yt_dlp/extractor/naver.py

   1 import itertools
   2 import re
   3 from urllib.parse import urlparse, parse_qs
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     ExtractorError,
   8     clean_html,
   9     dict_get,
  10     int_or_none,
  11     merge_dicts,
  12     parse_duration,
  13     traverse_obj,
  14     try_call,
  15     try_get,
  16     unified_timestamp,
  17     update_url_query,
  18 )
  19
  20
  21 class NaverBaseIE(InfoExtractor):
  22     _CAPTION_EXT_RE = r'\.(?:ttml|vtt)'
  23
  24     def _extract_video_info(self, video_id, vid, key):
  25         video_data = self._download_json(
  26             'http://play.rmcnmv.naver.com/vod/play/v2.0/' + vid,
  27             video_id, query={
  28                 'key': key,
  29             })
  30         meta = video_data['meta']
  31         title = meta['subject']
  32         formats = []
  33         get_list = lambda x: try_get(video_data, lambda y: y[x + 's']['list'], list) or []
  34
  35         def extract_formats(streams, stream_type, query={}):
  36             for stream in streams:
  37                 stream_url = stream.get('source')
  38                 if not stream_url:
  39                     continue
  40                 stream_url = update_url_query(stream_url, query)
  41                 encoding_option = stream.get('encodingOption', {})
  42                 bitrate = stream.get('bitrate', {})
  43                 formats.append({
  44                     'format_id': '%s_%s' % (stream.get('type') or stream_type, dict_get(encoding_option, ('name', 'id'))),
  45                     'url': stream_url,
  46                     'ext': 'mp4',
  47                     'width': int_or_none(encoding_option.get('width')),
  48                     'height': int_or_none(encoding_option.get('height')),
  49                     'vbr': int_or_none(bitrate.get('video')),
  50                     'abr': int_or_none(bitrate.get('audio')),
  51                     'filesize': int_or_none(stream.get('size')),
  52                     'protocol': 'm3u8_native' if stream_type == 'HLS' else None,
  53                 })
  54
  55         extract_formats(get_list('video'), 'H264')
  56         for stream_set in video_data.get('streams', []):
  57             query = {}
  58             for param in stream_set.get('keys', []):
  59                 query[param['name']] = param['value']
  60             stream_type = stream_set.get('type')
  61             videos = stream_set.get('videos')
  62             if videos:
  63                 extract_formats(videos, stream_type, query)
  64             elif stream_type == 'HLS':
  65                 stream_url = stream_set.get('source')
  66                 if not stream_url:
  67                     continue
  68                 formats.extend(self._extract_m3u8_formats(
  69                     update_url_query(stream_url, query), video_id,
  70                     'mp4', 'm3u8_native', m3u8_id=stream_type, fatal=False))
  71         self._sort_formats(formats)
  72
  73         replace_ext = lambda x, y: re.sub(self._CAPTION_EXT_RE, '.' + y, x)
  74
  75         def get_subs(caption_url):
  76             if re.search(self._CAPTION_EXT_RE, caption_url):
  77                 return [{
  78                     'url': replace_ext(caption_url, 'ttml'),
  79                 }, {
  80                     'url': replace_ext(caption_url, 'vtt'),
  81                 }]
  82             else:
  83                 return [{'url': caption_url}]
  84
  85         automatic_captions = {}
  86         subtitles = {}
  87         for caption in get_list('caption'):
  88             caption_url = caption.get('source')
  89             if not caption_url:
  90                 continue
  91             sub_dict = automatic_captions if caption.get('type') == 'auto' else subtitles
  92             sub_dict.setdefault(dict_get(caption, ('locale', 'language')), []).extend(get_subs(caption_url))
  93
  94         user = meta.get('user', {})
  95
  96         return {
  97             'id': video_id,
  98             'title': title,
  99             'formats': formats,
 100             'subtitles': subtitles,
