Issue 13496: Fix bisect.bisect overflow bug for large collections.

diff --git a/Lib/test/test_bisect.py b/Lib/test/test_bisect.py
index 93b86137ace2be715deb1e92faa54636968c200b..c24a1a2ad904777cc148120d0bd334157f386cfb 100644 (file)
--- a/Lib/test/test_bisect.py
+++ b/Lib/test/test_bisect.py
@@ -122,6 +122,13 @@ class TestBisect(unittest.TestCase):
         self.assertRaises(ValueError, mod.insort_left, [1, 2, 3], 5, -1, 3),
         self.assertRaises(ValueError, mod.insort_right, [1, 2, 3], 5, -1, 3),
 
+    def test_large_range(self):
+        # Issue 13496
+        mod = self.module
+        data = range(sys.maxsize-1)
+        self.assertEqual(mod.bisect_left(data, sys.maxsize-3), sys.maxsize-3)
+        self.assertEqual(mod.bisect_right(data, sys.maxsize-3), sys.maxsize-2)
+
     def test_random(self, n=25):
         from random import randrange
         for i in range(n):

diff --git a/Misc/NEWS b/Misc/NEWS
index b045742195ab21e58d83be2d353ac4c25720654e..91b0c65692e762ce131074a905d9ee1d0d578586 100644 (file)
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -43,6 +43,9 @@ Core and Builtins
 Library
 -------
 
+- Issue #13496: Fix potential overflow in bisect.bisect algorithm when applied
+  to a collection of size > sys.maxsize / 2.
+
 - Issue #14399: zipfile now recognizes that the archive has been modified even
   if only the comment is changed.  In addition, the TypeError that results from
   trying to set a non-binary value as a comment is now now raised at the time

diff --git a/Modules/_bisectmodule.c b/Modules/_bisectmodule.c
index 7fecfc6937b808476a8ec2b94003bf74fc351b03..93d0eed41e4a173fcdf367eff2a5f9d79934d261 100644 (file)
--- a/Modules/_bisectmodule.c
+++ b/Modules/_bisectmodule.c
@@ -21,7 +21,10 @@ internal_bisect_right(PyObject *list, PyObject *item, Py_ssize_t lo, Py_ssize_t
             return -1;
     }
     while (lo < hi) {
-        mid = (lo + hi) / 2;
+        /* The (size_t)cast ensures that the addition and subsequent division
+           are performed as unsigned operations, avoiding difficulties from
+           signed overflow.  (See issue 13496.) */
+        mid = ((size_t)lo + hi) / 2;
         litem = PySequence_GetItem(list, mid);
         if (litem == NULL)
             return -1;
@@ -121,7 +124,10 @@ internal_bisect_left(PyObject *list, PyObject *item, Py_ssize_t lo, Py_ssize_t h
             return -1;
     }
     while (lo < hi) {
-        mid = (lo + hi) / 2;
+        /* The (size_t)cast ensures that the addition and subsequent division
+           are performed as unsigned operations, avoiding difficulties from
+           signed overflow.  (See issue 13496.) */
+        mid = ((size_t)lo + hi) / 2;
         litem = PySequence_GetItem(list, mid);
         if (litem == NULL)
             return -1;