]> git.bitcoin.ninja Git - youtube-dl/commitdiff
[WDR] complete overhaul after relaunch of the site
authorBoris Wachtmeister <boris-code@gmx.com>
Sat, 12 Mar 2016 17:00:26 +0000 (18:00 +0100)
committerBoris Wachtmeister <boris-code@gmx.com>
Sat, 23 Apr 2016 09:42:18 +0000 (11:42 +0200)
The WDR relaunched their site on 2016-02-23 which not only changed the
URL-schema completely but also the layout of their pages.

Apparently the whole "mediathek" now runs on the wdr-domain, so no
separate URL for funkhauseuropa anymore.
There seems to be no explicit handling of video-sizes on the page or in
the URLs anymore. There seems to be only one size for HTML5, but still
several sizes for flash. The extractor adds all to the list of formats.

There is no metadata for the HTML5-stream, so that the best flash-stream
will always be considered as the "best" format. At least in my tests
this seemed to be true anyway.

youtube_dl/extractor/wdr.py

index 31c90430327da895ffc974c1d489cb4c92689d2f..f881b7300e986e2a9ac7e39335c64b3beee3ebb8 100644 (file)
@@ -1,7 +1,6 @@
 # -*- coding: utf-8 -*-
 from __future__ import unicode_literals
 
-import itertools
 import re
 
 from .common import InfoExtractor
@@ -11,204 +10,156 @@ from ..compat import (
 )
 from ..utils import (
     unified_strdate,
-    qualities,
+    ExtractorError,
 )
 
 
 class WDRIE(InfoExtractor):
-    _PLAYER_REGEX = '-(?:video|audio)player(?:_size-[LMS])?'
-    _VALID_URL = r'(?P<url>https?://www\d?\.(?:wdr\d?|funkhauseuropa)\.de/)(?P<id>.+?)(?P<player>%s)?\.html' % _PLAYER_REGEX
+    _PAGE_REGEX = r'/mediathek/(?P<media_type>[^/]+)/(?P<type>[^/]+)/(?P<display_id>.+)\.html'
+    _VALID_URL = r'(?P<page_url>https?://(?:www\d\.)?wdr\d?\.de)' + _PAGE_REGEX
+
+    _JS_URL_REGEX = r'(https?://deviceids-medp.wdr.de/ondemand/\d+/\d+\.js)'
 
