youtube_dl/extractor/wistia.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     ExtractorError,
   8     float_or_none,
   9     int_or_none,
  10     try_get,
  11     unescapeHTML,
  12 )
  13
  14
  15 class WistiaBaseIE(InfoExtractor):
  16     _VALID_ID_REGEX = r'(?P<id>[a-z0-9]{10})'
  17     _VALID_URL_BASE = r'https?://(?:fast\.)?wistia\.(?:net|com)/embed/'
  18     _EMBED_BASE_URL = 'http://fast.wistia.com/embed/'
  19
  20     def _download_embed_config(self, config_type, config_id, referer):
  21         base_url = self._EMBED_BASE_URL + '%ss/%s' % (config_type, config_id)
  22         embed_config = self._download_json(
  23             base_url + '.json', config_id, headers={
  24                 'Referer': referer if referer.startswith('http') else base_url,  # Some videos require this.
  25             })
  26
  27         if isinstance(embed_config, dict) and embed_config.get('error'):
  28             raise ExtractorError(
  29                 'Error while getting the playlist', expected=True)
  30
  31         return embed_config
  32
  33     def _extract_media(self, embed_config):
  34         data = embed_config['media']
  35         video_id = data['hashedId']
  36         title = data['name']
  37
  38         formats = []
  39         thumbnails = []
  40         for a in data['assets']:
  41             aurl = a.get('url')
  42             if not aurl:
  43                 continue
  44             astatus = a.get('status')
  45             atype = a.get('type')
  46             if (astatus is not None and astatus != 2) or atype in ('preview', 'storyboard'):
  47                 continue
  48             elif atype in ('still', 'still_image'):
  49                 thumbnails.append({
  50                     'url': aurl,
  51                     'width': int_or_none(a.get('width')),
  52                     'height': int_or_none(a.get('height')),
  53                     'filesize': int_or_none(a.get('size')),
  54                 })
  55             else:
  56                 aext = a.get('ext')
  57                 display_name = a.get('display_name')
  58                 format_id = atype
  59                 if atype and atype.endswith('_video') and display_name:
  60                     format_id = '%s-%s' % (atype[:-6], display_name)
  61                 f = {
  62                     'format_id': format_id,
  63                     'url': aurl,
  64                     'tbr': int_or_none(a.get('bitrate')) or None,
  65                     'preference': 1 if atype == 'original' else None,
  66                 }
  67                 if display_name == 'Audio':
  68                     f.update({
  69                         'vcodec': 'none',
  70                     })
  71                 else:
  72                     f.update({
  73                         'width': int_or_none(a.get('width')),
  74                         'height': int_or_none(a.get('height')),
  75                         'vcodec': a.get('codec'),
  76                     })
  77                 if a.get('container') == 'm3u8' or aext == 'm3u8':
  78                     ts_f = f.copy()
  79                     ts_f.update({
  80                         'ext': 'ts',
  81                         'format_id': f['format_id'].replace('hls-', 'ts-'),
  82                         'url': f['url'].replace('.bin', '.ts'),
  83                     })
  84                     formats.append(ts_f)
  85                     f.update({
  86                         'ext': 'mp4',
  87                         'protocol': 'm3u8_native',
  88                     })
  89                 else:
  90                     f.update({
  91                         'container': a.get('container'),
  92                         'ext': aext,
  93                         'filesize': int_or_none(a.get('size')),
  94                     })
  95                 formats.append(f)
  96
  97         self._sort_formats(formats)
  98
  99         subtitles = {}
 100         for caption in data.get('captions', []):
