From: Serhiy Storchaka <storchaka@gmail.com>
Date: Fri, 24 Jul 2026 12:54:47 +0000 (+0300)
Subject: [3.13] gh-154582: Fix test_invalid_utf8_arg in non-UTF-8 multibyte locales (GH-154584... 
X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=3871923d65c30cb8cd8306820a6f235547c0bcb4;p=thirdparty%2FPython%2Fcpython.git

[3.13] gh-154582: Fix test_invalid_utf8_arg in non-UTF-8 multibyte locales (GH-154584) (GH-154605)

Arbitrary bytes round-trip through surrogateescape only in UTF-8 and
single-byte encodings, not in a stateful multibyte encoding such as EUC-JP.
(cherry picked from commit b1e530ebf3b68e884ab37882d690d95e5d82aef5)

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
---

diff --git a/Lib/test/test_cmd_line.py b/Lib/test/test_cmd_line.py
index 797615c915c3..7c2e1b98c28d 100644
--- a/Lib/test/test_cmd_line.py
+++ b/Lib/test/test_cmd_line.py
@@ -326,8 +326,26 @@ class CmdLineTest(unittest.TestCase):
         )
         test_args = [valid_utf8, invalid_utf8]
 
-        for run_cmd in (run_default, run_c_locale, run_utf8_mode):
-            with self.subTest(run_cmd=run_cmd):
+        for run_cmd, encoding in (
+            (run_default, sys.getfilesystemencoding()),
+            (run_c_locale, None),
+            (run_utf8_mode, None),
+        ):
+            with self.subTest(run_cmd=run_cmd.__name__):
+                # Arbitrary bytes round-trip through surrogateescape only in
+                # UTF-8 and single-byte encodings, not in a multibyte encoding
+                # such as EUC-JP.
+                if encoding is not None:
+                    try:
+                        lossless = len(bytes(range(256)).decode(
+                            encoding, 'surrogateescape')) == 256
+                    except UnicodeError:
+                        lossless = False
+                else:
+                    lossless = True
+                if not lossless:
+                    self.skipTest(f'{encoding} cannot losslessly '
+                                  f'round-trip arbitrary bytes')
                 for arg in test_args:
                     proc = run_cmd(arg)
                     self.assertEqual(proc.stdout.rstrip(), ascii(arg))