[generic] Add nowvideo test hidden behind percent encoding

[youtube-dl] / youtube_dl / extractor / generic.py

diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py
index 595ba268fe5c223108ede65f520c78a77d58008c..0d02f836e122054c8eea8f658d6a4b13a5b97973 100644 (file)
--- a/youtube_dl/extractor/generic.py
+++ b/youtube_dl/extractor/generic.py
@@ -4,7 +4,6 @@ from __future__ import unicode_literals
 
 import os
 import re
-import xml.etree.ElementTree
 
 from .common import InfoExtractor
 from .youtube import YoutubeIE
@@ -17,6 +16,7 @@ from ..utils import (
 
     ExtractorError,
     HEADRequest,
+    parse_xml,
     smuggle_url,
     unescapeHTML,
     unified_strdate,
@@ -116,6 +116,45 @@ class GenericIE(InfoExtractor):
             'params': {
                 'skip_download': False,
             }
+        },
+        # embed.ly video
+        {
+            'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
+            'info_dict': {
+                'id': '9ODmcdjQcHQ',
+                'ext': 'mp4',
+                'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
+                'upload_date': '20140225',
+                'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
+                'uploader': 'Tested',
+                'uploader_id': 'testedcom',
+            },
+            # No need to test YoutubeIE here
+            'params': {
+                'skip_download': True,
+            },
+        },
+        # funnyordie embed
+        {
+            'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
+            'md5': '7cf780be104d40fea7bae52eed4a470e',
+            'info_dict': {
+                'id': '18e820ec3f',
+                'ext': 'mp4',
+                'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
+                'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
+            }
+        },
+        # nowvideo embed hidden behind percent encoding
+        {
+            'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
+            'md5': '2baf4ddd70f697d94b1c18cf796d5107',
+            'info_dict': {
+                'id': '06e53103ca9aa',
+                'ext': 'flv',
+                'title': 'Macross Episode 001  Watch Macross Episode 001 onl',
+                'description': 'No description',
+            },
         }
     ]
 
@@ -211,7 +250,7 @@ class GenericIE(InfoExtractor):
             else:
                 assert ':' in default_search
                 return self.url_result(default_search + url)
-        video_id = os.path.splitext(url.split('/')[-1])[0]
+        video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 
         self.to_screen('%s: Requesting header' % video_id)
 
@@ -257,12 +296,17 @@ class GenericIE(InfoExtractor):
 
         # Is it an RSS feed?
         try:
-            doc = xml.etree.ElementTree.fromstring(webpage.encode('utf-8'))
+            doc = parse_xml(webpage)
             if doc.tag == 'rss':
                 return self._extract_rss(url, video_id, doc)
         except compat_xml_parse_error:
             pass
 
+        # Sometimes embedded video player is hidden behind percent encoding
+        # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
+        # Unescaping the whole page allows to handle those cases in a generic way
+        webpage = compat_urllib_parse.unquote(webpage)
+
         # it's tempting to parse this further, but you would
         # have to take into account all the variations like
         #   Video Title - Site Name
@@ -396,12 +440,33 @@ class GenericIE(InfoExtractor):
         if mobj is not None:
             return self.url_result(mobj.group('url'), 'Facebook')
 
+        # Look for embedded VK player
+        mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'), 'VK')
+
         # Look for embedded Huffington Post player
         mobj = re.search(
             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
         if mobj is not None:
             return self.url_result(mobj.group('url'), 'HuffPost')
 
+        # Look for embed.ly
+        mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'))
+        mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
+        if mobj is not None:
+            return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
+
+        # Look for funnyordie embed
+        matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
+        if matches:
+            urlrs = [self.url_result(unescapeHTML(eurl), 'FunnyOrDie')
+                     for eurl in matches]
+            return self.playlist_result(
+                urlrs, playlist_id=video_id, playlist_title=video_title)
+
         # Start with something easy: JW Player in SWFObject
         mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
         if mobj is None: