youtube_dl/extractor/bandcamp.py

   1 import json
   2 import re
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     ExtractorError,
   7 )
   8
   9
  10 class BandcampIE(InfoExtractor):
  11     _VALID_URL = r'http://.*?\.bandcamp\.com/track/(?P<title>.*)'
  12     _TEST = {
  13         u'url': u'http://youtube-dl.bandcamp.com/track/youtube-dl-test-song',
  14         u'file': u'1812978515.mp3',
  15         u'md5': u'cdeb30cdae1921719a3cbcab696ef53c',
  16         u'info_dict': {
  17             u"title": u"youtube-dl test song \"'/\\\u00e4\u21ad"
  18         },
  19         u'skip': u'There is a limit of 200 free downloads / month for the test song'
  20     }
  21
  22     def _real_extract(self, url):
  23         mobj = re.match(self._VALID_URL, url)
  24         title = mobj.group('title')
  25         webpage = self._download_webpage(url, title)
  26         # We get the link to the free download page
  27         m_download = re.search(r'freeDownloadPage: "(.*?)"', webpage)
  28         if m_download is None:
  29             raise ExtractorError(u'No free songs found')
  30
  31         download_link = m_download.group(1)
  32         id = re.search(r'var TralbumData = {(.*?)id: (?P<id>\d*?)$',
  33                        webpage, re.MULTILINE|re.DOTALL).group('id')
  34
  35         download_webpage = self._download_webpage(download_link, id,
  36                                                   'Downloading free downloads page')
  37         # We get the dictionary of the track from some javascrip code
  38         info = re.search(r'items: (.*?),$',
  39                          download_webpage, re.MULTILINE).group(1)
  40         info = json.loads(info)[0]
  41         # We pick mp3-320 for now, until format selection can be easily implemented.
  42         mp3_info = info[u'downloads'][u'mp3-320']
  43         # If we try to use this url it says the link has expired
  44         initial_url = mp3_info[u'url']
  45         re_url = r'(?P<server>http://(.*?)\.bandcamp\.com)/download/track\?enc=mp3-320&fsig=(?P<fsig>.*?)&id=(?P<id>.*?)&ts=(?P<ts>.*)$'
  46         m_url = re.match(re_url, initial_url)
  47         #We build the url we will use to get the final track url
  48         # This url is build in Bandcamp in the script download_bunde_*.js
  49         request_url = '%s/statdownload/track?enc=mp3-320&fsig=%s&id=%s&ts=%s&.rand=665028774616&.vrs=1' % (m_url.group('server'), m_url.group('fsig'), id, m_url.group('ts'))
  50         final_url_webpage = self._download_webpage(request_url, id, 'Requesting download url')
  51         # If we could correctly generate the .rand field the url would be
  52         #in the "download_url" key
  53         final_url = re.search(r'"retry_url":"(.*?)"', final_url_webpage).group(1)
  54
  55         track_info = {'id':id,
  56                       'title' : info[u'title'],
  57                       'ext' :   'mp3',
  58                       'url' :   final_url,
  59                       'thumbnail' : info[u'thumb_url'],
  60                       'uploader' :  info[u'artist']
  61                       }
  62
  63         return [track_info]