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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import xml.etree.ElementTree as ET
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     int_or_none,
  10     ExtractorError,
  11 )
  12
  13
  14 class BiliBiliIE(InfoExtractor):
  15     _VALID_URL = r'http://www\.bilibili\.(?:tv|com)/video/av(?P<id>[0-9]+)/'
  16
  17     _TESTS = [{
  18         'url': 'http://www.bilibili.tv/video/av1074402/',
  19         'md5': '2c301e4dab317596e837c3e7633e7d86',
  20         'info_dict': {
  21             'id': '1554319',
  22             'ext': 'flv',
  23             'title': '【金坷垃】金泡沫',
  24             'duration': 308313,
  25             'upload_date': '20140420',
  26             'thumbnail': 're:^https?://.+\.jpg',
  27             'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
  28             'timestamp': 1397983878,
  29             'uploader': '菊子桑',
  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': 12,
  38     }]
  39
  40     def _extract_video_info(self, cid, view_data, page_num=1, num_pages=1):
  41         title = view_data['title']
  42
  43         page = self._download_webpage(
  44             'http://interface.bilibili.com/v_cdn_play?appkey=8e9fc618fbd41e28&cid=%s' % cid,
  45             cid,
  46             'Downloading page %d/%d' % (page_num, num_pages)
  47         )
  48         try:
  49             err_info = json.loads(page)
  50             raise ExtractorError(
  51                 'BiliBili said: ' + err_info['error_text'], expected=True)
  52         except ValueError:
  53             pass
  54
  55         doc = ET.fromstring(page)
  56         durls = doc.findall('./durl')
  57
  58         entries = []
  59
  60         for durl in durls:
  61             formats = []
  62             backup_url = durl.find('./backup_url')
  63             if backup_url is not None:
  64                 formats.append({'url': backup_url.find('./url').text})
  65             size = durl.find('./filesize|./size')
  66             formats.append({
  67                 'url': durl.find('./url').text,
  68                 'filesize': int_or_none(size.text) if size else None,
  69                 'ext': 'flv',
  70             })
  71             entries.append({
  72                 'id': '%s_part%s' % (cid, durl.find('./order').text),
  73                 'title': title,
  74                 'duration': int_or_none(durl.find('./length').text) // 1000,
  75                 'formats': formats,
  76             })
  77
  78         info = {
  79             'id': cid,
  80             'title': title,
  81             'description': view_data.get('description'),
  82             'thumbnail': view_data.get('pic'),
  83             'uploader': view_data.get('author'),
  84             'timestamp': int_or_none(view_data.get('created')),
  85             'view_count': view_data.get('play'),
  86             'duration': int_or_none(doc.find('./timelength').text),
  87         }
  88
  89         if len(entries) == 1:
  90             entries[0].update(info)
  91             return entries[0]
  92         else:
  93             info.update({
  94                 '_type': 'multi_video',
  95                 'entries': entries,
  96             })
  97             return info
  98
  99     def _real_extract(self, url):
 100         video_id = self._match_id(url)
 101         view_data = self._download_json('http://api.bilibili.com/view?type=json&appkey=8e9fc618fbd41e28&id=%s' % video_id, video_id)
 102
 103         num_pages = int_or_none(view_data['pages'])
 104         if num_pages > 1:
 105             play_list_title = view_data['title']
 106             page_list = self._download_json('http://www.bilibili.com/widget/getPageList?aid=%s' % video_id, video_id, 'Downloading page list metadata')
 107             entries = []
 108             for page in page_list:
 109                 view_data['title'] = page['pagename']
 110                 entries.append(self._extract_video_info(str(page['cid']), view_data, page['page'], num_pages))
 111             return self.playlist_result(entries, video_id, play_list_title, view_data.get('description'))
 112         else:
 113             return self._extract_video_info(str(view_data['cid']), view_data)