Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/tumblr.py

   1 import re
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     ExtractorError,
   6 )
   7
   8
   9 class TumblrIE(InfoExtractor):
  10     _VALID_URL = r'http://(?P<blog_name>.*?)\.tumblr\.com/((post)|(video))/(?P<id>\d*)/(.*?)'
  11
  12     def _real_extract(self, url):
  13         m_url = re.match(self._VALID_URL, url)
  14         video_id = m_url.group('id')
  15         blog = m_url.group('blog_name')
  16
  17         url = 'http://%s.tumblr.com/post/%s/' % (blog, video_id)
  18         webpage = self._download_webpage(url, video_id)
  19
  20         re_video = r'src=\\x22(?P<video_url>http://%s\.tumblr\.com/video_file/%s/(.*?))\\x22 type=\\x22video/(?P<ext>.*?)\\x22' % (blog, video_id)
  21         video = re.search(re_video, webpage)
  22         if video is None:
  23            raise ExtractorError(u'Unable to extract video')
  24         video_url = video.group('video_url')
  25         ext = video.group('ext')
  26
  27         video_thumbnail = self._search_regex(r'posters(.*?)\[\\x22(?P<thumb>.*?)\\x22',
  28             webpage, u'thumbnail', fatal=False)  # We pick the first poster
  29         if video_thumbnail: video_thumbnail = video_thumbnail.replace('\\', '')
  30
  31         # The only place where you can get a title, it's not complete,
  32         # but searching in other places doesn't work for all videos
  33         video_title = self._html_search_regex(r'<title>(?P<title>.*?)</title>',
  34             webpage, u'title', flags=re.DOTALL)
  35
  36         return [{'id': video_id,
  37                  'url': video_url,
  38                  'title': video_title,
  39                  'thumbnail': video_thumbnail,
  40                  'ext': ext
  41                  }]