Fix extraction from youtube.

[youtubedl] / youtube_dl / extractor / prosiebensat1.py

diff --git a/youtube_dl/extractor/prosiebensat1.py b/youtube_dl/extractor/prosiebensat1.py
index 7cc07a2ad5b88c51aa9f5d339839fd743727e17e..d0955d07903cca42806b4c53ca72eb87157a09fc 100644 (file)
--- a/youtube_dl/extractor/prosiebensat1.py
+++ b/youtube_dl/extractor/prosiebensat1.py
@@ -85,6 +85,9 @@ class ProSiebenSat1BaseIE(InfoExtractor):
                     formats.extend(self._extract_m3u8_formats(
                         source_url, clip_id, 'mp4', 'm3u8_native',
                         m3u8_id='hls', fatal=False))
+                elif mimetype == 'application/dash+xml':
+                    formats.extend(self._extract_mpd_formats(
+                        source_url, clip_id, mpd_id='dash', fatal=False))
                 else:
                     tbr = fix_bitrate(source['bitrate'])
                     if protocol in ('rtmp', 'rtmpe'):
@@ -126,10 +129,11 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
                     https?://
                         (?:www\.)?
                         (?:
+                            (?:beta\.)?
                             (?:
                                 prosieben(?:maxx)?|sixx|sat1(?:gold)?|kabeleins(?:doku)?|the-voice-of-germany|7tv|advopedia
                             )\.(?:de|at|ch)|
-                            ran\.de|fem\.com|advopedia\.de
+                            ran\.de|fem\.com|advopedia\.de|galileo\.tv/video
                         )
                         /(?P<id>.+)
                     '''
@@ -144,16 +148,12 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
             'url': 'http://www.prosieben.de/tv/circus-halligalli/videos/218-staffel-2-episode-18-jahresrueckblick-ganze-folge',
             'info_dict': {
                 'id': '2104602',
-                'ext': 'flv',
+                'ext': 'mp4',
                 'title': 'Episode 18 - Staffel 2',
                 'description': 'md5:8733c81b702ea472e069bc48bb658fc1',
                 'upload_date': '20131231',
                 'duration': 5845.04,
             },
-            'params': {
-                # rtmp download
-                'skip_download': True,
-            },
         },
         {
             'url': 'http://www.prosieben.de/videokatalog/Gesellschaft/Leben/Trends/video-Lady-Umstyling-f%C3%BCr-Audrina-Rebekka-Audrina-Fergen-billig-aussehen-Battal-Modica-700544.html',
@@ -255,7 +255,7 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
             'url': 'http://www.the-voice-of-germany.de/video/31-andreas-kuemmert-rocket-man-clip',
             'info_dict': {
                 'id': '2572814',
-                'ext': 'flv',
+                'ext': 'mp4',
                 'title': 'Andreas Kümmert: Rocket Man',
                 'description': 'md5:6ddb02b0781c6adf778afea606652e38',
                 'upload_date': '20131017',
@@ -269,7 +269,7 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
             'url': 'http://www.fem.com/wellness/videos/wellness-video-clip-kurztripps-zum-valentinstag.html',
             'info_dict': {
                 'id': '2156342',
-                'ext': 'flv',
+                'ext': 'mp4',
                 'title': 'Kurztrips zum Valentinstag',
                 'description': 'Romantischer Kurztrip zum Valentinstag? Nina Heinemann verrät, was sich hier wirklich lohnt.',
                 'duration': 307.24,
@@ -286,12 +286,13 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
                 'description': 'md5:63b8963e71f481782aeea877658dec84',
             },
             'playlist_count': 2,
+            'skip': 'This video is unavailable',
         },
         {
             'url': 'http://www.7tv.de/circus-halligalli/615-best-of-circus-halligalli-ganze-folge',
             'info_dict': {
                 'id': '4187506',
-                'ext': 'flv',
+                'ext': 'mp4',
                 'title': 'Best of Circus HalliGalli',
                 'description': 'md5:8849752efd90b9772c9db6fdf87fb9e9',
                 'upload_date': '20151229',
@@ -300,6 +301,21 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
