gh-118263: Generalize path_t for C level optimizations

Lib/ntpath.py

@@ -168,19 +168,12 @@ def splitdrive(p):
 
 
 try:
-    from nt import _path_splitroot_ex
+    from nt import _path_splitroot_ex as splitroot
 except ImportError:
     def splitroot(p):
-        """Split a pathname into drive, root and tail. The drive is defined
-        exactly as in splitdrive(). On Windows, the root may be a single path
-        separator or an empty string. The tail contains anything after the root.
-        For example:
-
-            splitroot('//server/share/') == ('//server/share', '/', '')
-            splitroot('C:/Users/Barney') == ('C:', '/', 'Users/Barney')
-            splitroot('C:///spam///ham') == ('C:', '/', '//spam///ham')
-            splitroot('Windows/notepad') == ('', '', 'Windows/notepad')
-        """
+        """Split a pathname into drive, root and tail.
+
+        The tail contains anything after the root."""
         p = os.fspath(p)
         if isinstance(p, bytes):
             sep = b'\\'
@@ -220,23 +213,6 @@ def splitroot(p):
         else:
             # Relative path, e.g. Windows
             return empty, empty, p
-else:
-    def splitroot(p):
-        """Split a pathname into drive, root and tail. The drive is defined
-        exactly as in splitdrive(). On Windows, the root may be a single path
-        separator or an empty string. The tail contains anything after the root.
-        For example:
-
-            splitroot('//server/share/') == ('//server/share', '/', '')
-            splitroot('C:/Users/Barney') == ('C:', '/', 'Users/Barney')
-            splitroot('C:///spam///ham') == ('C:', '/', '//spam///ham')
-            splitroot('Windows/notepad') == ('', '', 'Windows/notepad')
-        """
-        p = os.fspath(p)
-        if isinstance(p, bytes):
-            drive, root, tail = _path_splitroot_ex(os.fsdecode(p))
-            return os.fsencode(drive), os.fsencode(root), os.fsencode(tail)
-        return _path_splitroot_ex(p)
 
 
 # Split a path in head (everything up to the last '/') and tail (the
@@ -538,7 +514,7 @@ def expandvars(path):
 # Previously, this function also truncated pathnames to 8+3 format,
 # but as this module is called "ntpath", that's obviously wrong!
 try:
-    from nt import _path_normpath
+    from nt import _path_normpath as normpath
 
 except ImportError:
     def normpath(path):
@@ -577,14 +553,6 @@ def normpath(path):
             comps.append(curdir)
         return prefix + sep.join(comps)
 
-else:
-    def normpath(path):
-        """Normalize path, eliminating double slashes, etc."""
-        path = os.fspath(path)
-        if isinstance(path, bytes):
-            return os.fsencode(_path_normpath(os.fsdecode(path))) or b"."
-        return _path_normpath(path) or "."
-
 
 def _abspath_fallback(path):
     """Return the absolute version of a path as a fallback function in case

Lib/posixpath.py

@@ -135,18 +135,12 @@ def splitdrive(p):
 
 
 try:
-    from posix import _path_splitroot_ex
+    from posix import _path_splitroot_ex as splitroot
 except ImportError:
     def splitroot(p):
-        """Split a pathname into drive, root and tail. On Posix, drive is always
-        empty; the root may be empty, a single slash, or two slashes. The tail
-        contains anything after the root. For example:
-
-            splitroot('foo/bar') == ('', '', 'foo/bar')
-            splitroot('/foo/bar') == ('', '/', 'foo/bar')
-            splitroot('//foo/bar') == ('', '//', 'foo/bar')
-            splitroot('///foo/bar') == ('', '/', '//foo/bar')
-        """
+        """Split a pathname into drive, root and tail.
+
+        The tail contains anything after the root."""
         p = os.fspath(p)
         if isinstance(p, bytes):
             sep = b'/'
@@ -164,23 +158,6 @@ def splitroot(p):
             # Precisely two leading slashes, e.g.: '//foo'. Implementation defined per POSIX, see
             # https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap04.html#tag_04_13
             return empty, p[:2], p[2:]
-else:
-    def splitroot(p):
-        """Split a pathname into drive, root and tail. On Posix, drive is always
-        empty; the root may be empty, a single slash, or two slashes. The tail
-        contains anything after the root. For example:
-
-            splitroot('foo/bar') == ('', '', 'foo/bar')
-            splitroot('/foo/bar') == ('', '/', 'foo/bar')
-            splitroot('//foo/bar') == ('', '//', 'foo/bar')
-            splitroot('///foo/bar') == ('', '/', '//foo/bar')
-        """
-        p = os.fspath(p)
-        if isinstance(p, bytes):
-            # Optimisation: the drive is always empty
-            _, root, tail = _path_splitroot_ex(os.fsdecode(p))
-            return b'', os.fsencode(root), os.fsencode(tail)
-        return _path_splitroot_ex(p)
 
