trunk/Src/wptUTF8.cpp

/* wptUTF8.cpp - UTF8 conversation
 *      Copyright (C) 2002, 2004, 2005, 2006, 2009 Timo Schulz
 *
 * This file is part of WinPT.
 *
 * WinPT 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 2 of the License, or
 * (at your option) any later version.
 * 
 * WinPT is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
 * GNU General Public License for more details.
 */
#ifdef HAVE_CONFIG_H
#include <config.h>
#endif

#include <windows.h>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <ctype.h>

#include "wptTypes.h"
#include "wptErrors.h"


/**
 * Converts the given intput string, which is encoded with the locale 
 * setting, into UTF-8 representation.
 */
char*
native_to_utf8 (const char *string)
{
    wchar_t *result;
    char *native;
    int n;

    n = MultiByteToWideChar (GetACP (), 0, string, -1, NULL, 0);
    if (n < 0)
        return NULL;

    result = new wchar_t[n+1];
    if (!result)
        BUG (0);

    n = MultiByteToWideChar (GetACP (), 0, string, -1, result, n);
    if (n < 0) {
        free_if_alloc (result);
        return NULL;
    }

    n = WideCharToMultiByte (CP_UTF8, 0, result, -1, NULL, 0, NULL, NULL);
    if (n < 0)
        return NULL;

    native = new char[n+1];
    if (!native)
        BUG (0);

    n = WideCharToMultiByte (CP_UTF8, 0, result, -1, native, n, NULL, NULL);
    if (n < 0) {
        free_if_alloc (result);
        return NULL;
    }

    free_if_alloc (result);
    //native[n] = '\0';        
    return native;
}


/**
 * Converts the given string, which is encoded in UTF-8,
 * into the locale setting.
 */
char*
utf8_to_native (const char *string)
{
    wchar_t *result;
    char *native;    
    int n;

    n = MultiByteToWideChar (CP_UTF8, 0, string, -1, NULL, 0);
    if (n < 0)
        return NULL;

    result = new wchar_t[n+1];
    if (!result)
        BUG (0);

    n = MultiByteToWideChar (CP_UTF8, 0, string, -1, result, n);
    if (n < 0) {
        free_if_alloc (result);
        return NULL;
    }

    n = WideCharToMultiByte (GetACP (), 0, result, -1, NULL, 0, NULL, NULL);
    if (n < 0)
        return NULL;

    native = new char[n+1];
    if (!native)
        BUG (0);

    n = WideCharToMultiByte (GetACP (), 0, result, -1, native, n, NULL, NULL);
    if (n < 0) {
        free_if_alloc (result);
        return NULL;
    }
    
