[aparat] Improve extraction and extract more metadata (closes #17445, closes #18008)

youtube_dl/extractor/aparat.py

@@ -4,6 +4,7 @@ from __future__ import unicode_literals
 from .common import InfoExtractor
 from ..utils import (
     int_or_none,
+    merge_dicts,
     mimetype2ext,
     url_or_none,
 )
@@ -12,21 +13,32 @@ from ..utils import (
 class AparatIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?aparat\.com/(?:v/|video/video/embed/videohash/)(?P<id>[a-zA-Z0-9]+)'
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.aparat.com/v/wP8On',
         'md5': '131aca2e14fe7c4dcb3c4877ba300c89',
         'info_dict': {
             'id': 'wP8On',
             'ext': 'mp4',
             'title': 'تیم گلکسی 11 - زومیت',
-            'age_limit': 0,
+            'description': 'md5:096bdabcdcc4569f2b8a5e903a3b3028',
+            'duration': 231,
+            'timestamp': 1387394859,
+            'upload_date': '20131218',
+            'view_count': int,
         },
-        # 'skip': 'Extremely unreliable',
+    }, {
-    }
+        # multiple formats
+        'url': 'https://www.aparat.com/v/8dflw/',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
+        # Provides more metadata
+        webpage = self._download_webpage(url, video_id, fatal=False)
+
+        if not webpage:
             # Note: There is an easier-to-parse configuration at
             # http://www.aparat.com/video/video/config/videohash/%video_id
             # but the URL in there does not work
@@ -34,37 +46,50 @@ class AparatIE(InfoExtractor):
                 'http://www.aparat.com/video/video/embed/vt/frame/showvideo/yes/videohash/' + video_id,
                 video_id)
 
-        file_list = self._parse_json(
+        options = self._parse_json(
             self._search_regex(
-                r'var options\s*=\s*JSON\.parse\(\'([^\']+)\'\)', webpage,
+                r'options\s*=\s*JSON\.parse\(\s*(["\'])(?P<value>(?:(?!\1).)+)\1\s*\)',
-                'file list'),
+                webpage, 'options', group='value'),
             video_id)
 
-        title = file_list['plugins']['sabaPlayerPlugin']['title']
+        player = options['plugins']['sabaPlayerPlugin']
 
         formats = []
-        for list in file_list['plugins']['sabaPlayerPlugin']['multiSRC']:
+        for sources in player['multiSRC']:
-            for item in list:
+            for item in sources:
+                if not isinstance(item, dict):
+                    continue
                 file_url = url_or_none(item.get('src'))
                 if not file_url:
                     continue
+                item_type = item.get('type')
+                if item_type == 'application/vnd.apple.mpegurl':
+                    formats.extend(self._extract_m3u8_formats(
+                        file_url, video_id, 'mp4',
+                        entry_protocol='m3u8_native', m3u8_id='hls',
+                        fatal=False))
+                else:
                     ext = mimetype2ext(item.get('type'))
                     label = item.get('label')
                     formats.append({
                         'url': file_url,
                         'ext': ext,
-                    'format_id': label or ext,
+                        'format_id': 'http-%s' % (label or ext),
                         'height': int_or_none(self._search_regex(
-                        r'(\d+)[pP]', label or '', 'height', default=None)),
+                            r'(\d+)[pP]', label or '', 'height',
+                            default=None)),
                     })
-        self._sort_formats(formats)
+        self._sort_formats(
+            formats, field_preference=('height', 'width', 'tbr', 'format_id'))
 
-        thumbnail = file_list['poster']
+        info = self._search_json_ld(webpage, video_id, default={})
 
-        return {
+        if not info.get('title'):
+            info['title'] = player['title']
+
+        return merge_dicts(info, {
             'id': video_id,
-            'title': title,
+            'thumbnail': url_or_none(options.get('poster')),
-            'thumbnail': thumbnail,
+            'duration': int_or_none(player.get('duration')),
-            'age_limit': self._family_friendly_search(webpage),
             'formats': formats,
-        }
+        })