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