_ 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 json
   6 import xml.etree.ElementTree as ET
   7
   8 from .common import InfoExtractor
   9 from ..utils import (
  10     int_or_none,
  11     unescapeHTML,
  12     ExtractorError,
  13 )
  14
  15
  16 class BiliBiliIE(InfoExtractor):
  17     _VALID_URL = r'http://www\.bilibili\.(?:tv|com)/video/av(?P<id>\d+)(?:/index_(?P<page_num>\d+).html)?'
  18
  19     _TESTS = [{
  20         'url': 'http://www.bilibili.tv/video/av1074402/',
  21         'md5': '2c301e4dab317596e837c3e7633e7d86',
  22         'info_dict': {
  23             'id': '1554319',
  24             'ext': 'flv',
  25             'title': '【金坷垃】金泡沫',
  26             'duration': 308313,
  27             'upload_date': '20140420',
  28             'thumbnail': 're:^https?://.+\.jpg',
  29             'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
  30             'timestamp': 1397983878,
  31             'uploader': '菊子桑',
  32         },
  33     }, {
  34         'url': 'http://www.bilibili.com/video/av1041170/',
  35         'info_dict': {
  36             'id': '1041170',
  37             'title': '【BD1080P】刀语【诸神&异域】',
  38             'description': '这是个神奇的故事~每个人不留弹幕不给走哦~切利哦！~',
  39             'uploader': '枫叶逝去',
  40             'timestamp': 1396501299,
  41         },
  42         'playlist_count': 9,
  43     }]
  44
  45     def _real_extract(self, url):
  46         mobj = re.match(self._VALID_URL, url)
  47         video_id = mobj.group('id')
  48         page_num = mobj.group('page_num') or '1'
  49
  50         view_data = self._download_json(
  51             'http://api.bilibili.com/view?type=json&appkey=8e9fc618fbd41e28&id=%s&page=%s' % (video_id, page_num),
  52             video_id)
  53         if 'error' in view_data:
  54             raise ExtractorError('%s said: %s' % (self.IE_NAME, view_data['error']), expected=True)
  55
  56         cid = view_data['cid']
  57         title = unescapeHTML(view_data['title'])
  58
  59         page = self._download_webpage(
  60             'http://interface.bilibili.com/v_cdn_play?appkey=8e9fc618fbd41e28&cid=%s' % cid,
  61             cid,
  62             'Downloading page %s/%s' % (page_num, view_data['pages'])
  63         )
  64         try:
  65             err_info = json.loads(page)
  66             raise ExtractorError(
  67                 'BiliBili said: ' + err_info['error_text'], expected=True)
  68         except ValueError:
  69             pass
  70
  71         doc = ET.fromstring(page)
  72
  73         entries = []
  74
  75         for durl in doc.findall('./durl'):
  76             size = durl.find('./filesize|./size')
  77             formats = [{
  78                 'url': durl.find('./url').text,
  79                 'filesize': int_or_none(size.text) if size else None,
  80                 'ext': 'flv',
  81             }]
  82             backup_urls = durl.find('./backup_url')
  83             if backup_urls is not None:
  84                 for backup_url in backup_urls.findall('./url'):
  85                     formats.append({'url': backup_url.text})
  86             formats.reverse()
  87
  88             entries.append({
  89                 'id': '%s_part%s' % (cid, durl.find('./order').text),
  90                 'title': title,
  91                 'duration': int_or_none(durl.find('./length').text) // 1000,
  92                 'formats': formats,
  93             })
  94
  95         info = {
  96             'id': str(cid),
  97             'title': title,
  98             'description': view_data.get('description'),
  99             'thumbnail': view_data.get('pic'),
 100             'uploader': view_data.get('author'),
 101             'timestamp': int_or_none(view_data.get('created')),
 102             'view_count': view_data.get('play'),
 103             'duration': int_or_none(doc.find('./timelength').text),
 104         }
 105
 106         if len(entries) == 1:
 107             entries[0].update(info)
 108             return entries[0]
 109         else:
 110             info.update({
 111                 '_type': 'multi_video',
 112                 'id': video_id,
 113                 'entries': entries,
 114             })
 115             return info