1 # -*- coding: utf-8 -*-
2 from __future__ import unicode_literals
11 from .common import InfoExtractor
12 from ..compat import (
14 compat_urllib_parse_urlparse,
26 default_ns = 'http://www.w3.org/2005/SMIL21/Language'
27 _x = lambda p: xpath_with_ns(p, {'smil': default_ns})
30 class ThePlatformBaseIE(InfoExtractor):
31 def _extract_theplatform_smil_formats(self, smil_url, video_id, note='Downloading SMIL data'):
32 meta = self._download_xml(smil_url, video_id, note=note)
36 for n in meta.findall(_x('.//smil:ref'))
37 if n.attrib.get('title') == 'Geographic Restriction' or n.attrib.get('title') == 'Expired')
41 raise ExtractorError(error_msg, expected=True)
43 formats = self._parse_smil_formats(
44 meta, smil_url, video_id, namespace=default_ns,
45 # the parameters are from syfy.com, other sites may use others,
46 # they also work for nbc.com
47 f4m_params={'g': 'UXWGVKRWHFSP', 'hdcore': '3.0.3'},
48 transform_rtmp_url=lambda streamer, src: (streamer, 'mp4:' + src))
50 for _format in formats:
51 ext = determine_ext(_format['url'])
53 _format['ext'] = 'mp4'
55 self._sort_formats(formats)
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': 'srt' if mime == 'text/srt' else 'ttml',
74 'title': info['title'],
75 'subtitles': subtitles,
76 'description': info['description'],
77 'thumbnail': info['defaultThumbnailUrl'],
78 'duration': int_or_none(info.get('duration'), 1000),
82 class ThePlatformIE(ThePlatformBaseIE):
84 (?:https?://(?:link|player)\.theplatform\.com/[sp]/(?P<provider_id>[^/]+)/
85 (?:(?P<media>(?:[^/]+/)+select/media/)|(?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.',
100 'skip_download': True,
103 # from http://www.cnet.com/videos/tesla-model-s-a-second-step-towards-a-cleaner-motoring-future/
104 'url': 'http://link.theplatform.com/s/kYEXFC/22d_qsQ6MIRT',
106 'id': '22d_qsQ6MIRT',
108 'description': 'md5:ac330c9258c04f9d7512cf26b9595409',
109 'title': 'Tesla Model S: A second step towards a cleaner motoring future',
113 'skip_download': True,
116 'url': 'https://player.theplatform.com/p/D6x-PC/pulse_preview/embed/select/media/yMBg9E8KFxZD',
118 'id': 'yMBg9E8KFxZD',
120 'description': 'md5:644ad9188d655b742f942bf2e06b002d',
121 'title': 'HIGHLIGHTS: USA bag first ever series Cup win',
124 'url': 'http://player.theplatform.com/p/NnzsPC/widget/select/media/4Y0TlYUr_ZT7',
125 'only_matching': True,
127 'url': 'http://player.theplatform.com/p/2E2eJC/nbcNewsOffsite?guid=tdy_or_siri_150701',
128 'md5': '734f3790fb5fc4903da391beeebc4836',
130 'id': 'tdy_or_siri_150701',
132 'title': 'iPhone Siri’s sassy response to a math question has people talking',
133 'description': 'md5:a565d1deadd5086f3331d57298ec6333',
135 'thumbnail': 're:^https?://.*\.jpg$',
136 'timestamp': 1435752600,
137 'upload_date': '20150701',
138 'categories': ['Today/Shows/Orange Room', 'Today/Sections/Money', 'Today/Topics/Tech', "Today/Topics/Editor's picks"],
143 def _sign_url(url, sig_key, sig_secret, life=600, include_qs=False):
144 flags = '10' if include_qs else '00'
145 expiration_date = '%x' % (int(time.time()) + life)
148 return binascii.b2a_hex(str.encode('ascii')).decode('ascii')
151 return binascii.a2b_hex(hex)
153 relative_path = url.split('http://link.theplatform.com/s/')[1].split('?')[0]
154 clear_text = hex_to_str(flags + expiration_date + str_to_hex(relative_path))
155 checksum = hmac.new(sig_key.encode('ascii'), clear_text, hashlib.sha1).hexdigest()
156 sig = flags + expiration_date + checksum + str_to_hex(sig_secret)
157 return '%s&sig=%s' % (url, sig)
159 def _real_extract(self, url):
160 url, smuggled_data = unsmuggle_url(url, {})
162 mobj = re.match(self._VALID_URL, url)
163 provider_id = mobj.group('provider_id')
164 video_id = mobj.group('id')
167 provider_id = 'dJ5BDC'
170 if mobj.group('media'):
172 path += '/' + video_id
174 qs_dict = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
175 if 'guid' in qs_dict:
176 webpage = self._download_webpage(url, video_id)
177 scripts = re.findall(r'<script[^>]+src="([^"]+)"', webpage)
179 # feed id usually locates in the last script.
