]> gitweb @ CieloNegro.org - youtube-dl.git/blob - youtube_dl/extractor/ard.py
[ard] Improve _VALID_URL (closes #25134) (#25198)
[youtube-dl.git] / youtube_dl / extractor / ard.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import json
5 import re
6
7 from .common import InfoExtractor
8 from .generic import GenericIE
9 from ..utils import (
10     determine_ext,
11     ExtractorError,
12     int_or_none,
13     parse_duration,
14     qualities,
15     str_or_none,
16     try_get,
17     unified_strdate,
18     unified_timestamp,
19     update_url_query,
20     url_or_none,
21     xpath_text,
22 )
23 from ..compat import compat_etree_fromstring
24
25
26 class ARDMediathekBaseIE(InfoExtractor):
27     _GEO_COUNTRIES = ['DE']
28
29     def _extract_media_info(self, media_info_url, webpage, video_id):
30         media_info = self._download_json(
31             media_info_url, video_id, 'Downloading media JSON')
32         return self._parse_media_info(media_info, video_id, '"fsk"' in webpage)
33
34     def _parse_media_info(self, media_info, video_id, fsk):
35         formats = self._extract_formats(media_info, video_id)
36
37         if not formats:
38             if fsk:
39                 raise ExtractorError(
40                     'This video is only available after 20:00', expected=True)
41             elif media_info.get('_geoblocked'):
42                 self.raise_geo_restricted(
43                     'This video is not available due to geoblocking',
44                     countries=self._GEO_COUNTRIES)
45
46         self._sort_formats(formats)
47
48         subtitles = {}
49         subtitle_url = media_info.get('_subtitleUrl')
50         if subtitle_url:
51             subtitles['de'] = [{
52                 'ext': 'ttml',
53                 'url': subtitle_url,
54             }]
55
56         return {
57             'id': video_id,
58             'duration': int_or_none(media_info.get('_duration')),
59             'thumbnail': media_info.get('_previewImage'),
60             'is_live': media_info.get('_isLive') is True,
61             'formats': formats,
62             'subtitles': subtitles,
63         }
64
65     def _extract_formats(self, media_info, video_id):
66         type_ = media_info.get('_type')
67         media_array = media_info.get('_mediaArray', [])
68         formats = []
69         for num, media in enumerate(media_array):
70             for stream in media.get('_mediaStreamArray', []):
71                 stream_urls = stream.get('_stream')
72                 if not stream_urls:
73                     continue
74                 if not isinstance(stream_urls, list):
75                     stream_urls = [stream_urls]
76                 quality = stream.get('_quality')
77                 server = stream.get('_server')
78                 for stream_url in stream_urls:
79                     if not url_or_none(stream_url):
80                         continue
81                     ext = determine_ext(stream_url)
82                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
83                         continue
84                     if ext == 'f4m':
85                         formats.extend(self._extract_f4m_formats(
86                             update_url_query(stream_url, {
87                                 'hdcore': '3.1.1',
88                                 'plugin': 'aasp-3.1.1.69.124'
89                             }), video_id, f4m_id='hds', fatal=False))
90                     elif ext == 'm3u8':
91                         formats.extend(self._extract_m3u8_formats(
92                             stream_url, video_id, 'mp4', 'm3u8_native',
93                             m3u8_id='hls', fatal=False))
94                     else:
95                         if server and server.startswith('rtmp'):
96                             f = {
97                                 'url': server,
98                                 'play_path': stream_url,
99                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
100                             }
101                         else:
102                             f = {
103                                 'url': stream_url,
104                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
105                             }
106                         m = re.search(
107                             r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$',
108                             stream_url)
109                         if m:
110                             f.update({
111                                 'width': int(m.group('width')),
112                                 'height': int(m.group('height')),
113                             })
114                         if type_ == 'audio':
115                             f['vcodec'] = 'none'
116                         formats.append(f)
117         return formats
118
119
120 class ARDMediathekIE(ARDMediathekBaseIE):
121     IE_NAME = 'ARD:mediathek'
122     _VALID_URL = r'^https?://(?:(?:(?:www|classic)\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
123
124     _TESTS = [{
125         # available till 26.07.2022
126         'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',
127         'info_dict': {
128             'id': '44726822',
129             'ext': 'mp4',
130             'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
131             'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
132             'duration': 1740,
133         },
134         'params': {
135             # m3u8 download
136             'skip_download': True,
137         }
138     }, {
139         'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
140         'only_matching': True,
141     }, {
142         # audio
143         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
144         'only_matching': True,
145     }, {
146         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
147         'only_matching': True,
148     }, {
149         # audio
150         'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
151         'only_matching': True,
152     }, {
153         'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
154         'only_matching': True,
155     }]
156
157     @classmethod
158     def suitable(cls, url):
159         return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)
160
161     def _real_extract(self, url):
162         # determine video id from url
163         m = re.match(self._VALID_URL, url)
164
165         document_id = None
166
167         numid = re.search(r'documentId=([0-9]+)', url)
