ytdl-nightly/youtube_dl/extractor/audiomack.py

126 lines
5.1 KiB
Python
Raw Normal View History

2014-10-23 23:55:39 +02:00
# coding: utf-8
from __future__ import unicode_literals
from .common import InfoExtractor
2014-10-24 06:54:59 +02:00
from .soundcloud import SoundcloudIE
2014-10-25 04:07:01 +02:00
from ..utils import ExtractorError
2014-10-26 23:13:42 +01:00
2014-10-23 23:55:39 +02:00
import time
class AudiomackIE(InfoExtractor):
2015-01-09 01:35:33 +01:00
_VALID_URL = r'https?://(?:www\.)?audiomack\.com/song/(?P<id>[\w/-]+)'
2014-10-24 06:54:59 +02:00
IE_NAME = 'audiomack'
_TESTS = [
2015-01-09 01:35:33 +01:00
# hosted on audiomack
2014-10-24 06:54:59 +02:00
{
'url': 'http://www.audiomack.com/song/roosh-williams/extraordinary',
'info_dict':
{
'id': '310086',
2015-01-09 01:35:33 +01:00
'ext': 'mp3',
'artist': 'Roosh Williams',
'title': 'Extraordinary'
2014-10-24 06:54:59 +02:00
}
},
2015-01-09 01:35:33 +01:00
# audiomack wrapper around soundcloud song
2014-10-24 06:54:59 +02:00
{
2014-12-03 14:53:12 +01:00
'add_ie': ['Soundcloud'],
2014-10-24 06:54:59 +02:00
'url': 'http://www.audiomack.com/song/xclusiveszone/take-kare',
'info_dict': {
'id': '172419696',
'ext': 'mp3',
2014-12-03 14:53:12 +01:00
'description': 'md5:1fc3272ed7a635cce5be1568c2822997',
2014-10-24 06:54:59 +02:00
'title': 'Young Thug ft Lil Wayne - Take Kare',
'uploader': 'Young Thug World',
'upload_date': '20141016',
2014-10-24 06:54:59 +02:00
}
2014-12-03 14:53:12 +01:00
},
2014-10-24 06:54:59 +02:00
]
2014-10-23 23:55:39 +02:00
@staticmethod
def create_song_dictionary(api_response, album_url_tag, track_no=0):
# All keys are the same in audiomack api and InfoExtractor format
2015-01-09 01:35:33 +01:00
entry = {key: api_response[key] for key in ['title', 'artist', 'id', 'url'] if key in api_response}
# Fudge values in the face of missing metadata
2015-01-09 01:35:33 +01:00
if 'id' not in entry:
entry['id'] = track_no
if 'title' not in entry:
entry['title'] = album_url_tag
return entry
2014-10-23 23:55:39 +02:00
def _real_extract(self, url):
# URLs end with [uploader name]/[uploader title]
# this title is whatever the user types in, and is rarely
# the proper song title. Real metadata is in the api response
album_url_tag = self._match_id(url)
2014-10-23 23:55:39 +02:00
# Request the extended version of the api for extra fields like artist and title
2014-10-25 08:58:03 +02:00
api_response = self._download_json(
2015-01-09 01:35:33 +01:00
'http://www.audiomack.com/api/music/url/song/%s?extended=1&_=%d' % (
album_url_tag, time.time()),
album_url_tag)
2014-10-25 04:07:01 +02:00
# API is inconsistent with errors
2015-01-09 01:35:33 +01:00
if 'url' not in api_response or not api_response['url'] or 'error' in api_response:
raise ExtractorError('Invalid url %s', url)
2014-10-24 06:54:59 +02:00
2014-11-23 20:41:03 +01:00
# Audiomack wraps a lot of soundcloud tracks in their branded wrapper
# if so, pass the work off to the soundcloud extractor
2015-01-09 01:35:33 +01:00
if SoundcloudIE.suitable(api_response['url']):
return {'_type': 'url', 'url': api_response['url'], 'ie_key': 'Soundcloud'}
2014-10-25 08:58:03 +02:00
return self.create_song_dictionary(api_response, album_url_tag)
2014-10-25 08:58:03 +02:00
class AudiomackAlbumIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?audiomack\.com/album/(?P<id>[\w/-]+)'
IE_NAME = 'audiomack:album'
_TESTS = [
# Standard album playlist
{
'url': 'http://www.audiomack.com/album/flytunezcom/tha-tour-part-2-mixtape',
2015-01-09 01:35:33 +01:00
'playlist_count': 15,
'info_dict':
{
2015-01-09 01:35:33 +01:00
'id': '812251',
'title': 'Tha Tour: Part 2 (Official Mixtape)'
}
},
# Album playlist ripped from fakeshoredrive with no metadata
{
2015-01-09 01:35:33 +01:00
'url': 'http://www.audiomack.com/album/fakeshoredrive/ppp-pistol-p-project',
'playlist_count': 10
2014-10-25 08:58:03 +02:00
}
]
def _real_extract(self, url):
# URLs end with [uploader name]/[uploader title]
# this title is whatever the user types in, and is rarely
# the proper song title. Real metadata is in the api response
album_url_tag = self._match_id(url)
2015-01-09 01:35:33 +01:00
result = {'_type': 'playlist', 'entries': []}
# There is no one endpoint for album metadata - instead it is included/repeated in each song's metadata
# Therefore we don't know how many songs the album has and must infi-loop until failure
track_no = 0
while True:
# Get song's metadata
2015-01-09 01:35:33 +01:00
api_response = self._download_json('http://www.audiomack.com/api/music/url/album/%s/%d?extended=1&_=%d'
% (album_url_tag, track_no, time.time()), album_url_tag)
# Total failure, only occurs when url is totally wrong
# Won't happen in middle of valid playlist (next case)
2015-01-09 01:35:33 +01:00
if 'url' not in api_response or 'error' in api_response:
raise ExtractorError('Invalid url for track %d of album url %s' % (track_no, url))
# URL is good but song id doesn't exist - usually means end of playlist
2015-01-09 01:35:33 +01:00
elif not api_response['url']:
break
else:
# Pull out the album metadata and add to result (if it exists)
2015-01-09 01:35:33 +01:00
for resultkey, apikey in [('id', 'album_id'), ('title', 'album_title')]:
if apikey in api_response and resultkey not in result:
result[resultkey] = api_response[apikey]
2015-01-09 01:35:33 +01:00
result['entries'].append(AudiomackIE.create_song_dictionary(api_response, album_url_tag, track_no))
track_no += 1
return result