All platforms use same code to remove context-dependent codepoints from text input.

This commit introduces function fl_utf8_remove_context_dependent() that removes
from an UTF-8 string its context-dependent codepoints. Platforms macOS, Wayland
and X11 call this function to process UTF-8 text received from a character palette
as input to FLTK text. This makes sure FLTK text-editing widgets process textual input
equally and consistently across platforms, especially emojis entered via a palette.

Platform Windows creates a series of separate system events to input an emoji
via the character palette. For this reason, function fl_utf8_remove_context_dependent()
is not used by this platform which does internally the same filtering of context-
dependent codepoints.

4 files changed, 35 insertions, 17 deletions

diff --git a/src/Fl_cocoa.mm b/src/Fl_cocoa.mm
index eeeedf425..fed978e1e 100644
--- a/src/Fl_cocoa.mm
+++ b/src/Fl_cocoa.mm
@@ -67,6 +67,7 @@ extern "C" {
 // external functions
 extern void fl_fix_focus();
 extern int fl_send_system_handlers(void *e);
+extern int fl_utf8_remove_context_dependent(char *text, int len);
 
 // forward definition of functions in this file
 // converting cr lf converter function
@@ -2956,21 +2957,7 @@ static FLTextInputContext* fltextinputcontext_instance = nil;
   // insertText sent during handleEvent of a key without text cannot be processed in a single FL_KEYBOARD event.
   // Occurs with deadkey followed by non-text key. Occurs also with emoji palette.
   if (!in_key_event || !has_text_key) {
-    if (fl_utf_nb_char((const uchar*)Fl::e_text, Fl::e_length) > 1) {
-      // Some emojis are expressed by a series of Unicode points
-      const char *p = Fl::e_text, *end = Fl::e_text + Fl::e_length;
-      int len;
-      while (p < end) { // loop over all unicode points of the series
-        unsigned u = fl_utf8decode(p, end, &len); // extract one such unicode point
-        if ((u >= 0xFE00 && u <= 0xFE0F)      // variation selectors
-            || u == 0x200D   // zero-width joiner
-            || (u >= 0x1F3FB && u <= 0x1F3FF) // EMOJI MODIFIERS FITZPATRICK TYPE
-            ) { // remove context-dependent unicode points
-          memmove((void*)p, p + len, (end - (p+len)) + 1);
-          Fl::e_length -= len; end -= len;
-        } else p += len; // keep other unicode points
-      }
-    }
+    Fl::e_length = fl_utf8_remove_context_dependent(Fl::e_text, Fl::e_length);
     Fl::handle(FL_KEYBOARD, target);
     Fl::e_length = 0;
     }
diff --git a/src/Fl_x.cxx b/src/Fl_x.cxx
index fa58c8076..150438755 100644
--- a/src/Fl_x.cxx
+++ b/src/Fl_x.cxx
@@ -1,7 +1,7 @@
 //
 // X specific code for the Fast Light Tool Kit (FLTK).
 //
-// Copyright 1998-2025 by Bill Spitzak and others.
+// Copyright 1998-2026 by Bill Spitzak and others.
 //
 // This library is free software. Distribution and use rights are outlined in
 // the file "COPYING" which should have been included with this file.  If this
@@ -101,6 +101,7 @@ extern Fl_Window *fl_xmousewin;
 
 static void open_display_i(Display *d); // open display (internal)
 extern int fl_send_system_handlers(void *e);
+extern int fl_utf8_remove_context_dependent(char *text, int len);
 
 #if FLTK_CONSOLIDATE_MOTION
 static Fl_Window *send_motion;
@@ -1833,6 +1834,7 @@ int fl_handle(const XEvent& thisevent)
         Status status;
         len = XUtf8LookupString(Fl_X11_Screen_Driver::xim_ic, (XKeyPressedEvent *)&xevent.xkey,
                              kp_buffer, kp_buffer_len, &keysym, &status);
+        len = fl_utf8_remove_context_dependent(kp_buffer, len);
 
         while (status == XBufferOverflow && kp_buffer_len < 50000) {
           kp_buffer_len = kp_buffer_len * 5 + 1;
diff --git a/src/drivers/Wayland/Fl_Wayland_Screen_Driver.cxx b/src/drivers/Wayland/Fl_Wayland_Screen_Driver.cxx
index c474867d5..3816dff2c 100644
--- a/src/drivers/Wayland/Fl_Wayland_Screen_Driver.cxx
+++ b/src/drivers/Wayland/Fl_Wayland_Screen_Driver.cxx
@@ -53,6 +53,7 @@ extern "C" {
   bool fl_is_surface_from_GTK_titlebar (struct wl_surface *surface, struct libdecor_frame *frame,
                                         bool *using_GTK);
 }
+extern int fl_utf8_remove_context_dependent(char *text, int len);
 
 // set this to 1 for keyboard debug output, 0 for no debug output
 #define DEBUG_KEYBOARD 0
@@ -1017,6 +1018,7 @@ void text_input_commit_string(void *data, struct zwp_text_input_v3 *zwp_text_inp
 //printf("text_input_commit_string %s\n",text);
   free(pending_commit);
   pending_commit = text ? strdup(text) : NULL;
+  fl_utf8_remove_context_dependent(pending_commit, strlen(pending_commit));
 }
 
 
diff --git a/src/fl_utf8.cxx b/src/fl_utf8.cxx
index 7cc53f13b..05e45ded3 100644
--- a/src/fl_utf8.cxx
+++ b/src/fl_utf8.cxx
@@ -3,7 +3,7 @@
 //
 // Author: Jean-Marc Lienher ( http://oksid.ch )
 // Copyright 2000-2010 by O'ksi'D.
-// Copyright 2016-2022 by Bill Spitzak and others.
+// Copyright 2016-2026 by Bill Spitzak and others.
 //
 // This library is free software. Distribution and use rights are outlined in
 // the file "COPYING" which should have been included with this file.  If this
@@ -1630,3 +1630,30 @@ unsigned fl_utf8from_mb(char* dst, unsigned dstlen, const char* src, unsigned sr
 }
 
 /** @} */
+
+#ifndef FL_DOXYGEN
+
+/* This function removes from an UTF-8 string its context-dependent codepoints
+ when there are any, and returns the length of the possibly shortened string.
+ */
+int fl_utf8_remove_context_dependent(char *text, int len) {
+  if (len > 1 && fl_utf_nb_char((const uchar*)text, len) > 1) {
+    // Some emojis are expressed by a series of Unicode points
+    char *p = text, *end = text + len;
+    while (p < end) { // loop over all unicode points of the series
+      int l_point;
+      unsigned u = fl_utf8decode(p, end, &l_point); // extract one such unicode point
+      if ((u >= 0xFE00 && u <= 0xFE0F)      // variation selectors
+          || u == 0x200D                    // zero-width joiner
+          || (u >= 0x1F3FB && u <= 0x1F3FF) // EMOJI MODIFIERS FITZPATRICK TYPE
+          ) { // remove context-dependent unicode points
+        memmove((void*)p, p + l_point, (end - (p+l_point)) + 1);
+        end -= l_point;
+        len -= l_point;
+      } else p += l_point; // keep other unicode points
+    }
+  }
+  return len;
+}
+
+#endif // ! FL_DOXYGEN