Issue #23908: os functions now reject paths with embedded null character

on Windows instead of silently truncate them.

Removed no longer used _PyUnicode_HasNULChars().

Include/unicodeobject.h

@@ -2060,12 +2060,6 @@ PyAPI_FUNC(int) PyUnicode_Contains(
     PyObject *element           /* Element string */
     );
 
-/* Checks whether the string contains any NUL characters. */
-
-#ifndef Py_LIMITED_API
-PyAPI_FUNC(int) _PyUnicode_HasNULChars(PyObject *);
-#endif
-
 /* Checks whether argument is a valid identifier. */
 
 PyAPI_FUNC(int) PyUnicode_IsIdentifier(PyObject *s);

Lib/test/test_posix.py

@@ -1168,6 +1168,42 @@ class PosixTester(unittest.TestCase):
             else:
                 self.fail("No valid path_error2() test for os." + name)
 
+    def test_path_with_null_character(self):
+        fn = support.TESTFN
+        fn_with_NUL = fn + '\0'
+        self.addCleanup(support.unlink, fn)
+        support.unlink(fn)
+        fd = None
+        try:
+            with self.assertRaises(ValueError):
+                fd = os.open(fn_with_NUL, os.O_WRONLY | os.O_CREAT) # raises
+        finally:
+            if fd is not None:
+                os.close(fd)
+        self.assertFalse(os.path.exists(fn))
+        self.assertRaises(ValueError, os.mkdir, fn_with_NUL)
+        self.assertFalse(os.path.exists(fn))
+        open(fn, 'wb').close()
+        self.assertRaises(ValueError, os.stat, fn_with_NUL)
+
+    def test_path_with_null_byte(self):
+        fn = os.fsencode(support.TESTFN)
+        fn_with_NUL = fn + b'\0'
+        self.addCleanup(support.unlink, fn)
+        support.unlink(fn)
+        fd = None
+        try:
+            with self.assertRaises(ValueError):
+                fd = os.open(fn_with_NUL, os.O_WRONLY | os.O_CREAT) # raises
+        finally:
+            if fd is not None:
+                os.close(fd)
+        self.assertFalse(os.path.exists(fn))
+        self.assertRaises(ValueError, os.mkdir, fn_with_NUL)
+        self.assertFalse(os.path.exists(fn))
+        open(fn, 'wb').close()
+        self.assertRaises(ValueError, os.stat, fn_with_NUL)
+
 class PosixGroupsTester(unittest.TestCase):
 
     def setUp(self):

Misc/NEWS

@@ -37,6 +37,9 @@ Core and Builtins
 Library
 -------
 
+- Issue #23908: os functions now reject paths with embedded null character
+  on Windows instead of silently truncate them.
+
 - Issue #23728: binascii.crc_hqx() could return an integer outside of the range
   0-0xffff for empty data.
 

Modules/_io/fileio.c

@@ -281,15 +281,14 @@ _io_FileIO___init___impl(fileio *self, PyObject *nameobj, const char *mode,
 
 #ifdef MS_WINDOWS
     if (PyUnicode_Check(nameobj)) {
-        int rv = _PyUnicode_HasNULChars(nameobj);
+        Py_ssize_t length;
-        if (rv) {
+        widename = PyUnicode_AsUnicodeAndSize(nameobj, &length);
-            if (rv != -1)
-                PyErr_SetString(PyExc_ValueError, "embedded null character");
-            return -1;
-        }
-        widename = PyUnicode_AsUnicode(nameobj);
         if (widename == NULL)
             return -1;
+        if (wcslen(widename) != length) {
+            PyErr_SetString(PyExc_ValueError, "embedded null character");
+            return -1;
+        }
     } else
 #endif
     if (fd < 0)

Modules/posixmodule.c

@@ -866,6 +866,11 @@ path_converter(PyObject *o, void *p) {
             Py_DECREF(unicode);
             return 0;
         }
+        if (wcslen(wide) != length) {
+            FORMAT_EXCEPTION(PyExc_ValueError, "embedded null character");
+            Py_DECREF(unicode);
+            return 0;
+        }
 
         path->wide = wide;
         path->narrow = NULL;

Objects/unicodeobject.c

@@ -3606,21 +3606,6 @@ PyUnicode_DecodeFSDefaultAndSize(const char *s, Py_ssize_t size)
 }
 
 
-int
-_PyUnicode_HasNULChars(PyObject* str)
-{
-    Py_ssize_t pos;
-
-    if (PyUnicode_READY(str) == -1)
-        return -1;
-    pos = findchar(PyUnicode_DATA(str), PyUnicode_KIND(str),
-                   PyUnicode_GET_LENGTH(str), '\0', 1);
-    if (pos == -1)
-        return 0;
-    else
-        return 1;
-}
-
 int
 PyUnicode_FSConverter(PyObject* arg, void* addr)
 {