youtube_dl/extractor/voxmedia.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5 from .once import OnceIE
   6 from ..compat import compat_urllib_parse_unquote
   7 from ..utils import (
   8     ExtractorError,
   9     int_or_none,
  10 )
  11
  12
  13 class VoxMediaVolumeIE(OnceIE):
  14     _VALID_URL = r'https?://volume\.vox-cdn\.com/embed/(?P<id>[0-9a-f]{9})'
  15
  16     def _real_extract(self, url):
  17         video_id = self._match_id(url)
  18         webpage = self._download_webpage(url, video_id)
  19
  20         setup = self._parse_json(self._search_regex(
  21             r'setup\s*=\s*({.+});', webpage, 'setup'), video_id)
  22         video_data = setup.get('video') or {}
  23         info = {
  24             'id': video_id,
  25             'title': video_data.get('title_short'),
  26             'description': video_data.get('description_long') or video_data.get('description_short'),
  27             'thumbnail': video_data.get('brightcove_thumbnail')
  28         }
  29         asset = setup.get('asset') or setup.get('params') or {}
  30
  31         formats = []
  32         hls_url = asset.get('hls_url')
  33         if hls_url:
  34             formats.extend(self._extract_m3u8_formats(
  35                 hls_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
  36         mp4_url = asset.get('mp4_url')
  37         if mp4_url:
  38             tbr = self._search_regex(r'-(\d+)k\.', mp4_url, 'bitrate', default=None)
  39             format_id = 'http'
  40             if tbr:
  41                 format_id += '-' + tbr
  42             formats.append({
  43                 'format_id': format_id,
  44                 'url': mp4_url,
  45                 'tbr': int_or_none(tbr),
  46             })
  47         if formats:
  48             self._sort_formats(formats)
  49             info['formats'] = formats
  50             return info
  51
  52         for provider_video_type in ('ooyala', 'youtube', 'brightcove'):
  53             provider_video_id = video_data.get('%s_id' % provider_video_type)
  54             if not provider_video_id:
  55                 continue
  56             if provider_video_type == 'brightcove':
  57                 info['formats'] = self._extract_once_formats(provider_video_id)
  58                 self._sort_formats(info['formats'])
  59             else:
  60                 info.update({
  61                     '_type': 'url_transparent',
  62                     'url': provider_video_id if provider_video_type == 'youtube' else '%s:%s' % (provider_video_type, provider_video_id),
  63                     'ie_key': provider_video_type.capitalize(),
  64                 })
  65             return info
  66         raise ExtractorError('Unable to find provider video id')
  67
  68
  69 class VoxMediaIE(InfoExtractor):
  70     _VALID_URL = r'https?://(?:www\.)?(?:(?:theverge|vox|sbnation|eater|polygon|curbed|racked|funnyordie)\.com|recode\.net)/(?:[^/]+/)*(?P<id>[^/?]+)'
  71     _TESTS = [{
  72         # Volume embed, Youtube
  73         'url': 'http://www.theverge.com/2014/6/27/5849272/material-world-how-google-discovered-what-software-is-made-of',
  74         'info_dict': {
  75             'id': 'j4mLW6x17VM',
  76             'ext': 'mp4',
  77             'title': 'Material world: how Google discovered what software is made of',
  78             'description': 'md5:dfc17e7715e3b542d66e33a109861382',
  79             'upload_date': '20190710',
  80             'uploader_id': 'TheVerge',
  81             'uploader': 'The Verge',
  82         },
  83         'add_ie': ['Youtube'],
  84     }, {
  85         # Volume embed, Youtube
  86         'url': 'http://www.theverge.com/2014/10/21/7025853/google-nexus-6-hands-on-photos-video-android-phablet',
  87         'md5': '4c8f4a0937752b437c3ebc0ed24802b5',
  88         'info_dict': {
  89             'id': 'Gy8Md3Eky38',
  90             'ext': 'mp4',
  91             'title': 'The Nexus 6: hands-on with Google\'s phablet',
  92             'description': 'md5:d9f0216e5fb932dd2033d6db37ac3f1d',
  93             'uploader_id': 'TheVerge',
  94             'upload_date': '20141021',
  95             'uploader': 'The Verge',
  96         },
  97         'add_ie': ['Youtube'],
  98         'skip': 'similar to the previous test',
  99     }, {
 100         # Volume embed, Youtube
 101         'url': 'http://www.vox.com/2016/3/31/11336640/mississippi-lgbt-religious-freedom-bill',
 102         'info_dict': {
 103             'id': 'YCjDnX-Xzhg',
 104             'ext': 'mp4',
 105             'title': "Mississippi's laws are so bad that its anti-LGBTQ law isn't needed to allow discrimination",
 106             'description': 'md5:fc1317922057de31cd74bce91eb1c66c',
 107             'uploader_id': 'voxdotcom',
 108             'upload_date': '20150915',
 109             'uploader': 'Vox',
 110         },
 111         'add_ie': ['Youtube'],
 112         'skip': 'similar to the previous test',
 113     }, {
 114         # youtube embed
 115         'url': 'http://www.vox.com/2016/3/24/11291692/robot-dance',
 116         'md5': '83b3080489fb103941e549352d3e0977',