168         if numid:
169             document_id = video_id = numid.group(1)
170         else:
171             video_id = m.group('video_id')
172
173         webpage = self._download_webpage(url, video_id)
174
175         ERRORS = (
176             ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
177             ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
178              'Video %s is no longer available'),
179         )
180
181         for pattern, message in ERRORS:
182             if pattern in webpage:
183                 raise ExtractorError(message % video_id, expected=True)
184
185         if re.search(r'[\?&]rss($|[=&])', url):
186             doc = compat_etree_fromstring(webpage.encode('utf-8'))
187             if doc.tag == 'rss':
188                 return GenericIE()._extract_rss(url, video_id, doc)
189
190         title = self._html_search_regex(
191             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
192              r'<meta name="dcterms\.title" content="(.*?)"/>',
193              r'<h4 class="headline">(.*?)</h4>',
194              r'<title[^>]*>(.*?)</title>'],
195             webpage, 'title')
196         description = self._html_search_meta(
197             'dcterms.abstract', webpage, 'description', default=None)
198         if description is None:
199             description = self._html_search_meta(
200                 'description', webpage, 'meta description', default=None)
201         if description is None:
202             description = self._html_search_regex(
203                 r'<p\s+class="teasertext">(.+?)</p>',
204                 webpage, 'teaser text', default=None)
205
206         # Thumbnail is sometimes not present.
207         # It is in the mobile version, but that seems to use a different URL
208         # structure altogether.
209         thumbnail = self._og_search_thumbnail(webpage, default=None)
210
211         media_streams = re.findall(r'''(?x)
212             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
213             "([^"]+)"''', webpage)
214
215         if media_streams:
216             QUALITIES = qualities(['lo', 'hi', 'hq'])
217             formats = []
218             for furl in set(media_streams):
219                 if furl.endswith('.f4m'):
220                     fid = 'f4m'
221                 else:
222                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
223                     fid = fid_m.group(1) if fid_m else None
224                 formats.append({
225                     'quality': QUALITIES(fid),
226                     'format_id': fid,
227                     'url': furl,
228                 })
229             self._sort_formats(formats)
230             info = {
231                 'formats': formats,
232             }
233         else:  # request JSON file
234             if not document_id:
235                 video_id = self._search_regex(
236                     r'/play/(?:config|media)/(\d+)', webpage, 'media id')
237             info = self._extract_media_info(
238                 'http://www.ardmediathek.de/play/media/%s' % video_id,
239                 webpage, video_id)
240
241         info.update({
242             'id': video_id,
243             'title': self._live_title(title) if info.get('is_live') else title,
244             'description': description,
245             'thumbnail': thumbnail,
246         })
247
248         return info
249
250
251 class ARDIE(InfoExtractor):
252     _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos(?:extern)?/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
253     _TESTS = [{
254         # available till 14.02.2019
255         'url': 'http://www.daserste.de/information/talk/maischberger/videos/das-groko-drama-zerlegen-sich-die-volksparteien-video-102.html',
256         'md5': '8e4ec85f31be7c7fc08a26cdbc5a1f49',
257         'info_dict': {
258             'display_id': 'das-groko-drama-zerlegen-sich-die-volksparteien-video',
259             'id': '102',
260             'ext': 'mp4',
261             'duration': 4435.0,
262             'title': 'Das GroKo-Drama: Zerlegen sich die Volksparteien?',
263             'upload_date': '20180214',
264             'thumbnail': r're:^https?://.*\.jpg$',
265         },
266     }, {
267         'url': 'https://www.daserste.de/information/reportage-dokumentation/erlebnis-erde/videosextern/woelfe-und-herdenschutzhunde-ungleiche-brueder-102.html',
268         'only_matching': True,
269     }, {
270         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
271         'only_matching': True,
272     }]
273
274     def _real_extract(self, url):
275         mobj = re.match(self._VALID_URL, url)
276         display_id = mobj.group('display_id')
277
278         player_url = mobj.group('mainurl') + '~playerXml.xml'
279         doc = self._download_xml(player_url, display_id)
280         video_node = doc.find('./video')
281         upload_date = unified_strdate(xpath_text(
282             video_node, './broadcastDate'))
283         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
284
285         formats = []
286         for a in video_node.findall('.//asset'):
287             f = {
288                 'format_id': a.attrib['type'],
289                 'width': int_or_none(a.find('./frameWidth').text),
290                 'height': int_or_none(a.find('./frameHeight').text),
291                 'vbr': int_or_none(a.find('./bitrateVideo').text),
292                 'abr': int_or_none(a.find('./bitrateAudio').text),
293                 'vcodec': a.find('./codecVideo').text,
294                 'tbr': int_or_none(a.find('./totalBitrate').text),
295             }
296             if a.find('./serverPrefix').text:
297                 f['url'] = a.find('./serverPrefix').text
298                 f['playpath'] = a.find('./fileName').text
299             else:
300                 f['url'] = a.find('./fileName').text
301             formats.append(f)
302         self._sort_formats(formats)
303
304         return {
305             'id': mobj.group('id'),
306             'formats': formats,
307             'display_id': display_id,
308             'title': video_node.find('./title').text,
309             'duration': parse_duration(video_node.find('./duration').text),
310             'upload_date': upload_date,
311             'thumbnail': thumbnail,
312         }
313
314
315 class ARDBetaMediathekIE(ARDMediathekBaseIE):
316     _VALID_URL = r'https://(?:beta|www)\.ardmediathek\.de/(?P<client>[^/]+)/(?:player|live)/(?P<video_id>[a-zA-Z0-9]+)(?:/(?P<display_id>[^/?#]+))?'
