]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/ard.py
Update changelog.
[youtubedl] / youtube_dl / extractor / ard.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .common import InfoExtractor
7 from .generic import GenericIE
8 from ..utils import (
9 determine_ext,
10 ExtractorError,
11 qualities,
12 int_or_none,
13 parse_duration,
14 unified_strdate,
15 xpath_text,
16 update_url_query,
17 url_or_none,
18 )
19 from ..compat import compat_etree_fromstring
20
21
22 class ARDMediathekIE(InfoExtractor):
23 IE_NAME = 'ARD:mediathek'
24 _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
25
26 _TESTS = [{
27 # available till 26.07.2022
28 'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',
29 'info_dict': {
30 'id': '44726822',
31 'ext': 'mp4',
32 'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
33 'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
34 'duration': 1740,
35 },
36 'params': {
37 # m3u8 download
38 'skip_download': True,
39 }
40 }, {
41 'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
42 'only_matching': True,
43 }, {
44 # audio
45 'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
46 'only_matching': True,
47 }, {
48 'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
49 'only_matching': True,
50 }, {
51 # audio
52 'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
53 'only_matching': True,
54 }]
55
56 def _extract_media_info(self, media_info_url, webpage, video_id):
57 media_info = self._download_json(
58 media_info_url, video_id, 'Downloading media JSON')
59
60 formats = self._extract_formats(media_info, video_id)
61
62 if not formats:
63 if '"fsk"' in webpage:
64 raise ExtractorError(
65 'This video is only available after 20:00', expected=True)
66 elif media_info.get('_geoblocked'):
67 raise ExtractorError('This video is not available due to geo restriction', expected=True)
68
69 self._sort_formats(formats)
70
71 duration = int_or_none(media_info.get('_duration'))
72 thumbnail = media_info.get('_previewImage')
73 is_live = media_info.get('_isLive') is True
74
75 subtitles = {}
76 subtitle_url = media_info.get('_subtitleUrl')
77 if subtitle_url:
78 subtitles['de'] = [{
79 'ext': 'ttml',
80 'url': subtitle_url,
81 }]
82
83 return {
84 'id': video_id,
85 'duration': duration,
86 'thumbnail': thumbnail,
87 'is_live': is_live,
88 'formats': formats,
89 'subtitles': subtitles,
90 }
91
92 def _extract_formats(self, media_info, video_id):
93 type_ = media_info.get('_type')
94 media_array = media_info.get('_mediaArray', [])
95 formats = []
96 for num, media in enumerate(media_array):
97 for stream in media.get('_mediaStreamArray', []):
98 stream_urls = stream.get('_stream')
99 if not stream_urls:
100 continue
101 if not isinstance(stream_urls, list):
102 stream_urls = [stream_urls]
103 quality = stream.get('_quality')
104 server = stream.get('_server')
105 for stream_url in stream_urls:
106 if not url_or_none(stream_url):
107 continue
108 ext = determine_ext(stream_url)
109 if quality != 'auto' and ext in ('f4m', 'm3u8'):
110 continue
111 if ext == 'f4m':
112 formats.extend(self._extract_f4m_formats(
113 update_url_query(stream_url, {
114 'hdcore': '3.1.1',
115 'plugin': 'aasp-3.1.1.69.124'
116 }),
117 video_id, f4m_id='hds', fatal=False))
118 elif ext == 'm3u8':
119 formats.extend(self._extract_m3u8_formats(
120 stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
121 else:
122 if server and server.startswith('rtmp'):
123 f = {
124 'url': server,
125 'play_path': stream_url,
126 'format_id': 'a%s-rtmp-%s' % (num, quality),
127 }
128 else:
129 f = {
130 'url': stream_url,
131 'format_id': 'a%s-%s-%s' % (num, ext, quality)
132 }
