X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/975f9519f0ae48e780767bf543204c6626a493b0..0534c7ab3aafa6065faa9453651db8a8e11204b1:/youtube_dl/extractor/tudou.py?ds=inline

diff --git a/youtube_dl/extractor/tudou.py b/youtube_dl/extractor/tudou.py
index 1405b73..5f7ac4b 100644
--- a/youtube_dl/extractor/tudou.py
+++ b/youtube_dl/extractor/tudou.py
@@ -1,63 +1,94 @@
 # coding: utf-8
 
-import re
-import json
+from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from ..compat import compat_str
 
 
 class TudouIE(InfoExtractor):
-    _VALID_URL = r'(?:http://)?(?:www\.)?tudou\.com/(?:listplay|programs)/(?:view|(.+?))/(?:([^/]+)|([^/]+))(?:\.html)?'
-    _TEST = {
-        u'url': u'http://www.tudou.com/listplay/zzdE77v6Mmo/2xN2duXMxmw.html',
-        u'file': u'159448201.f4v',
-        u'md5': u'140a49ed444bd22f93330985d8475fcb',
-        u'info_dict': {
-            u"title": u"卡马乔国足开大脚长传冲吊集锦"
+    _VALID_URL = r'https?://(?:www\.)?tudou\.com/(?:listplay|programs(?:/view)?|albumplay)/([^/]+/)*(?P<id>[^/?#]+?)(?:\.html)?/?(?:$|[?#])'
+    _TESTS = [{
+        'url': 'http://www.tudou.com/listplay/zzdE77v6Mmo/2xN2duXMxmw.html',
+        'md5': '140a49ed444bd22f93330985d8475fcb',
+        'info_dict': {
+            'id': '159448201',
+            'ext': 'f4v',
+            'title': '卡马乔国足开大脚长传冲吊集锦',
+            'thumbnail': 're:^https?://.*\.jpg$',
         }
-    }
+    }, {
+        'url': 'http://www.tudou.com/programs/view/ajX3gyhL0pc/',
+        'info_dict': {
+            'id': '117049447',
+            'ext': 'f4v',
+            'title': 'La Sylphide-Bolshoi-Ekaterina Krysanova & Vyacheslav Lopatin 2012',
+            'thumbnail': 're:^https?://.*\.jpg$',
+        }
+    }, {
+        'url': 'http://www.tudou.com/albumplay/cJAHGih4yYg.html',
+        'only_matching': True,
+    }]
+
+    _PLAYER_URL = 'http://js.tudouui.com/bin/lingtong/PortalPlayer_177.swf'
 
-    def _url_for_id(self, id, quality = None):
-        info_url = "http://v2.tudou.com/f?id="+str(id)
+    def _url_for_id(self, video_id, quality=None):
+        info_url = 'http://v2.tudou.com/f?id=' + compat_str(video_id)
         if quality:
             info_url += '&hd' + quality
-        webpage = self._download_webpage(info_url, id, "Opening the info webpage")
-        final_url = self._html_search_regex('>(.+?)</f>',webpage, 'video url')
+        xml_data = self._download_xml(info_url, video_id, "Opening the info XML page")
+        final_url = xml_data.text
         return final_url
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group(2)
+        video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
-        title = re.search(",kw:\"(.+)\"",webpage)
-        if title is None:
-            title = re.search(",kw: \'(.+)\'",webpage)
-        title = title.group(1)
-        thumbnail_url = re.search(",pic: \'(.+?)\'",webpage)
-        if thumbnail_url is None:
-            thumbnail_url = re.search(",pic:\"(.+?)\"",webpage)
-        thumbnail_url = thumbnail_url.group(1)
-
-        segs_json = self._search_regex(r'segs: \'(.*)\'', webpage, 'segments')
-        segments = json.loads(segs_json)
+
+        youku_vcode = self._search_regex(
+            r'vcode\s*:\s*[\'"]([^\'"]*)[\'"]', webpage, 'youku vcode', default=None)
+        if youku_vcode:
+            return self.url_result('youku:' + youku_vcode, ie='Youku')
+
+        title = self._search_regex(
+            r',kw\s*:\s*[\'"]([^\'"]+)[\'"]', webpage, 'title')
+        thumbnail_url = self._search_regex(
+            r',pic\s*:\s*[\'"]([^\'"]+)[\'"]', webpage, 'thumbnail URL', fatal=False)
+
+        player_url = self._search_regex(
+            r'playerUrl\s*:\s*[\'"]([^\'"]+\.swf)[\'"]',
+            webpage, 'player URL', default=self._PLAYER_URL)
+
+        segments = self._parse_json(self._search_regex(
+            r'segs: \'([^\']+)\'', webpage, 'segments'), video_id)
         # It looks like the keys are the arguments that have to be passed as
         # the hd field in the request url, we pick the higher
-        quality = sorted(segments.keys())[-1]
+        # Also, filter non-number qualities (see issue #3643).
+        quality = sorted(filter(lambda k: k.isdigit(), segments.keys()),
+                         key=lambda k: int(k))[-1]
         parts = segments[quality]
         result = []
         len_parts = len(parts)
         if len_parts > 1:
-            self.to_screen(u'%s: found %s parts' % (video_id, len_parts))
+            self.to_screen('%s: found %s parts' % (video_id, len_parts))
         for part in parts:
             part_id = part['k']
             final_url = self._url_for_id(part_id, quality)
             ext = (final_url.split('?')[0]).split('.')[-1]
-            part_info = {'id': part_id,
-                          'url': final_url,
-                          'ext': ext,
-                          'title': title,
-                          'thumbnail': thumbnail_url,
-                          }
+            part_info = {
+                'id': '%s' % part_id,
+                'url': final_url,
+                'ext': ext,
+                'title': title,
+                'thumbnail': thumbnail_url,
+                'http_headers': {
+                    'Referer': player_url,
+                },
+            }
             result.append(part_info)
 
-        return result
+        return {
+            '_type': 'multi_video',
+            'entries': result,
+            'id': video_id,
+            'title': title,
+        }