[3.6] bpo-30708: Check for null characters in PyUnicode_AsWideCharString(). (GH-2285) (#2443)

Raise a ValueError if the second argument is NULL and the wchar_t\*
string contains null characters..
(cherry picked from commit e613e6add5f07ff6aad5802924596b631b707d2a)

12 files changed, 50 insertions, 20 deletions

diff --git a/Include/unicodeobject.h b/Include/unicodeobject.h
index 5b87718..cec2b7f 100644
--- a/Include/unicodeobject.h
+++ b/Include/unicodeobject.h
@@ -1063,6 +1063,12 @@ PyAPI_FUNC(wchar_t*) PyUnicode_AsWideCharString(
     );
 
 #ifndef Py_LIMITED_API
+/* Similar to PyUnicode_AsWideCharString(unicode, NULL), but check if
+   the string contains null characters. */
+PyAPI_FUNC(wchar_t*) _PyUnicode_AsWideCharString(
+    PyObject *unicode           /* Unicode object */
+    );
+
 PyAPI_FUNC(void*) _PyUnicode_AsKind(PyObject *s, unsigned int kind);
 #endif
 
diff --git a/Lib/ctypes/test/test_slicing.py b/Lib/ctypes/test/test_slicing.py
index 240dc0c..a3932f1 100644
--- a/Lib/ctypes/test/test_slicing.py
+++ b/Lib/ctypes/test/test_slicing.py
@@ -134,7 +134,7 @@ class SlicesTestCase(unittest.TestCase):
         dll.my_wcsdup.restype = POINTER(c_wchar)
         dll.my_wcsdup.argtypes = POINTER(c_wchar),
         dll.my_free.restype = None
-        res = dll.my_wcsdup(s)
+        res = dll.my_wcsdup(s[:-1])
         self.assertEqual(res[:len(s)], s)
         self.assertEqual(res[:len(s):], s)
         self.assertEqual(res[len(s)-1:-1:-1], s[::-1])
@@ -153,7 +153,7 @@ class SlicesTestCase(unittest.TestCase):
             dll.my_wcsdup.restype = POINTER(c_long)
         else:
             self.skipTest('Pointers to c_wchar are not supported')
-        res = dll.my_wcsdup(s)
+        res = dll.my_wcsdup(s[:-1])
         tmpl = list(range(ord("a"), ord("z")+1))
         self.assertEqual(res[:len(s)-1], tmpl)
         self.assertEqual(res[:len(s)-1:], tmpl)
diff --git a/Modules/_ctypes/callproc.c b/Modules/_ctypes/callproc.c
index 7d542fb..358f287 100644
--- a/Modules/_ctypes/callproc.c
+++ b/Modules/_ctypes/callproc.c
@@ -668,7 +668,7 @@ static int ConvParam(PyObject *obj, Py_ssize_t index, struct argument *pa)
 #ifdef CTYPES_UNICODE
     if (PyUnicode_Check(obj)) {
         pa->ffi_type = &ffi_type_pointer;
-        pa->value.p = PyUnicode_AsWideCharString(obj, NULL);
+        pa->value.p = _PyUnicode_AsWideCharString(obj);
         if (pa->value.p == NULL)
             return -1;
         pa->keep = PyCapsule_New(pa->value.p, CTYPES_CAPSULE_NAME_PYMEM, pymem_destructor);
diff --git a/Modules/_ctypes/cfield.c b/Modules/_ctypes/cfield.c
index a43585f..113a815 100644
--- a/Modules/_ctypes/cfield.c
+++ b/Modules/_ctypes/cfield.c
@@ -1372,7 +1372,7 @@ Z_set(void *ptr, PyObject *value, Py_ssize_t size)
 
     /* We must create a wchar_t* buffer from the unicode object,
        and keep it alive */
-    buffer = PyUnicode_AsWideCharString(value, NULL);
+    buffer = _PyUnicode_AsWideCharString(value);
     if (!buffer)
         return NULL;
     keep = PyCapsule_New(buffer, CTYPES_CFIELD_CAPSULE_NAME_PYMEM, pymem_destructor);
diff --git a/Modules/_cursesmodule.c b/Modules/_cursesmodule.c
index d88d06e..41b831e 100644
--- a/Modules/_cursesmodule.c
+++ b/Modules/_cursesmodule.c
@@ -345,7 +345,7 @@ PyCurses_ConvertToString(PyCursesWindowObject *win, PyObject *obj,
     if (PyUnicode_Check(obj)) {
 #ifdef HAVE_NCURSESW
         assert (wstr != NULL);
-        *wstr = PyUnicode_AsWideCharString(obj, NULL);
+        *wstr = _PyUnicode_AsWideCharString(obj);
         if (*wstr == NULL)
             return 0;
         return 2;
diff --git a/Modules/_io/winconsoleio.c b/Modules/_io/winconsoleio.c
index 94dfe3e..346c386 100644
--- a/Modules/_io/winconsoleio.c
+++ b/Modules/_io/winconsoleio.c
@@ -79,7 +79,7 @@ char _PyIO_get_console_type(PyObject *path_or_fd) {
         PyErr_Clear();
         return '\0';
     }
-    decoded_wstr = PyUnicode_AsWideCharString(decoded, NULL);
+    decoded_wstr = _PyUnicode_AsWideCharString(decoded);
     Py_CLEAR(decoded);
     if (!decoded_wstr) {
         PyErr_Clear();
@@ -311,8 +311,7 @@ _io__WindowsConsoleIO___init___impl(winconsoleio *self, PyObject *nameobj,
         if (!d)
             return -1;
 
