Issue #25318: Avoid sprintf() in backslashreplace()

Rewrite backslashreplace() to be closer to PyCodec_BackslashReplaceErrors().

Add also unit tests for non-BMP characters.

1 files changed, 4 insertions, 2 deletions

diff --git a/Lib/test/test_codecs.py b/Lib/test/test_codecs.py
index 7b6883f..ff314b1 100644
--- a/Lib/test/test_codecs.py
+++ b/Lib/test/test_codecs.py
@@ -3155,7 +3155,8 @@ class ASCIITest(unittest.TestCase):
             ('[\x80\xff\u20ac]', 'ignore', b'[]'),
             ('[\x80\xff\u20ac]', 'replace', b'[???]'),
             ('[\x80\xff\u20ac]', 'xmlcharrefreplace', b'[€ÿ€]'),
-            ('[\x80\xff\u20ac]', 'backslashreplace', b'[\\x80\\xff\\u20ac]'),
+            ('[\x80\xff\u20ac\U000abcde]', 'backslashreplace',
+             b'[\\x80\\xff\\u20ac\\U000abcde]'),
             ('[\udc80\udcff]', 'surrogateescape', b'[\x80\xff]'),
         ):
             with self.subTest(data=data, error_handler=error_handler,
@@ -3197,7 +3198,8 @@ class Latin1Test(unittest.TestCase):
         for data, error_handler, expected in (
             ('[\u20ac\udc80]', 'ignore', b'[]'),
             ('[\u20ac\udc80]', 'replace', b'[??]'),
-            ('[\u20ac\udc80]', 'backslashreplace', b'[\\u20ac\\udc80]'),
+            ('[\u20ac\U000abcde]', 'backslashreplace',
+             b'[\\u20ac\\U000abcde]'),
             ('[\u20ac\udc80]', 'xmlcharrefreplace', b'[€�]'),
             ('[\udc80\udcff]', 'surrogateescape', b'[\x80\xff]'),
         ):