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 subtitles = self._parse_smil_subtitles(meta, default_ns)
55 return formats, subtitles
57 def get_metadata(self, path, video_id):
58 info_url = 'http://link.theplatform.com/s/%s?format=preview' % path
59 info = self._download_json(info_url, video_id)
62 captions = info.get('captions')
63 if isinstance(captions, list):
64 for caption in captions:
65 lang, src, mime = caption.get('lang', 'en'), caption.get('src'), caption.get('type')
67 'ext': mimetype2ext(mime),
72 'title': info['title'],
73 'subtitles': subtitles,
74 'description': info['description'],
75 'thumbnail': info['defaultThumbnailUrl'],
76 'duration': int_or_none(info.get('duration'), 1000),
77 'timestamp': int_or_none(info.get('pubDate'), 1000) or None,
78 'uploader': info.get('billingCode'),
82 class ThePlatformIE(ThePlatformBaseIE):
84 (?:https?://(?:link|player)\.theplatform\.com/[sp]/(?P<provider_id>[^/]+)/
85 (?:(?:(?:[^/]+/)+select/)?(?P<media>media/(?:guid/\d+/)?)|(?P<config>(?:[^/\?]+/(?:swf|config)|onsite)/select/))?
86 |theplatform:)(?P<id>[^/\?&]+)'''
89 # from http://www.metacafe.com/watch/cb-e9I_cZgTgIPd/blackberrys_big_bold_z30/
90 'url': 'http://link.theplatform.com/s/dJ5BDC/e9I_cZgTgIPd/meta.smil?format=smil&Tracking=true&mbr=true',
94 'title': 'Blackberry\'s big, bold Z30',
95 'description': 'The Z30 is Blackberry\'s biggest, baddest mobile messaging device yet.',
97 'timestamp': 1383239700,
98 'upload_date': '20131031',
99 'uploader': 'CBSI-NEW',
103 'skip_download': True,
106 # from http://www.cnet.com/videos/tesla-model-s-a-second-step-towards-a-cleaner-motoring-future/
107 'url': 'http://link.theplatform.com/s/kYEXFC/22d_qsQ6MIRT',
109 'id': '22d_qsQ6MIRT',
111 'description': 'md5:ac330c9258c04f9d7512cf26b9595409',
112 'title': 'Tesla Model S: A second step towards a cleaner motoring future',
113 'timestamp': 1426176191,
114 'upload_date': '20150312',
115 'uploader': 'CBSI-NEW',
119 'skip_download': True,
122 'url': 'https://player.theplatform.com/p/D6x-PC/pulse_preview/embed/select/media/yMBg9E8KFxZD',
124 'id': 'yMBg9E8KFxZD',
126 'description': 'md5:644ad9188d655b742f942bf2e06b002d',
127 'title': 'HIGHLIGHTS: USA bag first ever series Cup win',
131 'url': 'http://player.theplatform.com/p/NnzsPC/widget/select/media/4Y0TlYUr_ZT7',
132 'only_matching': True,
134 'url': 'http://player.theplatform.com/p/2E2eJC/nbcNewsOffsite?guid=tdy_or_siri_150701',
135 'md5': 'fb96bb3d85118930a5b055783a3bd992',
137 'id': 'tdy_or_siri_150701',
139 'title': 'iPhone Siri’s sassy response to a math question has people talking',
140 'description': 'md5:a565d1deadd5086f3331d57298ec6333',
142 'thumbnail': 're:^https?://.*\.jpg$',
143 'timestamp': 1435752600,
144 'upload_date': '20150701',
145 'uploader': 'NBCU-NEWS',
148 # From http://www.nbc.com/the-blacklist/video/sir-crispin-crandall/2928790?onid=137781#vc137781=1
149 # geo-restricted (US), HLS encrypted with AES-128
150 'url': 'http://player.theplatform.com/p/NnzsPC/onsite_universal/select/media/guid/2410887629/2928790?fwsitesection=nbc_the_blacklist_video_library&autoPlay=true&carouselID=137781',
151 'only_matching': True,
