yt_dlp/extractor/npr.py

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