Merge branch 'master' of github.com:rg3/youtube-dl into weibo
[ytdl] / 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 ..compat import compat_str
9 from ..utils import (
10     determine_ext,
11     ExtractorError,
12     qualities,
13     int_or_none,
14     parse_duration,
15     unified_strdate,
16     xpath_text,
17     update_url_query,
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)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
25
26     _TESTS = [{
27         'url': 'http://www.ardmediathek.de/tv/Dokumentation-und-Reportage/Ich-liebe-das-Leben-trotzdem/rbb-Fernsehen/Video?documentId=29582122&bcastId=3822114',
28         'info_dict': {
29             'id': '29582122',
30             'ext': 'mp4',
31             'title': 'Ich liebe das Leben trotzdem',
32             'description': 'md5:45e4c225c72b27993314b31a84a5261c',
33             'duration': 4557,
34         },
35         'params': {
36             # m3u8 download
37             'skip_download': True,
38         },
39         'skip': 'HTTP Error 404: Not Found',
40     }, {
41         'url': 'http://www.ardmediathek.de/tv/Tatort/Tatort-Scheinwelten-H%C3%B6rfassung-Video/Das-Erste/Video?documentId=29522730&bcastId=602916',
42         'md5': 'f4d98b10759ac06c0072bbcd1f0b9e3e',
43         'info_dict': {
44             'id': '29522730',
45             'ext': 'mp4',
46             'title': 'Tatort: Scheinwelten - Hörfassung (Video tgl. ab 20 Uhr)',
47             'description': 'md5:196392e79876d0ac94c94e8cdb2875f1',
48             'duration': 5252,
49         },
50         'skip': 'HTTP Error 404: Not Found',
51     }, {
52         # audio
53         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
54         'md5': '219d94d8980b4f538c7fcb0865eb7f2c',
55         'info_dict': {
56             'id': '28488308',
57             'ext': 'mp3',
58             'title': 'Tod eines Fußballers',
59             'description': 'md5:f6e39f3461f0e1f54bfa48c8875c86ef',
60             'duration': 3240,
61         },
62         'skip': 'HTTP Error 404: Not Found',
63     }, {
64         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
65         'only_matching': True,
66     }, {
67         # audio
68         'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
69         'md5': '4e8f00631aac0395fee17368ac0e9867',
70         'info_dict': {
71             'id': '30796318',
72             'ext': 'mp3',
73             'title': 'Vor dem Fest',
74             'description': 'md5:c0c1c8048514deaed2a73b3a60eecacb',
75             'duration': 3287,
76         },
77         'skip': 'Video is no longer available',
78     }]
79
80     def _extract_media_info(self, media_info_url, webpage, video_id):
81         media_info = self._download_json(
82             media_info_url, video_id, 'Downloading media JSON')
83
84         formats = self._extract_formats(media_info, video_id)
85
86         if not formats:
87             if '"fsk"' in webpage:
88                 raise ExtractorError(
89                     'This video is only available after 20:00', expected=True)
90             elif media_info.get('_geoblocked'):
91                 raise ExtractorError('This video is not available due to geo restriction', expected=True)
92
93         self._sort_formats(formats)
94
95         duration = int_or_none(media_info.get('_duration'))
96         thumbnail = media_info.get('_previewImage')
97         is_live = media_info.get('_isLive') is True
98
99         subtitles = {}
100         subtitle_url = media_info.get('_subtitleUrl')
101         if subtitle_url:
102             subtitles['de'] = [{
103                 'ext': 'ttml',
104                 'url': subtitle_url,
105             }]
106
107         return {
108             'id': video_id,
109             'duration': duration,
110             'thumbnail': thumbnail,
111             'is_live': is_live,
112             'formats': formats,
113             'subtitles': subtitles,
114         }
115
116     def _extract_formats(self, media_info, video_id):
117         type_ = media_info.get('_type')
118         media_array = media_info.get('_mediaArray', [])
119         formats = []
120         for num, media in enumerate(media_array):
121             for stream in media.get('_mediaStreamArray', []):
122                 stream_urls = stream.get('_stream')
123                 if not stream_urls:
124                     continue
125                 if not isinstance(stream_urls, list):
126                     stream_urls = [stream_urls]
127                 quality = stream.get('_quality')
128                 server = stream.get('_server')
129                 for stream_url in stream_urls:
130                     if not isinstance(stream_url, compat_str) or '//' not in stream_url:
131                         continue
132                     ext = determine_ext(stream_url)
133                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
134                         continue
135                     if ext == 'f4m':
136                         formats.extend(self._extract_f4m_formats(
137                             update_url_query(stream_url, {
138                                 'hdcore': '3.1.1',
139                                 'plugin': 'aasp-3.1.1.69.124'
140                             }),
141                             video_id, f4m_id='hds', fatal=False))
142                     elif ext == 'm3u8':
143                         formats.extend(self._extract_m3u8_formats(
144                             stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
145                     else:
146                         if server and server.startswith('rtmp'):
147                             f = {
148                                 'url': server,
149                                 'play_path': stream_url,
150                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
151                             }
152                         else:
153                             f = {
154                                 'url': stream_url,
155                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
156                             }
157                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
158                         if m:
