youtube_dl/extractor/cbc.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_str,
  10     compat_HTTPError,
  11 )
  12 from ..utils import (
  13     js_to_json,
  14     smuggle_url,
  15     try_get,
  16     xpath_text,
  17     xpath_element,
  18     xpath_with_ns,
  19     find_xpath_attr,
  20     parse_duration,
  21     parse_iso8601,
  22     parse_age_limit,
  23     int_or_none,
  24     ExtractorError,
  25 )
  26
  27
  28 class CBCIE(InfoExtractor):
  29     IE_NAME = 'cbc.ca'
  30     _VALID_URL = r'https?://(?:www\.)?cbc\.ca/(?!player/)(?:[^/]+/)+(?P<id>[^/?#]+)'
  31     _TESTS = [{
  32         # with mediaId
  33         'url': 'http://www.cbc.ca/22minutes/videos/clips-season-23/don-cherry-play-offs',
  34         'md5': '97e24d09672fc4cf56256d6faa6c25bc',
  35         'info_dict': {
  36             'id': '2682904050',
  37             'ext': 'mp4',
  38             'title': 'Don Cherry – All-Stars',
  39             'description': 'Don Cherry has a bee in his bonnet about AHL player John Scott because that guy’s got heart.',
  40             'timestamp': 1454463000,
  41             'upload_date': '20160203',
  42             'uploader': 'CBCC-NEW',
  43         },
  44         'skip': 'Geo-restricted to Canada',
  45     }, {
  46         # with clipId, feed available via tpfeed.cbc.ca and feed.theplatform.com
  47         'url': 'http://www.cbc.ca/22minutes/videos/22-minutes-update/22-minutes-update-episode-4',
  48         'md5': '162adfa070274b144f4fdc3c3b8207db',
  49         'info_dict': {
  50             'id': '2414435309',
  51             'ext': 'mp4',
  52             'title': '22 Minutes Update: What Not To Wear Quebec',
  53             'description': "This week's latest Canadian top political story is What Not To Wear Quebec.",
  54             'upload_date': '20131025',
  55             'uploader': 'CBCC-NEW',
  56             'timestamp': 1382717907,
  57         },
  58     }, {
  59         # with clipId, feed only available via tpfeed.cbc.ca
  60         'url': 'http://www.cbc.ca/archives/entry/1978-robin-williams-freestyles-on-90-minutes-live',
  61         'md5': '0274a90b51a9b4971fe005c63f592f12',
  62         'info_dict': {
  63             'id': '2487345465',
  64             'ext': 'mp4',
  65             'title': 'Robin Williams freestyles on 90 Minutes Live',
  66             'description': 'Wacky American comedian Robin Williams shows off his infamous "freestyle" comedic talents while being interviewed on CBC\'s 90 Minutes Live.',
  67             'upload_date': '19780210',
  68             'uploader': 'CBCC-NEW',
  69             'timestamp': 255977160,
  70         },
  71     }, {
  72         # multiple iframes
  73         'url': 'http://www.cbc.ca/natureofthings/blog/birds-eye-view-from-vancouvers-burrard-street-bridge-how-we-got-the-shot',
  74         'playlist': [{
  75             'md5': '377572d0b49c4ce0c9ad77470e0b96b4',
  76             'info_dict': {
  77                 'id': '2680832926',
  78                 'ext': 'mp4',
  79                 'title': 'An Eagle\'s-Eye View Off Burrard Bridge',
  80                 'description': 'Hercules the eagle flies from Vancouver\'s Burrard Bridge down to a nearby park with a mini-camera strapped to his back.',
  81                 'upload_date': '20160201',
  82                 'timestamp': 1454342820,
  83                 'uploader': 'CBCC-NEW',
  84             },
  85         }, {
  86             'md5': '415a0e3f586113894174dfb31aa5bb1a',
  87             'info_dict': {
  88                 'id': '2658915080',
  89                 'ext': 'mp4',
  90                 'title': 'Fly like an eagle!',
  91                 'description': 'Eagle equipped with a mini camera flies from the world\'s tallest tower',
  92                 'upload_date': '20150315',
  93                 'timestamp': 1426443984,
  94                 'uploader': 'CBCC-NEW',
  95             },
  96         }],
  97         'skip': 'Geo-restricted to Canada',
  98     }, {
  99         # multiple CBC.APP.Caffeine.initInstance(...)
