changeset:   93068:449b1f427cc7
branch:      3.3
parent:      93032:433d10b195f2
user:        Benjamin Peterson <benjamin@python.org>
date:        Wed Oct 15 11:47:36 2014 -0400
files:       Lib/test/test_unicode.py Misc/NEWS Objects/unicodeobject.c
description:
fix integer overflow in unicode case operations (closes #22643)


diff -r 433d10b195f2 -r 449b1f427cc7 Lib/test/test_unicode.py
--- a/Lib/test/test_unicode.py	Mon Oct 13 11:54:50 2014 -0400
+++ b/Lib/test/test_unicode.py	Wed Oct 15 11:47:36 2014 -0400
@@ -661,6 +661,11 @@
         self.assertEqual('x'.center(4, '\U0010FFFF'),
                          '\U0010FFFFx\U0010FFFF\U0010FFFF')
 
+    @unittest.skipUnless(sys.maxsize == 2**31 - 1, "requires 32-bit system")
+    def test_case_operation_overflow(self):
+        # Issue #22643
+        self.assertRaises(OverflowError, ("ü"*(2**32//12 + 1)).upper)
+
     def test_contains(self):
         # Testing Unicode contains method
         self.assertIn('a', 'abdb')
diff -r 433d10b195f2 -r 449b1f427cc7 Misc/NEWS
--- a/Misc/NEWS	Mon Oct 13 11:54:50 2014 -0400
+++ b/Misc/NEWS	Wed Oct 15 11:47:36 2014 -0400
@@ -10,6 +10,9 @@
 Core and Builtins
 -----------------
 
+- Issue #22643: Fix integer overflow in Unicode case operations (upper, lower,
+  title, swapcase, casefold).
+
 - Issue #22518: Fixed integer overflow issues in "backslashreplace",
   "xmlcharrefreplace", and "surrogatepass" error handlers.
 
diff -r 433d10b195f2 -r 449b1f427cc7 Objects/unicodeobject.c
--- a/Objects/unicodeobject.c	Mon Oct 13 11:54:50 2014 -0400
+++ b/Objects/unicodeobject.c	Wed Oct 15 11:47:36 2014 -0400
@@ -9484,6 +9484,11 @@
     kind = PyUnicode_KIND(self);
     data = PyUnicode_DATA(self);
     length = PyUnicode_GET_LENGTH(self);
+    if (length > PY_SSIZE_T_MAX / 3 ||
+        length > PY_SIZE_MAX / (3 * sizeof(Py_UCS4))) {
+        PyErr_SetString(PyExc_OverflowError, "string is too long");
+        return NULL;
+    }
     tmp = PyMem_MALLOC(sizeof(Py_UCS4) * 3 * length);
     if (tmp == NULL)
         return PyErr_NoMemory();