[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/vidzi.py b/youtube_dl/extractor/vidzi.py
index d671e27decf973bb315bc306f94d89d4774e7ff2..42ea4952c381b497b6f01abcb44212339563e0a3 100644 (file)
--- a/youtube_dl/extractor/vidzi.py
+++ b/youtube_dl/extractor/vidzi.py
@@ -3,16 +3,18 @@ from __future__ import unicode_literals
 
 import re
 
-from .jwplatform import JWPlatformBaseIE
+from .common import InfoExtractor
 from ..utils import (
-    base_n,
+    decode_packed_codes,
     js_to_json,
+    NO_DEFAULT,
+    PACKED_CODES_RE,
 )
 
 
-class VidziIE(JWPlatformBaseIE):
-    _VALID_URL = r'https?://(?:www\.)?vidzi\.tv/(?P<id>\w+)'
-    _TEST = {
+class VidziIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?vidzi\.(?:tv|cc|si|nu)/(?:embed-)?(?P<id>[0-9a-zA-Z]+)'
+    _TESTS = [{
         'url': 'http://vidzi.tv/cghql9yq6emu.html',
         'md5': '4f16c71ca0c8c8635ab6932b5f3f1660',
         'info_dict': {
@@ -24,30 +26,41 @@ class VidziIE(JWPlatformBaseIE):
             # m3u8 download
             'skip_download': True,
         },
-    }
+    }, {
+        'url': 'http://vidzi.tv/embed-4z2yb0rzphe9-600x338.html',
+        'only_matching': True,
+    }, {
+        'url': 'http://vidzi.cc/cghql9yq6emu.html',
+        'only_matching': True,
+    }, {
+        'url': 'https://vidzi.si/rph9gztxj1et.html',
+        'only_matching': True,
+    }, {
+        'url': 'http://vidzi.nu/cghql9yq6emu.html',
+        'only_matching': 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('\\\'', '\'')
-        jwplayer_data = self._parse_json(
-            self._search_regex(r'setup\(([^)]+)\)', code, 'jwplayer data'),
-            video_id, transform_source=js_to_json)
+        codes = [webpage]
+        codes.extend([
+            decode_packed_codes(mobj.group(0)).replace('\\\'', '\'')
+            for mobj in re.finditer(PACKED_CODES_RE, webpage)])
+        for num, code in enumerate(codes, 1):
+            jwplayer_data = self._parse_json(
+                self._search_regex(
+                    r'setup\(([^)]+)\)', code, 'jwplayer data',
+                    default=NO_DEFAULT if num == len(codes) else '{}'),
+                video_id, transform_source=lambda s: js_to_json(
+                    re.sub(r'\s*\+\s*window\[.+?\]', '', s)))
+            if jwplayer_data:
+                break
 
         info_dict = self._parse_jwplayer_data(jwplayer_data, video_id, require_title=False)
         info_dict['title'] = title