_ Git - youtube-dl/blob - youtube_dl/extractor/bilibili.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import itertools
   6 import json
   7 import xml.etree.ElementTree as ET
   8
   9 from .common import InfoExtractor
  10 from ..utils import (
  11     int_or_none,
  12     unified_strdate,
  13     ExtractorError,
  14 )
  15
  16
  17 class BiliBiliIE(InfoExtractor):
  18     _VALID_URL = r'http://www\.bilibili\.(?:tv|com)/video/av(?P<id>[0-9]+)/'
  19
  20     _TESTS = [{
  21         'url': 'http://www.bilibili.tv/video/av1074402/',
  22         'md5': '2c301e4dab317596e837c3e7633e7d86',
  23         'info_dict': {
  24             'id': '1074402_part1',
  25             'ext': 'flv',
  26             'title': '【金坷垃】金泡沫',
  27             'duration': 308,
  28             'upload_date': '20140420',
  29             'thumbnail': 're:^https?://.+\.jpg',
  30         },
  31     }, {
  32         'url': 'http://www.bilibili.com/video/av1041170/',
  33         'info_dict': {
  34             'id': '1041170',
  35             'title': '【BD1080P】刀语【诸神&异域】',
  36         },
  37         'playlist_count': 9,
  38     }]
  39
  40     def _real_extract(self, url):
  41         video_id = self._match_id(url)
  42         webpage = self._download_webpage(url, video_id)
  43
  44         if self._search_regex(r'(此视频不存在或被删除)', webpage, 'error message', default=None):
  45             raise ExtractorError('The video does not exist or was deleted', expected=True)
  46         video_code = self._search_regex(
  47             r'(?s)<div itemprop="video".*?>(.*?)</div>', webpage, 'video code')
  48
  49         title = self._html_search_meta(
  50             'media:title', video_code, 'title', fatal=True)
  51         duration_str = self._html_search_meta(
  52             'duration', video_code, 'duration')
  53         if duration_str is None:
  54             duration = None
  55         else:
  56             duration_mobj = re.match(
  57                 r'^T(?:(?P<hours>[0-9]+)H)?(?P<minutes>[0-9]+)M(?P<seconds>[0-9]+)S$',
  58                 duration_str)
  59             duration = (
  60                 int_or_none(duration_mobj.group('hours'), default=0) * 3600 +
  61                 int(duration_mobj.group('minutes')) * 60 +
  62                 int(duration_mobj.group('seconds')))
  63         upload_date = unified_strdate(self._html_search_meta(
  64             'uploadDate', video_code, fatal=False))
  65         thumbnail = self._html_search_meta(
  66             'thumbnailUrl', video_code, 'thumbnail', fatal=False)
  67
  68         cid = self._search_regex(r'cid=(\d+)', webpage, 'cid')
  69
  70         entries = []
  71
  72         lq_page = self._download_webpage(
  73             'http://interface.bilibili.com/v_cdn_play?appkey=1&cid=%s' % cid,
  74             video_id,
  75             note='Downloading LQ video info'
  76         )
  77         try:
  78             err_info = json.loads(lq_page)
  79             raise ExtractorError(
  80                 'BiliBili said: ' + err_info['error_text'], expected=True)
  81         except ValueError:
  82             pass
  83
  84         lq_doc = ET.fromstring(lq_page)
  85         lq_durls = lq_doc.findall('./durl')
  86
  87         hq_doc = self._download_xml(
  88             'http://interface.bilibili.com/playurl?appkey=1&cid=%s' % cid,
  89             video_id,
  90             note='Downloading HQ video info',
  91             fatal=False,
  92         )
  93         if hq_doc is not False:
  94             hq_durls = hq_doc.findall('./durl')
  95             assert len(lq_durls) == len(hq_durls)
  96         else:
  97             hq_durls = itertools.repeat(None)
  98
  99         i = 1
 100         for lq_durl, hq_durl in zip(lq_durls, hq_durls):
 101             formats = [{
 102                 'format_id': 'lq',
 103                 'quality': 1,
 104                 'url': lq_durl.find('./url').text,
 105                 'filesize': int_or_none(
 106                     lq_durl.find('./size'), get_attr='text'),
 107             }]
 108             if hq_durl is not None:
 109                 formats.append({
 110                     'format_id': 'hq',
 111                     'quality': 2,
 112                     'ext': 'flv',
 113                     'url': hq_durl.find('./url').text,
 114                     'filesize': int_or_none(
 115                         hq_durl.find('./size'), get_attr='text'),
 116                 })
 117             self._sort_formats(formats)
 118
 119             entries.append({
 120                 'id': '%s_part%d' % (video_id, i),
 121                 'title': title,
 122                 'formats': formats,
 123                 'duration': duration,
 124                 'upload_date': upload_date,
 125                 'thumbnail': thumbnail,
 126             })
 127
 128             i += 1
 129
 130         return {
 131             '_type': 'multi_video',
 132             'entries': entries,
 133             'id': video_id,
 134             'title': title
 135         }