X-Git-Url: http://git.cielonegro.org/gitweb.cgi?a=blobdiff_plain;f=youtube_dl%2FInfoExtractors.py;h=9d885d238882c99752dc8b49fc5c2f81ce3d9f20;hb=1736dec629bd3da25bea21145794adfa7a835ea3;hp=789fd147c5be52db272ed5626a69cac45f27b7b8;hpb=7beb36a52974a2cf7c8b00ba1da339f4f3de3abc;p=youtube-dl.git
diff --git a/youtube_dl/InfoExtractors.py b/youtube_dl/InfoExtractors.py
index 789fd147c..9d885d238 100755
--- a/youtube_dl/InfoExtractors.py
+++ b/youtube_dl/InfoExtractors.py
@@ -31,103 +31,22 @@ from .extractor.gametrailers import GametrailersIE
from .extractor.generic import GenericIE
from .extractor.googleplus import GooglePlusIE
from .extractor.googlesearch import GoogleSearchIE
+from .extractor.infoq import InfoQIE
from .extractor.metacafe import MetacafeIE
+from .extractor.mtv import MTVIE
from .extractor.myvideo import MyVideoIE
+from .extractor.nba import NBAIE
from .extractor.statigram import StatigramIE
from .extractor.photobucket import PhotobucketIE
from .extractor.soundcloud import SoundcloudIE, SoundcloudSetIE
+from .extractor.stanfordoc import StanfordOpenClassroomIE
from .extractor.vimeo import VimeoIE
+from .extractor.xvideos import XVideosIE
from .extractor.yahoo import YahooIE, YahooSearchIE
from .extractor.youtube import YoutubeIE, YoutubePlaylistIE, YoutubeSearchIE, YoutubeUserIE, YoutubeChannelIE
from .extractor.zdf import ZDFIE
-class XVideosIE(InfoExtractor):
- """Information extractor for xvideos.com"""
-
- _VALID_URL = r'^(?:https?://)?(?:www\.)?xvideos\.com/video([0-9]+)(?:.*)'
- IE_NAME = u'xvideos'
-
- def _real_extract(self, url):
- mobj = re.match(self._VALID_URL, url)
- if mobj is None:
- raise ExtractorError(u'Invalid URL: %s' % url)
- video_id = mobj.group(1)
-
- webpage = self._download_webpage(url, video_id)
-
- self.report_extraction(video_id)
-
- # Extract video URL
- video_url = compat_urllib_parse.unquote(self._search_regex(r'flv_url=(.+?)&',
- webpage, u'video URL'))
-
- # Extract title
- video_title = self._html_search_regex(r'
(.*?)\s+-\s+XVID',
- webpage, u'title')
-
- # Extract video thumbnail
- video_thumbnail = self._search_regex(r'http://(?:img.*?\.)xvideos.com/videos/thumbs/[a-fA-F0-9]+/[a-fA-F0-9]+/[a-fA-F0-9]+/[a-fA-F0-9]+/([a-fA-F0-9.]+jpg)',
- webpage, u'thumbnail', fatal=False)
-
- info = {
- 'id': video_id,
- 'url': video_url,
- 'uploader': None,
- 'upload_date': None,
- 'title': video_title,
- 'ext': 'flv',
- 'thumbnail': video_thumbnail,
- 'description': None,
- }
-
- return [info]
-
-
-
-
-class InfoQIE(InfoExtractor):
- """Information extractor for infoq.com"""
- _VALID_URL = r'^(?:https?://)?(?:www\.)?infoq\.com/[^/]+/[^/]+$'
-
- def _real_extract(self, url):
- mobj = re.match(self._VALID_URL, url)
- if mobj is None:
- raise ExtractorError(u'Invalid URL: %s' % url)
-
- webpage = self._download_webpage(url, video_id=url)
- self.report_extraction(url)
-
- # Extract video URL
- mobj = re.search(r"jsclassref ?= ?'([^']*)'", webpage)
- if mobj is None:
- raise ExtractorError(u'Unable to extract video url')
- real_id = compat_urllib_parse.unquote(base64.b64decode(mobj.group(1).encode('ascii')).decode('utf-8'))
- video_url = 'rtmpe://video.infoq.com/cfx/st/' + real_id
-
- # Extract title
- video_title = self._search_regex(r'contentTitle = "(.*?)";',
- webpage, u'title')
-
- # Extract description
- video_description = self._html_search_regex(r'',
- webpage, u'description', fatal=False)
-
- video_filename = video_url.split('/')[-1]
- video_id, extension = video_filename.split('.')
-
- info = {
- 'id': video_id,
- 'url': video_url,
- 'uploader': None,
- 'upload_date': None,
- 'title': video_title,
- 'ext': extension, # Extension is always(?) mp4, but seems to be flv
- 'thumbnail': None,
- 'description': video_description,
- }
-
- return [info]
class MixcloudIE(InfoExtractor):
"""Information extractor for www.mixcloud.com"""
@@ -234,162 +153,7 @@ class MixcloudIE(InfoExtractor):
'player_url': player_url.decode('utf-8'),
}]
-class StanfordOpenClassroomIE(InfoExtractor):
- """Information extractor for Stanford's Open ClassRoom"""
- _VALID_URL = r'^(?:https?://)?openclassroom.stanford.edu(?P/?|(/MainFolder/(?:HomePage|CoursePage|VideoPage)\.php([?]course=(?P[^&]+)(&video=(?P