youtube_dl/extractor/yahoo.py

   1 from __future__ import unicode_literals
   2
   3 import itertools
   4 import json
   5 import re
   6
   7 from .common import InfoExtractor, SearchInfoExtractor
   8 from ..utils import (
   9     compat_urllib_parse,
  10     compat_urlparse,
  11     clean_html,
  12     int_or_none,
  13 )
  14
  15
  16 class YahooIE(InfoExtractor):
  17     IE_DESC = 'Yahoo screen and movies'
  18     _VALID_URL = r'(?P<url>https?://(?:screen|movies)\.yahoo\.com/.*?-(?P<id>[0-9]+)(?:-[a-z]+)?\.html)'
  19     _TESTS = [
  20         {
  21             'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
  22             'md5': '4962b075c08be8690a922ee026d05e69',
  23             'info_dict': {
  24                 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
  25                 'ext': 'mp4',
  26                 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
  27                 'description': 'Julian and Travis watch Julian Smith',
  28             },
  29         },
  30         {
  31             'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  32             'md5': 'd6e6fc6e1313c608f316ddad7b82b306',
  33             'info_dict': {
  34                 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
  35                 'ext': 'mp4',
  36                 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
  37                 'description': 'Agent Topple\'s mustache does its dirty work, and Nicole brokers a deal for peace. But why is the NSA collecting millions of Instagram brunch photos? And if your waffles have nothing to hide, what are they so worried about?',
  38             },
  39         },
  40         {
  41             'url': 'https://movies.yahoo.com/video/world-loves-spider-man-190819223.html',
  42             'md5': '410b7104aa9893b765bc22787a22f3d9',
  43             'info_dict': {
  44                 'id': '516ed8e2-2c4f-339f-a211-7a8b49d30845',
  45                 'ext': 'mp4',
  46                 'title': 'The World Loves Spider-Man',
  47                 'description': '''People all over the world are celebrating the release of \"The Amazing Spider-Man 2.\" We're taking a look at the enthusiastic response Spider-Man has received from viewers all over the world.''',
  48             }
  49         },
  50         {
  51             'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
  52             'md5': '60e8ac193d8fb71997caa8fce54c6460',
  53             'info_dict': {
  54                 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
  55                 'ext': 'mp4',
  56                 'title': "Yahoo Saves 'Community'",
  57                 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
  58             }
  59         },
  60     ]
  61
  62     def _real_extract(self, url):
  63         mobj = re.match(self._VALID_URL, url)
  64         video_id = mobj.group('id')
  65         url = mobj.group('url')
  66         webpage = self._download_webpage(url, video_id)
  67
  68         items_json = self._search_regex(
  69             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
  70             default=None)
  71         if items_json is None:
  72             CONTENT_ID_REGEXES = [
  73                 r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
  74                 r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
  75                 r'"first_videoid"\s*:\s*"([^"]+)"',
  76             ]
  77             long_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID')
  78             video_id = long_id
  79         else:
  80             items = json.loads(items_json)
  81             info = items['mediaItems']['query']['results']['mediaObj'][0]
  82             # The 'meta' field is not always in the video webpage, we request it
  83             # from another page
  84             long_id = info['id']
  85         return self._get_info(long_id, video_id, webpage)
  86
  87     def _get_info(self, long_id, video_id, webpage):
  88         query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
  89                  ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="US"'
  90                  ' AND protocol="http"' % long_id)
  91         data = compat_urllib_parse.urlencode({
  92             'q': query,
  93             'env': 'prod',
  94             'format': 'json',
  95         })
  96         query_result = self._download_json(
  97             'http://video.query.yahoo.com/v1/public/yql?' + data,
  98             video_id, 'Downloading video info')
  99         info = query_result['query']['results']['mediaObj'][0]
 100         meta = info['meta']
 101
 102         formats = []
 103         for s in info['streams']:
 104             format_info = {
 105                 'width': int_or_none(s.get('width')),
 106                 'height': int_or_none(s.get('height')),
 107                 'tbr': int_or_none(s.get('bitrate')),
 108             }
 109
 110             host = s['host']
 111             path = s['path']
 112             if host.startswith('rtmp'):
 113                 format_info.update({
 114                     'url': host,
 115                     'play_path': path,
 116                     'ext': 'flv',
 117                 })
 118             else:
 119                 format_url = compat_urlparse.urljoin(host, path)
 120                 format_info['url'] = format_url
 121             formats.append(format_info)
 122
 123         self._sort_formats(formats)
 124
 125         return {
 126             'id': video_id,
 127             'title': meta['title'],
 128             'formats': formats,
 129             'description': clean_html(meta['description']),
 130             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
 131         }
 132
 133
 134 class YahooNewsIE(YahooIE):
 135     IE_NAME = 'yahoo:news'
 136     _VALID_URL = r'http://news\.yahoo\.com/video/.*?-(?P<id>\d*?)\.html'
 137
 138     _TESTS = [{
 139         'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 140         'md5': '67010fdf3a08d290e060a4dd96baa07b',
 141         'info_dict': {
 142             'id': '104538833',
 143             'ext': 'mp4',
 144             'title': 'China Moses Is Crazy About the Blues',
 145             'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 146         },
 147     }]
 148
 149     def _real_extract(self, url):
 150         mobj = re.match(self._VALID_URL, url)
 151         video_id = mobj.group('id')
 152         webpage = self._download_webpage(url, video_id)
 153         long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, 'long id')
 154         return self._get_info(long_id, video_id, webpage)
 155
 156
 157 class YahooSearchIE(SearchInfoExtractor):
 158     IE_DESC = 'Yahoo screen search'
 159     _MAX_RESULTS = 1000
 160     IE_NAME = 'screen.yahoo:search'
 161     _SEARCH_KEY = 'yvsearch'
 162
 163     def _get_n_results(self, query, n):
 164         """Get a specified number of results for a query"""
 165         entries = []
 166         for pagenum in itertools.count(0):
 167             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)
 168             info = self._download_json(result_url, query,
 169                 note='Downloading results page '+str(pagenum+1))
 170             m = info['m']
 171             results = info['results']
 172
 173             for (i, r) in enumerate(results):
 174                 if (pagenum * 30) + i >= n:
 175                     break
 176                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 177                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 178                 entries.append(e)
 179             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
 180                 break
 181
 182         return {
 183             '_type': 'playlist',
 184             'id': query,
 185             'entries': entries,
 186         }