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     unescapeHTML,
  16     ExtractorError,
  17     int_or_none,
  18     mimetype2ext,
  19     determine_ext,
  20 )
  21
  22 from .brightcove import (
  23     BrightcoveLegacyIE,
  24     BrightcoveNewIE,
  25 )
  26 from .nbc import NBCSportsVPlayerIE
  27
  28
  29 class YahooIE(InfoExtractor):
  30     IE_DESC = 'Yahoo screen and movies'
  31     _VALID_URL = r'(?P<url>(?P<host>https?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P<display_id>.+)?-(?P<id>[0-9]+)(?:-[a-z]+)?(?:\.html)?)'
  32     _TESTS = [
  33         {
  34             'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
  35             'info_dict': {
  36                 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
  37                 'ext': 'mp4',
  38                 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
  39                 'description': 'Julian and Travis watch Julian Smith',
  40                 'duration': 6863,
  41             },
  42         },
  43         {
  44             'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  45             'md5': '251af144a19ebc4a033e8ba91ac726bb',
  46             'info_dict': {
  47                 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
  48                 'ext': 'mp4',
  49                 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
  50                 'description': 'md5:66b627ab0a282b26352136ca96ce73c1',
  51                 'duration': 151,
  52             },
  53         },
  54         {
  55             'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
  56             'md5': '7993e572fac98e044588d0b5260f4352',
  57             'info_dict': {
  58                 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
  59                 'ext': 'mp4',
  60                 'title': "Yahoo Saves 'Community'",
  61                 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
  62                 'duration': 170,
  63             }
  64         },
  65         {
  66             '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',
  67             'md5': '45c024bad51e63e9b6f6fad7a43a8c23',
  68             'info_dict': {
  69                 'id': 'cac903b3-fcf4-3c14-b632-643ab541712f',
  70                 'ext': 'mp4',
  71                 'title': '敢問市長／黃秀霜批賴清德「非常高傲」',
  72                 'description': '直言台南沒捷運 交通居五都之末',
  73                 'duration': 396,
  74             },
  75         },
  76         {
  77             'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
  78             'md5': '71298482f7c64cbb7fa064e4553ff1c1',
  79             'info_dict': {
  80                 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
  81                 'ext': 'webm',
  82                 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
  83                 'description': 'md5:f66c890e1490f4910a9953c941dee944',
  84                 'duration': 97,
  85             }
  86         },
  87         {
  88             'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html',
  89             'md5': '57e06440778b1828a6079d2f744212c4',
  90             'info_dict': {
  91                 'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73',
  92                 'ext': 'mp4',
  93                 'title': 'Program that makes hockey more affordable not offered in Manitoba',
  94                 'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
  95                 'duration': 121,
  96             },
  97             'skip': 'Video gone',
  98         }, {
  99             'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
 100             'info_dict': {
 101                 'id': '154609075',
 102             },
 103             'playlist': [{
 104                 'md5': '000887d0dc609bc3a47c974151a40fb8',
 105                 'info_dict': {
 106                     'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
 107                     'ext': 'mp4',
 108                     'title': '\'The Interview\' TV Spot: War',
 109                     'description': 'The Interview',
 110                     'duration': 30,
 111                 },
 112             }, {
 113                 'md5': '81bc74faf10750fe36e4542f9a184c66',
 114                 'info_dict': {
 115                     'id': '1fc8ada0-718e-3abe-a450-bf31f246d1a9',
 116                     'ext': 'mp4',
 117                     'title': '\'The Interview\' TV Spot: Guys',
 118                     'description': 'The Interview',
 119                     'duration': 30,
 120                 },
 121             }],
 122         }, {
 123             'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 124             'md5': '88e209b417f173d86186bef6e4d1f160',
 125             'info_dict': {
 126                 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
 127                 'ext': 'mp4',
 128                 'title': 'China Moses Is Crazy About the Blues',
 129                 'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 130                 'duration': 128,
 131             }
 132         }, {
 133             'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
