_ 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.screen.yahoo.com/election-2014-askmayor/敢問市長-黃秀霜批賴清德-非常高傲-033009720.html',
  63             'md5': '3a09cf59349cfaddae1797acc3c087fc',
  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         }, {
  94             'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
  95             'info_dict': {
  96                 'id': '154609075',
  97             },
  98             'playlist': [{
  99                 'md5': 'f8e336c6b66f503282e5f719641d6565',
 100                 'info_dict': {
 101                     'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
 102                     'ext': 'mp4',
 103                     'title': '\'The Interview\' TV Spot: War',
 104                     'description': 'The Interview',
 105                     'duration': 30,
 106                 },
 107             }, {
 108                 'md5': '958bcb90b4d6df71c56312137ee1cd5a',
 109                 'info_dict': {
 110                     'id': '1fc8ada0-718e-3abe-a450-bf31f246d1a9',
 111                     'ext': 'mp4',
 112                     'title': '\'The Interview\' TV Spot: Guys',
 113                     'description': 'The Interview',
 114                     'duration': 30,
 115                 },
 116             }],
 117         }, {
 118             'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 119             'md5': '88e209b417f173d86186bef6e4d1f160',
 120             'info_dict': {
 121                 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
 122                 'ext': 'mp4',
 123                 'title': 'China Moses Is Crazy About the Blues',
 124                 'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 125                 'duration': 128,
 126             }
 127         }, {
 128             'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
 129             'md5': 'd9a083ccf1379127bf25699d67e4791b',
 130             'info_dict': {
 131                 'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
 132                 'ext': 'mp4',
 133                 'title': 'Connect the Dots: Dark Side of Virgo',
 134                 'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
 135                 'duration': 201,
 136             }
 137         }, {
 138             'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
 139             'md5': 'b17ac378b1134fa44370fb27db09a744',
 140             'info_dict': {
 141                 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
 142                 'ext': 'mp4',
 143                 'title': '\'True Story\' Trailer',
 144                 'description': 'True Story',
 145                 'duration': 150,
 146             },
 147         }, {
 148             'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
 149             'only_matching': True,
 150         }, {
 151             'note': 'NBC Sports embeds',
 152             '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',
 153             'info_dict': {
 154                 'id': '9CsDKds0kvHI',
 155                 'ext': 'flv',
 156                 'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
 157                 'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
 158                 'upload_date': '20150313',
 159                 'uploader': 'NBCU-SPORTS',
 160                 'timestamp': 1426270238,
 161             }
 162         }, {
 163             'url': 'https://tw.news.yahoo.com/-100120367.html',
 164             'only_matching': True,
 165         }, {
 166             # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
 167             'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
 168             'md5': '1ddbf7c850777548438e5c4f147c7b8c',
 169             'info_dict': {
 170                 'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
 171                 'ext': 'mp4',
 172                 'title': 'Communitary - Community Episode 1: Ladders',
 173                 'description': 'md5:8fc39608213295748e1e289807838c97',
 174                 'duration': 1646,
 175             },
 176         }, {
 177             # it uses an alias to get the video_id
 178             'url': 'https://www.yahoo.com/movies/the-stars-of-daddys-home-have-very-different-212843197.html',
 179             'info_dict': {
 180                 'id': '40eda9c8-8e5f-3552-8745-830f67d0c737',
 181                 'ext': 'mp4',
 182                 'title': 'Will Ferrell & Mark Wahlberg Are Pro-Spanking',
 183                 'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
 184             },
 185         },
 186         {
 187             # config['models']['applet_model']['data']['sapi'] has no query
 188             'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
 189             'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
 190             'info_dict': {
 191                 'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
 192                 'ext': 'mp4',
 193                 'description': 'Galactic',
 194                 'title': 'Dolla Diva (feat. Maggie Koerner)',
 195             },
 196         },
 197     ]
 198
 199     def _real_extract(self, url):
 200         mobj = re.match(self._VALID_URL, url)
 201         display_id = mobj.group('display_id') or self._match_id(url)
 202         page_id = mobj.group('id')
 203         url = mobj.group('url')
 204         host = mobj.group('host')
 205         webpage = self._download_webpage(url, display_id)
 206
 207         # Look for iframed media first
 208         entries = []
 209         iframe_urls = re.findall(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
 210         for idx, iframe_url in enumerate(iframe_urls):
 211             iframepage = self._download_webpage(
 212                 host + iframe_url, display_id,
 213                 note='Downloading iframe webpage for video #%d' % idx)
 214             items_json = self._search_regex(
 215                 r'mediaItems: (\[.+?\])$', iframepage, 'items', flags=re.MULTILINE, default=None)
 216             if items_json:
 217                 items = json.loads(items_json)
 218                 video_id = items[0]['id']
 219                 entries.append(self._get_info(video_id, display_id, webpage))
 220         if entries:
 221             return self.playlist_result(entries, page_id)
 222
 223         # Look for NBCSports iframes
 224         nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
 225         if nbc_sports_url:
 226             return self.url_result(nbc_sports_url, 'NBCSportsVPlayer')
 227
 228         # Query result is often embedded in webpage as JSON. Sometimes explicit requests
 229         # to video API results in a failure with geo restriction reason therefore using
 230         # embedded query result when present sounds reasonable.
