yt_dlp/extractor/nosnl.py

   1 from .common import InfoExtractor
   2 from ..utils import parse_duration, parse_iso8601, traverse_obj
   3
   4
   5 class NOSNLArticleIE(InfoExtractor):
   6     _VALID_URL = r'https?://nos\.nl/(?P<type>video|(\w+/)?\w+)/?\d+-(?P<display_id>[\w-]+)'
   7     _TESTS = [
   8         {
   9             # only 1 video
  10             'url': 'https://nos.nl/nieuwsuur/artikel/2440353-verzakking-door-droogte-dreigt-tot-een-miljoen-kwetsbare-huizen',
  11             'info_dict': {
  12                 'id': '2440340',
  13                 'ext': 'mp4',
  14                 'description': 'md5:5f83185d902ac97af3af4bed7ece3db5',
  15                 'title': '\'We hebben een huis vol met scheuren\'',
  16                 'duration': 95.0,
  17                 'thumbnail': 'https://cdn.nos.nl/image/2022/08/12/887149/3840x2160a.jpg',
  18             },
  19         }, {
  20             # more than 1 video
  21             'url': 'https://nos.nl/artikel/2440409-vannacht-sliepen-weer-enkele-honderden-asielzoekers-in-ter-apel-buiten',
  22             'info_dict': {
  23                 'id': '2440409',
  24                 'title': 'Vannacht sliepen weer enkele honderden asielzoekers in Ter Apel buiten',
  25                 'description': 'md5:72b1e1674d798460e79d78fa37e9f56d',
  26                 'tags': ['aanmeldcentrum', 'Centraal Orgaan opvang asielzoekers', 'COA', 'asielzoekers', 'Ter Apel'],
  27                 'modified_timestamp': 1660452773,
  28                 'modified_date': '20220814',
  29                 'upload_date': '20220813',
  30                 'thumbnail': 'https://cdn.nos.nl/image/2022/07/18/880346/1024x576a.jpg',
  31                 'timestamp': 1660401384,
  32                 'categories': ['Regionaal nieuws', 'Binnenland'],
  33             },
  34             'playlist_count': 2,
  35         }, {
  36             # audio + video
  37             'url': 'https://nos.nl/artikel/2440789-wekdienst-16-8-groningse-acties-tien-jaar-na-zware-aardbeving-femke-bol-in-actie-op-ek-atletiek',
  38             'info_dict': {
  39                 'id': '2440789',
  40                 'title': 'Wekdienst 16/8: Groningse acties tien jaar na zware aardbeving • Femke Bol in actie op EK atletiek ',
  41                 'description': 'md5:0bd277ed7a44fc15cb12a9d27d8f6641',
  42                 'tags': ['wekdienst'],
  43                 'modified_date': '20220816',
  44                 'modified_timestamp': 1660625449,
  45                 'timestamp': 1660625449,
  46                 'upload_date': '20220816',
  47                 'thumbnail': 'https://cdn.nos.nl/image/2022/08/16/888178/1024x576a.jpg',
  48                 'categories': ['Binnenland', 'Buitenland'],
  49             },
  50             'playlist_count': 2,
  51         }, {
  52             # video url
  53             'url': 'https://nos.nl/video/2452718-xi-en-trudeau-botsen-voor-de-camera-op-g20-top-je-hebt-gelekt',
  54             'info_dict': {
  55                 'id': '2452718',
  56                 'title': 'Xi en Trudeau botsen voor de camera op G20-top: \'Je hebt gelekt\'',
  57                 'modified_date': '20221117',
  58                 'description': 'md5:61907dac576f75c11bf8ffffd4a3cc0f',
  59                 'tags': ['Xi', 'Trudeau', 'G20', 'indonesié'],
  60                 'upload_date': '20221117',
  61                 'thumbnail': 'https://cdn.nos.nl/image/2022/11/17/916155/1024x576a.jpg',
  62                 'modified_timestamp': 1668663388,
  63                 'timestamp': 1668663388,
  64                 'categories': ['Buitenland'],
  65             },
  66             'playlist_mincount': 1,
  67         },
  68     ]
  69
  70     def _entries(self, nextjs_json, display_id):
  71         for item in nextjs_json:
  72             if item.get('type') == 'video':
  73                 formats, subtitle = self._extract_m3u8_formats_and_subtitles(
  74                     traverse_obj(item, ('source', 'url')), display_id, ext='mp4')
  75                 yield {
  76                     'id': str(item['id']),
  77                     'title': item.get('title'),
  78                     'description': item.get('description'),
  79                     'formats': formats,
  80                     'subtitles': subtitle,
  81                     'duration': parse_duration(item.get('duration')),
  82                     'thumbnails': [{
  83                         'url': traverse_obj(image, ('url', ...), get_all=False),
  84                         'width': image.get('width'),
  85                         'height': image.get('height'),
  86                     } for image in traverse_obj(item, ('imagesByRatio', ...))[0]],
  87                 }
  88
  89             elif item.get('type') == 'audio':
  90                 yield {
  91                     'id': str(item['id']),
  92                     'title': item.get('title'),
  93                     'url': traverse_obj(item, ('media', 'src')),
  94                     'ext': 'mp3',
  95                 }
  96
  97     def _real_extract(self, url):
  98         site_type, display_id = self._match_valid_url(url).group('type', 'display_id')
  99         webpage = self._download_webpage(url, display_id)
 100
 101         nextjs_json = self._search_nextjs_data(webpage, display_id)['props']['pageProps']['data']
 102         return {
 103             '_type': 'playlist',
 104             'entries': self._entries(
 105                 [nextjs_json['video']] if site_type == 'video' else nextjs_json['items'], display_id),
 106             'id': str(nextjs_json['id']),
 107             'title': nextjs_json.get('title') or self._html_search_meta(['title', 'og:title', 'twitter:title'], webpage),
 108             'description': (nextjs_json.get('description')
 109                             or self._html_search_meta(['description', 'twitter:description', 'og:description'], webpage)),
 110             'tags': nextjs_json.get('keywords'),
 111             'modified_timestamp': parse_iso8601(nextjs_json.get('modifiedAt')),
 112             'thumbnail': nextjs_json.get('shareImageSrc') or self._html_search_meta(['og:image', 'twitter:image'], webpage),
 113             'timestamp': parse_iso8601(nextjs_json.get('publishedAt')),
 114             'categories': traverse_obj(nextjs_json, ('categories', ..., 'label')),
 115         }