_ 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 ..utils import (
  10     compat_urllib_parse,
  11     compat_urlparse,
  12     clean_html,
  13     int_or_none,
  14 )
  15
  16
  17 class YahooIE(InfoExtractor):
  18     IE_DESC = 'Yahoo screen and movies'
  19     _VALID_URL = r'(?P<url>https?://(?:.+?\.)?(?:screen|movies)\.yahoo\.com/.*?-(?P<id>[0-9]+)(?:-[a-z]+)?\.html)'
  20     _TESTS = [
  21         {
  22             'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
  23             'md5': '4962b075c08be8690a922ee026d05e69',
  24             'info_dict': {
  25                 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
  26                 'ext': 'mp4',
  27                 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
  28                 'description': 'Julian and Travis watch Julian Smith',
  29                 'duration': 6863,
  30             },
  31         },
  32         {
  33             'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  34             'md5': 'd6e6fc6e1313c608f316ddad7b82b306',
  35             'info_dict': {
  36                 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
  37                 'ext': 'mp4',
  38                 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
  39                 'description': 'md5:66b627ab0a282b26352136ca96ce73c1',
  40                 'duration': 151,
  41             },
  42         },
  43         {
  44             'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
  45             'md5': '60e8ac193d8fb71997caa8fce54c6460',
  46             'info_dict': {
  47                 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
  48                 'ext': 'mp4',
  49                 'title': "Yahoo Saves 'Community'",
  50                 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
  51                 'duration': 170,
  52             }
  53         },
  54         {
  55             'url': 'https://tw.screen.yahoo.com/taipei-opinion-poll/選情站報-街頭民調-台北市篇-102823042.html',
  56             'md5': '92a7fdd8a08783c68a174d7aa067dde8',
  57             'info_dict': {
  58                 'id': '7a23b569-7bea-36cb-85b9-bd5301a0a1fb',
  59                 'ext': 'mp4',
  60                 'title': '選情站報 街頭民調 台北市篇',
  61                 'description': '選情站報 街頭民調 台北市篇',
  62                 'duration': 429,
  63             }
  64         },
  65         {
  66             'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html  ',
  67             'md5': '0b51660361f0e27c9789e7037ef76f4b',
  68             'info_dict': {
  69                 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
  70                 'ext': 'mp4',
  71                 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
  72                 'description': 'md5:f66c890e1490f4910a9953c941dee944',
  73                 'duration': 97,
  74             }
  75         },
  76     ]
  77
  78     def _real_extract(self, url):
  79         mobj = re.match(self._VALID_URL, url)
  80         video_id = mobj.group('id')
  81         url = mobj.group('url')
  82         webpage = self._download_webpage(url, video_id)
  83
  84         items_json = self._search_regex(
  85             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
  86             default=None)
  87         if items_json is None:
  88             CONTENT_ID_REGEXES = [
  89                 r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
  90                 r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
  91                 r'"first_videoid"\s*:\s*"([^"]+)"',
  92             ]
  93             long_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID')
  94             video_id = long_id
  95         else:
  96             items = json.loads(items_json)
  97             info = items['mediaItems']['query']['results']['mediaObj'][0]
  98             # The 'meta' field is not always in the video webpage, we request it
  99             # from another page
 100             long_id = info['id']
 101         return self._get_info(long_id, video_id, webpage)
 102
 103     def _get_info(self, long_id, video_id, webpage):
 104         region = self._search_regex(
 105             r'"region"\s*:\s*"([^"]+)"', webpage, 'region', fatal=False, default='US')
 106         query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
 107                  ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="%s"'
 108                  ' AND protocol="http"' % (long_id, region))
 109         data = compat_urllib_parse.urlencode({
 110             'q': query,
 111             'env': 'prod',
 112             'format': 'json',
 113         })
 114         query_result = self._download_json(
 115             'http://video.query.yahoo.com/v1/public/yql?' + data,
 116             video_id, 'Downloading video info')
 117         info = query_result['query']['results']['mediaObj'][0]
 118         meta = info['meta']
 119
 120         formats = []
 121         for s in info['streams']:
 122             format_info = {
 123                 'width': int_or_none(s.get('width')),
 124                 'height': int_or_none(s.get('height')),
 125                 'tbr': int_or_none(s.get('bitrate')),
 126             }
 127
 128             host = s['host']
 129             path = s['path']
 130             if host.startswith('rtmp'):
 131                 format_info.update({
 132                     'url': host,
 133                     'play_path': path,
 134                     'ext': 'flv',
 135                 })
 136             else:
 137                 format_url = compat_urlparse.urljoin(host, path)
 138                 format_info['url'] = format_url
 139             formats.append(format_info)
 140
 141         self._sort_formats(formats)
 142
 143         return {
 144             'id': video_id,
 145             'title': meta['title'],
 146             'formats': formats,
 147             'description': clean_html(meta['description']),
 148             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
 149             'duration': int_or_none(meta.get('duration')),
 150         }
 151
 152
 153 class YahooNewsIE(YahooIE):
 154     IE_NAME = 'yahoo:news'
 155     _VALID_URL = r'http://news\.yahoo\.com/video/.*?-(?P<id>\d*?)\.html'
 156
 157     _TESTS = [{
 158         'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 159         'md5': '67010fdf3a08d290e060a4dd96baa07b',
 160         'info_dict': {
 161             'id': '104538833',
 162             'ext': 'mp4',
 163             'title': 'China Moses Is Crazy About the Blues',
 164             'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 165         },
 166     }]
 167
 168     def _real_extract(self, url):
 169         mobj = re.match(self._VALID_URL, url)
 170         video_id = mobj.group('id')
 171         webpage = self._download_webpage(url, video_id)
 172         long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, 'long id')
 173         return self._get_info(long_id, video_id, webpage)
 174
 175
 176 class YahooSearchIE(SearchInfoExtractor):
 177     IE_DESC = 'Yahoo screen search'
 178     _MAX_RESULTS = 1000
 179     IE_NAME = 'screen.yahoo:search'
 180     _SEARCH_KEY = 'yvsearch'
 181
 182     def _get_n_results(self, query, n):
 183         """Get a specified number of results for a query"""
 184         entries = []
 185         for pagenum in itertools.count(0):
 186             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)
 187             info = self._download_json(result_url, query,
 188                 note='Downloading results page '+str(pagenum+1))
 189             m = info['m']
 190             results = info['results']
 191
 192             for (i, r) in enumerate(results):
 193                 if (pagenum * 30) + i >= n:
 194                     break
 195                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 196                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 197                 entries.append(e)
 198             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
 199                 break
 200
 201         return {
 202             '_type': 'playlist',
 203             'id': query,
 204             'entries': entries,
 205         }