1 # -*- coding: utf-8 -*-
2 from __future__ import unicode_literals
11 from .once import OnceIE
12 from ..compat import (
14 compat_urllib_parse_urlparse,
27 default_ns = 'http://www.w3.org/2005/SMIL21/Language'
28 _x = lambda p: xpath_with_ns(p, {'smil': default_ns})
31 class ThePlatformBaseIE(OnceIE):
32 def _extract_theplatform_smil(self, smil_url, video_id, note='Downloading SMIL data'):
33 meta = self._download_xml(smil_url, video_id, note=note, query={'format': 'SMIL'})
34 error_element = find_xpath_attr(meta, _x('.//smil:ref'), 'src')
35 if error_element is not None and error_element.attrib['src'].startswith(
36 'http://link.theplatform.com/s/errorFiles/Unavailable.'):
37 raise ExtractorError(error_element.attrib['abstract'], expected=True)
39 smil_formats = self._parse_smil_formats(
40 meta, smil_url, video_id, namespace=default_ns,
41 # the parameters are from syfy.com, other sites may use others,
42 # they also work for nbc.com
43 f4m_params={'g': 'UXWGVKRWHFSP', 'hdcore': '3.0.3'},
44 transform_rtmp_url=lambda streamer, src: (streamer, 'mp4:' + src))
47 for _format in smil_formats:
48 if OnceIE.suitable(_format['url']):
49 formats.extend(self._extract_once_formats(_format['url']))
51 formats.append(_format)
53 self._sort_formats(formats)
55 subtitles = self._parse_smil_subtitles(meta, default_ns)
57 return formats, subtitles
59 def get_metadata(self, path, video_id):
60 info_url = 'http://link.theplatform.com/s/%s?format=preview' % path
61 info = self._download_json(info_url, video_id)
64 captions = info.get('captions')
65 if isinstance(captions, list):
66 for caption in captions:
67 lang, src, mime = caption.get('lang', 'en'), caption.get('src'), caption.get('type')
69 'ext': mimetype2ext(mime),
74 'title': info['title'],
75 'subtitles': subtitles,
76 'description': info['description'],
77 'thumbnail': info['defaultThumbnailUrl'],
78 'duration': int_or_none(info.get('duration'), 1000),
79 'timestamp': int_or_none(info.get('pubDate'), 1000) or None,
80 'uploader': info.get('billingCode'),
84 class ThePlatformIE(ThePlatformBaseIE):
86 (?:https?://(?:link|player)\.theplatform\.com/[sp]/(?P<provider_id>[^/]+)/
87 (?:(?:(?:[^/]+/)+select/)?(?P<media>media/(?:guid/\d+/)?)|(?P<config>(?:[^/\?]+/(?:swf|config)|onsite)/select/))?
88 |theplatform:)(?P<id>[^/\?&]+)'''
91 # from http://www.metacafe.com/watch/cb-e9I_cZgTgIPd/blackberrys_big_bold_z30/
92 'url': 'http://link.theplatform.com/s/dJ5BDC/e9I_cZgTgIPd/meta.smil?format=smil&Tracking=true&mbr=true',
96 'title': 'Blackberry\'s big, bold Z30',
97 'description': 'The Z30 is Blackberry\'s biggest, baddest mobile messaging device yet.',
99 'timestamp': 1383239700,
100 'upload_date': '20131031',
101 'uploader': 'CBSI-NEW',
105 'skip_download': True,
108 # from http://www.cnet.com/videos/tesla-model-s-a-second-step-towards-a-cleaner-motoring-future/
109 'url': 'http://link.theplatform.com/s/kYEXFC/22d_qsQ6MIRT',
111 'id': '22d_qsQ6MIRT',
113 'description': 'md5:ac330c9258c04f9d7512cf26b9595409',
114 'title': 'Tesla Model S: A second step towards a cleaner motoring future',
115 'timestamp': 1426176191,
116 'upload_date': '20150312',
117 'uploader': 'CBSI-NEW',
121 'skip_download': True,
124 'url': 'https://player.theplatform.com/p/D6x-PC/pulse_preview/embed/select/media/yMBg9E8KFxZD',
126 'id': 'yMBg9E8KFxZD',
128 'description': 'md5:644ad9188d655b742f942bf2e06b002d',
