changeset:   99420:9e800b2aeeac
parent:      99417:a89f92ac7d19
parent:      99419:a0e2376768dc
user:        Serhiy Storchaka <storchaka@gmail.com>
date:        Thu Dec 03 01:05:52 2015 +0200
files:       Misc/NEWS Objects/unicodeobject.c
description:
Issue #25709: Fixed problem with in-place string concatenation and utf-8 cache.


diff -r a89f92ac7d19 -r 9e800b2aeeac Lib/test/test_unicode.py
--- a/Lib/test/test_unicode.py	Thu Dec 03 00:58:13 2015 +0200
+++ b/Lib/test/test_unicode.py	Thu Dec 03 01:05:52 2015 +0200
@@ -2702,6 +2702,23 @@
         self.assertTrue(astral >= bmp2)
         self.assertFalse(astral >= astral2)
 
+    @support.cpython_only
+    def test_pep393_utf8_caching_bug(self):
+        # Issue #25709: Problem with string concatenation and utf-8 cache
+        from _testcapi import getargs_s_hash
+        for k in 0x24, 0xa4, 0x20ac, 0x1f40d:
+            s = ''
+            for i in range(5):
+                # Due to CPython specific optimization the 's' string can be
+                # resized in-place.
+                s += chr(k)
+                # Parsing with the "s#" format code calls indirectly
+                # PyUnicode_AsUTF8AndSize() which creates the UTF-8
+                # encoded string cached in the Unicode object.
+                self.assertEqual(getargs_s_hash(s), chr(k).encode() * (i + 1))
+                # Check that the second call returns the same result
+                self.assertEqual(getargs_s_hash(s), chr(k).encode() * (i + 1))
+
 
 class StringModuleTest(unittest.TestCase):
     def test_formatter_parser(self):
diff -r a89f92ac7d19 -r 9e800b2aeeac Misc/NEWS
--- a/Misc/NEWS	Thu Dec 03 00:58:13 2015 +0200
+++ b/Misc/NEWS	Thu Dec 03 01:05:52 2015 +0200
@@ -10,6 +10,8 @@
 Core and Builtins
 -----------------
 
+- Issue #25709: Fixed problem with in-place string concatenation and utf-8 cache.
+
 - Issue #5319: New Py_FinalizeEx() API allowing Python to set an exit status
   of 120 on failure to flush buffered streams.
 
diff -r a89f92ac7d19 -r 9e800b2aeeac Objects/unicodeobject.c
--- a/Objects/unicodeobject.c	Thu Dec 03 00:58:13 2015 +0200
+++ b/Objects/unicodeobject.c	Thu Dec 03 01:05:52 2015 +0200
@@ -885,6 +885,11 @@
     }
     new_size = (struct_size + (length + 1) * char_size);
 
+    if (_PyUnicode_HAS_UTF8_MEMORY(unicode)) {
+        PyObject_DEL(_PyUnicode_UTF8(unicode));
+        _PyUnicode_UTF8(unicode) = NULL;
+        _PyUnicode_UTF8_LENGTH(unicode) = 0;
+    }
     _Py_DEC_REFTOTAL;
     _Py_ForgetReference(unicode);