_ 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_urllib_parse_urlencode,
  12     compat_urlparse,
  13 )
  14 from ..utils import (
  15     clean_html,
  16     unescapeHTML,
  17     ExtractorError,
  18     int_or_none,
  19     mimetype2ext,
  20 )
  21
  22 from .nbc import NBCSportsVPlayerIE
  23
  24
  25 class YahooIE(InfoExtractor):
  26     IE_DESC = 'Yahoo screen and movies'
  27     _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)?)'
  28     _TESTS = [
  29         {
  30             'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
  31             'info_dict': {
  32                 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
  33                 'ext': 'mp4',
  34                 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
  35                 'description': 'Julian and Travis watch Julian Smith',
  36                 'duration': 6863,
  37             },
  38         },
  39         {
  40             'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  41             'md5': 'c3466d2b6d5dd6b9f41ba9ed04c24b23',
  42             'info_dict': {
  43                 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
  44                 'ext': 'mp4',
  45                 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
  46                 'description': 'md5:66b627ab0a282b26352136ca96ce73c1',
  47                 'duration': 151,
  48             },
  49         },
  50         {
  51             'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
  52             'md5': '75ffabdb87c16d4ffe8c036dc4d1c136',
  53             'info_dict': {
  54                 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
  55                 'ext': 'mp4',
  56                 'title': "Yahoo Saves 'Community'",
  57                 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
  58                 'duration': 170,
  59             }
  60         },
  61         {
  62             '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',
  63             'md5': '9035d38f88b1782682a3e89f985be5bb',
  64             'info_dict': {
  65                 'id': 'cac903b3-fcf4-3c14-b632-643ab541712f',
  66                 'ext': 'mp4',
  67                 'title': '敢問市長／黃秀霜批賴清德「非常高傲」',
  68                 'description': '直言台南沒捷運 交通居五都之末',
  69                 'duration': 396,
  70             },
  71         },
  72         {
  73             'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
  74             'md5': '0b51660361f0e27c9789e7037ef76f4b',
  75             'info_dict': {
  76                 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
  77                 'ext': 'mp4',
  78                 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
  79                 'description': 'md5:f66c890e1490f4910a9953c941dee944',
  80                 'duration': 97,
  81             }
  82         },
  83         {
  84             'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html',
  85             'md5': '57e06440778b1828a6079d2f744212c4',
  86             'info_dict': {
  87                 'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73',
  88                 'ext': 'mp4',
  89                 'title': 'Program that makes hockey more affordable not offered in Manitoba',
  90                 'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
  91                 'duration': 121,
  92             },
  93             'skip': 'Video gone',
  94         }, {
  95             'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
  96             'info_dict': {
  97                 'id': '154609075',
  98             },
  99             'playlist': [{
 100                 'md5': 'f8e336c6b66f503282e5f719641d6565',
 101                 'info_dict': {
 102                     'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
 103                     'ext': 'mp4',
 104                     'title': '\'The Interview\' TV Spot: War',
 105                     'description': 'The Interview',
 106                     'duration': 30,
 107                 },
 108             }, {
 109                 'md5': '958bcb90b4d6df71c56312137ee1cd5a',
 110                 'info_dict': {
 111                     'id': '1fc8ada0-718e-3abe-a450-bf31f246d1a9',
 112                     'ext': 'mp4',
 113                     'title': '\'The Interview\' TV Spot: Guys',
 114                     'description': 'The Interview',
 115                     'duration': 30,
 116                 },
 117             }],
 118         }, {
 119             'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 120             'md5': '88e209b417f173d86186bef6e4d1f160',
 121             'info_dict': {
 122                 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
 123                 'ext': 'mp4',
 124                 'title': 'China Moses Is Crazy About the Blues',
 125                 'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 126                 'duration': 128,
 127             }
 128         }, {
 129             'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