180 # Seems there's no pattern for the interested script filename, so
182 for script in reversed(scripts):
183 feed_script = self._download_webpage(script, video_id, 'Downloading feed script')
184 feed_id = self._search_regex(r'defaultFeedId\s*:\s*"([^"]+)"', feed_script, 'default feed id', default=None)
185 if feed_id is not None:
188 raise ExtractorError('Unable to find feed id')
189 return self.url_result('http://feed.theplatform.com/f/%s/%s?byGuid=%s' % (
190 provider_id, feed_id, qs_dict['guid'][0]))
192 if smuggled_data.get('force_smil_url', False):
194 elif mobj.group('config'):
195 config_url = url + '&form=json'
196 config_url = config_url.replace('swf/', 'config/')
197 config_url = config_url.replace('onsite/', 'onsite/config/')
198 config = self._download_json(config_url, video_id, 'Downloading config')
199 if 'releaseUrl' in config:
200 release_url = config['releaseUrl']
202 release_url = 'http://link.theplatform.com/s/%s?mbr=true' % path
203 smil_url = release_url + '&format=SMIL&formats=MPEG4&manifest=f4m'
205 smil_url = 'http://link.theplatform.com/s/%s/meta.smil?format=smil&mbr=true' % path
207 sig = smuggled_data.get('sig')
209 smil_url = self._sign_url(smil_url, sig['key'], sig['secret'])
211 formats = self._extract_theplatform_smil_formats(smil_url, video_id)
213 ret = self.get_metadata(path, video_id)
222 class ThePlatformFeedIE(ThePlatformBaseIE):
223 _URL_TEMPLATE = '%s//feed.theplatform.com/f/%s/%s?form=json&byGuid=%s'
224 _VALID_URL = r'https?://feed\.theplatform\.com/f/(?P<provider_id>[^/]+)/(?P<feed_id>[^?/]+)\?(?:[^&]+&)*byGuid=(?P<id>[a-zA-Z0-9_]+)'
226 # From http://player.theplatform.com/p/7wvmTC/MSNBCEmbeddedOffSite?guid=n_hardball_5biden_140207
227 'url': 'http://feed.theplatform.com/f/7wvmTC/msnbc_video-p-test?form=json&pretty=true&range=-40&byGuid=n_hardball_5biden_140207',
228 'md5': '22d2b84f058d3586efcd99e57d59d314',
230 'id': 'n_hardball_5biden_140207',
232 'title': 'The Biden factor: will Joe run in 2016?',
233 'description': 'Could Vice President Joe Biden be preparing a 2016 campaign? Mark Halperin and Sam Stein weigh in.',
234 'thumbnail': 're:^https?://.*\.jpg$',
235 'upload_date': '20140208',
236 'timestamp': 1391824260,
238 'categories': ['MSNBC/Issues/Democrats', 'MSNBC/Issues/Elections/Election 2016'],
242 def _real_extract(self, url):
243 mobj = re.match(self._VALID_URL, url)
245 video_id = mobj.group('id')
246 provider_id = mobj.group('provider_id')
247 feed_id = mobj.group('feed_id')
249 real_url = self._URL_TEMPLATE % (self.http_scheme(), provider_id, feed_id, video_id)
250 feed = self._download_json(real_url, video_id)
251 entry = feed['entries'][0]
254 first_video_id = None
256 for item in entry['media$content']:
257 smil_url = item['plfile$url'] + '&format=SMIL&Tracking=true&Embedded=true&formats=MPEG4,F4M'
258 cur_video_id = url_basename(smil_url)
259 if first_video_id is None:
260 first_video_id = cur_video_id
261 duration = float_or_none(item.get('plfile$duration'))
262 formats.extend(self._extract_theplatform_smil_formats(smil_url, video_id, 'Downloading SMIL data for %s' % cur_video_id))
264 self._sort_formats(formats)
267 'url': thumbnail['plfile$url'],
268 'width': int_or_none(thumbnail.get('plfile$width')),
269 'height': int_or_none(thumbnail.get('plfile$height')),
270 } for thumbnail in entry.get('media$thumbnails', [])]
272 timestamp = int_or_none(entry.get('media$availableDate'), scale=1000)
273 categories = [item['media$name'] for item in entry.get('media$categories', [])]
275 ret = self.get_metadata('%s/%s' % (provider_id, first_video_id), video_id)
279 'thumbnails': thumbnails,
280 'duration': duration,
281 'timestamp': timestamp,
282 'categories': categories,