抖音视频下载爬虫编写的完整过程!爬取app的过程

原创
2017/10/26 16:43
阅读数 8.1K

工具:

charles(类似fiddler) chrome debug调试台 jupyter notebook

分析APP情况

准备好charles的手机ssl证书,设置好代理启动charles开始抓取抖音app数据。经过分析可以看到如下结果 输入图片说明

可以获得两个get请求!

https://aweme.snssdk.com/aweme/v1/discover/search/?iid=15735436175&device_id=37063549497&os_api=18&app_name=aweme&channel=App%20Store&idfa=08621BB7-65C3-454D-908A-D02F565D85F1&device_platform=iphone&build_number=15805&vid=6BD753D7-C89A-4BEF-9C3D-7192E26CF330&openudid=ee5f41b63ff4704166b2f2d8920267fcd109136b&device_type=iPhone6,2&app_version=1.5.8&version_code=1.5.8&os_version=11.0.3&screen_width=640&aid=1128&ac=WIFI&count=10&cursor=0&keyword=关键词&type=1&cp=9646915915dce7fbe1&as=a115d95e314449fffd&ts=时间戳

https://aweme.snssdk.com/aweme/v1/aweme/post/?iid=15735436175&device_id=37063549497&os_api=18&app_name=aweme&channel=App%20Store&idfa=08621BB7-65C3-454D-908A-D02F565D85F1&device_platform=iphone&build_number=15805&vid=6BD753D7-C89A-4BEF-9C3D-7192E26CF330&openudid=ee5f41b63ff4704166b2f2d8920267fcd109136b&device_type=iPhone6,2&app_version=1.5.8&version_code=1.5.8&os_version=11.0.3&screen_width=640&aid=1128&ac=WIFI&count=视频数量&max_cursor=0&min_cursor=0&user_id=用户ID&cp=e7329b5ccceae5cbe1&as=a125ee5e8cd3393c4e&ts=时间戳
#以上链接进行了处理以便不能直接访问。

大体逻辑是通过搜索获得用户名列表,在判断用户名列表是不是想要的那位达人,获取他的user_id、count,然后通过第二链接获得用户的分享链接,再通过分享链接页面爬取视频。

编写代码

class DouYin(object):
    def __init__(self):
        print(sys.getdefaultencoding(),"抖音下载!")
        #urllib3连接错误时抛出exceptions.SSLError 
        http = urllib3.PoolManager(cert_reqs = 'CERT_REQUIRED', ca_certs = certifi.where())
    
    #获得总分享视频url地址,返回视频名字,视频数,分享视频链接列表
    def get_video_url(self,nickname):
        video_name = []
        video_urls = []
        user_agent = 'Aweme/1.5.8 (iPhone; iOS 11.0.3; Scale/2.00)'
        header = {'User-Agent':user_agent}
        search_url = "https://aweme.snssdk.com/aweme/v1/discover/search/?iid=15735436175&device_id=37063549497&os_api=18&app_name=aweme&channel=App%20Store&idfa=08621BB7-65C3-454D-908A-D02F565D85F1&device_platform=iphone&build_number=15805&vid=6BD753D7-C89A-4BEF-9C3D-7192E26CF330&openudid=ee5f41b63ff4704166b2f2d8920267fcd109136b&device_type=iPhone6,2&app_version=1.5.8&version_code=1.5.8&os_version=11.0.3&screen_width=640&aid=1128&ac=WIFI&count=10&cursor=0&keyword={0}&type=1&cp=9646915915dce7fbe1&as=a115d95e314449fffd&ts={1}".format(quote(nickname),int(time.time()))   
        print(search_url)
        req = http.request('GET',search_url)
        html = json.loads(req.data.decode('utf-8'))
        for each in html['user_list']:
            if each['user_info']['nickname'] == nickname:
                aweme_count = each['user_info']['aweme_count']
                user_id = each['user_info']['uid']
        user_url = 'https://aweme.snssdk.com/aweme/v1/aweme/post/?iid=15735436175&device_id=37063549497&os_api=18&app_name=aweme&channel=App%20Store&idfa=08621BB7-65C3-454D-908A-D02F565D85F1&device_platform=iphone&build_number=15805&vid=6BD753D7-C89A-4BEF-9C3D-7192E26CF330&openudid=ee5f41b63ff4704166b2f2d8920267fcd109136b&device_type=iPhone6,2&app_version=1.5.8&version_code=1.5.8&os_version=11.0.3&screen_width=640&aid=1128&ac=WIFI&count={0}&max_cursor=0&min_cursor=0&user_id={1}&cp=e7329b5ccceae5cbe1&as=a125ee5e8cd3393c4e&ts={2}'.format(aweme_count,user_id,int(time.time()))
        r = http.request('GET',user_url,headers = header)
        htm = json.loads(r.data.decode('utf-8'))
        for each in htm['aweme_list']:
            share_desc = each['share_info']['share_desc']
            if '抖音-原创音乐短视频社区'  == share_desc:
                print(each)
                video_names.append(each[cha_list][0]['cha_name']+'.mp4')
            else:
                video_names.append(share_desc+'.mp4')
            video_url.append(each['share_info']['share_url'])
        return video_names,video_urls,aweme_count
    
    #获得下载视频地址
    def get_download_url(self,video_url):
        #video_url = 'https://www.douyin.com/share/video/6479379598709624078/?region=CN&mid=6477398356422789901'
        req = http.request('GET',video_url).data.decode('utf-8')
        video_url_data = re.findall('var data = \[(.*?)\];', str(req))[0]
        video_html = json.loads(video_url_data)
        download_url = video_html['video']['real_play_addr']
        return download_url
       
     #下载视频   
    def vedio_download(self,video_name,video_url):
        size = 0
        with closing(http.request('GET',video_url,preload_content=False)) as response:
            chunk_size = 1024
            content_size =int(response.headers['content-length'])
            if response.status == 200:
                print('[文件大小]:%0.2f MB' % (content_size / chunk_size / 1024))
                with open(filename,'wb') as file:
                    for data in response.stream(chunk_size):
                        file.write(data)
                        size +=len(data)
                        file.flush()

                    sys.stdout.flush()
                response.release_conn()
    def run(self,nickname):
        vedio_names, video_urls, aweme_count = self.get_video_url(nickname)
        if nickname not in os.listdir():
            os.mkdir(nickname)
        print("视频下载中:\n")
        for num in range(aweme_count):
            video_url = self.get_download_url(video_url[num])
            self.video_download(video_names[num],video_url)
            
if __name__ == '__main__':
    douyin =DouYin()
    nickname = r'王同学'
    douyin.run(nickname)

浏览器老是崩溃,更多参考详情我的博客吧https://www.nigaea.com/dataanalysis/170.html

展开阅读全文
打赏
0
1 收藏
分享
加载中
更多评论
打赏
0 评论
1 收藏
0
分享
返回顶部
顶部