[dctptv] Fix extraction (closes #14599)

youtube_dl/extractor/dctp.py

@@ -2,53 +2,85 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
-from ..utils import unified_strdate
+from ..compat import compat_str
+from ..utils import (
+    float_or_none,
+    unified_strdate,
+)
 
 
 class DctpTvIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?dctp\.tv/(#/)?filme/(?P<id>.+?)/$'
+    _VALID_URL = r'https?://(?:www\.)?dctp\.tv/(?:#/)?filme/(?P<id>[^/?#&]+)'
     _TEST = {
         'url': 'http://www.dctp.tv/filme/videoinstallation-fuer-eine-kaufhausfassade/',
-        'md5': '174dd4a8a6225cf5655952f969cfbe24',
         'info_dict': {
             'id': '95eaa4f33dad413aa17b4ee613cccc6c',
             'display_id': 'videoinstallation-fuer-eine-kaufhausfassade',
-            'ext': 'mp4',
+            'ext': 'flv',
             'title': 'Videoinstallation für eine Kaufhausfassade',
             'description': 'Kurzfilm',
             'upload_date': '20110407',
             'thumbnail': r're:^https?://.*\.jpg$',
+            'duration': 71.24,
+        },
+        'params': {
+            # rtmp download
+            'skip_download': True,
         },
     }
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        display_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
 
-        object_id = self._html_search_meta('DC.identifier', webpage)
+        webpage = self._download_webpage(url, display_id)
 
-        servers_json = self._download_json(
+        video_id = self._html_search_meta(
-            'http://www.dctp.tv/elastic_streaming_client/get_streaming_server/',
+            'DC.identifier', webpage, 'video id',
-            video_id, note='Downloading server list')
+            default=None) or self._search_regex(
-        server = servers_json[0]['server']
+            r'id=["\']uuid[^>]+>([^<]+)<', webpage, 'video id')
-        m3u8_path = self._search_regex(
-            r'\'([^\'"]+/playlist\.m3u8)"', webpage, 'm3u8 path')
-        formats = self._extract_m3u8_formats(
-            'http://%s%s' % (server, m3u8_path), video_id, ext='mp4',
-            entry_protocol='m3u8_native')
 
         title = self._og_search_title(webpage)
+
+        servers = self._download_json(
+            'http://www.dctp.tv/streaming_servers/', display_id,
+            note='Downloading server list', fatal=False)
+
+        if servers:
+            endpoint = next(
+                server['endpoint']
+                for server in servers
+                if isinstance(server.get('endpoint'), compat_str) and
+                'cloudfront' in server['endpoint'])
+        else:
+            endpoint = 'rtmpe://s2pqqn4u96e4j8.cloudfront.net/cfx/st/'
+
+        app = self._search_regex(
+            r'^rtmpe?://[^/]+/(?P<app>.*)$', endpoint, 'app')
+
+        formats = [{
+            'url': endpoint,
+            'app': app,
+            'play_path': 'mp4:%s_dctp_0500_4x3.m4v' % video_id,
+            'page_url': url,
+            'player_url': 'http://svm-prod-dctptv-static.s3.amazonaws.com/dctptv-relaunch2012-109.swf',
+            'ext': 'flv',
+        }]
+
         description = self._html_search_meta('DC.description', webpage)
         upload_date = unified_strdate(
             self._html_search_meta('DC.date.created', webpage))
         thumbnail = self._og_search_thumbnail(webpage)
+        duration = float_or_none(self._search_regex(
+            r'id=["\']duration_in_ms[^+]>(\d+)', webpage, 'duration',
+            default=None), scale=1000)
 
         return {
-            'id': object_id,
+            'id': video_id,
             'title': title,
             'formats': formats,
-            'display_id': video_id,
+            'display_id': display_id,
             'description': description,
             'upload_date': upload_date,
             'thumbnail': thumbnail,
+            'duration': duration,
         }