youtube-dl

commit 8749477ed0a3cbc85d1726b6526fa5e794ce6072
parent 7088f5b5fa07381ec9d484c8ef83616724654e3f
Author: Sergey M․ <dstftw@gmail.com>
Date:   Fri, 10 Apr 2015 22:27:16 +0600

[rai] Fix extraction (Closes #5396)

Diffstat:
M
youtube_dl/extractor/rai.py
 | 
72
++++++++++++++++++++++++++++++++++++++++++++++++++++--------------------

1 file changed, 52 insertions(+), 20 deletions(-)
diff --git a/youtube_dl/extractor/rai.py b/youtube_dl/extractor/rai.py
@@ -13,7 +13,7 @@ from ..utils import (
 
 
 class RaiIE(InfoExtractor):
-    _VALID_URL = r'(?P<url>http://(?:.+?\.)?(?:rai\.it|rai\.tv|rainews\.it)/dl/.+?-(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})(?:-.+?)?\.html)'
+    _VALID_URL = r'(?P<url>(?P<host>http://(?:.+?\.)?(?:rai\.it|rai\.tv|rainews\.it))/dl/.+?-(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})(?:-.+?)?\.html)'
     _TESTS = [
         {
             'url': 'http://www.rai.tv/dl/RaiTV/programmi/media/ContentItem-cb27157f-9dd0-4aee-b788-b1f67643a391.html',
@@ -64,32 +64,65 @@ class RaiIE(InfoExtractor):
         },
     ]
 
+    def _extract_relinker_url(self, webpage):
+        return self._proto_relative_url(self._search_regex(
+            [r'name="videourl" content="([^"]+)"', r'var\s+videoURL(?:_MP4)?\s*=\s*"([^"]+)"'],
+            webpage, 'relinker url', default=None))
+
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
+        host = mobj.group('host')
+
+        webpage = self._download_webpage(url, video_id)
 
-        media = self._download_json('%s?json' % mobj.group('url'), video_id, 'Downloading video JSON')
+        relinker_url = self._extract_relinker_url(webpage)
 
-        title = media.get('name')
-        description = media.get('desc')
-        thumbnail = media.get('image_300') or media.get('image_medium') or media.get('image')
-        duration = parse_duration(media.get('length'))
-        uploader = media.get('author')
-        upload_date = unified_strdate(media.get('date'))
+        if not relinker_url:
+            iframe_path = self._search_regex(
+                r'<iframe[^>]+src="/?(dl/[^"]+\?iframe\b[^"]*)"',
+                webpage, 'iframe')
+            iframe_page = self._download_webpage(
+                '%s/%s' % (host, iframe_path), video_id)
+            relinker_url = self._extract_relinker_url(iframe_page)
 
-        formats = []
+        relinker = self._download_json(
+            '%s&output=47' % relinker_url, video_id)
 
-        for format_id in ['wmv', 'm3u8', 'mediaUri', 'h264']:
-            media_url = media.get(format_id)
-            if not media_url:
-                continue
-            formats.append({
+        media_url = relinker['video'][0]
+        ct = relinker.get('ct')
+        if ct == 'f4m':
+            formats = self._extract_f4m_formats(
+                media_url + '&hdcore=3.7.0&plugin=aasp-3.7.0.39.44', video_id)
+        else:
+            formats = [{
                 'url': media_url,
-                'format_id': format_id,
-                'ext': 'mp4',
-            })
+                'format_id': ct,
+            }]
+
+        json_link = self._html_search_meta(
+            'jsonlink', webpage, 'JSON link', default=None)
+        if json_link:
+            media = self._download_json(
+                host + json_link, video_id, 'Downloading video JSON')
+            title = media.get('name')
+            description = media.get('desc')
+            thumbnail = media.get('image_300') or media.get('image_medium') or media.get('image')
+            duration = parse_duration(media.get('length'))
+            uploader = media.get('author')
+            upload_date = unified_strdate(media.get('date'))
+        else:
+            title = self._search_regex(
+                r'var\s+videoTitolo\s*=\s*"([^"]+)";',
+                webpage, 'title', default=None) or self._og_search_title(webpage)
+            description = self._og_search_description(webpage)
+            thumbnail = self._og_search_thumbnail(webpage)
+            duration = None
+            uploader = self._html_search_meta('Editore', webpage, 'uploader')
+            upload_date = unified_strdate(self._html_search_meta(
+                'item-date', webpage, 'upload date'))
 
-        subtitles = self.extract_subtitles(video_id, url)
+        subtitles = self.extract_subtitles(video_id, webpage)
 
         return {
             'id': video_id,
@@ -103,8 +136,7 @@ class RaiIE(InfoExtractor):
             'subtitles': subtitles,
         }
 
-    def _get_subtitles(self, video_id, url):
-        webpage = self._download_webpage(url, video_id)
+    def _get_subtitles(self, video_id, webpage):
         subtitles = {}
         m = re.search(r'<meta name="closedcaption" content="(?P<captions>[^"]+)"', webpage)
         if m: