[europa] Improve extraction

1 files changed, 58 insertions, 28 deletions

diff --git a/youtube_dl/extractor/europa.py b/youtube_dl/extractor/europa.py
index c437c48..02ba8d6 100644
--- a/youtube_dl/extractor/europa.py
+++ b/youtube_dl/extractor/europa.py
@@ -2,59 +2,89 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from ..compat import compat_urlparse
 from ..utils import (
-    compat_urlparse,
+    int_or_none,
+    orderedSet,
+    parse_duration,
+    qualities,
+    unified_strdate,
     xpath_text
 )
 
 
 class EuropaIE(InfoExtractor):
-    _VALID_URL = r'https?://ec\.europa\.eu/avservices/video/player\.cfm\?(?:[^&]|&(?!ref))*ref=(?P<id>[A-Za-z0-9]+)'
-    _TEST = {
+    _VALID_URL = r'https?://ec\.europa\.eu/avservices/video/player\.cfm\?.*?\bref=(?P<id>[A-Za-z0-9]+)'
+    _TESTS = [{
         'url': 'http://ec.europa.eu/avservices/video/player.cfm?ref=I107758',
-        'md5': '728cca2fd41d5aa7350cec1141fbe620',
+        'md5': '574f080699ddd1e19a675b0ddf010371',
         'info_dict': {
             'id': 'I107758',
             'ext': 'mp4',
             'title': 'TRADE - Wikileaks on TTIP',
             'description': 'NEW  LIVE EC Midday press briefing of 11/08/2015',
-            'thumbnail': 're:^http://defiris\.ec\.streamcloud\.be/findmedia/18/107758/THUMB_[0-9A-Z]+\.jpg$'
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'upload_date': '20150811',
+            'duration': 34,
+            'view_count': int,
+            'formats': 'mincount:3',
         }
-    }
+    }, {
+        'url': 'http://ec.europa.eu/avservices/video/player.cfm?sitelang=en&ref=I107786',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
-        lang = query.get('sitelang', ['en'])[0]
 
-        playlist = self._download_xml('http://ec.europa.eu/avservices/video/player/playlist.cfm?ID=' + video_id, video_id)
-        videos = {}
-        formats = []
+        playlist = self._download_xml(
+            'http://ec.europa.eu/avservices/video/player/playlist.cfm?ID=%s' % video_id, video_id)
 
-        for item in playlist.findall('info/title/item'):
-            videos[xpath_text(item, 'lg')] = {'title': xpath_text(item, 'label').strip()}
+        def get_item(type_, preference):
+            items = {}
+            for item in playlist.findall('./info/%s/item' % type_):
+                lang, label = xpath_text(item, 'lg', default=None), xpath_text(item, 'label', default=None)
+                if lang and label:
+                    items[lang] = label.strip()
+            for p in preference:
+                if items.get(p):
+                    return items[p]
 
-        for item in playlist.findall('info/description/item'):
-            videos[xpath_text(item, 'lg')]['description'] = xpath_text(item, 'label').strip()
+        query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
+        preferred_lang = query.get('sitelang', ('en', ))[0]
 
-        for item in playlist.findall('files/file'):
-            lg = xpath_text(item, 'lg')
-            vid = videos[lg]
-            vid['format_note'] = xpath_text(item, 'lglabel')
-            vid['url'] = xpath_text(item, 'url')
+        preferred_langs = orderedSet((preferred_lang, 'en', 'int'))
 
-            if lg == lang:
-                vid['language_preference'] = 10
+        title = get_item('title', preferred_langs) or video_id
+        description = get_item('description', preferred_langs)
+        thumbnmail = xpath_text(playlist, './info/thumburl', 'thumbnail')
+        upload_date = unified_strdate(xpath_text(playlist, './info/date', 'upload date'))
+        duration = parse_duration(xpath_text(playlist, './info/duration', 'duration'))
+        view_count = int_or_none(xpath_text(playlist,'./info/views', 'views'))
 
-            formats.append(vid)
+        language_preference = qualities(preferred_langs[::-1])
 
-        formats.reverse()
-        def_video = videos.get(lang, videos['int'])
+        formats = []
+        for file_ in playlist.findall('./files/file'):
+            video_url = xpath_text(file_, './url')
+            if not video_url:
+                continue
+            lang = xpath_text(file_, './lg')
+            formats.append({
+                'url': video_url,
+                'format_id': lang,
+                'format_note': xpath_text(file_, './lglabel'),
+                'language_preference': language_preference(lang)
+            })
+        self._sort_formats(formats)
 
         return {
             'id': video_id,
-            'title': def_video['title'],
-            'description': def_video['description'],
-            'thumbnail': xpath_text(playlist, 'info/thumburl', 'thumburl'),
+            'title': title,
+            'description': description,
+            'thumbnail': thumbnmail,
+            'upload_date': upload_date,
+            'duration': duration,
+            'view_count': view_count,
             'formats': formats
         }