GametrailersIE: make it a subclass of MTVIE to reuse most of the extraction process

This commit is contained in:
Jaime Marquínez Ferrándiz 2013-07-14 14:29:15 +02:00
parent 300fcad8a6
commit ab2f744b90
2 changed files with 24 additions and 48 deletions

View File

@ -1,63 +1,36 @@
import re import re
import xml.etree.ElementTree
from .common import InfoExtractor from .mtv import MTVIE, _media_xml_tag
from ..utils import (
compat_urllib_parse,
ExtractorError, class GametrailersIE(MTVIE):
) """
Gametrailers use the same videos system as MTVIE, it just changes the feed
class GametrailersIE(InfoExtractor): url, where the uri is and the method to get the thumbnails.
"""
_VALID_URL = r'http://www.gametrailers.com/(?P<type>videos|reviews|full-episodes)/(?P<id>.*?)/(?P<title>.*)' _VALID_URL = r'http://www.gametrailers.com/(?P<type>videos|reviews|full-episodes)/(?P<id>.*?)/(?P<title>.*)'
_TEST = { _TEST = {
u'url': u'http://www.gametrailers.com/videos/zbvr8i/mirror-s-edge-2-e3-2013--debut-trailer', u'url': u'http://www.gametrailers.com/videos/zbvr8i/mirror-s-edge-2-e3-2013--debut-trailer',
u'file': u'70e9a5d7-cf25-4a10-9104-6f3e7342ae0d.flv', u'file': u'70e9a5d7-cf25-4a10-9104-6f3e7342ae0d.mp4',
u'md5': u'c3edbc995ab4081976e16779bd96a878', u'md5': u'4c8e67681a0ea7ec241e8c09b3ea8cf7',
u'info_dict': { u'info_dict': {
u"title": u"E3 2013: Debut Trailer" u'title': u'E3 2013: Debut Trailer',
u'description': u'Faith is back! Check out the World Premiere trailer for Mirror\'s Edge 2 straight from the EA Press Conference at E3 2013!',
}, },
u'skip': u'Requires rtmpdump'
} }
# Overwrite MTVIE properties we don't want
_TESTS = []
_FEED_URL = 'http://www.gametrailers.com/feeds/mrss'
def _get_thumbnail_url(self, uri, itemdoc):
search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
return itemdoc.find(search_path).attrib['url']
def _real_extract(self, url): def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url) mobj = re.match(self._VALID_URL, url)
if mobj is None:
raise ExtractorError(u'Invalid URL: %s' % url)
video_id = mobj.group('id') video_id = mobj.group('id')
webpage = self._download_webpage(url, video_id) webpage = self._download_webpage(url, video_id)
mgid = self._search_regex([r'data-video="(?P<mgid>mgid:.*?)"', mgid = self._search_regex([r'data-video="(?P<mgid>mgid:.*?)"',
r'data-contentId=\'(?P<mgid>mgid:.*?)\''], r'data-contentId=\'(?P<mgid>mgid:.*?)\''],
webpage, u'mgid') webpage, u'mgid')
return self._get_videos_info(mgid)
data = compat_urllib_parse.urlencode({'uri': mgid, 'acceptMethods': 'fms'})
info_page = self._download_webpage('http://www.gametrailers.com/feeds/mrss?' + data,
video_id, u'Downloading video info')
doc = xml.etree.ElementTree.fromstring(info_page.encode('utf-8'))
default_thumb = doc.find('./channel/image/url').text
media_namespace = {'media': 'http://search.yahoo.com/mrss/'}
parts = [{
'title': video_doc.find('title').text,
'ext': 'flv',
'id': video_doc.find('guid').text.rpartition(':')[2],
# Videos are actually flv not mp4
'url': self._get_video_url(video_doc.find('media:group/media:content', media_namespace).attrib['url'], video_id),
# The thumbnail may not be defined, it would be ''
'thumbnail': video_doc.find('media:group/media:thumbnail', media_namespace).attrib['url'] or default_thumb,
'description': video_doc.find('description').text,
} for video_doc in doc.findall('./channel/item')]
return parts
def _get_video_url(self, mediagen_url, video_id):
if 'acceptMethods' not in mediagen_url:
mediagen_url += '&acceptMethods=fms'
links_webpage = self._download_webpage(mediagen_url,
video_id, u'Downloading video urls info')
doc = xml.etree.ElementTree.fromstring(links_webpage)
urls = list(doc.iter('src'))
if len(urls) == 0:
raise ExtractorError(u'Unable to extract video url')
# They are sorted from worst to best quality
return urls[-1].text

View File

@ -49,7 +49,10 @@ class MTVIE(InfoExtractor):
if not m: if not m:
raise ExtractorError(u'Cannot transform RTMP url') raise ExtractorError(u'Cannot transform RTMP url')
base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/' base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/'
return base + m.group('finalid') return base + m.group('finalid')
def _get_thumbnail_url(self, uri, itemdoc):
return 'http://mtv.mtvnimages.com/uri/' + uri
def _extract_video_url(self, metadataXml): def _extract_video_url(self, metadataXml):
if '/error_country_block.swf' in metadataXml: if '/error_country_block.swf' in metadataXml:
@ -87,7 +90,7 @@ class MTVIE(InfoExtractor):
description = None description = None
video_info.update({'title': itemdoc.find('title').text, video_info.update({'title': itemdoc.find('title').text,
'id': video_id, 'id': video_id,
'thumbnail': 'http://mtv.mtvnimages.com/uri/' + uri, 'thumbnail': self._get_thumbnail_url(uri, itemdoc),
'description': description, 'description': description,
}) })
return video_info return video_info