[stitcher] Improve (Closes #7162, closes #7228)

youtube_dl/extractor/stitcher.py

@@ -1,37 +1,81 @@
-# coding: utf-8
 from __future__ import unicode_literals
+
+import re
+
 from .common import InfoExtractor
-from ..utils import int_or_none
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    js_to_json,
+    unescapeHTML,
+)
 
 
 class StitcherIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?stitcher\.com/podcast/[\/a-z\-]+(?P<id>\d+)'
-    _TEST = {
+    _VALID_URL = r'https?://(?:www\.)?stitcher\.com/podcast/(?:[^/]+/)+e/(?:(?P<display_id>[^/#?&]+?)-)?(?P<id>\d+)(?:[/#?&]|$)'
+    _TESTS = [{
         'url': 'http://www.stitcher.com/podcast/the-talking-machines/e/40789481?autoplay=true',
         'md5': '391dd4e021e6edeb7b8e68fbf2e9e940',
         'info_dict': {
             'id': '40789481',
             'ext': 'mp3',
-            'title': 'Machine Learning Mastery and Cancer Clusters from Talking Machines',
-        }
-    }
+            'title': 'Machine Learning Mastery and Cancer Clusters',
+            'description': 'md5:55163197a44e915a14a1ac3a1de0f2d3',
+            'duration': 1604,
+            'thumbnail': 're:^https?://.*\.jpg',
+        },
+    }, {
+        'url': 'http://www.stitcher.com/podcast/panoply/vulture-tv/e/the-rare-hourlong-comedy-plus-40846275?autoplay=true',
+        'info_dict': {
+            'id': '40846275',
+            'display_id': 'the-rare-hourlong-comedy-plus',
+            'ext': 'mp3',
+            'title': "The CW's 'Crazy Ex-Girlfriend'",
+            'description': 'md5:04f1e2f98eb3f5cbb094cea0f9e19b17',
+            'duration': 2235,
+            'thumbnail': 're:^https?://.*\.jpg',
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        # escaped title
+        'url': 'http://www.stitcher.com/podcast/marketplace-on-stitcher/e/40910226?autoplay=true',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.stitcher.com/podcast/panoply/getting-in/e/episode-2a-how-many-extracurriculars-should-i-have-40876278?autoplay=true',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
-        audio_id = self._match_id(url)
+        mobj = re.match(self._VALID_URL, url)
+        audio_id = mobj.group('id')
+        display_id = mobj.group('display_id') or audio_id
 
-        webpage = self._download_webpage(url, audio_id)
+        webpage = self._download_webpage(url, display_id)
 
-        title = self._og_search_title(webpage)
-        url = self._search_regex(r'episodeURL: "(.+?)"', webpage, 'url')
-        episode_image = self._search_regex(r'episodeImage: "(.+?)"', webpage, 'episode_image', fatal=False)
-        duration = int_or_none(self._search_regex(r'duration: (\d+?),', webpage, 'duration', fatal=False))
+        episode = self._parse_json(
+            js_to_json(self._search_regex(
+                r'(?s)var\s+stitcher\s*=\s*({.+?});\n', webpage, 'episode config')),
+            display_id)['config']['episode']
+
+        title = unescapeHTML(episode['title'])
+        formats = [{
+            'url': episode[episode_key],
+            'ext': determine_ext(episode[episode_key]) or 'mp3',
+            'vcodec': 'none',
+        } for episode_key in ('origEpisodeURL', 'episodeURL') if episode.get(episode_key)]
+        description = self._search_regex(
+            r'Episode Info:\s*</span>([^<]+)<', webpage, 'description', fatal=False)
+        duration = int_or_none(episode.get('duration'))
+        thumbnail = episode.get('episodeImage')
 
         return {
             'id': audio_id,
-            'url': url,
+            'display_id': display_id,
             'title': title,
+            'description': description,
             'duration': duration,
-            'thumbnail': episode_image,
-            'ext': 'mp3',
-            'vcodec': 'none',
+            'thumbnail': thumbnail,
+            'formats': formats,
         }