youtube_dl/extractor/cbc.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import compat_str
   8 from ..utils import (
   9     js_to_json,
  10     smuggle_url,
  11     try_get,
  12     xpath_text,
  13     xpath_element,
  14     xpath_with_ns,
  15     find_xpath_attr,
  16     parse_iso8601,
  17     parse_age_limit,
  18     int_or_none,
  19     ExtractorError,
  20 )
  21
  22
  23 class CBCIE(InfoExtractor):
  24     IE_NAME = 'cbc.ca'
  25     _VALID_URL = r'https?://(?:www\.)?cbc\.ca/(?!player/)(?:[^/]+/)+(?P<id>[^/?#]+)'
  26     _TESTS = [{
  27         # with mediaId
  28         'url': 'http://www.cbc.ca/22minutes/videos/clips-season-23/don-cherry-play-offs',
  29         'md5': '97e24d09672fc4cf56256d6faa6c25bc',
  30         'info_dict': {
  31             'id': '2682904050',
  32             'ext': 'mp4',
  33             'title': 'Don Cherry – All-Stars',
  34             'description': 'Don Cherry has a bee in his bonnet about AHL player John Scott because that guy’s got heart.',
  35             'timestamp': 1454463000,
  36             'upload_date': '20160203',
  37             'uploader': 'CBCC-NEW',
  38         },
  39         'skip': 'Geo-restricted to Canada',
  40     }, {
  41         # with clipId, feed available via tpfeed.cbc.ca and feed.theplatform.com
  42         'url': 'http://www.cbc.ca/22minutes/videos/22-minutes-update/22-minutes-update-episode-4',
  43         'md5': '162adfa070274b144f4fdc3c3b8207db',
  44         'info_dict': {
  45             'id': '2414435309',
  46             'ext': 'mp4',
  47             'title': '22 Minutes Update: What Not To Wear Quebec',
  48             'description': "This week's latest Canadian top political story is What Not To Wear Quebec.",
  49             'upload_date': '20131025',
  50             'uploader': 'CBCC-NEW',
  51             'timestamp': 1382717907,
  52         },
  53     }, {
  54         # with clipId, feed only available via tpfeed.cbc.ca
  55         'url': 'http://www.cbc.ca/archives/entry/1978-robin-williams-freestyles-on-90-minutes-live',
  56         'md5': '0274a90b51a9b4971fe005c63f592f12',
  57         'info_dict': {
  58             'id': '2487345465',
  59             'ext': 'mp4',
  60             'title': 'Robin Williams freestyles on 90 Minutes Live',
  61             'description': 'Wacky American comedian Robin Williams shows off his infamous "freestyle" comedic talents while being interviewed on CBC\'s 90 Minutes Live.',
  62             'upload_date': '19780210',
  63             'uploader': 'CBCC-NEW',
  64             'timestamp': 255977160,
  65         },
  66     }, {
  67         # multiple iframes
  68         'url': 'http://www.cbc.ca/natureofthings/blog/birds-eye-view-from-vancouvers-burrard-street-bridge-how-we-got-the-shot',
  69         'playlist': [{
  70             'md5': '377572d0b49c4ce0c9ad77470e0b96b4',
  71             'info_dict': {
  72                 'id': '2680832926',
  73                 'ext': 'mp4',
  74                 'title': 'An Eagle\'s-Eye View Off Burrard Bridge',
  75                 'description': 'Hercules the eagle flies from Vancouver\'s Burrard Bridge down to a nearby park with a mini-camera strapped to his back.',
  76                 'upload_date': '20160201',
  77                 'timestamp': 1454342820,
  78                 'uploader': 'CBCC-NEW',
  79             },
  80         }, {
  81             'md5': '415a0e3f586113894174dfb31aa5bb1a',
  82             'info_dict': {
  83                 'id': '2658915080',
  84                 'ext': 'mp4',
  85                 'title': 'Fly like an eagle!',
  86                 'description': 'Eagle equipped with a mini camera flies from the world\'s tallest tower',
  87                 'upload_date': '20150315',
  88                 'timestamp': 1426443984,
  89                 'uploader': 'CBCC-NEW',
  90             },
