Merge tag 'upstream/2015.06.04.1'

[youtubedl] / youtube_dl / extractor / ard.py

diff --git a/youtube_dl/extractor/ard.py b/youtube_dl/extractor/ard.py
index b88f71bc40b9803fb4ed0dea134738a7e1e07201..6a35ea463edcafe3b9d7db4c53b9bf0c53198fd0 100644 (file)
--- a/youtube_dl/extractor/ard.py
+++ b/youtube_dl/extractor/ard.py
@@ -4,26 +4,36 @@ 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,

+    qualities,
+    int_or_none,
+    parse_duration,
+    unified_strdate,
+    xpath_text,
+    parse_xml,

 )
 
 
 )
 
 

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

 
 

-    _TEST = {
-        'url': 'http://www.ardmediathek.de/das-erste/guenther-jauch/edward-snowden-im-interview-held-oder-verraeter?documentId=19288786',
-        'file': '19288786.mp4',
-        'md5': '515bf47ce209fb3f5a61b7aad364634c',
+    _TESTS = [{
+        'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.ardmediathek.de/tv/Tatort/Das-Wunder-von-Wolbeck-Video-tgl-ab-20/Das-Erste/Video?documentId=22490580&bcastId=602916',

         'info_dict': {
         'info_dict': {

-            'title': 'Edward Snowden im Interview - Held oder Verräter?',
-            'description': 'Edward Snowden hat alles aufs Spiel gesetzt, um die weltweite \xdcberwachung durch die Geheimdienste zu enttarnen. Nun stellt sich der ehemalige NSA-Mitarbeiter erstmals weltweit in einem TV-Interview den Fragen eines NDR-Journalisten. Die Sendung vom Sonntagabend.',
-            'thumbnail': 'http://www.ardmediathek.de/ard/servlet/contentblob/19/28/87/90/19288790/bild/2250037',
+            'id': '22490580',
+            '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.',

         },
         'skip': 'Blocked outside of Germany',
         },
         'skip': 'Blocked outside of Germany',

-    }
+    }]

 
     def _real_extract(self, url):
         # determine video id from url
 
     def _real_extract(self, url):
         # determine video id from url


@@ -37,40 +47,81 @@ class ARDIE(InfoExtractor):
 
         webpage = self._download_webpage(url, video_id)
 
 
         webpage = self._download_webpage(url, video_id)
 

+        if '>Der gewünschte Beitrag ist nicht mehr verfügbar.<' in webpage:
+            raise ExtractorError('Video %s is no longer available' % video_id, expected=True)
+
+        if 'Diese Sendung ist für Jugendliche unter 12 Jahren nicht geeignet. Der Clip ist deshalb nur von 20 bis 6 Uhr verfügbar.' in webpage:
+            raise ExtractorError('This program is only suitable for those aged 12 and older. Video %s is therefore only available between 20 pm and 6 am.' % video_id, expected=True)
+
+        if re.search(r'[\?&]rss($|[=&])', url):
+            doc = parse_xml(webpage)
+            if doc.tag == 'rss':
+                return GenericIE()._extract_rss(url, video_id, doc)
+

         title = self._html_search_regex(
         title = self._html_search_regex(

-            r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>', webpage, 'title')
+            [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
+             r'<meta name="dcterms.title" content="(.*?)"/>',
+             r'<h4 class="headline">(.*?)</h4>'],
+            webpage, 'title')

         description = self._html_search_meta(
         description = self._html_search_meta(

-            'dcterms.abstract', webpage, 'description')
-        thumbnail = self._og_search_thumbnail(webpage)
+            'dcterms.abstract', webpage, 'description', default=None)
+        if description is None:
+            description = self._html_search_meta(
+                'description', webpage, 'meta description')

 
 

-        streams = [
-            mo.groupdict()
-            for mo in re.finditer(
-                r'mediaCollection\.addMediaStream\((?P<media_type>\d+), (?P<quality>\d+), "(?P<rtmp_url>[^"]*)", "(?P<video_url>[^"]*)", "[^"]*"\)', webpage)]
-        if not streams:
-            if '"fsk"' in webpage:
-                raise ExtractorError('This video is only available after 20:00')
+        # Thumbnail is sometimes not present.
+        # It is in the mobile version, but that seems to use a different URL
+        # structure altogether.
+        thumbnail = self._og_search_thumbnail(webpage, default=None)

 
 

-        formats = []
-        for s in streams:
-            format = {
-                'quality': int(s['quality']),
-            }
-            if s.get('rtmp_url'):
-                format['protocol'] = 'rtmp'
-                format['url'] = s['rtmp_url']
-                format['playpath'] = s['video_url']
-            else:
-                format['url'] = s['video_url']
+        media_streams = re.findall(r'''(?x)
+            mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
+            "([^"]+)"''', webpage)
+
+        if media_streams:
+            QUALITIES = qualities(['lo', 'hi', 'hq'])
+            formats = []
+            for furl in set(media_streams):
+                if furl.endswith('.f4m'):
+                    fid = 'f4m'
+                else:
+                    fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
+                    fid = fid_m.group(1) if fid_m else None
+                formats.append({
+                    'quality': QUALITIES(fid),
+                    'format_id': fid,
+                    'url': furl,
+                })
+        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')
+
+            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
+
+                format = {
+                    'quality': s['_quality'],
+                    'url': s['_stream'],
+                }

 
 

-            quality_name = self._search_regex(
-                r'[,.]([a-zA-Z0-9_-]+),?\.mp4', format['url'],
-                'quality name', default='NA')
-            format['format_id'] = '%s-%s-%s-%s' % (
-                determine_ext(format['url']), quality_name, s['media_type'],
-                s['quality'])
+                format['format_id'] = '%s-%s' % (
+                    determine_ext(format['url']), format['quality'])

 
 

-            formats.append(format)
+                formats.append(format)

 
         self._sort_formats(formats)
 
 
         self._sort_formats(formats)
 


@@ -81,3 +132,60 @@ class ARDIE(InfoExtractor):
             'formats': formats,
             'thumbnail': thumbnail,
         }
             'formats': formats,
             'thumbnail': thumbnail,
         }

+
+
+class ARDIE(InfoExtractor):
+    _VALID_URL = '(?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': 're:^https?://.*\.jpg$',
+        }
+    }
+
+    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 {
+            'id': mobj.group('id'),
+            '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,
+        }