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://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
  42             'md5': '60e8ac193d8fb71997caa8fce54c6460',
  43             'info_dict': {
  44                 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
  45                 'ext': 'mp4',
  46                 'title': "Yahoo Saves 'Community'",
  47                 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
  48             }
  49         },
  50     ]
  51
  52     def _real_extract(self, url):
  53         mobj = re.match(self._VALID_URL, url)
  54         video_id = mobj.group('id')
  55         url = mobj.group('url')
  56         webpage = self._download_webpage(url, video_id)
  57
  58         items_json = self._search_regex(
  59             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
  60             default=None)
  61         if items_json is None:
  62             CONTENT_ID_REGEXES = [
  63                 r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
  64                 r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
  65                 r'"first_videoid"\s*:\s*"([^"]+)"',
  66             ]
  67             long_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID')
  68             video_id = long_id
  69         else:
  70             items = json.loads(items_json)
  71             info = items['mediaItems']['query']['results']['mediaObj'][0]
  72             # The 'meta' field is not always in the video webpage, we request it
  73             # from another page
  74             long_id = info['id']
  75         return self._get_info(long_id, video_id, webpage)
  76
  77     def _get_info(self, long_id, video_id, webpage):
  78         query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
  79                  ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="US"'
  80                  ' AND protocol="http"' % long_id)
  81         data = compat_urllib_parse.urlencode({
  82             'q': query,
  83             'env': 'prod',
  84             'format': 'json',
  85         })
  86         query_result = self._download_json(
  87             'http://video.query.yahoo.com/v1/public/yql?' + data,
  88             video_id, 'Downloading video info')
  89         info = query_result['query']['results']['mediaObj'][0]
  90         meta = info['meta']
  91
  92         formats = []
  93         for s in info['streams']:
  94             format_info = {
  95                 'width': int_or_none(s.get('width')),
  96                 'height': int_or_none(s.get('height')),
  97                 'tbr': int_or_none(s.get('bitrate')),
  98             }
  99
 100             host = s['host']
 101             path = s['path']
 102             if host.startswith('rtmp'):
 103                 format_info.update({
 104                     'url': host,
 105                     'play_path': path,
 106                     'ext': 'flv',
 107                 })
 108             else:
 109                 format_url = compat_urlparse.urljoin(host, path)
 110                 format_info['url'] = format_url
 111             formats.append(format_info)
 112
 113         self._sort_formats(formats)
 114
 115         return {
 116             'id': video_id,
 117             'title': meta['title'],
 118             'formats': formats,
 119             'description': clean_html(meta['description']),
 120             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
 121         }
 122
 123
 124 class YahooNewsIE(YahooIE):
 125     IE_NAME = 'yahoo:news'
 126     _VALID_URL = r'http://news\.yahoo\.com/video/.*?-(?P<id>\d*?)\.html'
 127
 128     _TESTS = [{
 129         'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 130         'md5': '67010fdf3a08d290e060a4dd96baa07b',
 131         'info_dict': {
 132             'id': '104538833',
 133             'ext': 'mp4',
 134             'title': 'China Moses Is Crazy About the Blues',
 135             'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 136         },
 137     }]
 138
 139     def _real_extract(self, url):
 140         mobj = re.match(self._VALID_URL, url)
 141         video_id = mobj.group('id')
 142         webpage = self._download_webpage(url, video_id)
 143         long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, 'long id')
 144         return self._get_info(long_id, video_id, webpage)
 145
 146
 147 class YahooSearchIE(SearchInfoExtractor):
 148     IE_DESC = 'Yahoo screen search'
 149     _MAX_RESULTS = 1000
 150     IE_NAME = 'screen.yahoo:search'
 151     _SEARCH_KEY = 'yvsearch'
 152
 153     def _get_n_results(self, query, n):
 154         """Get a specified number of results for a query"""
 155         entries = []
 156         for pagenum in itertools.count(0):
 157             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)
 158             info = self._download_json(result_url, query,
 159                 note='Downloading results page '+str(pagenum+1))
 160             m = info['m']
 161             results = info['results']
 162
 163             for (i, r) in enumerate(results):
 164                 if (pagenum * 30) + i >= n:
 165                     break
 166                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 167                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 168                 entries.append(e)
 169             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
 170                 break
 171
 172         return {
 173             '_type': 'playlist',
 174             'id': query,
 175             'entries': entries,
 176         }