129 'title': 'HIGHLIGHTS: USA bag first ever series Cup win',
133 'url': 'http://player.theplatform.com/p/NnzsPC/widget/select/media/4Y0TlYUr_ZT7',
134 'only_matching': True,
136 'url': 'http://player.theplatform.com/p/2E2eJC/nbcNewsOffsite?guid=tdy_or_siri_150701',
137 'md5': 'fb96bb3d85118930a5b055783a3bd992',
139 'id': 'tdy_or_siri_150701',
141 'title': 'iPhone Siri’s sassy response to a math question has people talking',
142 'description': 'md5:a565d1deadd5086f3331d57298ec6333',
144 'thumbnail': 're:^https?://.*\.jpg$',
145 'timestamp': 1435752600,
146 'upload_date': '20150701',
147 'uploader': 'NBCU-NEWS',
150 # From http://www.nbc.com/the-blacklist/video/sir-crispin-crandall/2928790?onid=137781#vc137781=1
151 # geo-restricted (US), HLS encrypted with AES-128
152 'url': 'http://player.theplatform.com/p/NnzsPC/onsite_universal/select/media/guid/2410887629/2928790?fwsitesection=nbc_the_blacklist_video_library&autoPlay=true&carouselID=137781',
153 'only_matching': True,
157 def _sign_url(url, sig_key, sig_secret, life=600, include_qs=False):
158 flags = '10' if include_qs else '00'
159 expiration_date = '%x' % (int(time.time()) + life)
162 return binascii.b2a_hex(str.encode('ascii')).decode('ascii')
165 return binascii.a2b_hex(hex)
167 relative_path = re.match(r'https?://link.theplatform.com/s/([^?]+)', url).group(1)
168 clear_text = hex_to_str(flags + expiration_date + str_to_hex(relative_path))
169 checksum = hmac.new(sig_key.encode('ascii'), clear_text, hashlib.sha1).hexdigest()
170 sig = flags + expiration_date + checksum + str_to_hex(sig_secret)
171 return '%s&sig=%s' % (url, sig)
173 def _real_extract(self, url):
174 url, smuggled_data = unsmuggle_url(url, {})
176 mobj = re.match(self._VALID_URL, url)
177 provider_id = mobj.group('provider_id')
178 video_id = mobj.group('id')
181 provider_id = 'dJ5BDC'
183 path = provider_id + '/'
184 if mobj.group('media'):
185 path += mobj.group('media')
188 qs_dict = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
189 if 'guid' in qs_dict:
190 webpage = self._download_webpage(url, video_id)
191 scripts = re.findall(r'<script[^>]+src="([^"]+)"', webpage)
193 # feed id usually locates in the last script.
194 # Seems there's no pattern for the interested script filename, so
196 for script in reversed(scripts):
197 feed_script = self._download_webpage(
198 self._proto_relative_url(script, 'http:'),
199 video_id, 'Downloading feed script')
200 feed_id = self._search_regex(
201 r'defaultFeedId\s*:\s*"([^"]+)"', feed_script,
202 'default feed id', default=None)
203 if feed_id is not None:
206 raise ExtractorError('Unable to find feed id')
207 return self.url_result('http://feed.theplatform.com/f/%s/%s?byGuid=%s' % (
208 provider_id, feed_id, qs_dict['guid'][0]))
210 if smuggled_data.get('force_smil_url', False):
212 # Explicitly specified SMIL (see https://github.com/rg3/youtube-dl/issues/7385)
213 elif '/guid/' in url:
215 source_url = smuggled_data.get('source_url')
217 headers['Referer'] = source_url
218 request = sanitized_Request(url, headers=headers)
219 webpage = self._download_webpage(request, video_id)
220 smil_url = self._search_regex(
221 r'<link[^>]+href=(["\'])(?P<url>.+?)\1[^>]+type=["\']application/smil\+xml',
222 webpage, 'smil url', group='url')
223 path = self._search_regex(
224 r'link\.theplatform\.com/s/((?:[^/?#&]+/)+[^/?#&]+)', smil_url, 'path')
225 smil_url += '?' if '?' not in smil_url else '&' + 'formats=m3u,mpeg4'
