GH-118289: Fix handling of non-directories in os.path.realpath()

In strict mode, raise `NotADirectoryError` if a file path is given with a trailing slash, or subsequent dot segments. We use a `part_count` variable rather than `len(rest)` because the `rest` stack also contains markers for unresolved symlinks.
python · barneygale · Apr 25, 2024 · Apr 25, 2024 · Apr 25, 2024 · May 18, 2024
commit 59f7fde03aa8659a42b23fab72737d8e47c3cedc
diff --git a/Lib/posixpath.py b/Lib/posixpath.py
@@ -22,6 +22,7 @@
 altsep = None
 devnull = '/dev/null'
 
+import errno
 import os
 import sys
 import stat
@@ -421,6 +422,7 @@ def realpath(filename, *, strict=False):
     # symlink path can be retrieved by popping again. The [::-1] slice is a
     # very fast way of spelling list(reversed(...)).
     rest = filename.split(sep)[::-1]
+    part_count = len(rest)
 
     # The resolved path, which is absolute throughout this function.
     # Note: getcwd() returns a normalized and symlink-free path.
@@ -432,12 +434,13 @@ def realpath(filename, *, strict=False):
     # the same links.
     seen = {}
 
-    while rest:
+    while part_count:
         name = rest.pop()
         if name is None:
             # resolved symlink target
             seen[rest.pop()] = path
             continue
+        part_count -= 1
         if not name or name == curdir:
             # current dir
             continue
@@ -450,8 +453,12 @@ def realpath(filename, *, strict=False):
         else:
             newpath = path + sep + name
         try:
-            st = os.lstat(newpath)
-            if not stat.S_ISLNK(st.st_mode):
+            st_mode = os.lstat(newpath).st_mode
+            if stat.S_ISLNK(st_mode):
+                pass
+            elif part_count and not stat.S_ISDIR(st_mode):
+                raise NotADirectoryError(errno.ENOTDIR, "Not a directory", newpath)
+            else:
                 path = newpath
                 continue
         except OSError:
@@ -474,6 +481,7 @@ def realpath(filename, *, strict=False):
             continue
         seen[newpath] = None # not resolved symlink
         target = os.readlink(newpath)
+        target_parts = target.split(sep)
         if target.startswith(sep):
             # Symlink target is absolute; reset resolved path.
             path = sep
@@ -483,7 +491,8 @@ def realpath(filename, *, strict=False):
         rest.append(newpath)
         rest.append(None)
         # Push the unresolved symlink target parts onto the stack.
-        rest.extend(target.split(sep)[::-1])
+        rest.extend(reversed(target_parts))
+        part_count += len(target_parts)
 
     return path
 

diff --git a/Lib/test/test_posixpath.py b/Lib/test/test_posixpath.py
@@ -660,6 +660,20 @@ def test_realpath_resolve_first(self):
             safe_rmdir(ABSTFN + "/k")
             safe_rmdir(ABSTFN)
 
+    def test_realpath_strict_nondir(self):
+        try:
+            with open(ABSTFN, 'w') as f:
+                f.write('test_posixpath wuz ere')
+            self.assertEqual(realpath(ABSTFN, strict=True), ABSTFN)
+            with self.assertRaises(NotADirectoryError):
+                realpath(ABSTFN + "/", strict=True)
+            with self.assertRaises(NotADirectoryError):
+                realpath(ABSTFN + "/.", strict=True)
+            with self.assertRaises(NotADirectoryError):
+                realpath(ABSTFN + "/subdir", strict=True)
+        finally:
+            os_helper.unlink(ABSTFN)
+
     def test_relpath(self):
         (real_getcwd, os.getcwd) = (os.getcwd, lambda: r"/home/user/bar")
         try:

diff --git a/Misc/NEWS.d/next/Library/2024-04-25-18-15-53.gh-issue-118289.moL9_d.rst b/Misc/NEWS.d/next/Library/2024-04-25-18-15-53.gh-issue-118289.moL9_d.rst
@@ -0,0 +1,2 @@
+:func:`os.path.realpath` now raises :exc:`NotADirectoryError` when *strict*
+mode is enabled and a non-directory path with a trailing slash is supplied.
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1,2 @@
		:func:`os.path.realpath` now raises :exc:`NotADirectoryError` when strict
		mode is enabled and a non-directory path with a trailing slash is supplied.