gl/mbrtowc.c

   1 /* Convert multibyte character to wide character.
   2    Copyright (C) 1999-2002, 2005-2023 Free Software Foundation, Inc.
   3    Written by Bruno Haible <bruno@clisp.org>, 2008.
   4
   5    This file is free software: you can redistribute it and/or modify
   6    it under the terms of the GNU Lesser General Public License as
   7    published by the Free Software Foundation; either version 2.1 of the
   8    License, or (at your option) any later version.
   9
  10    This file 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 Lesser General Public License for more details.
  14
  15    You should have received a copy of the GNU Lesser 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 "attribute.h"
  54 # include "lc-charset-dispatch.h"
  55 # include "mbtowc-lock.h"
  56
  57 static_assert (sizeof (mbstate_t) >= 4);
  58 static char internal_state[4];
  59
  60 size_t
  61 mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
  62 {
  63 # define FITS_IN_CHAR_TYPE(wc)  ((wc) <= WCHAR_MAX)
  64 # include "mbrtowc-impl.h"
  65 }
  66
  67 #else
  68 /* Override the system's mbrtowc() function.  */
  69
  70 # if MBRTOWC_IN_C_LOCALE_MAYBE_EILSEQ
  71 #  include "hard-locale.h"
  72 #  include <locale.h>
  73 # endif
  74
  75 # undef mbrtowc
  76
  77 size_t
  78 rpl_mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
  79 {
  80   size_t ret;
  81   wchar_t wc;
  82
  83 # if MBRTOWC_NULL_ARG2_BUG || MBRTOWC_RETVAL_BUG || MBRTOWC_EMPTY_INPUT_BUG
  84   if (s == NULL)
  85     {
  86       pwc = NULL;
  87       s = "";
  88       n = 1;
  89     }
  90 # endif
  91
  92 # if MBRTOWC_EMPTY_INPUT_BUG
  93   if (n == 0)
  94     return (size_t) -2;
  95 # endif
  96
  97   if (! pwc)
  98     pwc = &wc;
  99
 100 # if MBRTOWC_RETVAL_BUG
 101   {
 102     static mbstate_t internal_state;
 103
 104     /* Override mbrtowc's internal state.  We cannot call mbsinit() on the
 105        hidden internal state, but we can call it on our variable.  */
 106     if (ps == NULL)
 107       ps = &internal_state;
 108
 109     if (!mbsinit (ps))
 110       {
 111         /* Parse the rest of the multibyte character byte for byte.  */
 112         size_t count = 0;
 113         for (; n > 0; s++, n--)
 114           {
 115             ret = mbrtowc (&wc, s, 1, ps);
 116
 117             if (ret == (size_t)(-1))
 118               return (size_t)(-1);
 119             count++;
 120             if (ret != (size_t)(-2))
 121               {
 122                 /* The multibyte character has been completed.  */
 123                 *pwc = wc;
 124                 return (wc == 0 ? 0 : count);
 125               }
 126           }
 127         return (size_t)(-2);
 128       }
 129   }
 130 # endif
 131
 132 # if MBRTOWC_STORES_INCOMPLETE_BUG
 133   ret = mbrtowc (&wc, s, n, ps);
 134   if (ret < (size_t) -2 && pwc != NULL)
 135     *pwc = wc;
 136 # else
 137   ret = mbrtowc (pwc, s, n, ps);
 138 # endif
 139
 140 # if MBRTOWC_NUL_RETVAL_BUG
 141   if (ret < (size_t) -2 && !*pwc)
 142     return 0;
 143 # endif
 144
 145 # if MBRTOWC_IN_C_LOCALE_MAYBE_EILSEQ
 146   if ((size_t) -2 <= ret && n != 0 && ! hard_locale (LC_CTYPE))
 147     {
 148       unsigned char uc = *s;
 149       *pwc = uc;
 150       return 1;
 151     }
 152 # endif
 153
 154   return ret;
 155 }
 156
 157 #endif