 117         'info_dict': {
 118             'id': 'FcNHTJU1ufM',
 119             'ext': 'mp4',
 120             'title': 'How "the robot" became the greatest novelty dance of all time',
 121             'description': 'md5:b081c0d588b8b2085870cda55e6da176',
 122             'upload_date': '20160324',
 123             'uploader_id': 'voxdotcom',
 124             'uploader': 'Vox',
 125         },
 126         'add_ie': ['Youtube'],
 127         'skip': 'Page no longer contain videos',
 128     }, {
 129         # SBN.VideoLinkset.entryGroup multiple ooyala embeds
 130         'url': 'http://www.sbnation.com/college-football-recruiting/2015/2/3/7970291/national-signing-day-rationalizations-itll-be-ok-itll-be-ok',
 131         'info_dict': {
 132             'id': 'national-signing-day-rationalizations-itll-be-ok-itll-be-ok',
 133             'title': '25 lies you will tell yourself on National Signing Day',
 134             'description': 'It\'s the most self-delusional time of the year, and everyone\'s gonna tell the same lies together!',
 135         },
 136         'playlist': [{
 137             'md5': '721fededf2ab74ae4176c8c8cbfe092e',
 138             'info_dict': {
 139                 'id': 'p3cThlMjE61VDi_SD9JlIteSNPWVDBB9',
 140                 'ext': 'mp4',
 141                 'title': 'Buddy Hield vs Steph Curry (and the world)',
 142                 'description': 'Let’s dissect only the most important Final Four storylines.',
 143             },
 144         }, {
 145             'md5': 'bf0c5cc115636af028be1bab79217ea9',
 146             'info_dict': {
 147                 'id': 'BmbmVjMjE6esPHxdALGubTrouQ0jYLHj',
 148                 'ext': 'mp4',
 149                 'title': 'Chasing Cinderella 2016: Syracuse basketball',
 150                 'description': 'md5:e02d56b026d51aa32c010676765a690d',
 151             },
 152         }],
 153         'skip': 'Page no longer contain videos',
 154     }, {
 155         # volume embed, Brightcove Once
 156         'url': 'https://www.recode.net/2014/6/17/11628066/post-post-pc-ceo-the-full-code-conference-video-of-microsofts-satya',
 157         'md5': '2dbc77b8b0bff1894c2fce16eded637d',
 158         'info_dict': {
 159             'id': '1231c973d',
 160             'ext': 'mp4',
 161             'title': 'Post-Post-PC CEO: The Full Code Conference Video of Microsoft\'s Satya Nadella',
 162             'description': 'The longtime veteran was chosen earlier this year as the software giant\'s third leader in its history.',
 163         },
 164         'add_ie': ['VoxMediaVolume'],
 165     }]
 166
 167     def _real_extract(self, url):
 168         display_id = self._match_id(url)
 169         webpage = compat_urllib_parse_unquote(self._download_webpage(url, display_id))
 170
 171         def create_entry(provider_video_id, provider_video_type, title=None, description=None):
 172             video_url = {
 173                 'youtube': '%s',
 174                 'ooyala': 'ooyala:%s',
 175                 'volume': 'http://volume.vox-cdn.com/embed/%s',
 176             }[provider_video_type] % provider_video_id
 177             return {
 178                 '_type': 'url_transparent',
 179                 'url': video_url,
 180                 'title': title or self._og_search_title(webpage),
 181                 'description': description or self._og_search_description(webpage),
 182             }
 183
 184         entries = []
 185         entries_data = self._search_regex([
 186             r'Chorus\.VideoContext\.addVideo\((\[{.+}\])\);',
 187             r'var\s+entry\s*=\s*({.+});',
 188             r'SBN\.VideoLinkset\.entryGroup\(\s*(\[.+\])',
 189         ], webpage, 'video data', default=None)
 190         if entries_data:
 191             entries_data = self._parse_json(entries_data, display_id)
 192             if isinstance(entries_data, dict):
 193                 entries_data = [entries_data]
 194             for video_data in entries_data:
 195                 provider_video_id = video_data.get('provider_video_id')
 196                 provider_video_type = video_data.get('provider_video_type')
 197                 if provider_video_id and provider_video_type:
 198                     entries.append(create_entry(
 199                         provider_video_id, provider_video_type,
 200                         video_data.get('title'), video_data.get('description')))
 201
 202         provider_video_id = self._search_regex(
 203             r'data-ooyala-id="([^"]+)"', webpage, 'ooyala id', default=None)
 204         if provider_video_id:
 205             entries.append(create_entry(provider_video_id, 'ooyala'))
 206
 207         volume_uuid = self._search_regex(
 208             r'data-volume-uuid="([^"]+)"', webpage, 'volume uuid', default=None)
 209         if volume_uuid:
 210             entries.append(create_entry(volume_uuid, 'volume'))
 211
 212         if len(entries) == 1:
 213             return entries[0]
 214         else:
 215             return self.playlist_result(entries, display_id, self._og_search_title(webpage), self._og_search_description(webpage))