317     _TESTS = [{
318         'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
319         'md5': 'dfdc87d2e7e09d073d5a80770a9ce88f',
320         'info_dict': {
321             'display_id': 'die-robuste-roswita',
322             'id': '70153354',
323             'title': 'Die robuste Roswita',
324             'description': r're:^Der Mord.*trüber ist als die Ilm.',
325             'duration': 5316,
326             'thumbnail': 'https://img.ardmediathek.de/standard/00/70/15/33/90/-1852531467/16x9/960?mandant=ard',
327             'timestamp': 1577047500,
328             'upload_date': '20191222',
329             'ext': 'mp4',
330         },
331     }, {
332         'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3N3ci5kZS9hZXgvbzEwNzE5MTU/',
333         'only_matching': True,
334     }, {
335         'url': 'https://www.ardmediathek.de/swr/live/Y3JpZDovL3N3ci5kZS8xMzQ4MTA0Mg',
336         'only_matching': True,
337     }]
338
339     def _real_extract(self, url):
340         mobj = re.match(self._VALID_URL, url)
341         video_id = mobj.group('video_id')
342         display_id = mobj.group('display_id') or video_id
343
344         player_page = self._download_json(
345             'https://api.ardmediathek.de/public-gateway',
346             display_id, data=json.dumps({
347                 'query': '''{
348   playerPage(client:"%s", clipId: "%s") {
349     blockedByFsk
350     broadcastedOn
351     maturityContentRating
352     mediaCollection {
353       _duration
354       _geoblocked
355       _isLive
356       _mediaArray {
357         _mediaStreamArray {
358           _quality
359           _server
360           _stream
361         }
362       }
363       _previewImage
364       _subtitleUrl
365       _type
366     }
367     show {
368       title
369     }
370     synopsis
371     title
372     tracking {
373       atiCustomVars {
374         contentId
375       }
376     }
377   }
378 }''' % (mobj.group('client'), video_id),
379             }).encode(), headers={
380                 'Content-Type': 'application/json'
381             })['data']['playerPage']
382         title = player_page['title']
383         content_id = str_or_none(try_get(
384             player_page, lambda x: x['tracking']['atiCustomVars']['contentId']))
385         media_collection = player_page.get('mediaCollection') or {}
386         if not media_collection and content_id:
387             media_collection = self._download_json(
388                 'https://www.ardmediathek.de/play/media/' + content_id,
389                 content_id, fatal=False) or {}
390         info = self._parse_media_info(
391             media_collection, content_id or video_id,
392             player_page.get('blockedByFsk'))
393         age_limit = None
394         description = player_page.get('synopsis')
395         maturity_content_rating = player_page.get('maturityContentRating')
396         if maturity_content_rating:
397             age_limit = int_or_none(maturity_content_rating.lstrip('FSK'))
398         if not age_limit and description:
399             age_limit = int_or_none(self._search_regex(
400                 r'\(FSK\s*(\d+)\)\s*$', description, 'age limit', default=None))
401         info.update({
402             'age_limit': age_limit,
403             'display_id': display_id,
404             'title': title,
405             'description': description,
406             'timestamp': unified_timestamp(player_page.get('broadcastedOn')),
407             'series': try_get(player_page, lambda x: x['show']['title']),
408         })
409         return info