[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/motherless.py b/youtube_dl/extractor/motherless.py
index 4adac691c09f49241897a63b26774e86c3a019bf..b1615b4d8e4bce8b580942f717477e6ed57ee92e 100644 (file)
--- a/youtube_dl/extractor/motherless.py
+++ b/youtube_dl/extractor/motherless.py
@@ -26,7 +26,7 @@ class MotherlessIE(InfoExtractor):
             'categories': ['Gaming', 'anal', 'reluctant', 'rough', 'Wife'],
             'upload_date': '20100913',
             'uploader_id': 'famouslyfuckedup',
-            'thumbnail': r're:http://.*\.jpg',
+            'thumbnail': r're:https?://.*\.jpg',
             'age_limit': 18,
         }
     }, {
@@ -40,7 +40,7 @@ class MotherlessIE(InfoExtractor):
                            'game', 'hairy'],
             'upload_date': '20140622',
             'uploader_id': 'Sulivana7x',
-            'thumbnail': r're:http://.*\.jpg',
+            'thumbnail': r're:https?://.*\.jpg',
             'age_limit': 18,
         },
         'skip': '404',
@@ -54,7 +54,7 @@ class MotherlessIE(InfoExtractor):
             'categories': ['superheroine heroine  superher'],
             'upload_date': '20140827',
             'uploader_id': 'shade0230',
-            'thumbnail': r're:http://.*\.jpg',
+            'thumbnail': r're:https?://.*\.jpg',
             'age_limit': 18,
         }
     }, {
@@ -76,19 +76,24 @@ class MotherlessIE(InfoExtractor):
             raise ExtractorError('Video %s is for friends only' % video_id, expected=True)
 
         title = self._html_search_regex(
-            r'id="view-upload-title">\s+([^<]+)<', webpage, 'title')
-        video_url = self._html_search_regex(
-            r'setup\(\{\s+"file".+: "([^"]+)",', webpage, 'video URL')
+            (r'(?s)<div[^>]+\bclass=["\']media-meta-title[^>]+>(.+?)</div>',
+             r'id="view-upload-title">\s+([^<]+)<'), webpage, 'title')
+        video_url = (self._html_search_regex(
+            (r'setup\(\{\s*["\']file["\']\s*:\s*(["\'])(?P<url>(?:(?!\1).)+)\1',
+             r'fileurl\s*=\s*(["\'])(?P<url>(?:(?!\1).)+)\1'),
+            webpage, 'video URL', default=None, group='url')
+            or 'http://cdn4.videos.motherlessmedia.com/videos/%s.mp4?fs=opencloud' % video_id)
         age_limit = self._rta_search(webpage)
         view_count = str_to_int(self._html_search_regex(
-            r'<strong>Views</strong>\s+([^<]+)<',
+            (r'>(\d+)\s+Views<', r'<strong>Views</strong>\s+([^<]+)<'),
             webpage, 'view count', fatal=False))
         like_count = str_to_int(self._html_search_regex(
-            r'<strong>Favorited</strong>\s+([^<]+)<',
+            (r'>(\d+)\s+Favorites<', r'<strong>Favorited</strong>\s+([^<]+)<'),
             webpage, 'like count', fatal=False))
 
         upload_date = self._html_search_regex(
-            r'<strong>Uploaded</strong>\s+([^<]+)<', webpage, 'upload date')
+            (r'class=["\']count[^>]+>(\d+\s+[a-zA-Z]{3}\s+\d{4})<',
+             r'<strong>Uploaded</strong>\s+([^<]+)<'), webpage, 'upload date')
         if 'Ago' in upload_date:
             days = int(re.search(r'([0-9]+)', upload_date).group(1))
             upload_date = (datetime.datetime.now() - datetime.timedelta(days=days)).strftime('%Y%m%d')
@@ -120,7 +125,7 @@ class MotherlessIE(InfoExtractor):
 
 
 class MotherlessGroupIE(InfoExtractor):
-    _VALID_URL = 'https?://(?:www\.)?motherless\.com/gv?/(?P<id>[a-z0-9_]+)'
+    _VALID_URL = r'https?://(?:www\.)?motherless\.com/gv?/(?P<id>[a-z0-9_]+)'
     _TESTS = [{
         'url': 'http://motherless.com/g/movie_scenes',
         'info_dict': {
@@ -148,14 +153,27 @@ class MotherlessGroupIE(InfoExtractor):
                 else super(MotherlessGroupIE, cls).suitable(url))
 
     def _extract_entries(self, webpage, base):
-        return [
-            self.url_result(
-                compat_urlparse.urljoin(base, video_path),
-                MotherlessIE.ie_key(), video_title=title)
-            for video_path, title in orderedSet(re.findall(
-                r'href="(/[^"]+)"[^>]+>\s+<img[^>]+alt="[^-]+-\s([^"]+)"',
-                webpage))
-        ]
+        entries = []
+        for mobj in re.finditer(
+                r'href="(?P<href>/[^"]+)"[^>]*>(?:\s*<img[^>]+alt="[^-]+-\s(?P<title>[^"]+)")?',
+                webpage):
+            video_url = compat_urlparse.urljoin(base, mobj.group('href'))
+            if not MotherlessIE.suitable(video_url):
+                continue
+            video_id = MotherlessIE._match_id(video_url)
+            title = mobj.group('title')
+            entries.append(self.url_result(
+                video_url, ie=MotherlessIE.ie_key(), video_id=video_id,
+                video_title=title))
+        # Alternative fallback
+        if not entries:
+            entries = [
+                self.url_result(
+                    compat_urlparse.urljoin(base, '/' + entry_id),
+                    ie=MotherlessIE.ie_key(), video_id=entry_id)
+                for entry_id in orderedSet(re.findall(
+                    r'data-codename=["\']([A-Z0-9]+)', webpage))]
+        return entries
 
     def _real_extract(self, url):
         group_id = self._match_id(url)