[ndr] fix info extraction

youtube_dl/extractor/ndr.py

@@ -1,68 +1,66 @@
 # encoding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 from ..utils import (
     ExtractorError,
     int_or_none,
-    qualities,
-    parse_duration,
 )
 
 
+preferences = {'xl': 4, 'l': 3, 'm': 2, 's': 1, 'xs': 0,}
+
+
 class NDRBaseIE(InfoExtractor):
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
+        video_id = self._match_id(url)
-        video_id = mobj.group('id')
 
-        page = self._download_webpage(url, video_id, 'Downloading page')
+        json_data = self._download_json('http://www.ndr.de/%s-ppjson.json' % video_id, video_id, 'Downloading page')
-
-        title = self._og_search_title(page).strip()
-        description = self._og_search_description(page)
-        if description:
-            description = description.strip()
-
-        duration = int_or_none(self._html_search_regex(r'duration: (\d+),\n', page, 'duration', default=None))
-        if not duration:
-            duration = parse_duration(self._html_search_regex(
-                r'(<span class="min">\d+</span>:<span class="sec">\d+</span>)',
-                page, 'duration', default=None))
 
         formats = []
-
+        objetType = json_data.get('config').get('objectType')
-        mp3_url = re.search(r'''\{src:'(?P<audio>[^']+)', type:"audio/mp3"},''', page)
+        if objetType == 'video':
-        if mp3_url:
+            for key, f in json_data.get('playlist').items():
+                if key != 'config':
+                    src = f['src']
+                    if '.f4m' in src:
+                        formats.extend(self._extract_f4m_formats(src, video_id))
+                    elif '.m3u8' in src:
+                        formats.extend(self._extract_m3u8_formats(src, video_id))
+                    else:
+                        quality = f.get('quality')
                         formats.append({
-                'url': mp3_url.group('audio'),
+                            'url': src,
+                            'format_id': quality,
+                            'preference': preferences.get(quality),
+                        })
+        elif objetType == 'audio':
+            for key, f in json_data.get('playlist').items():
+                if key != 'config':
+                    formats.append({
+                        'url': f['src'],
                         'format_id': 'mp3',
+                        
                     })
-
+        else:
-        thumbnail = None
-
-        video_url = re.search(r'''3: \{src:'(?P<video>.+?)\.(lo|hi|hq)\.mp4', type:"video/mp4"},''', page)
-        if video_url:
-            thumbnails = re.findall(r'''\d+: \{src: "([^"]+)"(?: \|\| '[^']+')?, quality: '([^']+)'}''', page)
-            if thumbnails:
-                quality_key = qualities(['xs', 's', 'm', 'l', 'xl'])
-                largest = max(thumbnails, key=lambda thumb: quality_key(thumb[1]))
-                thumbnail = 'http://www.ndr.de' + largest[0]
-
-            for format_id in 'lo', 'hi', 'hq':
-                formats.append({
-                    'url': '%s.%s.mp4' % (video_url.group('video'), format_id),
-                    'format_id': format_id,
-                })
-
-        if not formats:
             raise ExtractorError('No media links available for %s' % video_id)
 
+        self._sort_formats(formats)
+
+        config = json_data.get('playlist').get('config')
+
+        title = config['title']
+        duration = int_or_none(config.get('duration'))
+        thumbnails = [{
+            'id': thumbnail.get('quality'),
+            'url': thumbnail.get('src'),
+            'preference': preferences.get(thumbnail.get('quality'))
+        } for thumbnail in config.get('poster').values()]
+
         return {
             'id': video_id,
             'title': title,
-            'description': description,
+            'thumbnails': thumbnails,
-            'thumbnail': thumbnail,
             'duration': duration,
             'formats': formats,
         }
@@ -71,7 +69,7 @@ class NDRBaseIE(InfoExtractor):
 class NDRIE(NDRBaseIE):
     IE_NAME = 'ndr'
     IE_DESC = 'NDR.de - Mediathek'
-    _VALID_URL = r'https?://www\.ndr\.de/.+?(?P<id>\d+)\.html'
+    _VALID_URL = r'https?://www\.ndr\.de/.+?,(?P<id>\w+)\.html'
 
     _TESTS = [
         {
@@ -79,10 +77,9 @@ class NDRIE(NDRBaseIE):
             'md5': '5bc5f5b92c82c0f8b26cddca34f8bb2c',
             'note': 'Video file',
             'info_dict': {
-                'id': '25866',
+                'id': 'nordmagazin25866',
                 'ext': 'mp4',
                 'title': 'Kartoffeltage in der Lewitz',
-                'description': 'md5:48c4c04dde604c8a9971b3d4e3b9eaa8',
                 'duration': 166,
             },
             'skip': '404 Not found',
@@ -91,22 +88,20 @@ class NDRIE(NDRBaseIE):
             'url': 'http://www.ndr.de/fernsehen/Party-Poette-und-Parade,hafengeburtstag988.html',
             'md5': 'dadc003c55ae12a5d2f6bd436cd73f59',
             'info_dict': {
-                'id': '988',
+                'id': 'hafengeburtstag988',
                 'ext': 'mp4',
                 'title': 'Party, Pötte und Parade',
-                'description': 'Hunderttausende feiern zwischen Speicherstadt und St. Pauli den 826. Hafengeburtstag. Die NDR Sondersendung zeigt die schönsten und spektakulärsten Bilder vom Auftakt.',
                 'duration': 3498,
             },
         },
         {
-            'url': 'http://www.ndr.de/info/audio51535.html',
+            'url': 'http://www.ndr.de/info/La-Valette-entgeht-der-Hinrichtung,audio51535.html',
             'md5': 'bb3cd38e24fbcc866d13b50ca59307b8',
             'note': 'Audio file',
             'info_dict': {
-                'id': '51535',
+                'id': 'audio51535',
                 'ext': 'mp3',
                 'title': 'La Valette entgeht der Hinrichtung',
-                'description': 'md5:22f9541913a40fe50091d5cdd7c9f536',
                 'duration': 884,
             }
         }
@@ -115,7 +110,7 @@ class NDRIE(NDRBaseIE):
 
 class NJoyIE(NDRBaseIE):
     IE_NAME = 'N-JOY'
-    _VALID_URL = r'https?://www\.n-joy\.de/.+?(?P<id>\d+)\.html'
+    _VALID_URL = r'https?://www\.n-joy\.de/.+?,(?P<id>\w+)\.html'
 
     _TEST = {
         'url': 'http://www.n-joy.de/entertainment/comedy/comedy_contest/Benaissa-beim-NDR-Comedy-Contest,comedycontest2480.html',
@@ -124,7 +119,6 @@ class NJoyIE(NDRBaseIE):
             'id': '2480',
             'ext': 'mp4',
             'title': 'Benaissa beim NDR Comedy Contest',
-            'description': 'Von seinem sehr "behaarten" Leben lässt sich Benaissa trotz aller Schwierigkeiten nicht unterkriegen.',
             'duration': 654,
         }
     }