]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/funnyordie.py
Merge tag 'upstream/2015.05.15'
[youtubedl] / youtube_dl / extractor / funnyordie.py
index 3045978f1b35cff1552b9b0dbf12f7a7ff522989..dd87257c465983dcda30a6faf5dbd7bc0950560c 100644 (file)
@@ -1,10 +1,37 @@
+from __future__ import unicode_literals
+
+import json
 import re
 
 from .common import InfoExtractor
+from ..utils import ExtractorError
 
 
 class FunnyOrDieIE(InfoExtractor):
-    _VALID_URL = r'^(?:https?://)?(?:www\.)?funnyordie\.com/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',
+        'info_dict': {
+            'id': '0732f586d7',
+            'ext': 'mp4',
+            'title': 'Heart-Shaped Box: Literal Video Version',
+            'description': 'md5:ea09a01bc9a1c46d9ab696c01747c338',
+            'thumbnail': 're:^http:.*\.jpg$',
+        },
+    }, {
+        'url': 'http://www.funnyordie.com/embed/e402820827',
+        'info_dict': {
+            'id': 'e402820827',
+            'ext': 'mp4',
+            'title': 'Please Use This Song (Jon Lajoie)',
+            'description': 'Please use this to sell something.  www.jonlajoie.com',
+            'thumbnail': 're:^http:.*\.jpg$',
+        },
+    }, {
+        'url': 'http://www.funnyordie.com/articles/ebf5e34fc8/10-hours-of-walking-in-nyc-as-a-man',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
@@ -12,20 +39,41 @@ class FunnyOrDieIE(InfoExtractor):
         video_id = mobj.group('id')
         webpage = self._download_webpage(url, video_id)
 
-        video_url = self._html_search_regex(r'<video[^>]*>\s*<source[^>]*>\s*<source src="(?P<url>[^"]+)"',
-            webpage, u'video URL', flags=re.DOTALL)
+        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)
 
-        title = self._html_search_regex((r"<h1 class='player_page_h1'.*?>(?P<title>.*?)</h1>",
-            r'<title>(?P<title>[^<]+?)</title>'), webpage, 'title', flags=re.DOTALL)
+        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()
 
-        video_description = self._html_search_regex(r'<meta property="og:description" content="(?P<desc>.*?)"',
-            webpage, u'description', fatal=False, flags=re.DOTALL)
+        formats = []
+        for bitrate in bitrates:
+            for link in links:
+                formats.append({
+                    'url': '%s%d.%s' % (link[0], bitrate, link[1]),
+                    'format_id': '%s-%d' % (link[1], bitrate),
+                    'vbr': bitrate,
+                })
 
-        info = {
+        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)
+
+        return {
             'id': video_id,
-            'url': video_url,
-            'ext': 'mp4',
-            'title': title,
-            'description': video_description,
+            'title': post['name'],
+            'description': post.get('description'),
+            'thumbnail': post.get('picture'),
+            'formats': formats,
+            'subtitles': subtitles,
         }
-        return [info]