 100         'url': 'http://www.cbc.ca/news/canada/calgary/dog-indoor-exercise-winter-1.3928238',
 101         'info_dict': {
 102             'title': 'Keep Rover active during the deep freeze with doggie pushups and other fun indoor tasks',
 103             'id': 'dog-indoor-exercise-winter-1.3928238',
 104             'description': 'md5:c18552e41726ee95bd75210d1ca9194c',
 105         },
 106         'playlist_mincount': 6,
 107     }]
 108
 109     @classmethod
 110     def suitable(cls, url):
 111         return False if CBCPlayerIE.suitable(url) else super(CBCIE, cls).suitable(url)
 112
 113     def _extract_player_init(self, player_init, display_id):
 114         player_info = self._parse_json(player_init, display_id, js_to_json)
 115         media_id = player_info.get('mediaId')
 116         if not media_id:
 117             clip_id = player_info['clipId']
 118             feed = self._download_json(
 119                 'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={:mpsReleases}{%s}' % clip_id,
 120                 clip_id, fatal=False)
 121             if feed:
 122                 media_id = try_get(feed, lambda x: x['entries'][0]['guid'], compat_str)
 123             if not media_id:
 124                 media_id = self._download_json(
 125                     'http://feed.theplatform.com/f/h9dtGB/punlNGjMlc1F?fields=id&byContent=byReleases%3DbyId%253D' + clip_id,
 126                     clip_id)['entries'][0]['id'].split('/')[-1]
 127         return self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
 128
 129     def _real_extract(self, url):
 130         display_id = self._match_id(url)
 131         webpage = self._download_webpage(url, display_id)
 132         entries = [
 133             self._extract_player_init(player_init, display_id)
 134             for player_init in re.findall(r'CBC\.APP\.Caffeine\.initInstance\(({.+?})\);', webpage)]
 135         entries.extend([
 136             self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
 137             for media_id in re.findall(r'<iframe[^>]+src="[^"]+?mediaId=(\d+)"', webpage)])
 138         return self.playlist_result(
 139             entries, display_id,
 140             self._og_search_title(webpage, fatal=False),
 141             self._og_search_description(webpage))
 142
 143
 144 class CBCPlayerIE(InfoExtractor):
 145     IE_NAME = 'cbc.ca:player'
 146     _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/|i/caffeine/syndicate/\?mediaId=))(?P<id>\d+)'
 147     _TESTS = [{
 148         'url': 'http://www.cbc.ca/player/play/2683190193',
 149         'md5': '64d25f841ddf4ddb28a235338af32e2c',
 150         'info_dict': {
 151             'id': '2683190193',
 152             'ext': 'mp4',
 153             'title': 'Gerry Runs a Sweat Shop',
 154             'description': 'md5:b457e1c01e8ff408d9d801c1c2cd29b0',
 155             'timestamp': 1455071400,
 156             'upload_date': '20160210',
 157             'uploader': 'CBCC-NEW',
 158         },
 159         'skip': 'Geo-restricted to Canada',
 160     }, {
 161         # Redirected from http://www.cbc.ca/player/AudioMobile/All%20in%20a%20Weekend%20Montreal/ID/2657632011/
 162         'url': 'http://www.cbc.ca/player/play/2657631896',
 163         'md5': 'e5e708c34ae6fca156aafe17c43e8b75',
 164         'info_dict': {
 165             'id': '2657631896',
 166             'ext': 'mp3',
 167             'title': 'CBC Montreal is organizing its first ever community hackathon!',
 168             'description': 'The modern technology we tend to depend on so heavily, is never without it\'s share of hiccups and headaches. Next weekend - CBC Montreal will be getting members of the public for its first Hackathon.',
 169             'timestamp': 1425704400,
 170             'upload_date': '20150307',
 171             'uploader': 'CBCC-NEW',
 172         },
 173     }, {
 174         'url': 'http://www.cbc.ca/player/play/2164402062',
 175         'md5': '33fcd8f6719b9dd60a5e73adcb83b9f6',
 176         'info_dict': {
