Raphaël G. Git Repositories - youtubedl/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         },
  26         {
  27             u'url': u'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  28             u'file': u'103000935.flv',
  29             u'info_dict': {
  30                 u'title': u'The Cougar Lies with Spanish Moss',
  31                 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?',
  32             },
  33             u'params': {
  34                 # Requires rtmpdump
  35                 u'skip_download': True,
  36             },
  37         },
  38     ]
  39
  40     def _real_extract(self, url):
  41         mobj = re.match(self._VALID_URL, url)
  42         video_id = mobj.group('id')
  43         webpage = self._download_webpage(url, video_id)
  44
  45         items_json = self._search_regex(r'YVIDEO_INIT_ITEMS = ({.*?});$',
  46             webpage, u'items', flags=re.MULTILINE)
  47         items = json.loads(items_json)
  48         info = items['mediaItems']['query']['results']['mediaObj'][0]
  49         meta = info['meta']
  50
  51         formats = []
  52         for s in info['streams']:
  53             format_info = {
  54                 'width': s.get('width'),
  55                 'height': s.get('height'),
  56                 'bitrate': s.get('bitrate'),
  57             }
  58
  59             host = s['host']
  60             path = s['path']
  61             if host.startswith('rtmp'):
  62                 format_info.update({
  63                     'url': host,
  64                     'play_path': path,
  65                     'ext': 'flv',
  66                 })
  67             else:
  68                 format_url = compat_urlparse.urljoin(host, path)
  69                 format_info['url'] = format_url
  70                 format_info['ext'] = determine_ext(format_url)
  71
  72             formats.append(format_info)
  73         formats = sorted(formats, key=lambda f:(f['height'], f['width']))
  74
  75         info = {
  76             'id': video_id,
  77             'title': meta['title'],
  78             'formats': formats,
  79             'description': clean_html(meta['description']),
  80             'thumbnail': meta['thumbnail'],
  81         }
  82         # TODO: Remove when #980 has been merged
  83         info.update(formats[-1])
  84
  85         return info
  86
  87
  88 class YahooSearchIE(SearchInfoExtractor):
  89     IE_DESC = u'Yahoo screen search'
  90     _MAX_RESULTS = 1000
  91     IE_NAME = u'screen.yahoo:search'
  92     _SEARCH_KEY = 'yvsearch'
  93
  94     def _get_n_results(self, query, n):
  95         """Get a specified number of results for a query"""
  96
  97         res = {
  98             '_type': 'playlist',
  99             'id': query,
 100             'entries': []
 101         }
 102         for pagenum in itertools.count(0):
 103             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)
 104             webpage = self._download_webpage(result_url, query,
 105                                              note='Downloading results page '+str(pagenum+1))
 106             info = json.loads(webpage)
 107             m = info[u'm']
 108             results = info[u'results']
 109
 110             for (i, r) in enumerate(results):
 111                 if (pagenum * 30) +i >= n:
 112                     break
 113                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 114                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 115                 res['entries'].append(e)
 116             if (pagenum * 30 +i >= n) or (m[u'last'] >= (m[u'total'] -1 )):
 117                 break
 118
 119         return res