Raphaël G. Git Repositories - youtubedl/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     unescapeHTML,
  16     ExtractorError,
  17     int_or_none,
  18     mimetype2ext,
  19 )
  20
  21 from .nbc import NBCSportsVPlayerIE
  22
  23
  24 class YahooIE(InfoExtractor):
  25     IE_DESC = 'Yahoo screen and movies'
  26     _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)'
  27     _TESTS = [
  28         {
  29             'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
  30             'info_dict': {
  31                 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
  32                 'ext': 'mp4',
  33                 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
  34                 'description': 'Julian and Travis watch Julian Smith',
  35                 'duration': 6863,
  36             },
  37         },
  38         {
  39             'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  40             'md5': 'd6e6fc6e1313c608f316ddad7b82b306',
  41             'info_dict': {
  42                 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
  43                 'ext': 'mp4',
  44                 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
  45                 'description': 'md5:66b627ab0a282b26352136ca96ce73c1',
  46                 'duration': 151,
  47             },
  48         },
  49         {
  50             'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
  51             'md5': '60e8ac193d8fb71997caa8fce54c6460',
  52             'info_dict': {
  53                 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
  54                 'ext': 'mp4',
  55                 'title': "Yahoo Saves 'Community'",
  56                 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
  57                 'duration': 170,
  58             }
  59         },
  60         {
  61             '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',
  62             'md5': '3a09cf59349cfaddae1797acc3c087fc',
  63             'info_dict': {
  64                 'id': 'cac903b3-fcf4-3c14-b632-643ab541712f',
  65                 'ext': 'mp4',
  66                 'title': '敢問市長／黃秀霜批賴清德「非常高傲」',
  67                 'description': '直言台南沒捷運 交通居五都之末',
  68                 'duration': 396,
  69             }
  70         },
  71         {
  72             'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
  73             'md5': '0b51660361f0e27c9789e7037ef76f4b',
  74             'info_dict': {
  75                 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
  76                 'ext': 'mp4',
  77                 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
  78                 'description': 'md5:f66c890e1490f4910a9953c941dee944',
  79                 'duration': 97,
  80             }
  81         },
  82         {
  83             'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html',
  84             'md5': '57e06440778b1828a6079d2f744212c4',
  85             'info_dict': {
  86                 'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73',
  87                 'ext': 'mp4',
  88                 'title': 'Program that makes hockey more affordable not offered in Manitoba',
  89                 'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
  90                 'duration': 121,
  91             }
  92         }, {
  93             'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
  94             'md5': '226a895aae7e21b0129e2a2006fe9690',
  95             'info_dict': {
  96                 'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
  97                 'ext': 'mp4',
  98                 'title': '\'The Interview\' TV Spot: War',
  99                 'description': 'The Interview',
 100                 'duration': 30,
 101             }
 102         }, {
 103             'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 104             'md5': '88e209b417f173d86186bef6e4d1f160',
 105             'info_dict': {
 106                 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
 107                 'ext': 'mp4',
 108                 'title': 'China Moses Is Crazy About the Blues',
 109                 'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 110                 'duration': 128,
 111             }
 112         }, {
 113             'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
 114             'md5': 'd9a083ccf1379127bf25699d67e4791b',
 115             'info_dict': {
 116                 'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
 117                 'ext': 'mp4',
 118                 'title': 'Connect the Dots: Dark Side of Virgo',
 119                 'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
 120                 'duration': 201,
 121             }
 122         }, {
 123             'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
 124             'md5': '989396ae73d20c6f057746fb226aa215',
 125             'info_dict': {
 126                 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
 127                 'ext': 'mp4',
 128                 'title': '\'True Story\' Trailer',
 129                 'description': 'True Story',
 130                 'duration': 150,
 131             },
 132         }, {
 133             'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
 134             'only_matching': True,
 135         }, {
 136             'note': 'NBC Sports embeds',
 137             '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',
 138             'info_dict': {
 139                 'id': '9CsDKds0kvHI',
 140                 'ext': 'flv',
 141                 'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
 142                 'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
 143             }
 144         }, {
 145             'url': 'https://tw.news.yahoo.com/-100120367.html',
 146             'only_matching': True,
 147         }, {
 148             # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
 149             'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
 150             'md5': '4fbafb9c9b6f07aa8f870629f6671b35',
 151             'info_dict': {
 152                 'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
 153                 'ext': 'mp4',
 154                 'title': 'Communitary - Community Episode 1: Ladders',
 155                 'description': 'md5:8fc39608213295748e1e289807838c97',
 156                 'duration': 1646,
 157             },
 158         }
 159     ]
 160
 161     def _real_extract(self, url):
 162         mobj = re.match(self._VALID_URL, url)
 163         display_id = mobj.group('display_id') or self._match_id(url)
 164         page_id = mobj.group('id')
 165         url = mobj.group('url')
 166         host = mobj.group('host')
 167         webpage = self._download_webpage(url, display_id)
 168
 169         # Look for iframed media first
 170         iframe_m = re.search(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
 171         if iframe_m:
 172             iframepage = self._download_webpage(
 173                 host + iframe_m.group(1), display_id, 'Downloading iframe webpage')
 174             items_json = self._search_regex(
 175                 r'mediaItems: (\[.+?\])$', iframepage, 'items', flags=re.MULTILINE, default=None)
 176             if items_json:
 177                 items = json.loads(items_json)
 178                 video_id = items[0]['id']
 179                 return self._get_info(video_id, display_id, webpage)
 180         # Look for NBCSports iframes
 181         nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
 182         if nbc_sports_url:
 183             return self.url_result(nbc_sports_url, 'NBCSportsVPlayer')
 184
 185         # Query result is often embedded in webpage as JSON. Sometimes explicit requests
 186         # to video API results in a failure with geo restriction reason therefore using
 187         # embedded query result when present sounds reasonable.