 101             'automatic_captions': automatic_captions,
 102             'thumbnail': try_get(meta, lambda x: x['cover']['source']),
 103             'view_count': int_or_none(meta.get('count')),
 104             'uploader_id': user.get('id'),
 105             'uploader': user.get('name'),
 106             'uploader_url': user.get('url'),
 107         }
 108
 109
 110 class NaverIE(NaverBaseIE):
 111     _VALID_URL = r'https?://(?:m\.)?tv(?:cast)?\.naver\.com/(?:v|embed)/(?P<id>\d+)'
 112     _GEO_BYPASS = False
 113     _TESTS = [{
 114         'url': 'http://tv.naver.com/v/81652',
 115         'info_dict': {
 116             'id': '81652',
 117             'ext': 'mp4',
 118             'title': '[9월 모의고사 해설강의][수학_김상희] 수학 A형 16~20번',
 119             'description': '메가스터디 수학 김상희 선생님이 9월 모의고사 수학A형 16번에서 20번까지 해설강의를 공개합니다.',
 120             'timestamp': 1378200754,
 121             'upload_date': '20130903',
 122             'uploader': '메가스터디, 합격불변의 법칙',
 123             'uploader_id': 'megastudy',
 124         },
 125     }, {
 126         'url': 'http://tv.naver.com/v/395837',
 127         'md5': '8a38e35354d26a17f73f4e90094febd3',
 128         'info_dict': {
 129             'id': '395837',
 130             'ext': 'mp4',
 131             'title': '9년이 지나도 아픈 기억, 전효성의 아버지',
 132             'description': 'md5:eb6aca9d457b922e43860a2a2b1984d3',
 133             'timestamp': 1432030253,
 134             'upload_date': '20150519',
 135             'uploader': '4가지쇼 시즌2',
 136             'uploader_id': 'wrappinguser29',
 137         },
 138         'skip': 'Georestricted',
 139     }, {
 140         'url': 'http://tvcast.naver.com/v/81652',
 141         'only_matching': True,
 142     }]
 143
 144     def _real_extract(self, url):
 145         video_id = self._match_id(url)
 146         content = self._download_json(
 147             'https://tv.naver.com/api/json/v/' + video_id,
 148             video_id, headers=self.geo_verification_headers())
 149         player_info_json = content.get('playerInfoJson') or {}
 150         current_clip = player_info_json.get('currentClip') or {}
 151
 152         vid = current_clip.get('videoId')
 153         in_key = current_clip.get('inKey')
 154
 155         if not vid or not in_key:
 156             player_auth = try_get(player_info_json, lambda x: x['playerOption']['auth'])
 157             if player_auth == 'notCountry':
 158                 self.raise_geo_restricted(countries=['KR'])
 159             elif player_auth == 'notLogin':
 160                 self.raise_login_required()
 161             raise ExtractorError('couldn\'t extract vid and key')
 162         info = self._extract_video_info(video_id, vid, in_key)
 163         info.update({
 164             'description': clean_html(current_clip.get('description')),
 165             'timestamp': int_or_none(current_clip.get('firstExposureTime'), 1000),
 166             'duration': parse_duration(current_clip.get('displayPlayTime')),
 167             'like_count': int_or_none(current_clip.get('recommendPoint')),
 168             'age_limit': 19 if current_clip.get('adult') else None,
 169         })
 170         return info
 171
 172
 173 class NaverLiveIE(InfoExtractor):
 174     IE_NAME = 'Naver:live'
 175     _VALID_URL = r'https?://(?:m\.)?tv(?:cast)?\.naver\.com/l/(?P<id>\d+)'
 176     _GEO_BYPASS = False
 177     _TESTS = [{
 178         'url': 'https://tv.naver.com/l/52010',
 179         'info_dict': {
 180             'id': '52010',
 181             'ext': 'mp4',
 182             'title': '[LIVE] 뉴스특보 : "수도권 거리두기, 2주간 2단계로 조정"',
 183             'description': 'md5:df7f0c237a5ed5e786ce5c91efbeaab3',
 184             'channel_id': 'NTV-ytnnews24-0',
 185             'start_time': 1597026780000,
 186         },
 187     }, {
 188         'url': 'https://tv.naver.com/l/51549',
 189         'info_dict': {
 190             'id': '51549',
 191             'ext': 'mp4',
 192             'title': '연합뉴스TV - 코로나19 뉴스특보',
 193             'description': 'md5:c655e82091bc21e413f549c0eaccc481',
 194             'channel_id': 'NTV-yonhapnewstv-0',
 195             'start_time': 1596406380000,
