[mtv] Use compat_xpath and fix FutureWarning
[youtube-dl] / 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_urlencode,
8     compat_str,
9     compat_xpath,
10 )
11 from ..utils import (
12     ExtractorError,
13     find_xpath_attr,
14     fix_xml_ampersands,
15     float_or_none,
16     HEADRequest,
17     sanitized_Request,
18     unescapeHTML,
19     url_basename,
20     RegexNotFoundError,
21     xpath_text,
22 )
23
24
25 def _media_xml_tag(tag):
26     return '{http://search.yahoo.com/mrss/}%s' % tag
27
28
29 class MTVServicesInfoExtractor(InfoExtractor):
30     _MOBILE_TEMPLATE = None
31     _LANG = None
32
33     @staticmethod
34     def _id_from_uri(uri):
35         return uri.split(':')[-1]
36
37     # This was originally implemented for ComedyCentral, but it also works here
38     @staticmethod
39     def _transform_rtmp_url(rtmp_video_url):
40         m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp\..+?/.*)$', rtmp_video_url)
41         if not m:
42             return rtmp_video_url
43         base = 'http://viacommtvstrmfs.fplive.net/'
44         return base + m.group('finalid')
45
46     def _get_feed_url(self, uri):
47         return self._FEED_URL
48
49     def _get_thumbnail_url(self, uri, itemdoc):
50         search_path = '%s/%s' % (_media_xml_tag('group'), _media_xml_tag('thumbnail'))
51         thumb_node = itemdoc.find(search_path)
52         if thumb_node is None:
53             return None
54         else:
55             return thumb_node.attrib['url']
56
57     def _extract_mobile_video_formats(self, mtvn_id):
58         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
59         req = sanitized_Request(webpage_url)
60         # Otherwise we get a webpage that would execute some javascript
61         req.add_header('User-Agent', 'curl/7')
62         webpage = self._download_webpage(req, mtvn_id,
63                                          'Downloading mobile page')
64         metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
65         req = HEADRequest(metrics_url)
66         response = self._request_webpage(req, mtvn_id, 'Resolving url')
67         url = response.geturl()
68         # Transform the url to get the best quality:
69         url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
70         return [{'url': url, 'ext': 'mp4'}]
71
72     def _extract_video_formats(self, mdoc, mtvn_id):
73         if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4|copyright_error\.flv(?:\?geo\b.+?)?)$', mdoc.find('.//src').text) is not None:
74             if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
75                 self.to_screen('The normal version is not available from your '
76                                'country, trying with the mobile version')
77                 return self._extract_mobile_video_formats(mtvn_id)
78             raise ExtractorError('This video is not available from your country.',
79                                  expected=True)
80
81         formats = []
82         for rendition in mdoc.findall('.//rendition'):
83             try:
84                 _, _, ext = rendition.attrib['type'].partition('/')
85                 rtmp_video_url = rendition.find('./src').text
86                 if rtmp_video_url.endswith('siteunavail.png'):
87                     continue
88                 formats.append({
89                     'ext': ext,
90                     'url': self._transform_rtmp_url(rtmp_video_url),
91                     'format_id': rendition.get('bitrate'),
92                     'width': int(rendition.get('width')),
93                     'height': int(rendition.get('height')),
94                 })
95             except (KeyError, TypeError):
96                 raise ExtractorError('Invalid rendition field.')
97         self._sort_formats(formats)
98         return formats
99
100     def _extract_subtitles(self, mdoc, mtvn_id):
101         subtitles = {}
102         for transcript in mdoc.findall('.//transcript'):
103             if transcript.get('kind') != 'captions':
104                 continue
105             lang = transcript.get('srclang')
106             subtitles[lang] = [{
107                 'url': compat_str(typographic.get('src')),
108                 'ext': typographic.get('format')
109             } for typographic in transcript.findall('./typographic')]
110         return subtitles
111
112     def _get_video_info(self, itemdoc):
113         uri = itemdoc.find('guid').text
114         video_id = self._id_from_uri(uri)
115         self.report_extraction(video_id)
116         content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
117         mediagen_url = content_el.attrib['url']
118         # Remove the templates, like &device={device}
119         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
120         if 'acceptMethods' not in mediagen_url:
121             mediagen_url += '&' if '?' in mediagen_url else '?'
