[puls4] fix extraction(closes #10583)
[youtube-dl] / youtube_dl / extractor / vidzi.py
index d671e27decf973bb315bc306f94d89d4774e7ff2..d49cc6cbc567a8a0219f304a52707bd4129d1119 100644 (file)
@@ -1,18 +1,16 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .jwplatform import JWPlatformBaseIE
 from ..utils import (
-    base_n,
+    decode_packed_codes,
     js_to_json,
 )
 
 
 class VidziIE(JWPlatformBaseIE):
-    _VALID_URL = r'https?://(?:www\.)?vidzi\.tv/(?P<id>\w+)'
-    _TEST = {
+    _VALID_URL = r'https?://(?:www\.)?vidzi\.tv/(?:embed-)?(?P<id>[0-9a-zA-Z]+)'
+    _TESTS = [{
         'url': 'http://vidzi.tv/cghql9yq6emu.html',
         'md5': '4f16c71ca0c8c8635ab6932b5f3f1660',
         'info_dict': {
@@ -24,27 +22,20 @@ class VidziIE(JWPlatformBaseIE):
             # m3u8 download
             'skip_download': True,
         },
-    }
+    }, {
+        'url': 'http://vidzi.tv/embed-4z2yb0rzphe9-600x338.html',
+        'skip_download': True,
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        webpage = self._download_webpage(url, video_id)
+        webpage = self._download_webpage(
+            'http://vidzi.tv/%s' % video_id, video_id)
         title = self._html_search_regex(
             r'(?s)<h2 class="video-title">(.*?)</h2>', webpage, 'title')
 
-        mobj = re.search(r"}\('(.+)',36,(\d+),'([^']+)'\.split\('\|'\)", webpage)
-        code, count, symbols = mobj.groups()
-
-        count = int(count)
-        symbols = symbols.split('|')
-
-        while count:
-            count -= 1
-            if symbols[count]:
-                code = re.sub(r'\b%s\b' % base_n(count, 36), symbols[count], code)
-
-        code = code.replace('\\\'', '\'')
+        code = decode_packed_codes(webpage).replace('\\\'', '\'')
         jwplayer_data = self._parse_json(
             self._search_regex(r'setup\(([^)]+)\)', code, 'jwplayer data'),
             video_id, transform_source=js_to_json)