InputText: restore truncating, now between UTF-8 codepoint. (#9029)

Amend e612536, 2a194e2.

docs/CHANGELOG.txt

@@ -49,6 +49,9 @@ Other Changes:
   EndTable() was mistakenly restoring a wrong current table.
 - Disabled: fixed a bug when a previously enabled item that got nav focus
   and then turns disabled could still be activated using keyboard. (#9036)
+- InputText: when buffer is not resizable, trying to paste contents that
+  cannot fit will now truncate text to nearest UTF-8 codepoint boundaries,
+  instead of completely ignoring the paste. (#9029)
 - InputText: avoid continuously overwriting ownership of ImGuiKey_Enter/_KeypadEnter
   keys in order to allow e.g. external Shortcut override behavior. (#9004)
 - InputText: when using a callback to reduce/manipulate the value of BufTextLen,

imgui.cpp

@@ -2663,17 +2663,29 @@ int ImTextCountUtf8BytesFromStr(const ImWchar* in_text, const ImWchar* in_text_e
     return bytes_count;
 }
 
-const char* ImTextFindPreviousUtf8Codepoint(const char* in_text_start, const char* in_text_curr)
+const char* ImTextFindPreviousUtf8Codepoint(const char* in_text_start, const char* in_p)
 {
-    while (in_text_curr > in_text_start)
+    while (in_p > in_text_start)
     {
-        in_text_curr--;
+        in_p--;
-        if ((*in_text_curr & 0xC0) != 0x80)
+        if ((*in_p & 0xC0) != 0x80)
-            return in_text_curr;
+            return in_p;
     }
     return in_text_start;
 }
 
+const char* ImTextFindValidUtf8CodepointEnd(const char* in_text_start, const char* in_text_end, const char* in_p)
+{
+    if (in_text_start == in_p)
+        return in_text_start;
+    const char* prev = ImTextFindPreviousUtf8Codepoint(in_text_start, in_p);
+    unsigned int prev_c;
+    int prev_c_len = ImTextCharFromUtf8(&prev_c, prev, in_text_end);
+    if (prev_c != IM_UNICODE_CODEPOINT_INVALID && prev_c_len <= (int)(in_p - prev))
+        return in_p;
+    return prev;
+}
+
 int ImTextCountLines(const char* in_text, const char* in_text_end)
 {
     if (in_text_end == NULL)

imgui_internal.h

@@ -428,7 +428,8 @@ IMGUI_API int           ImTextStrFromUtf8(ImWchar* out_buf, int out_buf_size, co
 IMGUI_API int           ImTextCountCharsFromUtf8(const char* in_text, const char* in_text_end);                                 // return number of UTF-8 code-points (NOT bytes count)
 IMGUI_API int           ImTextCountUtf8BytesFromChar(const char* in_text, const char* in_text_end);                             // return number of bytes to express one char in UTF-8
 IMGUI_API int           ImTextCountUtf8BytesFromStr(const ImWchar* in_text, const ImWchar* in_text_end);                        // return number of bytes to express string in UTF-8
-IMGUI_API const char*   ImTextFindPreviousUtf8Codepoint(const char* in_text_start, const char* in_text_curr);                   // return previous UTF-8 code-point.
+IMGUI_API const char*   ImTextFindPreviousUtf8Codepoint(const char* in_text_start, const char* in_p);                           // return previous UTF-8 code-point.
+IMGUI_API const char*   ImTextFindValidUtf8CodepointEnd(const char* in_text_start, const char* in_text_end, const char* in_p);  // return previous UTF-8 code-point if 'in_p' is not the end of a valid one.
 IMGUI_API int           ImTextCountLines(const char* in_text, const char* in_text_end);                                         // return number of lines taken by text. trailing carriage return doesn't count as an extra line.
 
 // Helpers: High-level text functions (DO NOT USE!!! THIS IS A MINIMAL SUBSET OF LARGER UPCOMING CHANGES)

imgui_widgets.cpp

@@ -4155,7 +4155,7 @@ static int STB_TEXTEDIT_INSERTCHARS(ImGuiInputTextState* obj, int pos, const cha
     // We support partial insertion (with a mod in stb_textedit.h)
     const int avail = obj->BufCapacity - 1 - obj->TextLen;
     if (!is_resizable && new_text_len > avail)
-        new_text_len = 0; // avail
+        new_text_len = (int)(ImTextFindValidUtf8CodepointEnd(new_text, new_text + new_text_len, new_text + avail) - new_text); // Truncate to closest UTF-8 codepoint. Alternative: return 0 to cancel insertion.
     if (new_text_len == 0)
         return 0;
 
@@ -4314,7 +4314,7 @@ void ImGuiInputTextCallbackData::InsertChars(int pos, const char* new_text, cons
     // We support partial insertion (with a mod in stb_textedit.h)
     const int avail = BufSize - 1 - BufTextLen;
     if (!is_resizable && new_text_len > avail)
-        new_text_len = 0; // avail
+        new_text_len = (int)(ImTextFindValidUtf8CodepointEnd(new_text, new_text + new_text_len, new_text + avail) - new_text); // Truncate to closest UTF-8 codepoint. Alternative: return 0 to cancel insertion.
     if (new_text_len == 0)
         return;