2 * Unicode utility routines
4 * Wireshark - Network traffic analyzer
5 * By Gerald Combs <gerald@wireshark.org>
6 * Copyright 2006 Gerald Combs
8 * SPDX-License-Identifier: GPL-2.0-or-later
11 #include "unicode-utils.h"
14 ws_utf8_char_len(guint8 ch)
16 if (ch >= 0xfe) return -1;
17 if (ch >= 0xfc) return 6;
18 if (ch >= 0xf8) return 5;
19 if (ch >= 0xf0) return 4;
20 if (ch >= 0xe0) return 3;
21 if (ch >= 0xc0) return 2;
32 * Unicode utilities (internal interface)
34 * We define UNICODE and _UNICODE under Windows. This means that
35 * Windows SDK routines expect UTF-16 strings, in contrast to newer
36 * versions of Glib and GTK+ which expect UTF-8. This module provides
37 * convenience routines for converting between UTF-8 and UTF-16.
40 #define INITIAL_UTFBUF_SIZE 128
43 * XXX - Should we use g_utf8_to_utf16() and g_utf16_to_utf8()
44 * instead? The goal of the functions below was to provide simple
45 * wrappers for UTF-8 <-> UTF-16 conversion without making the
46 * caller worry about freeing up memory afterward.
49 /* Convert from UTF-8 to UTF-16. */
51 utf_8to16(const char *utf8str)
53 static wchar_t *utf16buf[3];
54 static int utf16buf_len[3];
63 * Allocate the buffer if it's not already allocated.
65 if (utf16buf[idx] == NULL) {
66 utf16buf_len[idx] = INITIAL_UTFBUF_SIZE;
67 utf16buf[idx] = g_malloc(utf16buf_len[idx] * sizeof(wchar_t));
70 while (MultiByteToWideChar(CP_UTF8, 0, utf8str,
71 -1, NULL, 0) >= utf16buf_len[idx]) {
73 * Double the buffer's size if it's not big enough.
74 * The size of the buffer starts at 128, so doubling its size
75 * adds at least another 128 bytes, which is more than enough
76 * for one more character plus a terminating '\0'.
78 utf16buf_len[idx] *= 2;
79 utf16buf[idx] = g_realloc(utf16buf[idx], utf16buf_len[idx] * sizeof(wchar_t));
82 if (MultiByteToWideChar(CP_UTF8, 0, utf8str,
83 -1, utf16buf[idx], utf16buf_len[idx]) == 0)
90 utf_8to16_snprintf(TCHAR *utf16buf, gint utf16buf_len, const gchar* fmt, ...)
96 dst = g_strdup_vprintf(fmt, ap);
99 StringCchPrintf(utf16buf, utf16buf_len, _T("%s"), utf_8to16(dst));
104 /* Convert from UTF-16 to UTF-8. */
106 utf_16to8(const wchar_t *utf16str)
108 static gchar *utf8buf[3];
109 static int utf8buf_len[3];
112 if (utf16str == NULL)
118 * Allocate the buffer if it's not already allocated.
120 if (utf8buf[idx] == NULL) {
121 utf8buf_len[idx] = INITIAL_UTFBUF_SIZE;
122 utf8buf[idx] = g_malloc(utf8buf_len[idx]);
125 while (WideCharToMultiByte(CP_UTF8, 0, utf16str, -1,
126 NULL, 0, NULL, NULL) >= utf8buf_len[idx]) {
128 * Double the buffer's size if it's not big enough.
129 * The size of the buffer starts at 128, so doubling its size
130 * adds at least another 128 bytes, which is more than enough
131 * for one more character plus a terminating '\0'.
133 utf8buf_len[idx] *= 2;
134 utf8buf[idx] = g_realloc(utf8buf[idx], utf8buf_len[idx]);
137 if (WideCharToMultiByte(CP_UTF8, 0, utf16str, -1,
138 utf8buf[idx], utf8buf_len[idx], NULL, NULL) == 0)
144 /* Convert our argument list from UTF-16 to UTF-8. */
146 arg_list_utf_16to8(int argc, char *argv[]) {
150 /* Convert our arg list to UTF-8. */
151 wc_argv = CommandLineToArgvW(GetCommandLineW(), &wc_argc);
152 if (wc_argv && wc_argc == argc) {
153 for (i = 0; i < argc; i++) {
154 argv[i] = g_utf16_to_utf8(wc_argv[i], -1, NULL, NULL, NULL);
156 } /* XXX else bail because something is horribly, horribly wrong? */
163 * Editor modelines - http://www.wireshark.org/tools/modelines.html
168 * indent-tabs-mode: nil
171 * ex: set shiftwidth=2 tabstop=8 expandtab:
172 * :indentSize=2:tabSize=8:noTabs=true: