yt_dlp/extractor/npr.py

   1 from .common import InfoExtractor
   2 from ..utils import int_or_none, qualities, traverse_obj, url_or_none
   3
   4
   5 class NprIE(InfoExtractor):
   6     _VALID_URL = r'https?://(?:www\.)?npr\.org/(?:sections/[^/]+/)?\d{4}/\d{2}/\d{2}/(?P<id>\d+)'
   7     _TESTS = [{
   8         'url': 'https://www.npr.org/sections/allsongs/2015/10/21/449974205/new-music-from-beach-house-chairlift-cmj-discoveries-and-more',
   9         'info_dict': {
  10             'id': '449974205',
  11             'title': 'New Music From Beach House, Chairlift, CMJ Discoveries And More',
  12         },
  13         'playlist_count': 7,
  14     }, {
  15         'url': 'https://www.npr.org/sections/deceptivecadence/2015/10/09/446928052/music-from-the-shadows-ancient-armenian-hymns-and-piano-jazz',
  16         'info_dict': {
  17             'id': '446928052',
  18             'title': "Songs We Love: Tigran Hamasyan, 'Your Mercy is Boundless'",
  19         },
  20         'playlist': [{
  21             'md5': '12fa60cb2d3ed932f53609d4aeceabf1',
  22             'info_dict': {
  23                 'id': '446929930',
  24                 'ext': 'mp3',
  25                 'title': 'Your Mercy is Boundless (Bazum en Qo gtutyunqd)',
  26                 'duration': 402,
  27             },
  28         }],
  29     }, {
  30         # multimedia, not media title
  31         'url': 'https://www.npr.org/2017/06/19/533198237/tigers-jaw-tiny-desk-concert',
  32         'info_dict': {
  33             'id': '533198237',
  34             'title': 'Tigers Jaw: Tiny Desk Concert',
  35         },
  36         'playlist': [{
  37             'md5': '12fa60cb2d3ed932f53609d4aeceabf1',
  38             'info_dict': {
  39                 'id': '533201718',
  40                 'ext': 'mp4',
  41                 'title': 'Tigers Jaw: Tiny Desk Concert',
  42                 'duration': 402,
  43             },
  44         }],
  45         'expected_warnings': ['Failed to download m3u8 information'],
  46     }, {
  47         # multimedia, no formats, stream
  48         'url': 'https://www.npr.org/2020/02/14/805476846/laura-stevenson-tiny-desk-concert',
  49         'only_matching': True,
  50     }, {
  51         'url': 'https://www.npr.org/2022/03/15/1084896560/bonobo-tiny-desk-home-concert',
  52         'info_dict': {
  53             'id': '1086468851',
  54             'ext': 'mp4',
  55             'title': 'Bonobo: Tiny Desk (Home) Concert',
  56             'duration': 1061,
  57             'thumbnail': r're:^https?://media.npr.org/assets/img/.*\.jpg$',
  58         },
  59     }]
  60
  61     def _real_extract(self, url):
  62         playlist_id = self._match_id(url)
  63
  64         story = self._download_json(
  65             'http://api.npr.org/query', playlist_id, query={
  66                 'id': playlist_id,
  67                 'fields': 'audio,multimedia,title',
  68                 'format': 'json',
  69                 'apiKey': 'MDAzMzQ2MjAyMDEyMzk4MTU1MDg3ZmM3MQ010',
  70             })['list']['story'][0]
  71         playlist_title = story.get('title', {}).get('$text')
  72
  73         KNOWN_FORMATS = ('threegp', 'm3u8', 'smil', 'mp4', 'mp3')
  74         quality = qualities(KNOWN_FORMATS)
  75
  76         entries = []
  77         for media in story.get('audio', []) + story.get('multimedia', []):
  78             media_id = media['id']
  79
  80             formats = []
  81             for format_id, formats_entry in media.get('format', {}).items():
  82                 if not formats_entry:
  83                     continue
  84                 if isinstance(formats_entry, list):
  85                     formats_entry = formats_entry[0]
  86                 format_url = formats_entry.get('$text')
  87                 if not format_url:
  88                     continue
  89                 if format_id in KNOWN_FORMATS:
  90                     if format_id == 'm3u8':
  91                         formats.extend(self._extract_m3u8_formats(
  92                             format_url, media_id, 'mp4', 'm3u8_native',
  93                             m3u8_id='hls', fatal=False))
  94                     elif format_id == 'smil':
  95                         smil_formats = self._extract_smil_formats(
  96                             format_url, media_id, transform_source=lambda s: s.replace(
  97                                 'rtmp://flash.npr.org/ondemand/', 'https://ondemand.npr.org/'),
  98                             fatal=False)
  99                         self._check_formats(smil_formats, media_id)
 100                         formats.extend(smil_formats)
 101                     else:
 102                         formats.append({
 103                             'url': format_url,
 104                             'format_id': format_id,
 105                             'quality': quality(format_id),
 106                         })
 107             for stream_id, stream_entry in media.get('stream', {}).items():
 108                 if not isinstance(stream_entry, dict):
 109                     continue
 110                 if stream_id != 'hlsUrl':
 111                     continue
 112                 stream_url = url_or_none(stream_entry.get('$text'))
 113                 if not stream_url:
 114                     continue
 115                 formats.extend(self._extract_m3u8_formats(
 116                     stream_url, stream_id, 'mp4', 'm3u8_native',
 117                     m3u8_id='hls', fatal=False))
 118
 119             if not formats:
 120                 raw_json_ld = self._yield_json_ld(self._download_webpage(url, playlist_id), playlist_id, fatal=False)
 121                 m3u8_url = traverse_obj(list(raw_json_ld), (..., 'subjectOf', ..., 'embedUrl'), get_all=False)
 122                 formats = self._extract_m3u8_formats(m3u8_url, media_id, 'mp4', m3u8_id='hls', fatal=False)
 123
 124             entries.append({
 125                 'id': media_id,
 126                 'title': media.get('title', {}).get('$text') or playlist_title,
 127                 'thumbnail': media.get('altImageUrl', {}).get('$text'),
 128                 'duration': int_or_none(media.get('duration', {}).get('$text')),
 129                 'formats': formats,
 130             })
 131
 132         return self.playlist_result(entries, playlist_id, playlist_title)