 134             'md5': 'd9a083ccf1379127bf25699d67e4791b',
 135             'info_dict': {
 136                 'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
 137                 'ext': 'mp4',
 138                 'title': 'Connect the Dots: Dark Side of Virgo',
 139                 'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
 140                 'duration': 201,
 141             },
 142             'skip': 'Domain name in.lifestyle.yahoo.com gone',
 143         }, {
 144             'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
 145             'md5': '2a9752f74cb898af5d1083ea9f661b58',
 146             'info_dict': {
 147                 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
 148                 'ext': 'mp4',
 149                 'title': '\'True Story\' Trailer',
 150                 'description': 'True Story',
 151                 'duration': 150,
 152             },
 153         }, {
 154             'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
 155             'only_matching': True,
 156         }, {
 157             'note': 'NBC Sports embeds',
 158             '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',
 159             'info_dict': {
 160                 'id': '9CsDKds0kvHI',
 161                 'ext': 'flv',
 162                 'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
 163                 'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
 164                 'upload_date': '20150313',
 165                 'uploader': 'NBCU-SPORTS',
 166                 'timestamp': 1426270238,
 167             }
 168         }, {
 169             'url': 'https://tw.news.yahoo.com/-100120367.html',
 170             'only_matching': True,
 171         }, {
 172             # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
 173             'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
 174             'md5': '4fbafb9c9b6f07aa8f870629f6671b35',
 175             'info_dict': {
 176                 'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
 177                 'ext': 'mp4',
 178                 'title': 'Communitary - Community Episode 1: Ladders',
 179                 'description': 'md5:8fc39608213295748e1e289807838c97',
 180                 'duration': 1646,
 181             },
 182         }, {
 183             # it uses an alias to get the video_id
 184             'url': 'https://www.yahoo.com/movies/the-stars-of-daddys-home-have-very-different-212843197.html',
 185             'info_dict': {
 186                 'id': '40eda9c8-8e5f-3552-8745-830f67d0c737',
 187                 'ext': 'mp4',
 188                 'title': 'Will Ferrell & Mark Wahlberg Are Pro-Spanking',
 189                 'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
 190             },
 191         },
 192         {
 193             # config['models']['applet_model']['data']['sapi'] has no query
 194             'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
 195             'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
 196             'info_dict': {
 197                 'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
 198                 'ext': 'mp4',
 199                 'description': 'Galactic',
 200                 'title': 'Dolla Diva (feat. Maggie Koerner)',
 201             },
 202             'skip': 'redirect to https://www.yahoo.com/music',
 203         },
 204     ]
 205
 206     def _real_extract(self, url):
 207         mobj = re.match(self._VALID_URL, url)
 208         display_id = mobj.group('display_id') or self._match_id(url)
 209         page_id = mobj.group('id')
 210         url = mobj.group('url')
 211         host = mobj.group('host')
 212         webpage, urlh = self._download_webpage_handle(url, display_id)
 213         if 'err=404' in urlh.geturl():
 214             raise ExtractorError('Video gone', expected=True)
 215
 216         # Look for iframed media first
 217         entries = []
 218         iframe_urls = re.findall(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
 219         for idx, iframe_url in enumerate(iframe_urls):
 220             entries.append(self.url_result(host + iframe_url, 'Yahoo'))
 221         if entries:
 222             return self.playlist_result(entries, page_id)
 223
 224         # Look for NBCSports iframes
 225         nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
 226         if nbc_sports_url:
 227             return self.url_result(nbc_sports_url, NBCSportsVPlayerIE.ie_key())
 228
 229         # Look for Brightcove Legacy Studio embeds
 230         bc_url = BrightcoveLegacyIE._extract_brightcove_url(webpage)
 231         if bc_url:
 232             return self.url_result(bc_url, BrightcoveLegacyIE.ie_key())
 233
 234         # Look for Brightcove New Studio embeds
 235         bc_url = BrightcoveNewIE._extract_url(webpage)
 236         if bc_url:
 237             return self.url_result(bc_url, BrightcoveNewIE.ie_key())
 238
 239         # Query result is often embedded in webpage as JSON. Sometimes explicit requests
 240         # to video API results in a failure with geo restriction reason therefore using
 241         # embedded query result when present sounds reasonable.
