yt_dlp/extractor/weibo.py

   1 from .common import InfoExtractor
   2
   3 import json
   4 import random
   5 import re
   6
   7 from ..compat import (
   8     compat_parse_qs,
   9     compat_str,
  10 )
  11 from ..utils import (
  12     js_to_json,
  13     strip_jsonp,
  14     urlencode_postdata,
  15 )
  16
  17
  18 class WeiboIE(InfoExtractor):
  19     _VALID_URL = r'https?://(?:www\.)?weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'
  20     _TEST = {
  21         'url': 'https://weibo.com/6275294458/Fp6RGfbff?type=comment',
  22         'info_dict': {
  23             'id': 'Fp6RGfbff',
  24             'ext': 'mp4',
  25             'title': 'You should have servants to massage you,... 来自Hosico_猫 - 微博',
  26         }
  27     }
  28
  29     def _real_extract(self, url):
  30         video_id = self._match_id(url)
  31         # to get Referer url for genvisitor
  32         webpage, urlh = self._download_webpage_handle(url, video_id)
  33
  34         visitor_url = urlh.url
  35
  36         if 'passport.weibo.com' in visitor_url:
  37             # first visit
  38             visitor_data = self._download_json(
  39                 'https://passport.weibo.com/visitor/genvisitor', video_id,
  40                 note='Generating first-visit data',
  41                 transform_source=strip_jsonp,
  42                 headers={'Referer': visitor_url},
  43                 data=urlencode_postdata({
  44                     'cb': 'gen_callback',
  45                     'fp': json.dumps({
  46                         'os': '2',
  47                         'browser': 'Gecko57,0,0,0',
  48                         'fonts': 'undefined',
  49                         'screenInfo': '1440*900*24',
  50                         'plugins': '',
  51                     }),
  52                 }))
  53
  54             tid = visitor_data['data']['tid']
  55             cnfd = '%03d' % visitor_data['data']['confidence']
  56
  57             self._download_webpage(
  58                 'https://passport.weibo.com/visitor/visitor', video_id,
  59                 note='Running first-visit callback',
  60                 query={
  61                     'a': 'incarnate',
  62                     't': tid,
  63                     'w': 2,
  64                     'c': cnfd,
  65                     'cb': 'cross_domain',
  66                     'from': 'weibo',
  67                     '_rand': random.random(),
  68                 })
  69
  70             webpage = self._download_webpage(
  71                 url, video_id, note='Revisiting webpage')
  72
  73         title = self._html_extract_title(webpage)
  74
  75         video_formats = compat_parse_qs(self._search_regex(
  76             r'video-sources=\\\"(.+?)\"', webpage, 'video_sources'))
  77
  78         formats = []
  79         supported_resolutions = (480, 720)
  80         for res in supported_resolutions:
  81             vid_urls = video_formats.get(compat_str(res))
  82             if not vid_urls or not isinstance(vid_urls, list):
  83                 continue
  84
  85             vid_url = vid_urls[0]
  86             formats.append({
  87                 'url': vid_url,
  88                 'height': res,
  89             })
  90
  91         uploader = self._og_search_property(
  92             'nick-name', webpage, 'uploader', default=None)
  93
  94         return {
  95             'id': video_id,
  96             'title': title,
  97             'uploader': uploader,
  98             'formats': formats
  99         }
 100
 101
 102 class WeiboMobileIE(InfoExtractor):
 103     _VALID_URL = r'https?://m\.weibo\.cn/status/(?P<id>[0-9]+)(\?.+)?'
 104     _TEST = {
 105         'url': 'https://m.weibo.cn/status/4189191225395228?wm=3333_2001&sourcetype=weixin&featurecode=newtitle&from=singlemessage&isappinstalled=0',
 106         'info_dict': {
 107             'id': '4189191225395228',
 108             'ext': 'mp4',
 109             'title': '午睡当然是要甜甜蜜蜜的啦',
 110             'uploader': '柴犬柴犬'
 111         }
 112     }
 113
 114     def _real_extract(self, url):
 115         video_id = self._match_id(url)
 116         # to get Referer url for genvisitor
 117         webpage = self._download_webpage(url, video_id, note='visit the page')
 118
 119         weibo_info = self._parse_json(self._search_regex(
 120             r'var\s+\$render_data\s*=\s*\[({.*})\]\[0\]\s*\|\|\s*{};',
 121             webpage, 'js_code', flags=re.DOTALL),
 122             video_id, transform_source=js_to_json)
 123
 124         status_data = weibo_info.get('status', {})
 125         page_info = status_data.get('page_info')
 126         title = status_data['status_title']
 127         uploader = status_data.get('user', {}).get('screen_name')
 128
 129         return {
 130             'id': video_id,
 131             'title': title,
 132             'uploader': uploader,
 133             'url': page_info['media_info']['stream_url']
 134         }