yt_dlp/extractor/weibo.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5
   6 import json
   7 import random
   8 import re
   9
  10 from ..compat import (
  11     compat_parse_qs,
  12     compat_str,
  13 )
  14 from ..utils import (
  15     js_to_json,
  16     strip_jsonp,
  17     urlencode_postdata,
  18 )
  19
  20
  21 class WeiboIE(InfoExtractor):
  22     _VALID_URL = r'https?://(?:www\.)?weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'
  23     _TEST = {
  24         'url': 'https://weibo.com/6275294458/Fp6RGfbff?type=comment',
  25         'info_dict': {
  26             'id': 'Fp6RGfbff',
  27             'ext': 'mp4',
  28             'title': 'You should have servants to massage you,... 来自Hosico_猫 - 微博',
  29         }
  30     }
  31
  32     def _real_extract(self, url):
  33         video_id = self._match_id(url)
  34         # to get Referer url for genvisitor
  35         webpage, urlh = self._download_webpage_handle(url, video_id)
  36
  37         visitor_url = urlh.geturl()
  38
  39         if 'passport.weibo.com' in visitor_url:
  40             # first visit
  41             visitor_data = self._download_json(
  42                 'https://passport.weibo.com/visitor/genvisitor', video_id,
  43                 note='Generating first-visit data',
  44                 transform_source=strip_jsonp,
  45                 headers={'Referer': visitor_url},
  46                 data=urlencode_postdata({
  47                     'cb': 'gen_callback',
  48                     'fp': json.dumps({
  49                         'os': '2',
  50                         'browser': 'Gecko57,0,0,0',
  51                         'fonts': 'undefined',
  52                         'screenInfo': '1440*900*24',
  53                         'plugins': '',
  54                     }),
  55                 }))
  56
  57             tid = visitor_data['data']['tid']
  58             cnfd = '%03d' % visitor_data['data']['confidence']
  59
  60             self._download_webpage(
  61                 'https://passport.weibo.com/visitor/visitor', video_id,
  62                 note='Running first-visit callback',
  63                 query={
  64                     'a': 'incarnate',
  65                     't': tid,
  66                     'w': 2,
  67                     'c': cnfd,
  68                     'cb': 'cross_domain',
  69                     'from': 'weibo',
  70                     '_rand': random.random(),
  71                 })
  72
  73             webpage = self._download_webpage(
  74                 url, video_id, note='Revisiting webpage')
  75
  76         title = self._html_extract_title(webpage)
  77
  78         video_formats = compat_parse_qs(self._search_regex(
  79             r'video-sources=\\\"(.+?)\"', webpage, 'video_sources'))
  80
  81         formats = []
  82         supported_resolutions = (480, 720)
  83         for res in supported_resolutions:
  84             vid_urls = video_formats.get(compat_str(res))
  85             if not vid_urls or not isinstance(vid_urls, list):
  86                 continue
  87
  88             vid_url = vid_urls[0]
  89             formats.append({
  90                 'url': vid_url,
  91                 'height': res,
  92             })
  93
  94         self._sort_formats(formats)
  95
  96         uploader = self._og_search_property(
  97             'nick-name', webpage, 'uploader', default=None)
  98
  99         return {
 100             'id': video_id,
 101             'title': title,
 102             'uploader': uploader,
 103             'formats': formats
 104         }
 105
 106
 107 class WeiboMobileIE(InfoExtractor):
 108     _VALID_URL = r'https?://m\.weibo\.cn/status/(?P<id>[0-9]+)(\?.+)?'
 109     _TEST = {
 110         'url': 'https://m.weibo.cn/status/4189191225395228?wm=3333_2001&sourcetype=weixin&featurecode=newtitle&from=singlemessage&isappinstalled=0',
 111         'info_dict': {
 112             'id': '4189191225395228',
 113             'ext': 'mp4',
 114             'title': '午睡当然是要甜甜蜜蜜的啦',
 115             'uploader': '柴犬柴犬'
 116         }
 117     }
 118
 119     def _real_extract(self, url):
 120         video_id = self._match_id(url)
 121         # to get Referer url for genvisitor
 122         webpage = self._download_webpage(url, video_id, note='visit the page')
 123
 124         weibo_info = self._parse_json(self._search_regex(
 125             r'var\s+\$render_data\s*=\s*\[({.*})\]\[0\]\s*\|\|\s*{};',
 126             webpage, 'js_code', flags=re.DOTALL),
 127             video_id, transform_source=js_to_json)
 128
 129         status_data = weibo_info.get('status', {})
 130         page_info = status_data.get('page_info')
 131         title = status_data['status_title']
 132         uploader = status_data.get('user', {}).get('screen_name')
 133
 134         return {
 135             'id': video_id,
 136             'title': title,
 137             'uploader': uploader,
 138             'url': page_info['media_info']['stream_url']
 139         }