]> gitweb @ CieloNegro.org - youtube-dl.git/blob - youtube_dl/extractor/bilibili.py
[vimeo] improve format extraction and sorting(closes #25285)
[youtube-dl.git] / youtube_dl / extractor / bilibili.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import hashlib
5 import re
6
7 from .common import InfoExtractor
8 from ..compat import (
9     compat_parse_qs,
10     compat_urlparse,
11 )
12 from ..utils import (
13     ExtractorError,
14     int_or_none,
15     float_or_none,
16     parse_iso8601,
17     smuggle_url,
18     str_or_none,
19     strip_jsonp,
20     unified_timestamp,
21     unsmuggle_url,
22     urlencode_postdata,
23 )
24
25
26 class BiliBiliIE(InfoExtractor):
27     _VALID_URL = r'''(?x)
28                     https?://
29                         (?:(?:www|bangumi)\.)?
30                         bilibili\.(?:tv|com)/
31                         (?:
32                             (?:
33                                 video/[aA][vV]|
34                                 anime/(?P<anime_id>\d+)/play\#
35                             )(?P<id_bv>\d+)|
36                             video/[bB][vV](?P<id>[^/?#&]+)
37                         )
38                     '''
39
40     _TESTS = [{
41         'url': 'http://www.bilibili.tv/video/av1074402/',
42         'md5': '5f7d29e1a2872f3df0cf76b1f87d3788',
43         'info_dict': {
44             'id': '1074402',
45             'ext': 'flv',
46             'title': '【金坷垃】金泡沫',
47             'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
48             'duration': 308.067,
49             'timestamp': 1398012678,
50             'upload_date': '20140420',
51             'thumbnail': r're:^https?://.+\.jpg',
52             'uploader': '菊子桑',
53             'uploader_id': '156160',
54         },
55     }, {
56         # Tested in BiliBiliBangumiIE
57         'url': 'http://bangumi.bilibili.com/anime/1869/play#40062',
58         'only_matching': True,
59     }, {
60         'url': 'http://bangumi.bilibili.com/anime/5802/play#100643',
61         'md5': '3f721ad1e75030cc06faf73587cfec57',
62         'info_dict': {
63             'id': '100643',
64             'ext': 'mp4',
65             'title': 'CHAOS;CHILD',
66             'description': '如果你是神明,并且能够让妄想成为现实。那你会进行怎么样的妄想?是淫靡的世界?独裁社会?毁灭性的制裁?还是……2015年,涩谷。从6年前发生的大灾害“涩谷地震”之后复兴了的这个街区里新设立的私立高中...',
67         },
68         'skip': 'Geo-restricted to China',
69     }, {
70         # Title with double quotes
71         'url': 'http://www.bilibili.com/video/av8903802/',
72         'info_dict': {
73             'id': '8903802',
74             'title': '阿滴英文|英文歌分享#6 "Closer',
75             'description': '滴妹今天唱Closer給你聽! 有史以来,被推最多次也是最久的歌曲,其实歌词跟我原本想像差蛮多的,不过还是好听! 微博@阿滴英文',
76         },
77         'playlist': [{
78             'info_dict': {
79                 'id': '8903802_part1',
80                 'ext': 'flv',
81                 'title': '阿滴英文|英文歌分享#6 "Closer',
82                 'description': 'md5:3b1b9e25b78da4ef87e9b548b88ee76a',
83                 'uploader': '阿滴英文',
84                 'uploader_id': '65880958',
85                 'timestamp': 1488382634,
86                 'upload_date': '20170301',
87             },
88             'params': {
89                 'skip_download': True,  # Test metadata only
90             },
91         }, {
92             'info_dict': {
93                 'id': '8903802_part2',
94                 'ext': 'flv',
95                 'title': '阿滴英文|英文歌分享#6 "Closer',
96                 'description': 'md5:3b1b9e25b78da4ef87e9b548b88ee76a',
97                 'uploader': '阿滴英文',
98                 'uploader_id': '65880958',
99                 'timestamp': 1488382634,
100                 'upload_date': '20170301',
101             },
102             'params': {
103                 'skip_download': True,  # Test metadata only
104             },
105         }]
106     }, {
107         # new BV video id format
108         'url': 'https://www.bilibili.com/video/BV1JE411F741',
109         'only_matching': True,
110     }]
111
112     _APP_KEY = 'iVGUTjsxvpLeuDCf'
113     _BILIBILI_KEY = 'aHRmhWMLkdeMuILqORnYZocwMBpMEOdt'
114
115     def _report_error(self, result):
116         if 'message' in result:
117             raise ExtractorError('%s said: %s' % (self.IE_NAME, result['message']), expected=True)
118         elif 'code' in result:
119             raise ExtractorError('%s returns error %d' % (self.IE_NAME, result['code']), expected=True)