  91         }],
  92         'skip': 'Geo-restricted to Canada',
  93     }]
  94
  95     @classmethod
  96     def suitable(cls, url):
  97         return False if CBCPlayerIE.suitable(url) else super(CBCIE, cls).suitable(url)
  98
  99     def _real_extract(self, url):
 100         display_id = self._match_id(url)
 101         webpage = self._download_webpage(url, display_id)
 102         player_init = self._search_regex(
 103             r'CBC\.APP\.Caffeine\.initInstance\(({.+?})\);', webpage, 'player init',
 104             default=None)
 105         if player_init:
 106             player_info = self._parse_json(player_init, display_id, js_to_json)
 107             media_id = player_info.get('mediaId')
 108             if not media_id:
 109                 clip_id = player_info['clipId']
 110                 feed = self._download_json(
 111                     'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={:mpsReleases}{%s}' % clip_id,
 112                     clip_id, fatal=False)
 113                 if feed:
 114                     media_id = try_get(feed, lambda x: x['entries'][0]['guid'], compat_str)
 115                 if not media_id:
 116                     media_id = self._download_json(
 117                         'http://feed.theplatform.com/f/h9dtGB/punlNGjMlc1F?fields=id&byContent=byReleases%3DbyId%253D' + clip_id,
 118                         clip_id)['entries'][0]['id'].split('/')[-1]
 119             return self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
 120         else:
 121             entries = [self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id) for media_id in re.findall(r'<iframe[^>]+src="[^"]+?mediaId=(\d+)"', webpage)]
 122             return self.playlist_result(entries)
 123
 124
 125 class CBCPlayerIE(InfoExtractor):
 126     IE_NAME = 'cbc.ca:player'
 127     _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/|i/caffeine/syndicate/\?mediaId=))(?P<id>\d+)'
 128     _TESTS = [{
 129         'url': 'http://www.cbc.ca/player/play/2683190193',
 130         'md5': '64d25f841ddf4ddb28a235338af32e2c',
 131         'info_dict': {
 132             'id': '2683190193',
 133             'ext': 'mp4',
 134             'title': 'Gerry Runs a Sweat Shop',
 135             'description': 'md5:b457e1c01e8ff408d9d801c1c2cd29b0',
 136             'timestamp': 1455071400,
 137             'upload_date': '20160210',
 138             'uploader': 'CBCC-NEW',
 139         },
 140         'skip': 'Geo-restricted to Canada',
 141     }, {
 142         # Redirected from http://www.cbc.ca/player/AudioMobile/All%20in%20a%20Weekend%20Montreal/ID/2657632011/
 143         'url': 'http://www.cbc.ca/player/play/2657631896',
 144         'md5': 'e5e708c34ae6fca156aafe17c43e8b75',
 145         'info_dict': {
 146             'id': '2657631896',
 147             'ext': 'mp3',
 148             'title': 'CBC Montreal is organizing its first ever community hackathon!',
 149             '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.',
 150             'timestamp': 1425704400,
 151             'upload_date': '20150307',
 152             'uploader': 'CBCC-NEW',
 153         },
 154     }, {
 155         # available only when we add `formats=MPEG4,FLV,MP3` to theplatform url
 156         'url': 'http://www.cbc.ca/player/play/2164402062',
 157         'md5': '17a61eb813539abea40618d6323a7f82',
 158         'info_dict': {
 159             'id': '2164402062',
 160             'ext': 'flv',
 161             'title': 'Cancer survivor four times over',
 162             'description': 'Tim Mayer has beaten three different forms of cancer four times in five years.',
 163             'timestamp': 1320410746,
 164             'upload_date': '20111104',
 165             'uploader': 'CBCC-NEW',
 166         },
 167     }]
 168
 169     def _real_extract(self, url):
 170         video_id = self._match_id(url)
