Start new release.

[youtubedl] / youtube_dl / extractor / ard.py

diff --git a/youtube_dl/extractor/ard.py b/youtube_dl/extractor/ard.py
index 957bdefcbec7666473530815df9e3eb5dc88e096..8adae46449232fe487ec6796b5b60345f144d310 100644 (file)
--- a/youtube_dl/extractor/ard.py
+++ b/youtube_dl/extractor/ard.py
@@ -4,63 +4,197 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
 import re
 
 from .common import InfoExtractor

+from .generic import GenericIE

 from ..utils import (
     determine_ext,
     ExtractorError,
 from ..utils import (
     determine_ext,
     ExtractorError,

+    int_or_none,
+    parse_duration,

     qualities,
     qualities,

-    compat_urllib_parse_urlparse,
-    compat_urllib_parse,
+    str_or_none,
+    try_get,
+    unified_strdate,
+    unified_timestamp,
+    update_url_query,
+    url_or_none,
+    xpath_text,

 )
 )

+from ..compat import compat_etree_fromstring

 
 
 
 

-class ARDIE(InfoExtractor):
-    _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
+class ARDMediathekIE(InfoExtractor):
+    IE_NAME = 'ARD:mediathek'
+    _VALID_URL = r'^https?://(?:(?:(?:www|classic)\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'

 
     _TESTS = [{
 
     _TESTS = [{

-        'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
-        'file': '22429276.mp4',
-        'md5': '469751912f1de0816a9fc9df8336476c',
-        'info_dict': {
-            'title': 'Vertrauen ist gut, Spionieren ist besser - Geht so deutsch-amerikanische Freundschaft?',
-            'description': 'Das Erste Mediathek [ARD]: Vertrauen ist gut, Spionieren ist besser - Geht so deutsch-amerikanische Freundschaft?, Anne Will, Über die Spionage-Affäre diskutieren Clemens Binninger, Katrin Göring-Eckardt, Georg Mascolo, Andrew B. Denison und Constanze Kurz.. Das Video zur Sendung Anne Will am Mittwoch, 16.07.2014',
-        },
-        'skip': 'Blocked outside of Germany',
-    }, {
-        'url': 'http://www.ardmediathek.de/tv/Tatort/Das-Wunder-von-Wolbeck-Video-tgl-ab-20/Das-Erste/Video?documentId=22490580&bcastId=602916',
+        # available till 26.07.2022
+        'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',

         'info_dict': {
         'info_dict': {

-            'id': '22490580',
+            'id': '44726822',

             'ext': 'mp4',
             'ext': 'mp4',

-            'title': 'Das Wunder von Wolbeck (Video tgl. ab 20 Uhr)',
-            'description': 'Auf einem restaurierten Hof bei Wolbeck wird der Heilpraktiker Raffael Lembeck eines morgens von seiner Frau Stella tot aufgefunden. Das Opfer war offensichtlich in seiner Praxis zu Fall gekommen und ist dann verblutet, erklärt Prof. Boerne am Tatort.',
+            'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
+            'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
+            'duration': 1740,

         },
         },

-        'skip': 'Blocked outside of Germany',
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        }
+    }, {
+        'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
+        'only_matching': True,
+    }, {
+        # audio
+        'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
+        'only_matching': True,
+    }, {
+        'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
+        'only_matching': True,
+    }, {
+        # audio
+        'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
+        'only_matching': True,
+    }, {
+        'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
+        'only_matching': True,

     }]
 
     }]
 

