[BlipTVIE] Fix and simplify extraction of embedded videos
authorrzhxeo <rzhxeot7z81b4700@mailcatch.com>
Mon, 30 Dec 2013 05:07:52 +0000 (06:07 +0100)
committerrzhxeo <rzhxeot7z81b4700@mailcatch.com>
Mon, 30 Dec 2013 05:14:10 +0000 (06:14 +0100)
youtube_dl/extractor/bliptv.py

index c7e0a53c12d61bd450438d67bf1f5b538c13e665..8a8c2e7a88583ec49bc87000712aa3cd8d5fed1a 100644 (file)
@@ -20,7 +20,7 @@ from ..utils import (
 class BlipTVIE(InfoExtractor):
     """Information extractor for blip.tv"""
 
-    _VALID_URL = r'^(?:https?://)?(?:www\.)?blip\.tv/((.+/)|(play/)|(api\.swf#))(.+)$'
+    _VALID_URL = r'^(?:https?://)?(?:\w+\.)?blip\.tv/((.+/)|(play/)|(api\.swf#))(.+)$'
     _URL_EXT = r'^.*\.([a-z0-9]+)$'
     IE_NAME = u'blip.tv'
     _TEST = {
@@ -45,17 +45,12 @@ class BlipTVIE(InfoExtractor):
             raise ExtractorError(u'Invalid URL: %s' % url)
 
         # See https://github.com/rg3/youtube-dl/issues/857
-        api_mobj = re.match(r'http://a\.blip\.tv/api\.swf#(?P<video_id>[\d\w]+)', url)
-        if api_mobj is not None:
-            url = 'http://blip.tv/play/g_%s' % api_mobj.group('video_id')
-        urlp = compat_urllib_parse_urlparse(url)
-        if urlp.path.startswith('/play/'):
-            response = self._request_webpage(url, None, False)
-            redirecturl = response.geturl()
-            rurlp = compat_urllib_parse_urlparse(redirecturl)
-            file_id = compat_parse_qs(rurlp.fragment)['file'][0].rpartition('/')[2]
-            url = 'http://blip.tv/a/a-' + file_id
-            return self._real_extract(url)
+        embed_mobj = re.search(r'^(?:https?://)?(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)([a-zA-Z0-9]+)', url)
+        if embed_mobj:
+            info_url = 'http://blip.tv/play/%s.x?p=1' % embed_mobj.group(1)
+            info_page = self._download_webpage(info_url, embed_mobj.group(1))
+            video_id = self._search_regex(r'data-episode-id="(\d+)', info_page, u'video_id')
+            return self.url_result('http://blip.tv/a/a-'+video_id, 'BlipTV')
 
         if '?' in url:
             cchar = '&'