122             mediagen_url += 'acceptMethods=fms'
123
124         mediagen_doc = self._download_xml(mediagen_url, video_id,
125                                           'Downloading video urls')
126
127         item = mediagen_doc.find('./video/item')
128         if item is not None and item.get('type') == 'text':
129             message = '%s returned error: ' % self.IE_NAME
130             if item.get('code') is not None:
131                 message += '%s - ' % item.get('code')
132             message += item.text
133             raise ExtractorError(message, expected=True)
134
135         description = xpath_text(itemdoc, 'description')
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(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
144         if title_el is None:
145             title_el = itemdoc.find(compat_xpath('.//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             'duration': float_or_none(content_el.attrib.get('duration')),
169         }
170
171     def _get_feed_query(self, uri):
172         data = {'uri': uri}
173         if self._LANG:
174             data['lang'] = self._LANG
175         return compat_urllib_parse_urlencode(data)
176
177     def _get_videos_info(self, uri):
178         video_id = self._id_from_uri(uri)
179         feed_url = self._get_feed_url(uri)
180         info_url = feed_url + '?' + self._get_feed_query(uri)
181         return self._get_videos_info_from_url(info_url, video_id)
182
183     def _get_videos_info_from_url(self, url, video_id):
184         idoc = self._download_xml(
185             url, video_id,
186             'Downloading info', transform_source=fix_xml_ampersands)
187         return self.playlist_result(
188             [self._get_video_info(item) for item in idoc.findall('.//item')])
189
190     def _extract_mgid(self, webpage):
191         try:
192             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
193             # or http://media.mtvnservices.com/{mgid}
194             og_url = self._og_search_video_url(webpage)
195             mgid = url_basename(og_url)
196             if mgid.endswith('.swf'):
197                 mgid = mgid[:-4]
198         except RegexNotFoundError:
199             mgid = None
200
201         if mgid is None or ':' not in mgid:
202             mgid = self._search_regex(
203                 [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
204                 webpage, 'mgid', default=None)
205
206         if not mgid:
207             sm4_embed = self._html_search_meta(
208                 'sm4:video:embed', webpage, 'sm4 embed', default='')
209             mgid = self._search_regex(
210                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid')
211         return mgid
212
213     def _real_extract(self, url):
214         title = url_basename(url)
215         webpage = self._download_webpage(url, title)
216         mgid = self._extract_mgid(webpage)
217         videos_info = self._get_videos_info(mgid)
218         return videos_info
219
220
221 class MTVServicesEmbeddedIE(MTVServicesInfoExtractor):
222     IE_NAME = 'mtvservices:embedded'
223     _VALID_URL = r'https?://media\.mtvnservices\.com/embed/(?P<mgid>.+?)(\?|/|$)'
224
225     _TEST = {
226         # From http://www.thewrap.com/peter-dinklage-sums-up-game-of-thrones-in-45-seconds-video/
227         'url': 'http://media.mtvnservices.com/embed/mgid:uma:video:mtv.com:1043906/cp~vid%3D1043906%26uri%3Dmgid%3Auma%3Avideo%3Amtv.com%3A1043906',
228         'md5': 'cb349b21a7897164cede95bd7bf3fbb9',
229         'info_dict': {
230             'id': '1043906',
231             'ext': 'mp4',
232             'title': 'Peter Dinklage Sums Up \'Game Of Thrones\' In 45 Seconds',
233             'description': '"Sexy sexy sexy, stabby stabby stabby, beautiful language," says Peter Dinklage as he tries summarizing "Game of Thrones" in under a minute.',
234         },
235     }
236
237     @staticmethod
238     def _extract_url(webpage):
239         mobj = re.search(
240             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//media.mtvnservices.com/embed/.+?)\1', webpage)
241         if mobj:
242             return mobj.group('url')
243
244     def _get_feed_url(self, uri):
245         video_id = self._id_from_uri(uri)
246         site_id = uri.replace(video_id, '')
247         config_url = ('http://media.mtvnservices.com/pmt/e1/players/{0}/'
248                       'context4/context5/config.xml'.format(site_id))
249         config_doc = self._download_xml(config_url, video_id)
250         feed_node = config_doc.find('.//feed')
251         feed_url = feed_node.text.strip().split('?')[0]
252         return feed_url
253
254     def _real_extract(self, url):
255         mobj = re.match(self._VALID_URL, url)
256         mgid = mobj.group('mgid')
