youtube_dl/extractor/npr.py

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