Raphaël G. Git Repositories - youtubedl/blob - 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'
  18     _VALID_URL = r'http://screen\.yahoo\.com/.*?-(?P<id>\d*?)\.html'
  19     _TESTS = [
  20         {
  21             'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
  22             'file': '214727115.mp4',
  23             'md5': '4962b075c08be8690a922ee026d05e69',
  24             'info_dict': {
  25                 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
  26                 'description': 'Julian and Travis watch Julian Smith',
  27             },
  28         },
  29         {
  30             'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
  31             'file': '103000935.mp4',
  32             'md5': 'd6e6fc6e1313c608f316ddad7b82b306',
  33             'info_dict': {
  34                 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
  35                 '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?',
  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'mediaItems: ({.*?})$',
  46             webpage, 'items', flags=re.MULTILINE)
  47         items = json.loads(items_json)
  48         info = items['mediaItems']['query']['results']['mediaObj'][0]
  49         # The 'meta' field is not always in the video webpage, we request it
  50         # from another page
  51         long_id = info['id']
  52         return self._get_info(long_id, video_id)
  53
  54     def _get_info(self, long_id, video_id):
  55         query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
  56                  ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="US"'
  57                  ' AND protocol="http"' % 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, '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': int_or_none(s.get('width')),
  74                 'height': int_or_none(s.get('height')),
  75                 'tbr': int_or_none(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
  90             formats.append(format_info)
  91
  92         self._sort_formats(formats)
  93
  94         return {
  95             'id': video_id,
  96             'title': meta['title'],
  97             'formats': formats,
  98             'description': clean_html(meta['description']),
  99             'thumbnail': meta['thumbnail'],
 100         }
 101
 102
 103 class YahooNewsIE(YahooIE):
 104     IE_NAME = 'yahoo:news'
 105     _VALID_URL = r'http://news\.yahoo\.com/video/.*?-(?P<id>\d*?)\.html'
 106
 107     _TEST = {
 108         'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
 109         'md5': '67010fdf3a08d290e060a4dd96baa07b',
 110         'info_dict': {
 111             'id': '104538833',
 112             'ext': 'mp4',
 113             'title': 'China Moses Is Crazy About the Blues',
 114             'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
 115         },
 116     }
 117
 118     # Overwrite YahooIE properties we don't want
 119     _TESTS = []
 120
 121     def _real_extract(self, url):
 122         mobj = re.match(self._VALID_URL, url)
 123         video_id = mobj.group('id')
 124         webpage = self._download_webpage(url, video_id)
 125         long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, 'long id')
 126         return self._get_info(long_id, video_id)
 127
 128
 129 class YahooSearchIE(SearchInfoExtractor):
 130     IE_DESC = 'Yahoo screen search'
 131     _MAX_RESULTS = 1000
 132     IE_NAME = 'screen.yahoo:search'
 133     _SEARCH_KEY = 'yvsearch'
 134
 135     def _get_n_results(self, query, n):
 136         """Get a specified number of results for a query"""
 137
 138         res = {
 139             '_type': 'playlist',
 140             'id': query,
 141             'entries': []
 142         }
 143         for pagenum in itertools.count(0):
 144             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)
 145             webpage = self._download_webpage(result_url, query,
 146                                              note='Downloading results page '+str(pagenum+1))
 147             info = json.loads(webpage)
 148             m = info['m']
 149             results = info['results']
 150
 151             for (i, r) in enumerate(results):
 152                 if (pagenum * 30) +i >= n:
 153                     break
 154                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
 155                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
 156                 res['entries'].append(e)
 157             if (pagenum * 30 +i >= n) or (m['last'] >= (m['total'] -1)):
 158                 break
 159
 160         return res