 130             'md5': 'd9a083ccf1379127bf25699d67e4791b',
 131             'info_dict': {
 132                 'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
 133                 'ext': 'mp4',
 134                 'title': 'Connect the Dots: Dark Side of Virgo',
 135                 'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
 136                 'duration': 201,
 137             },
 138             'skip': 'Domain name in.lifestyle.yahoo.com gone',
 139         }, {
 140             'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
 141             'md5': 'b17ac378b1134fa44370fb27db09a744',
 142             'info_dict': {
 143                 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
 144                 'ext': 'mp4',
 145                 'title': '\'True Story\' Trailer',
 146                 'description': 'True Story',
 147                 'duration': 150,
 148             },
 149         }, {
 150             'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
 151             'only_matching': True,
 152         }, {
 153             'note': 'NBC Sports embeds',
 154             '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',
 155             'info_dict': {
 156                 'id': '9CsDKds0kvHI',
 157                 'ext': 'flv',
 158                 'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
 159                 'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
 160                 'upload_date': '20150313',
 161                 'uploader': 'NBCU-SPORTS',
 162                 'timestamp': 1426270238,
 163             }
 164         }, {
 165             'url': 'https://tw.news.yahoo.com/-100120367.html',
 166             'only_matching': True,
 167         }, {
 168             # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
 169             'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
 170             'md5': '1ddbf7c850777548438e5c4f147c7b8c',
 171             'info_dict': {
 172                 'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
 173                 'ext': 'mp4',
 174                 'title': 'Communitary - Community Episode 1: Ladders',
 175                 'description': 'md5:8fc39608213295748e1e289807838c97',
 176                 'duration': 1646,
 177             },
 178         }, {
 179             # it uses an alias to get the video_id
 180             'url': 'https://www.yahoo.com/movies/the-stars-of-daddys-home-have-very-different-212843197.html',
 181             'info_dict': {
 182                 'id': '40eda9c8-8e5f-3552-8745-830f67d0c737',
 183                 'ext': 'mp4',
 184                 'title': 'Will Ferrell & Mark Wahlberg Are Pro-Spanking',
 185                 'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
 186             },
 187         },
 188         {
 189             # config['models']['applet_model']['data']['sapi'] has no query
 190             'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
 191             'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
 192             'info_dict': {
 193                 'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
 194                 'ext': 'mp4',
 195                 'description': 'Galactic',
 196                 'title': 'Dolla Diva (feat. Maggie Koerner)',
 197             },
 198         },
 199     ]
 200
 201     def _real_extract(self, url):
 202         mobj = re.match(self._VALID_URL, url)
 203         display_id = mobj.group('display_id') or self._match_id(url)
 204         page_id = mobj.group('id')
 205         url = mobj.group('url')
 206         host = mobj.group('host')
 207         webpage, urlh = self._download_webpage_handle(url, display_id)
 208         if 'err=404' in urlh.geturl():
 209             raise ExtractorError('Video gone', expected=True)
 210
 211         # Look for iframed media first
 212         entries = []
 213         iframe_urls = re.findall(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
 214         for idx, iframe_url in enumerate(iframe_urls):
 215             iframepage = self._download_webpage(
 216                 host + iframe_url, display_id,
 217                 note='Downloading iframe webpage for video #%d' % idx)
 218             items_json = self._search_regex(
 219                 r'mediaItems: (\[.+?\])$', iframepage, 'items', flags=re.MULTILINE, default=None)
 220             if items_json:
 221                 items = json.loads(items_json)
 222                 video_id = items[0]['id']
 223                 entries.append(self._get_info(video_id, display_id, webpage))
 224         if entries:
 225             return self.playlist_result(entries, page_id)
 226
 227         # Look for NBCSports iframes
 228         nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
 229         if nbc_sports_url:
 230             return self.url_result(nbc_sports_url, 'NBCSportsVPlayer')
 231
 232         # Query result is often embedded in webpage as JSON. Sometimes explicit requests
 233         # to video API results in a failure with geo restriction reason therefore using
 234         # embedded query result when present sounds reasonable.
