X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fvk.py;h=dc2719cf987981999a31487966508e589c86f64b;hb=3995d37da58ed071b54b7f81757cff4d534f5b9b;hp=77f5cebcfb053823c61d3f1b578bb1e5e80af891;hpb=a7ee8a00f4af9853d06ed895c5023cc6b573fd57;p=youtube-dl

diff --git a/youtube_dl/extractor/vk.py b/youtube_dl/extractor/vk.py
index 77f5cebcf..dc2719cf9 100644
--- a/youtube_dl/extractor/vk.py
+++ b/youtube_dl/extractor/vk.py
@@ -1,9 +1,8 @@
-# encoding: utf-8
+# coding: utf-8
 from __future__ import unicode_literals
 
 import collections
 import re
-import json
 import sys
 
 from .common import InfoExtractor
@@ -246,7 +245,7 @@ class VKIE(VKBaseIE):
             },
         },
         {
-            # finished live stream, live_mp4
+            # finished live stream, postlive_mp4
             'url': 'https://vk.com/videos-387766?z=video-387766_456242764%2Fpl_-387766_-2',
             'md5': '90d22d051fccbbe9becfccc615be6791',
             'info_dict': {
@@ -255,10 +254,11 @@ class VKIE(VKBaseIE):
                 'title': 'ИгроМир 2016 — день 1',
                 'uploader': 'Игромания',
                 'duration': 5239,
+                'view_count': int,
             },
         },
         {
-            # live stream, hls and rtmp links,most likely already finished live
+            # live stream, hls and rtmp links, most likely already finished live
             # stream by the time you are reading this comment
             'url': 'https://vk.com/video-140332_456239111',
             'only_matching': True,
@@ -281,6 +281,11 @@ class VKIE(VKBaseIE):
         {
             'url': 'http://new.vk.com/video205387401_165548505',
             'only_matching': True,
+        },
+        {
+            # This video is no longer available, because its author has been blocked.
+            'url': 'https://vk.com/video-10639516_456240611',
+            'only_matching': True,
         }
     ]
 
@@ -328,6 +333,12 @@ class VKIE(VKBaseIE):
 
             r'<!>Access denied':
             'Access denied to video %s.',
+
+            r'<!>Видеозапись недоступна, так как её автор был заблокирован.':
+            'Video %s is no longer available, because its author has been blocked.',
+
+            r'<!>This video is no longer available, because its author has been blocked.':
+            'Video %s is no longer available, because its author has been blocked.',
         }
 
         for error_re, error_msg in ERRORS.items():
@@ -340,7 +351,7 @@ class VKIE(VKBaseIE):
         if youtube_url:
             return self.url_result(youtube_url, 'Youtube')
 
-        vimeo_url = VimeoIE._extract_vimeo_url(url, info_page)
+        vimeo_url = VimeoIE._extract_url(url, info_page)
         if vimeo_url is not None:
             return self.url_result(vimeo_url)
 
@@ -368,28 +379,50 @@ class VKIE(VKBaseIE):
                     opts_url = 'http:' + opts_url
                 return self.url_result(opts_url)
 
-        data_json = self._search_regex(r'var\s+vars\s*=\s*({.+?});', info_page, 'vars')
-        data = json.loads(data_json)
+        # vars does not look to be served anymore since 24.10.2016
+        data = self._parse_json(
+            self._search_regex(
+                r'var\s+vars\s*=\s*({.+?});', info_page, 'vars', default='{}'),
+            video_id, fatal=False)
+
+        # <!json> is served instead
+        if not data:
+            data = self._parse_json(
+                self._search_regex(
+                    r'<!json>\s*({.+?})\s*<!>', info_page, 'json', default='{}'),
+                video_id)
+            if data:
+                data = data['player']['params'][0]
+
+        if not data:
+            data = self._parse_json(
+                self._search_regex(
+                    r'var\s+playerParams\s*=\s*({.+?})\s*;\s*\n', info_page,
+                    'player params'),
+                video_id)['params'][0]
 
         title = unescapeHTML(data['md_title'])
 
-        if data.get('live') == 2:
+        # 2 = live
+        # 3 = post live (finished live)
+        is_live = data.get('live') == 2
+        if is_live:
             title = self._live_title(title)
 
-        # Extract upload date
         timestamp = unified_timestamp(self._html_search_regex(
-            r'class=["\']mv_info_date[^>]*>([^<]+)(?:<|from)', info_page,
+            r'class=["\']mv_info_date[^>]+>([^<]+)(?:<|from)', info_page,
             'upload date', fatal=False))
 
-        view_count = str_to_int(self._html_search_regex(
-            r'class="mv_views_count[^>]*>([\d,.]+)',
-            info_page, 'view count', default=None))
+        view_count = str_to_int(self._search_regex(
+            r'class=["\']mv_views_count[^>]+>\s*([\d,.]+)',
+            info_page, 'view count', fatal=False))
 
         formats = []
         for format_id, format_url in data.items():
             if not isinstance(format_url, compat_str) or not format_url.startswith(('http', '//', 'rtmp')):
                 continue
-            if format_id.startswith(('url', 'cache')) or format_id in ('extra_data', 'live_mp4'):
+            if (format_id.startswith(('url', 'cache')) or
+                    format_id in ('extra_data', 'live_mp4', 'postlive_mp4')):
                 height = int_or_none(self._search_regex(
                     r'^(?:url|cache)(\d+)', format_id, 'height', default=None))
                 formats.append({
@@ -399,8 +432,8 @@ class VKIE(VKBaseIE):
                 })
             elif format_id == 'hls':
                 formats.extend(self._extract_m3u8_formats(
-                    format_url, video_id, 'mp4', m3u8_id=format_id,
-                    fatal=False, live=True))
+                    format_url, video_id, 'mp4', 'm3u8_native',
+                    m3u8_id=format_id, fatal=False, live=is_live))
             elif format_id == 'rtmp':
                 formats.append({
                     'format_id': format_id,
@@ -418,6 +451,7 @@ class VKIE(VKBaseIE):
             'duration': data.get('duration'),
             'timestamp': timestamp,
             'view_count': view_count,
+            'is_live': is_live,
         }