 101             language = caption.get('language')
 102             if not language:
 103                 continue
 104             subtitles[language] = [{
 105                 'url': self._EMBED_BASE_URL + 'captions/' + video_id + '.vtt?language=' + language,
 106             }]
 107
 108         return {
 109             'id': video_id,
 110             'title': title,
 111             'description': data.get('seoDescription'),
 112             'formats': formats,
 113             'thumbnails': thumbnails,
 114             'duration': float_or_none(data.get('duration')),
 115             'timestamp': int_or_none(data.get('createdAt')),
 116             'subtitles': subtitles,
 117         }
 118
 119
 120 class WistiaIE(WistiaBaseIE):
 121     _VALID_URL = r'(?:wistia:|%s(?:iframe|medias)/)%s' % (WistiaBaseIE._VALID_URL_BASE, WistiaBaseIE._VALID_ID_REGEX)
 122
 123     _TESTS = [{
 124         # with hls video
 125         'url': 'wistia:807fafadvk',
 126         'md5': 'daff0f3687a41d9a71b40e0e8c2610fe',
 127         'info_dict': {
 128             'id': '807fafadvk',
 129             'ext': 'mp4',
 130             'title': 'Drip Brennan Dunn Workshop',
 131             'description': 'a JV Webinars video',
 132             'upload_date': '20160518',
 133             'timestamp': 1463607249,
 134             'duration': 4987.11,
 135         },
 136     }, {
 137         'url': 'wistia:sh7fpupwlt',
 138         'only_matching': True,
 139     }, {
 140         'url': 'http://fast.wistia.net/embed/iframe/sh7fpupwlt',
 141         'only_matching': True,
 142     }, {
 143         'url': 'http://fast.wistia.com/embed/iframe/sh7fpupwlt',
 144         'only_matching': True,
 145     }, {
 146         'url': 'http://fast.wistia.net/embed/medias/sh7fpupwlt.json',
 147         'only_matching': True,
 148     }]
 149
 150     # https://wistia.com/support/embed-and-share/video-on-your-website
 151     @staticmethod
 152     def _extract_url(webpage):
 153         urls = WistiaIE._extract_urls(webpage)
 154         return urls[0] if urls else None
 155
 156     @staticmethod
 157     def _extract_urls(webpage):
 158         urls = []
 159         for match in re.finditer(
 160                 r'<(?:meta[^>]+?content|(?:iframe|script)[^>]+?src)=["\'](?P<url>(?:https?:)?//(?:fast\.)?wistia\.(?:net|com)/embed/(?:iframe|medias)/[a-z0-9]{10})', webpage):
 161             urls.append(unescapeHTML(match.group('url')))
 162         for match in re.finditer(
 163                 r'''(?sx)
 164                     <div[^>]+class=(["'])(?:(?!\1).)*?\bwistia_async_(?P<id>[a-z0-9]{10})\b(?:(?!\1).)*?\1
 165                 ''', webpage):
 166             urls.append('wistia:%s' % match.group('id'))
 167         for match in re.finditer(r'(?:data-wistia-?id=["\']|Wistia\.embed\(["\']|id=["\']wistia_)(?P<id>[a-z0-9]{10})', webpage):
 168             urls.append('wistia:%s' % match.group('id'))
 169         return urls
 170
 171     def _real_extract(self, url):
 172         video_id = self._match_id(url)
 173         embed_config = self._download_embed_config('media', video_id, url)
 174         return self._extract_media(embed_config)
 175
 176
 177 class WistiaPlaylistIE(WistiaBaseIE):
 178     _VALID_URL = r'%splaylists/%s' % (WistiaIE._VALID_URL_BASE, WistiaIE._VALID_ID_REGEX)
 179
 180     _TEST = {
 181         'url': 'https://fast.wistia.net/embed/playlists/aodt9etokc',
 182         'info_dict': {
 183             'id': 'aodt9etokc',
 184         },
 185         'playlist_count': 3,
 186     }
 187
 188     def _real_extract(self, url):
 189         playlist_id = self._match_id(url)
 190         playlist = self._download_embed_config('playlist', playlist_id, url)
 191
 192         entries = []
 193         for media in (try_get(playlist, lambda x: x[0]['medias']) or []):
 194             embed_config = media.get('embed_config')
 195             if not embed_config:
 196                 continue
 197             entries.append(self._extract_media(embed_config))
 198
 199         return self.playlist_result(entries, playlist_id)