[youtube] Fix extraction.

[youtube-dl] / youtube_dl / extractor / c56.py

diff --git a/youtube_dl/extractor/c56.py b/youtube_dl/extractor/c56.py
index dc3a8d47d164912590d88d42c0fc072a99faeee4..cac8fdcba4a9967b9aa028c29bac1d539af458ca 100644 (file)
--- a/youtube_dl/extractor/c56.py
+++ b/youtube_dl/extractor/c56.py
@@ -1,36 +1,65 @@
 # coding: utf-8
+from __future__ import unicode_literals
 
 import re
-import json
 
 from .common import InfoExtractor
-from ..utils import determine_ext
+from ..utils import js_to_json
+
 
 class C56IE(InfoExtractor):
-    _VALID_URL = r'https?://((www|player)\.)?56\.com/(.+?/)?(v_|(play_album.+-))(?P<textid>.+?)\.(html|swf)'
-    IE_NAME = u'56.com'
-
-    _TEST ={
-        u'url': u'http://www.56.com/u39/v_OTM0NDA3MTY.html',
-        u'file': u'93440716.flv',
-        u'md5': u'e59995ac63d0457783ea05f93f12a866',
-        u'info_dict': {
-            u'title': u'网事知多少 第32期：车怒',
+    _VALID_URL = r'https?://(?:(?:www|player)\.)?56\.com/(?:.+?/)?(?:v_|(?:play_album.+-))(?P<textid>.+?)\.(?:html|swf)'
+    IE_NAME = '56.com'
+    _TESTS = [{
+        'url': 'http://www.56.com/u39/v_OTM0NDA3MTY.html',
+        'md5': 'e59995ac63d0457783ea05f93f12a866',
+        'info_dict': {
+            'id': '93440716',
+            'ext': 'flv',
+            'title': '网事知多少 第32期：车怒',
+            'duration': 283.813,
+        },
+    }, {
+        'url': 'http://www.56.com/u47/v_MTM5NjQ5ODc2.html',
+        'md5': '',
+        'info_dict': {
+            'id': '82247482',
+            'title': '爱的诅咒之杜鹃花开',
         },
-    }
+        'playlist_count': 7,
+        'add_ie': ['Sohu'],
+    }]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
         text_id = mobj.group('textid')
-        info_page = self._download_webpage('http://vxml.56.com/json/%s/' % text_id,
-                                           text_id, u'Downloading video info')
-        info = json.loads(info_page)['info']
-        best_format = sorted(info['rfiles'], key=lambda f: int(f['filesize']))[-1]
-        video_url = best_format['url']
-
-        return {'id': info['vid'],
-                'title': info['Subject'],
-                'url': video_url,
-                'ext': determine_ext(video_url),
-                'thumbnail': info.get('bimg') or info.get('img'),
-                }
+
+        webpage = self._download_webpage(url, text_id)
+        sohu_video_info_str = self._search_regex(
+            r'var\s+sohuVideoInfo\s*=\s*({[^}]+});', webpage, 'Sohu video info', default=None)
+        if sohu_video_info_str:
+            sohu_video_info = self._parse_json(
+                sohu_video_info_str, text_id, transform_source=js_to_json)
+            return self.url_result(sohu_video_info['url'], 'Sohu')
+
+        page = self._download_json(
+            'http://vxml.56.com/json/%s/' % text_id, text_id, 'Downloading video info')
+
+        info = page['info']
+
+        formats = [
+            {
+                'format_id': f['type'],
+                'filesize': int(f['filesize']),
+                'url': f['url']
+            } for f in info['rfiles']
+        ]
+        self._sort_formats(formats)
+
+        return {
+            'id': info['vid'],
+            'title': info['Subject'],
+            'duration': int(info['duration']) / 1000.0,
+            'formats': formats,
+            'thumbnail': info.get('bimg') or info.get('img'),
+        }