 177             'id': '2164402062',
 178             'ext': 'mp4',
 179             'title': 'Cancer survivor four times over',
 180             'description': 'Tim Mayer has beaten three different forms of cancer four times in five years.',
 181             'timestamp': 1320410746,
 182             'upload_date': '20111104',
 183             'uploader': 'CBCC-NEW',
 184         },
 185     }]
 186
 187     def _real_extract(self, url):
 188         video_id = self._match_id(url)
 189         return {
 190             '_type': 'url_transparent',
 191             'ie_key': 'ThePlatform',
 192             'url': smuggle_url(
 193                 'http://link.theplatform.com/s/ExhSPC/media/guid/2655402169/%s?mbr=true&formats=MPEG4,FLV,MP3' % video_id, {
 194                     'force_smil_url': True
 195                 }),
 196             'id': video_id,
 197         }
 198
 199
 200 class CBCWatchBaseIE(InfoExtractor):
 201     _device_id = None
 202     _device_token = None
 203     _API_BASE_URL = 'https://api-cbc.cloud.clearleap.com/cloffice/client/'
 204     _NS_MAP = {
 205         'media': 'http://search.yahoo.com/mrss/',
 206         'clearleap': 'http://www.clearleap.com/namespace/clearleap/1.0/',
 207     }
 208     _GEO_COUNTRIES = ['CA']
 209
 210     def _call_api(self, path, video_id):
 211         url = path if path.startswith('http') else self._API_BASE_URL + path
 212         for _ in range(2):
 213             try:
 214                 result = self._download_xml(url, video_id, headers={
 215                     'X-Clearleap-DeviceId': self._device_id,
 216                     'X-Clearleap-DeviceToken': self._device_token,
 217                 })
 218             except ExtractorError as e:
 219                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
 220                     # Device token has expired, re-acquiring device token
 221                     self._register_device()
 222                     continue
 223                 raise
 224         error_message = xpath_text(result, 'userMessage') or xpath_text(result, 'systemMessage')
 225         if error_message:
 226             raise ExtractorError('%s said: %s' % (self.IE_NAME, error_message))
 227         return result
 228
 229     def _real_initialize(self):
 230         if self._valid_device_token():
 231             return
 232         device = self._downloader.cache.load('cbcwatch', 'device') or {}
 233         self._device_id, self._device_token = device.get('id'), device.get('token')
 234         if self._valid_device_token():
 235             return
 236         self._register_device()
 237
 238     def _valid_device_token(self):
 239         return self._device_id and self._device_token
 240
 241     def _register_device(self):
 242         self._device_id = self._device_token = None
 243         result = self._download_xml(
 244             self._API_BASE_URL + 'device/register',
 245             None, 'Acquiring device token',
 246             data=b'<device><type>web</type></device>')
 247         self._device_id = xpath_text(result, 'deviceId', fatal=True)
 248         self._device_token = xpath_text(result, 'deviceToken', fatal=True)
 249         self._downloader.cache.store(
 250             'cbcwatch', 'device', {
 251                 'id': self._device_id,
 252                 'token': self._device_token,
 253             })
 254
 255     def _parse_rss_feed(self, rss):
 256         channel = xpath_element(rss, 'channel', fatal=True)
 257
 258         def _add_ns(path):
 259             return xpath_with_ns(path, self._NS_MAP)
 260
 261         entries = []
 262         for item in channel.findall('item'):
 263             guid = xpath_text(item, 'guid', fatal=True)
 264             title = xpath_text(item, 'title', fatal=True)
 265
 266             media_group = xpath_element(item, _add_ns('media:group'), fatal=True)
 267             content = xpath_element(media_group, _add_ns('media:content'), fatal=True)
 268             content_url = content.attrib['url']
 269
 270             thumbnails = []
