yt_dlp/extractor/voxmedia.py

   1 import urllib.parse
   2
   3 from .common import InfoExtractor
   4 from .once import OnceIE
   5 from ..utils import (
   6     ExtractorError,
   7     int_or_none,
   8     try_get,
   9     unified_timestamp,
  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         player_setup = setup.get('player_setup') or setup
  23         video_data = player_setup.get('video') or {}
  24         formatted_metadata = video_data.get('formatted_metadata') or {}
  25         info = {
  26             'id': video_id,
  27             'title': player_setup.get('title') or video_data.get('title_short'),
  28             'description': video_data.get('description_long') or video_data.get('description_short'),
  29             'thumbnail': formatted_metadata.get('thumbnail') or video_data.get('brightcove_thumbnail'),
  30             'timestamp': unified_timestamp(formatted_metadata.get('video_publish_date')),
  31         }
  32         asset = try_get(setup, lambda x: x['embed_assets']['chorus'], dict) or {}
  33
  34         formats = []
  35         hls_url = asset.get('hls_url')
  36         if hls_url:
  37             formats.extend(self._extract_m3u8_formats(
  38                 hls_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
  39         mp4_url = asset.get('mp4_url')
  40         if mp4_url:
  41             tbr = self._search_regex(r'-(\d+)k\.', mp4_url, 'bitrate', default=None)
  42             format_id = 'http'
  43             if tbr:
  44                 format_id += '-' + tbr
  45             formats.append({
  46                 'format_id': format_id,
  47                 'url': mp4_url,
  48                 'tbr': int_or_none(tbr),
  49             })
  50         if formats:
  51             info['formats'] = formats
  52             info['duration'] = int_or_none(asset.get('duration'))
  53             return info
  54
  55         for provider_video_type in ('youtube', 'brightcove'):
  56             provider_video_id = video_data.get(f'{provider_video_type}_id')
  57             if not provider_video_id:
  58                 continue
  59             if provider_video_type == 'brightcove':
  60                 info['formats'] = self._extract_once_formats(provider_video_id)
  61             else:
  62                 info.update({
  63                     '_type': 'url_transparent',
  64                     'url': provider_video_id if provider_video_type == 'youtube' else f'{provider_video_type}:{provider_video_id}',
  65                     'ie_key': provider_video_type.capitalize(),
  66                 })
  67             return info
  68         raise ExtractorError('Unable to find provider video id')
  69
  70
  71 class VoxMediaIE(InfoExtractor):
  72     _VALID_URL = r'https?://(?:www\.)?(?:(?:theverge|vox|sbnation|eater|polygon|curbed|racked|funnyordie)\.com|recode\.net)/(?:[^/]+/)*(?P<id>[^/?]+)'
  73     _EMBED_REGEX = [r'<iframe[^>]+?src="(?P<url>https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"']
  74     _TESTS = [{
  75         # Volume embed, Youtube
  76         'url': 'http://www.theverge.com/2014/6/27/5849272/material-world-how-google-discovered-what-software-is-made-of',
  77         'info_dict': {
  78             'id': 'j4mLW6x17VM',
  79             'ext': 'mp4',
  80             'title': 'Material world: how Google discovered what software is made of',
  81             'description': 'md5:dfc17e7715e3b542d66e33a109861382',
  82             'upload_date': '20190710',
  83             'uploader_id': 'TheVerge',
  84             'uploader': 'The Verge',
  85         },
  86         'add_ie': ['Youtube'],
  87     }, {
  88         # Volume embed, Youtube
  89         'url': 'http://www.theverge.com/2014/10/21/7025853/google-nexus-6-hands-on-photos-video-android-phablet',
  90         'md5': 'fd19aa0cf3a0eea515d4fd5c8c0e9d68',
  91         'info_dict': {
  92             'id': 'Gy8Md3Eky38',
  93             'ext': 'mp4',
  94             'title': 'The Nexus 6: hands-on with Google\'s phablet',
  95             'description': 'md5:d9f0216e5fb932dd2033d6db37ac3f1d',
  96             'uploader_id': 'TheVerge',
  97             'upload_date': '20141021',
  98             'uploader': 'The Verge',
  99             'timestamp': 1413907200,
 100         },
 101         'add_ie': ['Youtube'],
 102         'skip': 'similar to the previous test',
 103     }, {
 104         # Volume embed, Youtube
 105         'url': 'http://www.vox.com/2016/3/31/11336640/mississippi-lgbt-religious-freedom-bill',
 106         'info_dict': {
 107             'id': '22986359b',
 108             'ext': 'mp4',
 109             'title': "Mississippi's laws are so bad that its anti-LGBTQ law isn't needed to allow discrimination",
 110             'description': 'md5:fc1317922057de31cd74bce91eb1c66c',
 111             'upload_date': '20150915',
 112             'timestamp': 1442332800,
 113             'duration': 285,
 114         },
 115         'add_ie': ['Youtube'],
 116         'skip': 'similar to the previous test',
 117     }, {
 118         # youtube embed
