, 'r', encoding="utf-8") as file:content = file.read()fileName = os.path.basename(file_path).split(".")[0]pattern = r'([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}\/[\d]+\.ts\?auth_key=[\d\w-]+)'matches = re.findall(pattern, content)urls = getMp4Url(matches, fileName, matches[0], 0)# 处理每个文件中的链接列表并添加进度条for item in tqdm(urls, desc=f"处理 {fileName} 文件内链接进度"):response = requests.get(item, verify=False)time.sleep(2)# with open(r"E:\杰哥数学\{0}.mp4".format(fileName), "ab", ) as f:with open("/disk/data/杰哥数学/{0}.mp4".format(fileName), "ab", ) as f:f.write(response.content)log.close()# 创建线程列表
threads = []# 遍历文件路径列表，为每个文件创建一个线程进行处理
for file_path in file_paths:thread = threading.Thread(target=process_file, args=(file_path,))threads.append(thread)thread.start()# 等待所有线程完成
for thread in threads:thread.join()

效果

注意事项

• 不要忘记开系统代理
• chrome浏览器需要全部关闭才可以chrome.exe --remote-debugging-port=9222这个命令,不然selenium会没反应
• 不要忘记安装证书
  • 要是下载请求失败的话，请注意访问的频率、更换请求头和IP