Added new IE for Grooveshark
authorKeith Beckman <kbeckman@becknet.com>
Tue, 20 May 2014 06:47:34 +0000 (02:47 -0400)
committerKeith Beckman <kbeckman@becknet.com>
Tue, 20 May 2014 06:47:34 +0000 (02:47 -0400)
youtube_dl/__init__.py
youtube_dl/extractor/__init__.py
youtube_dl/extractor/grooveshark.py [new file with mode: 0644]

index 4e657e297d1a5ddb2aac4a448a8a94a0ac4cf180..cd4e9b484601212af1e73b631dde6cb97ec30279 100644 (file)
@@ -56,6 +56,7 @@ __authors__  = (
     'Nicolas Évrard',
     'Jason Normore',
     'Hoje Lee',
+    'Keith Beckman'
 )
 
 __license__ = 'Public Domain'
index 3503c76b7c61515d6b42dae8a7afebdc5cfa6712..3f6c67fa568a8134ea2fceb3ee0a17e4133ec338 100644 (file)
@@ -109,6 +109,7 @@ from .gdcvault import GDCVaultIE
 from .generic import GenericIE
 from .googleplus import GooglePlusIE
 from .googlesearch import GoogleSearchIE
+from .grooveshark import GroovesharkIE
 from .hark import HarkIE
 from .helsinki import HelsinkiIE
 from .hentaistigma import HentaiStigmaIE
diff --git a/youtube_dl/extractor/grooveshark.py b/youtube_dl/extractor/grooveshark.py
new file mode 100644 (file)
index 0000000..6798add
--- /dev/null
@@ -0,0 +1,200 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import time
+import math
+import re
+from urlparse import urlparse, urlunparse, urldefrag
+from urllib import quote, urlencode
+from os.path import basename
+
+from .common import InfoExtractor
+from ..utils import ExtractorError, compat_urllib_request, compat_html_parser
+
+class GroovesharkHtmlParser(compat_html_parser.HTMLParser):
+    def __init__(self):
+        self._current_object = None
+        self.objects = []
+        compat_html_parser.HTMLParser.__init__(self)
+
+    def handle_starttag(self, tag, attrs):
+        attrs = dict((k, v) for k, v in attrs)
+        if tag == 'object':
+            self._current_object = { 'attrs': attrs, 'params': [] }
+        elif tag == 'param':
+            self._current_object['params'].append(attrs)
+    
+    def handle_endtag(self, tag):
+        if tag == 'object':
+            self.objects.append(self._current_object)
+            self._current_object = None
+
+    @classmethod
+    def extract_object_tags(cls, html):
+        p = cls()
+        p.feed(html)
+        p.close()
+        return p.objects
+
+class GroovesharkIE(InfoExtractor):
+    _VALID_URL = r'https?://(www\.)?grooveshark\.com/#!/s/([^/]+)/([^/]+)'
+    _TEST = {
+        'url': 'http://grooveshark.com/#!/s/Jolene+Tenth+Key+Remix+Ft+Will+Sessions/6SS1DW?src=5',
+        'md5': 'bbccc50b19daca23b8f961152c1dc95b',
+        'info_dict': {
+            'id': '6SS1DW',
+            'title': 'Jolene (Tenth Key Remix ft. Will Sessions)',
+            'ext': 'mp3',
+            'duration': 227
+        }
+    }
+
+    do_playerpage_request = True
+    do_bootstrap_request = True
+
+    def _parse_target(self, target):
+        uri = urlparse(target)
+        hash = uri.fragment[1:].split('?')[0]
+        token = basename(hash.rstrip('/'))
+        return (uri, hash, token)
+
+    def _build_bootstrap_url(self, target):
+        (uri, hash, token) = self._parse_target(target)
+        query = 'getCommunicationToken=1&hash=%s&%d' % (quote(hash, safe=''), self.ts)
+        return (urlunparse((uri.scheme, uri.netloc, '/preload.php', None, query, None)), token)
+
+    def _build_meta_url(self, target):
+        (uri, hash, token) = self._parse_target(target)
+        query = 'hash=%s&%d' % (quote(hash, safe=''), self.ts)
+        return (urlunparse((uri.scheme, uri.netloc, '/preload.php', None, query, None)), token)
+
+    def _build_stream_url(self, meta):
+        return urlunparse(('http', meta['streamKey']['ip'], '/stream.php', None, None, None))
+
+    def _build_swf_referer(self, target, obj):
+        (uri, _, _) = self._parse_target(target)
+        return urlunparse((uri.scheme, uri.netloc, obj['attrs']['data'], None, None, None))
+
+    def _transform_bootstrap(self, js):
+        return re.split('^\s*try\s*{', js, flags=re.M)[0] \
+                 .split(' = ', 1)[1].strip().rstrip(';')
+
+    def _transform_meta(self, js):
+        return js.split('\n')[0].split('=')[1].rstrip(';')
+
+    def _get_meta(self, target):
+        (meta_url, token) = self._build_meta_url(target)
+        self.to_screen('Metadata URL: %s' % meta_url)
+
+        headers = {'Referer': urldefrag(target)[0]}
+        req = compat_urllib_request.Request(meta_url, headers=headers)
+        res = self._download_json(req, token,
+                                  transform_source=self._transform_meta)
+
+        if 'getStreamKeyWithSong' not in res:
+            raise ExtractorError(
+                'Metadata not found. URL may be malformed, or Grooveshark API may have changed.')