 196         },
 197     }, {
 198         'url': 'https://tv.naver.com/l/54887',
 199         'only_matching': True,
 200     }]
 201
 202     def _real_extract(self, url):
 203         video_id = self._match_id(url)
 204         page = self._download_webpage(url, video_id, 'Downloading Page', 'Unable to download Page')
 205         secure_url = self._search_regex(r'sApiF:\s+(?:"|\')([^"\']+)', page, 'secureurl')
 206
 207         info = self._extract_video_info(video_id, secure_url)
 208         info.update({
 209             'description': self._og_search_description(page)
 210         })
 211
 212         return info
 213
 214     def _extract_video_info(self, video_id, url):
 215         video_data = self._download_json(url, video_id, headers=self.geo_verification_headers())
 216         meta = video_data.get('meta')
 217         status = meta.get('status')
 218
 219         if status == 'CLOSED':
 220             raise ExtractorError('Stream is offline.', expected=True)
 221         elif status != 'OPENED':
 222             raise ExtractorError('Unknown status %s' % status)
 223
 224         title = meta.get('title')
 225         stream_list = video_data.get('streams')
 226
 227         if stream_list is None:
 228             raise ExtractorError('Could not get stream data.', expected=True)
 229
 230         formats = []
 231         for quality in stream_list:
 232             if not quality.get('url'):
 233                 continue
 234
 235             prop = quality.get('property')
 236             if prop.get('abr'):  # This abr doesn't mean Average audio bitrate.
 237                 continue
 238
 239             formats.extend(self._extract_m3u8_formats(
 240                 quality.get('url'), video_id, 'mp4',
 241                 m3u8_id=quality.get('qualityId'), live=True
 242             ))
 243         self._sort_formats(formats)
 244
 245         return {
 246             'id': video_id,
 247             'title': title,
 248             'formats': formats,
 249             'channel_id': meta.get('channelId'),
 250             'channel_url': meta.get('channelUrl'),
 251             'thumbnail': meta.get('imgUrl'),
 252             'start_time': meta.get('startTime'),
 253             'categories': [meta.get('categoryId')],
 254             'is_live': True
 255         }
 256
 257
 258 class NaverNowIE(NaverBaseIE):
 259     IE_NAME = 'navernow'
 260     _VALID_URL = r'https?://now\.naver\.com/show/(?P<id>[0-9]+)'
 261     _PAGE_SIZE = 30
 262     _API_URL = 'https://apis.naver.com/now_web/nowcms-api-xhmac/cms/v1'
 263     _TESTS = [{
 264         'url': 'https://now.naver.com/show/4759?shareReplayId=5901#replay=',
 265         'md5': 'e05854162c21c221481de16b2944a0bc',
 266         'info_dict': {
 267             'id': '4759-5901',
 268             'title': '아이키X노제\r\n💖꽁냥꽁냥💖(1)',
 269             'ext': 'mp4',
 270             'thumbnail': r're:^https?://.*\.jpg',
 271             'timestamp': 1650369600,
 272             'upload_date': '20220419',
 273             'uploader_id': 'now',
 274             'view_count': int,
 275         },
 276         'params': {
 277             'noplaylist': True,
 278         }
 279     }, {
 280         'url': 'https://now.naver.com/show/4759?shareHightlight=1078#highlight=',
 281         'md5': '9f6118e398aa0f22b2152f554ea7851b',
 282         'info_dict': {
 283             'id': '4759-1078',
 284             'title': '아이키: 나 리정한테 흔들렸어,,, 질투 폭발하는 노제 여보😾 [아이키의 떰즈업]ㅣ네이버 NOW.',
 285             'ext': 'mp4',
 286             'thumbnail': r're:^https?://.*\.jpg',
 287             'upload_date': '20220504',
 288             'timestamp': 1651648042,
 289             'uploader_id': 'now',
 290             'view_count': int,
 291         },
 292         'params': {
 293             'noplaylist': True,
 294         },
 295     }, {
 296         'url': 'https://now.naver.com/show/4759',
 297         'info_dict': {
 298             'id': '4759',
 299             'title': '아이키의 떰즈업',
 300         },
 301         'playlist_mincount': 48
 302     }, {
 303         'url': 'https://now.naver.com/show/4759?shareReplayId=5901#replay',
 304         'info_dict': {
