]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/mtv.py
302c9bf35bc6bb533c10f35ea11cd500012bba0a
[youtubedl] / youtube_dl / extractor / mtv.py
1 from __future__ import unicode_literals
2
3 import re
4
5 from .common import InfoExtractor
6 from ..compat import (
7 compat_urllib_parse,
8 compat_urllib_request,
9 compat_str,
10 )
11 from ..utils import (
12 ExtractorError,
13 find_xpath_attr,
14 fix_xml_ampersands,
15 HEADRequest,
16 unescapeHTML,
17 url_basename,
18 RegexNotFoundError,
19 )
20
21
22 def _media_xml_tag(tag):
23 return '{http://search.yahoo.com/mrss/}%s' % tag
24
25
26 class MTVServicesInfoExtractor(InfoExtractor):
27 _MOBILE_TEMPLATE = None
28 _LANG = None
29
30 @staticmethod
31 def _id_from_uri(uri):
32 return uri.split(':')[-1]
33
34 # This was originally implemented for ComedyCentral, but it also works here
35 @staticmethod
36 def _transform_rtmp_url(rtmp_video_url):
37 m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
38 if not m:
39 return rtmp_video_url
40 base = 'http://viacommtvstrmfs.fplive.net/'
41 return base + m.group('finalid')
42
43 def _get_feed_url(self, uri):
44 return self._FEED_URL
45
46 def _get_thumbnail_url(self, uri, itemdoc):
47 search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
48 thumb_node = itemdoc.find(search_path)
49 if thumb_node is None:
50 return None
51 else:
52 return thumb_node.attrib['url']
53
54 def _extract_mobile_video_formats(self, mtvn_id):
55 webpage_url = self._MOBILE_TEMPLATE % mtvn_id
56 req = compat_urllib_request.Request(webpage_url)
57 # Otherwise we get a webpage that would execute some javascript
58 req.add_header('User-Agent', 'curl/7')
59 webpage = self._download_webpage(req, mtvn_id,
60 'Downloading mobile page')
61 metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
62 req = HEADRequest(metrics_url)
63 response = self._request_webpage(req, mtvn_id, 'Resolving url')
64 url = response.geturl()
65 # Transform the url to get the best quality:
66 url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
67 return [{'url': url, 'ext': 'mp4'}]
68
69 def _extract_video_formats(self, mdoc, mtvn_id):
70 if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None:
71 if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
72 self.to_screen('The normal version is not available from your '
73 'country, trying with the mobile version')
74 return self._extract_mobile_video_formats(mtvn_id)
75 raise ExtractorError('This video is not available from your country.',
76 expected=True)
77
78 formats = []
79 for rendition in mdoc.findall('.//rendition'):
80 try:
81 _, _, ext = rendition.attrib['type'].partition('/')
82 rtmp_video_url = rendition.find('./src').text
83 if rtmp_video_url.endswith('siteunavail.png'):
84 continue
85 formats.append({
86 'ext': ext,
87 'url': self._transform_rtmp_url(rtmp_video_url),
88 'format_id': rendition.get('bitrate'),
89 'width': int(rendition.get('width')),
90 'height': int(rendition.get('height')),
91 })
92 except (KeyError, TypeError):
93 raise ExtractorError('Invalid rendition field.')
94 self._sort_formats(formats)
95 return formats
96
97 def _extract_subtitles(self, mdoc, mtvn_id):
98 subtitles = {}
99 for transcript in mdoc.findall('.//transcript'):
100 if transcript.get('kind') != 'captions':
101 continue
102 lang = transcript.get('srclang')
103 subtitles[lang] = [{
104 'url': compat_str(typographic.get('src')),
105 'ext': typographic.get('format')
106 } for typographic in transcript.findall('./typographic')]
107 return subtitles
108
109 def _get_video_info(self, itemdoc):
110 uri = itemdoc.find('guid').text
111 video_id = self._id_from_uri(uri)
112 self.report_extraction(video_id)
113 mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
114 # Remove the templates, like &device={device}
115 mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
116 if 'acceptMethods' not in mediagen_url:
117 mediagen_url += '&' if '?' in mediagen_url else '?'
118 mediagen_url += 'acceptMethods=fms'
119
120 mediagen_doc = self._download_xml(mediagen_url, video_id,
121 'Downloading video urls')
122
123 item = mediagen_doc.find('./video/item')
124 if item is not None and item.get('type') == 'text':
125 message = '%s returned error: ' % self.IE_NAME
126 if item.get('code') is not None:
127 message += '%s - ' % item.get('code')
128 message += item.text
129 raise ExtractorError(message, expected=True)
130
131 description_node = itemdoc.find('description')
132 if description_node is not None:
133 description = description_node.text.strip()
134 else:
135 description = None
136
137 title_el = None
138 if title_el is None:
139 title_el = find_xpath_attr(
140 itemdoc, './/{http://search.yahoo.com/mrss/}category',
141 'scheme', 'urn:mtvn:video_title')
142 if title_el is None:
143 title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
144 if title_el is None:
145 title_el = itemdoc.find('.//title') or itemdoc.find('./title')
146 if title_el.text is None:
147 title_el = None
148
149 title = title_el.text
150 if title is None:
151 raise ExtractorError('Could not find video title')
152 title = title.strip()
153
154 # This a short id that's used in the webpage urls
155 mtvn_id = None
156 mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
157 'scheme', 'urn:mtvn:id')
158 if mtvn_id_node is not None:
159 mtvn_id = mtvn_id_node.text
160
161 return {
162 'title': title,
163 'formats': self._extract_video_formats(mediagen_doc, mtvn_id),
164 'subtitles': self._extract_subtitles(mediagen_doc, mtvn_id),
165 'id': video_id,
166 'thumbnail': self._get_thumbnail_url(uri, itemdoc),
167 'description': description,
168 }
169
170 def _get_videos_info(self, uri):
171 video_id = self._id_from_uri(uri)
172 feed_url = self._get_feed_url(uri)
173 data = compat_urllib_parse.urlencode({'uri': uri})
174 info_url = feed_url + '?'
