--- lbbs/src/str_process.c	2025/07/02 03:08:10	1.19
+++ lbbs/src/str_process.c	2025/11/07 13:42:45	1.25
@@ -1,25 +1,75 @@
-/***************************************************************************
-						  str_process.c  -  description
-							 -------------------
-	Copyright            : (C) 2004-2025 by Leaflet
-	Email                : leaflet@leafok.com
- ***************************************************************************/
-
-/***************************************************************************
- *                                                                         *
- *   This program is free software; you can redistribute it and/or modify  *
- *   it under the terms of the GNU General Public License as published by  *
- *   the Free Software Foundation; either version 3 of the License, or     *
- *   (at your option) any later version.                                   *
- *                                                                         *
- ***************************************************************************/
+/* SPDX-License-Identifier: GPL-3.0-or-later */
+/*
+ * str_process
+ *   - common string processing features with UTF-8 support
+ *
+ * Copyright (C) 2004-2025  Leaflet <leaflet@leafok.com>
+ */
 
 #include "common.h"
 #include "log.h"
 #include "str_process.h"
+#include <ctype.h>
 #include <stdio.h>
 #include <string.h>
 
+int str_length(const char *str, int skip_ctrl_seq)
+{
+	int i;
+	char c;
+	int ret = 0;
+
+	for (i = 0; str[i] != '\0'; i++)
+	{
+		c = str[i];
+
+		if (c == '\r' || c == '\7') // skip
+		{
+			continue;
+		}
+
+		if (skip_ctrl_seq && c == '\033' && str[i + 1] == '[') // Skip control sequence
+		{
+			for (i = i + 2; isdigit(str[i]) || str[i] == ';' || str[i] == '?'; i++)
+				;
+
+			if (str[i] == 'm') // valid
+			{
+				// skip
+			}
+			else if (isalpha(str[i]))
+			{
+				// unsupported ANSI CSI command
+			}
+			else
+			{
+				i--;
+			}
+
+			continue;
+		}
+
+		// Process UTF-8 Chinese characters
+		if (c & 0x80) // head of multi-byte character
+		{
+			c = (c & 0x70) << 1;
+			while (c & 0x80)
+			{
+				i++;
+				c = (c & 0x7f) << 1;
+			}
+
+			ret += 2;
+		}
+		else
+		{
+			ret++;
+		}
+	}
+
+	return ret;
+}
+
 int split_line(const char *buffer, int max_display_len, int *p_eol, int *p_display_len, int skip_ctrl_seq)
 {
 	int i;
@@ -46,13 +96,20 @@ int split_line(const char *buffer, int m
 			continue;
 		}
 
-		if (c < 0 || c > 127) // GBK chinese character
+		if (c & 0x80) // head of multi-byte character
 		{
 			if (*p_display_len + 2 > max_display_len)
 			{
 				break;
 			}
-			i++;
+
+			c = (c & 0x70) << 1;
+			while (c & 0x80)
+			{
+				i++;
+				c = (c & 0x7f) << 1;
+			}
+
 			(*p_display_len) += 2;
 		}
 		else
@@ -105,7 +162,7 @@ long split_data_lines(const char *p_buf,
 		p_line_offsets[line_cnt + 1] = p_line_offsets[line_cnt] + len;
 		line_cnt++;
 		p_buf += len;
-	} while (p_buf[0] != '\0');
+	} while (len > 0);
 
 	return line_cnt;
 }