gh-98393: os module reject bytes-like, only accept bytes (#98394)

The os module and the PyUnicode_FSDecoder() function no longer accept
bytes-like paths, like bytearray and memoryview types: only the exact
bytes type is accepted for bytes strings.

diff --git a/Doc/whatsnew/3.12.rst b/Doc/whatsnew/3.12.rst
index ad0270ae19b825393d68846b33867221b06d5da5..525efc405c852042b77fd34ef2c5d1e3a5629f11 100644 (file)
--- a/Doc/whatsnew/3.12.rst
+++ b/Doc/whatsnew/3.12.rst
@@ -515,6 +515,11 @@ Changes in the Python API
   in Python 3.9.
   (Contributed by Victor Stinner in :gh:`94352`.)
 
+* The :mod:`os` module no longer accepts bytes-like paths, like
+  :class:`bytearray` and :class:`memoryview` types: only the exact
+  :class:`bytes` type is accepted for bytes strings.
+  (Contributed by Victor Stinner in :gh:`98393`.)
+
 
 Build Changes
 =============
@@ -631,6 +636,11 @@ Porting to Python 3.12
   to traverse and clear their instance's dictionaries.
   To clear weakrefs, call :c:func:`PyObject_ClearWeakRefs`, as before.
 
+* The :c:func:`PyUnicode_FSDecoder` function no longer accepts bytes-like
+  paths, like :class:`bytearray` and :class:`memoryview` types: only the exact
+  :class:`bytes` type is accepted for bytes strings.
+  (Contributed by Victor Stinner in :gh:`98393`.)
+
 Deprecated
 ----------
 

diff --git a/Lib/test/test_compile.py b/Lib/test/test_compile.py
index 8bf8470ff16feb9829da31f266478bcedca3db30..f5c0c76478323a1edc873d0ef0fd1d651515e1d4 100644 (file)
--- a/Lib/test/test_compile.py
+++ b/Lib/test/test_compile.py
@@ -494,9 +494,8 @@ if 1:
             code = compile('pass', filename, 'exec')
             self.assertEqual(code.co_filename, 'file.py')
         for filename in bytearray(b'file.py'), memoryview(b'file.py'):
-            with self.assertWarns(DeprecationWarning):
-                code = compile('pass', filename, 'exec')
-            self.assertEqual(code.co_filename, 'file.py')
+            with self.assertRaises(TypeError):
+                compile('pass', filename, 'exec')
         self.assertRaises(TypeError, compile, 'pass', list(b'file.py'), 'exec')
 
     @support.cpython_only

diff --git a/Lib/test/test_os.py b/Lib/test/test_os.py
index 2dafb78ebe0ec2fe4ad462e64f5f7d176ebfe5a6..10dbb2bd15dfab0e777ca9dbc48ded1a62533251 100644 (file)
--- a/Lib/test/test_os.py
+++ b/Lib/test/test_os.py
@@ -3860,18 +3860,18 @@ class OSErrorTests(unittest.TestCase):
 
         for filenames, func, *func_args in funcs:
             for name in filenames:
-                try:
-                    if isinstance(name, (str, bytes)):
+                if not isinstance(name, (str, bytes)):
+                    with self.assertRaises(TypeError):
                         func(name, *func_args)
-                    else:
-                        with self.assertWarnsRegex(DeprecationWarning, 'should be'):
-                            func(name, *func_args)
-                except OSError as err:
-                    self.assertIs(err.filename, name, str(func))
-                except UnicodeDecodeError:
-                    pass
                 else:
-                    self.fail("No exception thrown by {}".format(func))
+                    try:
+                        func(name, *func_args)
+                    except OSError as err:
+                        self.assertIs(err.filename, name, str(func))
+                    except UnicodeDecodeError:
+                        pass
+                    else:
+                        self.fail("No exception thrown by {}".format(func))
 
 class CPUCountTests(unittest.TestCase):
     def test_cpu_count(self):
@@ -4350,16 +4350,8 @@ class TestScandir(unittest.TestCase):
 
         for cls in bytearray, memoryview:
             path_bytes = cls(os.fsencode(self.path))
-            with self.assertWarns(DeprecationWarning):
-                entries = list(os.scandir(path_bytes))
-            self.assertEqual(len(entries), 1, entries)
-            entry = entries[0]
-
-            self.assertEqual(entry.name, b'file.txt')
-            self.assertEqual(entry.path,
-                             os.fsencode(os.path.join(self.path, 'file.txt')))
-            self.assertIs(type(entry.name), bytes)
-            self.assertIs(type(entry.path), bytes)
+            with self.assertRaises(TypeError):
+                list(os.scandir(path_bytes))
 
     @unittest.skipUnless(os.listdir in os.supports_fd,
                          'fd support for listdir required for this test.')