 305             'id': '4759',
 306             'title': '아이키의 떰즈업',
 307         },
 308         'playlist_mincount': 48,
 309     }, {
 310         'url': 'https://now.naver.com/show/4759?shareHightlight=1078#highlight=',
 311         'info_dict': {
 312             'id': '4759',
 313             'title': '아이키의 떰즈업',
 314         },
 315         'playlist_mincount': 48,
 316     }]
 317
 318     def _extract_replay(self, show_id, replay_id):
 319         vod_info = self._download_json(f'{self._API_URL}/shows/{show_id}/vod/{replay_id}', replay_id)
 320         in_key = self._download_json(f'{self._API_URL}/shows/{show_id}/vod/{replay_id}/inkey', replay_id)['inKey']
 321         return merge_dicts({
 322             'id': f'{show_id}-{replay_id}',
 323             'title': traverse_obj(vod_info, ('episode', 'title')),
 324             'timestamp': unified_timestamp(traverse_obj(vod_info, ('episode', 'start_time'))),
 325             'thumbnail': vod_info.get('thumbnail_image_url'),
 326         }, self._extract_video_info(replay_id, vod_info['video_id'], in_key))
 327
 328     def _extract_show_replays(self, show_id):
 329         page = 0
 330         while True:
 331             show_vod_info = self._download_json(
 332                 f'{self._API_URL}/vod-shows/{show_id}', show_id,
 333                 query={'offset': page * self._PAGE_SIZE, 'limit': self._PAGE_SIZE},
 334                 note=f'Downloading JSON vod list for show {show_id} - page {page}'
 335             )['response']['result']
 336             for v in show_vod_info.get('vod_list') or []:
 337                 yield self._extract_replay(show_id, v['id'])
 338
 339             if try_call(lambda: show_vod_info['count'] <= self._PAGE_SIZE * (page + 1)):
 340                 break
 341             page += 1
 342
 343     def _extract_show_highlights(self, show_id, highlight_id=None):
 344         page = 0
 345         while True:
 346             highlights_videos = self._download_json(
 347                 f'{self._API_URL}/shows/{show_id}/highlights/videos/', show_id,
 348                 query={'offset': page * self._PAGE_SIZE, 'limit': self._PAGE_SIZE},
 349                 note=f'Downloading JSON highlights for show {show_id} - page {page}')
 350
 351             for highlight in highlights_videos.get('results') or []:
 352                 if highlight_id and highlight.get('id') != int(highlight_id):
 353                     continue
 354                 yield merge_dicts({
 355                     'id': f'{show_id}-{highlight["id"]}',
 356                     'title': highlight.get('title'),
 357                     'timestamp': unified_timestamp(highlight.get('regdate')),
 358                     'thumbnail': highlight.get('thumbnail_url'),
 359                 }, self._extract_video_info(highlight['id'], highlight['video_id'], highlight['video_inkey']))
 360
 361             if try_call(lambda: highlights_videos['count'] <= self._PAGE_SIZE * (page + 1)):
 362                 break
 363             page += 1
 364
 365     def _extract_highlight(self, show_id, highlight_id):
 366         try:
 367             return next(self._extract_show_highlights(show_id, highlight_id))
 368         except StopIteration:
 369             raise ExtractorError(f'Unable to find highlight {highlight_id} for show {show_id}')
 370
 371     def _real_extract(self, url):
 372         show_id = self._match_id(url)
 373         qs = parse_qs(urlparse(url).query)
 374
 375         if not self._yes_playlist(show_id, qs.get('shareHightlight')):
 376             return self._extract_highlight(show_id, qs['shareHightlight'][0])
 377         elif not self._yes_playlist(show_id, qs.get('shareReplayId')):
 378             return self._extract_replay(show_id, qs['shareReplayId'][0])
 379
 380         show_info = self._download_json(
 381             f'{self._API_URL}/shows/{show_id}', show_id,
 382             note=f'Downloading JSON vod list for show {show_id}')
 383
 384         return self.playlist_result(
 385             itertools.chain(self._extract_show_replays(show_id), self._extract_show_highlights(show_id)),
 386             show_id, show_info.get('title'))