youtube_dl/extractor/nhk.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import urljoin
   8
   9
  10 class NhkBaseIE(InfoExtractor):
  11     _API_URL_TEMPLATE = 'https://nwapi.nhk.jp/nhkworld/%sod%slist/v7b/%s/%s/%s/all%s.json'
  12     _BASE_URL_REGEX = r'https?://www3\.nhk\.or\.jp/nhkworld/(?P<lang>[a-z]{2})/ondemand'
  13     _TYPE_REGEX = r'/(?P<type>video|audio)/'
  14
  15     def _call_api(self, m_id, lang, is_video, is_episode, is_clip):
  16         return self._download_json(
  17             self._API_URL_TEMPLATE % (
  18                 'v' if is_video else 'r',
  19                 'clip' if is_clip else 'esd',
  20                 'episode' if is_episode else 'program',
  21                 m_id, lang, '/all' if is_video else ''),
  22             m_id, query={'apikey': 'EJfK8jdS57GqlupFgAfAAwr573q01y6k'})['data']['episodes'] or []
  23
  24     def _extract_episode_info(self, url, episode=None):
  25         fetch_episode = episode is None
  26         lang, m_type, episode_id = re.match(NhkVodIE._VALID_URL, url).groups()
  27         if len(episode_id) == 7:
  28             episode_id = episode_id[:4] + '-' + episode_id[4:]
  29
  30         is_video = m_type == 'video'
  31         if fetch_episode:
  32             episode = self._call_api(
  33                 episode_id, lang, is_video, True, episode_id[:4] == '9999')[0]
  34         title = episode.get('sub_title_clean') or episode['sub_title']
  35
  36         def get_clean_field(key):
  37             return episode.get(key + '_clean') or episode.get(key)
  38
  39         series = get_clean_field('title')
  40
  41         thumbnails = []
  42         for s, w, h in [('', 640, 360), ('_l', 1280, 720)]:
  43             img_path = episode.get('image' + s)
  44             if not img_path:
  45                 continue
  46             thumbnails.append({
  47                 'id': '%dp' % h,
  48                 'height': h,
  49                 'width': w,
  50                 'url': 'https://www3.nhk.or.jp' + img_path,
  51             })
  52
  53         info = {
  54             'id': episode_id + '-' + lang,
  55             'title': '%s - %s' % (series, title) if series and title else title,
  56             'description': get_clean_field('description'),
  57             'thumbnails': thumbnails,
  58             'series': series,
  59             'episode': title,
  60         }
  61         if is_video:
  62             vod_id = episode['vod_id']
  63             info.update({
  64                 '_type': 'url_transparent',
  65                 'ie_key': 'Piksel',
  66                 'url': 'https://player.piksel.com/v/refid/nhkworld/prefid/' + vod_id,
  67                 'id': vod_id,
  68             })
  69         else:
  70             if fetch_episode:
  71                 audio_path = episode['audio']['audio']
  72                 info['formats'] = self._extract_m3u8_formats(
  73                     'https://nhkworld-vh.akamaihd.net/i%s/master.m3u8' % audio_path,
  74                     episode_id, 'm4a', entry_protocol='m3u8_native',
  75                     m3u8_id='hls', fatal=False)
  76                 for f in info['formats']:
  77                     f['language'] = lang
  78             else:
  79                 info.update({
  80                     '_type': 'url_transparent',
  81                     'ie_key': NhkVodIE.ie_key(),
  82                     'url': url,
  83                 })
  84         return info
  85
  86
  87 class NhkVodIE(NhkBaseIE):
  88     # the 7-character IDs can have alphabetic chars too: assume [a-z] rather than just [a-f], eg
  89     _VALID_URL = r'%s%s(?P<id>[0-9a-z]{7}|[^/]+?-\d{8}-[0-9a-z]+)' % (NhkBaseIE._BASE_URL_REGEX, NhkBaseIE._TYPE_REGEX)
  90     # Content available only for a limited period of time. Visit
  91     # https://www3.nhk.or.jp/nhkworld/en/ondemand/ for working samples.