diff --git a/Lib/test/test_posix.py b/Lib/test/test_posix.py
index e643d8e5a4ce63b97b14b29f4e351e06b6ea23cc..be561afc4cffa788686f727ddbd9facf1e7419f3 100644 (file)
--- a/Lib/test/test_posix.py
+++ b/Lib/test/test_posix.py
@@ -634,7 +634,7 @@ class PosixTester(unittest.TestCase):
         self.assertTrue(posix.stat(os_helper.TESTFN))
         self.assertTrue(posix.stat(os.fsencode(os_helper.TESTFN)))
 
-        self.assertWarnsRegex(DeprecationWarning,
+        self.assertRaisesRegex(TypeError,
                 'should be string, bytes, os.PathLike or integer, not',
                 posix.stat, bytearray(os.fsencode(os_helper.TESTFN)))
         self.assertRaisesRegex(TypeError,
@@ -841,11 +841,8 @@ class PosixTester(unittest.TestCase):
 
     def test_listdir_bytes_like(self):
         for cls in bytearray, memoryview:
-            with self.assertWarns(DeprecationWarning):
-                names = posix.listdir(cls(b'.'))
-            self.assertIn(os.fsencode(os_helper.TESTFN), names)
-            for name in names:
-                self.assertIs(type(name), bytes)
+            with self.assertRaises(TypeError):
+                posix.listdir(cls(b'.'))
 
     @unittest.skipUnless(posix.listdir in os.supports_fd,
                          "test needs fd support for posix.listdir()")

diff --git a/Lib/test/test_symtable.py b/Lib/test/test_symtable.py
index 819354e4eee9b5e03852d34daf1b17691f9b6f1b..25714aecda3a1525571b7e4b68ab523154c24d7d 100644 (file)
--- a/Lib/test/test_symtable.py
+++ b/Lib/test/test_symtable.py
@@ -222,10 +222,9 @@ class SymtableTest(unittest.TestCase):
         checkfilename("def f(x): foo)(", 14)  # parse-time
         checkfilename("def f(x): global x", 11)  # symtable-build-time
         symtable.symtable("pass", b"spam", "exec")
-        with self.assertWarns(DeprecationWarning), \
-             self.assertRaises(TypeError):
+        with self.assertRaises(TypeError):
             symtable.symtable("pass", bytearray(b"spam"), "exec")
-        with self.assertWarns(DeprecationWarning):
+        with self.assertRaises(TypeError):
             symtable.symtable("pass", memoryview(b"spam"), "exec")
         with self.assertRaises(TypeError):
             symtable.symtable("pass", list(b"spam"), "exec")

diff --git a/Misc/NEWS.d/next/C API/2022-10-18-16-16-27.gh-issue-98393.55u4BF.rst b/Misc/NEWS.d/next/C API/2022-10-18-16-16-27.gh-issue-98393.55u4BF.rst
new file mode 100644 (file)
index 0000000..658f587
--- /dev/null
+++ b/Misc/NEWS.d/next/C API/2022-10-18-16-16-27.gh-issue-98393.55u4BF.rst
@@ -0,0 +1,3 @@
+The :c:func:`PyUnicode_FSDecoder` function no longer accepts bytes-like
+paths, like :class:`bytearray` and :class:`memoryview` types: only the exact
+:class:`bytes` type is accepted for bytes strings. Patch by Victor Stinner.

diff --git a/Misc/NEWS.d/next/Library/2022-10-18-15-41-37.gh-issue-98393.vhPu4L.rst b/Misc/NEWS.d/next/Library/2022-10-18-15-41-37.gh-issue-98393.vhPu4L.rst
new file mode 100644 (file)
index 0000000..54e45bb
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2022-10-18-15-41-37.gh-issue-98393.vhPu4L.rst
@@ -0,0 +1,3 @@
+The :mod:`os` module no longer accepts bytes-like paths, like
+:class:`bytearray` and :class:`memoryview` types: only the exact
+:class:`bytes` type is accepted for bytes strings. Patch by Victor Stinner.

diff --git a/Modules/posixmodule.c b/Modules/posixmodule.c
index 1e556fc904b77bbf9dc721fb4e012346bde6b846..39198cb7ddcdac81e2c3271b335a4dab297d0621 100644 (file)
--- a/Modules/posixmodule.c
+++ b/Modules/posixmodule.c
@@ -1120,7 +1120,7 @@ path_converter(PyObject *o, void *p)
     path_t *path = (path_t *)p;
     PyObject *bytes = NULL;
     Py_ssize_t length = 0;
-    int is_index, is_buffer, is_bytes, is_unicode;
+    int is_index, is_bytes, is_unicode;
     const char *narrow;
 #ifdef MS_WINDOWS
     PyObject *wo = NULL;
@@ -1158,11 +1158,10 @@ path_converter(PyObject *o, void *p)
     /* Only call this here so that we don't treat the return value of
        os.fspath() as an fd or buffer. */
     is_index = path->allow_fd && PyIndex_Check(o);
-    is_buffer = PyObject_CheckBuffer(o);
     is_bytes = PyBytes_Check(o);
     is_unicode = PyUnicode_Check(o);
 
