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     strip_jsonp,
  19     unified_timestamp,
  20     unsmuggle_url,
  21     urlencode_postdata,
  22 )
  23
  24
  25 class BiliBiliIE(InfoExtractor):
  26     _VALID_URL = r'https?://(?:www\.|bangumi\.|)bilibili\.(?:tv|com)/(?:video/av|anime/(?P<anime_id>\d+)/play#)(?P<id>\d+)'
  27
  28     _TESTS = [{
  29         'url': 'http://www.bilibili.tv/video/av1074402/',
  30         'md5': '9fa226fe2b8a9a4d5a69b4c6a183417e',
  31         'info_dict': {
  32             'id': '1074402',
  33             'ext': 'mp4',
  34             'title': '【金坷垃】金泡沫',
  35             'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
  36             'duration': 308.315,
  37             'timestamp': 1398012660,
  38             'upload_date': '20140420',
  39             'thumbnail': r're:^https?://.+\.jpg',
  40             'uploader': '菊子桑',
  41             'uploader_id': '156160',
  42         },
  43     }, {
  44         # Tested in BiliBiliBangumiIE
  45         'url': 'http://bangumi.bilibili.com/anime/1869/play#40062',
  46         'only_matching': True,
  47     }, {
  48         'url': 'http://bangumi.bilibili.com/anime/5802/play#100643',
  49         'md5': '3f721ad1e75030cc06faf73587cfec57',
  50         'info_dict': {
  51             'id': '100643',
  52             'ext': 'mp4',
  53             'title': 'CHAOS;CHILD',
  54             '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...',
  55         },
  56         'skip': 'Geo-restricted to China',
  57     }]
  58
  59     _APP_KEY = '84956560bc028eb7'
  60     _BILIBILI_KEY = '94aba54af9065f71de72f5508f1cd42e'
  61
  62     def _report_error(self, result):
  63         if 'message' in result:
  64             raise ExtractorError('%s said: %s' % (self.IE_NAME, result['message']), expected=True)
  65         elif 'code' in result:
  66             raise ExtractorError('%s returns error %d' % (self.IE_NAME, result['code']), expected=True)
  67         else:
  68             raise ExtractorError('Can\'t extract Bangumi episode ID')
  69
  70     def _real_extract(self, url):
  71         url, smuggled_data = unsmuggle_url(url, {})
  72
  73         mobj = re.match(self._VALID_URL, url)
  74         video_id = mobj.group('id')
  75         anime_id = mobj.group('anime_id')
  76         webpage = self._download_webpage(url, video_id)
  77
  78         if 'anime/' not in url:
  79             cid = compat_parse_qs(self._search_regex(
  80                 [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
  81                  r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
  82                 webpage, 'player parameters'))['cid'][0]
  83         else:
  84             if 'no_bangumi_tip' not in smuggled_data:
  85                 self.to_screen('Downloading episode %s. To download all videos in anime %s, re-run youtube-dl with %s' % (
  86                     video_id, anime_id, compat_urlparse.urljoin(url, '//bangumi.bilibili.com/anime/%s' % anime_id)))
  87             headers = {
  88                 'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
  89             }
  90             headers.update(self.geo_verification_headers())
  91
  92             js = self._download_json(
  93                 'http://bangumi.bilibili.com/web_api/get_source', video_id,
  94                 data=urlencode_postdata({'episode_id': video_id}),
  95                 headers=headers)
  96             if 'result' not in js:
  97                 self._report_error(js)
  98             cid = js['result']['cid']
  99
 100         payload = 'appkey=%s&cid=%s&otype=json&quality=2&type=mp4' % (self._APP_KEY, cid)
 101         sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()
 102
 103         video_info = self._download_json(
 104             'http://interface.bilibili.com/playurl?%s&sign=%s' % (payload, sign),
 105             video_id, note='Downloading video info page',
 106             headers=self.geo_verification_headers())
 107
 108         if 'durl' not in video_info:
 109             self._report_error(video_info)
 110
 111         entries = []
 112
 113         for idx, durl in enumerate(video_info['durl']):
 114             formats = [{
 115                 'url': durl['url'],
 116                 'filesize': int_or_none(durl['size']),
 117             }]
 118             for backup_url in durl.get('backup_url', []):
 119                 formats.append({
 120                     'url': backup_url,
 121                     # backup URLs have lower priorities
 122                     'preference': -2 if 'hd.mp4' in backup_url else -3,
 123                 })
 124
 125             for a_format in formats:
 126                 a_format.setdefault('http_headers', {}).update({
