youtube_dl/extractor/yahoo.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import itertools
   5 import json
   6 import re
   7
   8 from .common import InfoExtractor, SearchInfoExtractor
   9 from ..compat import (
  10     compat_urllib_parse,
  11     compat_urlparse,
  12 )
  13 from ..utils import (
  14     clean_html,
  15     determine_ext,
  16     ExtractorError,
  17     extract_attributes,
  18     int_or_none,
  19     mimetype2ext,
  20     smuggle_url,
  21     unescapeHTML,
  22 )
  23
  24 from .brightcove import (
  25     BrightcoveLegacyIE,
  26     BrightcoveNewIE,
  27 )
  28 from .nbc import NBCSportsVPlayerIE
  29
  30
  31 class YahooIE(InfoExtractor):
  32     IE_DESC = 'Yahoo screen and movies'
  33     _VALID_URL = r'(?P<host>https?://(?:(?P<country>[a-zA-Z]{2})\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?:(?P<display_id>.+)?-)?(?P<id>[0-9]+)(?:-[a-z]+)?(?:\.html)?'
  34     _TESTS = [
  35         {
  36             'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
  37             'info_dict': {
  38                 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
  39                 'ext': 'mp4',
  40                 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
  41                 'description': 'Julian and Travis watch Julian Smith',
  42                 'duration': 6863,
  43             },
  44         },
  45         {
  46             'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  47             'md5': '251af144a19ebc4a033e8ba91ac726bb',
  48             'info_dict': {
  49                 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
  50                 'ext': 'mp4',
  51                 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
  52                 'description': 'md5:66b627ab0a282b26352136ca96ce73c1',
  53                 'duration': 151,
  54             },
  55             'skip': 'HTTP Error 404',
  56         },
  57         {
  58             'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
  59             'md5': '7993e572fac98e044588d0b5260f4352',
  60             'info_dict': {
  61                 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
  62                 'ext': 'mp4',
  63                 'title': "Yahoo Saves 'Community'",
  64                 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
  65                 'duration': 170,
  66             }
  67         },
  68         {
  69             'url': 'https://tw.news.yahoo.com/%E6%95%A2%E5%95%8F%E5%B8%82%E9%95%B7%20%E9%BB%83%E7%A7%80%E9%9C%9C%E6%89%B9%E8%B3%B4%E6%B8%85%E5%BE%B7%20%E9%9D%9E%E5%B8%B8%E9%AB%98%E5%82%B2-034024051.html',
  70             'md5': '45c024bad51e63e9b6f6fad7a43a8c23',
  71             'info_dict': {
  72                 'id': 'cac903b3-fcf4-3c14-b632-643ab541712f',
  73                 'ext': 'mp4',
  74                 'title': '敢問市長／黃秀霜批賴清德「非常高傲」',
  75                 'description': '直言台南沒捷運 交通居五都之末',
  76                 'duration': 396,
  77             },
  78         },
  79         {
  80             'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
  81             'md5': '71298482f7c64cbb7fa064e4553ff1c1',
  82             'info_dict': {
  83                 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
  84                 'ext': 'webm',
  85                 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
  86                 'description': 'md5:f66c890e1490f4910a9953c941dee944',
  87                 'duration': 97,
  88             }
  89         },
  90         {
  91             'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html',
  92             'md5': '57e06440778b1828a6079d2f744212c4',
  93             'info_dict': {
  94                 'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73',
  95                 'ext': 'mp4',
  96                 'title': 'Program that makes hockey more affordable not offered in Manitoba',
  97                 'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
  98                 'duration': 121,
  99             },
 100             'skip': 'Video gone',
 101         }, {
 102             'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
 103             'info_dict': {
 104                 'id': '154609075',
 105             },
 106             'playlist': [{
 107                 'md5': '000887d0dc609bc3a47c974151a40fb8',
 108                 'info_dict': {
 109                     'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
 110                     'ext': 'mp4',
 111                     'title': '\'The Interview\' TV Spot: War',
 112                     'description': 'The Interview',
 113                     'duration': 30,
 114                 },
 115             }, {
 116                 'md5': '81bc74faf10750fe36e4542f9a184c66',
 117                 'info_dict': {
 118                     'id': '1fc8ada0-718e-3abe-a450-bf31f246d1a9',
 119                     'ext': 'mp4',
 120                     'title': '\'The Interview\' TV Spot: Guys',
 121                     'description': 'The Interview',
 122                     'duration': 30,
 123                 },
 124             }],
 125         }, {
 126             'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 127             'md5': '88e209b417f173d86186bef6e4d1f160',
 128             'info_dict': {
 129                 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
 130                 'ext': 'mp4',
 131                 'title': 'China Moses Is Crazy About the Blues',