 271             for thumbnail in media_group.findall(_add_ns('media:thumbnail')):
 272                 thumbnail_url = thumbnail.get('url')
 273                 if not thumbnail_url:
 274                     continue
 275                 thumbnails.append({
 276                     'id': thumbnail.get('profile'),
 277                     'url': thumbnail_url,
 278                     'width': int_or_none(thumbnail.get('width')),
 279                     'height': int_or_none(thumbnail.get('height')),
 280                 })
 281
 282             timestamp = None
 283             release_date = find_xpath_attr(
 284                 item, _add_ns('media:credit'), 'role', 'releaseDate')
 285             if release_date is not None:
 286                 timestamp = parse_iso8601(release_date.text)
 287
 288             entries.append({
 289                 '_type': 'url_transparent',
 290                 'url': content_url,
 291                 'id': guid,
 292                 'title': title,
 293                 'description': xpath_text(item, 'description'),
 294                 'timestamp': timestamp,
 295                 'duration': int_or_none(content.get('duration')),
 296                 'age_limit': parse_age_limit(xpath_text(item, _add_ns('media:rating'))),
 297                 'episode': xpath_text(item, _add_ns('clearleap:episode')),
 298                 'episode_number': int_or_none(xpath_text(item, _add_ns('clearleap:episodeInSeason'))),
 299                 'series': xpath_text(item, _add_ns('clearleap:series')),
 300                 'season_number': int_or_none(xpath_text(item, _add_ns('clearleap:season'))),
 301                 'thumbnails': thumbnails,
 302                 'ie_key': 'CBCWatchVideo',
 303             })
 304
 305         return self.playlist_result(
 306             entries, xpath_text(channel, 'guid'),
 307             xpath_text(channel, 'title'),
 308             xpath_text(channel, 'description'))
 309
 310
 311 class CBCWatchVideoIE(CBCWatchBaseIE):
 312     IE_NAME = 'cbc.ca:watch:video'
 313     _VALID_URL = r'https?://api-cbc\.cloud\.clearleap\.com/cloffice/client/web/play/?\?.*?\bcontentId=(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
 314     _TEST = {
 315         # geo-restricted to Canada, bypassable
 316         'url': 'https://api-cbc.cloud.clearleap.com/cloffice/client/web/play/?contentId=3c84472a-1eea-4dee-9267-2655d5055dcf&categoryId=ebc258f5-ee40-4cca-b66b-ba6bd55b7235',
 317         'only_matching': True,
 318     }
 319
 320     def _real_extract(self, url):
 321         video_id = self._match_id(url)
 322         result = self._call_api(url, video_id)
 323
 324         m3u8_url = xpath_text(result, 'url', fatal=True)
 325         formats = self._extract_m3u8_formats(re.sub(r'/([^/]+)/[^/?]+\.m3u8', r'/\1/\1.m3u8', m3u8_url), video_id, 'mp4', fatal=False)
 326         if len(formats) < 2:
 327             formats = self._extract_m3u8_formats(m3u8_url, video_id, 'mp4')
 328         for f in formats:
 329             format_id = f.get('format_id')
 330             if format_id.startswith('AAC'):
 331                 f['acodec'] = 'aac'
 332             elif format_id.startswith('AC3'):
 333                 f['acodec'] = 'ac-3'
 334         self._sort_formats(formats)
 335
 336         info = {
 337             'id': video_id,
 338             'title': video_id,
 339             'formats': formats,
 340         }
 341
 342         rss = xpath_element(result, 'rss')
 343         if rss:
 344             info.update(self._parse_rss_feed(rss)['entries'][0])
 345             del info['url']
 346             del info['_type']
 347             del info['ie_key']
 348         return info
 349
 350
 351 class CBCWatchIE(CBCWatchBaseIE):
 352     IE_NAME = 'cbc.ca:watch'
 353     _VALID_URL = r'https?://watch\.cbc\.ca/(?:[^/]+/)+(?P<id>[0-9a-f-]+)'
 354     _TESTS = [{
 355         # geo-restricted to Canada, bypassable
 356         'url': 'http://watch.cbc.ca/doc-zone/season-6/customer-disservice/38e815a-009e3ab12e4',
 357         'info_dict': {
 358             'id': '9673749a-5e77-484c-8b62-a1092a6b5168',
 359             'ext': 'mp4',
 360             'title': 'Customer (Dis)Service',
 361             'description': 'md5:8bdd6913a0fe03d4b2a17ebe169c7c87',
 362             'upload_date': '20160219',
 363             'timestamp': 1455840000,
 364         },
 365         'params': {
 366             # m3u8 download
 367             'skip_download': True,
 368             'format': 'bestvideo',
 369         },
 370     }, {
 371         # geo-restricted to Canada, bypassable
 372         'url': 'http://watch.cbc.ca/arthur/all/1ed4b385-cd84-49cf-95f0-80f004680057',
 373         'info_dict': {
 374             'id': '1ed4b385-cd84-49cf-95f0-80f004680057',
 375             'title': 'Arthur',
 376             'description': 'Arthur, the sweetest 8-year-old aardvark, and his pals solve all kinds of problems with humour, kindness and teamwork.',
 377         },
 378         'playlist_mincount': 30,
 379     }]
 380
 381     def _real_extract(self, url):
 382         video_id = self._match_id(url)
 383         rss = self._call_api('web/browse/' + video_id, video_id)
 384         return self._parse_rss_feed(rss)
 385
 386
 387 class CBCOlympicsIE(InfoExtractor):
 388     IE_NAME = 'cbc.ca:olympics'
 389     _VALID_URL = r'https?://olympics\.cbc\.ca/video/[^/]+/(?P<id>[^/?#]+)'
 390     _TESTS = [{
 391         'url': 'https://olympics.cbc.ca/video/whats-on-tv/olympic-morning-featuring-the-opening-ceremony/',
 392         'only_matching': True,
 393     }]
 394
 395     def _real_extract(self, url):
 396         display_id = self._match_id(url)
 397         webpage = self._download_webpage(url, display_id)
 398         video_id = self._hidden_inputs(webpage)['videoId']
 399         video_doc = self._download_xml(
 400             'https://olympics.cbc.ca/videodata/%s.xml' % video_id, video_id)
 401         title = xpath_text(video_doc, 'title', fatal=True)
 402         is_live = xpath_text(video_doc, 'kind') == 'Live'
 403         if is_live:
 404             title = self._live_title(title)
 405
 406         formats = []
 407         for video_source in video_doc.findall('videoSources/videoSource'):
 408             uri = xpath_text(video_source, 'uri')
 409             if not uri:
 410                 continue
 411             tokenize = self._download_json(
 412                 'https://olympics.cbc.ca/api/api-akamai/tokenize',
 413                 video_id, data=json.dumps({
 414                     'VideoSource': uri,
 415                 }).encode(), headers={
 416                     'Content-Type': 'application/json',
 417                     'Referer': url,
 418                     # d3.VideoPlayer._init in https://olympics.cbc.ca/components/script/base.js
 419                     'Cookie': '_dvp=TK:C0ObxjerU',  # AKAMAI CDN cookie
 420                 }, fatal=False)
 421             if not tokenize:
 422                 continue
 423             content_url = tokenize['ContentUrl']
 424             video_source_format = video_source.get('format')
 425             if video_source_format == 'IIS':
 426                 formats.extend(self._extract_ism_formats(
 427                     content_url, video_id, ism_id=video_source_format, fatal=False))
 428             else:
 429                 formats.extend(self._extract_m3u8_formats(
 430                     content_url, video_id, 'mp4',
 431                     'm3u8' if is_live else 'm3u8_native',
 432                     m3u8_id=video_source_format, fatal=False))
 433         self._sort_formats(formats)
 434
 435         return {
 436             'id': video_id,
 437             'display_id': display_id,
 438             'title': title,
 439             'description': xpath_text(video_doc, 'description'),
 440             'thumbnail': xpath_text(video_doc, 'thumbnailUrl'),
 441             'duration': parse_duration(xpath_text(video_doc, 'duration')),
 442             'formats': formats,
 443             'is_live': is_live,
 444         }