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.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'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         # The 'meta' field is not always in the video webpage, we request it
  54         # from another page
  55         long_id = info['id']
  56         query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
  57                  ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2"' % long_id)
  58         data = compat_urllib_parse.urlencode({
  59             'q': query,
  60             'env': 'prod',
  61             'format': 'json',
  62         })
  63         query_result_json = self._download_webpage(
  64             'http://video.query.yahoo.com/v1/public/yql?' + data,
  65             video_id, u'Downloading video info')
  66         query_result = json.loads(query_result_json)
  67         info = query_result['query']['results']['mediaObj'][0]
  68         meta = info['meta']
  69
  70         formats = []
  71         for s in info['streams']:
  72             format_info = {
  73                 'width': s.get('width'),
  74                 'height': s.get('height'),
  75                 'bitrate': s.get('bitrate'),
  76             }
  77
  78             host = s['host']
  79             path = s['path']
  80             if host.startswith('rtmp'):
  81                 format_info.update({
  82                     'url': host,
  83                     'play_path': path,
  84                     'ext': 'flv',
  85                 })
  86             else:
  87                 format_url = compat_urlparse.urljoin(host, path)
  88                 format_info['url'] = format_url
  89                 format_info['ext'] = determine_ext(format_url)
  90
  91             formats.append(format_info)
  92         formats = sorted(formats, key=lambda f:(f['height'], f['width']))
  93
  94         info = {
  95             'id': video_id,
  96             'title': meta['title'],
  97             'formats': formats,
  98             'description': clean_html(meta['description']),
  99             'thumbnail': meta['thumbnail'],
 100         }
 101         # TODO: Remove when #980 has been merged
 102         info.update(formats[-1])
 103
 104         return info
 105
 106
 107 class YahooSearchIE(SearchInfoExtractor):
 108     IE_DESC = u'Yahoo screen search'
 109     _MAX_RESULTS = 1000
 110     IE_NAME = u'screen.yahoo:search'
 111     _SEARCH_KEY = 'yvsearch'
 112
 113     def _get_n_results(self, query, n):
 114         """Get a specified number of results for a query"""
 115
 116         res = {
 117             '_type': 'playlist',
 118             'id': query,
 119             'entries': []
 120         }
 121         for pagenum in itertools.count(0):
 122             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)
 123             webpage = self._download_webpage(result_url, query,
 124                                              note='Downloading results page '+str(pagenum+1))
 125             info = json.loads(webpage)
 126             m = info[u'm']
 127             results = info[u'results']
 128
 129             for (i, r) in enumerate(results):
 130                 if (pagenum * 30) +i >= n:
 131                     break
 132                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 133                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 134                 res['entries'].append(e)
 135             if (pagenum * 30 +i >= n) or (m[u'last'] >= (m[u'total'] -1)):
 136                 break
 137
 138         return res