 127                     'Referer': url,
 128                 })
 129
 130             self._sort_formats(formats)
 131
 132             entries.append({
 133                 'id': '%s_part%s' % (video_id, idx),
 134                 'duration': float_or_none(durl.get('length'), 1000),
 135                 'formats': formats,
 136             })
 137
 138         title = self._html_search_regex('<h1[^>]+title="([^"]+)">', webpage, 'title')
 139         description = self._html_search_meta('description', webpage)
 140         timestamp = unified_timestamp(self._html_search_regex(
 141             r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time', default=None))
 142         thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
 143
 144         # TODO 'view_count' requires deobfuscating Javascript
 145         info = {
 146             'id': video_id,
 147             'title': title,
 148             'description': description,
 149             'timestamp': timestamp,
 150             'thumbnail': thumbnail,
 151             'duration': float_or_none(video_info.get('timelength'), scale=1000),
 152         }
 153
 154         uploader_mobj = re.search(
 155             r'<a[^>]+href="(?:https?:)?//space\.bilibili\.com/(?P<id>\d+)"[^>]+title="(?P<name>[^"]+)"',
 156             webpage)
 157         if uploader_mobj:
 158             info.update({
 159                 'uploader': uploader_mobj.group('name'),
 160                 'uploader_id': uploader_mobj.group('id'),
 161             })
 162
 163         for entry in entries:
 164             entry.update(info)
 165
 166         if len(entries) == 1:
 167             return entries[0]
 168         else:
 169             for idx, entry in enumerate(entries):
 170                 entry['id'] = '%s_part%d' % (video_id, (idx + 1))
 171
 172             return {
 173                 '_type': 'multi_video',
 174                 'id': video_id,
 175                 'title': title,
 176                 'description': description,
 177                 'entries': entries,
 178             }
 179
 180
 181 class BiliBiliBangumiIE(InfoExtractor):
 182     _VALID_URL = r'https?://bangumi\.bilibili\.com/anime/(?P<id>\d+)'
 183
 184     IE_NAME = 'bangumi.bilibili.com'
 185     IE_DESC = 'BiliBili番剧'
 186
 187     _TESTS = [{
 188         'url': 'http://bangumi.bilibili.com/anime/1869',
 189         'info_dict': {
 190             'id': '1869',
 191             'title': '混沌武士',
 192             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
 193         },
 194         'playlist_count': 26,
 195     }, {
 196         'url': 'http://bangumi.bilibili.com/anime/1869',
 197         'info_dict': {
 198             'id': '1869',
 199             'title': '混沌武士',
 200             'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
 201         },
 202         'playlist': [{
 203             'md5': '91da8621454dd58316851c27c68b0c13',
 204             'info_dict': {
 205                 'id': '40062',
 206                 'ext': 'mp4',
 207                 'title': '混沌武士',
 208                 '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...',
 209                 'timestamp': 1414538739,
 210                 'upload_date': '20141028',
 211                 'episode': '疾风怒涛 Tempestuous Temperaments',
 212                 'episode_number': 1,
 213             },
 214         }],
 215         'params': {
 216             'playlist_items': '1',
 217         },
 218     }]
 219
 220     @classmethod
 221     def suitable(cls, url):
 222         return False if BiliBiliIE.suitable(url) else super(BiliBiliBangumiIE, cls).suitable(url)
 223
 224     def _real_extract(self, url):
 225         bangumi_id = self._match_id(url)
 226
 227         # Sometimes this API returns a JSONP response
 228         season_info = self._download_json(
 229             'http://bangumi.bilibili.com/jsonp/seasoninfo/%s.ver' % bangumi_id,
 230             bangumi_id, transform_source=strip_jsonp)['result']
 231
 232         entries = [{
 233             '_type': 'url_transparent',
 234             'url': smuggle_url(episode['webplay_url'], {'no_bangumi_tip': 1}),
 235             'ie_key': BiliBiliIE.ie_key(),
 236             'timestamp': parse_iso8601(episode.get('update_time'), delimiter=' '),
 237             'episode': episode.get('index_title'),
 238             'episode_number': int_or_none(episode.get('index')),
 239         } for episode in season_info['episodes']]
 240
 241         entries = sorted(entries, key=lambda entry: entry.get('episode_number'))
 242
 243         return self.playlist_result(
 244             entries, bangumi_id,
 245             season_info.get('bangumi_title'), season_info.get('evaluate'))