120         else:
121             raise ExtractorError('Can\'t extract Bangumi episode ID')
122
123     def _real_extract(self, url):
124         url, smuggled_data = unsmuggle_url(url, {})
125
126         mobj = re.match(self._VALID_URL, url)
127         video_id = mobj.group('id') or mobj.group('id_bv')
128         anime_id = mobj.group('anime_id')
129         webpage = self._download_webpage(url, video_id)
130
131         if 'anime/' not in url:
132             cid = self._search_regex(
133                 r'\bcid(?:["\']:|=)(\d+)', webpage, 'cid',
134                 default=None
135             ) or compat_parse_qs(self._search_regex(
136                 [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
137                  r'EmbedPlayer\([^)]+,\s*\\"([^"]+)\\"\)',
138                  r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
139                 webpage, 'player parameters'))['cid'][0]
140         else:
141             if 'no_bangumi_tip' not in smuggled_data:
142                 self.to_screen('Downloading episode %s. To download all videos in anime %s, re-run youtube-dl with %s' % (
143                     video_id, anime_id, compat_urlparse.urljoin(url, '//bangumi.bilibili.com/anime/%s' % anime_id)))
144             headers = {
145                 'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
146                 'Referer': url
147             }
148             headers.update(self.geo_verification_headers())
149
150             js = self._download_json(
151                 'http://bangumi.bilibili.com/web_api/get_source', video_id,
152                 data=urlencode_postdata({'episode_id': video_id}),
153                 headers=headers)
154             if 'result' not in js:
155                 self._report_error(js)
156             cid = js['result']['cid']
157
158         headers = {
159             'Referer': url
160         }
161         headers.update(self.geo_verification_headers())
162
163         entries = []
164
165         RENDITIONS = ('qn=80&quality=80&type=', 'quality=2&type=mp4')
166         for num, rendition in enumerate(RENDITIONS, start=1):
167             payload = 'appkey=%s&cid=%s&otype=json&%s' % (self._APP_KEY, cid, rendition)
168             sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()
169
170             video_info = self._download_json(
171                 'http://interface.bilibili.com/v2/playurl?%s&sign=%s' % (payload, sign),
172                 video_id, note='Downloading video info page',
173                 headers=headers, fatal=num == len(RENDITIONS))
174
175             if not video_info:
176                 continue
177
178             if 'durl' not in video_info:
179                 if num < len(RENDITIONS):
180                     continue
181                 self._report_error(video_info)
182
183             for idx, durl in enumerate(video_info['durl']):
184                 formats = [{
185                     'url': durl['url'],
186                     'filesize': int_or_none(durl['size']),
187                 }]
188                 for backup_url in durl.get('backup_url', []):
189                     formats.append({
190                         'url': backup_url,
191                         # backup URLs have lower priorities
192                         'preference': -2 if 'hd.mp4' in backup_url else -3,
193                     })
194
195                 for a_format in formats:
196                     a_format.setdefault('http_headers', {}).update({
197                         'Referer': url,
198                     })
199
200                 self._sort_formats(formats)
201
202                 entries.append({
203                     'id': '%s_part%s' % (video_id, idx),
204                     'duration': float_or_none(durl.get('length'), 1000),
205                     'formats': formats,
206                 })
207             break
208
209         title = self._html_search_regex(
210             ('<h1[^>]+\btitle=(["\'])(?P<title>(?:(?!\1).)+)\1',
211              '(?s)<h1[^>]*>(?P<title>.+?)</h1>'), webpage, 'title',
212             group='title')
213         description = self._html_search_meta('description', webpage)
214         timestamp = unified_timestamp(self._html_search_regex(
215             r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time',
216             default=None) or self._html_search_meta(
