fix: resolve Windows UTF-8 encoding issue for international characters

[eLyiN]

Summary

• Fixes Windows UTF-8 encoding issue where international characters (Chinese, etc.) were corrupted
• Improves Windows CLI detection with additional paths and extensions
• Enhances error diagnostics with platform-specific messages

Problem

Issue #1 reported that UTF-8 characters passed through MCP were corrupted on Windows. This was caused by the subprocess module using the system's default code page (e.g., cp1252, cp936) instead of UTF-8.

Solution

UTF-8 Encoding Fix

• Use explicit UTF-8 encoding for subprocess I/O on Windows
• Set PYTHONUTF8=1 and PYTHONIOENCODING=utf-8 environment variables
• Encode input as UTF-8 bytes and decode output with errors='replace'

Enhanced Windows CLI Detection

• Added .ps1 to Windows executable extensions
• Added fallback checks for common Windows installation paths:
  • %LOCALAPPDATA%\Programs\codex\codex.exe
  • %APPDATA%\npm\codex.cmd
  • %USERPROFILE%\.cargo\bin\codex.exe

Improved Error Diagnostics

• Added FileNotFoundError specific handling
• Error responses now include platform and exception_type fields
• Clear instructions for Windows users to verify installation

Test Plan

• Python syntax validation passes (python -m py_compile)
• All version numbers updated to 1.2.3
• CHANGELOG.md updated with v1.2.2 and v1.2.3 entries
• Windows user verification (requires feedback from @bianoengpeng)

Files Changed

• src/mcp_server.py - Core fix for UTF-8 encoding and error handling
• src/__init__.py - Version bump to 1.2.3
• pyproject.toml - Version bump to 1.2.3
• CHANGELOG.md - Added v1.2.2 and v1.2.3 entries

Closes #1

[Copilot]

Pull request overview

This pull request fixes a Windows UTF-8 encoding issue where international characters (Chinese, etc.) were corrupted when using the MCP server on Windows. The fix explicitly sets UTF-8 encoding for subprocess I/O and adds enhanced Windows CLI detection with additional installation paths and PowerShell support.

• Implements UTF-8 encoding for subprocess I/O on Windows to replace system default code page handling
• Enhances Windows CLI detection by adding .ps1 extension support and checking common installation paths
• Improves error diagnostics with FileNotFoundError handling and platform-specific metadata

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File	Description
src/mcp_server.py	Core UTF-8 encoding fix in _run_codex_command(), enhanced Windows CLI detection in _get_codex_command(), new _build_codex_exec_command() helper, and improved FileNotFoundError handling
src/init.py	Version bump from 1.2.2 to 1.2.3 with updated description
pyproject.toml	Version bump from 1.2.2 to 1.2.3
CHANGELOG.md	Added entries for v1.2.3 and v1.2.2 with detailed changelog of UTF-8 fixes and Windows compatibility improvements

Comments suppressed due to low confidence (1)

src/mcp_server.py:699

• The general Exception handler in the batch function is missing the exception_type and platform fields in its metadata, which are included in the same handler for consult_codex (line 420) and consult_codex_with_stdin (line 555). For consistency and better debugging, the metadata should include these fields: "platform": platform.system() and "exception_type": type(e).name.

        except Exception as e:
            results.append({
                "status": "error",
                "index": i,
                "query": query[:100] + "..." if len(query) > 100 else query,
                "error": f"Error executing query: {str(e)}",
                "metadata": {}
            })

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.