youtube-dl

commit ecfe623422f61df886b83e6eae98a0e4a297802e
parent 4a6c94288ae42402f4a61a924318cedc5d1bb511
Author: Philipp Hagemeister <phihag@phihag.de>
Date:   Mon, 27 Oct 2014 01:33:49 +0100

[heise] Fix extraction

Now they use an XML format instead of JSON.

Diffstat:
M
youtube_dl/extractor/heise.py
 | 
48
+++++++++++++++++++++---------------------------

1 file changed, 21 insertions(+), 27 deletions(-)
diff --git a/youtube_dl/extractor/heise.py b/youtube_dl/extractor/heise.py
@@ -4,6 +4,7 @@ from __future__ import unicode_literals
 from .common import InfoExtractor
 from ..utils import (
     get_meta_content,
+    int_or_none,
     parse_iso8601,
 )
 
@@ -28,20 +29,26 @@ class HeiseIE(InfoExtractor):
             'timestamp': 1411812600,
             'upload_date': '20140927',
             'description': 'In uplink-Episode 3.3 geht es darum, wie man sich von Cloud-Anbietern emanzipieren kann, worauf man beim Kauf einer Tastatur achten sollte und was Smartphones über uns verraten.',
+            'thumbnail': 're:https?://.*\.jpg$',
         }
     }
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-
         webpage = self._download_webpage(url, video_id)
-        json_url = self._search_regex(
-            r'json_url:\s*"([^"]+)"', webpage, 'json URL')
-        config = self._download_json(json_url, video_id)
+
+        container_id = self._search_regex(
+            r'<div class="videoplayerjw".*?data-container="([0-9]+)"',
+            webpage, 'container ID')
+        sequenz_id = self._search_regex(
+            r'<div class="videoplayerjw".*?data-sequenz="([0-9]+)"',
+            webpage, 'sequenz ID')
+        data_url = 'http://www.heise.de/videout/feed?container=%s&sequenz=%s' % (container_id, sequenz_id)
+        doc = self._download_xml(data_url, video_id)
 
         info = {
             'id': video_id,
-            'thumbnail': config.get('poster'),
+            'thumbnail': self._og_search_thumbnail(webpage),
             'timestamp': parse_iso8601(get_meta_content('date', webpage)),
             'description': self._og_search_description(webpage),
         }
@@ -49,32 +56,19 @@ class HeiseIE(InfoExtractor):
         title = get_meta_content('fulltitle', webpage)
         if title:
             info['title'] = title
-        elif config.get('title'):
-            info['title'] = config['title']
         else:
             info['title'] = self._og_search_title(webpage)
 
         formats = []
-        for t, rs in config['formats'].items():
-            if not rs or not hasattr(rs, 'items'):
-                self._downloader.report_warning(
-                    'formats: {0}: no resolutions'.format(t))
-                continue
-
-            for height_str, obj in rs.items():
-                format_id = '{0}_{1}'.format(t, height_str)
-
-                if not obj or not obj.get('url'):
-                    self._downloader.report_warning(
-                        'formats: {0}: no url'.format(format_id))
-                    continue
-
-                formats.append({
-                    'url': obj['url'],
-                    'format_id': format_id,
-                    'height': self._int(height_str, 'height'),
-                })
-
+        for source_node in doc.findall('.//{http://rss.jwpcdn.com/}source'):
+            label = source_node.attrib['label']
+            height = int_or_none(self._search_regex(
+                r'^(.*?_)?([0-9]+)p$', label, 'height', default=None))
+            formats.append({
+                'url': source_node.attrib['file'],
+                'format_note': label,
+                'height': height,
+            })
         self._sort_formats(formats)
         info['formats'] = formats