From 8e2dcca9dc43b9d24dfb8874a629279b50b6645d Mon Sep 17 00:00:00 2001
From: ycjcl868 <chaolinjin@gmail.com>
Date: Sat, 21 Mar 2026 22:25:38 +0800
Subject: [PATCH] fix: dense cjk utf-8 content misdetected as binary

---
 openhands_aci/editor/editor.py      | 13 ++++++++++++-
 tests/integration/test_oh_editor.py |  9 +++++++++
 tests/unit/test_file_validation.py  |  8 ++++++++
 3 files changed, 29 insertions(+), 1 deletion(-)

diff --git a/openhands_aci/editor/editor.py b/openhands_aci/editor/editor.py
index a968726..c291b91 100644
--- a/openhands_aci/editor/editor.py
+++ b/openhands_aci/editor/editor.py
@@ -608,7 +608,18 @@ def validate_file(self, path: Path) -> None:
         if self.is_supported_binary_file(path):
             return
 
-        # Check file type
+        # Try encoding-based text detection first (handles CJK/multi-byte UTF-8)
+        try:
+            encoding = self._encoding_manager.get_encoding(path)
+            with open(path, 'r', encoding=encoding) as f:
+                chunk = f.read(8192)
+            # Null characters indicate binary content even if decoding succeeded
+            if '\x00' not in chunk:
+                return  # Successfully decoded as text — not binary
+        except (UnicodeDecodeError, ValueError, OSError, TypeError):
+            pass  # Fall through to binaryornot check
+
+        # Fallback to binaryornot
         if is_binary(str(path)):
             raise FileValidationError(
                 path=str(path),
diff --git a/tests/integration/test_oh_editor.py b/tests/integration/test_oh_editor.py
index 9a95a82..3ac55e1 100644
--- a/tests/integration/test_oh_editor.py
+++ b/tests/integration/test_oh_editor.py
@@ -699,3 +699,12 @@ def test_str_replace_and_insert_snippet_output_on_a_large_file(editor):
         new_str='Inserted line at 500',
     )
     assert '   500\tInserted line at 500' in result.output
+
+
+def test_view_dense_chinese_markdown(tmp_path):
+    """view should handle UTF-8 markdown files with dense CJK content."""
+    md_file = tmp_path / 'test.md'
+    md_file.write_text('中文测试内容。' * 50, encoding='utf-8')
+    editor = OHEditor()
+    result = editor(command='view', path=str(md_file))
+    assert '中文测试内容' in result.output
diff --git a/tests/unit/test_file_validation.py b/tests/unit/test_file_validation.py
index 422f371..aba5249 100644
--- a/tests/unit/test_file_validation.py
+++ b/tests/unit/test_file_validation.py
@@ -95,3 +95,11 @@ def test_validate_image_file():
         editor.validate_file(image_file)
 
     assert 'file appears to be binary' in str(exc_info.value).lower()
+
+
+def test_validate_dense_cjk_utf8_not_binary(tmp_path):
+    """Dense CJK UTF-8 text should not be rejected as binary."""
+    f = tmp_path / 'chinese.md'
+    f.write_text('中文测试内容。' * 50, encoding='utf-8')
+    editor = OHEditor()
+    editor.validate_file(f)  # Should not raise