[vrak] Improve and update test (closes #11452)

youtube_dl/extractor/vrak.py

@@ -4,65 +4,77 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-
 from .brightcove import BrightcoveNewIE
+from ..utils import (
+    int_or_none,
+    parse_age_limit,
+    smuggle_url,
+    unescapeHTML,
+)
 
 
 class VrakIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?vrak\.tv/videos\?.*?target=(?P<id>[0-9\.]+).*'
+    _VALID_URL = r'https?://(?:www\.)?vrak\.tv/videos\?.*?\btarget=(?P<id>[\d.]+)'
     _TEST = {
-        'url': 'http://www.vrak.tv/videos?target=1.2240923&filtre=emission&id=1.1806721',
-        'md5': 'c5d5ce237bca3b1e990ce1b48d1f0948',
+        'url': 'http://www.vrak.tv/videos?target=1.2306782&filtre=emission&id=1.1806721',
         'info_dict': {
-            'id': '5231040869001',
+            'id': '5345661243001',
             'ext': 'mp4',
-            'title': 'Référendums américains, animés japonais et hooligans russes',
-            'upload_date': '20161201',
-            'description': 'This video file has been uploaded automatically using Oprah. It should be updated with real description soon.',
-            'timestamp': 1480628425,
+            'title': 'Obésité, film de hockey et Roseline Filion',
+            'timestamp': 1488492126,
+            'upload_date': '20170302',
             'uploader_id': '2890187628001',
+            'creator': 'VRAK.TV',
+            'age_limit': 8,
+            'series': 'ALT (Actualité Légèrement Tordue)',
+            'episode': 'Obésité, film de hockey et Roseline Filion',
+            'tags': list,
+        },
+        'params': {
+            'skip_download': True,
+        },
     }
-    }
+    BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/2890187628001/default_default/index.html?videoId=%s'
 
     def _real_extract(self, url):
-        url_id = self._match_id(url)
-        webpage = self._download_webpage(url, url_id)
+        video_id = self._match_id(url)
 
-        result = {}
-        result['title'] = self._html_search_regex(
-            r'<h3 class="videoTitle">(.+?)</h3>', webpage, 'title')
+        webpage = self._download_webpage(url, video_id)
 
-        # Inspired from BrightcoveNewIE._extract_url()
-        entries = []
-        for account_id, player_id, _, video_id in re.findall(
-                # account_id, player_id and embed from:
-                #   <div class="video-player [...]
-                #     data-publisher-id="2890187628001"
-                #     data-player-id="VkSnGw3cx"
-                # video id is extracted from weird CMS Java/Javascript notation:
-                #   RW java.lang.String value = '5231040869001';
-                # Need to use backtrack to pin to a ref since video is in grid
-                # layout with others
-                r'''(?sx)
-                    <div[^>]+
-                        data-publisher-id=["\'](\d+)["\']
-                        [^>]*
-                        data-player-id=["\']([^"\']+)["\']
-                        [^>]*
-                        refId&quot;:&quot;([^&]+)&quot;
-                        [^>]*
-                        >.*?
-                    </div>.*?
-                    RW\ java\.lang\.String\ value\ =\ \'brightcove\.article\.\d+\.\3\'
-                    [^>]*
-                    RW\ java\.lang\.String\ value\ =\ \'(\d+)\'
-                ''', webpage):
+        title = self._html_search_regex(
+            r'<h\d\b[^>]+\bclass=["\']videoTitle["\'][^>]*>([^<]+)',
+            webpage, 'title', default=None) or self._og_search_title(webpage)
+
+        content = self._parse_json(
+            self._search_regex(
+                r'data-player-options-content=(["\'])(?P<content>{.+?})\1',
+                webpage, 'content', default='{}', group='content'),
+            video_id, transform_source=unescapeHTML)
 
-            entries.append(
-                'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s'
-                % (account_id, player_id, 'default', video_id))
+        ref_id = content.get('refId') or self._search_regex(
+            r'refId&quot;:&quot;([^&]+)&quot;', webpage, 'ref id')
 
-        if entries:
-            result = self.url_result(entries[0], BrightcoveNewIE.ie_key())
+        brightcove_id = self._search_regex(
+            r'''(?x)
+                java\.lang\.String\s+value\s*=\s*["']brightcove\.article\.\d+\.%s
+                [^>]*
+                java\.lang\.String\s+value\s*=\s*["'](\d+)
+            ''' % re.escape(ref_id), webpage, 'brightcove id')
 
-        return result
+        return {
+            '_type': 'url_transparent',
+            'ie_key': BrightcoveNewIE.ie_key(),
+            'url': smuggle_url(
+                self.BRIGHTCOVE_URL_TEMPLATE % brightcove_id,
+                {'geo_countries': ['CA']}),
+            'id': brightcove_id,
+            'description': content.get('description'),
+            'creator': content.get('brand'),
+            'age_limit': parse_age_limit(content.get('rating')),
+            'series': content.get('showName') or content.get(
+                'episodeName'),  # this is intentional
+            'season_number': int_or_none(content.get('seasonNumber')),
+            'episode': title,
+            'episode_number': int_or_none(content.get('episodeNumber')),
+            'tags': content.get('tags', []),
+        }