yt_dlp/extractor/sohu.py

   1 import re
   2
   3 from .common import InfoExtractor
   4 from ..compat import (
   5     compat_str,
   6     compat_urllib_parse_urlencode,
   7 )
   8 from ..utils import (
   9     ExtractorError,
  10     int_or_none,
  11     try_get,
  12 )
  13
  14
  15 class SohuIE(InfoExtractor):
  16     _VALID_URL = r'https?://(?P<mytv>my\.)?tv\.sohu\.com/.+?/(?(mytv)|n)(?P<id>\d+)\.shtml.*?'
  17
  18     # Sohu videos give different MD5 sums on Travis CI and my machine
  19     _TESTS = [{
  20         'note': 'This video is available only in Mainland China',
  21         'url': 'http://tv.sohu.com/20130724/n382479172.shtml#super',
  22         'info_dict': {
  23             'id': '382479172',
  24             'ext': 'mp4',
  25             'title': 'MV：Far East Movement《The Illest》',
  26         },
  27         'skip': 'On available in China',
  28     }, {
  29         'url': 'http://tv.sohu.com/20150305/n409385080.shtml',
  30         'info_dict': {
  31             'id': '409385080',
  32             'ext': 'mp4',
  33             'title': '《2015湖南卫视羊年元宵晚会》唐嫣《花好月圆》',
  34         }
  35     }, {
  36         'url': 'http://my.tv.sohu.com/us/232799889/78693464.shtml',
  37         'info_dict': {
  38             'id': '78693464',
  39             'ext': 'mp4',
  40             'title': '【爱范品】第31期：MWC见不到的奇葩手机',
  41         }
  42     }, {
  43         'note': 'Multipart video',
  44         'url': 'http://my.tv.sohu.com/pl/8384802/78910339.shtml',
  45         'info_dict': {
  46             'id': '78910339',
  47             'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
  48         },
  49         'playlist': [{
  50             'info_dict': {
  51                 'id': '78910339_part1',
  52                 'ext': 'mp4',
  53                 'duration': 294,
  54                 'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
  55             }
  56         }, {
  57             'info_dict': {
  58                 'id': '78910339_part2',
  59                 'ext': 'mp4',
  60                 'duration': 300,
  61                 'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
  62             }
  63         }, {
  64             'info_dict': {
  65                 'id': '78910339_part3',
  66                 'ext': 'mp4',
  67                 'duration': 150,
  68                 'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
  69             }
  70         }]
  71     }, {
  72         'note': 'Video with title containing dash',
  73         'url': 'http://my.tv.sohu.com/us/249884221/78932792.shtml',
  74         'info_dict': {
  75             'id': '78932792',
  76             'ext': 'mp4',
  77             'title': 'youtube-dl testing video',
  78         },
  79         'params': {
  80             'skip_download': True
  81         }
  82     }]
  83
  84     def _real_extract(self, url):
  85
  86         def _fetch_data(vid_id, mytv=False):
  87             if mytv:
  88                 base_data_url = 'http://my.tv.sohu.com/play/videonew.do?vid='
  89             else:
  90                 base_data_url = 'http://hot.vrs.sohu.com/vrs_flash.action?vid='
  91
  92             return self._download_json(
  93                 base_data_url + vid_id, video_id,
  94                 'Downloading JSON data for %s' % vid_id,
  95                 headers=self.geo_verification_headers())
  96
  97         mobj = self._match_valid_url(url)
  98         video_id = mobj.group('id')
  99         mytv = mobj.group('mytv') is not None
 100
 101         webpage = self._download_webpage(url, video_id)
 102
 103         title = re.sub(r' - 搜狐视频$', '', self._og_search_title(webpage))
 104
 105         vid = self._html_search_regex(
 106             r'var vid ?= ?["\'](\d+)["\']',
 107             webpage, 'video path')
 108         vid_data = _fetch_data(vid, mytv)
 109         if vid_data['play'] != 1:
 110             if vid_data.get('status') == 12:
 111                 raise ExtractorError(
 112                     '%s said: There\'s something wrong in the video.' % self.IE_NAME,
 113                     expected=True)
 114             else:
 115                 self.raise_geo_restricted(
 116                     '%s said: The video is only licensed to users in Mainland China.' % self.IE_NAME)
 117
 118         formats_json = {}
 119         for format_id in ('nor', 'high', 'super', 'ori', 'h2644k', 'h2654k'):
 120             vid_id = vid_data['data'].get('%sVid' % format_id)
 121             if not vid_id:
 122                 continue
 123             vid_id = compat_str(vid_id)
 124             formats_json[format_id] = vid_data if vid == vid_id else _fetch_data(vid_id, mytv)
 125
 126         part_count = vid_data['data']['totalBlocks']
 127
 128         playlist = []
 129         for i in range(part_count):
 130             formats = []
 131             for format_id, format_data in formats_json.items():
 132                 allot = format_data['allot']
 133
 134                 data = format_data['data']
 135                 clips_url = data['clipsURL']
 136                 su = data['su']
 137
 138                 video_url = 'newflv.sohu.ccgslb.net'
 139                 cdnId = None
 140                 retries = 0
 141
 142                 while 'newflv.sohu.ccgslb.net' in video_url:
 143                     params = {
 144                         'prot': 9,
 145                         'file': clips_url[i],
 146                         'new': su[i],
 147                         'prod': 'flash',
 148                         'rb': 1,
 149                     }
 150
 151                     if cdnId is not None:
 152                         params['idc'] = cdnId
 153
 154                     download_note = 'Downloading %s video URL part %d of %d' % (
 155                         format_id, i + 1, part_count)
 156
 157                     if retries > 0:
 158                         download_note += ' (retry #%d)' % retries
 159                     part_info = self._parse_json(self._download_webpage(
 160                         'http://%s/?%s' % (allot, compat_urllib_parse_urlencode(params)),
 161                         video_id, download_note), video_id)
 162
 163                     video_url = part_info['url']
 164                     cdnId = part_info.get('nid')
 165
 166                     retries += 1
 167                     if retries > 5:
 168                         raise ExtractorError('Failed to get video URL')
 169
 170                 formats.append({
 171                     'url': video_url,
 172                     'format_id': format_id,
 173                     'filesize': int_or_none(
 174                         try_get(data, lambda x: x['clipsBytes'][i])),
 175                     'width': int_or_none(data.get('width')),
 176                     'height': int_or_none(data.get('height')),
 177                     'fps': int_or_none(data.get('fps')),
 178                 })
 179
 180             playlist.append({
 181                 'id': '%s_part%d' % (video_id, i + 1),
 182                 'title': title,
 183                 'duration': vid_data['data']['clipsDuration'][i],
 184                 'formats': formats,
 185             })
 186
 187         if len(playlist) == 1:
 188             info = playlist[0]
 189             info['id'] = video_id
 190         else:
 191             info = {
 192                 '_type': 'multi_video',
 193                 'entries': playlist,
 194                 'id': video_id,
 195                 'title': title,
 196             }
 197
 198         return info