 171         return {
 172             '_type': 'url_transparent',
 173             'ie_key': 'ThePlatform',
 174             'url': smuggle_url(
 175                 'http://link.theplatform.com/s/ExhSPC/media/guid/2655402169/%s?mbr=true&formats=MPEG4,FLV,MP3' % video_id, {
 176                     'force_smil_url': True
 177                 }),
 178             'id': video_id,
 179         }
 180
 181
 182 class CBCWatchBaseIE(InfoExtractor):
 183     _device_id = None
 184     _device_token = None
 185     _API_BASE_URL = 'https://api-cbc.cloud.clearleap.com/cloffice/client/'
 186     _NS_MAP = {
 187         'media': 'http://search.yahoo.com/mrss/',
 188         'clearleap': 'http://www.clearleap.com/namespace/clearleap/1.0/',
 189     }
 190
 191     def _call_api(self, path, video_id):
 192         url = path if path.startswith('http') else self._API_BASE_URL + path
 193         result = self._download_xml(url, video_id, headers={
 194             'X-Clearleap-DeviceId': self._device_id,
 195             'X-Clearleap-DeviceToken': self._device_token,
 196         })
 197         error_message = xpath_text(result, 'userMessage') or xpath_text(result, 'systemMessage')
 198         if error_message:
 199             raise ExtractorError('%s said: %s' % (self.IE_NAME, error_message))
 200         return result
 201
 202     def _real_initialize(self):
 203         if not self._device_id or not self._device_token:
 204             device = self._downloader.cache.load('cbcwatch', 'device') or {}
 205             self._device_id, self._device_token = device.get('id'), device.get('token')
 206             if not self._device_id or not self._device_token:
 207                 result = self._download_xml(
 208                     self._API_BASE_URL + 'device/register',
 209                     None, data=b'<device><type>web</type></device>')
 210                 self._device_id = xpath_text(result, 'deviceId', fatal=True)
 211                 self._device_token = xpath_text(result, 'deviceToken', fatal=True)
 212                 self._downloader.cache.store(
 213                     'cbcwatch', 'device', {
 214                         'id': self._device_id,
 215                         'token': self._device_token,
 216                     })
 217
 218     def _parse_rss_feed(self, rss):
 219         channel = xpath_element(rss, 'channel', fatal=True)
 220
 221         def _add_ns(path):
 222             return xpath_with_ns(path, self._NS_MAP)
 223
 224         entries = []
 225         for item in channel.findall('item'):
 226             guid = xpath_text(item, 'guid', fatal=True)
 227             title = xpath_text(item, 'title', fatal=True)
 228
 229             media_group = xpath_element(item, _add_ns('media:group'), fatal=True)
 230             content = xpath_element(media_group, _add_ns('media:content'), fatal=True)
 231             content_url = content.attrib['url']
 232
 233             thumbnails = []
 234             for thumbnail in media_group.findall(_add_ns('media:thumbnail')):
 235                 thumbnail_url = thumbnail.get('url')
 236                 if not thumbnail_url:
 237                     continue
 238                 thumbnails.append({
 239                     'id': thumbnail.get('profile'),
 240                     'url': thumbnail_url,
 241                     'width': int_or_none(thumbnail.get('width')),
 242                     'height': int_or_none(thumbnail.get('height')),
 243                 })
 244
 245             timestamp = None
 246             release_date = find_xpath_attr(
 247                 item, _add_ns('media:credit'), 'role', 'releaseDate')
 248             if release_date is not None:
 249                 timestamp = parse_iso8601(release_date.text)
 250
 251             entries.append({
 252                 '_type': 'url_transparent',
 253                 'url': content_url,
