fix: soundcloud: HTTP Error 401: Unauthorized

feat: soundcloud download playlist

src/you_get/common.py

@@ -1064,6 +1064,20 @@ def download_urls(
                 for part in parts:
                     os.remove(part)
 
+        elif ext == 'mp3':
+            try:
+                from .processor.ffmpeg import has_ffmpeg_installed
+
+                assert has_ffmpeg_installed()
+                from .processor.ffmpeg import ffmpeg_concat_mp3_to_mp3
+                ffmpeg_concat_mp3_to_mp3(parts, output_filepath)
+                print('Merged into %s' % output_filename)
+            except:
+                raise
+            else:
+                for part in parts:
+                    os.remove(part)
+
         else:
             print("Can't merge %s files" % ext)
 

src/you_get/extractors/soundcloud.py

@@ -1,44 +1,80 @@
 #!/usr/bin/env python
 
-__all__ = ['soundcloud_download', 'soundcloud_download_by_id']
+__all__ = ['sndcd_download']
 
 from ..common import *
+import re
 import json
 import urllib.error
 
-client_id = 'WKcQQdEZw7Oi01KqtHWxeVSxNyRzgT8M'
 
-def soundcloud_download_by_id(id, title=None, output_dir='.', merge=True, info_only=False):
+def get_sndcd_apikey():
-    assert title
+    home_page = get_content('https://soundcloud.com')
-    url = 'https://api.soundcloud.com/tracks/{}/{}?client_id={}'.format(id, 'stream', client_id)
+    js_url = re.findall(r'script crossorigin src="(.+?)"></script>', home_page)[-1]
 
-    type, ext, size = url_info(url)
+    client_id = get_content(js_url)
+    return re.search(r'client_id:"(.+?)"', client_id).group(1)
 
-    print_info(site_info, title, type, size)
+
+def get_resource_info(resource_url, client_id):
+    cont = get_content(resource_url, decoded=True)
+
+    x = re.escape('forEach(function(e){n(e)})}catch(t){}})},')
+    x = re.search(r'' + x + r'(.*)\);</script>', cont)
+
+    info = json.loads(x.group(1))[-1]['data'][0]
+
+    info = info['tracks'] if info.get('track_count') else [info]
+
+    ids = [i['id'] for i in info if i.get('comment_count') is None]
+    ids = list(map(str, ids))
+    ids_split = ['%2C'.join(ids[i:i+10]) for i in range(0, len(ids), 10)]
+    api_url = 'https://api-v2.soundcloud.com/tracks?ids={ids}&client_id={client_id}&%5Bobject%20Object%5D=&app_version=1584348206&app_locale=en'
+
+    res = []
+    for ids in ids_split:
+        uri = api_url.format(ids=ids, client_id=client_id)
+        cont = get_content(uri, decoded=True)
+        res += json.loads(cont)
+
+    res = iter(res)
+    info = [next(res) if i.get('comment_count') is None else i for i in info]
+
+    return info
+
+
+def sndcd_download(url, output_dir='.', merge=True, info_only=False, **kwargs):
+    client_id = get_sndcd_apikey()
+
+    r_info = get_resource_info(url, client_id)
+
+    for info in r_info:
+        title = info['title']
+        metadata = info.get('publisher_metadata')
+
+        transcodings = info['media']['transcodings']
+        sq = [i for i in transcodings if i['quality'] == 'sq']
+        hq = [i for i in transcodings if i['quality'] == 'hq']
+        # source url
+        surl = sq[0] if hq == [] else hq[0]
+        surl = surl['url']
+
+        uri = surl + '?client_id=' + client_id
+        r = get_content(uri)
+        surl = json.loads(r)['url']
+
+        m3u8 = get_content(surl)
+        # url list
+        urll = re.findall(r'http.*?(?=\n)', m3u8)
+
+        size = urls_size(urll)
+        print_info(site_info, title, 'audio/mpeg', size)
+        print(end='', flush=True)
 
         if not info_only:
-        download_urls([url], title, ext, size, output_dir, merge = merge)
+            download_urls(urll, title=title, ext='mp3', total_size=size, output_dir=output_dir, merge=True)
 
-def soundcloud_i1_api(track_id):
-    url = 'https://api.soundcloud.com/i1/tracks/{}/streams?client_id={}'.format(track_id, client_id)
-    return json.loads(get_content(url))['http_mp3_128_url']
-
-def soundcloud_download(url, output_dir='.', merge=True, info_only=False, **kwargs):
-    url = 'https://api.soundcloud.com/resolve.json?url={}&client_id={}'.format(url, client_id)
-    metadata = get_content(url)
-    info = json.loads(metadata)
-    title = info["title"]
-    real_url = info.get('download_url')
-    if real_url is None:
-        real_url = info.get('steram_url')
-    if real_url is None:
-        raise Exception('Cannot get media URI for {}'.format(url))
-    real_url = soundcloud_i1_api(info['id'])
-    mime, ext, size = url_info(real_url)
-    print_info(site_info, title, mime, size)
-    if not info_only:
-        download_urls([real_url], title, ext, size, output_dir, merge=merge)
 
 site_info = "SoundCloud.com"
-download = soundcloud_download
+download = sndcd_download
-download_playlist = playlist_not_supported('soundcloud')
+download_playlist = sndcd_download

src/you_get/processor/ffmpeg.py

@@ -180,6 +180,18 @@ def ffmpeg_concat_flv_to_mp4(files, output='output.mp4'):
     else:
         raise
 
+def ffmpeg_concat_mp3_to_mp3(files, output='output.mp3'):
+    print('Merging video parts... ', end="", flush=True)
+
+    files = 'concat:' + '|'.join(files)
+
+    params = [FFMPEG] + LOGLEVEL + ['-y']
+    params += ['-i', files, '-acodec', 'copy', output]
+
+    subprocess.call(params)
+
+    return True
+
 def ffmpeg_concat_mp4_to_mp4(files, output='output.mp4'):
     print('Merging video parts... ', end="", flush=True)
     # Use concat demuxer on FFmpeg >= 1.1

tests/test.py

@@ -8,7 +8,8 @@ from you_get.extractors import (
     youtube,
     missevan,
     acfun,
-    bilibili
+    bilibili,
+    soundcloud
 )
 
 
@@ -45,5 +46,16 @@ class YouGetTests(unittest.TestCase):
         bilibili.download(
             "https://www.bilibili.com/watchlater/#/av74906671/p6", info_only=True
         )
+
+    def test_soundcloud(self):
+        ## single song
+        soundcloud.download(
+            'https://soundcloud.com/keiny-pham/impure-bird', info_only=True
+        )
+        ## playlist
+        soundcloud.download(
+            'https://soundcloud.com/anthony-flieger/sets/cytus', info_only=True
+        )
+
 if __name__ == '__main__':
     unittest.main()