send_unicode_string(): Add support for code points > 0xFFFF (#8236)

docs/feature_unicode.md

@@ -195,7 +195,7 @@ By default, when the keyboard boots, it will initialize the input mode to the la
 
 ## `send_unicode_string()`
 
-This function is much like `send_string()` but allows you to input UTF-8 characters directly, currently up to code point U+FFFF. Make sure your `keymap.c` is formatted in UTF-8 encoding.
+This function is much like `send_string()` but allows you to input UTF-8 characters directly, and supports all code points (provided the selected input method also supports it). Make sure your `keymap.c` is formatted in UTF-8 encoding.
 
 ```c
 send_unicode_string("(ノಠ痊ಠ)ノ彡┻━┻");
@@ -210,7 +210,6 @@ send_unicode_hex_string("0028 30CE 0CA0 75CA 0CA0 0029 30CE 5F61 253B 2501 253B"
 ```
 
 An easy way to convert your Unicode string to this format is by using [this site](https://r12a.github.io/app-conversion/), and taking the result in the "Hex/UTF-32" section.
-Unlike `send_unicode_string()` this function supports code points up to U+10FFFF.
 
 ## Additional Language Support
 

quantum/process_keycode/process_unicode_common.c

@@ -150,6 +150,24 @@ void register_hex(uint16_t hex) {
     }
 }
 
+void register_hex32(uint32_t hex) {
+    bool onzerostart = true;
+    for (int i = 7; i >= 0; i--) {
+        if (i <= 3) {
+            onzerostart = false;
+        }
+        uint8_t digit = ((hex >> (i * 4)) & 0xF);
+        if (digit == 0) {
+            if (!onzerostart) {
+                tap_code(hex_to_keycode(digit));
+            }
+        } else {
+            tap_code(hex_to_keycode(digit));
+            onzerostart = false;
+        }
+    }
+}
+
 void send_unicode_hex_string(const char *str) {
     if (!str) {
         return;
@@ -192,9 +210,7 @@ const char *decode_utf8(const char *str, int32_t *code_point) {
         *code_point = ((int32_t)(str[0] & 0x0F) << 12) | ((int32_t)(str[1] & 0x3F) << 6) | ((int32_t)(str[2] & 0x3F) << 0);
         next        = str + 3;
     } else if ((str[0] & 0xF8) == 0xF0 && (str[0] <= 0xF4)) {  // U+10000-10FFFF
-        // Skip for now - register_hex() only takes a uint16
+        *code_point = ((int32_t)(str[0] & 0x07) << 18) | ((int32_t)(str[1] & 0x3F) << 12) | ((int32_t)(str[2] & 0x3F) << 6) | ((int32_t)(str[3] & 0x3F) << 0);
-        //*code_point = ((int32_t)(str[0] & 0x07) << 18) | ((int32_t)(str[1] & 0x3F) << 12) | ((int32_t)(str[2] & 0x3F) << 6) | ((int32_t)(str[3] & 0x3F) << 0);
-        *code_point = -1;
         next        = str + 4;
     } else {
         *code_point = -1;
@@ -221,7 +237,7 @@ void send_unicode_string(const char *str) {
 
         if (code_point >= 0) {
             unicode_input_start();
-            register_hex(code_point);
+            register_hex32(code_point);
             unicode_input_finish();
         }
     }

quantum/process_keycode/process_unicode_common.h

@@ -79,6 +79,7 @@ void unicode_input_finish(void);
 void unicode_input_cancel(void);
 
 void register_hex(uint16_t hex);
+void register_hex32(uint32_t hex);
 void send_unicode_hex_string(const char *str);
 void send_unicode_string(const char *str);
 

quantum/process_keycode/process_unicodemap.c

@@ -16,26 +16,6 @@
 
 #include "process_unicodemap.h"
 
-void register_hex32(uint32_t hex) {
-    bool onzerostart = true;
-    for (int i = 7; i >= 0; i--) {
-        if (i <= 3) {
-            onzerostart = false;
-        }
-        uint8_t digit = ((hex >> (i * 4)) & 0xF);
-        if (digit == 0) {
-            if (!onzerostart) {
-                register_code(hex_to_keycode(digit));
-                unregister_code(hex_to_keycode(digit));
-            }
-        } else {
-            register_code(hex_to_keycode(digit));
-            unregister_code(hex_to_keycode(digit));
-            onzerostart = false;
-        }
-    }
-}
-
 __attribute__((weak)) uint16_t unicodemap_index(uint16_t keycode) {
     if (keycode >= QK_UNICODEMAP_PAIR) {
         // Keycode is a pair: extract index based on Shift / Caps Lock state

quantum/process_keycode/process_unicodemap.h

@@ -20,6 +20,5 @@
 
 extern const uint32_t PROGMEM unicode_map[];
 
-void     register_hex32(uint32_t hex);
 uint16_t unicodemap_index(uint16_t keycode);
 bool     process_unicodemap(uint16_t keycode, keyrecord_t *record);