 235         config_json = self._search_regex(
 236             r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
 237             webpage, 'videoplayer applet', default=None)
 238         if config_json:
 239             config = self._parse_json(config_json, display_id, fatal=False)
 240             if config:
 241                 sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
 242                 if sapi and 'query' in sapi:
 243                     return self._extract_info(display_id, sapi, webpage)
 244
 245         items_json = self._search_regex(
 246             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
 247             default=None)
 248         if items_json is None:
 249             alias = self._search_regex(
 250                 r'"aliases":{"video":"(.*?)"', webpage, 'alias', default=None)
 251             if alias is not None:
 252                 alias_info = self._download_json(
 253                     'https://www.yahoo.com/_td/api/resource/VideoService.videos;video_aliases=["%s"]' % alias,
 254                     display_id, 'Downloading alias info')
 255                 video_id = alias_info[0]['id']
 256             else:
 257                 CONTENT_ID_REGEXES = [
 258                     r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
 259                     r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
 260                     r'"first_videoid"\s*:\s*"([^"]+)"',
 261                     r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
 262                     r'<article[^>]data-uuid=["\']([^"\']+)',
 263                     r'yahoo://article/view\?.*\buuid=([^&"\']+)',
 264                 ]
 265                 video_id = self._search_regex(
 266                     CONTENT_ID_REGEXES, webpage, 'content ID')
 267         else:
 268             items = json.loads(items_json)
 269             info = items['mediaItems']['query']['results']['mediaObj'][0]
 270             # The 'meta' field is not always in the video webpage, we request it
 271             # from another page
 272             video_id = info['id']
 273         return self._get_info(video_id, display_id, webpage)
 274
 275     def _extract_info(self, display_id, query, webpage):
 276         info = query['query']['results']['mediaObj'][0]
 277         meta = info.get('meta')
 278         video_id = info.get('id')
 279
 280         if not meta:
 281             msg = info['status'].get('msg')
 282             if msg:
 283                 raise ExtractorError(
 284                     '%s returned error: %s' % (self.IE_NAME, msg), expected=True)
 285             raise ExtractorError('Unable to extract media object meta')
 286
 287         formats = []
 288         for s in info['streams']:
 289             format_info = {
 290                 'width': int_or_none(s.get('width')),
 291                 'height': int_or_none(s.get('height')),
 292                 'tbr': int_or_none(s.get('bitrate')),
 293             }
 294
 295             host = s['host']
 296             path = s['path']
 297             if host.startswith('rtmp'):
 298                 format_info.update({
 299                     'url': host,
 300                     'play_path': path,
 301                     'ext': 'flv',
 302                 })
 303             else:
 304                 if s.get('format') == 'm3u8_playlist':
 305                     format_info['protocol'] = 'm3u8_native'
 306                     format_info['ext'] = 'mp4'
 307                 format_url = compat_urlparse.urljoin(host, path)
 308                 format_info['url'] = format_url
 309             formats.append(format_info)
 310
 311         self._sort_formats(formats)
 312
 313         closed_captions = self._html_search_regex(
 314             r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
 315             default='[]')
 316
 317         cc_json = self._parse_json(closed_captions, video_id, fatal=False)
 318         subtitles = {}
 319         if cc_json:
 320             for closed_caption in cc_json:
 321                 lang = closed_caption['lang']
 322                 if lang not in subtitles:
 323                     subtitles[lang] = []
 324                 subtitles[lang].append({
 325                     'url': closed_caption['url'],
 326                     'ext': mimetype2ext(closed_caption['content_type']),
 327                 })
 328
 329         return {
 330             'id': video_id,
 331             'display_id': display_id,
 332             'title': unescapeHTML(meta['title']),
 333             'formats': formats,
 334             'description': clean_html(meta['description']),
 335             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
 336             'duration': int_or_none(meta.get('duration')),
 337             'subtitles': subtitles,
 338         }
 339
 340     def _get_info(self, video_id, display_id, webpage):
 341         region = self._search_regex(
 342             r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
 343             webpage, 'region', fatal=False, default='US')
 344         data = compat_urllib_parse_urlencode({
 345             'protocol': 'http',
 346             'region': region,
 347         })
 348         query_url = (
 349             'https://video.media.yql.yahoo.com/v1/video/sapi/streams/'
 350             '{id}?{data}'.format(id=video_id, data=data))
 351         query_result = self._download_json(
 352             query_url, display_id, 'Downloading video info')
 353         return self._extract_info(display_id, query_result, webpage)
 354
 355
 356 class YahooSearchIE(SearchInfoExtractor):
 357     IE_DESC = 'Yahoo screen search'
 358     _MAX_RESULTS = 1000
 359     IE_NAME = 'screen.yahoo:search'
 360     _SEARCH_KEY = 'yvsearch'
 361
 362     def _get_n_results(self, query, n):
 363         """Get a specified number of results for a query"""
 364         entries = []
 365         for pagenum in itertools.count(0):
 366             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)
 367             info = self._download_json(result_url, query,
 368                                        note='Downloading results page ' + str(pagenum + 1))
 369             m = info['m']
 370             results = info['results']
 371
 372             for (i, r) in enumerate(results):
 373                 if (pagenum * 30) + i >= n:
 374                     break
 375                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 376                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 377                 entries.append(e)
 378             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
 379                 break
 380
 381         return {
 382             '_type': 'playlist',
 383             'id': query,
 384             'entries': entries,
 385         }