     _TESTS = [
         {
-            'url': 'http://www1.wdr.de/mediathek/video/sendungen/servicezeit/videoservicezeit560-videoplayer_size-L.html',
-            'info_dict': {
-                'id': 'mdb-362427',
-                'ext': 'flv',
-                'title': 'Servicezeit',
-                'description': 'md5:c8f43e5e815eeb54d0b96df2fba906cb',
-                'upload_date': '20140310',
-                'is_live': False
-            },
-            'params': {
-                'skip_download': True,
-            },
-            'skip': 'Page Not Found',
-        },
-        {
-            'url': 'http://www1.wdr.de/themen/av/videomargaspiegelisttot101-videoplayer.html',
+            'url': 'http://www1.wdr.de/mediathek/video/sendungen/doku-am-freitag/video-geheimnis-aachener-dom-100.html',
+            'md5': 'e58c39c3e30077141d258bf588700a7b',
             'info_dict': {
-                'id': 'mdb-363194',
+                'id': 'mdb-1058683',
                 'ext': 'flv',
-                'title': 'Marga Spiegel ist tot',
-                'description': 'md5:2309992a6716c347891c045be50992e4',
-                'upload_date': '20140311',
-                'is_live': False
-            },
-            'params': {
-                'skip_download': True,
+                'display_id': 'doku-am-freitag/video-geheimnis-aachener-dom-100',
+                'title': 'Geheimnis Aachener Dom',
+                'alt_title': 'Doku am Freitag',
+                'upload_date': '20160304',
+                'description': 'md5:87be8ff14d8dfd7a7ee46f0299b52318',
+                'is_live': False,
+                'subtitles': {'de': [{
+                    'url': 'http://ondemand-ww.wdr.de/medp/fsk0/105/1058683/1058683_12220974.xml'
+                }]},
             },
             'skip': 'Page Not Found',
         },
         {
-            'url': 'http://www1.wdr.de/themen/kultur/audioerlebtegeschichtenmargaspiegel100-audioplayer.html',
-            'md5': '83e9e8fefad36f357278759870805898',
+            'url': 'http://www1.wdr.de/mediathek/audio/wdr3/wdr3-gespraech-am-samstag/audio-schriftstellerin-juli-zeh-100.html',
+            'md5': 'f4c1f96d01cf285240f53ea4309663d8',
             'info_dict': {
-                'id': 'mdb-194332',
+                'id': 'mdb-1072000',
                 'ext': 'mp3',
-                'title': 'Erlebte Geschichten: Marga Spiegel (29.11.2009)',
-                'description': 'md5:2309992a6716c347891c045be50992e4',
-                'upload_date': '20091129',
-                'is_live': False
-            },
-        },
-        {
-            'url': 'http://www.funkhauseuropa.de/av/audioflaviacoelhoamaramar100-audioplayer.html',
-            'md5': '99a1443ff29af19f6c52cf6f4dc1f4aa',
-            'info_dict': {
-                'id': 'mdb-478135',
-                'ext': 'mp3',
-                'title': 'Flavia Coelho: Amar é Amar',
-                'description': 'md5:7b29e97e10dfb6e265238b32fa35b23a',
-                'upload_date': '20140717',
-                'is_live': False
+                'display_id': 'wdr3-gespraech-am-samstag/audio-schriftstellerin-juli-zeh-100',
+                'title': 'Schriftstellerin Juli Zeh',
+                'alt_title': 'WDR 3 Gespräch am Samstag',
+                'upload_date': '20160312',
+                'description': 'md5:e127d320bc2b1f149be697ce044a3dd7',
+                'is_live': False,
+                'subtitles': {}
             },
             'skip': 'Page Not Found',
         },
         {
-            'url': 'http://www1.wdr.de/mediathek/video/sendungen/quarks_und_co/filterseite-quarks-und-co100.html',
-            'playlist_mincount': 146,
+            'url': 'http://www1.wdr.de/mediathek/video/live/index.html',
             'info_dict': {
-                'id': 'mediathek/video/sendungen/quarks_und_co/filterseite-quarks-und-co100',
+                'id': 'mdb-103364',
+                'ext': 'flv',
+                'display_id': 'index',
+                'title': r're:^WDR Fernsehen im Livestream [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
+                'alt_title': 'WDR Fernsehen Live',
+                'upload_date': None,
+                'description': 'md5:ae2ff888510623bf8d4b115f95a9b7c9',
+                'is_live': True,
+                'subtitles': {}
             }
         },
         {
-            'url': 'http://www1.wdr.de/mediathek/video/livestream/index.html',
+            'url': 'http://www1.wdr.de/mediathek/video/sendungen/aktuelle-stunde/aktuelle-stunde-120.html',
+            'playlist_mincount': 10,
             'info_dict': {
-                'id': 'mdb-103364',
-                'title': 're:^WDR Fernsehen Live [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
-                'description': 'md5:ae2ff888510623bf8d4b115f95a9b7c9',
-                'ext': 'flv',
-                'upload_date': '20150101',
-                'is_live': True
-            },
-            'params': {
-                'skip_download': True,
+                'id': 'aktuelle-stunde/aktuelle-stunde-120',
             },
         }
     ]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        page_url = mobj.group('url')
-        page_id = mobj.group('id')
+        url_type = mobj.group('type')
+        page_url = mobj.group('page_url')
+        display_id = mobj.group('display_id')
+        webpage = self._download_webpage(url, display_id)
 
-        webpage = self._download_webpage(url, page_id)
+        js_url = self._search_regex(self._JS_URL_REGEX, webpage, 'js_url', default=None)
 
-        if mobj.group('player') is None:
+        if not js_url:
             entries = [
-                self.url_result(page_url + href, 'WDR')
+                self.url_result(page_url + href[0], 'WDR')
                 for href in re.findall(
-                    r'<a href="/?(.+?%s\.html)" rel="nofollow"' % self._PLAYER_REGEX,
+                    r'<a href="(%s)"' % self._PAGE_REGEX,
                     webpage)
             ]
 
             if entries:  # Playlist page
-                return self.playlist_result(entries, page_id)
+                return self.playlist_result(entries, playlist_id=display_id)
 
-            # Overview page
-            entries = []
-            for page_num in itertools.count(2):
-                hrefs = re.findall(
-                    r'<li class="mediathekvideo"\s*>\s*<img[^>]*>\s*<a href="(/mediathek/video/[^"]+)"',
-                    webpage)
-                entries.extend(
-                    self.url_result(page_url + href, 'WDR')
-                    for href in hrefs)
-                next_url_m = re.search(
-                    r'<li class="nextToLast">\s*<a href="([^"]+)"', webpage)
-                if not next_url_m:
-                    break
-                next_url = page_url + next_url_m.group(1)
-                webpage = self._download_webpage(
-                    next_url, page_id,
-                    note='Downloading playlist page %d' % page_num)
-            return self.playlist_result(entries, page_id)
+            raise ExtractorError('No downloadable streams found', expected=True)
+
+        js_data = self._download_webpage(js_url, 'metadata')
+        json_data = self._search_regex(r'\(({.*})\)', js_data, 'json')
+        metadata = self._parse_json(json_data, display_id)
 
