]> git.bitcoin.ninja Git - youtube-dl/blob - youtube_dl/extractor/bilibili.py
[pornhub] Improve locked videos detection (closes #22449, closes #22780)
[youtube-dl] / 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'https?://(?:www\.|bangumi\.|)bilibili\.(?:tv|com)/(?:video/av|anime/(?P<anime_id>\d+)/play#)(?P<id>\d+)'
28
29     _TESTS = [{
30         'url': 'http://www.bilibili.tv/video/av1074402/',
31         'md5': '5f7d29e1a2872f3df0cf76b1f87d3788',
32         'info_dict': {
33             'id': '1074402',
34             'ext': 'flv',
35             'title': '【金坷垃】金泡沫',
36             'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
37             'duration': 308.067,
38             'timestamp': 1398012678,
39             'upload_date': '20140420',
40             'thumbnail': r're:^https?://.+\.jpg',
41             'uploader': '菊子桑',
42             'uploader_id': '156160',
43         },
44     }, {
45         # Tested in BiliBiliBangumiIE
46         'url': 'http://bangumi.bilibili.com/anime/1869/play#40062',
47         'only_matching': True,
48     }, {
49         'url': 'http://bangumi.bilibili.com/anime/5802/play#100643',
50         'md5': '3f721ad1e75030cc06faf73587cfec57',
51         'info_dict': {
52             'id': '100643',
53             'ext': 'mp4',
54             'title': 'CHAOS;CHILD',
55             '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...',
56         },
57         'skip': 'Geo-restricted to China',
58     }, {
59         # Title with double quotes
60         'url': 'http://www.bilibili.com/video/av8903802/',
61         'info_dict': {
62             'id': '8903802',
63             'title': '阿滴英文|英文歌分享#6 "Closer',
64             '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',
65         },
66         'playlist': [{
67             'info_dict': {
68                 'id': '8903802_part1',
69                 'ext': 'flv',
70                 'title': '阿滴英文|英文歌分享#6 "Closer',
71                 'description': 'md5:3b1b9e25b78da4ef87e9b548b88ee76a',
72                 'uploader': '阿滴英文',
73                 'uploader_id': '65880958',
74                 'timestamp': 1488382634,
75                 'upload_date': '20170301',
76             },
77             'params': {
78                 'skip_download': True,  # Test metadata only
79             },
80         }, {
81             'info_dict': {
82                 'id': '8903802_part2',
83                 'ext': 'flv',
84                 'title': '阿滴英文|英文歌分享#6 "Closer',
85                 'description': 'md5:3b1b9e25b78da4ef87e9b548b88ee76a',
86                 'uploader': '阿滴英文',
87                 'uploader_id': '65880958',
88                 'timestamp': 1488382634,
89                 'upload_date': '20170301',
90             },
91             'params': {
92                 'skip_download': True,  # Test metadata only
93             },
94         }]
95     }]
96
97     _APP_KEY = 'iVGUTjsxvpLeuDCf'
98     _BILIBILI_KEY = 'aHRmhWMLkdeMuILqORnYZocwMBpMEOdt'
99
100     def _report_error(self, result):
101         if 'message' in result:
102             raise ExtractorError('%s said: %s' % (self.IE_NAME, result['message']), expected=True)
103         elif 'code' in result:
104             raise ExtractorError('%s returns error %d' % (self.IE_NAME, result['code']), expected=True)
105         else:
106             raise ExtractorError('Can\'t extract Bangumi episode ID')
107
108     def _real_extract(self, url):
109         url, smuggled_data = unsmuggle_url(url, {})
110
111         mobj = re.match(self._VALID_URL, url)
112         video_id = mobj.group('id')
113         anime_id = mobj.group('anime_id')
114         webpage = self._download_webpage(url, video_id)
115
116         if 'anime/' not in url:
117             cid = self._search_regex(
118                 r'\bcid(?:["\']:|=)(\d+)', webpage, 'cid',
119                 default=None
120             ) or compat_parse_qs(self._search_regex(
121                 [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
122                  r'EmbedPlayer\([^)]+,\s*\\"([^"]+)\\"\)',
123                  r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
124                 webpage, 'player parameters'))['cid'][0]
125         else:
126             if 'no_bangumi_tip' not in smuggled_data:
127                 self.to_screen('Downloading episode %s. To download all videos in anime %s, re-run youtube-dl with %s' % (
128                     video_id, anime_id, compat_urlparse.urljoin(url, '//bangumi.bilibili.com/anime/%s' % anime_id)))
129             headers = {
130                 'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
131                 'Referer': url
132             }
133             headers.update(self.geo_verification_headers())
134
135             js = self._download_json(
136                 'http://bangumi.bilibili.com/web_api/get_source', video_id,
137                 data=urlencode_postdata({'episode_id': video_id}),
138                 headers=headers)
139             if 'result' not in js:
140                 self._report_error(js)
141             cid = js['result']['cid']
142
143         headers = {
144             'Referer': url
145         }
146         headers.update(self.geo_verification_headers())
147
148         entries = []
149