+    @classmethod
+    def suitable(cls, url):
+        return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)
+
+    def _extract_media_info(self, media_info_url, webpage, video_id):
+        media_info = self._download_json(
+            media_info_url, video_id, 'Downloading media JSON')
+
+        formats = self._extract_formats(media_info, video_id)
+
+        if not formats:
+            if '"fsk"' in webpage:
+                raise ExtractorError(
+                    'This video is only available after 20:00', expected=True)
+            elif media_info.get('_geoblocked'):
+                raise ExtractorError('This video is not available due to geo restriction', expected=True)
+
+        self._sort_formats(formats)
+
+        duration = int_or_none(media_info.get('_duration'))
+        thumbnail = media_info.get('_previewImage')
+        is_live = media_info.get('_isLive') is True
+
+        subtitles = {}
+        subtitle_url = media_info.get('_subtitleUrl')
+        if subtitle_url:
+            subtitles['de'] = [{
+                'ext': 'ttml',
+                'url': subtitle_url,
+            }]
+
+        return {
+            'id': video_id,
+            'duration': duration,
+            'thumbnail': thumbnail,
+            'is_live': is_live,
+            'formats': formats,
+            'subtitles': subtitles,
+        }
+
+    def _extract_formats(self, media_info, video_id):
+        type_ = media_info.get('_type')
+        media_array = media_info.get('_mediaArray', [])
+        formats = []
+        for num, media in enumerate(media_array):
+            for stream in media.get('_mediaStreamArray', []):
+                stream_urls = stream.get('_stream')
+                if not stream_urls:
+                    continue
+                if not isinstance(stream_urls, list):
+                    stream_urls = [stream_urls]
+                quality = stream.get('_quality')
+                server = stream.get('_server')
+                for stream_url in stream_urls:
+                    if not url_or_none(stream_url):
+                        continue
+                    ext = determine_ext(stream_url)
+                    if quality != 'auto' and ext in ('f4m', 'm3u8'):
+                        continue
+                    if ext == 'f4m':
+                        formats.extend(self._extract_f4m_formats(
+                            update_url_query(stream_url, {
+                                'hdcore': '3.1.1',
+                                'plugin': 'aasp-3.1.1.69.124'
+                            }),
+                            video_id, f4m_id='hds', fatal=False))
+                    elif ext == 'm3u8':
+                        formats.extend(self._extract_m3u8_formats(
+                            stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
+                    else:
+                        if server and server.startswith('rtmp'):
+                            f = {
+                                'url': server,
+                                'play_path': stream_url,
+                                'format_id': 'a%s-rtmp-%s' % (num, quality),
+                            }
+                        else:
+                            f = {
+                                'url': stream_url,
+                                'format_id': 'a%s-%s-%s' % (num, ext, quality)
+                            }
+                        m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
+                        if m:
+                            f.update({
+                                'width': int(m.group('width')),
+                                'height': int(m.group('height')),
+                            })
+                        if type_ == 'audio':
+                            f['vcodec'] = 'none'
+                        formats.append(f)
+        return formats
+

     def _real_extract(self, url):
         # determine video id from url
         m = re.match(self._VALID_URL, url)
 
     def _real_extract(self, url):
         # determine video id from url
         m = re.match(self._VALID_URL, url)
 

+        document_id = None
+

         numid = re.search(r'documentId=([0-9]+)', url)
         if numid:
         numid = re.search(r'documentId=([0-9]+)', url)
         if numid:

-            video_id = numid.group(1)
+            document_id = video_id = numid.group(1)

         else:
             video_id = m.group('video_id')
 
         else:
             video_id = m.group('video_id')
 

-        urlp = compat_urllib_parse_urlparse(url)
-        url = urlp._replace(path=compat_urllib_parse.quote(urlp.path.encode('utf-8'))).geturl()
-

         webpage = self._download_webpage(url, video_id)
 
         webpage = self._download_webpage(url, video_id)
 

+        ERRORS = (
+            ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
+            ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
+             'Video %s is no longer available'),
+        )
+
+        for pattern, message in ERRORS:
+            if pattern in webpage:
+                raise ExtractorError(message % video_id, expected=True)
+
+        if re.search(r'[\?&]rss($|[=&])', url):
+            doc = compat_etree_fromstring(webpage.encode('utf-8'))
+            if doc.tag == 'rss':
+                return GenericIE()._extract_rss(url, video_id, doc)
+

         title = self._html_search_regex(
             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
         title = self._html_search_regex(
             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',

-             r'<meta name="dcterms.title" content="(.*?)"/>',
-             r'<h4 class="headline">(.*?)</h4>'],
+             r'<meta name="dcterms\.title" content="(.*?)"/>',
+             r'<h4 class="headline">(.*?)</h4>',
+             r'<title[^>]*>(.*?)</title>'],

             webpage, 'title')
         description = self._html_search_meta(
             'dcterms.abstract', webpage, 'description', default=None)
         if description is None:
             description = self._html_search_meta(
             webpage, 'title')
         description = self._html_search_meta(
             'dcterms.abstract', webpage, 'description', default=None)
         if description is None:
             description = self._html_search_meta(