 188         config_json = self._search_regex(
 189             r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
 190             webpage, 'videoplayer applet', default=None)
 191         if config_json:
 192             config = self._parse_json(config_json, display_id, fatal=False)
 193             if config:
 194                 sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
 195                 if sapi:
 196                     return self._extract_info(display_id, sapi, webpage)
 197
 198         items_json = self._search_regex(
 199             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
 200             default=None)
 201         if items_json is None:
 202             CONTENT_ID_REGEXES = [
 203                 r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
 204                 r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
 205                 r'"first_videoid"\s*:\s*"([^"]+)"',
 206                 r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
 207             ]
 208             video_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID')
 209         else:
 210             items = json.loads(items_json)
 211             info = items['mediaItems']['query']['results']['mediaObj'][0]
 212             # The 'meta' field is not always in the video webpage, we request it
 213             # from another page
 214             video_id = info['id']
 215         return self._get_info(video_id, display_id, webpage)
 216
 217     def _extract_info(self, display_id, query, webpage):
 218         info = query['query']['results']['mediaObj'][0]
 219         meta = info.get('meta')
 220         video_id = info.get('id')
 221
 222         if not meta:
 223             msg = info['status'].get('msg')
 224             if msg:
 225                 raise ExtractorError(
 226                     '%s returned error: %s' % (self.IE_NAME, msg), expected=True)
 227             raise ExtractorError('Unable to extract media object meta')
 228
 229         formats = []
 230         for s in info['streams']:
 231             format_info = {
 232                 'width': int_or_none(s.get('width')),
 233                 'height': int_or_none(s.get('height')),
 234                 'tbr': int_or_none(s.get('bitrate')),
 235             }
 236
 237             host = s['host']
 238             path = s['path']
 239             if host.startswith('rtmp'):
 240                 format_info.update({
 241                     'url': host,
 242                     'play_path': path,
 243                     'ext': 'flv',
 244                 })
 245             else:
 246                 if s.get('format') == 'm3u8_playlist':
 247                     format_info['protocol'] = 'm3u8_native'
 248                     format_info['ext'] = 'mp4'
 249                 format_url = compat_urlparse.urljoin(host, path)
 250                 format_info['url'] = format_url
 251             formats.append(format_info)
 252
 253         self._sort_formats(formats)
 254
 255         closed_captions = self._html_search_regex(
 256             r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
 257             default='[]')
 258
 259         cc_json = self._parse_json(closed_captions, video_id, fatal=False)
 260         subtitles = {}
 261         if cc_json:
 262             for closed_caption in cc_json:
 263                 lang = closed_caption['lang']
 264                 if lang not in subtitles:
 265                     subtitles[lang] = []
 266                 subtitles[lang].append({
 267                     'url': closed_caption['url'],
 268                     'ext': mimetype2ext(closed_caption['content_type']),
 269                 })
 270
 271         return {
 272             'id': video_id,
 273             'display_id': display_id,
 274             'title': unescapeHTML(meta['title']),
 275             'formats': formats,
 276             'description': clean_html(meta['description']),
 277             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
 278             'duration': int_or_none(meta.get('duration')),
 279             'subtitles': subtitles,
 280         }
 281
 282     def _get_info(self, video_id, display_id, webpage):
 283         region = self._search_regex(
 284             r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
 285             webpage, 'region', fatal=False, default='US')
 286         data = compat_urllib_parse.urlencode({
 287             'protocol': 'http',
 288             'region': region,
 289         })
 290         query_url = (
 291             'https://video.media.yql.yahoo.com/v1/video/sapi/streams/'
 292             '{id}?{data}'.format(id=video_id, data=data))
 293         query_result = self._download_json(
 294             query_url, display_id, 'Downloading video info')
 295         return self._extract_info(display_id, query_result, webpage)
 296
 297
 298 class YahooSearchIE(SearchInfoExtractor):
 299     IE_DESC = 'Yahoo screen search'
 300     _MAX_RESULTS = 1000
 301     IE_NAME = 'screen.yahoo:search'
 302     _SEARCH_KEY = 'yvsearch'
 303
 304     def _get_n_results(self, query, n):
 305         """Get a specified number of results for a query"""
 306         entries = []
 307         for pagenum in itertools.count(0):
 308             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)
 309             info = self._download_json(result_url, query,
 310                                        note='Downloading results page ' + str(pagenum + 1))
 311             m = info['m']
 312             results = info['results']
 313
 314             for (i, r) in enumerate(results):
 315                 if (pagenum * 30) + i >= n:
 316                     break
 317                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 318                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 319                 entries.append(e)
 320             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
 321                 break
 322
 323         return {
 324             '_type': 'playlist',
 325             'id': query,
 326             'entries': entries,
 327         }