155 def _extract_urls(cls, webpage):
159 property=(["'])(?:og:video(?::(?:secure_)?url)?|twitter:player)\1\s+
160 content=(["'])(?P<url>https?://player\.theplatform\.com/p/.+?)\2
163 return [m.group('url')]
165 matches = re.findall(
166 r'<(?:iframe|script)[^>]+src=(["\'])((?:https?:)?//player\.theplatform\.com/p/.+?)\1', webpage)
168 return list(zip(*matches))[1]
171 def _sign_url(url, sig_key, sig_secret, life=600, include_qs=False):
172 flags = '10' if include_qs else '00'
173 expiration_date = '%x' % (int(time.time()) + life)
176 return binascii.b2a_hex(str.encode('ascii')).decode('ascii')
178 def hex_to_bytes(hex):
179 return binascii.a2b_hex(hex.encode('ascii'))
181 relative_path = re.match(r'https?://link.theplatform.com/s/([^?]+)', url).group(1)
182 clear_text = hex_to_bytes(flags + expiration_date + str_to_hex(relative_path))
183 checksum = hmac.new(sig_key.encode('ascii'), clear_text, hashlib.sha1).hexdigest()
184 sig = flags + expiration_date + checksum + str_to_hex(sig_secret)
185 return '%s&sig=%s' % (url, sig)
187 def _real_extract(self, url):
188 url, smuggled_data = unsmuggle_url(url, {})
190 mobj = re.match(self._VALID_URL, url)
191 provider_id = mobj.group('provider_id')
192 video_id = mobj.group('id')
195 provider_id = 'dJ5BDC'
197 path = provider_id + '/'
198 if mobj.group('media'):
199 path += mobj.group('media')
202 qs_dict = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
203 if 'guid' in qs_dict:
204 webpage = self._download_webpage(url, video_id)
205 scripts = re.findall(r'<script[^>]+src="([^"]+)"', webpage)
207 # feed id usually locates in the last script.
208 # Seems there's no pattern for the interested script filename, so
210 for script in reversed(scripts):
211 feed_script = self._download_webpage(
212 self._proto_relative_url(script, 'http:'),
213 video_id, 'Downloading feed script')
214 feed_id = self._search_regex(
215 r'defaultFeedId\s*:\s*"([^"]+)"', feed_script,
216 'default feed id', default=None)
217 if feed_id is not None:
220 raise ExtractorError('Unable to find feed id')
221 return self.url_result('http://feed.theplatform.com/f/%s/%s?byGuid=%s' % (
222 provider_id, feed_id, qs_dict['guid'][0]))
224 if smuggled_data.get('force_smil_url', False):
226 # Explicitly specified SMIL (see https://github.com/rg3/youtube-dl/issues/7385)
227 elif '/guid/' in url:
229 source_url = smuggled_data.get('source_url')
231 headers['Referer'] = source_url
232 request = sanitized_Request(url, headers=headers)
233 webpage = self._download_webpage(request, video_id)
234 smil_url = self._search_regex(
235 r'<link[^>]+href=(["\'])(?P<url>.+?)\1[^>]+type=["\']application/smil\+xml',
236 webpage, 'smil url', group='url')
237 path = self._search_regex(
238 r'link\.theplatform\.com/s/((?:[^/?#&]+/)+[^/?#&]+)', smil_url, 'path')
239 smil_url += '?' if '?' not in smil_url else '&' + 'formats=m3u,mpeg4'
240 elif mobj.group('config'):
241 config_url = url + '&form=json'
242 config_url = config_url.replace('swf/', 'config/')
243 config_url = config_url.replace('onsite/', 'onsite/config/')
244 config = self._download_json(config_url, video_id, 'Downloading config')
245 if 'releaseUrl' in config:
