Fix extraction from youtube.

[youtubedl] / youtube_dl / extractor / voxmedia.py

diff --git a/youtube_dl/extractor/voxmedia.py b/youtube_dl/extractor/voxmedia.py
index b1b32ad44ecfd796e46219a87ea71caa3587face..c7a0a88fe896ac408f0cf1b559929d98b7f6a8b7 100644 (file)
--- a/youtube_dl/extractor/voxmedia.py
+++ b/youtube_dl/extractor/voxmedia.py
@@ -2,20 +2,56 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from .once import OnceIE
 from ..compat import compat_urllib_parse_unquote
+from ..utils import ExtractorError
+
+
+class VoxMediaVolumeIE(OnceIE):
+    _VALID_URL = r'https?://volume\.vox-cdn\.com/embed/(?P<id>[0-9a-f]{9})'
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        video_data = self._parse_json(self._search_regex(
+            r'Volume\.createVideo\(({.+})\s*,\s*{.*}\s*,\s*\[.*\]\s*,\s*{.*}\);', webpage, 'video data'), video_id)
+        for provider_video_type in ('ooyala', 'youtube', 'brightcove'):
+            provider_video_id = video_data.get('%s_id' % provider_video_type)
+            if not provider_video_id:
+                continue
+            info = {
+                'id': video_id,
+                'title': video_data.get('title_short'),
+                'description': video_data.get('description_long') or video_data.get('description_short'),
+                'thumbnail': video_data.get('brightcove_thumbnail')
+            }
+            if provider_video_type == 'brightcove':
+                info['formats'] = self._extract_once_formats(provider_video_id)
+                self._sort_formats(info['formats'])
+            else:
+                info.update({
+                    '_type': 'url_transparent',
+                    'url': provider_video_id if provider_video_type == 'youtube' else '%s:%s' % (provider_video_type, provider_video_id),
+                    'ie_key': provider_video_type.capitalize(),
+                })
+            return info
+        raise ExtractorError('Unable to find provider video id')
 
 
 class VoxMediaIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?(?:theverge|vox|sbnation|eater|polygon|curbed|racked)\.com/(?:[^/]+/)*(?P<id>[^/?]+)'
+    _VALID_URL = r'https?://(?:www\.)?(?:(?:theverge|vox|sbnation|eater|polygon|curbed|racked)\.com|recode\.net)/(?:[^/]+/)*(?P<id>[^/?]+)'
     _TESTS = [{
         'url': 'http://www.theverge.com/2014/6/27/5849272/material-world-how-google-discovered-what-software-is-made-of',
-        'md5': '73856edf3e89a711e70d5cf7cb280b37',
         'info_dict': {
             'id': '11eXZobjrG8DCSTgrNjVinU-YmmdYjhe',
             'ext': 'mp4',
             'title': 'Google\'s new material design direction',
             'description': 'md5:2f44f74c4d14a1f800ea73e1c6832ad2',
         },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
         'add_ie': ['Ooyala'],
     }, {
         # data-ooyala-id
@@ -28,16 +64,20 @@ class VoxMediaIE(InfoExtractor):
             'description': 'md5:87a51fe95ff8cea8b5bdb9ac7ae6a6af',
         },
         'add_ie': ['Ooyala'],
+        'skip': 'Video Not Found',
     }, {
         # volume embed
         'url': 'http://www.vox.com/2016/3/31/11336640/mississippi-lgbt-religious-freedom-bill',
-        'md5': '375c483c5080ab8cd85c9c84cfc2d1e4',
         'info_dict': {
             'id': 'wydzk3dDpmRz7PQoXRsTIX6XTkPjYL0b',
             'ext': 'mp4',
             'title': 'The new frontier of LGBTQ civil rights, explained',
             'description': 'md5:0dc58e94a465cbe91d02950f770eb93f',
         },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
         'add_ie': ['Ooyala'],
     }, {
         # youtube embed
@@ -78,6 +118,17 @@ class VoxMediaIE(InfoExtractor):
                 'description': 'md5:e02d56b026d51aa32c010676765a690d',
             },
         }],
+    }, {
+        # volume embed, Brightcove Once
+        'url': 'https://www.recode.net/2014/6/17/11628066/post-post-pc-ceo-the-full-code-conference-video-of-microsofts-satya',
+        'md5': '01571a896281f77dc06e084138987ea2',
+        'info_dict': {
+            'id': '1231c973d',
+            'ext': 'mp4',
+            'title': 'Post-Post-PC CEO: The Full Code Conference Video of Microsoft\'s Satya Nadella',
+            'description': 'The longtime veteran was chosen earlier this year as the software giant\'s third leader in its history.',
+        },
+        'add_ie': ['VoxMediaVolume'],
     }]
 
     def _real_extract(self, url):
@@ -85,9 +136,14 @@ class VoxMediaIE(InfoExtractor):
         webpage = compat_urllib_parse_unquote(self._download_webpage(url, display_id))
 
         def create_entry(provider_video_id, provider_video_type, title=None, description=None):
+            video_url = {
+                'youtube': '%s',
+                'ooyala': 'ooyala:%s',
+                'volume': 'http://volume.vox-cdn.com/embed/%s',
+            }[provider_video_type] % provider_video_id
             return {
                 '_type': 'url_transparent',
-                'url': provider_video_id if provider_video_type == 'youtube' else '%s:%s' % (provider_video_type, provider_video_id),
+                'url': video_url,
                 'title': title or self._og_search_title(webpage),
                 'description': description or self._og_search_description(webpage),
             }
@@ -118,17 +174,7 @@ class VoxMediaIE(InfoExtractor):
         volume_uuid = self._search_regex(
             r'data-volume-uuid="([^"]+)"', webpage, 'volume uuid', default=None)
         if volume_uuid:
-            volume_webpage = self._download_webpage(
-                'http://volume.vox-cdn.com/embed/%s' % volume_uuid, volume_uuid)
-            video_data = self._parse_json(self._search_regex(
-                r'Volume\.createVideo\(({.+})\s*,\s*{.*}\s*,\s*\[.*\]\s*,\s*{.*}\);', volume_webpage, 'video data'), volume_uuid)
-            for provider_video_type in ('ooyala', 'youtube'):
-                provider_video_id = video_data.get('%s_id' % provider_video_type)
-                if provider_video_id:
-                    description = video_data.get('description_long') or video_data.get('description_short')
-                    entries.append(create_entry(
-                        provider_video_id, provider_video_type, video_data.get('title_short'), description))
-                    break
+            entries.append(create_entry(volume_uuid, 'volume'))
 
         if len(entries) == 1:
             return entries[0]