    free_if_alloc (result);
    //native[n] = '\0';
    return native;
}


/**
 * Returns -1 if the given string contains any 8-bit characters.
 * This is a helper to decide when to use UTF8 encoding.
 */
int
is_8bit_string (const char *str)
{
    for (size_t i = 0; i < strlen (str); i++) {
        if (str[i] & 0x80)
            return -1;
    }
    return 0;
}
1	twoaday	193	/* wptUTF8.cpp - UTF8 conversation
2	twoaday	328	* Copyright (C) 2002, 2004, 2005, 2006, 2009 Timo Schulz
3	twoaday	2	*
4	werner	36	* This file is part of WinPT.
5			*
6			* WinPT is free software; you can redistribute it and/or modify
7			* it under the terms of the GNU General Public License as published by
8			* the Free Software Foundation; either version 2 of the License, or
9			* (at your option) any later version.
10			*
11			* WinPT is distributed in the hope that it will be useful,
12			* but WITHOUT ANY WARRANTY; without even the implied warranty of
13			* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14			* GNU General Public License for more details.
15	twoaday	2	*/
16	werner	36	#ifdef HAVE_CONFIG_H
17			#include <config.h>
18			#endif
19
20	twoaday	2	#include <windows.h>
21	werner	36	#include <stdlib.h>
22	twoaday	2	#include <stdio.h>
23			#include <string.h>
24			#include <ctype.h>
25	werner	36
26			#include "wptTypes.h"
27	twoaday	2	#include "wptErrors.h"
28
29
30	twoaday	328	/**
31			* Converts the given intput string, which is encoded with the locale
32			* setting, into UTF-8 representation.
33			*/
34	twoaday	187	char*
35			native_to_utf8 (const char *string)
36	werner	36	{
37	twoaday	187	wchar_t *result;
38			char *native;
39			int n;
40	werner	36
41	twoaday	187	n = MultiByteToWideChar (GetACP (), 0, string, -1, NULL, 0);
42			if (n < 0)
43			return NULL;
44
45	twoaday	201	result = new wchar_t[n+1];
46	twoaday	187	if (!result)
47			BUG (0);
48
49			n = MultiByteToWideChar (GetACP (), 0, string, -1, result, n);
50			if (n < 0) {
51	twoaday	201	free_if_alloc (result);
52	twoaday	187	return NULL;
53	werner	36	}
54	twoaday	187
55			n = WideCharToMultiByte (CP_UTF8, 0, result, -1, NULL, 0, NULL, NULL);
56			if (n < 0)
57			return NULL;
58
59	twoaday	201	native = new char[n+1];
60	twoaday	187	if (!native)
61			BUG (0);
62
63			n = WideCharToMultiByte (CP_UTF8, 0, result, -1, native, n, NULL, NULL);
64			if (n < 0) {
65	twoaday	201	free_if_alloc (result);
66	twoaday	187	return NULL;
67			}
68
69	twoaday	201	free_if_alloc (result);
70	twoaday	328	//native[n] = '\0';
71	twoaday	187	return native;
72	twoaday	128	}
73	werner	36
74	twoaday	185
75	twoaday	328	/**
76			* Converts the given string, which is encoded in UTF-8,
77			* into the locale setting.
78			*/
79	twoaday	187	char*
80	twoaday	185	utf8_to_native (const char *string)
81	twoaday	2	{
82	twoaday	185	wchar_t *result;
83			char *native;
84			int n;
85	twoaday	2
86	twoaday	185	n = MultiByteToWideChar (CP_UTF8, 0, string, -1, NULL, 0);
87			if (n < 0)
88			return NULL;
89	twoaday	2
90	twoaday	328	result = new wchar_t[n+1];
91	twoaday	185	if (!result)
92			BUG (0);
93	twoaday	2
94	twoaday	185	n = MultiByteToWideChar (CP_UTF8, 0, string, -1, result, n);
95			if (n < 0) {
96	twoaday	328	free_if_alloc (result);
97	twoaday	185	return NULL;
98			}
99	twoaday	128
100	twoaday	185	n = WideCharToMultiByte (GetACP (), 0, result, -1, NULL, 0, NULL, NULL);
101			if (n < 0)
102			return NULL;
103	twoaday	2
104	twoaday	328	native = new char[n+1];
105	twoaday	185	if (!native)
106			BUG (0);
107	twoaday	2
108	twoaday	187	n = WideCharToMultiByte (GetACP (), 0, result, -1, native, n, NULL, NULL);
109	twoaday	185	if (n < 0) {
110	twoaday	328	free_if_alloc (result);
111	twoaday	185	return NULL;
112			}
113	twoaday	328
114			free_if_alloc (result);
115			//native[n] = '\0';
116	twoaday	185	return native;
117	twoaday	2	}
118	werner	36
119
120	twoaday	328	/**
121			* Returns -1 if the given string contains any 8-bit characters.
122			* This is a helper to decide when to use UTF8 encoding.
123			*/
124	werner	36	int
125	twoaday	193	is_8bit_string (const char *str)
126	werner	36	{
127	twoaday	328	for (size_t i = 0; i < strlen (str); i++) {
128	werner	36	if (str[i] & 0x80)
129			return -1;
130			}
131			return 0;
132	twoaday	128	}