[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/hotnewhiphop.py b/youtube_dl/extractor/hotnewhiphop.py
index 82752f9128ccd2cf56cb663740904d232bccf954..4703e1894f08f70057ace62fc1411980ea7728b4 100644 (file)
--- a/youtube_dl/extractor/hotnewhiphop.py
+++ b/youtube_dl/extractor/hotnewhiphop.py
@@ -1,40 +1,66 @@
-import re
-import base64
+from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from ..compat import compat_b64decode
+from ..utils import (
+    ExtractorError,
+    HEADRequest,
+    sanitized_Request,
+    urlencode_postdata,
+)
 
 
 class HotNewHipHopIE(InfoExtractor):
-    _VALID_URL = r'http://www\.hotnewhiphop.com/.*\.(?P<id>.*)\.html'
+    _VALID_URL = r'https?://(?:www\.)?hotnewhiphop\.com/.*\.(?P<id>.*)\.html'
+    _TEST = {
+        'url': 'http://www.hotnewhiphop.com/freddie-gibbs-lay-it-down-song.1435540.html',
+        'md5': '2c2cd2f76ef11a9b3b581e8b232f3d96',
+        'info_dict': {
+            'id': '1435540',
+            'ext': 'mp3',
+            'title': 'Freddie Gibbs - Lay It Down'
+        }
+    }
 
     def _real_extract(self, url):
-        m = re.match(self._VALID_URL, url)
-        video_id = m.group('id')
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
 
-        webpage_src = self._download_webpage(url, video_id)
+        video_url_base64 = self._search_regex(
+            r'data-path="(.*?)"', webpage, 'video URL', default=None)
 
-        video_url_base64 = self._search_regex(r'data-path="(.*?)"',
-            webpage_src, u'video URL', fatal=False)
-
-        if video_url_base64 == None:
-            video_url = self._search_regex(r'"contentUrl" content="(.*?)"', webpage_src,
-                u'video URL')
+        if video_url_base64 is None:
+            video_url = self._search_regex(
+                r'"contentUrl" content="(.*?)"', webpage, 'content URL')
             return self.url_result(video_url, ie='Youtube')
 
-        video_url = base64.b64decode(video_url_base64).decode('utf-8')
-
-        video_title = self._html_search_regex(r"<title>(.*)</title>",
-            webpage_src, u'title')
-        
-        # Getting thumbnail and if not thumbnail sets correct title for WSHH candy video.
-        thumbnail = self._html_search_regex(r'"og:image" content="(.*)"',
-            webpage_src, u'thumbnail', fatal=False)
-
-        results = [{
-                    'id': video_id,
-                    'url' : video_url,
-                    'title' : video_title,
-                    'thumbnail' : thumbnail,
-                    'ext' : 'mp3',
-                    }]
-        return results
\ No newline at end of file
+        reqdata = urlencode_postdata([
+            ('mediaType', 's'),
+            ('mediaId', video_id),
+        ])
+        r = sanitized_Request(
+            'http://www.hotnewhiphop.com/ajax/media/getActions/', data=reqdata)
+        r.add_header('Content-Type', 'application/x-www-form-urlencoded')
+        mkd = self._download_json(
+            r, video_id, note='Requesting media key',
+            errnote='Could not download media key')
+        if 'mediaKey' not in mkd:
+            raise ExtractorError('Did not get a media key')
+
+        redirect_url = compat_b64decode(video_url_base64).decode('utf-8')
+        redirect_req = HEADRequest(redirect_url)
+        req = self._request_webpage(
+            redirect_req, video_id,
+            note='Resolving final URL', errnote='Could not resolve final URL')
+        video_url = req.geturl()
+        if video_url.endswith('.html'):
+            raise ExtractorError('Redirect failed')
+
+        video_title = self._og_search_title(webpage).strip()
+
+        return {
+            'id': video_id,
+            'url': video_url,
+            'title': video_title,
+            'thumbnail': self._og_search_thumbnail(webpage),
+        }