Index: PC/pyconfig.h =================================================================== --- PC/pyconfig.h (revision 75508) +++ PC/pyconfig.h (working copy) @@ -641,6 +641,11 @@ #define HAVE_WCSCOLL 1 #endif +/* Define to 1 if you have the `wcsxfrm' function. */ +#ifndef MS_WINCE +#define HAVE_WCSXFRM 1 +#endif + /* Define if you have the header file. */ /* #undef HAVE_DLFCN_H */ Index: Lib/test/test_locale.py =================================================================== --- Lib/test/test_locale.py (revision 75508) +++ Lib/test/test_locale.py (working copy) @@ -8,12 +8,10 @@ def get_enUS_locale(): global enUS_locale - if sys.platform == 'darwin': - raise unittest.SkipTest("Locale support on MacOSX is minimal") if sys.platform.startswith("win"): tlocs = ("En", "English") else: - tlocs = ("en_US.UTF-8", "en_US.US-ASCII", "en_US") + tlocs = ("en_US.UTF-8", "en_US.ISO8859-1", "en_US.US-ASCII", "en_US") oldlocale = locale.setlocale(locale.LC_NUMERIC) for tloc in tlocs: try: @@ -309,6 +307,39 @@ grouping=True, international=True) +class TestCollation(unittest.TestCase): + # Test string collation functions + + def test_strcoll(self): + self.assertLess(locale.strcoll('a', 'b'), 0) + self.assertEqual(locale.strcoll('a', 'a'), 0) + self.assertGreater(locale.strcoll('b', 'a'), 0) + + def test_strxfrm(self): + self.assertLess(locale.strxfrm('a'), locale.strxfrm('b')) + + +class TestEnUSCollation(BaseLocalizedTest, TestCollation): + # Test string collation functions with a real English locale + + locale_type = locale.LC_ALL + + def setUp(self): + BaseLocalizedTest.setUp(self) + enc = codecs.lookup(locale.getpreferredencoding(False) or 'ascii').name + if enc not in ('utf-8', 'iso8859-1', 'cp1252'): + raise unittest.SkipTest('encoding not suitable') + if enc != 'iso8859-1' and (sys.platform == 'darwin' or + sys.platform.startswith('freebsd')): + raise unittest.SkipTest('wcscoll/wcsxfrm have known bugs') + + def test_strcoll_with_diacritic(self): + self.assertLess(locale.strcoll('à', 'b'), 0) + + def test_strxfrm_with_diacritic(self): + self.assertLess(locale.strxfrm('à'), locale.strxfrm('b')) + + class TestMiscellaneous(unittest.TestCase): def test_getpreferredencoding(self): # Invoke getpreferredencoding to make sure it does not cause exceptions. @@ -317,11 +348,10 @@ # If encoding non-empty, make sure it is valid codecs.lookup(enc) - if hasattr(locale, "strcoll"): - def test_strcoll_3303(self): - # test crasher from bug #3303 - self.assertRaises(TypeError, locale.strcoll, "a", None) - self.assertRaises(TypeError, locale.strcoll, b"a", None) + def test_strcoll_3303(self): + # test crasher from bug #3303 + self.assertRaises(TypeError, locale.strcoll, "a", None) + self.assertRaises(TypeError, locale.strcoll, b"a", None) def test_main(): @@ -331,6 +361,7 @@ TestEnUSNumberFormatting, TestCNumberFormatting, TestFrFRNumberFormatting, + TestCollation ] # SkipTest can't be raised inside unittests, handle it manually instead try: @@ -339,7 +370,7 @@ if verbose: print("Some tests will be disabled: %s" % e) else: - tests += [TestNumberFormatting] + tests += [TestNumberFormatting, TestEnUSCollation] run_unittest(*tests) if __name__ == '__main__': Index: Lib/locale.py =================================================================== --- Lib/locale.py (revision 75508) +++ Lib/locale.py (working copy) @@ -578,14 +578,13 @@ # returning nothing will crash the # interpreter. result = 'UTF-8' - setlocale(LC_CTYPE, oldloc) - return result else: result = nl_langinfo(CODESET) if not result and sys.platform == 'darwin': # See above for explanation result = 'UTF-8' + return result ### Database Index: Modules/_localemodule.c =================================================================== --- Modules/_localemodule.c (revision 75508) +++ Modules/_localemodule.c (working copy) @@ -9,6 +9,7 @@ ******************************************************************/ +#define PY_SSIZE_T_CLEAN #include "Python.h" #include @@ -315,7 +316,7 @@ result = PyUnicode_FromWideChar(buf, n2); exit: if (buf) PyMem_Free(buf); -#ifdef HAVE_USABLE_WCHAR_T +#ifndef HAVE_USABLE_WCHAR_T PyMem_Free(s); #endif return result;