Prepare for release.

[youtubedl] / youtube_dl / extractor / bilibili.py

diff --git a/youtube_dl/extractor/bilibili.py b/youtube_dl/extractor/bilibili.py
index 2103ed73aad860738bba5108ee27a86fd921d29c..2d174e6f9a81da7412cd58ac316c7b5924dcde78 100644 (file)
--- a/youtube_dl/extractor/bilibili.py
+++ b/youtube_dl/extractor/bilibili.py
@@ -1,135 +1,125 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
 # coding: utf-8
 from __future__ import unicode_literals
 

+import hashlib

 import re
 import re

-import itertools
-import json
-import xml.etree.ElementTree as ET

 
 from .common import InfoExtractor
 
 from .common import InfoExtractor

+from ..compat import compat_parse_qs

 from ..utils import (
     int_or_none,
 from ..utils import (
     int_or_none,

-    unified_strdate,
-    ExtractorError,
+    float_or_none,
+    unified_timestamp,
+    urlencode_postdata,

 )
 
 
 class BiliBiliIE(InfoExtractor):
 )
 
 
 class BiliBiliIE(InfoExtractor):

-    _VALID_URL = r'http://www\.bilibili\.(?:tv|com)/video/av(?P<id>[0-9]+)/'
+    _VALID_URL = r'https?://(?:www\.|bangumi\.|)bilibili\.(?:tv|com)/(?:video/av|anime/v/)(?P<id>\d+)'

 
 

-    _TESTS = [{
+    _TEST = {

         'url': 'http://www.bilibili.tv/video/av1074402/',
         'url': 'http://www.bilibili.tv/video/av1074402/',

-        'md5': '2c301e4dab317596e837c3e7633e7d86',
+        'md5': '9fa226fe2b8a9a4d5a69b4c6a183417e',

         'info_dict': {
         'info_dict': {

-            'id': '1074402_part1',
-            'ext': 'flv',
+            'id': '1074402',
+            'ext': 'mp4',

             'title': '【金坷垃】金泡沫',
             'title': '【金坷垃】金泡沫',

-            'duration': 308,
+            'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
+            'duration': 308.315,
+            'timestamp': 1398012660,

             'upload_date': '20140420',
             'thumbnail': 're:^https?://.+\.jpg',
             'upload_date': '20140420',
             'thumbnail': 're:^https?://.+\.jpg',

+            'uploader': '菊子桑',
+            'uploader_id': '156160',

         },
         },

-    }, {
-        'url': 'http://www.bilibili.com/video/av1041170/',
-        'info_dict': {
-            'id': '1041170',
-            'title': '【BD1080P】刀语【诸神&异域】',
-        },
-        'playlist_count': 9,
-    }]
+    }
+
+    _APP_KEY = '6f90a59ac58a4123'
+    _BILIBILI_KEY = '0bfd84cc3940035173f35e6777508326'

 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
 
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
 

-        if self._search_regex(r'(此视频不存在或被删除)', webpage, 'error message', default=None):
-            raise ExtractorError('The video does not exist or was deleted', expected=True)
-        video_code = self._search_regex(
-            r'(?s)<div itemprop="video".*?>(.*?)</div>', webpage, 'video code')
-
-        title = self._html_search_meta(
-            'media:title', video_code, 'title', fatal=True)
-        duration_str = self._html_search_meta(
-            'duration', video_code, 'duration')
-        if duration_str is None:
-            duration = None
+        if 'anime/v' not in url:
+            cid = compat_parse_qs(self._search_regex(
+                [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
+                 r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
+                webpage, 'player parameters'))['cid'][0]

         else:
         else:

-            duration_mobj = re.match(
-                r'^T(?:(?P<hours>[0-9]+)H)?(?P<minutes>[0-9]+)M(?P<seconds>[0-9]+)S$',
-                duration_str)
-            duration = (
-                int_or_none(duration_mobj.group('hours'), default=0) * 3600 +
-                int(duration_mobj.group('minutes')) * 60 +
-                int(duration_mobj.group('seconds')))
-        upload_date = unified_strdate(self._html_search_meta(
-            'uploadDate', video_code, fatal=False))
-        thumbnail = self._html_search_meta(
-            'thumbnailUrl', video_code, 'thumbnail', fatal=False)
-
-        cid = self._search_regex(r'cid=(\d+)', webpage, 'cid')
+            js = self._download_json(
+                'http://bangumi.bilibili.com/web_api/get_source', video_id,
+                data=urlencode_postdata({'episode_id': video_id}),
+                headers={'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8'})
+            cid = js['result']['cid']