257         return self._get_videos_info(mgid)
258
259
260 class MTVIE(MTVServicesInfoExtractor):
261     _VALID_URL = r'''(?x)^https?://
262         (?:(?:www\.)?mtv\.com/videos/.+?/(?P<videoid>[0-9]+)/[^/]+$|
263            m\.mtv\.com/videos/video\.rbml\?.*?id=(?P<mgid>[^&]+))'''
264
265     _FEED_URL = 'http://www.mtv.com/player/embed/AS3/rss/'
266
267     _TESTS = [
268         {
269             'url': 'http://www.mtv.com/videos/misc/853555/ours-vh1-storytellers.jhtml',
270             'md5': '850f3f143316b1e71fa56a4edfd6e0f8',
271             'info_dict': {
272                 'id': '853555',
273                 'ext': 'mp4',
274                 'title': 'Taylor Swift - "Ours (VH1 Storytellers)"',
275                 'description': 'Album: Taylor Swift performs "Ours" for VH1 Storytellers at Harvey Mudd College.',
276             },
277         },
278     ]
279
280     def _get_thumbnail_url(self, uri, itemdoc):
281         return 'http://mtv.mtvnimages.com/uri/' + uri
282
283     def _real_extract(self, url):
284         mobj = re.match(self._VALID_URL, url)
285         video_id = mobj.group('videoid')
286         uri = mobj.groupdict().get('mgid')
287         if uri is None:
288             webpage = self._download_webpage(url, video_id)
289
290             # Some videos come from Vevo.com
291             m_vevo = re.search(
292                 r'(?s)isVevoVideo = true;.*?vevoVideoId = "(.*?)";', webpage)
293             if m_vevo:
294                 vevo_id = m_vevo.group(1)
295                 self.to_screen('Vevo video detected: %s' % vevo_id)
296                 return self.url_result('vevo:%s' % vevo_id, ie='Vevo')
297
298             uri = self._html_search_regex(r'/uri/(.*?)\?', webpage, 'uri')
299         return self._get_videos_info(uri)
300
301
302 class MTVIggyIE(MTVServicesInfoExtractor):
303     IE_NAME = 'mtviggy.com'
304     _VALID_URL = r'https?://www\.mtviggy\.com/videos/.+'
305     _TEST = {
306         'url': 'http://www.mtviggy.com/videos/arcade-fire-behind-the-scenes-at-the-biggest-music-experiment-yet/',
307         'info_dict': {
308             'id': '984696',
309             'ext': 'mp4',
310             'title': 'Arcade Fire: Behind the Scenes at the Biggest Music Experiment Yet',
311         }
312     }
313     _FEED_URL = 'http://all.mtvworldverticals.com/feed-xml/'
314
315
316 class MTVDEIE(MTVServicesInfoExtractor):
317     IE_NAME = 'mtv.de'
318     _VALID_URL = r'https?://(?:www\.)?mtv\.de/(?:artists|shows|news)/(?:[^/]+/)*(?P<id>\d+)-[^/#?]+/*(?:[#?].*)?$'
319     _TESTS = [{
320         'url': 'http://www.mtv.de/artists/10571-cro/videos/61131-traum',
321         'info_dict': {
322             'id': 'music_video-a50bc5f0b3aa4b3190aa',
323             'ext': 'mp4',
324             'title': 'MusicVideo_cro-traum',
325             'description': 'Cro - Traum',
326         },
327         'params': {
328             # rtmp download
329             'skip_download': True,
330         },
331     }, {
332         # mediagen URL without query (e.g. http://videos.mtvnn.com/mediagen/e865da714c166d18d6f80893195fcb97)
333         'url': 'http://www.mtv.de/shows/933-teen-mom-2/staffeln/5353/folgen/63565-enthullungen',
334         'info_dict': {
335             'id': 'local_playlist-f5ae778b9832cc837189',
336             'ext': 'mp4',
337             'title': 'Episode_teen-mom-2_shows_season-5_episode-1_full-episode_part1',
338         },
339         'params': {
340             # rtmp download
341             'skip_download': True,
342         },
343     }, {
344         # single video in pagePlaylist with different id
345         'url': 'http://www.mtv.de/news/77491-mtv-movies-spotlight-pixels-teil-3',
346         'info_dict': {
347             'id': 'local_playlist-4e760566473c4c8c5344',
348             'ext': 'mp4',
349             'title': 'Article_mtv-movies-spotlight-pixels-teil-3_short-clips_part1',
350             'description': 'MTV Movies Supercut',
351         },
352         'params': {
353             # rtmp download
354             'skip_download': True,
355         },
356     }]
357
358     def _real_extract(self, url):
359         video_id = self._match_id(url)
360
361         webpage = self._download_webpage(url, video_id)
362
363         playlist = self._parse_json(
364             self._search_regex(
365                 r'window\.pagePlaylist\s*=\s*(\[.+?\]);\n', webpage, 'page playlist'),
366             video_id)
367
368         # news pages contain single video in playlist with different id
369         if len(playlist) == 1:
370             return self._get_videos_info_from_url(playlist[0]['mrss'], video_id)
371
372         for item in playlist:
373             item_id = item.get('id')
374             if item_id and compat_str(item_id) == video_id:
375                 return self._get_videos_info_from_url(item['mrss'], video_id)