[vk] Updated the regex for matching user video pages. It now matches optional URL...

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

diff --git a/youtube_dl/extractor/vk.py b/youtube_dl/extractor/vk.py
index 918bd10988a2c49fac2fc76bb7481e1a3ed7fdf1..3bcf50e280a7730c5650f93c1325718d935afd1a 100644 (file)
--- a/youtube_dl/extractor/vk.py
+++ b/youtube_dl/extractor/vk.py
@@ -16,7 +16,7 @@ from ..utils import (
 
 class VKIE(InfoExtractor):
     IE_NAME = 'vk.com'
-    _VALID_URL = r'https?://(?:m\.)?vk\.com/(?:video_ext\.php\?.*?\boid=(?P<oid>-?\d+).*?\bid=(?P<id>\d+)|(?:.+?\?.*?z=)?video(?P<videoid>.*?)(?:\?|%2F|$))'
+    _VALID_URL = r'https?://(?:m\.)?vk\.com/(?:video_ext\.php\?.*?\boid=(?P<oid>-?\d+).*?\bid=(?P<id>\d+)|(?:.+?\?.*?z=)?video(?P<videoid>[^s].*?)(?:\?|%2F|$))'
     _NETRC_MACHINE = 'vk'
 
     _TESTS = [
@@ -138,9 +138,19 @@ class VKIE(InfoExtractor):
         info_url = 'http://vk.com/al_video.php?act=show&al=1&video=%s' % video_id
         info_page = self._download_webpage(info_url, video_id)
 
-        if re.search(r'<!>Please log in or <', info_page):
-            raise ExtractorError('This video is only available for registered users, '
-                'use --username and --password options to provide account credentials.', expected=True)
+        ERRORS = {
+            r'>Видеозапись .*? была изъята из публичного доступа в связи с обращением правообладателя.<':
+                'Video %s has been removed from public access due to rightholder complaint.',
+            r'<!>Please log in or <':
+                'Video %s is only available for registered users, '
+                'use --username and --password options to provide account credentials.',
+            '<!>Unknown error':
+                'Video %s does not exist.'
+        }
+
+        for error_re, error_msg in ERRORS.items():
+            if re.search(error_re, info_page):
+                raise ExtractorError(error_msg % video_id, expected=True)
 
         m_yt = re.search(r'src="(http://www.youtube.com/.*?)"', info_page)
         if m_yt is not None:
@@ -175,3 +185,38 @@ class VKIE(InfoExtractor):
             'uploader': data.get('md_author'),
             'duration': data.get('duration')
         }
+
+
+class VKUserVideosIE(InfoExtractor):
+    IE_NAME = 'vk.com:user-videos'
+    IE_DESC = 'All of a user\'s videos'
+    _VALID_URL = r'https?://(?:m\.)?vk\.com/videos([0-9]+)(?:m\?.*)?'
+    _TEMPLATE_URL = 'https://vk.com/videos'
+    _TEST = {
+        'url': 'http://vk.com/videos205387401',
+        'playlist_mincount': 4,
+    }
+
+    def extract_videos_from_page(self, page):
+        ids_in_page = []
+        for mobj in re.finditer(r'href="/video([0-9_]+)"', page):
+            if mobj.group(1) not in ids_in_page:
+                ids_in_page.append(mobj.group(1))
+        return ids_in_page
+
+    def _real_extract(self, url):
+        # Extract page id
+        mobj = re.match(self._VALID_URL, url)
+        if mobj is None:
+            raise ExtractorError('Invalid URL: %s' % url)
+
+        # Download page and get video ids
+        page_id = mobj.group(1)
+        page = self._download_webpage(url, page_id)
+        video_ids = self.extract_videos_from_page(page)
+
+        self._downloader.to_screen('[vk] User videos %s: Found %i videos' % (page_id, len(video_ids)))
+
+        url_entries = [self.url_result('http://vk.com/video' + video_id, 'VK', video_id=video_id)
+                       for video_id in video_ids]
+        return self.playlist_result(url_entries, page_id)
\ No newline at end of file