159                             f.update({
160                                 'width': int(m.group('width')),
161                                 'height': int(m.group('height')),
162                             })
163                         if type_ == 'audio':
164                             f['vcodec'] = 'none'
165                         formats.append(f)
166         return formats
167
168     def _real_extract(self, url):
169         # determine video id from url
170         m = re.match(self._VALID_URL, url)
171
172         document_id = None
173
174         numid = re.search(r'documentId=([0-9]+)', url)
175         if numid:
176             document_id = video_id = numid.group(1)
177         else:
178             video_id = m.group('video_id')
179
180         webpage = self._download_webpage(url, video_id)
181
182         ERRORS = (
183             ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
184             ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
185              'Video %s is no longer available'),
186         )
187
188         for pattern, message in ERRORS:
189             if pattern in webpage:
190                 raise ExtractorError(message % video_id, expected=True)
191
192         if re.search(r'[\?&]rss($|[=&])', url):
193             doc = compat_etree_fromstring(webpage.encode('utf-8'))
194             if doc.tag == 'rss':
195                 return GenericIE()._extract_rss(url, video_id, doc)
196
197         title = self._html_search_regex(
198             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
199              r'<meta name="dcterms\.title" content="(.*?)"/>',
200              r'<h4 class="headline">(.*?)</h4>'],
201             webpage, 'title')
202         description = self._html_search_meta(
203             'dcterms.abstract', webpage, 'description', default=None)
204         if description is None:
205             description = self._html_search_meta(
206                 'description', webpage, 'meta description')
207
208         # Thumbnail is sometimes not present.
209         # It is in the mobile version, but that seems to use a different URL
210         # structure altogether.
211         thumbnail = self._og_search_thumbnail(webpage, default=None)
212
213         media_streams = re.findall(r'''(?x)
214             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
215             "([^"]+)"''', webpage)
216
217         if media_streams:
218             QUALITIES = qualities(['lo', 'hi', 'hq'])
219             formats = []
220             for furl in set(media_streams):
221                 if furl.endswith('.f4m'):
222                     fid = 'f4m'
223                 else:
224                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
225                     fid = fid_m.group(1) if fid_m else None
226                 formats.append({
227                     'quality': QUALITIES(fid),
228                     'format_id': fid,
229                     'url': furl,
230                 })
231             self._sort_formats(formats)
232             info = {
233                 'formats': formats,
234             }
235         else:  # request JSON file
236             if not document_id:
237                 video_id = self._search_regex(
238                     r'/play/(?:config|media)/(\d+)', webpage, 'media id')
239             info = self._extract_media_info(
240                 'http://www.ardmediathek.de/play/media/%s' % video_id,
241                 webpage, video_id)
242
243         info.update({
244             'id': video_id,
245             'title': self._live_title(title) if info.get('is_live') else title,
246             'description': description,
247             'thumbnail': thumbnail,
248         })
249
250         return info
251
252
253 class ARDIE(InfoExtractor):
254     _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
255     _TEST = {
256         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
257         'md5': 'd216c3a86493f9322545e045ddc3eb35',
258         'info_dict': {
259             'display_id': 'die-story-im-ersten-mission-unter-falscher-flagge',
260             'id': '100',
261             'ext': 'mp4',
262             'duration': 2600,
263             'title': 'Die Story im Ersten: Mission unter falscher Flagge',
264             'upload_date': '20140804',
265             'thumbnail': r're:^https?://.*\.jpg$',
266         },
267         'skip': 'HTTP Error 404: Not Found',
268     }
269
270     def _real_extract(self, url):
271         mobj = re.match(self._VALID_URL, url)
272         display_id = mobj.group('display_id')
273
274         player_url = mobj.group('mainurl') + '~playerXml.xml'
275         doc = self._download_xml(player_url, display_id)
276         video_node = doc.find('./video')
277         upload_date = unified_strdate(xpath_text(
278             video_node, './broadcastDate'))
279         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
280
281         formats = []
282         for a in video_node.findall('.//asset'):
283             f = {
284                 'format_id': a.attrib['type'],
285                 'width': int_or_none(a.find('./frameWidth').text),
286                 'height': int_or_none(a.find('./frameHeight').text),
287                 'vbr': int_or_none(a.find('./bitrateVideo').text),
288                 'abr': int_or_none(a.find('./bitrateAudio').text),
289                 'vcodec': a.find('./codecVideo').text,
290                 'tbr': int_or_none(a.find('./totalBitrate').text),
291             }
292             if a.find('./serverPrefix').text:
293                 f['url'] = a.find('./serverPrefix').text
294                 f['playpath'] = a.find('./fileName').text
295             else:
296                 f['url'] = a.find('./fileName').text
297             formats.append(f)
298         self._sort_formats(formats)
299
300         return {
301             'id': mobj.group('id'),
302             'formats': formats,
303             'display_id': display_id,
304             'title': video_node.find('./title').text,
305             'duration': parse_duration(video_node.find('./duration').text),
306             'upload_date': upload_date,
307             'thumbnail': thumbnail,
308         }