2 from __future__ import unicode_literals
8 from .common import InfoExtractor
10 compat_etree_fromstring,
13 compat_xml_parse_error,
23 class BiliBiliIE(InfoExtractor):
24 _VALID_URL = r'https?://www\.bilibili\.(?:tv|com)/video/av(?P<id>\d+)'
27 'url': 'http://www.bilibili.tv/video/av1074402/',
28 'md5': '5f7d29e1a2872f3df0cf76b1f87d3788',
33 'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
35 'timestamp': 1398012660,
36 'upload_date': '20140420',
37 'thumbnail': 're:^https?://.+\.jpg',
39 'uploader_id': '156160',
42 'url': 'http://www.bilibili.com/video/av1041170/',
45 'title': '【BD1080P】刀语【诸神&异域】',
46 'description': '这是个神奇的故事~每个人不留弹幕不给走哦~切利哦!~',
50 'url': 'http://www.bilibili.com/video/av4808130/',
53 'title': '【长篇】哆啦A梦443【钉铛】',
54 'description': '(2016.05.27)来组合客人的脸吧&amp;寻母六千里锭 抱歉,又轮到周日上班现在才到家 封面www.pixiv.net/member_illust.php?mode=medium&amp;illust_id=56912929',
57 'md5': '55cdadedf3254caaa0d5d27cf20a8f9c',
59 'id': '4808130_part1',
61 'title': '【长篇】哆啦A梦443【钉铛】',
62 'description': '(2016.05.27)来组合客人的脸吧&amp;寻母六千里锭 抱歉,又轮到周日上班现在才到家 封面www.pixiv.net/member_illust.php?mode=medium&amp;illust_id=56912929',
63 'timestamp': 1464564180,
64 'upload_date': '20160529',
66 'uploader_id': '151066',
69 'md5': '926f9f67d0c482091872fbd8eca7ea3d',
71 'id': '4808130_part2',
73 'title': '【长篇】哆啦A梦443【钉铛】',
74 'description': '(2016.05.27)来组合客人的脸吧&amp;寻母六千里锭 抱歉,又轮到周日上班现在才到家 封面www.pixiv.net/member_illust.php?mode=medium&amp;illust_id=56912929',
75 'timestamp': 1464564180,
76 'upload_date': '20160529',
78 'uploader_id': '151066',
81 'md5': '4b7b225b968402d7c32348c646f1fd83',
83 'id': '4808130_part3',
85 'title': '【长篇】哆啦A梦443【钉铛】',
86 'description': '(2016.05.27)来组合客人的脸吧&amp;寻母六千里锭 抱歉,又轮到周日上班现在才到家 封面www.pixiv.net/member_illust.php?mode=medium&amp;illust_id=56912929',
87 'timestamp': 1464564180,
88 'upload_date': '20160529',
90 'uploader_id': '151066',
93 'md5': '7b795e214166501e9141139eea236e91',
95 'id': '4808130_part4',
97 'title': '【长篇】哆啦A梦443【钉铛】',
98 'description': '(2016.05.27)来组合客人的脸吧&amp;寻母六千里锭 抱歉,又轮到周日上班现在才到家 封面www.pixiv.net/member_illust.php?mode=medium&amp;illust_id=56912929',
99 'timestamp': 1464564180,
100 'upload_date': '20160529',
102 'uploader_id': '151066',
106 # Missing upload time
107 'url': 'http://www.bilibili.com/video/av1867637/',
111 'title': '【HDTV】【喜剧】岳父岳母真难当 (2014)【法国票房冠军】',
112 'description': '一个信奉天主教的法国旧式传统资产阶级家庭中有四个女儿。三个女儿却分别找了阿拉伯、犹太、中国丈夫,老夫老妻唯独期盼剩下未嫁的小女儿能找一个信奉天主教的法国白人,结果没想到小女儿找了一位非裔黑人……【这次应该不会跳帧了】',
114 'uploader_id': '610729',
117 # Just to test metadata extraction
118 'skip_download': True,
120 'expected_warnings': ['upload time'],
123 # BiliBili blocks keys from time to time. The current key is extracted from
125 # TODO: find the sign algorithm used in the flash player
126 _APP_KEY = '86385cdc024c0f6c'
128 def _real_extract(self, url):
129 mobj = re.match(self._VALID_URL, url)
130 video_id = mobj.group('id')
132 webpage = self._download_webpage(url, video_id)
134 params = compat_parse_qs(self._search_regex(
135 [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
136 r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
137 webpage, 'player parameters'))
138 cid = params['cid'][0]
140 info_xml_str = self._download_webpage(
141 'http://interface.bilibili.com/v_cdn_play',
142 cid, query={'appkey': self._APP_KEY, 'cid': cid},
143 note='Downloading video info page')
149 info_xml = compat_etree_fromstring(info_xml_str.encode('utf-8'))
150 except compat_xml_parse_error:
151 info_json = self._parse_json(info_xml_str, video_id, fatal=False)
152 err_msg = (info_json or {}).get('error_text')
154 err_msg = xpath_text(info_xml, './message')
156 if info_xml is not None:
157 durls = info_xml.findall('./durl')
160 raise ExtractorError('%s said: %s' % (self.IE_NAME, err_msg), expected=True)
162 raise ExtractorError('No videos found!')
167 size = xpath_text(durl, ['./filesize', './size'])
169 'url': durl.find('./url').text,
170 'filesize': int_or_none(size),
172 for backup_url in durl.findall('./backup_url/url'):
174 'url': backup_url.text,
175 # backup URLs have lower priorities
176 'preference': -2 if 'hd.mp4' in backup_url.text else -3,
179 self._sort_formats(formats)
182 'id': '%s_part%s' % (cid, xpath_text(durl, './order')),
183 'duration': int_or_none(xpath_text(durl, './length'), 1000),
187 title = self._html_search_regex('<h1[^>]+title="([^"]+)">', webpage, 'title')
188 description = self._html_search_meta('description', webpage)
189 datetime_str = self._html_search_regex(
190 r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time', fatal=False)
193 timestamp = calendar.timegm(datetime.datetime.strptime(datetime_str, '%Y-%m-%dT%H:%M').timetuple())
195 # TODO 'view_count' requires deobfuscating Javascript
197 'id': compat_str(cid),
199 'description': description,
200 'timestamp': timestamp,
201 'thumbnail': self._html_search_meta('thumbnailUrl', webpage),
202 'duration': float_or_none(xpath_text(info_xml, './timelength'), scale=1000),
205 uploader_mobj = re.search(
206 r'<a[^>]+href="https?://space\.bilibili\.com/(?P<id>\d+)"[^>]+title="(?P<name>[^"]+)"',
210 'uploader': uploader_mobj.group('name'),
211 'uploader_id': uploader_mobj.group('id'),
214 for entry in entries:
217 if len(entries) == 1:
220 for idx, entry in enumerate(entries):
221 entry['id'] = '%s_part%d' % (video_id, (idx + 1))
224 '_type': 'multi_video',
227 'description': description,