 231         config_json = self._search_regex(
 232             r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
 233             webpage, 'videoplayer applet', default=None)
 234         if config_json:
 235             config = self._parse_json(config_json, display_id, fatal=False)
 236             if config:
 237                 sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
 238                 if sapi and 'query' in sapi:
 239                     return self._extract_info(display_id, sapi, webpage)
 240
 241         items_json = self._search_regex(
 242             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
 243             default=None)
 244         if items_json is None:
 245             alias = self._search_regex(
 246                 r'"aliases":{"video":"(.*?)"', webpage, 'alias', default=None)
 247             if alias is not None:
 248                 alias_info = self._download_json(
 249                     'https://www.yahoo.com/_td/api/resource/VideoService.videos;video_aliases=["%s"]' % alias,
 250                     display_id, 'Downloading alias info')
 251                 video_id = alias_info[0]['id']
 252             else:
 253                 CONTENT_ID_REGEXES = [
 254                     r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
 255                     r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
 256                     r'"first_videoid"\s*:\s*"([^"]+)"',
 257                     r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
 258                     r'<article[^>]data-uuid=["\']([^"\']+)',
 259                     r'yahoo://article/view\?.*\buuid=([^&"\']+)',
 260                 ]
 261                 video_id = self._search_regex(
 262                     CONTENT_ID_REGEXES, webpage, 'content ID')
 263         else:
 264             items = json.loads(items_json)
 265             info = items['mediaItems']['query']['results']['mediaObj'][0]
 266             # The 'meta' field is not always in the video webpage, we request it
 267             # from another page
 268             video_id = info['id']
 269         return self._get_info(video_id, display_id, webpage)
 270
 271     def _extract_info(self, display_id, query, webpage):
 272         info = query['query']['results']['mediaObj'][0]
 273         meta = info.get('meta')
 274         video_id = info.get('id')
 275
 276         if not meta:
 277             msg = info['status'].get('msg')
 278             if msg:
 279                 raise ExtractorError(
 280                     '%s returned error: %s' % (self.IE_NAME, msg), expected=True)
 281             raise ExtractorError('Unable to extract media object meta')
 282
 283         formats = []
 284         for s in info['streams']:
 285             format_info = {
 286                 'width': int_or_none(s.get('width')),
 287                 'height': int_or_none(s.get('height')),
 288                 'tbr': int_or_none(s.get('bitrate')),
 289             }
 290
 291             host = s['host']
 292             path = s['path']
 293             if host.startswith('rtmp'):
 294                 format_info.update({
 295                     'url': host,
 296                     'play_path': path,
 297                     'ext': 'flv',
 298                 })
 299             else:
 300                 if s.get('format') == 'm3u8_playlist':
 301                     format_info['protocol'] = 'm3u8_native'
 302                     format_info['ext'] = 'mp4'
 303                 format_url = compat_urlparse.urljoin(host, path)
 304                 format_info['url'] = format_url
 305             formats.append(format_info)
 306
 307         self._sort_formats(formats)
 308
 309         closed_captions = self._html_search_regex(
 310             r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
 311             default='[]')
 312
 313         cc_json = self._parse_json(closed_captions, video_id, fatal=False)
 314         subtitles = {}
 315         if cc_json:
 316             for closed_caption in cc_json:
 317                 lang = closed_caption['lang']
 318                 if lang not in subtitles:
 319                     subtitles[lang] = []
 320                 subtitles[lang].append({
 321                     'url': closed_caption['url'],
 322                     'ext': mimetype2ext(closed_caption['content_type']),
 323                 })
 324
 325         return {
 326             'id': video_id,
 327             'display_id': display_id,
 328             'title': unescapeHTML(meta['title']),
 329             'formats': formats,
 330             'description': clean_html(meta['description']),
 331             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
 332             'duration': int_or_none(meta.get('duration')),
 333             'subtitles': subtitles,
 334         }
 335
 336     def _get_info(self, video_id, display_id, webpage):
 337         region = self._search_regex(
 338             r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
 339             webpage, 'region', fatal=False, default='US')
 340         data = compat_urllib_parse_urlencode({
 341             'protocol': 'http',
 342             'region': region,
 343         })
 344         query_url = (
 345             'https://video.media.yql.yahoo.com/v1/video/sapi/streams/'
 346             '{id}?{data}'.format(id=video_id, data=data))
 347         query_result = self._download_json(
 348             query_url, display_id, 'Downloading video info')
 349         return self._extract_info(display_id, query_result, webpage)
 350
 351
 352 class YahooSearchIE(SearchInfoExtractor):
 353     IE_DESC = 'Yahoo screen search'
 354     _MAX_RESULTS = 1000
 355     IE_NAME = 'screen.yahoo:search'
 356     _SEARCH_KEY = 'yvsearch'
 357
 358     def _get_n_results(self, query, n):
 359         """Get a specified number of results for a query"""
 360         entries = []
 361         for pagenum in itertools.count(0):
 362             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)
 363             info = self._download_json(result_url, query,
 364                                        note='Downloading results page ' + str(pagenum + 1))
 365             m = info['m']
 366             results = info['results']
 367
 368             for (i, r) in enumerate(results):
 369                 if (pagenum * 30) + i >= n:
 370                     break
 371                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 372                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 373                 entries.append(e)
 374             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
 375                 break
 376
 377         return {
 378             '_type': 'playlist',
 379             'id': query,
 380             'entries': entries,
 381         }