217             'uploadDate', webpage, 'timestamp', default=None))
218         thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
219
220         # TODO 'view_count' requires deobfuscating Javascript
221         info = {
222             'id': video_id,
223             'title': title,
224             'description': description,
225             'timestamp': timestamp,
226             'thumbnail': thumbnail,
227             'duration': float_or_none(video_info.get('timelength'), scale=1000),
228         }
229
230         uploader_mobj = re.search(
231             r'<a[^>]+href="(?:https?:)?//space\.bilibili\.com/(?P<id>\d+)"[^>]*>(?P<name>[^<]+)',
232             webpage)
233         if uploader_mobj:
234             info.update({
235                 'uploader': uploader_mobj.group('name'),
236                 'uploader_id': uploader_mobj.group('id'),
237             })
238         if not info.get('uploader'):
239             info['uploader'] = self._html_search_meta(
240                 'author', webpage, 'uploader', default=None)
241
242         for entry in entries:
243             entry.update(info)
244
245         if len(entries) == 1:
246             return entries[0]
247         else:
248             for idx, entry in enumerate(entries):
249                 entry['id'] = '%s_part%d' % (video_id, (idx + 1))
250
251             return {
252                 '_type': 'multi_video',
253                 'id': video_id,
254                 'title': title,
255                 'description': description,
256                 'entries': entries,
257             }
258
259
260 class BiliBiliBangumiIE(InfoExtractor):
261     _VALID_URL = r'https?://bangumi\.bilibili\.com/anime/(?P<id>\d+)'
262
263     IE_NAME = 'bangumi.bilibili.com'
264     IE_DESC = 'BiliBili番剧'
265
266     _TESTS = [{
267         'url': 'http://bangumi.bilibili.com/anime/1869',
268         'info_dict': {
269             'id': '1869',
270             'title': '混沌武士',
271             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
272         },
273         'playlist_count': 26,
274     }, {
275         'url': 'http://bangumi.bilibili.com/anime/1869',
276         'info_dict': {
277             'id': '1869',
278             'title': '混沌武士',
279             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
280         },
281         'playlist': [{
282             'md5': '91da8621454dd58316851c27c68b0c13',
283             'info_dict': {
284                 'id': '40062',
285                 'ext': 'mp4',
286                 'title': '混沌武士',
287                 'description': '故事发生在日本的江户时代。风是一个小酒馆的打工女。一日,酒馆里来了一群恶霸,虽然他们的举动令风十分不满,但是毕竟风只是一届女流,无法对他们采取什么行动,只能在心里嘟哝。这时,酒家里又进来了个“不良份子...',
288                 'timestamp': 1414538739,
289                 'upload_date': '20141028',
290                 'episode': '疾风怒涛 Tempestuous Temperaments',
291                 'episode_number': 1,
292             },
293         }],
294         'params': {
295             'playlist_items': '1',
296         },
297     }]
298
299     @classmethod
300     def suitable(cls, url):
301         return False if BiliBiliIE.suitable(url) else super(BiliBiliBangumiIE, cls).suitable(url)
302
303     def _real_extract(self, url):
304         bangumi_id = self._match_id(url)
305
306         # Sometimes this API returns a JSONP response
307         season_info = self._download_json(
308             'http://bangumi.bilibili.com/jsonp/seasoninfo/%s.ver' % bangumi_id,
309             bangumi_id, transform_source=strip_jsonp)['result']
310
311         entries = [{
312             '_type': 'url_transparent',
313             'url': smuggle_url(episode['webplay_url'], {'no_bangumi_tip': 1}),
314             'ie_key': BiliBiliIE.ie_key(),
315             'timestamp': parse_iso8601(episode.get('update_time'), delimiter=' '),
316             'episode': episode.get('index_title'),
317             'episode_number': int_or_none(episode.get('index')),
318         } for episode in season_info['episodes']]
319
320         entries = sorted(entries, key=lambda entry: entry.get('episode_number'))
321
322         return self.playlist_result(
323             entries, bangumi_id,
324             season_info.get('bangumi_title'), season_info.get('evaluate'))
325
326
327 class BilibiliAudioBaseIE(InfoExtractor):
328     def _call_api(self, path, sid, query=None):
329         if not query:
