1 /* Convert multibyte character to wide character.
2 Copyright (C) 1999-2002, 2005-2020 Free Software Foundation, Inc.
3 Written by Bruno Haible <bruno@clisp.org>, 2008.
4
5 This program is free software: you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published by
7 the Free Software Foundation; either version 3 of the License, or
8 (at your option) any later version.
9
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 GNU General Public License for more details.
14
15 You should have received a copy of the GNU General Public License
16 along with this program. If not, see <https://www.gnu.org/licenses/>. */
17
18 #include <config.h>
19
20 /* Specification. */
21 #include <wchar.h>
22
23 #if GNULIB_defined_mbstate_t
24 /* Implement mbrtowc() on top of mbtowc() for the non-UTF-8 locales
25 and directly for the UTF-8 locales. */
26
27 # include <errno.h>
28 # include <stdint.h>
29 # include <stdlib.h>
30
31 # if defined _WIN32 && !defined __CYGWIN__
32
33 # define WIN32_LEAN_AND_MEAN /* avoid including junk */
34 # include <windows.h>
35
36 # elif HAVE_PTHREAD_API
37
38 # include <pthread.h>
39 # if HAVE_THREADS_H && HAVE_WEAK_SYMBOLS
40 # include <threads.h>
41 # pragma weak thrd_exit
42 # define c11_threads_in_use() (thrd_exit != NULL)
43 # else
44 # define c11_threads_in_use() 0
45 # endif
46
47 # elif HAVE_THREADS_H
48
49 # include <threads.h>
50
51 # endif
52
53 # include "verify.h"
54 # include "lc-charset-dispatch.h"
55 # include "mbtowc-lock.h"
56
57 # ifndef FALLTHROUGH
58 # if __GNUC__ < 7
59 # define FALLTHROUGH ((void) 0)
60 # else
61 # define FALLTHROUGH __attribute__ ((__fallthrough__))
62 # endif
63 # endif
64
65 verify (sizeof (mbstate_t) >= 4);
66 static char internal_state[4];
67
68 size_t
mbrtowc(wchar_t * pwc,const char * s,size_t n,mbstate_t * ps)69 mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
70 {
71 # define FITS_IN_CHAR_TYPE(wc) ((wc) <= WCHAR_MAX)
72 # include "mbrtowc-impl.h"
73 }
74
75 #else
76 /* Override the system's mbrtowc() function. */
77
78 # if MBRTOWC_IN_C_LOCALE_MAYBE_EILSEQ
79 # include "hard-locale.h"
80 # include <locale.h>
81 # endif
82
83 # undef mbrtowc
84
85 size_t
rpl_mbrtowc(wchar_t * pwc,const char * s,size_t n,mbstate_t * ps)86 rpl_mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
87 {
88 size_t ret;
89 wchar_t wc;
90
91 # if MBRTOWC_NULL_ARG2_BUG || MBRTOWC_RETVAL_BUG || MBRTOWC_EMPTY_INPUT_BUG
92 if (s == NULL)
93 {
94 pwc = NULL;
95 s = "";
96 n = 1;
97 }
98 # endif
99
100 # if MBRTOWC_EMPTY_INPUT_BUG
101 if (n == 0)
102 return (size_t) -2;
103 # endif
104
105 if (! pwc)
106 pwc = &wc;
107
108 # if MBRTOWC_RETVAL_BUG
109 {
110 static mbstate_t internal_state;
111
112 /* Override mbrtowc's internal state. We cannot call mbsinit() on the
113 hidden internal state, but we can call it on our variable. */
114 if (ps == NULL)
115 ps = &internal_state;
116
117 if (!mbsinit (ps))
118 {
119 /* Parse the rest of the multibyte character byte for byte. */
120 size_t count = 0;
121 for (; n > 0; s++, n--)
122 {
123 ret = mbrtowc (&wc, s, 1, ps);
124
125 if (ret == (size_t)(-1))
126 return (size_t)(-1);
127 count++;
128 if (ret != (size_t)(-2))
129 {
130 /* The multibyte character has been completed. */
131 *pwc = wc;
132 return (wc == 0 ? 0 : count);
133 }
134 }
135 return (size_t)(-2);
136 }
137 }
138 # endif
139
140 # if MBRTOWC_STORES_INCOMPLETE_BUG
141 ret = mbrtowc (&wc, s, n, ps);
142 if (ret < (size_t) -2 && pwc != NULL)
143 *pwc = wc;
144 # else
145 ret = mbrtowc (pwc, s, n, ps);
146 # endif
147
148 # if MBRTOWC_NUL_RETVAL_BUG
149 if (ret < (size_t) -2 && !*pwc)
150 return 0;
151 # endif
152
153 # if MBRTOWC_IN_C_LOCALE_MAYBE_EILSEQ
154 if ((size_t) -2 <= ret && n != 0 && ! hard_locale (LC_CTYPE))
155 {
156 unsigned char uc = *s;
157 *pwc = uc;
158 return 1;
159 }
160 # endif
161
162 return ret;
163 }
164
165 #endif
166