150         RENDITIONS = ('qn=80&quality=80&type=', 'quality=2&type=mp4')
151         for num, rendition in enumerate(RENDITIONS, start=1):
152             payload = 'appkey=%s&cid=%s&otype=json&%s' % (self._APP_KEY, cid, rendition)
153             sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()
154
155             video_info = self._download_json(
156                 'http://interface.bilibili.com/v2/playurl?%s&sign=%s' % (payload, sign),
157                 video_id, note='Downloading video info page',
158                 headers=headers, fatal=num == len(RENDITIONS))
159
160             if not video_info:
161                 continue
162
163             if 'durl' not in video_info:
164                 if num < len(RENDITIONS):
165                     continue
166                 self._report_error(video_info)
167
168             for idx, durl in enumerate(video_info['durl']):
169                 formats = [{
170                     'url': durl['url'],
171                     'filesize': int_or_none(durl['size']),
172                 }]
173                 for backup_url in durl.get('backup_url', []):
174                     formats.append({
175                         'url': backup_url,
176                         # backup URLs have lower priorities
177                         'preference': -2 if 'hd.mp4' in backup_url else -3,
178                     })
179
180                 for a_format in formats:
181                     a_format.setdefault('http_headers', {}).update({
182                         'Referer': url,
183                     })
184
185                 self._sort_formats(formats)
186
187                 entries.append({
188                     'id': '%s_part%s' % (video_id, idx),
189                     'duration': float_or_none(durl.get('length'), 1000),
190                     'formats': formats,
191                 })
192             break
193
194         title = self._html_search_regex(
195             ('<h1[^>]+\btitle=(["\'])(?P<title>(?:(?!\1).)+)\1',
196              '(?s)<h1[^>]*>(?P<title>.+?)</h1>'), webpage, 'title',
197             group='title')
198         description = self._html_search_meta('description', webpage)
199         timestamp = unified_timestamp(self._html_search_regex(
200             r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time',
201             default=None) or self._html_search_meta(
202             'uploadDate', webpage, 'timestamp', default=None))
203         thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
204
205         # TODO 'view_count' requires deobfuscating Javascript
206         info = {
207             'id': video_id,
208             'title': title,
209             'description': description,
210             'timestamp': timestamp,
211             'thumbnail': thumbnail,
212             'duration': float_or_none(video_info.get('timelength'), scale=1000),
213         }
214
215         uploader_mobj = re.search(
216             r'<a[^>]+href="(?:https?:)?//space\.bilibili\.com/(?P<id>\d+)"[^>]*>(?P<name>[^<]+)',
217             webpage)
218         if uploader_mobj:
219             info.update({
220                 'uploader': uploader_mobj.group('name'),
221                 'uploader_id': uploader_mobj.group('id'),
222             })
223         if not info.get('uploader'):
224             info['uploader'] = self._html_search_meta(
225                 'author', webpage, 'uploader', default=None)
226
227         for entry in entries:
228             entry.update(info)
229
230         if len(entries) == 1:
231             return entries[0]
232         else:
233             for idx, entry in enumerate(entries):
234                 entry['id'] = '%s_part%d' % (video_id, (idx + 1))
235
236             return {
237                 '_type': 'multi_video',
238                 'id': video_id,
239                 'title': title,
240                 'description': description,
241                 'entries': entries,
242             }
243
244
245 class BiliBiliBangumiIE(InfoExtractor):
246     _VALID_URL = r'https?://bangumi\.bilibili\.com/anime/(?P<id>\d+)'
247
248     IE_NAME = 'bangumi.bilibili.com'
249     IE_DESC = 'BiliBili番剧'
250
251     _TESTS = [{
252         'url': 'http://bangumi.bilibili.com/anime/1869',
253         'info_dict': {
254             'id': '1869',
255             'title': '混沌武士',
256             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
257         },
258         'playlist_count': 26,
259     }, {
260         'url': 'http://bangumi.bilibili.com/anime/1869',
261         'info_dict': {
262             'id': '1869',
263             'title': '混沌武士',
264             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
265         },
266         'playlist': [{
267             'md5': '91da8621454dd58316851c27c68b0c13',
268             'info_dict': {
269                 'id': '40062',
270                 'ext': 'mp4',
271                 'title': '混沌武士',
272                 '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...',
273                 'timestamp': 1414538739,
274                 'upload_date': '20141028',
275                 'episode': '疾风怒涛 Tempestuous Temperaments',
276                 'episode_number': 1,
277             },
278         }],
279         'params': {