133 m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
134 if m:
135 f.update({
136 'width': int(m.group('width')),
137 'height': int(m.group('height')),
138 })
139 if type_ == 'audio':
140 f['vcodec'] = 'none'
141 formats.append(f)
142 return formats
143
144 def _real_extract(self, url):
145 # determine video id from url
146 m = re.match(self._VALID_URL, url)
147
148 document_id = None
149
150 numid = re.search(r'documentId=([0-9]+)', url)
151 if numid:
152 document_id = video_id = numid.group(1)
153 else:
154 video_id = m.group('video_id')
155
156 webpage = self._download_webpage(url, video_id)
157
158 ERRORS = (
159 ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
160 ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
161 'Video %s is no longer available'),
162 )
163
164 for pattern, message in ERRORS:
165 if pattern in webpage:
166 raise ExtractorError(message % video_id, expected=True)
167
168 if re.search(r'[\?&]rss($|[=&])', url):
169 doc = compat_etree_fromstring(webpage.encode('utf-8'))
170 if doc.tag == 'rss':
171 return GenericIE()._extract_rss(url, video_id, doc)
172
173 title = self._html_search_regex(
174 [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
175 r'<meta name="dcterms\.title" content="(.*?)"/>',
176 r'<h4 class="headline">(.*?)</h4>'],
177 webpage, 'title')
178 description = self._html_search_meta(
179 'dcterms.abstract', webpage, 'description', default=None)
180 if description is None:
181 description = self._html_search_meta(
182 'description', webpage, 'meta description')
183
184 # Thumbnail is sometimes not present.
185 # It is in the mobile version, but that seems to use a different URL
186 # structure altogether.
187 thumbnail = self._og_search_thumbnail(webpage, default=None)
188
189 media_streams = re.findall(r'''(?x)
190 mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
191 "([^"]+)"''', webpage)
192
193 if media_streams:
194 QUALITIES = qualities(['lo', 'hi', 'hq'])
195 formats = []
196 for furl in set(media_streams):
197 if furl.endswith('.f4m'):
198 fid = 'f4m'
199 else:
200 fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
201 fid = fid_m.group(1) if fid_m else None
202 formats.append({
203 'quality': QUALITIES(fid),
204 'format_id': fid,
205 'url': furl,
206 })
207 self._sort_formats(formats)
208 info = {
209 'formats': formats,
210 }
211 else: # request JSON file
212 if not document_id:
213 video_id = self._search_regex(
214 r'/play/(?:config|media)/(\d+)', webpage, 'media id')
215 info = self._extract_media_info(
216 'http://www.ardmediathek.de/play/media/%s' % video_id,
217 webpage, video_id)
218
219 info.update({
220 'id': video_id,
221 'title': self._live_title(title) if info.get('is_live') else title,
222 'description': description,
223 'thumbnail': thumbnail,
224 })
225
226 return info
227
228
229 class ARDIE(InfoExtractor):
230 _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
231 _TESTS = [{
232 # available till 14.02.2019
233 'url': 'http://www.daserste.de/information/talk/maischberger/videos/das-groko-drama-zerlegen-sich-die-volksparteien-video-102.html',
234 'md5': '8e4ec85f31be7c7fc08a26cdbc5a1f49',
235 'info_dict': {
236 'display_id': 'das-groko-drama-zerlegen-sich-die-volksparteien-video',
237 'id': '102',
238 'ext': 'mp4',
239 'duration': 4435.0,
240 'title': 'Das GroKo-Drama: Zerlegen sich die Volksparteien?',
241 'upload_date': '20180214',
242 'thumbnail': r're:^https?://.*\.jpg$',
243 },
244 }, {
245 'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
246 'only_matching': True,
247 }]
248
249 def _real_extract(self, url):
250 mobj = re.match(self._VALID_URL, url)
251 display_id = mobj.group('display_id')
252
253 player_url = mobj.group('mainurl') + '~playerXml.xml'