226 elif mobj.group('config'):
227 config_url = url + '&form=json'
228 config_url = config_url.replace('swf/', 'config/')
229 config_url = config_url.replace('onsite/', 'onsite/config/')
230 config = self._download_json(config_url, video_id, 'Downloading config')
231 if 'releaseUrl' in config:
232 release_url = config['releaseUrl']
234 release_url = 'http://link.theplatform.com/s/%s?mbr=true' % path
235 smil_url = release_url + '&formats=MPEG4&manifest=f4m'
237 smil_url = 'http://link.theplatform.com/s/%s?mbr=true' % path
239 sig = smuggled_data.get('sig')
241 smil_url = self._sign_url(smil_url, sig['key'], sig['secret'])
243 formats, subtitles = self._extract_theplatform_smil(smil_url, video_id)
245 ret = self.get_metadata(path, video_id)
246 combined_subtitles = self._merge_subtitles(ret.get('subtitles', {}), subtitles)
250 'subtitles': combined_subtitles,
256 class ThePlatformFeedIE(ThePlatformBaseIE):
257 _URL_TEMPLATE = '%s//feed.theplatform.com/f/%s/%s?form=json&byGuid=%s'
258 _VALID_URL = r'https?://feed\.theplatform\.com/f/(?P<provider_id>[^/]+)/(?P<feed_id>[^?/]+)\?(?:[^&]+&)*byGuid=(?P<id>[a-zA-Z0-9_]+)'
260 # From http://player.theplatform.com/p/7wvmTC/MSNBCEmbeddedOffSite?guid=n_hardball_5biden_140207
261 'url': 'http://feed.theplatform.com/f/7wvmTC/msnbc_video-p-test?form=json&pretty=true&range=-40&byGuid=n_hardball_5biden_140207',
262 'md5': '6e32495b5073ab414471b615c5ded394',
264 'id': 'n_hardball_5biden_140207',
266 'title': 'The Biden factor: will Joe run in 2016?',
267 'description': 'Could Vice President Joe Biden be preparing a 2016 campaign? Mark Halperin and Sam Stein weigh in.',
268 'thumbnail': 're:^https?://.*\.jpg$',
269 'upload_date': '20140208',
270 'timestamp': 1391824260,
272 'categories': ['MSNBC/Issues/Democrats', 'MSNBC/Issues/Elections/Election 2016'],
276 def _real_extract(self, url):
277 mobj = re.match(self._VALID_URL, url)
279 video_id = mobj.group('id')
280 provider_id = mobj.group('provider_id')
281 feed_id = mobj.group('feed_id')
283 real_url = self._URL_TEMPLATE % (self.http_scheme(), provider_id, feed_id, video_id)
284 feed = self._download_json(real_url, video_id)
285 entry = feed['entries'][0]
289 first_video_id = None
291 for item in entry['media$content']:
292 smil_url = item['plfile$url'] + '&mbr=true'
293 cur_video_id = ThePlatformIE._match_id(smil_url)
294 if first_video_id is None:
295 first_video_id = cur_video_id
296 duration = float_or_none(item.get('plfile$duration'))
297 cur_formats, cur_subtitles = self._extract_theplatform_smil(smil_url, video_id, 'Downloading SMIL data for %s' % cur_video_id)
298 formats.extend(cur_formats)
299 subtitles = self._merge_subtitles(subtitles, cur_subtitles)
301 self._sort_formats(formats)
304 'url': thumbnail['plfile$url'],
305 'width': int_or_none(thumbnail.get('plfile$width')),
306 'height': int_or_none(thumbnail.get('plfile$height')),
307 } for thumbnail in entry.get('media$thumbnails', [])]
309 timestamp = int_or_none(entry.get('media$availableDate'), scale=1000)
310 categories = [item['media$name'] for item in entry.get('media$categories', [])]
312 ret = self.get_metadata('%s/%s' % (provider_id, first_video_id), video_id)
313 subtitles = self._merge_subtitles(subtitles, ret['subtitles'])
317 'subtitles': subtitles,
318 'thumbnails': thumbnails,
319 'duration': duration,
320 'timestamp': timestamp,
321 'categories': categories,