Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/grooveshark.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import time
   5 import math
   6 import os.path
   7 import re
   8
   9
  10 from .common import InfoExtractor
  11 from ..compat import (
  12     compat_html_parser,
  13     compat_urllib_parse,
  14     compat_urllib_request,
  15     compat_urlparse,
  16 )
  17 from ..utils import ExtractorError
  18
  19
  20 class GroovesharkHtmlParser(compat_html_parser.HTMLParser):
  21     def __init__(self):
  22         self._current_object = None
  23         self.objects = []
  24         compat_html_parser.HTMLParser.__init__(self)
  25
  26     def handle_starttag(self, tag, attrs):
  27         attrs = dict((k, v) for k, v in attrs)
  28         if tag == 'object':
  29             self._current_object = {'attrs': attrs, 'params': []}
  30         elif tag == 'param':
  31             self._current_object['params'].append(attrs)
  32
  33     def handle_endtag(self, tag):
  34         if tag == 'object':
  35             self.objects.append(self._current_object)
  36             self._current_object = None
  37
  38     @classmethod
  39     def extract_object_tags(cls, html):
  40         p = cls()
  41         p.feed(html)
  42         p.close()
  43         return p.objects
  44
  45
  46 class GroovesharkIE(InfoExtractor):
  47     _VALID_URL = r'https?://(www\.)?grooveshark\.com/#!/s/([^/]+)/([^/]+)'
  48     _TEST = {
  49         'url': 'http://grooveshark.com/#!/s/Jolene+Tenth+Key+Remix+Ft+Will+Sessions/6SS1DW?src=5',
  50         'md5': '7ecf8aefa59d6b2098517e1baa530023',
  51         'info_dict': {
  52             'id': '6SS1DW',
  53             'title': 'Jolene (Tenth Key Remix ft. Will Sessions)',
  54             'ext': 'mp3',
  55             'duration': 227,
  56         }
  57     }
  58
  59     do_playerpage_request = True
  60     do_bootstrap_request = True
  61
  62     def _parse_target(self, target):
  63         uri = compat_urlparse.urlparse(target)
  64         hash = uri.fragment[1:].split('?')[0]
  65         token = os.path.basename(hash.rstrip('/'))
  66         return (uri, hash, token)
  67
  68     def _build_bootstrap_url(self, target):
  69         (uri, hash, token) = self._parse_target(target)
  70         query = 'getCommunicationToken=1&hash=%s&%d' % (compat_urllib_parse.quote(hash, safe=''), self.ts)
  71         return (compat_urlparse.urlunparse((uri.scheme, uri.netloc, '/preload.php', None, query, None)), token)
  72
  73     def _build_meta_url(self, target):
  74         (uri, hash, token) = self._parse_target(target)
  75         query = 'hash=%s&%d' % (compat_urllib_parse.quote(hash, safe=''), self.ts)
  76         return (compat_urlparse.urlunparse((uri.scheme, uri.netloc, '/preload.php', None, query, None)), token)
  77
  78     def _build_stream_url(self, meta):
  79         return compat_urlparse.urlunparse(('http', meta['streamKey']['ip'], '/stream.php', None, None, None))
  80
  81     def _build_swf_referer(self, target, obj):
  82         (uri, _, _) = self._parse_target(target)
  83         return compat_urlparse.urlunparse((uri.scheme, uri.netloc, obj['attrs']['data'], None, None, None))
  84
  85     def _transform_bootstrap(self, js):
  86         return re.split('(?m)^\s*try\s*\{', js)[0] \
  87                  .split(' = ', 1)[1].strip().rstrip(';')
  88
  89     def _transform_meta(self, js):
  90         return js.split('\n')[0].split('=')[1].rstrip(';')
  91
  92     def _get_meta(self, target):
  93         (meta_url, token) = self._build_meta_url(target)
  94         self.to_screen('Metadata URL: %s' % meta_url)
  95
  96         headers = {'Referer': compat_urlparse.urldefrag(target)[0]}
  97         req = compat_urllib_request.Request(meta_url, headers=headers)
  98         res = self._download_json(req, token,
  99                                   transform_source=self._transform_meta)
 100
 101         if 'getStreamKeyWithSong' not in res:
 102             raise ExtractorError(
 103                 'Metadata not found. URL may be malformed, or Grooveshark API may have changed.')
