[youtube] Fix extraction.

[youtube-dl] / youtube_dl / extractor / dbtv.py

diff --git a/youtube_dl/extractor/dbtv.py b/youtube_dl/extractor/dbtv.py
index cf76dbf0533652de1e22fbfabebb904238c3aeec..aaedf2e3d36084d90699277a52aebb6c3a71a335 100644 (file)
--- a/youtube_dl/extractor/dbtv.py
+++ b/youtube_dl/extractor/dbtv.py
@@ -5,72 +5,53 @@ import re
 
 from .common import InfoExtractor
 
-from ..utils import (
-  ExtractorError
-)
 
 class DBTVIE(InfoExtractor):
-  _VALID_URL = r'http://dbtv.no/(?P<id>[0-9]+)/?(?P<slug>.*)$'
-  _TEST = {
-    'url': 'http://dbtv.no/3649835190001#Skulle_teste_ut_fornøyelsespark,_men_kollegaen_var_bare_opptatt_av_bikinikroppen',
-    'md5': 'b89953ed25dacb6edb3ef6c6f430f8bc',
-    'info_dict': {
-      'id': '3649835190001',
-      'ext': 'mp4',
-      'title': 'Skulle teste ut fornøyelsespark, men kollegaen var bare opptatt av bikinikroppen',
-      'description': 'md5:d681bf2bb7dd3503892cedb9c2d0e6f2',
-      'thumbnail': 'http://gfx.dbtv.no/thumbs/still/33100.jpg',
-      'timestamp': 1404039863,
-      'upload_date': '20140629',
-      'duration': 69544,
-    }
-  }
-
-  def _real_extract(self, url):
-    mobj = re.match(self._VALID_URL, url)
-    video_id = mobj.group('id')
-
-    # Download JSON file containing video info.
-    data = self._download_json('http://api.dbtv.no/discovery/%s' % video_id, video_id, 'Downloading media JSON')
-    # We only want the first video in the JSON API file.
-    video = data['playlist'][0]
-
-    # Check for full HD video, else use the standard video URL
-    for i in range(0, len(video['renditions'])):
-      if int(video['renditions'][i]['width']) == 1280:
-        video_url = video['renditions'][i]['URL']
-        break
-      else:
-        video_url = video['URL']
-
-    # Add access token to image or it will fail.
-    thumbnail = video['splash']
-
-    # Duration int.
-    duration = int(video['length'])
-
-    # Timestamp is given in milliseconds.
-    timestamp = float(str(video['publishedAt'])[0:-3])
-
-    formats = []
-
-    # Video URL.
-    if video['URL'] is not None:
-      formats.append({
-        'url': video_url,
-        'format_id': 'mp4',
-        'ext': 'mp4'
-      })
-    else:
-      raise ExtractorError('No download URL found for video: %s.' % video_id, expected=True)
-
-    return {
-      'id': video_id,
-      'title': video['title'],
-      'description': video['desc'],
-      'thumbnail': thumbnail,
-      'timestamp': timestamp,
-      'duration': duration,
-      'view_count': video['views'],
-      'formats': formats,
-    }
+    _VALID_URL = r'https?://(?:www\.)?dagbladet\.no/video/(?:(?:embed|(?P<display_id>[^/]+))/)?(?P<id>[0-9A-Za-z_-]{11}|[a-zA-Z0-9]{8})'
+    _TESTS = [{
+        'url': 'https://www.dagbladet.no/video/PynxJnNWChE/',
+        'md5': 'b8f850ba1860adbda668d367f9b77699',
+        'info_dict': {
+            'id': 'PynxJnNWChE',
+            'ext': 'mp4',
+            'title': 'Skulle teste ut fornøyelsespark, men kollegaen var bare opptatt av bikinikroppen',
+            'description': 'md5:49cc8370e7d66e8a2ef15c3b4631fd3f',
+            'thumbnail': r're:https?://.*\.jpg',
+            'upload_date': '20160916',
+            'duration': 69,
+            'uploader_id': 'UCk5pvsyZJoYJBd7_oFPTlRQ',
+            'uploader': 'Dagbladet',
+        },
+        'add_ie': ['Youtube']
+    }, {
+        'url': 'https://www.dagbladet.no/video/embed/xlGmyIeN9Jo/?autoplay=false',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.dagbladet.no/video/truer-iran-bor-passe-dere/PalfB2Cw',
+        'only_matching': True,
+    }]
+
+    @staticmethod
+    def _extract_urls(webpage):
+        return [url for _, url in re.findall(
+            r'<iframe[^>]+src=(["\'])((?:https?:)?//(?:www\.)?dagbladet\.no/video/embed/(?:[0-9A-Za-z_-]{11}|[a-zA-Z0-9]{8}).*?)\1',
+            webpage)]
+
+    def _real_extract(self, url):
+        display_id, video_id = re.match(self._VALID_URL, url).groups()
+        info = {
+            '_type': 'url_transparent',
+            'id': video_id,
+            'display_id': display_id,
+        }
+        if len(video_id) == 11:
+            info.update({
+                'url': video_id,
+                'ie_key': 'Youtube',
+            })
+        else:
+            info.update({
+                'url': 'jwplatform:' + video_id,
+                'ie_key': 'JWPlatform',
+            })
+        return info