175 if self._LANG:
176 info_url += 'lang=%s&' % self._LANG
177 info_url += data
178 return self._get_videos_info_from_url(info_url, video_id)
179
180 def _get_videos_info_from_url(self, url, video_id):
181 idoc = self._download_xml(
182 url, video_id,
183 'Downloading info', transform_source=fix_xml_ampersands)
184 return self.playlist_result(
185 [self._get_video_info(item) for item in idoc.findall('.//item')])
186
187 def _real_extract(self, url):
188 title = url_basename(url)
189 webpage = self._download_webpage(url, title)
190 try:
191 # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
192 # or http://media.mtvnservices.com/{mgid}
193 og_url = self._og_search_video_url(webpage)
194 mgid = url_basename(og_url)
195 if mgid.endswith('.swf'):
196 mgid = mgid[:-4]
197 except RegexNotFoundError:
198 mgid = None
199
200 if mgid is None or ':' not in mgid:
201 mgid = self._search_regex(
202 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
203 webpage, 'mgid', default=None)
204
205 if not mgid:
206 sm4_embed = self._html_search_meta(
207 'sm4:video:embed', webpage, 'sm4 embed', default='')
208 mgid = self._search_regex(
209 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid')
210
211 videos_info = self._get_videos_info(mgid)
212 return videos_info
213
214
215 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
216 IE_NAME = 'mtvservices:embedded'
217 _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
218
219 _TEST = {
220 # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
221 'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
222 'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
223 'info_dict': {
224 'id': '1043906',
225 'ext': 'mp4',
226 'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
227 'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
228 },
229 }
230
231 @staticmethod
232 def _extract_url(webpage):
233 mobj = re.search(
234 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
235 if mobj:
236 return mobj.group('url')
237
238 def _get_feed_url(self, uri):
239 video_id = self._id_from_uri(uri)
240 site_id = uri.replace(video_id, '')
241 config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/'
242 'context4/context5/config.xml'.format(site_id))
243 config_doc = self._download_xml(config_url, video_id)
244 feed_node = config_doc.find('.//feed')
245 feed_url = feed_node.text.strip().split('?')[0]
246 return feed_url
247
248 def _real_extract(self, url):
249 mobj = re.match(self._VALID_URL, url)
250 mgid = mobj.group('mgid')
251 return self._get_videos_info(mgid)
252
253
254 class MTVIE(MTVServicesInfoExtractor):
255 _VALID_URL = r'''(?x)^https?://
256 (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
257 m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
258
259 _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
260
261 _TESTS = [
262 {
263 'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
264 'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
265 'info_dict': {
266 'id': '853555',
267 'ext': 'mp4',
268 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
269 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
270 },
271 },
272 ]
273
274 def _get_thumbnail_url(self, uri, itemdoc):
275 return 'http://mtv.mtvnimages.com/uri/' + uri
276
277 def _real_extract(self, url):
278 mobj = re.match(self._VALID_URL, url)
279 video_id = mobj.group('videoid')
280 uri = mobj.groupdict().get('mgid')
281 if uri is None:
282 webpage = self._download_webpage(url, video_id)
283
284 # Some videos come from Vevo.com
285 m_vevo = re.search(
286 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
287 if m_vevo:
288 vevo_id = m_vevo.group(1)
289 self.to_screen('Vevo video detected: %s' % vevo_id)
290 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
291
292 uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
293 return self._get_videos_info(uri)
294
295
296 class MTVIggyIE(MTVServicesInfoExtractor):
297 IE_NAME = 'mtviggy.com'
298 _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
299 _TEST = {
300 'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
301 'info_dict': {
302 'id': '984696',
303 'ext': 'mp4',
304 'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
305 }
306 }
307 _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'
308
309
310 class MTVDEIE(MTVServicesInfoExtractor):
311 IE_NAME = 'mtv.de'
312 _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
313 _TESTS = [{
314 'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
315 'info_dict': {
316 'id': 'music_video-a50bc5f0b3aa4b3190aa',
317 'ext': 'mp4',
318 'title': 'MusicVideo_cro-traum',
319 'description': 'Cro - Traum',
320 },
321 'params': {
322 # rtmp download
323 'skip_download': True,
324 },
325 }, {
326 # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
327 'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
328 'info_dict': {
329 'id': 'local_playlist-f5ae778b9832cc837189',
330 'ext': 'mp4',
331 'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
332 },
333 'params': {
334 # rtmp download
335 'skip_download': True,
336 },
337 }, {
338 # single video in pagePlaylist with different id
339 'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
340 'info_dict': {
341 'id': 'local_playlist-4e760566473c4c8c5344',
342 'ext': 'mp4',
343 'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
344 'description': 'MTV Movies Supercut',
345 },
346 'params': {
347 # rtmp download
348 'skip_download': True,
349 },
350 }]
351
352 def _real_extract(self, url):
353 video_id = self._match_id(url)
354
355 webpage = self._download_webpage(url, video_id)
356
357 playlist = self._parse_json(
358 self._search_regex(
359 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
360 video_id)
361
362 # news pages contain single video in playlist with different id
363 if len(playlist) == 1:
364 return self._get_videos_info_from_url(playlist[0]['mrss'], video_id)
365
366 for item in playlist:
367 item_id = item.get('id')
368 if item_id and compat_str(item_id) == video_id:
369 return self._get_videos_info_from_url(item['mrss'], video_id)