From 6ce207039240a63120a91ee4ba4d39bd0a3ef8c8 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Camilla=20L=C3=B6wy?= <elmindreda@elmindreda.org>
Date: Mon, 29 Jun 2020 20:43:28 +0200
Subject: [PATCH] Win32: Fix non-BMP Unicode codepoint input

Supplementary Plane codepoints from WM_CHAR and WM_SYSCHAR messages were
reported as UTF-16 surrogate pairs.

Related to #1635.
---
 README.md            |  1 +
 src/win32_platform.h |  2 ++
 src/win32_window.c   | 39 +++++++++++++++++++++++++++++++--------
 3 files changed, 34 insertions(+), 8 deletions(-)

diff --git a/README.md b/README.md
index 90a1196f..64416958 100644
--- a/README.md
+++ b/README.md
@@ -152,6 +152,7 @@ information on what to include when reporting a bug.
    invalid pointer
  - [Win32] Bugfix: Some synthetic key events were reported as `GLFW_KEY_UNKNOWN`
    (#1623)
+ - [Win32] Bugfix: Non-BMP Unicode codepoint input was reported as UTF-16
  - [Cocoa] Added support for `VK_EXT_metal_surface` (#1619)
  - [Cocoa] Added locating the Vulkan loader at runtime in an application bundle
  - [Cocoa] Moved main menu creation to GLFW initialization time (#1649)
diff --git a/src/win32_platform.h b/src/win32_platform.h
index 5b260b82..76ea609a 100644
--- a/src/win32_platform.h
+++ b/src/win32_platform.h
@@ -311,6 +311,8 @@ typedef struct _GLFWwindowWin32
 
     // The last received cursor position, regardless of source
     int                 lastCursorPosX, lastCursorPosY;
+    // The last recevied high surrogate when decoding pairs of UTF-16 messages
+    WCHAR               highSurrogate;
 
 } _GLFWwindowWin32;
 
diff --git a/src/win32_window.c b/src/win32_window.c
index d9c30c55..9dc52bab 100644
--- a/src/win32_window.c
+++ b/src/win32_window.c
@@ -646,11 +646,38 @@ static LRESULT CALLBACK windowProc(HWND hWnd, UINT uMsg,
 
         case WM_CHAR:
         case WM_SYSCHAR:
+        {
+            if (wParam >= 0xd800 && wParam <= 0xdbff)
+                window->win32.highSurrogate = (WCHAR) wParam;
+            else
+            {
+                unsigned int codepoint = 0;
+
+                if (wParam >= 0xdc00 && wParam <= 0xdfff)
+                {
+                    if (window->win32.highSurrogate)
+                    {
+                        codepoint += (window->win32.highSurrogate - 0xd800) << 10;
+                        codepoint += (WCHAR) wParam - 0xdc00;
+                        codepoint += 0x10000;
+                    }
+                }
+                else
+                    codepoint = (WCHAR) wParam;
+
+                window->win32.highSurrogate = 0;
+                _glfwInputChar(window, codepoint, getKeyMods(), uMsg != WM_SYSCHAR);
+            }
+
+            if (uMsg == WM_SYSCHAR && window->win32.keymenu)
+                break;
+
+            return 0;
+        }
+
         case WM_UNICHAR:
         {
-            const GLFWbool plain = (uMsg != WM_SYSCHAR);
-
-            if (uMsg == WM_UNICHAR && wParam == UNICODE_NOCHAR)
+            if (wParam == UNICODE_NOCHAR)
             {
                 // WM_UNICHAR is not sent by Windows, but is sent by some
                 // third-party input method engine
@@ -658,11 +685,7 @@ static LRESULT CALLBACK windowProc(HWND hWnd, UINT uMsg,
                 return TRUE;
             }
 
-            _glfwInputChar(window, (unsigned int) wParam, getKeyMods(), plain);
-
-            if (uMsg == WM_SYSCHAR && window->win32.keymenu)
-                break;
-
+            _glfwInputChar(window, (unsigned int) wParam, getKeyMods(), GLFW_TRUE);
             return 0;
         }