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