_ 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.mp4',
  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'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'YVIDEO_INIT_ITEMS = ({.*?});$',
  50             webpage, u'items', flags=re.MULTILINE)
  51         items = json.loads(items_json)
  52         info = items['mediaItems']['query']['results']['mediaObj'][0]
  53         meta = info['meta']
  54
  55         formats = []
  56         for s in info['streams']:
  57             format_info = {
  58                 'width': s.get('width'),
  59                 'height': s.get('height'),
  60                 'bitrate': s.get('bitrate'),
  61             }
  62
  63             host = s['host']
  64             path = s['path']
  65             if host.startswith('rtmp'):
  66                 format_info.update({
  67                     'url': host,
  68                     'play_path': path,
  69                     'ext': 'flv',
  70                 })
  71             else:
  72                 format_url = compat_urlparse.urljoin(host, path)
  73                 format_info['url'] = format_url
  74                 format_info['ext'] = determine_ext(format_url)
  75
  76             formats.append(format_info)
  77         formats = sorted(formats, key=lambda f:(f['height'], f['width']))
  78
  79         info = {
  80             'id': video_id,
  81             'title': meta['title'],
  82             'formats': formats,
  83             'description': clean_html(meta['description']),
  84             'thumbnail': meta['thumbnail'],
  85         }
  86         # TODO: Remove when #980 has been merged
  87         info.update(formats[-1])
  88
  89         return info
  90
  91
  92 class YahooSearchIE(SearchInfoExtractor):
  93     IE_DESC = u'Yahoo screen search'
  94     _MAX_RESULTS = 1000
  95     IE_NAME = u'screen.yahoo:search'
  96     _SEARCH_KEY = 'yvsearch'
  97
  98     def _get_n_results(self, query, n):
  99         """Get a specified number of results for a query"""
 100
 101         res = {
 102             '_type': 'playlist',
 103             'id': query,
 104             'entries': []
 105         }
 106         for pagenum in itertools.count(0):
 107             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)
 108             webpage = self._download_webpage(result_url, query,
 109                                              note='Downloading results page '+str(pagenum+1))
 110             info = json.loads(webpage)
 111             m = info[u'm']
 112             results = info[u'results']
 113
 114             for (i, r) in enumerate(results):
 115                 if (pagenum * 30) +i >= n:
 116                     break
 117                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 118                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 119                 res['entries'].append(e)
 120             if (pagenum * 30 +i >= n) or (m[u'last'] >= (m[u'total'] -1 )):
 121                 break
 122
 123         return res