 
 # Return the tail (basename) part of a path, same as split(path)[1].
@@ -363,7 +340,7 @@ def expandvars(path):
 # if it contains symbolic links!
 
 try:
-    from posix import _path_normpath
+    from posix import _path_normpath as normpath
 
 except ImportError:
     def normpath(path):
@@ -394,14 +371,6 @@ def normpath(path):
         path = initial_slashes + sep.join(comps)
         return path or dot
 
-else:
-    def normpath(path):
-        """Normalize path, eliminating double slashes, etc."""
-        path = os.fspath(path)
-        if isinstance(path, bytes):
-            return os.fsencode(_path_normpath(os.fsdecode(path))) or b"."
-        return _path_normpath(path) or "."
-
 
 def abspath(path):
     """Return an absolute path."""

Lib/test/test_ntpath.py

@@ -1129,6 +1129,10 @@ def test_fast_paths_in_use(self):
         # There are fast paths of these functions implemented in posixmodule.c.
         # Confirm that they are being used, and not the Python fallbacks in
         # genericpath.py.
+        self.assertTrue(os.path.splitroot is nt._path_splitroot_ex)
+        self.assertFalse(inspect.isfunction(os.path.splitroot))
+        self.assertTrue(os.path.normpath is nt._path_normpath)
+        self.assertFalse(inspect.isfunction(os.path.normpath))
         self.assertTrue(os.path.isdir is nt._path_isdir)
         self.assertFalse(inspect.isfunction(os.path.isdir))
         self.assertTrue(os.path.isfile is nt._path_isfile)

Lib/test/test_posixpath.py

@@ -1,11 +1,12 @@
+import inspect
 import os
 import posixpath
 import sys
 import unittest
 from posixpath import realpath, abspath, dirname, basename
 from test import test_genericpath
 from test.support import import_helper
-from test.support import os_helper
+from test.support import cpython_only, os_helper
 from test.support.os_helper import FakePath
 from unittest import mock
 
@@ -283,6 +284,16 @@ def fake_lstat(path):
     def test_isjunction(self):
         self.assertFalse(posixpath.isjunction(ABSTFN))
 
+    @unittest.skipIf(sys.platform == 'win32', "Fast paths are not for win32")
+    @cpython_only
+    def test_fast_paths_in_use(self):
+        # There are fast paths of these functions implemented in posixmodule.c.
+        # Confirm that they are being used, and not the Python fallbacks
+        self.assertTrue(os.path.splitroot is posix._path_splitroot_ex)
+        self.assertFalse(inspect.isfunction(os.path.splitroot))
+        self.assertTrue(os.path.normpath is posix._path_normpath)
+        self.assertFalse(inspect.isfunction(os.path.normpath))
+
     def test_expanduser(self):
         self.assertEqual(posixpath.expanduser("foo"), "foo")
         self.assertEqual(posixpath.expanduser(b"foo"), b"foo")

Misc/NEWS.d/next/Core and Builtins/2024-04-28-19-51-00.gh-issue-118263.Gaap3S.rst

@@ -0,0 +1 @@
+Speed up :func:`os.path.splitroot` & :func:`os.path.normpath` with a direct C call.