youtube-dl/youtube_dl/extractor/grooveshark.py

# 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(stream_url, post_data, headers)

        info_dict = {
            'id': token,
            'title': meta['song']['Name'],
            'http_method': 'POST',
            'url': stream_url,
            'ext': 'mp3',
            'format': 'mp3 audio',
            'duration': duration,

            # various ways of supporting the download request.
            # remove keys unnecessary to the eventual post implementation
            '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