]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/ard.py
Imported Upstream version 2016.02.22
[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 get_element_by_attribute,
12 qualities,
13 int_or_none,
14 parse_duration,
15 unified_strdate,
16 xpath_text,
17 )
18 from ..compat import compat_etree_fromstring
19
20
21 class ARDMediathekIE(InfoExtractor):
22 IE_NAME = 'ARD:mediathek'
23 _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
24
25 _TESTS = [{
26 'url': 'http://www.ardmediathek.de/tv/Dokumentation-und-Reportage/Ich-liebe-das-Leben-trotzdem/rbb-Fernsehen/Video?documentId=29582122&bcastId=3822114',
27 'info_dict': {
28 'id': '29582122',
29 'ext': 'mp4',
30 'title': 'Ich liebe das Leben trotzdem',
31 'description': 'md5:45e4c225c72b27993314b31a84a5261c',
32 'duration': 4557,
33 },
34 'params': {
35 # m3u8 download
36 'skip_download': True,
37 },
38 }, {
39 'url': 'http://www.ardmediathek.de/tv/Tatort/Tatort-Scheinwelten-H%C3%B6rfassung-Video/Das-Erste/Video?documentId=29522730&bcastId=602916',
40 'md5': 'f4d98b10759ac06c0072bbcd1f0b9e3e',
41 'info_dict': {
42 'id': '29522730',
43 'ext': 'mp4',
44 'title': 'Tatort: Scheinwelten - Hörfassung (Video tgl. ab 20 Uhr)',
45 'description': 'md5:196392e79876d0ac94c94e8cdb2875f1',
46 'duration': 5252,
47 },
48 }, {
49 # audio
50 'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
51 'md5': '219d94d8980b4f538c7fcb0865eb7f2c',
52 'info_dict': {
53 'id': '28488308',
54 'ext': 'mp3',
55 'title': 'Tod eines Fußballers',
56 'description': 'md5:f6e39f3461f0e1f54bfa48c8875c86ef',
57 'duration': 3240,
58 },
59 }, {
60 'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
61 'only_matching': True,
62 }]
63
64 def _extract_media_info(self, media_info_url, webpage, video_id):
65 media_info = self._download_json(
66 media_info_url, video_id, 'Downloading media JSON')
67
68 formats = self._extract_formats(media_info, video_id)
69
70 if not formats:
71 if '"fsk"' in webpage:
72 raise ExtractorError(
73 'This video is only available after 20:00', expected=True)
74 elif media_info.get('_geoblocked'):
75 raise ExtractorError('This video is not available due to geo restriction', expected=True)
76
77 self._sort_formats(formats)
78
79 duration = int_or_none(media_info.get('_duration'))
80 thumbnail = media_info.get('_previewImage')
81
82 subtitles = {}
83 subtitle_url = media_info.get('_subtitleUrl')
84 if subtitle_url:
85 subtitles['de'] = [{
86 'ext': 'srt',
87 'url': subtitle_url,
88 }]
89
90 return {
91 'id': video_id,
92 'duration': duration,
93 'thumbnail': thumbnail,
94 'formats': formats,
95 'subtitles': subtitles,
96 }
97
98 def _extract_formats(self, media_info, video_id):
99 type_ = media_info.get('_type')
100 media_array = media_info.get('_mediaArray', [])
101 formats = []
102 for num, media in enumerate(media_array):
103 for stream in media.get('_mediaStreamArray', []):
104 stream_urls = stream.get('_stream')
105 if not stream_urls:
106 continue
107 if not isinstance(stream_urls, list):
108 stream_urls = [stream_urls]
109 quality = stream.get('_quality')
110 server = stream.get('_server')
111 for stream_url in stream_urls:
112 ext = determine_ext(stream_url)
113 if quality != 'auto' and ext in ('f4m', 'm3u8'):
114 continue
115 if ext == 'f4m':
116 formats.extend(self._extract_f4m_formats(
117 stream_url + '?hdcore=3.1.1&plugin=aasp-3.1.1.69.124',
118 video_id, preference=-1, f4m_id='hds', fatal=False))
119 elif ext == 'm3u8':
120 formats.extend(self._extract_m3u8_formats(
121 stream_url, video_id, 'mp4', preference=1, m3u8_id='hls', fatal=False))
122 else:
123 if server and server.startswith('rtmp'):
124 f = {
125 'url': server,
126 'play_path': stream_url,
127 'format_id': 'a%s-rtmp-%s' % (num, quality),
128 }
129 elif stream_url.startswith('http'):
130 f = {
131 'url': stream_url,
132 'format_id': 'a%s-%s-%s' % (num, ext, quality)
133 }
134 else:
135 continue
136 m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
137 if m:
138 f.update({
139 'width': int(m.group('width')),
140 'height': int(m.group('height')),
141 })
142 if type_ == 'audio':
143 f['vcodec'] = 'none'
144 formats.append(f)
145 return formats
146
147 def _real_extract(self, url):
148 # determine video id from url
149 m = re.match(self._VALID_URL, url)
150
151 numid = re.search(r'documentId=([0-9]+)', url)
152 if numid:
153 video_id = numid.group(1)
154 else:
155 video_id = m.group('video_id')
156
157 webpage = self._download_webpage(url, video_id)
158
159 if '>Der gewünschte Beitrag ist nicht mehr verfügbar.<' in webpage:
160 raise ExtractorError('Video %s is no longer available' % video_id, expected=True)
161
162 if 'Diese Sendung ist für Jugendliche unter 12 Jahren nicht geeignet. Der Clip ist deshalb nur von 20 bis 6 Uhr verfügbar.' in webpage:
163 raise ExtractorError('This program is only suitable for those aged 12 and older. Video %s is therefore only available between 20 pm and 6 am.' % video_id, expected=True)
164
165 if re.search(r'[\?&]rss($|[=&])', url):
166 doc = compat_etree_fromstring(webpage.encode('utf-8'))
167 if doc.tag == 'rss':
168 return GenericIE()._extract_rss(url, video_id, doc)
169
170 title = self._html_search_regex(
171 [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
172 r'<meta name="dcterms.title" content="(.*?)"/>',
173 r'<h4 class="headline">(.*?)</h4>'],
174 webpage, 'title')
175 description = self._html_search_meta(
176 'dcterms.abstract', webpage, 'description', default=None)
177 if description is None:
178 description = self._html_search_meta(
179 'description', webpage, 'meta description')
180
181 # Thumbnail is sometimes not present.