330             query = {'sid': sid}
331         return self._download_json(
332             'https://www.bilibili.com/audio/music-service-c/web/' + path,
333             sid, query=query)['data']
334
335
336 class BilibiliAudioIE(BilibiliAudioBaseIE):
337     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/au(?P<id>\d+)'
338     _TEST = {
339         'url': 'https://www.bilibili.com/audio/au1003142',
340         'md5': 'fec4987014ec94ef9e666d4d158ad03b',
341         'info_dict': {
342             'id': '1003142',
343             'ext': 'm4a',
344             'title': '【tsukimi】YELLOW / 神山羊',
345             'artist': 'tsukimi',
346             'comment_count': int,
347             'description': 'YELLOW的mp3版!',
348             'duration': 183,
349             'subtitles': {
350                 'origin': [{
351                     'ext': 'lrc',
352                 }],
353             },
354             'thumbnail': r're:^https?://.+\.jpg',
355             'timestamp': 1564836614,
356             'upload_date': '20190803',
357             'uploader': 'tsukimi-つきみぐー',
358             'view_count': int,
359         },
360     }
361
362     def _real_extract(self, url):
363         au_id = self._match_id(url)
364
365         play_data = self._call_api('url', au_id)
366         formats = [{
367             'url': play_data['cdns'][0],
368             'filesize': int_or_none(play_data.get('size')),
369         }]
370
371         song = self._call_api('song/info', au_id)
372         title = song['title']
373         statistic = song.get('statistic') or {}
374
375         subtitles = None
376         lyric = song.get('lyric')
377         if lyric:
378             subtitles = {
379                 'origin': [{
380                     'url': lyric,
381                 }]
382             }
383
384         return {
385             'id': au_id,
386             'title': title,
387             'formats': formats,
388             'artist': song.get('author'),
389             'comment_count': int_or_none(statistic.get('comment')),
390             'description': song.get('intro'),
391             'duration': int_or_none(song.get('duration')),
392             'subtitles': subtitles,
393             'thumbnail': song.get('cover'),
394             'timestamp': int_or_none(song.get('passtime')),
395             'uploader': song.get('uname'),
396             'view_count': int_or_none(statistic.get('play')),
397         }
398
399
400 class BilibiliAudioAlbumIE(BilibiliAudioBaseIE):
401     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/am(?P<id>\d+)'
402     _TEST = {
403         'url': 'https://www.bilibili.com/audio/am10624',
404         'info_dict': {
405             'id': '10624',
406             'title': '每日新曲推荐(每日11:00更新)',
407             'description': '每天11:00更新,为你推送最新音乐',
408         },
409         'playlist_count': 19,
410     }
411
412     def _real_extract(self, url):
413         am_id = self._match_id(url)
414
415         songs = self._call_api(
416             'song/of-menu', am_id, {'sid': am_id, 'pn': 1, 'ps': 100})['data']
417
418         entries = []
419         for song in songs:
420             sid = str_or_none(song.get('id'))
421             if not sid:
422                 continue
423             entries.append(self.url_result(
424                 'https://www.bilibili.com/audio/au' + sid,
425                 BilibiliAudioIE.ie_key(), sid))
426
427         if entries:
428             album_data = self._call_api('menu/info', am_id) or {}
429             album_title = album_data.get('title')
430             if album_title:
431                 for entry in entries:
432                     entry['album'] = album_title
433                 return self.playlist_result(
434                     entries, am_id, album_title, album_data.get('intro'))
435
436         return self.playlist_result(entries, am_id)
437
438
439 class BiliBiliPlayerIE(InfoExtractor):
440     _VALID_URL = r'https?://player\.bilibili\.com/player\.html\?.*?\baid=(?P<id>\d+)'
441     _TEST = {
442         'url': 'http://player.bilibili.com/player.html?aid=92494333&cid=157926707&page=1',
443         'only_matching': True,
444     }
445
446     def _real_extract(self, url):
447         video_id = self._match_id(url)
448         return self.url_result(
449             'http://www.bilibili.tv/video/av%s/' % video_id,
450             ie=BiliBiliIE.ie_key(), video_id=video_id)