[tagesschau] Add support for playlists

youtube_dl/extractor/tagesschau.py

@@ -8,11 +8,11 @@
 
 
 class TagesschauIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?tagesschau\.de/multimedia/(?:[^/]+/)*?[^/#?]+?(?P<id>-?[0-9]+)(?:~_?[^/#?]+?)?\.html'
+    _VALID_URL = r'https?://(?:www\.)?tagesschau\.de/[^/]+/(?:[^/]+/)*?[^/#?]+?(?P<id>-?[0-9]+)(?:~_?[^/#?]+?)?\.html'
 
     _TESTS = [{
         'url': 'http://www.tagesschau.de/multimedia/video/video-102143.html',
-        'md5': '917a228bc7df7850783bc47979673a09',
+        'md5': 'f7c27a0eff3bfe8c7727e65f8fe1b1e6',
         'info_dict': {
             'id': '102143',
             'ext': 'mp4',
@@ -40,6 +40,13 @@ class TagesschauIE(InfoExtractor):
             'description': 'Flüchtlingsdebatte: Hitzig, aber wenig hilfreich',
             'thumbnail': 're:^https?:.*\.jpg$',
         },
+    }, {
+        'url': 'http://www.tagesschau.de/inland/afd-parteitag-135.html',
+        'info_dict': {
+            'id': '135',
+            'title': 'Möchtegern-Underdog mit Machtanspruch',
+        },
+        'playlist_count': 2,
     }, {
         'url': 'http://www.tagesschau.de/multimedia/sendung/tsg-3771.html',
         'only_matching': True,
@@ -75,43 +82,7 @@ class TagesschauIE(InfoExtractor):
         'xxl': {'quality': 5},
     }
 
-    def _real_extract(self, url):
+    def _extract_formats(self, download_text):
-        video_id = self._match_id(url)
-        display_id = video_id.lstrip('-')
-        webpage = self._download_webpage(url, display_id)
-
-        player_url = self._html_search_meta(
-            'twitter:player', webpage, 'player URL', default=None)
-        if player_url:
-            playerpage = self._download_webpage(
-                player_url, display_id, 'Downloading player page')
-
-            formats = []
-            for media in re.finditer(
-                    r'''(?x)
-                        (?P<q_url>["\'])(?P<url>http://media.+?)(?P=q_url)
-                        ,\s*type:(?P<q_type>["\'])(?P<type>video|audio)/(?P<ext>.+?)(?P=q_type)
-                        (?:,\s*quality:(?P<q_quality>["\'])(?P<quality>.+?)(?P=q_quality))?
-                    ''', playerpage):
-                url = media.group('url')
-                type_ = media.group('type')
-                ext = media.group('ext')
-                res = media.group('quality')
-                f = {
-                    'format_id': '%s_%s' % (res, ext) if res else ext,
-                    'url': url,
-                    'ext': ext,
-                    'vcodec': 'none' if type_ == 'audio' else None,
-                }
-                f.update(self._FORMATS.get(res, {}))
-                formats.append(f)
-            thumbnail = self._og_search_thumbnail(playerpage)
-            title = self._og_search_title(webpage).strip()
-            description = self._og_search_description(webpage).strip()
-        else:
-            download_text = self._search_regex(
-                r'(?s)<p>Wir bieten dieses Video in folgenden Formaten zum Download an:</p>\s*<div class="controls">(.*?)</div>\s*<p>',
-                webpage, 'download links')
         links = re.finditer(
             r'<div class="button" title="(?P<title>[^"]*)"><a href="(?P<url>[^"]+)">(?P<name>.+?)</a></div>',
             download_text)
@@ -143,12 +114,67 @@ def _real_extract(self, url):
                     'filesize_approx': parse_filesize(m.group('filesize_approx')),
                 })
             formats.append(format)
+        self._sort_formats(formats)
+        return formats
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        display_id = video_id.lstrip('-')
+        webpage = self._download_webpage(url, display_id)
+
+        player_url = self._html_search_meta(
+            'twitter:player', webpage, 'player URL', default=None)
+        if player_url:
+            playerpage = self._download_webpage(
+                player_url, display_id, 'Downloading player page')
+
+            formats = []
+            for media in re.finditer(
+                    r'''(?x)
+                        (?P<q_url>["\'])(?P<url>http://media.+?)(?P=q_url)
+                        ,\s*type:(?P<q_type>["\'])(?P<type>video|audio)/(?P<ext>.+?)(?P=q_type)
+                        (?:,\s*quality:(?P<q_quality>["\'])(?P<quality>.+?)(?P=q_quality))?
+                    ''', playerpage):
+                url = media.group('url')
+                webpage_type = media.group('type')
+                ext = media.group('ext')
+                res = media.group('quality')
+                f = {
+                    'format_id': '%s_%s' % (res, ext) if res else ext,
+                    'url': url,
+                    'ext': ext,
+                    'vcodec': 'none' if webpage_type == 'audio' else None,
+                }
+                f.update(self._FORMATS.get(res, {}))
+                formats.append(f)
+            thumbnail = self._og_search_thumbnail(playerpage)
+            title = self._og_search_title(webpage).strip()
+            description = self._og_search_description(webpage).strip()
+        else:
+            title = self._html_search_regex(
+                r'<span class="headline".*?>(.*?)</span>', webpage, 'title')
+
+            DOWNLOAD_REGEX = r'(?s)<p>Wir bieten dieses Video in folgenden Formaten zum Download an:</p>\s*<div class="controls">(.*?)</div>\s*<p>'
+
+            webpage_type = self._og_search_property('type', webpage, default=None)
+            if webpage_type == 'website':  # Article
+                entries = []
+                for num, (entry_title, download_text) in enumerate(re.findall(
+                        r'(?s)<p[^>]+class="infotext"[^>]*>.*?<strong>(.+?)</strong>.*?</p>.*?%s' % DOWNLOAD_REGEX,
+                        webpage)):
+                    entries.append({
+                        'id': display_id,
+                        'title': '%s-%d' % (entry_title, num),
+                        'formats': self._extract_formats(download_text),
+                    })
+                return self.playlist_result(entries, display_id, title)
+            else:  # Assume single video
+                download_text = self._search_regex(DOWNLOAD_REGEX, webpage, 'download links')
+                formats = self._extract_formats(download_text)
                 thumbnail = self._og_search_thumbnail(webpage)
                 description = self._html_search_regex(
                     r'(?s)<p class="teasertext">(.*?)</p>',
                     webpage, 'description', default=None)
-            title = self._html_search_regex(
-                r'<span class="headline".*?>(.*?)</span>', webpage, 'title')
 
         self._sort_formats(formats)