From 0b7b8ba4a22786c327a739be67e285ded9348b34 Mon Sep 17 00:00:00 2001
From: Jeroen van Rijn <Kelimion@users.noreply.github.com>
Date: Sun, 16 Aug 2026 16:16:35 +0200
Subject: [PATCH] Set console codepage to utf-8 (on Windows), and restore old
 one on exit

Avoid mojibake.
---
 base/runtime/entry_windows.odin | 59 ++++++++++++++++++++++++++++-----
 src/common.cpp                  | 14 ++++++++
 src/main.cpp                    |  7 ++++
 3 files changed, 72 insertions(+), 8 deletions(-)

diff --git a/base/runtime/entry_windows.odin b/base/runtime/entry_windows.odin
index d3b38bc9c..56cb9063d 100644
--- a/base/runtime/entry_windows.odin
+++ b/base/runtime/entry_windows.odin
@@ -5,6 +5,14 @@ package runtime
 
 import "base:intrinsics"
 
+when !ODIN_BEDROCK {
+	@(private="file")
+	old_console_codepage: u32
+
+	@(private="file")
+	UTF_8 :: 65001
+}
+
 when ODIN_BUILD_MODE == .Dynamic {
 	@(link_name="DllMain", linkage="strong", require)
 	DllMain :: proc "system" (hinstDLL: rawptr, fdwReason: u32, lpReserved: rawptr) -> b32 {
@@ -16,10 +24,17 @@ when ODIN_BUILD_MODE == .Dynamic {
 
 		switch dll_forward_reason {
 		case .Process_Attach:
-			when !ODIN_BEDROCK { #force_no_inline _startup_runtime() }
+			when !ODIN_BEDROCK {
+				#force_no_inline _startup_runtime()
+				old_console_codepage = GetConsoleOutputCP()
+				SetConsoleOutputCP(UTF_8)
+			}
 			intrinsics.__entry_point()
 		case .Process_Detach:
-			when !ODIN_BEDROCK { #force_no_inline _cleanup_runtime() }
+			when !ODIN_BEDROCK {
+				#force_no_inline _cleanup_runtime()
+				SetConsoleOutputCP(old_console_codepage)
+			}
 		case .Thread_Attach:
 			break
 		case .Thread_Detach:
@@ -35,18 +50,32 @@ when ODIN_BUILD_MODE == .Dynamic {
 		main :: proc "c" (argc: i32, argv: [^]cstring) -> i32 {
 			args__ = argv[:argc]
 			context = default_context()
-			when !ODIN_BEDROCK { #force_no_inline _startup_runtime() }
+			when !ODIN_BEDROCK {
+				#force_no_inline _startup_runtime()
+				old_console_codepage = GetConsoleOutputCP()
+				SetConsoleOutputCP(UTF_8)
+			}
 			intrinsics.__entry_point()
-			when !ODIN_BEDROCK { #force_no_inline _cleanup_runtime() }
+			when !ODIN_BEDROCK {
+				#force_no_inline _cleanup_runtime()
+				SetConsoleOutputCP(old_console_codepage)
+			}
 			return 0
 		}
 	} else when ODIN_NO_CRT {
 		@(link_name="mainCRTStartup", linkage="strong", require)
 		mainCRTStartup :: proc "system" () -> i32 {
 			context = default_context()
-			when !ODIN_BEDROCK { #force_no_inline _startup_runtime() }
+			when !ODIN_BEDROCK {
+				#force_no_inline _startup_runtime()
+				old_console_codepage = GetConsoleOutputCP()
+				SetConsoleOutputCP(UTF_8)
+			}
 			intrinsics.__entry_point()
-			when !ODIN_BEDROCK { #force_no_inline _cleanup_runtime() }
+			when !ODIN_BEDROCK {
+				#force_no_inline _cleanup_runtime()
+				SetConsoleOutputCP(old_console_codepage)
+			}
 			return 0
 		}
 	} else {
@@ -54,10 +83,24 @@ when ODIN_BUILD_MODE == .Dynamic {
 		main :: proc "c" (argc: i32, argv: [^]cstring) -> i32 {
 			args__ = argv[:argc]
 			context = default_context()
-			when !ODIN_BEDROCK { #force_no_inline _startup_runtime() }
+			when !ODIN_BEDROCK {
+				#force_no_inline _startup_runtime()
+				old_console_codepage = GetConsoleOutputCP()
+				SetConsoleOutputCP(UTF_8)
+			}
 			intrinsics.__entry_point()
-			when !ODIN_BEDROCK { #force_no_inline _cleanup_runtime() }
+			when !ODIN_BEDROCK {
+				#force_no_inline _cleanup_runtime()
+				SetConsoleOutputCP(old_console_codepage)
+			}
 			return 0
 		}
 	}
+}
+
+foreign import kernel32 "system:Kernel32.lib"
+@(default_calling_convention="system")
+foreign kernel32 {
+	GetConsoleOutputCP :: proc() -> u32 ---
+	SetConsoleOutputCP :: proc(codepage: u32) -> b32 ---
 }
\ No newline at end of file
diff --git a/src/common.cpp b/src/common.cpp
index 8338ec707..3b4557cde 100644
--- a/src/common.cpp
+++ b/src/common.cpp
@@ -65,6 +65,20 @@ template <typename T> struct TypeIs64BitInteger { enum {value = false}; };
 template <> struct TypeIs64BitInteger<u64> { enum {value = true}; };
 template <> struct TypeIs64BitInteger<i64> { enum {value = true}; };
 
+#if defined(GB_SYSTEM_WINDOWS)
+uint32_t old_console_codepage = 0;
+void set_utf8_codepage() {
+	old_console_codepage = GetConsoleOutputCP();
+}
+
+void restore_old_codepage() {
+	// Nothing we can do if this fails, so we're not asserting or anything.
+	SetConsoleOutputCP(old_console_codepage);
+}
+#else
+void set_utf8_codepage() {}
+void restore_old_codepage() {}
+#endif
 
 
 #include "unicode.cpp"
diff --git a/src/main.cpp b/src/main.cpp
index 74900587f..7d2f4d713 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -3794,6 +3794,13 @@ int main(int arg_count, char const **arg_ptr) {
 	defer (timings_destroy(&global_timings));
 
 	MAIN_TIME_SECTION("initialization");
+	// NOTE(Jeroen): Set codepage to UTF-8 (Windows only) and restore on exit.
+	//               Keep in mind this is for the compiler's own output only.
+	//               Like error messages on lines containing unicode.
+	//               Child processes will inherit the default codepage,
+	//               and so must do their own codepage management if they want.
+	set_utf8_codepage();
+	defer (restore_old_codepage());
 
 	init_string_interner();
 	init_global_error_collector();