+
+        if res['getStreamKeyWithSong'] is None:
+            raise ExtractorError(
+                'Metadata download failed, probably due to Grooveshark anti-abuse throttling. Wait at least an hour before retrying from this IP.',
+                expected=True)
+
+        return res['getStreamKeyWithSong']
+
+    def _get_bootstrap(self, target):
+        (bootstrap_url, token) = self._build_bootstrap_url(target)
+        
+        headers = {'Referer': urldefrag(target)[0]}
+        req = compat_urllib_request.Request(bootstrap_url, headers=headers)
+        res = self._download_json(req, token, fatal=False,
+                                  note='Downloading player bootstrap data',
+                                  errnote='Unable to download player bootstrap data',
+                                  transform_source=self._transform_bootstrap)
+        return res
+    
+    def _get_playerpage(self, target):
+        (_, _, token) = self._parse_target(target)
+        
+        res = self._download_webpage(
+            target, token,
+            note='Downloading player page',
+            errnote='Unable to download player page',
+            fatal=False)
+        
+        if res is not None:
+            o = GroovesharkHtmlParser.extract_object_tags(res)
+            return (res, [x for x in o if x['attrs']['id'] == 'jsPlayerEmbed'])
+        
+        return (res, None)
+    
+    def _real_extract(self, url):
+        (target_uri, _, token) = self._parse_target(url)
+    
+        # 1. Fill cookiejar by making a request to the player page
+        if self.do_playerpage_request:
+            (_, player_objs) = self._get_playerpage(url)
+            if player_objs is not None:
+                swf_referer = self._build_swf_referer(url, player_objs[0])
+                self.to_screen('SWF Referer: %s' % swf_referer)
+
+        # 2. Ask preload.php for swf bootstrap data to better mimic webapp
+        if self.do_bootstrap_request:
+            bootstrap = self._get_bootstrap(url)
+            self.to_screen('CommunicationToken: %s' % bootstrap['getCommunicationToken'])
+        
+        # 3. Ask preload.php for track metadata.
+        meta = self._get_meta(url)
+        
+        # 4. Construct stream request for track.
+        stream_url = self._build_stream_url(meta)
+        duration = int(math.ceil(float(meta['streamKey']['uSecs']) / 1000000))
+        post_dict = {'streamKey': meta['streamKey']['streamKey']}
+        post_data = urlencode(post_dict).encode('utf-8')
+        headers = {
+            'Content-Length': len(post_data),
+            'Content-Type': 'application/x-www-form-urlencoded'
+            }
+
+        if 'swf_referer' in locals():
+            headers['Referer'] = swf_referer        
+
+        req = compat_urllib_request.Request(streamurl, post_data, headers)
+        
+        info_dict = {
+            'id': token,
+            'title': meta['song']['Name'],
+            'http_method': 'POST',
+            'url': stream_url,
+            'ext': 'mp3',
+            'format': 'mp3 audio',
+            'duration': duration,
+            
+            'post_data': post_data,
+            'post_dict': post_dict,
+            'headers': headers,
+            'request': req
+            }
+        
+        if 'swf_referer' in locals():
+            info_dict['http_referer'] = swf_referer        
+        
+        return info_dict
+
+    def _real_initialize(self):
+        self.ts = int(time.time() * 1000)  # timestamp in millis
+
+    def _download_json(self, url_or_request, video_id,
+                       note=u'Downloading JSON metadata',
+                       errnote=u'Unable to download JSON metadata',
+                       fatal=True,
+                       transform_source=None):
+        try:
+            out = super(GroovesharkIE, self)._download_json(
+                url_or_request, video_id, note, errnote, transform_source)
+            return out
+        except ExtractorError as ee:
+            if fatal:
+                raise ee
+        return None
+