  92     _TESTS = [{
  93         # video clip
  94         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999011/',
  95         'md5': '7a90abcfe610ec22a6bfe15bd46b30ca',
  96         'info_dict': {
  97             'id': 'a95j5iza',
  98             'ext': 'mp4',
  99             'title': "Dining with the Chef - Chef Saito's Family recipe: MENCHI-KATSU",
 100             'description': 'md5:5aee4a9f9d81c26281862382103b0ea5',
 101             'timestamp': 1565965194,
 102             'upload_date': '20190816',
 103         },
 104     }, {
 105         # audio clip
 106         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/r_inventions-20201104-1/',
 107         'info_dict': {
 108             'id': 'r_inventions-20201104-1-en',
 109             'ext': 'm4a',
 110             'title': "Japan's Top Inventions - Miniature Video Cameras",
 111             'description': 'md5:07ea722bdbbb4936fdd360b6a480c25b',
 112         },
 113         'params': {
 114             # m3u8 download
 115             'skip_download': True,
 116         },
 117     }, {
 118         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2015173/',
 119         'only_matching': True,
 120     }, {
 121         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/plugin-20190404-1/',
 122         'only_matching': True,
 123     }, {
 124         'url': 'https://www3.nhk.or.jp/nhkworld/fr/ondemand/audio/plugin-20190404-1/',
 125         'only_matching': True,
 126     }, {
 127         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/audio/j_art-20150903-1/',
 128         'only_matching': True,
 129     }, {
 130         # video, alphabetic character in ID #29670
 131         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/video/9999a34/',
 132         'only_matching': True,
 133         'info_dict': {
 134             'id': 'qfjay6cg',
 135             'ext': 'mp4',
 136             'title': 'DESIGN TALKS plus - Fishermen’s Finery',
 137             'description': 'md5:8a8f958aaafb0d7cb59d38de53f1e448',
 138             'thumbnail': r're:^https?:/(/[a-z0-9.-]+)+\.jpg\?w=1920&h=1080$',
 139             'upload_date': '20210615',
 140             'timestamp': 1623722008,
 141         }
 142     }]
 143
 144     def _real_extract(self, url):
 145         return self._extract_episode_info(url)
 146
 147
 148 class NhkVodProgramIE(NhkBaseIE):
 149     _VALID_URL = r'%s/program%s(?P<id>[0-9a-z]+)(?:.+?\btype=(?P<episode_type>clip|(?:radio|tv)Episode))?' % (NhkBaseIE._BASE_URL_REGEX, NhkBaseIE._TYPE_REGEX)
 150     _TESTS = [{
 151         # video program episodes
 152         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway',
 153         'info_dict': {
 154             'id': 'japanrailway',
 155             'title': 'Japan Railway Journal',
 156         },
 157         'playlist_mincount': 1,
 158     }, {
 159         # video program clips
 160         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/japanrailway/?type=clip',
 161         'info_dict': {
 162             'id': 'japanrailway',
 163             'title': 'Japan Railway Journal',
 164         },
 165         'playlist_mincount': 5,
 166     }, {
 167         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/video/10yearshayaomiyazaki/',
 168         'only_matching': True,
 169     }, {
 170         # audio program
 171         'url': 'https://www3.nhk.or.jp/nhkworld/en/ondemand/program/audio/listener/',
 172         'only_matching': True,
 173     }]
 174
 175     def _real_extract(self, url):
 176         lang, m_type, program_id, episode_type = re.match(self._VALID_URL, url).groups()
 177
 178         episodes = self._call_api(
 179             program_id, lang, m_type == 'video', False, episode_type == 'clip')
 180
 181         entries = []
 182         for episode in episodes:
 183             episode_path = episode.get('url')
 184             if not episode_path:
 185                 continue
 186             entries.append(self._extract_episode_info(
 187                 urljoin(url, episode_path), episode))
 188
 189         program_title = None
 190         if entries:
 191             program_title = entries[0].get('series')
 192
 193         return self.playlist_result(entries, program_id, program_title)