[npr] Fix extractor (Closes #7218)

youtube_dl/extractor/npr.py

@@ -1,71 +1,82 @@
-# coding: utf-8
 from __future__ import unicode_literals
 
-import os.path
-import re
-
-from ..compat import compat_urllib_parse_unquote
-from ..utils import url_basename
 from .common import InfoExtractor
+from ..compat import compat_urllib_parse
+from ..utils import (
+    int_or_none,
+    qualities,
+)
+
 
 class NprIE(InfoExtractor):
-    _VALID_URL = r'http://(?:www\.)?npr\.org/player/v2/mediaPlayer.html?.*id=(?P<id>[0-9]+)'
+    _VALID_URL = r'http://(?:www\.)?npr\.org/player/v2/mediaPlayer\.html\?.*\bid=(?P<id>\d+)'
-    _TEST = {
+    _TESTS = [{
-    'url': 'http://www.npr.org/player/v2/mediaPlayer.html?id=449974205',
+        'url': 'http://www.npr.org/player/v2/mediaPlayer.html?id=449974205',
-    'info_dict': {
+        'info_dict': {
-        'id': '449974205',
+            'id': '449974205',
-        'ext': 'mp4',
+            'title': 'New Music From Beach House, Chairlift, CMJ Discoveries And More'
-        'title': 'New Music From Beach House, Chairlift, CMJ Discoveries And More'
+        },
-    }
+        'playlist_count': 7,
-}
+    }, {
-
+        'url': 'http://www.npr.org/player/v2/mediaPlayer.html?action=1&t=1&islist=false&id=446928052&m=446929930&live=1',
+        'info_dict': {
+            'id': '446928052',
+            'title': "Songs We Love: Tigran Hamasyan, 'Your Mercy is Boundless'"
+        },
+        'playlist': [{
+            'md5': '12fa60cb2d3ed932f53609d4aeceabf1',
+            'info_dict': {
+                'id': '446929930',
+                'ext': 'mp3',
+                'title': 'Your Mercy is Boundless (Bazum en Qo gtutyunqd)',
+                'duration': 402,
+            },
+        }],
+    }]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
+        playlist_id = self._match_id(url)
-        video_id = mobj.group('id')
+
-        webpage_url = 'http://www.npr.org/player/v2/mediaPlayer.html?id=' + video_id
+        config = self._download_json(
-        webpage = self._download_webpage(webpage_url, video_id)
+            'http://api.npr.org/query?%s' % compat_urllib_parse.urlencode({
-        key = 'MDAzMzQ2MjAyMDEyMzk4MTU1MDg3ZmM3MQ010'
+                'id': playlist_id,
-        xml_url = 'http://api.npr.org/query?id=%s&apiKey=%s' % (video_id, key)
+                'fields': 'titles,audio,show',
-        json_url = 'http://api.npr.org/query?id=%s&apiKey=%s&format=json' % (video_id, key)
+                'format': 'json',
+                'apiKey': 'MDAzMzQ2MjAyMDEyMzk4MTU1MDg3ZmM3MQ010',
+            }), playlist_id)
+
+        story = config['list']['story'][0]
+
+        KNOWN_FORMATS = ('threegp', 'mp4', 'mp3')
+        quality = qualities(KNOWN_FORMATS)
 
-        formats = []
         entries = []
+        for audio in story.get('audio', []):
+            title = audio.get('title', {}).get('$text')
+            duration = int_or_none(audio.get('duration', {}).get('$text'))
+            formats = []
+            for format_id, formats_entry in audio.get('format', {}).items():
+                if not formats_entry:
+                    continue
+                if isinstance(formats_entry, list):
+                    formats_entry = formats_entry[0]
+                format_url = formats_entry.get('$text')
+                if not format_url:
+                    continue
+                if format_id in KNOWN_FORMATS:
+                    formats.append({
+                        'url': format_url,
+                        'format_id': format_id,
+                        'ext': formats_entry.get('type'),
+                        'quality': quality(format_id),
+                    })
+            self._sort_formats(formats)
+            entries.append({
+                'id': audio['id'],
+                'title': title,
+                'duration': duration,
+                'formats': formats,
+            })
 
-        config = self._download_json(json_url, video_id)
+        playlist_title = story.get('title', {}).get('$text')
-
+        return self.playlist_result(entries, playlist_id, playlist_title)
-        content = config["list"]["story"]
-
-        album_title = config["list"]["story"][0]['song'][0]['album']['albumTitle']
-        print album_title['$text']
-
-        for key in content:
-            if "audio" in key:
-                for x in key['audio']:
-                    if x['type'] == 'standard':
-                        playlist = True
-                        song_duration = x["duration"]['$text']
-                        song_title = x["title"]["$text"]
-                        song_id = x["id"]
-
-                        for k in x["format"]:
-                            if type(x["format"][k]) is list:
-                                for z in x["format"][k]:
-                                    formats.append({ 'format': z['type'], 
-                                                     'url'   : z['$text']
-                                              })
-                            else:
-                                formats.append({ 'format': k, 
-                                                 'url'   : x["format"][k]['$text']
-                                      })
-
-                        entries.append({ "title":song_title,
-                                         "id":song_id,
-                                         "duration": song_duration ,
-                                         "formats":formats})
-                        formats = []
-
-        return {    '_type': 'playlist',
-                    'id' : video_id,
-                    'title' : album_title,
-                    'entries': entries  }