Fix regex to find throttle function name (#1222)

* Fix regex to find throttle function name

The javascript now stores the throttling function name in an array.
Fix https://github.com/pytube/pytube/issues/1218

* Fix array parsing

Strip whitespaces around symbol names for future-proofing.

The variable name might be "b" right now, but it could change in the
future.

pytube/cipher.py

@@ -263,9 +263,14 @@ def get_throttling_function_name(js: str) -> str:
     """
     function_patterns = [
         # https://github.com/ytdl-org/youtube-dl/issues/29326#issuecomment-865985377
-        # a.C&&(b=a.get("n"))&&(b=Dea(b),a.set("n",b))}};
-        # In above case, `Dea` is the relevant function name
-        r'a\.[A-Z]&&\(b=a\.get\("n"\)\)&&\(b=([^(]+)\(b\)',
+        # https://github.com/yt-dlp/yt-dlp/commit/48416bc4a8f1d5ff07d5977659cb8ece7640dcd8
+        # var Bpa = [iha];
+        # ...
+        # a.C && (b = a.get("n")) && (b = Bpa[0](b), a.set("n", b),
+        # Bpa.length || iha("")) }};
+        # In the above case, `iha` is the relevant function name
+        r'a\.[a-zA-Z]\s*&&\s*\([a-z]\s*=\s*a\.get\("n"\)\)\s*&&\s*'
+        r'\([a-z]\s*=\s*([a-zA-Z0-9$]{3})(\[\d+\])?\([a-z]\)',
     ]
     logger.debug('Finding throttling function name')
     for pattern in function_patterns:
@@ -273,7 +278,20 @@ def get_throttling_function_name(js: str) -> str:
         function_match = regex.search(js)
         if function_match:
             logger.debug("finished regex search, matched: %s", pattern)
-            return function_match.group(1)
+            if len(function_match.groups()) == 1:
+                return function_match.group(1)
+            idx = function_match.group(2)
+            if idx:
+                idx = idx.strip("[]")
+                array = re.search(
+                    r'var {nfunc}\s*=\s*(\[.+?\]);'.format(
+                        nfunc=function_match.group(1)),
+                    js
+                )
+                if array:
+                    array = array.group(1).strip("[]").split(",")
+                    array = [x.strip() for x in array]
+                    return array[int(idx)]
 
     raise RegexMatchError(
         caller="get_throttling_function_name", pattern="multiple"

tests/conftest.py

@@ -146,3 +146,17 @@ def channel_videos_html():
     )
     with gzip.open(file_path, 'rb') as f:
         return f.read().decode('utf-8')
+
+
+@pytest.fixture
+def base_js():
+    """Youtube base.js retrieved on 2022-02-04 from
+    https://www.youtube.com/watch?v=vmzxpUsN0uA
+    """
+    file_path = os.path.join(
+        os.path.dirname(os.path.realpath(__file__)),
+        "mocks",
+        "base.js.gz",
+    )
+    with gzip.open(file_path, 'rb') as f:
+        return f.read().decode('utf-8')

tests/test_cipher.py

@@ -77,3 +77,14 @@ def test_js_splice():
     for args, result in mapping.items():
         a = [1, 2, 3, 4]
         assert cipher.js_splice(a, *args) == result
+
+
+def test_get_throttling_function_name(base_js):
+    # Values expected as of 2022/02/04:
+    raw_var = r'var Apa=[hha]'
+    assert raw_var in base_js
+    raw_code = r'a.url="";a.C&&(b=a.get("n"))&&(b=Apa[0](b),a.set("n",b),'\
+               r'Apa.length||hha(""))}};'
+    assert raw_code in base_js
+    func_name = cipher.get_throttling_function_name(base_js)
+    assert func_name == "hha"