Merge branch 'srgssr' of https://github.com/remitamine/youtube-dl into remitamine...
[youtube-dl] / youtube_dl / extractor / cspan.py
index c74b35fd9f01cb6f17729793063f2d6e87682042..b3ee670188e2e4b7fb1e50d173ee091e74a05654 100644 (file)
@@ -14,10 +14,6 @@ from ..utils import (
 from .senateisvp import SenateISVPIE
 
 
-def get_text_attr(d, attr):
-    return d.get(attr, {}).get('#text')
-
-
 class CSpanIE(InfoExtractor):
     _VALID_URL = r'http://(?:www\.)?c-span\.org/video/\?(?P<id>[0-9a-f]+)'
     IE_DESC = 'C-SPAN'
@@ -62,18 +58,26 @@ class CSpanIE(InfoExtractor):
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
+        video_type = None
         webpage = self._download_webpage(url, video_id)
-        matches = re.search(r'data-(prog|clip)id=\'([0-9]+)\'', webpage)
-        if matches:
+        # We first look for clipid, because clipprog always appears before
+        patterns = [r'id=\'clip(%s)\'\s*value=\'([0-9]+)\'' % t for t in ('id', 'prog')]
+        results = list(filter(None, (re.search(p, webpage) for p in patterns)))
+        if results:
+            matches = results[0]
             video_type, video_id = matches.groups()
-            if video_type == 'prog':
-                video_type = 'program'
+            video_type = 'clip' if video_type == 'id' else 'program'
         else:
             senate_isvp_url = SenateISVPIE._search_iframe_url(webpage)
             if senate_isvp_url:
                 title = self._og_search_title(webpage)
                 surl = smuggle_url(senate_isvp_url, {'force_title': title})
                 return self.url_result(surl, 'SenateISVP', video_id, title)
+        if video_type is None or video_id is None:
+            raise ExtractorError('unable to find video id and type')
+
+        def get_text_attr(d, attr):
+            return d.get(attr, {}).get('#text')
 
         data = self._download_json(
             'http://www.c-span.org/assets/player/ajax-player.php?os=android&html5=%s&id=%s' % (video_type, video_id),