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

   1 import re
   2 import xml.etree.ElementTree
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     compat_urllib_parse,
   7
   8     ExtractorError,
   9 )
  10
  11 class GametrailersIE(InfoExtractor):
  12     _VALID_URL = r'http://www.gametrailers.com/(?P<type>videos|reviews|full-episodes)/(?P<id>.*?)/(?P<title>.*)'
  13     _TEST = {
  14         u'url': u'http://www.gametrailers.com/videos/zbvr8i/mirror-s-edge-2-e3-2013--debut-trailer',
  15         u'file': u'70e9a5d7-cf25-4a10-9104-6f3e7342ae0d.flv',
  16         u'md5': u'c3edbc995ab4081976e16779bd96a878',
  17         u'info_dict': {
  18             u"title": u"E3 2013: Debut Trailer"
  19         },
  20         u'skip': u'Requires rtmpdump'
  21     }
  22
  23     def _real_extract(self, url):
  24         mobj = re.match(self._VALID_URL, url)
  25         if mobj is None:
  26             raise ExtractorError(u'Invalid URL: %s' % url)
  27         video_id = mobj.group('id')
  28         webpage = self._download_webpage(url, video_id)
  29         mgid = self._search_regex([r'data-video="(?P<mgid>mgid:.*?)"',
  30                                    r'data-contentId=\'(?P<mgid>mgid:.*?)\''],
  31                                   webpage, u'mgid')
  32
  33         data = compat_urllib_parse.urlencode({'uri': mgid, 'acceptMethods': 'fms'})
  34         info_page = self._download_webpage('http://www.gametrailers.com/feeds/mrss?' + data,
  35                                            video_id, u'Downloading video info')
  36         doc = xml.etree.ElementTree.fromstring(info_page.encode('utf-8'))
  37         default_thumb = doc.find('./channel/image/url').text
  38
  39         media_namespace = {'media': 'http://search.yahoo.com/mrss/'}
  40         parts = [{
  41             'title': video_doc.find('title').text,
  42             'ext': 'flv',
  43             'id': video_doc.find('guid').text.rpartition(':')[2],
  44             # Videos are actually flv not mp4
  45             'url': self._get_video_url(video_doc.find('media:group/media:content', media_namespace).attrib['url'], video_id),
  46             # The thumbnail may not be defined, it would be ''
  47             'thumbnail': video_doc.find('media:group/media:thumbnail', media_namespace).attrib['url'] or default_thumb,
  48             'description': video_doc.find('description').text,
  49         } for video_doc in doc.findall('./channel/item')]
  50         return parts
  51
  52     def _get_video_url(self, mediagen_url, video_id):
  53         if 'acceptMethods' not in mediagen_url:
  54             mediagen_url += '&acceptMethods=fms'
  55         links_webpage = self._download_webpage(mediagen_url,
  56                                                video_id, u'Downloading video urls info')
  57         doc = xml.etree.ElementTree.fromstring(links_webpage)
  58         urls = list(doc.iter('src'))
  59         if len(urls) == 0:
  60             raise ExtractorError(u'Unable to extract video url')
  61         # They are sorted from worst to best quality
  62         return urls[-1].text
  63