Support Youtube Shorts URL format

From yt-dl commit 48416bc

test/test_youtube_signature.py

@@ -82,6 +82,14 @@ _NSIG_TESTS = [
         'https://www.youtube.com/s/player/f1ca6900/player_ias.vflset/en_US/base.js',
         'cu3wyu6LQn2hse', 'jvxetvmlI9AN9Q',
     ),
+    (
+        'https://www.youtube.com/s/player/8040e515/player_ias.vflset/en_US/base.js',
+        'wvOFaY-yjgDuIEg5', 'HkfBFDHmgw4rsw',
+    ),
+    (
+        'https://www.youtube.com/s/player/e06dea74/player_ias.vflset/en_US/base.js',
+        'AiuodmaDDYw8d3y4bf', 'ankd8eza2T6Qmw',
+    ),
 ]
 
 
@@ -110,10 +118,17 @@ class TestPlayerInfo(unittest.TestCase):
 class TestSignature(unittest.TestCase):
     def setUp(self):
         TEST_DIR = os.path.dirname(os.path.abspath(__file__))
-        self.TESTDATA_DIR = os.path.join(TEST_DIR, 'testdata')
+        self.TESTDATA_DIR = os.path.join(TEST_DIR, 'testdata/sigs')
         if not os.path.exists(self.TESTDATA_DIR):
             os.mkdir(self.TESTDATA_DIR)
 
+    def tearDown(self):
+        try:
+            for f in os.listdir(self.TESTDATA_DIR):
+                os.remove(f)
+        except OSError:
+            pass
+
 
 def t_factory(name, sig_func, url_pattern):
     def make_tfunc(url, sig_input, expected_sig):
@@ -145,12 +160,7 @@ def signature(jscode, sig_input):
 
 
 def n_sig(jscode, sig_input):
-    # Pending implementation of _extract_n_function_name() or similar in
-    # youtube.py, hard-code here
-    # funcname = YoutubeIE(FakeYDL())._extract_n_function_name(jscode)
-    import re
-    funcname = re.search(r'[=(,&|](\w+)\(\w+\),\w+\.set\("n",', jscode)
-    funcname = funcname and funcname.group(1)
+    funcname = YoutubeIE(FakeYDL())._extract_n_function_name(jscode)
     return JSInterpreter(jscode).call_function(funcname, sig_input)
 
 

youtube_dl/extractor/youtube.py

@@ -28,6 +28,7 @@ from ..utils import (
     dict_get,
     float_or_none,
     int_or_none,
+    js_to_json,
     mimetype2ext,
     parse_codecs,
     parse_duration,
@@ -416,6 +417,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
                          (?:                                                  # the various things that can precede the ID:
                              (?:(?:v|embed|e)/(?!videoseries))                # v/ or embed/ or e/
+                             |shorts/
                              |(?:                                             # or the v= param in all its forms
                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)?  # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
@@ -1118,6 +1120,22 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'skip_download': True,
             },
         },
+        {
+            # YT 'Shorts'
+            'url': 'https://youtube.com/shorts/4L2J27mJ3Dc',
+            'info_dict': {
+                'id': '4L2J27mJ3Dc',
+                'ext': 'mp4',
+                'upload_date': '20211025',
+                'uploader': 'Charlie Berens',
+                'description': 'md5:976512b8a29269b93bbd8a61edc45a6d',
+                'uploader_id': 'fivedlrmilkshake',
+                'title': 'Midwest Squid Game #Shorts',
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },
     ]
     _formats = {
         '5': {'ext': 'flv', 'width': 400, 'height': 240, 'acodec': 'mp3', 'abr': 64, 'vcodec': 'h263'},
@@ -1391,9 +1409,16 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
     # 2. https://code.videolan.org/videolan/vlc/-/blob/4fb284e5af69aa9ac2100ccbdd3b88debec9987f/share/lua/playlist/youtube.lua#L116
     # 3. https://github.com/ytdl-org/youtube-dl/issues/30097#issuecomment-950157377
     def _extract_n_function_name(self, jscode):
-        return self._search_regex(
-            (r'\.get\("n"\)\)&&\(b=(?P<nfunc>[a-zA-Z0-9$]{3})\([a-zA-Z0-9]\)',),
-            jscode, 'Initial JS player n function name', group='nfunc')
+        target = r'(?P<nfunc>[a-zA-Z0-9$]{3})(?:\[(?P<idx>\d+)\])?'
+        nfunc_and_idx = self._search_regex(
+            r'\.get\("n"\)\)&&\(b=(%s)\([a-zA-Z0-9]\)' % (target, ),
+            jscode, 'Initial JS player n function name')
+        nfunc, idx = re.match(target, nfunc_and_idx).group('nfunc', 'idx')
+        if not idx:
+            return nfunc
+        return self._parse_json(self._search_regex(
+            r'var %s\s*=\s*(\[.+?\]);' % (nfunc, ), jscode,
+            'Initial JS player n function list ({nfunc}[{idx}])'.format(**locals())), nfunc, transform_source=js_to_json)[int(idx)]
 
     def _extract_n_function(self, video_id, player_url):
         player_id = self._extract_player_info(player_url)