-                'description', webpage, 'meta description')
+                'description', webpage, 'meta description', default=None)
+        if description is None:
+            description = self._html_search_regex(
+                r'<p\s+class="teasertext">(.+?)</p>',
+                webpage, 'teaser text', default=None)

 
         # Thumbnail is sometimes not present.
         # It is in the mobile version, but that seems to use a different URL
 
         # Thumbnail is sometimes not present.
         # It is in the mobile version, but that seems to use a different URL


@@ -85,43 +219,182 @@ class ARDIE(InfoExtractor):
                     'format_id': fid,
                     'url': furl,
                 })
                     'format_id': fid,
                     'url': furl,
                 })

+            self._sort_formats(formats)
+            info = {
+                'formats': formats,
+            }

         else:  # request JSON file
         else:  # request JSON file

-            media_info = self._download_json(
-                'http://www.ardmediathek.de/play/media/%s' % video_id, video_id)
-            # The second element of the _mediaArray contains the standard http urls
-            streams = media_info['_mediaArray'][1]['_mediaStreamArray']
-            if not streams:
-                if '"fsk"' in webpage:
-                    raise ExtractorError('This video is only available after 20:00')
+            if not document_id:
+                video_id = self._search_regex(
+                    r'/play/(?:config|media)/(\d+)', webpage, 'media id')
+            info = self._extract_media_info(
+                'http://www.ardmediathek.de/play/media/%s' % video_id,
+                webpage, video_id)

 
 

-            formats = []
-            for s in streams:
-                if type(s['_stream']) == list:
-                    for index, url in enumerate(s['_stream'][::-1]):
-                        quality = s['_quality'] + index
-                        formats.append({
-                            'quality': quality,
-                            'url': url,
-                            'format_id': '%s-%s' % (determine_ext(url), quality)
-                        })
-                    continue
+        info.update({
+            'id': video_id,
+            'title': self._live_title(title) if info.get('is_live') else title,
+            'description': description,
+            'thumbnail': thumbnail,
+        })

 
 

-                format = {
-                    'quality': s['_quality'],
-                    'url': s['_stream'],
-                }
+        return info

 
 

-                format['format_id'] = '%s-%s' % (
-                    determine_ext(format['url']), format['quality'])

 
 

-                formats.append(format)
+class ARDIE(InfoExtractor):
+    _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
+    _TESTS = [{
+        # available till 14.02.2019
+        'url': 'http://www.daserste.de/information/talk/maischberger/videos/das-groko-drama-zerlegen-sich-die-volksparteien-video-102.html',
+        'md5': '8e4ec85f31be7c7fc08a26cdbc5a1f49',
+        'info_dict': {
+            'display_id': 'das-groko-drama-zerlegen-sich-die-volksparteien-video',
+            'id': '102',
+            'ext': 'mp4',
+            'duration': 4435.0,
+            'title': 'Das GroKo-Drama: Zerlegen sich die Volksparteien?',
+            'upload_date': '20180214',
+            'thumbnail': r're:^https?://.*\.jpg$',
+        },
+    }, {
+        'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
+        'only_matching': True,
+    }]

 
 

+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        display_id = mobj.group('display_id')
+
+        player_url = mobj.group('mainurl') + '~playerXml.xml'
+        doc = self._download_xml(player_url, display_id)
+        video_node = doc.find('./video')
+        upload_date = unified_strdate(xpath_text(
+            video_node, './broadcastDate'))
+        thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
+
+        formats = []
+        for a in video_node.findall('.//asset'):
+            f = {
+                'format_id': a.attrib['type'],
+                'width': int_or_none(a.find('./frameWidth').text),
+                'height': int_or_none(a.find('./frameHeight').text),
+                'vbr': int_or_none(a.find('./bitrateVideo').text),
+                'abr': int_or_none(a.find('./bitrateAudio').text),
+                'vcodec': a.find('./codecVideo').text,
+                'tbr': int_or_none(a.find('./totalBitrate').text),
+            }
+            if a.find('./serverPrefix').text:
+                f['url'] = a.find('./serverPrefix').text
+                f['playpath'] = a.find('./fileName').text
+            else:
+                f['url'] = a.find('./fileName').text
+            formats.append(f)

         self._sort_formats(formats)
 
         return {
         self._sort_formats(formats)
 
         return {

-            'id': video_id,
-            'title': title,
-            'description': description,
+            'id': mobj.group('id'),