 119         'url': 'http://www.vox.com/2016/3/24/11291692/robot-dance',
 120         'md5': '83b3080489fb103941e549352d3e0977',
 121         'info_dict': {
 122             'id': 'FcNHTJU1ufM',
 123             'ext': 'mp4',
 124             'title': 'How "the robot" became the greatest novelty dance of all time',
 125             'description': 'md5:b081c0d588b8b2085870cda55e6da176',
 126             'upload_date': '20160324',
 127             'uploader_id': 'voxdotcom',
 128             'uploader': 'Vox',
 129         },
 130         'add_ie': ['Youtube'],
 131         'skip': 'Page no longer contain videos',
 132     }, {
 133         # SBN.VideoLinkset.entryGroup multiple ooyala embeds
 134         'url': 'http://www.sbnation.com/college-football-recruiting/2015/2/3/7970291/national-signing-day-rationalizations-itll-be-ok-itll-be-ok',
 135         'info_dict': {
 136             'id': 'national-signing-day-rationalizations-itll-be-ok-itll-be-ok',
 137             'title': '25 lies you will tell yourself on National Signing Day',
 138             'description': 'It\'s the most self-delusional time of the year, and everyone\'s gonna tell the same lies together!',
 139         },
 140         'playlist': [{
 141             'md5': '721fededf2ab74ae4176c8c8cbfe092e',
 142             'info_dict': {
 143                 'id': 'p3cThlMjE61VDi_SD9JlIteSNPWVDBB9',
 144                 'ext': 'mp4',
 145                 'title': 'Buddy Hield vs Steph Curry (and the world)',
 146                 'description': 'Let’s dissect only the most important Final Four storylines.',
 147             },
 148         }, {
 149             'md5': 'bf0c5cc115636af028be1bab79217ea9',
 150             'info_dict': {
 151                 'id': 'BmbmVjMjE6esPHxdALGubTrouQ0jYLHj',
 152                 'ext': 'mp4',
 153                 'title': 'Chasing Cinderella 2016: Syracuse basketball',
 154                 'description': 'md5:e02d56b026d51aa32c010676765a690d',
 155             },
 156         }],
 157         'skip': 'Page no longer contain videos',
 158     }, {
 159         # volume embed, Brightcove Once
 160         'url': 'https://www.recode.net/2014/6/17/11628066/post-post-pc-ceo-the-full-code-conference-video-of-microsofts-satya',
 161         'md5': '2dbc77b8b0bff1894c2fce16eded637d',
 162         'info_dict': {
 163             'id': '1231c973d',
 164             'ext': 'mp4',
 165             'title': 'Post-Post-PC CEO: The Full Code Conference Video of Microsoft\'s Satya Nadella',
 166             'description': 'The longtime veteran was chosen earlier this year as the software giant\'s third leader in its history.',
 167             'timestamp': 1402938000,
 168             'upload_date': '20140616',
 169             'duration': 4114,
 170         },
 171         'add_ie': ['VoxMediaVolume'],
 172     }]
 173
 174     def _real_extract(self, url):
 175         display_id = self._match_id(url)
 176         webpage = urllib.parse.unquote(self._download_webpage(url, display_id))
 177
 178         def create_entry(provider_video_id, provider_video_type, title=None, description=None):
 179             video_url = {
 180                 'youtube': '%s',
 181                 'volume': 'http://volume.vox-cdn.com/embed/%s',
 182             }[provider_video_type] % provider_video_id
 183             return {
 184                 '_type': 'url_transparent',
 185                 'url': video_url,
 186                 'title': title or self._og_search_title(webpage),
 187                 'description': description or self._og_search_description(webpage),
 188             }
 189
 190         entries = []
 191         entries_data = self._search_regex([
 192             r'Chorus\.VideoContext\.addVideo\((\[{.+}\])\);',
 193             r'var\s+entry\s*=\s*({.+});',
 194             r'SBN\.VideoLinkset\.entryGroup\(\s*(\[.+\])',
 195         ], webpage, 'video data', default=None)
 196         if entries_data:
 197             entries_data = self._parse_json(entries_data, display_id)
 198             if isinstance(entries_data, dict):
 199                 entries_data = [entries_data]
 200             for video_data in entries_data:
 201                 provider_video_id = video_data.get('provider_video_id')
 202                 provider_video_type = video_data.get('provider_video_type')
 203                 if provider_video_id and provider_video_type:
 204                     entries.append(create_entry(
 205                         provider_video_id, provider_video_type,
 206                         video_data.get('title'), video_data.get('description')))
 207
 208         volume_uuid = self._search_regex(
 209             r'data-volume-uuid="([^"]+)"', webpage, 'volume uuid', default=None)
 210         if volume_uuid:
 211             entries.append(create_entry(volume_uuid, 'volume'))
 212
 213         if len(entries) == 1:
 214             return entries[0]
 215         else:
 216             return self.playlist_result(entries, display_id, self._og_search_title(webpage), self._og_search_description(webpage))