182 # It is in the mobile version, but that seems to use a different URL
183 # structure altogether.
184 thumbnail = self._og_search_thumbnail(webpage, default=None)
185
186 media_streams = re.findall(r'''(?x)
187 mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
188 "([^"]+)"''', webpage)
189
190 if media_streams:
191 QUALITIES = qualities(['lo', 'hi', 'hq'])
192 formats = []
193 for furl in set(media_streams):
194 if furl.endswith('.f4m'):
195 fid = 'f4m'
196 else:
197 fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
198 fid = fid_m.group(1) if fid_m else None
199 formats.append({
200 'quality': QUALITIES(fid),
201 'format_id': fid,
202 'url': furl,
203 })
204 self._sort_formats(formats)
205 info = {
206 'formats': formats,
207 }
208 else: # request JSON file
209 info = self._extract_media_info(
210 'http://www.ardmediathek.de/play/media/%s' % video_id, webpage, video_id)
211
212 info.update({
213 'id': video_id,
214 'title': title,
215 'description': description,
216 'thumbnail': thumbnail,
217 })
218
219 return info
220
221
222 class ARDIE(InfoExtractor):
223 _VALID_URL = '(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
224 _TEST = {
225 'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
226 'md5': 'd216c3a86493f9322545e045ddc3eb35',
227 'info_dict': {
228 'display_id': 'die-story-im-ersten-mission-unter-falscher-flagge',
229 'id': '100',
230 'ext': 'mp4',
231 'duration': 2600,
232 'title': 'Die Story im Ersten: Mission unter falscher Flagge',
233 'upload_date': '20140804',
234 'thumbnail': 're:^https?://.*\.jpg$',
235 }
236 }
237
238 def _real_extract(self, url):
239 mobj = re.match(self._VALID_URL, url)
240 display_id = mobj.group('display_id')
241
242 player_url = mobj.group('mainurl') + '~playerXml.xml'
243 doc = self._download_xml(player_url, display_id)
244 video_node = doc.find('./video')
245 upload_date = unified_strdate(xpath_text(
246 video_node, './broadcastDate'))
247 thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
248
249 formats = []
250 for a in video_node.findall('.//asset'):
251 f = {
252 'format_id': a.attrib['type'],
253 'width': int_or_none(a.find('./frameWidth').text),
254 'height': int_or_none(a.find('./frameHeight').text),
255 'vbr': int_or_none(a.find('./bitrateVideo').text),
256 'abr': int_or_none(a.find('./bitrateAudio').text),
257 'vcodec': a.find('./codecVideo').text,
258 'tbr': int_or_none(a.find('./totalBitrate').text),
259 }
260 if a.find('./serverPrefix').text:
261 f['url'] = a.find('./serverPrefix').text
262 f['playpath'] = a.find('./fileName').text
263 else:
264 f['url'] = a.find('./fileName').text
265 formats.append(f)
266 self._sort_formats(formats)
267
268 return {
269 'id': mobj.group('id'),
270 'formats': formats,
271 'display_id': display_id,
272 'title': video_node.find('./title').text,
273 'duration': parse_duration(video_node.find('./duration').text),
274 'upload_date': upload_date,
275 'thumbnail': thumbnail,
276 }
277
278
279 class SportschauIE(ARDMediathekIE):
280 IE_NAME = 'Sportschau'
281 _VALID_URL = r'(?P<baseurl>https?://(?:www\.)?sportschau\.de/(?:[^/]+/)+video(?P<id>[^/#?]+))\.html'
282 _TESTS = [{
283 'url': 'http://www.sportschau.de/tourdefrance/videoseppeltkokainhatnichtsmitklassischemdopingzutun100.html',
284 'info_dict': {
285 'id': 'seppeltkokainhatnichtsmitklassischemdopingzutun100',
286 'ext': 'mp4',
287 'title': 'Seppelt: "Kokain hat nichts mit klassischem Doping zu tun"',
288 'thumbnail': 're:^https?://.*\.jpg$',
289 'description': 'Der ARD-Doping Experte Hajo Seppelt gibt seine Einschätzung zum ersten Dopingfall der diesjährigen Tour de France um den Italiener Luca Paolini ab.',
290 },
291 'params': {
292 # m3u8 download
293 'skip_download': True,
294 },
295 }]
296
297 def _real_extract(self, url):
298 mobj = re.match(self._VALID_URL, url)
299 video_id = mobj.group('id')
300 base_url = mobj.group('baseurl')
301
302 webpage = self._download_webpage(url, video_id)
303 title = get_element_by_attribute('class', 'headline', webpage)
304 description = self._html_search_meta('description', webpage, 'description')
305
306 info = self._extract_media_info(
307 base_url + '-mc_defaultQuality-h.json', webpage, video_id)
308
309 info.update({
310 'title': title,
311 'description': description,
312 })
313
314 return info