 132                 'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 133                 'duration': 128,
 134             }
 135         }, {
 136             'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
 137             'md5': 'd9a083ccf1379127bf25699d67e4791b',
 138             'info_dict': {
 139                 'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
 140                 'ext': 'mp4',
 141                 'title': 'Connect the Dots: Dark Side of Virgo',
 142                 'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
 143                 'duration': 201,
 144             },
 145             'skip': 'Domain name in.lifestyle.yahoo.com gone',
 146         }, {
 147             'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
 148             'md5': '989396ae73d20c6f057746fb226aa215',
 149             'info_dict': {
 150                 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
 151                 'ext': 'mp4',
 152                 'title': '\'True Story\' Trailer',
 153                 'description': 'True Story',
 154                 'duration': 150,
 155             },
 156         }, {
 157             'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
 158             'only_matching': True,
 159         }, {
 160             'note': 'NBC Sports embeds',
 161             'url': 'http://sports.yahoo.com/blogs/ncaab-the-dagger/tyler-kalinoski-s-buzzer-beater-caps-davidson-s-comeback-win-185609842.html?guid=nbc_cbk_davidsonbuzzerbeater_150313',
 162             'info_dict': {
 163                 'id': '9CsDKds0kvHI',
 164                 'ext': 'flv',
 165                 'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
 166                 'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
 167                 'upload_date': '20150313',
 168                 'uploader': 'NBCU-SPORTS',
 169                 'timestamp': 1426270238,
 170             }
 171         }, {
 172             'url': 'https://tw.news.yahoo.com/-100120367.html',
 173             'only_matching': True,
 174         }, {
 175             # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
 176             'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
 177             'md5': '4fbafb9c9b6f07aa8f870629f6671b35',
 178             'info_dict': {
 179                 'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
 180                 'ext': 'mp4',
 181                 'title': 'Communitary - Community Episode 1: Ladders',
 182                 'description': 'md5:8fc39608213295748e1e289807838c97',
 183                 'duration': 1646,
 184             },
 185         }, {
 186             # it uses an alias to get the video_id
 187             'url': 'https://www.yahoo.com/movies/the-stars-of-daddys-home-have-very-different-212843197.html',
 188             'info_dict': {
 189                 'id': '40eda9c8-8e5f-3552-8745-830f67d0c737',
 190                 'ext': 'mp4',
 191                 'title': 'Will Ferrell & Mark Wahlberg Are Pro-Spanking',
 192                 'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
 193             },
 194         },
 195         {
 196             # config['models']['applet_model']['data']['sapi'] has no query
 197             'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
 198             'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
 199             'info_dict': {
 200                 'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
 201                 'ext': 'mp4',
 202                 'description': 'Galactic',
 203                 'title': 'Dolla Diva (feat. Maggie Koerner)',
 204             },
 205             'skip': 'redirect to https://www.yahoo.com/music',
 206         },
 207         {
 208             # yahoo://article/
 209             'url': 'https://www.yahoo.com/movies/video/true-story-trailer-173000497.html',
 210             'info_dict': {
 211                 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
 212                 'ext': 'mp4',
 213                 'title': "'True Story' Trailer",
 214                 'description': 'True Story',
 215             },
 216             'params': {
 217                 'skip_download': True,
 218             },
 219         },
 220         {
 221             # ytwnews://cavideo/
 222             'url': 'https://tw.video.yahoo.com/movie-tw/單車天使-中文版預-092316541.html',
 223             'info_dict': {
 224                 'id': 'ba133ff2-0793-3510-b636-59dfe9ff6cff',
 225                 'ext': 'mp4',
 226                 'title': '單車天使 - 中文版預',
 227                 'description': '中文版預',
 228             },
 229             'params': {
 230                 'skip_download': True,
 231             },
 232         },
 233         {
 234             # custom brightcove
 235             'url': 'https://au.tv.yahoo.com/plus7/sunrise/-/watch/37083565/clown-entertainers-say-it-is-hurting-their-business/',
 236             'info_dict': {
 237                 'id': '5575377707001',
 238                 'ext': 'mp4',
 239                 'title': "Clown entertainers say 'It' is hurting their business",
 240                 'description': 'Stephen King s horror film has much to answer for. Jelby and Mr Loopy the Clowns join us.',
 241                 'timestamp': 1505341164,
 242                 'upload_date': '20170913',
 243                 'uploader_id': '2376984109001',
 244             },
 245             'params': {
 246                 'skip_download': True,
