youtube_dl/extractor/howstuffworks.py

   1 from __future__ import unicode_literals
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     find_xpath_attr,
   6     int_or_none,
   7     js_to_json,
   8     unescapeHTML,
   9     determine_ext,
  10 )
  11
  12
  13 class HowStuffWorksIE(InfoExtractor):
  14     _VALID_URL = r'https?://[\da-z-]+\.(?:howstuffworks|stuff(?:(?:youshould|theydontwantyouto)know|toblowyourmind|momnevertoldyou)|(?:brain|car)stuffshow|fwthinking|geniusstuff)\.com/(?:[^/]+/)*(?:\d+-)?(?P<id>.+?)-video\.htm'
  15     _TESTS = [
  16         {
  17             'url': 'http://www.stufftoblowyourmind.com/videos/optical-illusions-video.htm',
  18             'md5': '76646a5acc0c92bf7cd66751ca5db94d',
  19             'info_dict': {
  20                 'id': '855410',
  21                 'ext': 'mp4',
  22                 'title': 'Your Trickster Brain: Optical Illusions -- Science on the Web',
  23                 'description': 'md5:e374ff9561f6833ad076a8cc0a5ab2fb',
  24             },
  25         },
  26         {
  27             'url': 'http://shows.howstuffworks.com/more-shows/why-does-balloon-stick-to-hair-video.htm',
  28             'only_matching': True,
  29         }
  30     ]
  31
  32     def _real_extract(self, url):
  33         display_id = self._match_id(url)
  34         webpage = self._download_webpage(url, display_id)
  35         clip_js = self._search_regex(
  36             r'(?s)var clip = ({.*?});', webpage, 'clip info')
  37         clip_info = self._parse_json(
  38             clip_js, display_id, transform_source=js_to_json)
  39
  40         video_id = clip_info['content_id']
  41         formats = []
  42         m3u8_url = clip_info.get('m3u8')
  43         if m3u8_url and determine_ext(m3u8_url) == 'm3u8':
  44             formats.extend(self._extract_m3u8_formats(m3u8_url, video_id, 'mp4', format_id='hls', fatal=True))
  45         flv_url = clip_info.get('flv_url')
  46         if flv_url:
  47             formats.append({
  48                 'url': flv_url,
  49                 'format_id': 'flv',
  50             })
  51         for video in clip_info.get('mp4', []):
  52             formats.append({
  53                 'url': video['src'],
  54                 'format_id': 'mp4-%s' % video['bitrate'],
  55                 'vbr': int_or_none(video['bitrate'].rstrip('k')),
  56             })
  57
  58         if not formats:
  59             smil = self._download_xml(
  60                 'http://services.media.howstuffworks.com/videos/%s/smil-service.smil' % video_id,
  61                 video_id, 'Downloading video SMIL')
  62
  63             http_base = find_xpath_attr(
  64                 smil,
  65                 './{0}head/{0}meta'.format('{http://www.w3.org/2001/SMIL20/Language}'),
  66                 'name',
  67                 'httpBase').get('content')
  68
  69             URL_SUFFIX = '?v=2.11.3&fp=LNX 11,2,202,356&r=A&g=A'
  70
  71             for video in smil.findall(
  72                     './{0}body/{0}switch/{0}video'.format('{http://www.w3.org/2001/SMIL20/Language}')):
  73                 vbr = int_or_none(video.attrib['system-bitrate'], scale=1000)
  74                 formats.append({
  75                     'url': '%s/%s%s' % (http_base, video.attrib['src'], URL_SUFFIX),
  76                     'format_id': '%dk' % vbr,
  77                     'vbr': vbr,
  78                 })
  79
  80         self._sort_formats(formats)
  81
  82         return {
  83             'id': '%s' % video_id,
  84             'display_id': display_id,
  85             'title': unescapeHTML(clip_info['clip_title']),
  86             'description': unescapeHTML(clip_info.get('caption')),
  87             'thumbnail': clip_info.get('video_still_url'),
  88             'duration': int_or_none(clip_info.get('duration')),
  89             'formats': formats,
  90         }