_ Git - youtube-dl/blob - youtube_dl/extractor/yahoo.py

   1 import itertools
   2 import json
   3 import re
   4
   5 from .common import InfoExtractor, SearchInfoExtractor
   6 from ..utils import (
   7     compat_urllib_parse,
   8     compat_urlparse,
   9     determine_ext,
  10     clean_html,
  11 )
  12
  13
  14 class YahooIE(InfoExtractor):
  15     IE_DESC = u'Yahoo screen'
  16     _VALID_URL = r'http://screen\.yahoo\.com/.*?-(?P<id>\d*?)\.html'
  17     _TESTS = [
  18         {
  19             u'url': u'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
  20             u'file': u'214727115.flv',
  21             u'info_dict': {
  22                 u'title': u'Julian Smith & Travis Legg Watch Julian Smith',
  23                 u'description': u'Julian and Travis watch Julian Smith',
  24             },
  25             u'params': {
  26                 # Requires rtmpdump
  27                 u'skip_download': True,
  28             },
  29         },
  30         {
  31             u'url': u'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  32             u'file': u'103000935.flv',
  33             u'info_dict': {
  34                 u'title': u'Codefellas - The Cougar Lies with Spanish Moss',
  35                 u'description': u'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?',
  36             },
  37             u'params': {
  38                 # Requires rtmpdump
  39                 u'skip_download': True,
  40             },
  41         },
  42     ]
  43
  44     def _real_extract(self, url):
  45         mobj = re.match(self._VALID_URL, url)
  46         video_id = mobj.group('id')
  47         webpage = self._download_webpage(url, video_id)
  48
  49         items_json = self._search_regex(r'mediaItems: ({.*?})$',
  50             webpage, u'items', flags=re.MULTILINE)
  51         items = json.loads(items_json)
  52         info = items['mediaItems']['query']['results']['mediaObj'][0]
  53         # The 'meta' field is not always in the video webpage, we request it
  54         # from another page
  55         long_id = info['id']
  56         return self._get_info(info['id'], video_id)
  57
  58     def _get_info(self, long_id, video_id):
  59         query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
  60                  ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="US"' % long_id)
  61         data = compat_urllib_parse.urlencode({
  62             'q': query,
  63             'env': 'prod',
  64             'format': 'json',
  65         })
  66         query_result_json = self._download_webpage(
  67             'http://video.query.yahoo.com/v1/public/yql?' + data,
  68             video_id, u'Downloading video info')
  69         query_result = json.loads(query_result_json)
  70         info = query_result['query']['results']['mediaObj'][0]
  71         meta = info['meta']
  72
  73         formats = []
  74         for s in info['streams']:
  75             format_info = {
  76                 'width': s.get('width'),
  77                 'height': s.get('height'),
  78                 'bitrate': s.get('bitrate'),
  79             }
  80
  81             host = s['host']
  82             path = s['path']
  83             if host.startswith('rtmp'):
  84                 format_info.update({
  85                     'url': host,
  86                     'play_path': path,
  87                     'ext': 'flv',
  88                 })
  89             else:
  90                 format_url = compat_urlparse.urljoin(host, path)
  91                 format_info['url'] = format_url
  92                 format_info['ext'] = determine_ext(format_url)
  93
  94             formats.append(format_info)
  95         formats = sorted(formats, key=lambda f:(f['height'], f['width']))
  96
  97         return {
  98             'id': video_id,
  99             'title': meta['title'],
 100             'formats': formats,
 101             'description': clean_html(meta['description']),
 102             'thumbnail': meta['thumbnail'],
 103         }
 104
 105
 106 class YahooNewsIE(YahooIE):
 107     IE_NAME = 'yahoo:news'
 108     _VALID_URL = r'http://news\.yahoo\.com/video/.*?-(?P<id>\d*?)\.html'
 109
 110     _TEST = {
 111         u'url': u'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 112         u'info_dict': {
 113             u'id': u'104538833',
 114             u'ext': u'flv',
 115             u'title': u'China Moses Is Crazy About the Blues',
 116             u'description': u'md5:9900ab8cd5808175c7b3fe55b979bed0',
 117         },
 118         u'params': {
 119             # Requires rtmpdump
 120             u'skip_download': True,
 121         },
 122     }
 123
 124     # Overwrite YahooIE properties we don't want
 125     _TESTS = []
 126
 127     def _real_extract(self, url):
 128         mobj = re.match(self._VALID_URL, url)
 129         video_id = mobj.group('id')
 130         webpage = self._download_webpage(url, video_id)
 131         long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, u'long id')
 132         return self._get_info(long_id, video_id)
 133
 134
 135 class YahooSearchIE(SearchInfoExtractor):
 136     IE_DESC = u'Yahoo screen search'
 137     _MAX_RESULTS = 1000
 138     IE_NAME = u'screen.yahoo:search'
 139     _SEARCH_KEY = 'yvsearch'
 140
 141     def _get_n_results(self, query, n):
 142         """Get a specified number of results for a query"""
 143
 144         res = {
 145             '_type': 'playlist',
 146             'id': query,
 147             'entries': []
 148         }
 149         for pagenum in itertools.count(0):
 150             result_url = u'http://video.search.yahoo.com/search/?p=%s&fr=screen&o=js&gs=0&b=%d' % (compat_urllib_parse.quote_plus(query), pagenum * 30)
 151             webpage = self._download_webpage(result_url, query,
 152                                              note='Downloading results page '+str(pagenum+1))
 153             info = json.loads(webpage)
 154             m = info[u'm']
 155             results = info[u'results']
 156
 157             for (i, r) in enumerate(results):
 158                 if (pagenum * 30) +i >= n:
 159                     break
 160                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 161                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 162                 res['entries'].append(e)
 163             if (pagenum * 30 +i >= n) or (m[u'last'] >= (m[u'total'] -1)):
 164                 break
 165
 166         return res