[ard] Support multiple formats (Closes #2247)

youtube_dl/extractor/ard.py

@@ -1,22 +1,28 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
 import re
 
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
     ExtractorError,
 )
 
+
 class ARDIE(InfoExtractor):
-    _VALID_URL = r'^(?:https?://)?(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[^/\?]+)(?:\?.*)?'
-    _TITLE = r'<h1(?: class="boxTopHeadline")?>(?P<title>.*)</h1>'
-    _MEDIA_STREAM = r'mediaCollection\.addMediaStream\((?P<media_type>\d+), (?P<quality>\d+), "(?P<rtmp_url>[^"]*)", "(?P<video_url>[^"]*)", "[^"]*"\)'
+    _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[^/\?]+)(?:\?.*)?'
+
     _TEST = {
-        u'url': u'http://www.ardmediathek.de/das-erste/tagesschau-in-100-sek?documentId=14077640',
-        u'file': u'14077640.mp4',
-        u'md5': u'6ca8824255460c787376353f9e20bbd8',
-        u'info_dict': {
-            u"title": u"11.04.2013 09:23 Uhr - Tagesschau in 100 Sekunden"
+        'url': 'http://www.ardmediathek.de/das-erste/guenther-jauch/edward-snowden-im-interview-held-oder-verraeter?documentId=19288786',
+        'file': '19288786.mp4',
+        'md5': '515bf47ce209fb3f5a61b7aad364634c',
+        'info_dict': {
+            'title': 'Edward Snowden im Interview - Held oder Verräter?',
+            'description': 'Edward Snowden hat alles aufs Spiel gesetzt, um die weltweite \xdcberwachung durch die Geheimdienste zu enttarnen. Nun stellt sich der ehemalige NSA-Mitarbeiter erstmals weltweit in einem TV-Interview den Fragen eines NDR-Journalisten. Die Sendung vom Sonntagabend.',
+            'thumbnail': 'http://www.ardmediathek.de/ard/servlet/contentblob/19/28/87/90/19288790/bild/2250037',
         },
-        u'skip': u'Requires rtmpdump'
+        'skip': 'Blocked outside of Germany',
     }
 
     def _real_extract(self, url):
@@ -29,26 +35,49 @@ class ARDIE(InfoExtractor):
         else:
             video_id = m.group('video_id')
 
-        # determine title and media streams from webpage
-        html = self._download_webpage(url, video_id)
-        title = re.search(self._TITLE, html).group('title')
-        streams = [mo.groupdict() for mo in re.finditer(self._MEDIA_STREAM, html)]
+        webpage = self._download_webpage(url, video_id)
+
+        title = self._html_search_regex(
+            r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>', webpage, 'title')
+        description = self._html_search_meta(
+            'dcterms.abstract', webpage, 'description')
+        thumbnail = self._og_search_thumbnail(webpage)
+
+        streams = [
+            mo.groupdict()
+            for mo in re.finditer(
+                r'mediaCollection\.addMediaStream\((?P<media_type>\d+), (?P<quality>\d+), "(?P<rtmp_url>[^"]*)", "(?P<video_url>[^"]*)", "[^"]*"\)', webpage)]
         if not streams:
-            assert '"fsk"' in html
-            raise ExtractorError(u'This video is only available after 8:00 pm')
-
-        # choose default media type and highest quality for now
-        stream = max([s for s in streams if int(s["media_type"]) == 0],
-                     key=lambda s: int(s["quality"]))
-
-        # there's two possibilities: RTMP stream or HTTP download
-        info = {'id': video_id, 'title': title, 'ext': 'mp4'}
-        if stream['rtmp_url']:
-            self.to_screen(u'RTMP download detected')
-            assert stream['video_url'].startswith('mp4:')
-            info["url"] = stream["rtmp_url"]
-            info["play_path"] = stream['video_url']
-        else:
-            assert stream["video_url"].endswith('.mp4')
-            info["url"] = stream["video_url"]
-        return [info]
+            if '"fsk"' in webpage:
+                raise ExtractorError('This video is only available after 20:00')
+
+        formats = []
+        for s in streams:
+            format = {
+                'quality': int(s['quality']),
+            }
+            if s.get('rtmp_url'):
+                format['protocol'] = 'rtmp'
+                format['url'] = s['rtmp_url']
+                format['playpath'] = s['video_url']
+            else:
+                format['url'] = s['video_url']
+
+            quality_name = self._search_regex(
+                r'[,.]([a-zA-Z0-9_-]+),?\.mp4', format['url'],
+                'quality name', default='NA')
+            format['format_id'] = '%s-%s-%s-%s' % (
+                determine_ext(format['url']), quality_name, s['media_type'],
+                s['quality'])
+
+            formats.append(format)
+
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': description,
+            'formats': formats,
+            'thumbnail': thumbnail,
+        }