Prepare for upload.

[youtubedl] / youtube_dl / extractor / ard.py

diff --git a/youtube_dl/extractor/ard.py b/youtube_dl/extractor/ard.py
index 957bdefcbec7666473530815df9e3eb5dc88e096..2d5599456688eba9756e28c2ffe9dbae48decb2c 100644 (file)
--- a/youtube_dl/extractor/ard.py
+++ b/youtube_dl/extractor/ard.py
@@ -4,38 +4,164 @@ 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,
     qualities,
 from ..utils import (
     determine_ext,
     ExtractorError,
     qualities,

-    compat_urllib_parse_urlparse,
-    compat_urllib_parse,
+    int_or_none,
+    parse_duration,
+    unified_strdate,
+    xpath_text,
+    update_url_query,

 )
 )

+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\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.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',
+        'url': 'http://www.ardmediathek.de/tv/Dokumentation-und-Reportage/Ich-liebe-das-Leben-trotzdem/rbb-Fernsehen/Video?documentId=29582122&bcastId=3822114',

         'info_dict': {
         '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',
+            'id': '29582122',
+            'ext': 'mp4',
+            'title': 'Ich liebe das Leben trotzdem',
+            'description': 'md5:45e4c225c72b27993314b31a84a5261c',
+            'duration': 4557,

         },
         },

-        'skip': 'Blocked outside of Germany',
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
+        'skip': 'HTTP Error 404: Not Found',

     }, {
     }, {

-        'url': 'http://www.ardmediathek.de/tv/Tatort/Das-Wunder-von-Wolbeck-Video-tgl-ab-20/Das-Erste/Video?documentId=22490580&bcastId=602916',
+        'url': 'http://www.ardmediathek.de/tv/Tatort/Tatort-Scheinwelten-H%C3%B6rfassung-Video/Das-Erste/Video?documentId=29522730&bcastId=602916',
+        'md5': 'f4d98b10759ac06c0072bbcd1f0b9e3e',

         'info_dict': {
         'info_dict': {

-            'id': '22490580',
+            'id': '29522730',

             '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': 'Tatort: Scheinwelten - Hörfassung (Video tgl. ab 20 Uhr)',
+            'description': 'md5:196392e79876d0ac94c94e8cdb2875f1',
+            'duration': 5252,
+        },
+        'skip': 'HTTP Error 404: Not Found',
+    }, {
+        # 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',
+        'md5': '219d94d8980b4f538c7fcb0865eb7f2c',
+        'info_dict': {
+            'id': '28488308',
+            'ext': 'mp3',
+            'title': 'Tod eines Fußballers',
+            'description': 'md5:f6e39f3461f0e1f54bfa48c8875c86ef',
+            'duration': 3240,
+        },
+        'skip': 'HTTP Error 404: Not Found',
+    }, {
+        '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',
+        'md5': '4e8f00631aac0395fee17368ac0e9867',
+        'info_dict': {
+            'id': '30796318',
+            'ext': 'mp3',
+            'title': 'Vor dem Fest',
+            'description': 'md5:c0c1c8048514deaed2a73b3a60eecacb',
+            'duration': 3287,

         },
         },

-        'skip': 'Blocked outside of Germany',
+        'skip': 'Video is no longer available',

     }]
 
     }]
 

+    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')
+
+        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,
+            '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:
+                    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),
+                            }
+                        elif stream_url.startswith('http'):
+                            f = {
+                                'url': stream_url,
+                                'format_id': 'a%s-%s-%s' % (num, ext, quality)
+                            }
+                        else:
+                            continue
+                        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)


@@ -46,11 +172,23 @@ class ARDIE(InfoExtractor):
         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>',
              r'<meta name="dcterms.title" content="(.*?)"/>',
         title = self._html_search_regex(
             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
              r'<meta name="dcterms.title" content="(.*?)"/>',


@@ -85,43 +223,77 @@ 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')
+            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': 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'
+    _TEST = {
+        'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
+        'md5': 'd216c3a86493f9322545e045ddc3eb35',
+        'info_dict': {
+            'display_id': 'die-story-im-ersten-mission-unter-falscher-flagge',
+            'id': '100',
+            'ext': 'mp4',
+            'duration': 2600,
+            'title': 'Die Story im Ersten: Mission unter falscher Flagge',
+            'upload_date': '20140804',
+            'thumbnail': r're:^https?://.*\.jpg$',
+        },
+        'skip': 'HTTP Error 404: Not Found',
+    }

 
 

+    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,
         }