1
1
mirror of https://github.com/ytdl-org/youtube-dl synced 2024-12-27 12:35:52 +01:00
youtube-dl/youtube_dl/extractor/grooveshark.py

192 lines
7.0 KiB
Python
Raw Normal View History

2014-05-20 08:47:34 +02:00
# coding: utf-8
from __future__ import unicode_literals
import time
import math
2014-08-24 02:06:59 +02:00
import os.path
2014-05-20 08:47:34 +02:00
import re
2014-05-20 08:47:34 +02:00
from .common import InfoExtractor
2014-11-04 23:51:01 +01:00
from ..compat import (
compat_html_parser,
2014-08-24 02:06:59 +02:00
compat_urllib_parse,
2014-11-04 23:51:01 +01:00
compat_urllib_request,
2014-08-24 02:06:59 +02:00
compat_urlparse,
)
2014-11-04 23:51:01 +01:00
from ..utils import ExtractorError
2014-05-20 08:55:21 +02:00
2014-05-20 08:47:34 +02:00
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':
2014-05-20 08:55:21 +02:00
self._current_object = {'attrs': attrs, 'params': []}
2014-05-20 08:47:34 +02:00
elif tag == 'param':
self._current_object['params'].append(attrs)
2014-05-20 08:55:21 +02:00
2014-05-20 08:47:34 +02:00
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
2014-05-20 08:55:21 +02:00
2014-05-20 08:47:34 +02:00
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',
2014-08-24 01:32:12 +02:00
'md5': '7ecf8aefa59d6b2098517e1baa530023',
2014-05-20 08:47:34 +02:00
'info_dict': {
'id': '6SS1DW',
'title': 'Jolene (Tenth Key Remix ft. Will Sessions)',
'ext': 'mp3',
'duration': 227,
2014-05-20 08:47:34 +02:00
}
}
do_playerpage_request = True
do_bootstrap_request = True
def _parse_target(self, target):
uri = compat_urlparse.urlparse(target)
2014-05-20 08:47:34 +02:00
hash = uri.fragment[1:].split('?')[0]
2014-08-24 02:06:59 +02:00
token = os.path.basename(hash.rstrip('/'))
2014-05-20 08:47:34 +02:00
return (uri, hash, token)
def _build_bootstrap_url(self, target):
(uri, hash, token) = self._parse_target(target)
2014-08-24 02:06:59 +02:00
query = 'getCommunicationToken=1&hash=%s&%d' % (compat_urllib_parse.quote(hash, safe=''), self.ts)
return (compat_urlparse.urlunparse((uri.scheme, uri.netloc, '/preload.php', None, query, None)), token)
2014-05-20 08:47:34 +02:00
def _build_meta_url(self, target):
(uri, hash, token) = self._parse_target(target)
2014-08-24 02:06:59 +02:00
query = 'hash=%s&%d' % (compat_urllib_parse.quote(hash, safe=''), self.ts)
return (compat_urlparse.urlunparse((uri.scheme, uri.netloc, '/preload.php', None, query, None)), token)
2014-05-20 08:47:34 +02:00
def _build_stream_url(self, meta):
2014-08-24 02:06:59 +02:00
return compat_urlparse.urlunparse(('http', meta['streamKey']['ip'], '/stream.php', None, None, None))
2014-05-20 08:47:34 +02:00
def _build_swf_referer(self, target, obj):
(uri, _, _) = self._parse_target(target)
2014-08-24 02:06:59 +02:00
return compat_urlparse.urlunparse((uri.scheme, uri.netloc, obj['attrs']['data'], None, None, None))
2014-05-20 08:47:34 +02:00
def _transform_bootstrap(self, js):
return re.split('(?m)^\s*try\s*\{', js)[0] \
2014-05-20 08:47:34 +02:00
.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)
2014-08-24 02:06:59 +02:00
headers = {'Referer': compat_urlparse.urldefrag(target)[0]}
2014-05-20 08:47:34 +02:00
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)
2014-05-20 08:55:21 +02:00
2014-08-24 02:06:59 +02:00
headers = {'Referer': compat_urlparse.urldefrag(target)[0]}
2014-05-20 08:47:34 +02:00
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
2014-05-20 08:55:21 +02:00
2014-05-20 08:47:34 +02:00
def _get_playerpage(self, target):
(_, _, token) = self._parse_target(target)
2014-05-20 08:55:21 +02:00
webpage = self._download_webpage(
2014-05-20 08:47:34 +02:00
target, token,
note='Downloading player page',
errnote='Unable to download player page',
fatal=False)
2014-05-20 08:55:21 +02:00
if webpage is not None:
# Search (for example German) error message
error_msg = self._html_search_regex(
r'<div id="content">\s*<h2>(.*?)</h2>', webpage,
'error message', default=None)
if error_msg is not None:
error_msg = error_msg.replace('\n', ' ')
raise ExtractorError('Grooveshark said: %s' % error_msg)
2014-05-20 08:55:21 +02:00
if webpage is not None:
o = GroovesharkHtmlParser.extract_object_tags(webpage)
2015-03-17 18:13:43 +01:00
return webpage, [x for x in o if x['attrs']['id'] == 'jsPlayerEmbed']
2015-03-17 18:13:43 +01:00
return webpage, None
def _real_initialize(self):
self.ts = int(time.time() * 1000) # timestamp in millis
2014-05-20 08:55:21 +02:00
2014-05-20 08:47:34 +02:00
def _real_extract(self, url):
(target_uri, _, token) = self._parse_target(url)
2014-05-20 08:55:21 +02:00
2014-05-20 08:47:34 +02:00
# 1. Fill cookiejar by making a request to the player page
swf_referer = None
2014-05-20 08:47:34 +02:00
if self.do_playerpage_request:
(_, player_objs) = self._get_playerpage(url)
2015-03-17 18:13:43 +01:00
if player_objs:
2014-05-20 08:47:34 +02:00
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'])
2014-05-20 08:55:21 +02:00
2014-05-20 08:47:34 +02:00
# 3. Ask preload.php for track metadata.
meta = self._get_meta(url)
2014-05-20 08:55:21 +02:00
2014-05-20 08:47:34 +02:00
# 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']}
2014-08-24 02:06:59 +02:00
post_data = compat_urllib_parse.urlencode(post_dict).encode('utf-8')
2014-05-20 08:47:34 +02:00
headers = {
'Content-Length': len(post_data),
'Content-Type': 'application/x-www-form-urlencoded'
2014-05-20 08:55:21 +02:00
}
if swf_referer is not None:
2014-05-20 08:55:21 +02:00
headers['Referer'] = swf_referer
return {
2014-05-20 08:47:34 +02:00
'id': token,
'title': meta['song']['Name'],
'http_method': 'POST',
'url': stream_url,
'ext': 'mp3',
'format': 'mp3 audio',
'duration': duration,
'http_post_data': post_data,
'http_headers': headers,
2014-05-20 08:55:21 +02:00
}