-        flashvars = compat_parse_qs(self._html_search_regex(
-            r'<param name="flashvars" value="([^"]+)"', webpage, 'flashvars'))
+        metadata_tracker_data = metadata["trackerData"]
+        metadata_media_resource = metadata["mediaResource"]
+
+        formats = []
+
+        # check if the metadata contains a direct URL to a file
+        metadata_media_alt = metadata_media_resource.get("alt")
+        if metadata_media_alt:
+            for tag_name in ["videoURL", 'audioURL']:
+                if tag_name in metadata_media_alt:
+                    formats.append({
+                        'url': metadata_media_alt[tag_name]
+                    })
 
-        page_id = flashvars['trackerClipId'][0]
-        video_url = flashvars['dslSrc'][0]
-        title = flashvars['trackerClipTitle'][0]
-        thumbnail = flashvars['startPicture'][0] if 'startPicture' in flashvars else None
-        is_live = flashvars.get('isLive', ['0'])[0] == '1'
+        # check if there are flash-streams for this video
+        if "dflt" in metadata_media_resource and "videoURL" in metadata_media_resource["dflt"]:
+            video_url = metadata_media_resource["dflt"]["videoURL"]
+            if video_url.endswith('.f4m'):
+                full_video_url = video_url + '?hdcore=3.2.0&plugin=aasp-3.2.0.77.18'
+                formats.extend(self._extract_f4m_formats(full_video_url, display_id, f4m_id='hds', fatal=False))
+            elif video_url.endswith('.smil'):
+                formats.extend(self._extract_smil_formats(video_url, 'stream', fatal=False))
+
+        subtitles = {}
+        caption_url = metadata_media_resource.get("captionURL")
+        if caption_url:
+            subtitles['de'] = [{
+                'url': caption_url
+            }]
+
+        title = metadata_tracker_data.get("trackerClipTitle")
+        is_live = url_type == 'live'
 
         if is_live:
             title = self._live_title(title)
-
-        if 'trackerClipAirTime' in flashvars:
-            upload_date = flashvars['trackerClipAirTime'][0]
+            upload_date = None
+        elif 'trackerClipAirTime' in metadata_tracker_data:
+            upload_date = metadata_tracker_data['trackerClipAirTime']
         else:
-            upload_date = self._html_search_meta(
-                'DC.Date', webpage, 'upload date')
+            upload_date = self._html_search_meta('DC.Date', webpage, 'upload date')
 
         if upload_date:
             upload_date = unified_strdate(upload_date)
 
-        formats = []
-        preference = qualities(['S', 'M', 'L', 'XL'])
-
-        if video_url.endswith('.f4m'):
-            formats.extend(self._extract_f4m_formats(
-                video_url + '?hdcore=3.2.0&plugin=aasp-3.2.0.77.18', page_id,
-                f4m_id='hds', fatal=False))
-        elif video_url.endswith('.smil'):
-            formats.extend(self._extract_smil_formats(
-                video_url, page_id, False, {
-                    'hdcore': '3.3.0',
-                    'plugin': 'aasp-3.3.0.99.43',
-                }))
-        else:
-            formats.append({
-                'url': video_url,
-                'http_headers': {
-                    'User-Agent': 'mobile',
-                },
-            })
-
-        m3u8_url = self._search_regex(
-            r'rel="adaptiv"[^>]+href="([^"]+)"',
-            webpage, 'm3u8 url', default=None)
-        if m3u8_url:
-            formats.extend(self._extract_m3u8_formats(
-                m3u8_url, page_id, 'mp4', 'm3u8_native',
-                m3u8_id='hls', fatal=False))
-
-        direct_urls = re.findall(
-            r'rel="web(S|M|L|XL)"[^>]+href="([^"]+)"', webpage)
-        if direct_urls:
-            for quality, video_url in direct_urls:
-                formats.append({
-                    'url': video_url,
-                    'preference': preference(quality),
-                    'http_headers': {
-                        'User-Agent': 'mobile',
-                    },
-                })
-
         self._sort_formats(formats)
 
-        description = self._html_search_meta('Description', webpage, 'description')
-
         return {
-            'id': page_id,
-            'formats': formats,
+            'id': metadata_tracker_data.get("trackerClipId", display_id),
+            'display_id': display_id,
             'title': title,
-            'description': description,
-            'thumbnail': thumbnail,
+            'alt_title': metadata_tracker_data.get("trackerClipSubcategory"),
+            'formats': formats,
             'upload_date': upload_date,
-            'is_live': is_live
+            'description': self._html_search_meta("Description", webpage),
+            'is_live': is_live,
+            'subtitles': subtitles,
         }