[tvn24] Improve extraction (closes #11679)

youtube_dl/extractor/tvn24.py

@@ -2,11 +2,15 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from ..utils import (
+    int_or_none,
+    unescapeHTML,
+)
 
 
 class TVN24IE(InfoExtractor):
-    _VALID_URL = r'http://(?:tvn24bis|(?:www|fakty)\.tvn24)\.pl/.+/(?P<id>[^/]+)\.html'
+    _VALID_URL = r'https?://(?:(?:[^/]+)\.)?tvn24(?:bis)?\.pl/(?:[^/]+/)*(?P<id>[^/]+)\.html'
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.tvn24.pl/wiadomosci-z-kraju,3/oredzie-artura-andrusa,702428.html',
         'md5': 'fbdec753d7bc29d96036808275f2130c',
         'info_dict': {
@@ -16,28 +20,53 @@ class TVN24IE(InfoExtractor):
             'description': 'Wyjątkowe orędzie Artura Andrusa, jednego z gości "Szkła kontaktowego".',
             'thumbnail': 're:http://.*[.]jpeg',
         }
-    }
+    }, {
+        'url': 'http://fakty.tvn24.pl/ogladaj-online,60/53-konferencja-bezpieczenstwa-w-monachium,716431.html',
+        'only_matching': True,
+    }, {
+        'url': 'http://sport.tvn24.pl/pilka-nozna,105/ligue-1-kamil-glik-rozcial-glowe-monaco-tylko-remisuje-z-bastia,716522.html',
+        'only_matching': True,
+    }, {
+        'url': 'http://tvn24bis.pl/poranek,146,m/gen-koziej-w-tvn24-bis-wracamy-do-czasow-zimnej-wojny,715660.html',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
-        page_id = self._match_id(url)
+        video_id = self._match_id(url)
-        webpage = self._download_webpage(url, page_id)
+
+        webpage = self._download_webpage(url, video_id)
+
         title = self._og_search_title(webpage)
-        description = self._og_search_description(webpage)
+
-        thumbnail = self._html_search_regex(r'\bdata-poster="(.+?)"', webpage, 'data-poster')
+        def extract_json(attr, name, fatal=True):
-        share_params = self._html_search_regex(r'\bdata-share-params="(.+?)"', webpage, 'data-share-params')
+            return self._parse_json(
-        share_params = self._parse_json(share_params, page_id)
+                self._search_regex(
-        video_id = share_params['id']
+                    r'\b%s=(["\'])(?P<json>(?!\1).+?)\1' % attr, webpage,
-        quality_data = self._html_search_regex(r'\bdata-quality="(.+?)"', webpage, 'data-quality')
+                    name, group='json', fatal=fatal) or '{}',
-        quality_data = self._parse_json(quality_data, page_id)
+                video_id, transform_source=unescapeHTML, fatal=fatal)
+
+        quality_data = extract_json('data-quality', 'formats')
+
         formats = []
         for format_id, url in quality_data.items():
             formats.append({
-                'format_id': format_id,
-                'height': int(format_id.rstrip('p')),
                 'url': url,
-                'ext': 'mp4',
+                'format_id': format_id,
+                'height': int_or_none(format_id.rstrip('p')),
             })
         self._sort_formats(formats)
+
+        description = self._og_search_description(webpage)
+        thumbnail = self._og_search_thumbnail(
+            webpage, default=None) or self._html_search_regex(
+            r'\bdata-poster=(["\'])(?P<url>(?!\1).+?)\1', webpage,
+            'thumbnail', group='url')
+
+        share_params = extract_json(
+            'data-share-params', 'share params', fatal=False)
+        if isinstance(share_params, dict):
+            video_id = share_params.get('id') or video_id
+
         return {
             'id': video_id,
             'title': title,