246 release_url = config['releaseUrl']
248 release_url = 'http://link.theplatform.com/s/%s?mbr=true' % path
249 smil_url = release_url + '&formats=MPEG4&manifest=f4m'
251 smil_url = 'http://link.theplatform.com/s/%s?mbr=true' % path
253 sig = smuggled_data.get('sig')
255 smil_url = self._sign_url(smil_url, sig['key'], sig['secret'])
257 formats, subtitles = self._extract_theplatform_smil(smil_url, video_id)
258 self._sort_formats(formats)
260 ret = self.get_metadata(path, video_id)
261 combined_subtitles = self._merge_subtitles(ret.get('subtitles', {}), subtitles)
265 'subtitles': combined_subtitles,
271 class ThePlatformFeedIE(ThePlatformBaseIE):
272 _URL_TEMPLATE = '%s//feed.theplatform.com/f/%s/%s?form=json&byGuid=%s'
273 _VALID_URL = r'https?://feed\.theplatform\.com/f/(?P<provider_id>[^/]+)/(?P<feed_id>[^?/]+)\?(?:[^&]+&)*byGuid=(?P<id>[a-zA-Z0-9_]+)'
275 # From http://player.theplatform.com/p/7wvmTC/MSNBCEmbeddedOffSite?guid=n_hardball_5biden_140207
276 'url': 'http://feed.theplatform.com/f/7wvmTC/msnbc_video-p-test?form=json&pretty=true&range=-40&byGuid=n_hardball_5biden_140207',
277 'md5': '6e32495b5073ab414471b615c5ded394',
279 'id': 'n_hardball_5biden_140207',
281 'title': 'The Biden factor: will Joe run in 2016?',
282 'description': 'Could Vice President Joe Biden be preparing a 2016 campaign? Mark Halperin and Sam Stein weigh in.',
283 'thumbnail': 're:^https?://.*\.jpg$',
284 'upload_date': '20140208',
285 'timestamp': 1391824260,
287 'categories': ['MSNBC/Issues/Democrats', 'MSNBC/Issues/Elections/Election 2016'],
288 'uploader': 'NBCU-NEWS',
292 def _real_extract(self, url):
293 mobj = re.match(self._VALID_URL, url)
295 video_id = mobj.group('id')
296 provider_id = mobj.group('provider_id')
297 feed_id = mobj.group('feed_id')
299 real_url = self._URL_TEMPLATE % (self.http_scheme(), provider_id, feed_id, video_id)
300 feed = self._download_json(real_url, video_id)
301 entry = feed['entries'][0]
305 first_video_id = None
307 for item in entry['media$content']:
308 smil_url = item['plfile$url'] + '&mbr=true'
309 cur_video_id = ThePlatformIE._match_id(smil_url)
310 if first_video_id is None:
311 first_video_id = cur_video_id
312 duration = float_or_none(item.get('plfile$duration'))
313 cur_formats, cur_subtitles = self._extract_theplatform_smil(smil_url, video_id, 'Downloading SMIL data for %s' % cur_video_id)
314 formats.extend(cur_formats)
315 subtitles = self._merge_subtitles(subtitles, cur_subtitles)
317 self._sort_formats(formats)
320 'url': thumbnail['plfile$url'],
321 'width': int_or_none(thumbnail.get('plfile$width')),
322 'height': int_or_none(thumbnail.get('plfile$height')),
323 } for thumbnail in entry.get('media$thumbnails', [])]
325 timestamp = int_or_none(entry.get('media$availableDate'), scale=1000)
326 categories = [item['media$name'] for item in entry.get('media$categories', [])]
328 ret = self.get_metadata('%s/%s' % (provider_id, first_video_id), video_id)
329 subtitles = self._merge_subtitles(subtitles, ret['subtitles'])
333 'subtitles': subtitles,
334 'thumbnails': thumbnails,
335 'duration': duration,
336 'timestamp': timestamp,
337 'categories': categories,