youtube_dl/extractor/tumblr.py

   1 # -*- coding: utf-8 -*-
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     ExtractorError,
   9 )
  10
  11
  12 class TumblrIE(InfoExtractor):
  13     _VALID_URL = r'http://(?P<blog_name>.*?)\.tumblr\.com/((post)|(video))/(?P<id>\d*)($|/)'
  14     _TESTS = [{
  15         'url': 'http://tatianamaslanydaily.tumblr.com/post/54196191430/orphan-black-dvd-extra-behind-the-scenes',
  16         'md5': '479bb068e5b16462f5176a6828829767',
  17         'info_dict': {
  18             'id': '54196191430',
  19             'ext': 'mp4',
  20             'title': 'tatiana maslany news, Orphan Black || DVD extra - behind the scenes ↳...',
  21             'description': 'md5:dfac39636969fe6bf1caa2d50405f069',
  22             'thumbnail': 're:http://.*\.jpg',
  23         }
  24     }, {
  25         'url': 'http://5sostrum.tumblr.com/post/90208453769/yall-forgetting-the-greatest-keek-of-them-all',
  26         'md5': 'bf348ef8c0ef84fbf1cbd6fa6e000359',
  27         'info_dict': {
  28             'id': '90208453769',
  29             'ext': 'mp4',
  30             'title': '5SOS STRUM ;)',
  31             'description': 'md5:dba62ac8639482759c8eb10ce474586a',
  32             'thumbnail': 're:http://.*\.jpg',
  33         }
  34     }]
  35
  36     def _real_extract(self, url):
  37         m_url = re.match(self._VALID_URL, url)
  38         video_id = m_url.group('id')
  39         blog = m_url.group('blog_name')
  40
  41         url = 'http://%s.tumblr.com/post/%s/' % (blog, video_id)
  42         webpage = self._download_webpage(url, video_id)
  43
  44         re_video = r'src=\\x22(?P<video_url>http://%s\.tumblr\.com/video_file/%s/(.*?))\\x22 type=\\x22video/(?P<ext>.*?)\\x22' % (blog, video_id)
  45         video = re.search(re_video, webpage)
  46         if video is None:
  47             raise ExtractorError('Unable to extract video')
  48         video_url = video.group('video_url')
  49         ext = video.group('ext')
  50
  51         video_thumbnail = self._search_regex(
  52             r'posters.*?\[\\x22(.*?)\\x22',
  53             webpage, 'thumbnail', fatal=False)  # We pick the first poster
  54         if video_thumbnail:
  55             video_thumbnail = video_thumbnail.replace('\\\\/', '/')
  56
  57         # The only place where you can get a title, it's not complete,
  58         # but searching in other places doesn't work for all videos
  59         video_title = self._html_search_regex(r'<title>(?P<title>.*?)(?: \| Tumblr)?</title>',
  60             webpage, 'title', flags=re.DOTALL)
  61
  62         return [{'id': video_id,
  63                  'url': video_url,
  64                  'title': video_title,
  65                  'description': self._html_search_meta('description', webpage),
  66                  'thumbnail': video_thumbnail,
  67                  'ext': ext
  68                  }]