 254                 'id': guid,
 255                 'title': title,
 256                 'description': xpath_text(item, 'description'),
 257                 'timestamp': timestamp,
 258                 'duration': int_or_none(content.get('duration')),
 259                 'age_limit': parse_age_limit(xpath_text(item, _add_ns('media:rating'))),
 260                 'episode': xpath_text(item, _add_ns('clearleap:episode')),
 261                 'episode_number': int_or_none(xpath_text(item, _add_ns('clearleap:episodeInSeason'))),
 262                 'series': xpath_text(item, _add_ns('clearleap:series')),
 263                 'season_number': int_or_none(xpath_text(item, _add_ns('clearleap:season'))),
 264                 'thumbnails': thumbnails,
 265                 'ie_key': 'CBCWatchVideo',
 266             })
 267
 268         return self.playlist_result(
 269             entries, xpath_text(channel, 'guid'),
 270             xpath_text(channel, 'title'),
 271             xpath_text(channel, 'description'))
 272
 273
 274 class CBCWatchVideoIE(CBCWatchBaseIE):
 275     IE_NAME = 'cbc.ca:watch:video'
 276     _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})'
 277
 278     def _real_extract(self, url):
 279         video_id = self._match_id(url)
 280         result = self._call_api(url, video_id)
 281
 282         m3u8_url = xpath_text(result, 'url', fatal=True)
 283         formats = self._extract_m3u8_formats(re.sub(r'/([^/]+)/[^/?]+\.m3u8', r'/\1/\1.m3u8', m3u8_url), video_id, 'mp4', fatal=False)
 284         if len(formats) < 2:
 285             formats = self._extract_m3u8_formats(m3u8_url, video_id, 'mp4')
 286         # Despite metadata in m3u8 all video+audio formats are
 287         # actually video-only (no audio)
 288         for f in formats:
 289             if f.get('acodec') != 'none' and f.get('vcodec') != 'none':
 290                 f['acodec'] = 'none'
 291         self._sort_formats(formats)
 292
 293         info = {
 294             'id': video_id,
 295             'title': video_id,
 296             'formats': formats,
 297         }
 298
 299         rss = xpath_element(result, 'rss')
 300         if rss:
 301             info.update(self._parse_rss_feed(rss)['entries'][0])
 302             del info['url']
 303             del info['_type']
 304             del info['ie_key']
 305         return info
 306
 307
 308 class CBCWatchIE(CBCWatchBaseIE):
 309     IE_NAME = 'cbc.ca:watch'
 310     _VALID_URL = r'https?://watch\.cbc\.ca/(?:[^/]+/)+(?P<id>[0-9a-f-]+)'
 311     _TESTS = [{
 312         'url': 'http://watch.cbc.ca/doc-zone/season-6/customer-disservice/38e815a-009e3ab12e4',
 313         'info_dict': {
 314             'id': '38e815a-009e3ab12e4',
 315             'ext': 'mp4',
 316             'title': 'Customer (Dis)Service',
 317             'description': 'md5:8bdd6913a0fe03d4b2a17ebe169c7c87',
 318             'upload_date': '20160219',
 319             'timestamp': 1455840000,
 320         },
 321         'params': {
 322             # m3u8 download
 323             'skip_download': True,
 324             'format': 'bestvideo',
 325         },
 326         'skip': 'Geo-restricted to Canada',
 327     }, {
 328         'url': 'http://watch.cbc.ca/arthur/all/1ed4b385-cd84-49cf-95f0-80f004680057',
 329         'info_dict': {
 330             'id': '1ed4b385-cd84-49cf-95f0-80f004680057',
 331             'title': 'Arthur',
 332             'description': 'Arthur, the sweetest 8-year-old aardvark, and his pals solve all kinds of problems with humour, kindness and teamwork.',
 333         },
 334         'playlist_mincount': 30,
 335         'skip': 'Geo-restricted to Canada',
 336     }]
 337
 338     def _real_extract(self, url):
 339         video_id = self._match_id(url)
 340         rss = self._call_api('web/browse/' + video_id, video_id)
 341         return self._parse_rss_feed(rss)