             'formats': formats,
             'formats': formats,

+            'display_id': display_id,
+            'title': video_node.find('./title').text,
+            'duration': parse_duration(video_node.find('./duration').text),
+            'upload_date': upload_date,

             'thumbnail': thumbnail,
         }
             'thumbnail': thumbnail,
         }

+
+
+class ARDBetaMediathekIE(InfoExtractor):
+    _VALID_URL = r'https://(?:beta|www)\.ardmediathek\.de/[^/]+/(?:player|live)/(?P<video_id>[a-zA-Z0-9]+)(?:/(?P<display_id>[^/?#]+))?'
+    _TESTS = [{
+        'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
+        'md5': '2d02d996156ea3c397cfc5036b5d7f8f',
+        'info_dict': {
+            'display_id': 'die-robuste-roswita',
+            'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
+            'title': 'Tatort: Die robuste Roswita',
+            'description': r're:^Der Mord.*trüber ist als die Ilm.',
+            'duration': 5316,
+            'thumbnail': 'https://img.ardmediathek.de/standard/00/55/43/59/34/-1774185891/16x9/960?mandant=ard',
+            'upload_date': '20180826',
+            'ext': 'mp4',
+        },
+    }, {
+        'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3N3ci5kZS9hZXgvbzEwNzE5MTU/',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.ardmediathek.de/swr/live/Y3JpZDovL3N3ci5kZS8xMzQ4MTA0Mg',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('video_id')
+        display_id = mobj.group('display_id') or video_id
+
+        webpage = self._download_webpage(url, display_id)
+        data_json = self._search_regex(r'window\.__APOLLO_STATE__\s*=\s*(\{.*);\n', webpage, 'json')
+        data = self._parse_json(data_json, display_id)
+
+        res = {
+            'id': video_id,
+            'display_id': display_id,
+        }
+        formats = []
+        subtitles = {}
+        geoblocked = False
+        for widget in data.values():
+            if widget.get('_geoblocked') is True:
+                geoblocked = True
+            if '_duration' in widget:
+                res['duration'] = int_or_none(widget['_duration'])
+            if 'clipTitle' in widget:
+                res['title'] = widget['clipTitle']
+            if '_previewImage' in widget:
+                res['thumbnail'] = widget['_previewImage']
+            if 'broadcastedOn' in widget:
+                res['timestamp'] = unified_timestamp(widget['broadcastedOn'])
+            if 'synopsis' in widget:
+                res['description'] = widget['synopsis']
+            subtitle_url = url_or_none(widget.get('_subtitleUrl'))
+            if subtitle_url:
+                subtitles.setdefault('de', []).append({
+                    'ext': 'ttml',
+                    'url': subtitle_url,
+                })
+            if '_quality' in widget:
+                format_url = url_or_none(try_get(
+                    widget, lambda x: x['_stream']['json'][0]))
+                if not format_url:
+                    continue
+                ext = determine_ext(format_url)
+                if ext == 'f4m':
+                    formats.extend(self._extract_f4m_formats(
+                        format_url + '?hdcore=3.11.0',
+                        video_id, f4m_id='hds', fatal=False))
+                elif ext == 'm3u8':
+                    formats.extend(self._extract_m3u8_formats(
+                        format_url, video_id, 'mp4', m3u8_id='hls',
+                        fatal=False))
+                else:
+                    # HTTP formats are not available when geoblocked is True,
+                    # other formats are fine though
+                    if geoblocked:
+                        continue
+                    quality = str_or_none(widget.get('_quality'))
+                    formats.append({
+                        'format_id': ('http-' + quality) if quality else 'http',
+                        'url': format_url,
+                        'preference': 10,  # Plain HTTP, that's nice
+                    })
+
+        if not formats and geoblocked:
+            self.raise_geo_restricted(
+                msg='This video is not available due to geoblocking',
+                countries=['DE'])
+
+        self._sort_formats(formats)
+        res.update({
+            'subtitles': subtitles,
+            'formats': formats,
+        })
+
+        return res