[vk] Fix authentication (Closes #6105)

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

diff --git a/youtube_dl/extractor/vk.py b/youtube_dl/extractor/vk.py
index cc384adbf9837f35f90c64d0e8dc0396b0b601ec..f2ae109f99ddc2df40d43b6a25f311b22431f8af 100644 (file)
--- a/youtube_dl/extractor/vk.py
+++ b/youtube_dl/extractor/vk.py
@@ -13,6 +13,7 @@ from ..compat import (
 from ..utils import (
     ExtractorError,
     orderedSet,
+    str_to_int,
     unescapeHTML,
     unified_strdate,
 )
@@ -34,6 +35,7 @@ class VKIE(InfoExtractor):
                 'uploader': 're:(?:Noize MC|Alexander Ilyashenko).*',
                 'duration': 195,
                 'upload_date': '20120212',
+                'view_count': int,
             },
         },
         {
@@ -45,7 +47,8 @@ class VKIE(InfoExtractor):
                 'uploader': 'Tom Cruise',
                 'title': 'No name',
                 'duration': 9,
-                'upload_date': '20130721'
+                'upload_date': '20130721',
+                'view_count': int,
             }
         },
         {
@@ -59,6 +62,7 @@ class VKIE(InfoExtractor):
                 'title': 'Lin Dan',
                 'duration': 101,
                 'upload_date': '20120730',
+                'view_count': int,
             }
         },
         {
@@ -73,7 +77,8 @@ class VKIE(InfoExtractor):
                 'uploader': 'Триллеры',
                 'title': '► Бойцовский клуб / Fight Club 1999 [HD 720]',
                 'duration': 8352,
-                'upload_date': '20121218'
+                'upload_date': '20121218',
+                'view_count': int,
             },
             'skip': 'Requires vk account credentials',
         },
@@ -100,6 +105,7 @@ class VKIE(InfoExtractor):
                 'title': 'Книга Илая',
                 'duration': 6771,
                 'upload_date': '20140626',
+                'view_count': int,
             },
             'skip': 'Only works from Russia',
         },
@@ -115,20 +121,27 @@ class VKIE(InfoExtractor):
         if username is None:
             return
 
-        login_form = {
-            'act': 'login',
-            'role': 'al_frame',
-            'expire': '1',
-            'email': username,
-            'pass': password,
-        }
+        login_page = self._download_webpage(
+            'https://vk.com', None, 'Downloading login page')
+
+        login_form = dict(re.findall(
+            r'<input\s+type="hidden"\s+name="([^"]+)"\s+(?:id="[^"]+"\s+)?value="([^"]*)"',
+            login_page))
 
-        request = compat_urllib_request.Request('https://login.vk.com/?act=login',
-                                                compat_urllib_parse.urlencode(login_form).encode('utf-8'))
-        login_page = self._download_webpage(request, None, note='Logging in as %s' % username)
+        login_form.update({
+            'email': username.encode('cp1251'),
+            'pass': password.encode('cp1251'),
+        })
+
+        request = compat_urllib_request.Request(
+            'https://login.vk.com/?act=login',
+            compat_urllib_parse.urlencode(login_form).encode('utf-8'))
+        login_page = self._download_webpage(
+            request, None, note='Logging in as %s' % username)
 
         if re.search(r'onLoginFailed', login_page):
-            raise ExtractorError('Unable to login, incorrect username and/or password', expected=True)
+            raise ExtractorError(
+                'Unable to login, incorrect username and/or password', expected=True)
 
     def _real_initialize(self):
         self._login()
@@ -175,25 +188,29 @@ class VKIE(InfoExtractor):
                 m_rutube.group(1).replace('\\', ''))
             return self.url_result(rutube_url)
 
-        m_opts = re.search(r'(?s)var\s+opts\s*=\s*({.*?});', info_page)
+        m_opts = re.search(r'(?s)var\s+opts\s*=\s*({.+?});', info_page)
         if m_opts:
-            m_opts_url = re.search(r"url\s*:\s*'([^']+)", m_opts.group(1))
+            m_opts_url = re.search(r"url\s*:\s*'((?!/\b)[^']+)", m_opts.group(1))
             if m_opts_url:
                 opts_url = m_opts_url.group(1)
                 if opts_url.startswith('//'):
                     opts_url = 'http:' + opts_url
                 return self.url_result(opts_url)
 
-        data_json = self._search_regex(r'var vars = ({.*?});', info_page, 'vars')
+        data_json = self._search_regex(r'var\s+vars\s*=\s*({.+?});', info_page, 'vars')
         data = json.loads(data_json)
 
         # Extract upload date
         upload_date = None
-        mobj = re.search(r'id="mv_date_wrap".*?Added ([a-zA-Z]+ [0-9]+), ([0-9]+) at', info_page)
+        mobj = re.search(r'id="mv_date(?:_views)?_wrap"[^>]*>([a-zA-Z]+ [0-9]+), ([0-9]+) at', info_page)
         if mobj is not None:
             mobj.group(1) + ' ' + mobj.group(2)
             upload_date = unified_strdate(mobj.group(1) + ' ' + mobj.group(2))
 
+        view_count = str_to_int(self._search_regex(
+            r'"mv_views_count_number"[^>]*>([\d,.]+) views<',
+            info_page, 'view count', fatal=False))
+
         formats = [{
             'format_id': k,
             'url': v,
@@ -210,6 +227,7 @@ class VKIE(InfoExtractor):
             'uploader': data.get('md_author'),
             'duration': data.get('duration'),
             'upload_date': upload_date,
+            'view_count': view_count,
         }