[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/karaoketv.py b/youtube_dl/extractor/karaoketv.py
index 4d50308cc44de26269eb9727203c8662c857df7c..bfccf89b0fda0be1100764290681a53e022947e0 100644 (file)
--- a/youtube_dl/extractor/karaoketv.py
+++ b/youtube_dl/extractor/karaoketv.py
@@ -1,47 +1,64 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import re
-import json
-import sys
-
 from .common import InfoExtractor
-from ..utils import compat_urllib_parse, ExtractorError
 
 
 class KaraoketvIE(InfoExtractor):
-    _VALID_URL = r'http://karaoketv\.co\.il/\?container=songs&id=(?P<id>[0-9]+)'
+    _VALID_URL = r'https?://(?:www\.)?karaoketv\.co\.il/[^/]+/(?P<id>\d+)'
     _TEST = {
-        'url': 'http://karaoketv.co.il/?container=songs&id=171568',
+        'url': 'http://www.karaoketv.co.il/%D7%A9%D7%99%D7%A8%D7%99_%D7%A7%D7%A8%D7%99%D7%95%D7%A7%D7%99/58356/%D7%90%D7%99%D7%96%D7%95%D7%9F',
         'info_dict': {
-            'id': '171568',
-            'ext': 'mp4',
-            'title': 'אל העולם שלך - רותם כהן - שרים קריוקי',
+            'id': '58356',
+            'ext': 'flv',
+            'title': 'קריוקי של איזון',
+        },
+        'params': {
+            # rtmp download
+            'skip_download': True,
         }
     }
 
     def _real_extract(self, url):
+        video_id = self._match_id(url)
 
-        # BUG: SSL23_GET_SERVER_HELLO:unknown protocol 
-        if sys.hexversion < 0x03000000:
-            raise ExtractorError("Only python 3 supported.\n")
+        webpage = self._download_webpage(url, video_id)
+        api_page_url = self._search_regex(
+            r'<iframe[^>]+src=(["\'])(?P<url>https?://www\.karaoke\.co\.il/api_play\.php\?.+?)\1',
+            webpage, 'API play URL', group='url')
 
-        mobj = re.match(self._VALID_URL, url)
-        
-        video_id = mobj.group('id')
+        api_page = self._download_webpage(api_page_url, video_id)
+        video_cdn_url = self._search_regex(
+            r'<iframe[^>]+src=(["\'])(?P<url>https?://www\.video-cdn\.com/embed/iframe/.+?)\1',
+            api_page, 'video cdn URL', group='url')
 
-        webpage = self._download_webpage(url, video_id)
+        video_cdn = self._download_webpage(video_cdn_url, video_id)
+        play_path = self._parse_json(
+            self._search_regex(
+                r'var\s+options\s*=\s*({.+?});', video_cdn, 'options'),
+            video_id)['clip']['url']
 
-        settings_json = compat_urllib_parse.unquote_plus(self._search_regex(r'config=(.*)', self._og_search_video_url(webpage ,video_id), ''))
-        
-        urls_info_webpage = self._download_webpage(settings_json, 'Downloading settings json')
+        settings = self._parse_json(
+            self._search_regex(
+                r'var\s+settings\s*=\s*({.+?});', video_cdn, 'servers', default='{}'),
+            video_id, fatal=False) or {}
 
-        urls_info_json = json.loads(urls_info_webpage.replace('\'', '"'))
+        servers = settings.get('servers')
+        if not servers or not isinstance(servers, list):
+            servers = ('wowzail.video-cdn.com:80/vodcdn', )
 
-        url = urls_info_json['playlist'][0]['url']
+        formats = [{
+            'url': 'rtmp://%s' % server if not server.startswith('rtmp') else server,
+            'play_path': play_path,
+            'app': 'vodcdn',
+            'page_url': video_cdn_url,
+            'player_url': 'http://www.video-cdn.com/assets/flowplayer/flowplayer.commercial-3.2.18.swf',
+            'rtmp_real_time': True,
+            'ext': 'flv',
+        } for server in servers]
 
         return {
             'id': video_id,
             'title': self._og_search_title(webpage),
-            'url': url,
-        }
\ No newline at end of file
+            'formats': formats,
+        }