-    if (!is_index && !is_buffer && !is_unicode && !is_bytes) {
+    if (!is_index && !is_unicode && !is_bytes) {
         /* Inline PyOS_FSPath() for better error messages. */
         PyObject *func, *res;
 
@@ -1225,27 +1224,6 @@ path_converter(PyObject *o, void *p)
         bytes = o;
         Py_INCREF(bytes);
     }
-    else if (is_buffer) {
-        /* XXX Replace PyObject_CheckBuffer with PyBytes_Check in other code
-           after removing support of non-bytes buffer objects. */
-        if (PyErr_WarnFormat(PyExc_DeprecationWarning, 1,
-            "%s%s%s should be %s, not %.200s",
-            path->function_name ? path->function_name : "",
-            path->function_name ? ": "                : "",
-            path->argument_name ? path->argument_name : "path",
-            path->allow_fd && path->nullable ? "string, bytes, os.PathLike, "
-                                               "integer or None" :
-            path->allow_fd ? "string, bytes, os.PathLike or integer" :
-            path->nullable ? "string, bytes, os.PathLike or None" :
-                             "string, bytes or os.PathLike",
-            _PyType_Name(Py_TYPE(o)))) {
-            goto error_exit;
-        }
-        bytes = PyBytes_FromObject(o);
-        if (!bytes) {
-            goto error_exit;
-        }
-    }
     else if (is_index) {
         if (!_fd_converter(o, &path->fd)) {
             goto error_exit;
@@ -4126,8 +4104,8 @@ _posix_listdir(path_t *path, PyObject *list)
         const char *name;
         if (path->narrow) {
             name = path->narrow;
-            /* only return bytes if they specified a bytes-like object */
-            return_str = !PyObject_CheckBuffer(path->object);
+            /* only return bytes if they specified a bytes object */
+            return_str = !PyBytes_Check(path->object);
         }
         else {
             name = ".";
@@ -14049,7 +14027,7 @@ DirEntry_from_posix_info(PyObject *module, path_t *path, const char *name,
             goto error;
     }
 
-    if (!path->narrow || !PyObject_CheckBuffer(path->object)) {
+    if (!path->narrow || !PyBytes_Check(path->object)) {
         entry->name = PyUnicode_DecodeFSDefaultAndSize(name, name_len);
         if (joined_path)
             entry->path = PyUnicode_DecodeFSDefault(joined_path);

diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 5b737f1b596e3e1c44898289c46b22f6e9ada692..d090915146f804d06f11ccfa2cbdf19ad50cc83a 100644 (file)
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -3619,48 +3619,25 @@ PyUnicode_FSConverter(PyObject* arg, void* addr)
 int
 PyUnicode_FSDecoder(PyObject* arg, void* addr)
 {
-    int is_buffer = 0;
-    PyObject *path = NULL;
-    PyObject *output = NULL;
     if (arg == NULL) {
         Py_DECREF(*(PyObject**)addr);
         *(PyObject**)addr = NULL;
         return 1;
     }
 
-    is_buffer = PyObject_CheckBuffer(arg);
-    if (!is_buffer) {
-        path = PyOS_FSPath(arg);
-        if (path == NULL) {
-            return 0;
-        }
-    }
-    else {
-        path = arg;
-        Py_INCREF(arg);
+    PyObject *path = PyOS_FSPath(arg);
+    if (path == NULL) {
+        return 0;
     }
 
+    PyObject *output = NULL;
     if (PyUnicode_Check(path)) {
         output = path;
     }
-    else if (PyBytes_Check(path) || is_buffer) {
-        PyObject *path_bytes = NULL;
-
-        if (!PyBytes_Check(path) &&
-            PyErr_WarnFormat(PyExc_DeprecationWarning, 1,
-            "path should be string, bytes, or os.PathLike, not %.200s",
-            Py_TYPE(arg)->tp_name)) {
-                Py_DECREF(path);
-            return 0;
-        }
-        path_bytes = PyBytes_FromObject(path);
+    else if (PyBytes_Check(path)) {
+        output = PyUnicode_DecodeFSDefaultAndSize(PyBytes_AS_STRING(path),
+                                                  PyBytes_GET_SIZE(path));
         Py_DECREF(path);
-        if (!path_bytes) {
-            return 0;
-        }
-        output = PyUnicode_DecodeFSDefaultAndSize(PyBytes_AS_STRING(path_bytes),
-                                                  PyBytes_GET_SIZE(path_bytes));
-        Py_DECREF(path_bytes);
         if (!output) {
             return 0;
         }
@@ -3672,6 +3649,7 @@ PyUnicode_FSDecoder(PyObject* arg, void* addr)
         Py_DECREF(path);
         return 0;
     }
+
     if (findchar(PyUnicode_DATA(output), PyUnicode_KIND(output),
                  PyUnicode_GET_LENGTH(output), 0, 1) >= 0) {
         PyErr_SetString(PyExc_ValueError, "embedded null character");