 247             },
 248         },
 249         {
 250             # custom brightcove, geo-restricted to Australia, bypassable
 251             'url': 'https://au.tv.yahoo.com/plus7/sunrise/-/watch/37263964/sunrise-episode-wed-27-sep/',
 252             'only_matching': True,
 253         }
 254     ]
 255
 256     def _real_extract(self, url):
 257         mobj = re.match(self._VALID_URL, url)
 258         page_id = mobj.group('id')
 259         display_id = mobj.group('display_id') or page_id
 260         host = mobj.group('host')
 261         webpage, urlh = self._download_webpage_handle(url, display_id)
 262         if 'err=404' in urlh.geturl():
 263             raise ExtractorError('Video gone', expected=True)
 264
 265         # Look for iframed media first
 266         entries = []
 267         iframe_urls = re.findall(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
 268         for idx, iframe_url in enumerate(iframe_urls):
 269             entries.append(self.url_result(host + iframe_url, 'Yahoo'))
 270         if entries:
 271             return self.playlist_result(entries, page_id)
 272
 273         # Look for NBCSports iframes
 274         nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
 275         if nbc_sports_url:
 276             return self.url_result(nbc_sports_url, NBCSportsVPlayerIE.ie_key())
 277
 278         # Look for Brightcove Legacy Studio embeds
 279         bc_url = BrightcoveLegacyIE._extract_brightcove_url(webpage)
 280         if bc_url:
 281             return self.url_result(bc_url, BrightcoveLegacyIE.ie_key())
 282
 283         def brightcove_url_result(bc_url):
 284             return self.url_result(
 285                 smuggle_url(bc_url, {'geo_countries': [mobj.group('country')]}),
 286                 BrightcoveNewIE.ie_key())
 287
 288         # Look for Brightcove New Studio embeds
 289         bc_url = BrightcoveNewIE._extract_url(self, webpage)
 290         if bc_url:
 291             return brightcove_url_result(bc_url)
 292
 293         brightcove_iframe = self._search_regex(
 294             r'(<iframe[^>]+data-video-id=["\']\d+[^>]+>)', webpage,
 295             'brightcove iframe', default=None)
 296         if brightcove_iframe:
 297             attr = extract_attributes(brightcove_iframe)
 298             src = attr.get('src')
 299             if src:
 300                 parsed_src = compat_urlparse.urlparse(src)
 301                 qs = compat_urlparse.parse_qs(parsed_src.query)
 302                 account_id = qs.get('accountId', ['2376984109001'])[0]
 303                 brightcove_id = attr.get('data-video-id') or qs.get('videoId', [None])[0]
 304                 if account_id and brightcove_id:
 305                     return brightcove_url_result(
 306                         'http://players.brightcove.net/%s/default_default/index.html?videoId=%s'
 307                         % (account_id, brightcove_id))
 308
 309         # Query result is often embedded in webpage as JSON. Sometimes explicit requests
 310         # to video API results in a failure with geo restriction reason therefore using
 311         # embedded query result when present sounds reasonable.
 312         config_json = self._search_regex(
 313             r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
 314             webpage, 'videoplayer applet', default=None)
 315         if config_json:
 316             config = self._parse_json(config_json, display_id, fatal=False)
 317             if config:
 318                 sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
 319                 if sapi and 'query' in sapi:
 320                     info = self._extract_info(display_id, sapi, webpage)
 321                     self._sort_formats(info['formats'])
 322                     return info
 323
 324         items_json = self._search_regex(
 325             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
 326             default=None)
 327         if items_json is None:
 328             alias = self._search_regex(
 329                 r'"aliases":{"video":"(.*?)"', webpage, 'alias', default=None)
 330             if alias is not None:
 331                 alias_info = self._download_json(
 332                     'https://www.yahoo.com/_td/api/resource/VideoService.videos;video_aliases=["%s"]' % alias,
 333                     display_id, 'Downloading alias info')
 334                 video_id = alias_info[0]['id']
 335             else:
 336                 CONTENT_ID_REGEXES = [
 337                     r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
 338                     r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
 339                     r'"first_videoid"\s*:\s*"([^"]+)"',
 340                     r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
 341                     r'<article[^>]data-uuid=["\']([^"\']+)',
 342                     r'<meta[^<>]+yahoo://article/view\?.*\buuid=([^&"\']+)',
 343                     r'<meta[^<>]+["\']ytwnews://cavideo/(?:[^/]+/)+([\da-fA-F-]+)[&"\']',
 344                 ]
 345                 video_id = self._search_regex(
 346                     CONTENT_ID_REGEXES, webpage, 'content ID')
 347         else:
 348             items = json.loads(items_json)
 349             info = items['mediaItems']['query']['results']['mediaObj'][0]
 350             # The 'meta' field is not always in the video webpage, we request it
 351             # from another page
 352             video_id = info['id']
