]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/funnyordie.py
Merge tag 'upstream/2016.06.25'
[youtubedl] / youtube_dl / extractor / funnyordie.py
index 6e6b6666003d0837bffd6c25ddd12fe1ce892e50..8c5ffc9e84cec305e9fc813a6366b360b7e36230 100644 (file)
@@ -8,7 +8,7 @@ from ..utils import ExtractorError
 
 
 class FunnyOrDieIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?funnyordie\.com/(?P<type>embed|videos)/(?P<id>[0-9a-f]+)(?:$|[?#/])'
+    _VALID_URL = r'https?://(?:www\.)?funnyordie\.com/(?P<type>embed|articles|videos)/(?P<id>[0-9a-f]+)(?:$|[?#/])'
     _TESTS = [{
         'url': 'http://www.funnyordie.com/videos/0732f586d7/heart-shaped-box-literal-video-version',
         'md5': 'bcd81e0c4f26189ee09be362ad6e6ba9',
@@ -21,14 +21,16 @@ class FunnyOrDieIE(InfoExtractor):
         },
     }, {
         'url': 'http://www.funnyordie.com/embed/e402820827',
-        'md5': 'ff4d83318f89776ed0250634cfaa8d36',
         'info_dict': {
             'id': 'e402820827',
             'ext': 'mp4',
             'title': 'Please Use This Song (Jon Lajoie)',
-            'description': 'md5:2ed27d364f5a805a6dba199faaf6681d',
+            'description': 'Please use this to sell something.  www.jonlajoie.com',
             'thumbnail': 're:^http:.*\.jpg$',
         },
+    }, {
+        'url': 'http://www.funnyordie.com/articles/ebf5e34fc8/10-hours-of-walking-in-nyc-as-a-man',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
@@ -37,26 +39,39 @@ class FunnyOrDieIE(InfoExtractor):
         video_id = mobj.group('id')
         webpage = self._download_webpage(url, video_id)
 
-        links = re.findall(r'<source src="([^"]+/v)\d+\.([^"]+)" type=\'video', webpage)
+        links = re.findall(r'<source src="([^"]+/v)[^"]+\.([^"]+)" type=\'video', webpage)
         if not links:
             raise ExtractorError('No media links available for %s' % video_id)
 
         links.sort(key=lambda link: 1 if link[1] == 'mp4' else 0)
 
-        bitrates = self._html_search_regex(r'<source src="[^"]+/v,((?:\d+,)+)\.mp4\.csmil', webpage, 'video bitrates')
-        bitrates = [int(b) for b in bitrates.rstrip(',').split(',')]
-        bitrates.sort()
+        m3u8_url = self._search_regex(
+            r'<source[^>]+src=(["\'])(?P<url>.+?/master\.m3u8[^"\']*)\1',
+            webpage, 'm3u8 url', group='url')
 
         formats = []
 
+        formats.extend(self._extract_m3u8_formats(
+            m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
+
+        bitrates = [int(bitrate) for bitrate in re.findall(r'[,/]v(\d+)[,/]', m3u8_url)]
+        bitrates.sort()
+
         for bitrate in bitrates:
             for link in links:
                 formats.append({
-                    'url': '%s%d.%s' % (link[0], bitrate, link[1]),
+                    'url': self._proto_relative_url('%s%d.%s' % (link[0], bitrate, link[1])),
                     'format_id': '%s-%d' % (link[1], bitrate),
                     'vbr': bitrate,
                 })
 
+        subtitles = {}
+        for src, src_lang in re.findall(r'<track kind="captions" src="([^"]+)" srclang="([^"]+)"', webpage):
+            subtitles[src_lang] = [{
+                'ext': src.split('/')[-1],
+                'url': 'http://www.funnyordie.com%s' % src,
+            }]
+
         post_json = self._search_regex(
             r'fb_post\s*=\s*(\{.*?\});', webpage, 'post details')
         post = json.loads(post_json)
@@ -67,4 +82,5 @@ class FunnyOrDieIE(InfoExtractor):
             'description': post.get('description'),
             'thumbnail': post.get('picture'),
             'formats': formats,
+            'subtitles': subtitles,
         }