From 64f4f5c35e13afe8c2b86475f19e39cb1de21312 Mon Sep 17 00:00:00 2001
From: 0xacee <0xacee@users.noreply.github.com>
Date: Mon, 21 Sep 2026 16:03:24 +0000
Subject: [PATCH] sort-simple-yaml: read and write files as UTF-8

---
 pre_commit_hooks/sort_simple_yaml.py |  2 +-
 tests/sort_simple_yaml_test.py       | 12 ++++++++++++
 2 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/pre_commit_hooks/sort_simple_yaml.py b/pre_commit_hooks/sort_simple_yaml.py
index 65e6b7a6..7b94a402 100644
--- a/pre_commit_hooks/sort_simple_yaml.py
+++ b/pre_commit_hooks/sort_simple_yaml.py
@@ -107,7 +107,7 @@ def main(argv: Sequence[str] | None = None) -> int:
     retval = 0
 
     for filename in args.filenames:
-        with open(filename, 'r+') as f:
+        with open(filename, 'r+', encoding='utf-8', newline='') as f:
             lines = [line.rstrip() for line in f.readlines()]
             new_lines = sort(lines)
 
diff --git a/tests/sort_simple_yaml_test.py b/tests/sort_simple_yaml_test.py
index 6cbda857..62fad71b 100644
--- a/tests/sort_simple_yaml_test.py
+++ b/tests/sort_simple_yaml_test.py
@@ -52,6 +52,18 @@ def test_integration_good_bad_lines(tmpdir, bad_lines, good_lines, retval):
         assert [line.rstrip() for line in f.readlines()] == good_lines
 
 
+def test_non_ascii_content_roundtrip(tmpdir):
+    file_path = os.path.join(str(tmpdir), 'foo.yaml')
+
+    with open(file_path, 'w', encoding='utf-8') as f:
+        f.write('b: caf\u00e9\n\na: \u2603\n')
+
+    assert main([file_path]) == RETVAL_BAD
+
+    with open(file_path, encoding='utf-8') as f:
+        assert f.read() == 'a: \u2603\n\nb: caf\u00e9\n'
+
+
 def test_parse_header():
     lines = ['# some header', '# is here', '', 'this is not a header']
     assert parse_block(lines, header=True) == ['# some header', '# is here']