 242         config_json = self._search_regex(
 243             r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
 244             webpage, 'videoplayer applet', default=None)
 245         if config_json:
 246             config = self._parse_json(config_json, display_id, fatal=False)
 247             if config:
 248                 sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
 249                 if sapi and 'query' in sapi:
 250                     info = self._extract_info(display_id, sapi, webpage)
 251                     self._sort_formats(info['formats'])
 252                     return info
 253
 254         items_json = self._search_regex(
 255             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
 256             default=None)
 257         if items_json is None:
 258             alias = self._search_regex(
 259                 r'"aliases":{"video":"(.*?)"', webpage, 'alias', default=None)
 260             if alias is not None:
 261                 alias_info = self._download_json(
 262                     'https://www.yahoo.com/_td/api/resource/VideoService.videos;video_aliases=["%s"]' % alias,
 263                     display_id, 'Downloading alias info')
 264                 video_id = alias_info[0]['id']
 265             else:
 266                 CONTENT_ID_REGEXES = [
 267                     r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
 268                     r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
 269                     r'"first_videoid"\s*:\s*"([^"]+)"',
 270                     r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
 271                     r'<article[^>]data-uuid=["\']([^"\']+)',
 272                     r'yahoo://article/view\?.*\buuid=([^&"\']+)',
 273                 ]
 274                 video_id = self._search_regex(
 275                     CONTENT_ID_REGEXES, webpage, 'content ID')
 276         else:
 277             items = json.loads(items_json)
 278             info = items['mediaItems']['query']['results']['mediaObj'][0]
 279             # The 'meta' field is not always in the video webpage, we request it
 280             # from another page
 281             video_id = info['id']
 282         return self._get_info(video_id, display_id, webpage)
 283
 284     def _extract_info(self, display_id, query, webpage):
 285         info = query['query']['results']['mediaObj'][0]
 286         meta = info.get('meta')
 287         video_id = info.get('id')
 288
 289         if not meta:
 290             msg = info['status'].get('msg')
 291             if msg:
 292                 raise ExtractorError(
 293                     '%s returned error: %s' % (self.IE_NAME, msg), expected=True)
 294             raise ExtractorError('Unable to extract media object meta')
 295
 296         formats = []
 297         for s in info['streams']:
 298             tbr = int_or_none(s.get('bitrate'))
 299             format_info = {
 300                 'width': int_or_none(s.get('width')),
 301                 'height': int_or_none(s.get('height')),
 302                 'tbr': tbr,
 303             }
 304
 305             host = s['host']
 306             path = s['path']
 307             if host.startswith('rtmp'):
 308                 fmt = 'rtmp'
 309                 format_info.update({
 310                     'url': host,
 311                     'play_path': path,
 312                     'ext': 'flv',
 313                 })
 314             else:
 315                 if s.get('format') == 'm3u8_playlist':
 316                     fmt = 'hls'
 317                     format_info.update({
 318                         'protocol': 'm3u8_native',
 319                         'ext': 'mp4',
 320                     })
 321                 else:
 322                     fmt = format_info['ext'] = determine_ext(path)
 323                 format_url = compat_urlparse.urljoin(host, path)
 324                 format_info['url'] = format_url
 325             format_info['format_id'] = fmt + ('-%d' % tbr if tbr else '')
 326             formats.append(format_info)
 327
 328         closed_captions = self._html_search_regex(
 329             r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
 330             default='[]')
 331
 332         cc_json = self._parse_json(closed_captions, video_id, fatal=False)
 333         subtitles = {}
 334         if cc_json:
 335             for closed_caption in cc_json:
 336                 lang = closed_caption['lang']
 337                 if lang not in subtitles:
 338                     subtitles[lang] = []
 339                 subtitles[lang].append({
 340                     'url': closed_caption['url'],
 341                     'ext': mimetype2ext(closed_caption['content_type']),
 342                 })
 343
 344         return {
 345             'id': video_id,
 346             'display_id': display_id,
 347             'title': unescapeHTML(meta['title']),
 348             'formats': formats,
 349             'description': clean_html(meta['description']),
 350             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
 351             'duration': int_or_none(meta.get('duration')),
 352             'subtitles': subtitles,
 353         }
 354
 355     def _get_info(self, video_id, display_id, webpage):
 356         region = self._search_regex(
 357             r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
 358             webpage, 'region', fatal=False, default='US').upper()
 359         formats = []
 360         info = {}
 361         for fmt in ('webm', 'mp4'):
 362             query_result = self._download_json(
 363                 'https://video.media.yql.yahoo.com/v1/video/sapi/streams/' + video_id,
 364                 display_id, 'Downloading %s video info' % fmt, query={
 365                     'protocol': 'http',
 366                     'region': region,
 367                     'format': fmt,
 368                 })
 369             info = self._extract_info(display_id, query_result, webpage)
 370             formats.extend(info['formats'])
 371         formats.extend(self._extract_m3u8_formats(
 372             'http://video.media.yql.yahoo.com/v1/hls/%s?region=%s' % (video_id, region),
 373             video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
 374         self._sort_formats(formats)
 375         info['formats'] = formats
 376         return info
 377
 378
 379 class YahooSearchIE(SearchInfoExtractor):
 380     IE_DESC = 'Yahoo screen search'
 381     _MAX_RESULTS = 1000
 382     IE_NAME = 'screen.yahoo:search'
 383     _SEARCH_KEY = 'yvsearch'
 384
 385     def _get_n_results(self, query, n):
 386         """Get a specified number of results for a query"""
 387         entries = []
 388         for pagenum in itertools.count(0):
 389             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)
 390             info = self._download_json(result_url, query,
 391                                        note='Downloading results page ' + str(pagenum + 1))
 392             m = info['m']
 393             results = info['results']
 394
 395             for (i, r) in enumerate(results):
 396                 if (pagenum * 30) + i >= n:
 397                     break
 398                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 399                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 400                 entries.append(e)
 401             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
 402                 break
 403
 404         return {
 405             '_type': 'playlist',
 406             'id': query,
 407             'entries': entries,
 408         }