280             'playlist_items': '1',
281         },
282     }]
283
284     @classmethod
285     def suitable(cls, url):
286         return False if BiliBiliIE.suitable(url) else super(BiliBiliBangumiIE, cls).suitable(url)
287
288     def _real_extract(self, url):
289         bangumi_id = self._match_id(url)
290
291         # Sometimes this API returns a JSONP response
292         season_info = self._download_json(
293             'http://bangumi.bilibili.com/jsonp/seasoninfo/%s.ver' % bangumi_id,
294             bangumi_id, transform_source=strip_jsonp)['result']
295
296         entries = [{
297             '_type': 'url_transparent',
298             'url': smuggle_url(episode['webplay_url'], {'no_bangumi_tip': 1}),
299             'ie_key': BiliBiliIE.ie_key(),
300             'timestamp': parse_iso8601(episode.get('update_time'), delimiter=' '),
301             'episode': episode.get('index_title'),
302             'episode_number': int_or_none(episode.get('index')),
303         } for episode in season_info['episodes']]
304
305         entries = sorted(entries, key=lambda entry: entry.get('episode_number'))
306
307         return self.playlist_result(
308             entries, bangumi_id,
309             season_info.get('bangumi_title'), season_info.get('evaluate'))
310
311
312 class BilibiliAudioBaseIE(InfoExtractor):
313     def _call_api(self, path, sid, query=None):
314         if not query:
315             query = {'sid': sid}
316         return self._download_json(
317             'https://www.bilibili.com/audio/music-service-c/web/' + path,
318             sid, query=query)['data']
319
320
321 class BilibiliAudioIE(BilibiliAudioBaseIE):
322     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/au(?P<id>\d+)'
323     _TEST = {
324         'url': 'https://www.bilibili.com/audio/au1003142',
325         'md5': 'fec4987014ec94ef9e666d4d158ad03b',
326         'info_dict': {
327             'id': '1003142',
328             'ext': 'm4a',
329             'title': '【tsukimi】YELLOW / 神山羊',
330             'artist': 'tsukimi',
331             'comment_count': int,
332             'description': 'YELLOW的mp3版!',
333             'duration': 183,
334             'subtitles': {
335                 'origin': [{
336                     'ext': 'lrc',
337                 }],
338             },
339             'thumbnail': r're:^https?://.+\.jpg',
340             'timestamp': 1564836614,
341             'upload_date': '20190803',
342             'uploader': 'tsukimi-つきみぐー',
343             'view_count': int,
344         },
345     }
346
347     def _real_extract(self, url):
348         au_id = self._match_id(url)
349
350         play_data = self._call_api('url', au_id)
351         formats = [{
352             'url': play_data['cdns'][0],
353             'filesize': int_or_none(play_data.get('size')),
354         }]
355
356         song = self._call_api('song/info', au_id)
357         title = song['title']
358         statistic = song.get('statistic') or {}
359
360         subtitles = None
361         lyric = song.get('lyric')
362         if lyric:
363             subtitles = {
364                 'origin': [{
365                     'url': lyric,
366                 }]
367             }
368
369         return {
370             'id': au_id,
371             'title': title,
372             'formats': formats,
373             'artist': song.get('author'),
374             'comment_count': int_or_none(statistic.get('comment')),
375             'description': song.get('intro'),
376             'duration': int_or_none(song.get('duration')),
377             'subtitles': subtitles,
378             'thumbnail': song.get('cover'),
379             'timestamp': int_or_none(song.get('passtime')),
380             'uploader': song.get('uname'),
381             'view_count': int_or_none(statistic.get('play')),
382         }
383
384
385 class BilibiliAudioAlbumIE(BilibiliAudioBaseIE):
386     _VALID_URL = r'https?://(?:www\.)?bilibili\.com/audio/am(?P<id>\d+)'
387     _TEST = {
388         'url': 'https://www.bilibili.com/audio/am10624',
389         'info_dict': {
390             'id': '10624',
391             'title': '每日新曲推荐(每日11:00更新)',
392             'description': '每天11:00更新,为你推送最新音乐',
393         },
394         'playlist_count': 19,
395     }
396
397     def _real_extract(self, url):
398         am_id = self._match_id(url)
399
400         songs = self._call_api(
401             'song/of-menu', am_id, {'sid': am_id, 'pn': 1, 'ps': 100})['data']
402
403         entries = []
404         for song in songs:
405             sid = str_or_none(song.get('id'))
406             if not sid:
407                 continue
408             entries.append(self.url_result(
409                 'https://www.bilibili.com/audio/au' + sid,
410                 BilibiliAudioIE.ie_key(), sid))
411
412         if entries:
413             album_data = self._call_api('menu/info', am_id) or {}
414             album_title = album_data.get('title')
415             if album_title:
416                 for entry in entries:
417                     entry['album'] = album_title
418                 return self.playlist_result(
419                     entries, am_id, album_title, album_data.get('intro'))
420
421         return self.playlist_result(entries, am_id)