[libsyn] improve extraction(closes #20229)

youtube_dl/extractor/libsyn.py

@@ -1,12 +1,14 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import json
 import re
 
 from .common import InfoExtractor
 from ..utils import (
+    clean_html,
+    get_element_by_class,
     parse_duration,
+    strip_or_none,
     unified_strdate,
 )
 
@@ -21,7 +23,9 @@ class LibsynIE(InfoExtractor):
             'id': '6385796',
             'ext': 'mp3',
             'title': "Champion Minded - Developing a Growth Mindset",
-            'description': 'In this episode, Allistair talks about the importance of developing a growth mindset, not only in sports, but in life too.',
+            # description fetched using another request:
+            # http://html5-player.libsyn.com/embed/getitemdetails?item_id=6385796
+            # 'description': 'In this episode, Allistair talks about the importance of developing a growth mindset, not only in sports, but in life too.',
             'upload_date': '20180320',
             'thumbnail': 're:^https?://.*',
         },
@@ -38,22 +42,36 @@ class LibsynIE(InfoExtractor):
     }]
 
     def _real_extract(self, url):
-        m = re.match(self._VALID_URL, url)
+        url, video_id = re.match(self._VALID_URL, url).groups()
-        video_id = m.group('id')
-        url = m.group('mainurl')
         webpage = self._download_webpage(url, video_id)
 
-        podcast_title = self._search_regex(
+        data = self._parse_json(self._search_regex(
-            r'<h3>([^<]+)</h3>', webpage, 'podcast title', default=None)
+            r'var\s+playlistItem\s*=\s*({.+?});',
-        if podcast_title:
+            webpage, 'JSON data block'), video_id)
-            podcast_title = podcast_title.strip()
+
-        episode_title = self._search_regex(
+        episode_title = data.get('item_title') or get_element_by_class('episode-title', webpage)
-            r'(?:<div class="episode-title">|<h4>)([^<]+)</', webpage, 'episode title')
+        if not episode_title:
-        if episode_title:
+            self._search_regex(
-            episode_title = episode_title.strip()
+                [r'data-title="([^"]+)"', r'<title>(.+?)</title>'],
+                webpage, 'episode title')
+        episode_title = episode_title.strip()
+
+        podcast_title = strip_or_none(clean_html(self._search_regex(
+            r'<h3>([^<]+)</h3>', webpage, 'podcast title',
+            default=None) or get_element_by_class('podcast-title', webpage)))
 
         title = '%s - %s' % (podcast_title, episode_title) if podcast_title else episode_title
 
+        formats = []
+        for k, format_id in (('media_url_libsyn', 'libsyn'), ('media_url', 'main'), ('download_link', 'download')):
+            f_url = data.get(k)
+            if not f_url:
+                continue
+            formats.append({
+                'url': f_url,
+                'format_id': format_id,
+            })
+
         description = self._html_search_regex(
             r'<p\s+id="info_text_body">(.+?)</p>', webpage,
             'description', default=None)
@@ -61,27 +79,15 @@ class LibsynIE(InfoExtractor):
             # Strip non-breaking and normal spaces
             description = description.replace('\u00A0', ' ').strip()
         release_date = unified_strdate(self._search_regex(
-            r'<div class="release_date">Released: ([^<]+)<', webpage, 'release date', fatal=False))
+            r'<div class="release_date">Released: ([^<]+)<',
-
+            webpage, 'release date', default=None) or data.get('release_date'))
-        data_json = self._search_regex(r'var\s+playlistItem\s*=\s*(\{.*?\});\n', webpage, 'JSON data block')
-        data = json.loads(data_json)
-
-        formats = [{
-            'url': data['media_url'],
-            'format_id': 'main',
-        }, {
-            'url': data['media_url_libsyn'],
-            'format_id': 'libsyn',
-        }]
-        thumbnail = data.get('thumbnail_url')
-        duration = parse_duration(data.get('duration'))
 
         return {
             'id': video_id,
             'title': title,
             'description': description,
-            'thumbnail': thumbnail,
+            'thumbnail': data.get('thumbnail_url'),
             'upload_date': release_date,
-            'duration': duration,
+            'duration': parse_duration(data.get('duration')),
             'formats': formats,
         }