254 doc = self._download_xml(player_url, display_id)
255 video_node = doc.find('./video')
256 upload_date = unified_strdate(xpath_text(
257 video_node, './broadcastDate'))
258 thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
259
260 formats = []
261 for a in video_node.findall('.//asset'):
262 f = {
263 'format_id': a.attrib['type'],
264 'width': int_or_none(a.find('./frameWidth').text),
265 'height': int_or_none(a.find('./frameHeight').text),
266 'vbr': int_or_none(a.find('./bitrateVideo').text),
267 'abr': int_or_none(a.find('./bitrateAudio').text),
268 'vcodec': a.find('./codecVideo').text,
269 'tbr': int_or_none(a.find('./totalBitrate').text),
270 }
271 if a.find('./serverPrefix').text:
272 f['url'] = a.find('./serverPrefix').text
273 f['playpath'] = a.find('./fileName').text
274 else:
275 f['url'] = a.find('./fileName').text
276 formats.append(f)
277 self._sort_formats(formats)
278
279 return {
280 'id': mobj.group('id'),
281 'formats': formats,
282 'display_id': display_id,
283 'title': video_node.find('./title').text,
284 'duration': parse_duration(video_node.find('./duration').text),
285 'upload_date': upload_date,
286 'thumbnail': thumbnail,
287 }
288
289
290 class ARDBetaMediathekIE(InfoExtractor):
291 _VALID_URL = r'https://beta\.ardmediathek\.de/[a-z]+/player/(?P<video_id>[a-zA-Z0-9]+)/(?P<display_id>[^/?#]+)'
292 _TESTS = [{
293 'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
294 'md5': '2d02d996156ea3c397cfc5036b5d7f8f',
295 'info_dict': {
296 'display_id': 'die-robuste-roswita',
297 'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
298 'title': 'Tatort: Die robuste Roswita',
299 'description': r're:^Der Mord.*trüber ist als die Ilm.',
300 'duration': 5316,
301 'thumbnail': 'https://img.ardmediathek.de/standard/00/55/43/59/34/-1774185891/16x9/960?mandant=ard',
302 'upload_date': '20180826',
303 'ext': 'mp4',
304 },
305 }]
306
307 def _real_extract(self, url):
308 mobj = re.match(self._VALID_URL, url)
309 video_id = mobj.group('video_id')
310 display_id = mobj.group('display_id')
311
312 webpage = self._download_webpage(url, display_id)
313 data_json = self._search_regex(r'window\.__APOLLO_STATE__\s*=\s*(\{.*);\n', webpage, 'json')
314 data = self._parse_json(data_json, display_id)
315
316 res = {
317 'id': video_id,
318 'display_id': display_id,
319 }
320 formats = []
321 for widget in data.values():
322 if widget.get('_geoblocked'):
323 raise ExtractorError('This video is not available due to geoblocking', expected=True)
324
325 if '_duration' in widget:
326 res['duration'] = widget['_duration']
327 if 'clipTitle' in widget:
328 res['title'] = widget['clipTitle']
329 if '_previewImage' in widget:
330 res['thumbnail'] = widget['_previewImage']
331 if 'broadcastedOn' in widget:
332 res['upload_date'] = unified_strdate(widget['broadcastedOn'])
333 if 'synopsis' in widget:
334 res['description'] = widget['synopsis']
335 if '_subtitleUrl' in widget:
336 res['subtitles'] = {'de': [{
337 'ext': 'ttml',
338 'url': widget['_subtitleUrl'],
339 }]}
340 if '_quality' in widget:
341 format_url = widget['_stream']['json'][0]
342
343 if format_url.endswith('.f4m'):
344 formats.extend(self._extract_f4m_formats(
345 format_url + '?hdcore=3.11.0',
346 video_id, f4m_id='hds', fatal=False))
347 elif format_url.endswith('m3u8'):
348 formats.extend(self._extract_m3u8_formats(
349 format_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
350 else:
351 formats.append({
352 'format_id': 'http-' + widget['_quality'],
353 'url': format_url,
354 'preference': 10, # Plain HTTP, that's nice
355 })
356
357 self._sort_formats(formats)
358 res['formats'] = formats
359
360 return res