Fix extraction from youtube.

[youtubedl] / youtube_dl / extractor / onionstudios.py

diff --git a/youtube_dl/extractor/onionstudios.py b/youtube_dl/extractor/onionstudios.py
index d7b13a0f1fbc53deba74e47f6d1eefc0bdee8be2..c6e3d5640da8f4f0040e862a9641fb73deb391a4 100644 (file)
--- a/youtube_dl/extractor/onionstudios.py
+++ b/youtube_dl/extractor/onionstudios.py
@@ -7,83 +7,75 @@ from .common import InfoExtractor
 from ..utils import (
     determine_ext,
     int_or_none,
+    float_or_none,
+    mimetype2ext,
 )
 
 
 class OnionStudiosIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?onionstudios\.com/(?:videos/[^/]+-|embed\?.*\bid=)(?P<id>\d+)(?!-)'
+    _VALID_URL = r'https?://(?:www\.)?onionstudios\.com/(?:video(?:s/[^/]+-|/)|embed\?.*\bid=)(?P<id>\d+)(?!-)'
 
     _TESTS = [{
         'url': 'http://www.onionstudios.com/videos/hannibal-charges-forward-stops-for-a-cocktail-2937',
-        'md5': 'd4851405d31adfadf71cd7a487b765bb',
+        'md5': '719d1f8c32094b8c33902c17bcae5e34',
         'info_dict': {
             'id': '2937',
             'ext': 'mp4',
             'title': 'Hannibal charges forward, stops for a cocktail',
-            'description': 'md5:e786add7f280b7f0fe237b64cc73df76',
-            'thumbnail': 're:^https?://.*\.jpg$',
+            'thumbnail': r're:^https?://.*\.jpg$',
             'uploader': 'The A.V. Club',
-            'uploader_id': 'TheAVClub',
+            'uploader_id': 'the-av-club',
         },
     }, {
         'url': 'http://www.onionstudios.com/embed?id=2855&autoplay=true',
         'only_matching': True,
+    }, {
+        'url': 'http://www.onionstudios.com/video/6139.json',
+        'only_matching': True,
     }]
 
     @staticmethod
     def _extract_url(webpage):
         mobj = re.search(
-            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?onionstudios\.com/embed.+?)\1', webpage)
+            r'(?s)<(?:iframe|bulbs-video)[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?onionstudios\.com/(?:embed.+?|video/\d+\.json))\1', webpage)
         if mobj:
             return mobj.group('url')
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        webpage = self._download_webpage(
-            'http://www.onionstudios.com/embed?id=%s' % video_id, video_id)
+        video_data = self._download_json(
+            'http://www.onionstudios.com/video/%s.json' % video_id, video_id)
+
+        title = video_data['title']
 
         formats = []
-        for src in re.findall(r'<source[^>]+src="([^"]+)"', webpage):
-            ext = determine_ext(src)
+        for source in video_data.get('sources', []):
+            source_url = source.get('url')
+            if not source_url:
+                continue
+            ext = mimetype2ext(source.get('content_type')) or determine_ext(source_url)
             if ext == 'm3u8':
                 formats.extend(self._extract_m3u8_formats(
-                    src, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
+                    source_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
             else:
-                height = int_or_none(self._search_regex(
-                    r'/(\d+)\.%s' % ext, src, 'height', default=None))
+                tbr = int_or_none(source.get('bitrate'))
                 formats.append({
-                    'format_id': ext + ('-%sp' % height if height else ''),
-                    'url': src,
-                    'height': height,
+                    'format_id': ext + ('-%d' % tbr if tbr else ''),
+                    'url': source_url,
+                    'width': int_or_none(source.get('width')),
+                    'tbr': tbr,
                     'ext': ext,
-                    'preference': 1,
                 })
         self._sort_formats(formats)
 
-        title = self._search_regex(
-            r'share_title\s*=\s*(["\'])(?P<title>[^\1]+?)\1',
-            webpage, 'title', group='title')
-        description = self._search_regex(
-            r'share_description\s*=\s*(["\'])(?P<description>[^\'"]+?)\1',
-            webpage, 'description', default=None, group='description')
-        thumbnail = self._search_regex(
-            r'poster\s*=\s*(["\'])(?P<thumbnail>[^\1]+?)\1',
-            webpage, 'thumbnail', default=False, group='thumbnail')
-
-        uploader_id = self._search_regex(
-            r'twitter_handle\s*=\s*(["\'])(?P<uploader_id>[^\1]+?)\1',
-            webpage, 'uploader id', fatal=False, group='uploader_id')
-        uploader = self._search_regex(
-            r'window\.channelName\s*=\s*(["\'])Embedded:(?P<uploader>[^\1]+?)\1',
-            webpage, 'uploader', default=False, group='uploader')
-
         return {
             'id': video_id,
             'title': title,
-            'description': description,
-            'thumbnail': thumbnail,
-            'uploader': uploader,
-            'uploader_id': uploader_id,
+            'thumbnail': video_data.get('poster_url'),
+            'uploader': video_data.get('channel_name'),
+            'uploader_id': video_data.get('channel_slug'),
+            'duration': float_or_none(video_data.get('duration', 1000)),
+            'tags': video_data.get('tags'),
             'formats': formats,
         }