                 'skip_download': True,
             },
         },
+        {
+            # title in <h2 class="subtitle">
+            'url': 'http://www.prosieben.de/stars/oscar-award/videos/jetzt-erst-enthuellt-das-geheimnis-von-emma-stones-oscar-robe-clip',
+            'info_dict': {
+                'id': '4895826',
+                'ext': 'mp4',
+                'title': 'Jetzt erst enthüllt: Das Geheimnis von Emma Stones Oscar-Robe',
+                'description': 'md5:e5ace2bc43fadf7b63adc6187e9450b9',
+                'upload_date': '20170302',
+            },
+            'params': {
+                'skip_download': True,
+            },
+            'skip': 'geo restricted to Germany',
+        },
         {
             # geo restricted to Germany
             'url': 'http://www.kabeleinsdoku.de/tv/mayday-alarm-im-cockpit/video/102-notlandung-im-hudson-river-ganze-folge',
@@ -310,6 +326,11 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
             'url': 'http://www.sat1gold.de/tv/edel-starck/video/11-staffel-1-episode-1-partner-wider-willen-ganze-folge',
             'only_matching': True,
         },
+        {
+            # geo restricted to Germany
+            'url': 'https://www.galileo.tv/video/diese-emojis-werden-oft-missverstanden',
+            'only_matching': True,
+        },
         {
             'url': 'http://www.sat1gold.de/tv/edel-starck/playlist/die-gesamte-1-staffel',
             'only_matching': True,
@@ -327,8 +348,10 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
         r'"clip_id"\s*:\s+"(\d+)"',
         r'clipid: "(\d+)"',
         r'clip[iI]d=(\d+)',
-        r'clip[iI]d\s*=\s*["\'](\d+)',
+        r'clip[iI][dD]\s*=\s*["\'](\d+)',
         r"'itemImageUrl'\s*:\s*'/dynamic/thumbnails/full/\d+/(\d+)",
+        r'proMamsId&quot;\s*:\s*&quot;(\d+)',
+        r'proMamsId"\s*:\s*"(\d+)',
     ]
     _TITLE_REGEXES = [
         r'<h2 class="subtitle" itemprop="name">\s*(.+?)</h2>',
@@ -338,6 +361,7 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
         r'<header class="module_header">\s*<h2>([^<]+)</h2>\s*</header>',
         r'<h2 class="video-title" itemprop="name">\s*(.+?)</h2>',
         r'<div[^>]+id="veeseoTitle"[^>]*>(.+?)</div>',
+        r'<h2[^>]+class="subtitle"[^>]*>([^<]+)</h2>',
     ]
     _DESCRIPTION_REGEXES = [
         r'<p itemprop="description">\s*(.+?)</p>',
@@ -369,10 +393,14 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
     def _extract_clip(self, url, webpage):
         clip_id = self._html_search_regex(
             self._CLIPID_REGEXES, webpage, 'clip id')
-        title = self._html_search_regex(self._TITLE_REGEXES, webpage, 'title')
+        title = self._html_search_regex(
+            self._TITLE_REGEXES, webpage, 'title',
+            default=None) or self._og_search_title(webpage)
         info = self._extract_video_info(url, clip_id)
         description = self._html_search_regex(
-            self._DESCRIPTION_REGEXES, webpage, 'description', fatal=False)
+            self._DESCRIPTION_REGEXES, webpage, 'description', default=None)
+        if description is None:
+            description = self._og_search_description(webpage)
         thumbnail = self._og_search_thumbnail(webpage)
         upload_date = unified_strdate(self._html_search_regex(
             self._UPLOAD_DATE_REGEXES, webpage, 'upload date', default=None))
@@ -391,7 +419,7 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
             self._PLAYLIST_ID_REGEXES, webpage, 'playlist id')
         playlist = self._parse_json(
             self._search_regex(
-                'var\s+contentResources\s*=\s*(\[.+?\]);\s*</script',
+                r'var\s+contentResources\s*=\s*(\[.+?\]);\s*</script',
                 webpage, 'playlist'),
             playlist_id)
         entries = []
@@ -422,3 +450,6 @@ class ProSiebenSat1IE(ProSiebenSat1BaseIE):
             return self._extract_clip(url, webpage)
         elif page_type == 'playlist':
             return self._extract_playlist(url, webpage)
+        else:
+            raise ExtractorError(
+                'Unsupported page type %s' % page_type, expected=True)