 353         return self._get_info(video_id, display_id, webpage)
 354
 355     def _extract_info(self, display_id, query, webpage):
 356         info = query['query']['results']['mediaObj'][0]
 357         meta = info.get('meta')
 358         video_id = info.get('id')
 359
 360         if not meta:
 361             msg = info['status'].get('msg')
 362             if msg:
 363                 raise ExtractorError(
 364                     '%s returned error: %s' % (self.IE_NAME, msg), expected=True)
 365             raise ExtractorError('Unable to extract media object meta')
 366
 367         formats = []
 368         for s in info['streams']:
 369             tbr = int_or_none(s.get('bitrate'))
 370             format_info = {
 371                 'width': int_or_none(s.get('width')),
 372                 'height': int_or_none(s.get('height')),
 373                 'tbr': tbr,
 374             }
 375
 376             host = s['host']
 377             path = s['path']
 378             if host.startswith('rtmp'):
 379                 fmt = 'rtmp'
 380                 format_info.update({
 381                     'url': host,
 382                     'play_path': path,
 383                     'ext': 'flv',
 384                 })
 385             else:
 386                 if s.get('format') == 'm3u8_playlist':
 387                     fmt = 'hls'
 388                     format_info.update({
 389                         'protocol': 'm3u8_native',
 390                         'ext': 'mp4',
 391                     })
 392                 else:
 393                     fmt = format_info['ext'] = determine_ext(path)
 394                 format_url = compat_urlparse.urljoin(host, path)
 395                 format_info['url'] = format_url
 396             format_info['format_id'] = fmt + ('-%d' % tbr if tbr else '')
 397             formats.append(format_info)
 398
 399         closed_captions = self._html_search_regex(
 400             r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
 401             default='[]')
 402
 403         cc_json = self._parse_json(closed_captions, video_id, fatal=False)
 404         subtitles = {}
 405         if cc_json:
 406             for closed_caption in cc_json:
 407                 lang = closed_caption['lang']
 408                 if lang not in subtitles:
 409                     subtitles[lang] = []
 410                 subtitles[lang].append({
 411                     'url': closed_caption['url'],
 412                     'ext': mimetype2ext(closed_caption['content_type']),
 413                 })
 414
 415         return {
 416             'id': video_id,
 417             'display_id': display_id,
 418             'title': unescapeHTML(meta['title']),
 419             'formats': formats,
 420             'description': clean_html(meta['description']),
 421             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
 422             'duration': int_or_none(meta.get('duration')),
 423             'subtitles': subtitles,
 424         }
 425
 426     def _get_info(self, video_id, display_id, webpage):
 427         region = self._search_regex(
 428             r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
 429             webpage, 'region', fatal=False, default='US').upper()
 430         formats = []
 431         info = {}
 432         for fmt in ('webm', 'mp4'):
 433             query_result = self._download_json(
 434                 'https://video.media.yql.yahoo.com/v1/video/sapi/streams/' + video_id,
 435                 display_id, 'Downloading %s video info' % fmt, query={
 436                     'protocol': 'http',
 437                     'region': region,
 438                     'format': fmt,
 439                 })
 440             info = self._extract_info(display_id, query_result, webpage)
 441             formats.extend(info['formats'])
 442         formats.extend(self._extract_m3u8_formats(
 443             'http://video.media.yql.yahoo.com/v1/hls/%s?region=%s' % (video_id, region),
 444             video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
 445         self._sort_formats(formats)
 446         info['formats'] = formats
 447         return info
 448
 449
 450 class YahooSearchIE(SearchInfoExtractor):
 451     IE_DESC = 'Yahoo screen search'
 452     _MAX_RESULTS = 1000
 453     IE_NAME = 'screen.yahoo:search'
 454     _SEARCH_KEY = 'yvsearch'
 455
 456     def _get_n_results(self, query, n):
 457         """Get a specified number of results for a query"""
 458         entries = []
 459         for pagenum in itertools.count(0):
 460             result_url = 'http://video.search.yahoo.com/search/?p=%s&fr=screen&o=js&gs=0&b=%d' % (compat_urllib_parse.quote_plus(query), pagenum * 30)
 461             info = self._download_json(result_url, query,
 462                                        note='Downloading results page ' + str(pagenum + 1))
 463             m = info['m']
 464             results = info['results']
 465
 466             for (i, r) in enumerate(results):
 467                 if (pagenum * 30) + i >= n:
 468                     break
 469                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 470                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 471                 entries.append(e)
 472             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
 473                 break
 474
 475         return {
 476             '_type': 'playlist',
 477             'id': query,
 478             'entries': entries,
 479         }