 
 

-        entries = []
+        payload = 'appkey=%s&cid=%s&otype=json&quality=2&type=mp4' % (self._APP_KEY, cid)
+        sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()

 
 

-        lq_page = self._download_webpage(
-            'http://interface.bilibili.com/v_cdn_play?appkey=1&cid=%s' % cid,
-            video_id,
-            note='Downloading LQ video info'
-        )
-        try:
-            err_info = json.loads(lq_page)
-            raise ExtractorError(
-                'BiliBili said: ' + err_info['error_text'], expected=True)
-        except ValueError:
-            pass
-
-        lq_doc = ET.fromstring(lq_page)
-        lq_durls = lq_doc.findall('./durl')
-
-        hq_doc = self._download_xml(
-            'http://interface.bilibili.com/playurl?appkey=1&cid=%s' % cid,
-            video_id,
-            note='Downloading HQ video info',
-            fatal=False,
-        )
-        if hq_doc is not False:
-            hq_durls = hq_doc.findall('./durl')
-            assert len(lq_durls) == len(hq_durls)
-        else:
-            hq_durls = itertools.repeat(None)
+        video_info = self._download_json(
+            'http://interface.bilibili.com/playurl?%s&sign=%s' % (payload, sign),
+            video_id, note='Downloading video info page')
+
+        entries = []

 
 

-        i = 1
-        for lq_durl, hq_durl in zip(lq_durls, hq_durls):
+        for idx, durl in enumerate(video_info['durl']):

             formats = [{
             formats = [{

-                'format_id': 'lq',
-                'quality': 1,
-                'url': lq_durl.find('./url').text,
-                'filesize': int_or_none(
-                    lq_durl.find('./size'), get_attr='text'),
+                'url': durl['url'],
+                'filesize': int_or_none(durl['size']),

             }]
             }]

-            if hq_durl:
+            for backup_url in durl.get('backup_url', []):

                 formats.append({
                 formats.append({

-                    'format_id': 'hq',
-                    'quality': 2,
-                    'ext': 'flv',
-                    'url': hq_durl.find('./url').text,
-                    'filesize': int_or_none(
-                        hq_durl.find('./size'), get_attr='text'),
+                    'url': backup_url,
+                    # backup URLs have lower priorities
+                    'preference': -2 if 'hd.mp4' in backup_url else -3,

                 })
                 })

+

             self._sort_formats(formats)
 
             entries.append({
             self._sort_formats(formats)
 
             entries.append({

-                'id': '%s_part%d' % (video_id, i),
-                'title': title,
+                'id': '%s_part%s' % (video_id, idx),
+                'duration': float_or_none(durl.get('length'), 1000),

                 'formats': formats,
                 'formats': formats,

-                'duration': duration,
-                'upload_date': upload_date,
-                'thumbnail': thumbnail,

             })
 
             })
 

-            i += 1
+        title = self._html_search_regex('<h1[^>]+title="([^"]+)">', webpage, 'title')
+        description = self._html_search_meta('description', webpage)
+        timestamp = unified_timestamp(self._html_search_regex(
+            r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time', fatal=False))
+        thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)

 
 

-        return {
-            '_type': 'multi_video',
-            'entries': entries,
+        # TODO 'view_count' requires deobfuscating Javascript
+        info = {

             'id': video_id,
             'id': video_id,

-            'title': title
+            'title': title,
+            'description': description,
+            'timestamp': timestamp,
+            'thumbnail': thumbnail,
+            'duration': float_or_none(video_info.get('timelength'), scale=1000),

         }
         }

+
+        uploader_mobj = re.search(
+            r'<a[^>]+href="https?://space\.bilibili\.com/(?P<id>\d+)"[^>]+title="(?P<name>[^"]+)"',
+            webpage)
+        if uploader_mobj:
+            info.update({
+                'uploader': uploader_mobj.group('name'),
+                'uploader_id': uploader_mobj.group('id'),
+            })
+
+        for entry in entries:
+            entry.update(info)
+
+        if len(entries) == 1:
+            return entries[0]
+        else:
+            for idx, entry in enumerate(entries):
+                entry['id'] = '%s_part%d' % (video_id, (idx + 1))
+
+            return {
+                '_type': 'multi_video',
+                'id': video_id,
+                'title': title,
+                'description': description,
+                'entries': entries,
+            }