Extract the base_url for the XML download from the JS snippet's 'server' variable.

[youtube-dl.git] / youtube_dl / extractor / spiegel.py

diff --git a/youtube_dl/extractor/spiegel.py b/youtube_dl/extractor/spiegel.py
index 98a65b78bab945b999a1edae2ca088771622e125..87f1ee69462600f57c0bde46784307cb6e3e0963 100644 (file)
--- a/youtube_dl/extractor/spiegel.py
+++ b/youtube_dl/extractor/spiegel.py
@@ -1,11 +1,28 @@
+from __future__ import unicode_literals
+
 import re
-import xml.etree.ElementTree
 
 from .common import InfoExtractor
 
 
 class SpiegelIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?spiegel\.de/video/[^/]*-(?P<videoID>[0-9]+)(?:\.html)?(?:#.*)?$'
+    _TESTS = [{
+        'url': 'http://www.spiegel.de/video/vulkan-tungurahua-in-ecuador-ist-wieder-aktiv-video-1259285.html',
+        'file': '1259285.mp4',
+        'md5': '2c2754212136f35fb4b19767d242f66e',
+        'info_dict': {
+            'title': 'Vulkanausbruch in Ecuador: Der "Feuerschlund" ist wieder aktiv',
+        },
+    },
+    {
+        'url': 'http://www.spiegel.de/video/schach-wm-videoanalyse-des-fuenften-spiels-video-1309159.html',
+        'file': '1309159.mp4',
+        'md5': 'f2cdf638d7aa47654e251e1aee360af1',
+        'info_dict': {
+            'title': 'Schach-WM in der Videoanalyse: Carlsen nutzt die Fehlgriffe des Titelverteidigers',
+        },
+    }]
 
     def _real_extract(self, url):
         m = re.match(self._VALID_URL, url)
@@ -13,25 +30,42 @@ class SpiegelIE(InfoExtractor):
 
         webpage = self._download_webpage(url, video_id)
 
-        video_title = self._html_search_regex(r'<div class="module-title">(.*?)</div>',
-            webpage, u'title')
+        video_title = self._html_search_regex(
+            r'<div class="module-title">(.*?)</div>', webpage, 'title')
+
+        base_url = self._search_regex(
+            r'var\s+server\s+=\s+\"(http://video\d*\.spiegel\.de/flash/\d+/\d+/)\";',
+            webpage,
+            'base_url',
+        )
+
+        xml_url = base_url + video_id + '.xml'
+        idoc = self._download_xml(
+            xml_url, video_id,
+            note='Downloading XML', errnote='Failed to download XML from "{0}"'.format(xml_url))
 
-        xml_url = u'http://video2.spiegel.de/flash/' + video_id + u'.xml'
-        xml_code = self._download_webpage(xml_url, video_id,
-                    note=u'Downloading XML', errnote=u'Failed to download XML')
+        formats = [
+            {
+                'format_id': n.tag.rpartition('type')[2],
+                'url': 'http://video2.spiegel.de/flash/' + n.find('./filename').text,
+                'width': int(n.find('./width').text),
+                'height': int(n.find('./height').text),
+                'abr': int(n.find('./audiobitrate').text),
+                'vbr': int(n.find('./videobitrate').text),
+                'vcodec': n.find('./codec').text,
+                'acodec': 'MP4A',
+            }
+            for n in list(idoc)
+            # Blacklist type 6, it's extremely LQ and not available on the same server
+            if n.tag.startswith('type') and n.tag != 'type6'
+        ]
+        duration = float(idoc[0].findall('./duration')[0].text)
 
-        idoc = xml.etree.ElementTree.fromstring(xml_code)
-        last_type = idoc[-1]
-        filename = last_type.findall('./filename')[0].text
-        duration = float(last_type.findall('./duration')[0].text)
+        self._sort_formats(formats)
 
-        video_url = 'http://video2.spiegel.de/flash/' + filename
-        video_ext = filename.rpartition('.')[2]
-        info = {
+        return {
             'id': video_id,
-            'url': video_url,
-            'ext': video_ext,
             'title': video_title,
             'duration': duration,
+            'formats': formats,
         }
-        return [info]