 104
 105         if res['getStreamKeyWithSong'] is None:
 106             raise ExtractorError(
 107                 'Metadata download failed, probably due to Grooveshark anti-abuse throttling. Wait at least an hour before retrying from this IP.',
 108                 expected=True)
 109
 110         return res['getStreamKeyWithSong']
 111
 112     def _get_bootstrap(self, target):
 113         (bootstrap_url, token) = self._build_bootstrap_url(target)
 114
 115         headers = {'Referer': compat_urlparse.urldefrag(target)[0]}
 116         req = compat_urllib_request.Request(bootstrap_url, headers=headers)
 117         res = self._download_json(req, token, fatal=False,
 118                                   note='Downloading player bootstrap data',
 119                                   errnote='Unable to download player bootstrap data',
 120                                   transform_source=self._transform_bootstrap)
 121         return res
 122
 123     def _get_playerpage(self, target):
 124         (_, _, token) = self._parse_target(target)
 125
 126         webpage = self._download_webpage(
 127             target, token,
 128             note='Downloading player page',
 129             errnote='Unable to download player page',
 130             fatal=False)
 131
 132         if webpage is not None:
 133             # Search (for example German) error message
 134             error_msg = self._html_search_regex(
 135                 r'<div id="content">\s*<h2>(.*?)</h2>', webpage,
 136                 'error message', default=None)
 137             if error_msg is not None:
 138                 error_msg = error_msg.replace('\n', ' ')
 139                 raise ExtractorError('Grooveshark said: %s' % error_msg)
 140
 141         if webpage is not None:
 142             o = GroovesharkHtmlParser.extract_object_tags(webpage)
 143             return (webpage, [x for x in o if x['attrs']['id'] == 'jsPlayerEmbed'])
 144
 145         return (webpage, None)
 146
 147     def _real_initialize(self):
 148         self.ts = int(time.time() * 1000)  # timestamp in millis
 149
 150     def _real_extract(self, url):
 151         (target_uri, _, token) = self._parse_target(url)
 152
 153         # 1. Fill cookiejar by making a request to the player page
 154         swf_referer = None
 155         if self.do_playerpage_request:
 156             (_, player_objs) = self._get_playerpage(url)
 157             if player_objs is not None:
 158                 swf_referer = self._build_swf_referer(url, player_objs[0])
 159                 self.to_screen('SWF Referer: %s' % swf_referer)
 160
 161         # 2. Ask preload.php for swf bootstrap data to better mimic webapp
 162         if self.do_bootstrap_request:
 163             bootstrap = self._get_bootstrap(url)
 164             self.to_screen('CommunicationToken: %s' % bootstrap['getCommunicationToken'])
 165
 166         # 3. Ask preload.php for track metadata.
 167         meta = self._get_meta(url)
 168
 169         # 4. Construct stream request for track.
 170         stream_url = self._build_stream_url(meta)
 171         duration = int(math.ceil(float(meta['streamKey']['uSecs']) / 1000000))
 172         post_dict = {'streamKey': meta['streamKey']['streamKey']}
 173         post_data = compat_urllib_parse.urlencode(post_dict).encode('utf-8')
 174         headers = {
 175             'Content-Length': len(post_data),
 176             'Content-Type': 'application/x-www-form-urlencoded'
 177         }
 178         if swf_referer is not None:
 179             headers['Referer'] = swf_referer
 180
 181         return {
 182             'id': token,
 183             'title': meta['song']['Name'],
 184             'http_method': 'POST',
 185             'url': stream_url,
 186             'ext': 'mp3',
 187             'format': 'mp3 audio',
 188             'duration': duration,
 189             'http_post_data': post_data,
 190             'http_headers': headers,
 191         }