-        Py_ssize_t length;
-        name = PyUnicode_AsWideCharString(decodedname, &length);
+        name = _PyUnicode_AsWideCharString(decodedname);
         console_type = _PyIO_get_console_type(decodedname);
         Py_CLEAR(decodedname);
         if (name == NULL)
@@ -322,12 +321,6 @@ _io__WindowsConsoleIO___init___impl(winconsoleio *self, PyObject *nameobj,
                 "Cannot open non-console file");
             return -1;
         }
-
-        if (wcslen(name) != length) {
-            PyMem_Free(name);
-            PyErr_SetString(PyExc_ValueError, "embedded null character");
-            return -1;
-        }
     }
 
     s = mode;
diff --git a/Modules/_localemodule.c b/Modules/_localemodule.c
index feb3802..0c7c3cd 100644
--- a/Modules/_localemodule.c
+++ b/Modules/_localemodule.c
@@ -215,10 +215,10 @@ PyLocale_strcoll(PyObject* self, PyObject* args)
     if (!PyArg_ParseTuple(args, "UU:strcoll", &os1, &os2))
         return NULL;
     /* Convert the unicode strings to wchar[]. */
-    ws1 = PyUnicode_AsWideCharString(os1, NULL);
+    ws1 = _PyUnicode_AsWideCharString(os1);
     if (ws1 == NULL)
         goto done;
-    ws2 = PyUnicode_AsWideCharString(os2, NULL);
+    ws2 = _PyUnicode_AsWideCharString(os2);
     if (ws2 == NULL)
         goto done;
     /* Collate the strings. */
diff --git a/Modules/_tkinter.c b/Modules/_tkinter.c
index 3e96aa0..4a1324c 100644
--- a/Modules/_tkinter.c
+++ b/Modules/_tkinter.c
@@ -3615,7 +3615,7 @@ PyInit__tkinter(void)
                     return NULL;
                 }
                 if (str_path != NULL) {
-                    wcs_path = PyUnicode_AsWideCharString(str_path, NULL);
+                    wcs_path = _PyUnicode_AsWideCharString(str_path);
                     if (wcs_path == NULL) {
                         return NULL;
                     }
diff --git a/Modules/overlapped.c b/Modules/overlapped.c
index 0aa8657..17581e0 100644
--- a/Modules/overlapped.c
+++ b/Modules/overlapped.c
@@ -1151,7 +1151,7 @@ ConnectPipe(OverlappedObject *self, PyObject *args)
     if (!PyArg_ParseTuple(args, "U",  &AddressObj))
         return NULL;
 
-    Address = PyUnicode_AsWideCharString(AddressObj, NULL);
+    Address = _PyUnicode_AsWideCharString(AddressObj);
     if (Address == NULL)
         return NULL;
 
diff --git a/Modules/timemodule.c b/Modules/timemodule.c
index 328b84f..d6c34b9 100644
--- a/Modules/timemodule.c
+++ b/Modules/timemodule.c
@@ -607,7 +607,7 @@ time_strftime(PyObject *self, PyObject *args)
         buf.tm_isdst = 1;
 
 #ifdef HAVE_WCSFTIME
-    format = PyUnicode_AsWideCharString(format_arg, NULL);
+    format = _PyUnicode_AsWideCharString(format_arg);
     if (format == NULL)
         return NULL;
     fmt = format;
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 4e0c663..3767064 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -3017,6 +3017,37 @@ PyUnicode_AsWideCharString(PyObject *unicode,
     return buffer;
 }
 
+wchar_t*
+_PyUnicode_AsWideCharString(PyObject *unicode)
+{
+    const wchar_t *wstr;
+    wchar_t *buffer;
+    Py_ssize_t buflen;
+
+    if (unicode == NULL) {
+        PyErr_BadInternalCall();
+        return NULL;
+    }
+
+    wstr = PyUnicode_AsUnicodeAndSize(unicode, &buflen);
+    if (wstr == NULL) {
+        return NULL;
+    }
+    if (wcslen(wstr) != (size_t)buflen) {
+        PyErr_SetString(PyExc_ValueError,
+                        "embedded null character");
+        return NULL;
+    }
+
+    buffer = PyMem_NEW(wchar_t, buflen + 1);
+    if (buffer == NULL) {
+        PyErr_NoMemory();
+        return NULL;
+    }
+    memcpy(buffer, wstr, (buflen + 1) * sizeof(wchar_t));
+    return buffer;
+}
+
 #endif /* HAVE_WCHAR_H */
 
 PyObject *
diff --git a/PC/winsound.c b/PC/winsound.c
index 7feebcb..0d3f045 100644
--- a/PC/winsound.c
+++ b/PC/winsound.c
@@ -101,7 +101,7 @@ winsound_PlaySound_impl(PyObject *module, PyObject *sound, int flags)
                          Py_TYPE(sound)->tp_name);
             return NULL;
         }
-        wsound = PyUnicode_AsWideCharString(sound, NULL);
+        wsound = _PyUnicode_AsWideCharString(sound);
         if (wsound == NULL) {
             return NULL;
         }