[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/mdr.py b/youtube_dl/extractor/mdr.py
index 541ddd9099fec7a3997585e478f8bf89a9907aed..322e5b45a141bacd327288ac58ed658a46e628f7 100644 (file)
--- a/youtube_dl/extractor/mdr.py
+++ b/youtube_dl/extractor/mdr.py
@@ -14,12 +14,24 @@ from ..utils import (
 
 class MDRIE(InfoExtractor):
     IE_DESC = 'MDR.DE and KiKA'
-    _VALID_URL = r'https?://(?:www\.)?(?:mdr|kika)\.de/(?:.*)/[a-z]+(?P<id>\d+)(?:_.+?)?\.html'
+    _VALID_URL = r'https?://(?:www\.)?(?:mdr|kika)\.de/(?:.*)/[a-z-]+-?(?P<id>\d+)(?:_.+?)?\.html'
 
     _TESTS = [{
-        # MDR regularily deletes its videos
+        # MDR regularly deletes its videos
         'url': 'http://www.mdr.de/fakt/video189002.html',
         'only_matching': True,
+    }, {
+        # audio
+        'url': 'http://www.mdr.de/kultur/audio1312272_zc-15948bad_zs-86171fdd.html',
+        'md5': '64c4ee50f0a791deb9479cd7bbe9d2fa',
+        'info_dict': {
+            'id': '1312272',
+            'ext': 'mp3',
+            'title': 'Feuilleton vom 30. Oktober 2015',
+            'duration': 250,
+            'uploader': 'MITTELDEUTSCHER RUNDFUNK',
+        },
+        'skip': '404 not found',
     }, {
         'url': 'http://www.kika.de/baumhaus/videos/video19636.html',
         'md5': '4930515e36b06c111213e80d1e4aad0e',
@@ -30,6 +42,7 @@ class MDRIE(InfoExtractor):
             'duration': 134,
             'uploader': 'KIKA',
         },
+        'skip': '404 not found',
     }, {
         'url': 'http://www.kika.de/sendungen/einzelsendungen/weihnachtsprogramm/videos/video8182.html',
         'md5': '5fe9c4dd7d71e3b238f04b8fdd588357',
@@ -38,17 +51,30 @@ class MDRIE(InfoExtractor):
             'ext': 'mp4',
             'title': 'Beutolomäus und der geheime Weihnachtswunsch',
             'description': 'md5:b69d32d7b2c55cbe86945ab309d39bbd',
-            'timestamp': 1419047100,
-            'upload_date': '20141220',
+            'timestamp': 1482541200,
+            'upload_date': '20161224',
             'duration': 4628,
             'uploader': 'KIKA',
         },
+    }, {
+        # audio with alternative playerURL pattern
+        'url': 'http://www.mdr.de/kultur/videos-und-audios/audio-radio/operation-mindfuck-robert-wilson100.html',
+        'info_dict': {
+            'id': '100',
+            'ext': 'mp4',
+            'title': 'Feature: Operation Mindfuck - Robert Anton Wilson',
+            'duration': 3239,
+            'uploader': 'MITTELDEUTSCHER RUNDFUNK',
+        },
     }, {
         'url': 'http://www.kika.de/baumhaus/sendungen/video19636_zc-fea7f8a0_zs-4bf89c60.html',
         'only_matching': True,
     }, {
         'url': 'http://www.kika.de/sendungen/einzelsendungen/weihnachtsprogramm/einzelsendung2534.html',
         'only_matching': True,
+    }, {
+        'url': 'http://www.mdr.de/mediathek/mdr-videos/a/video-1334.html',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
@@ -57,14 +83,13 @@ class MDRIE(InfoExtractor):
         webpage = self._download_webpage(url, video_id)
 
         data_url = self._search_regex(
-            r'dataURL\s*:\s*(["\'])(?P<url>/.+/(?:video|audio)[0-9]+-avCustom\.xml)\1',
-            webpage, 'data url', group='url')
+            r'(?:dataURL|playerXml(?:["\'])?)\s*:\s*(["\'])(?P<url>.+?-avCustom\.xml)\1',
+            webpage, 'data url', group='url').replace(r'\/', '/')
 
         doc = self._download_xml(
             compat_urlparse.urljoin(url, data_url), video_id)
 
-        title = (xpath_text(doc, './title', 'title', default=None) or
-                 xpath_text(doc, './broadcast/broadcastName', 'title'))
+        title = xpath_text(doc, ['./title', './broadcast/broadcastName'], 'title', fatal=True)
 
         formats = []
         processed_urls = []
@@ -86,8 +111,6 @@ class MDRIE(InfoExtractor):
                 vbr = int_or_none(xpath_text(asset, './bitrateVideo', 'vbr'), 1000)
                 abr = int_or_none(xpath_text(asset, './bitrateAudio', 'abr'), 1000)
 
-                url_formats = []
-
                 ext = determine_ext(url_el.text)
                 if ext == 'm3u8':
                     url_formats = self._extract_m3u8_formats(
@@ -120,7 +143,10 @@ class MDRIE(InfoExtractor):
                             'height': height,
                         })
 
-                    url_formats.append(f)
+                    url_formats = [f]
+
+                if not url_formats:
+                    continue
 
                 if not vbr:
                     for f in url_formats:
@@ -132,14 +158,18 @@ class MDRIE(InfoExtractor):
                             'vcodec': 'none',
                         })
 
-                if url_formats:
-                    formats.extend(url_formats)
+                formats.extend(url_formats)
+
         self._sort_formats(formats)
 
         description = xpath_text(doc, './broadcast/broadcastDescription', 'description')
         timestamp = parse_iso8601(
-            xpath_text(doc, './broadcast/broadcastDate', 'timestamp', default=None) or
-            xpath_text(doc, './broadcast/broadcastStartDate', 'timestamp', default=None))
+            xpath_text(
+                doc, [
+                    './broadcast/broadcastDate',
+                    './broadcast/broadcastStartDate',
+                    './broadcast/broadcastEndDate'],
+                'timestamp', default=None))
         duration = parse_duration(xpath_text(doc, './duration', 'duration'))
         uploader = xpath_text(doc, './rights', 'uploader')