lib/mbrtoc32.c

   1 /* Convert multibyte character to 32-bit wide character.
   2    Copyright (C) 2020 Free Software Foundation, Inc.
   3
   4    This program is free software: you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation; either version 3 of the License, or
   7    (at your option) any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
  16
  17 /* Written by Bruno Haible <bruno@clisp.org>, 2020.  */
  18
  19 #include <config.h>
  20
  21 /* Specification.  */
  22 #include <uchar.h>
  23
  24 #include <errno.h>
  25 #include <stdlib.h>
  26
  27 # ifndef FALLTHROUGH
  28 #  if __GNUC__ < 7
  29 #   define FALLTHROUGH ((void) 0)
  30 #  else
  31 #   define FALLTHROUGH __attribute__ ((__fallthrough__))
  32 #  endif
  33 # endif
  34
  35 #if GNULIB_defined_mbstate_t /* AIX, IRIX */
  36 /* Implement mbrtoc32() on top of mbtowc() for the non-UTF-8 locales
  37    and directly for the UTF-8 locales.  */
  38
  39 # if defined _WIN32 && !defined __CYGWIN__
  40
  41 #  define WIN32_LEAN_AND_MEAN  /* avoid including junk */
  42 #  include <windows.h>
  43
  44 # elif HAVE_PTHREAD_API
  45
  46 #  include <pthread.h>
  47 #  if HAVE_THREADS_H && HAVE_WEAK_SYMBOLS
  48 #   include <threads.h>
  49 #   pragma weak thrd_exit
  50 #   define c11_threads_in_use() (thrd_exit != NULL)
  51 #  else
  52 #   define c11_threads_in_use() 0
  53 #  endif
  54
  55 # elif HAVE_THREADS_H
  56
  57 #  include <threads.h>
  58
  59 # endif
  60
  61 # include "verify.h"
  62 # include "lc-charset-dispatch.h"
  63 # include "mbtowc-lock.h"
  64
  65 verify (sizeof (mbstate_t) >= 4);
  66 static char internal_state[4];
  67
  68 size_t
  69 mbrtoc32 (char32_t *pwc, const char *s, size_t n, mbstate_t *ps)
  70 {
  71 # define FITS_IN_CHAR_TYPE(wc)  1
  72 # include "mbrtowc-impl.h"
  73 }
  74
  75 #else /* glibc, macOS, FreeBSD, NetBSD, OpenBSD, HP-UX, Solaris, Cygwin, mingw, MSVC, Minix, Android */
  76
  77 /* Implement mbrtoc32() based on mbrtowc().  */
  78
  79 # include <wchar.h>
  80
  81 # include "localcharset.h"
  82 # include "streq.h"
  83
  84 static mbstate_t internal_state;
  85
  86 size_t
  87 mbrtoc32 (char32_t *pwc, const char *s, size_t n, mbstate_t *ps)
  88 {
  89   /* It's simpler to handle the case s == NULL upfront, than to worry about
  90      this case later, before every test of pwc and n.  */
  91   if (s == NULL)
  92     {
  93       pwc = NULL;
  94       s = "";
  95       n = 1;
  96     }
  97
  98 # if MBRTOC32_EMPTY_INPUT_BUG || _GL_LARGE_CHAR32_T
  99   if (n == 0)
 100     return (size_t) -2;
 101 # endif
 102
 103   if (ps == NULL)
 104     ps = &internal_state;
 105
 106 # if _GL_LARGE_CHAR32_T
 107
 108   /* Special-case all encodings that may produce wide character values
 109      > WCHAR_MAX.  */
 110   const char *encoding = locale_charset ();
 111   if (STREQ_OPT (encoding, "UTF-8", 'U', 'T', 'F', '-', '8', 0, 0, 0, 0))
 112     {
 113       /* Special-case the UTF-8 encoding.  Assume that the wide-character
 114          encoding in a UTF-8 locale is UCS-2 or, equivalently, UTF-16.  */
 115       /* Here n > 0.  */
 116       char *pstate = (char *)ps;
 117       size_t nstate = pstate[0];
 118       char buf[4];
 119       const char *p;
 120       size_t m;
 121       int res;
 122
 123       switch (nstate)
 124         {
 125         case 0:
 126           p = s;
 127           m = n;
 128           break;
 129         case 3:
 130           buf[2] = pstate[3];
 131           FALLTHROUGH;
 132         case 2:
 133           buf[1] = pstate[2];
 134           FALLTHROUGH;
 135         case 1:
 136           buf[0] = pstate[1];
 137           p = buf;
 138           m = nstate;
 139           buf[m++] = s[0];
 140           if (n >= 2 && m < 4)
 141             {
 142               buf[m++] = s[1];
 143               if (n >= 3 && m < 4)
 144                 buf[m++] = s[2];
 145             }
 146           break;
 147         default:
 148           errno = EINVAL;
 149           return (size_t)(-1);
 150         }
 151
 152       /* Here m > 0.  */
 153
 154       {
 155 #  define FITS_IN_CHAR_TYPE(wc)  1
 156 #  include "mbrtowc-impl-utf8.h"
 157       }
 158
 159      success:
 160       if (nstate >= (res > 0 ? res : 1))
 161         abort ();
 162       res -= nstate;
 163       /* Set *ps to the initial state.  */
 164 #  if defined _WIN32 && !defined __CYGWIN__
 165       /* Native Windows.  */
 166       /* MSVC defines 'mbstate_t' as an 8-byte struct; the first 4 bytes matter.
 167          On mingw, 'mbstate_t' is sometimes defined as 'int', sometimes defined
 168          as an 8-byte struct, of which the first 4 bytes matter.  */
 169       *(unsigned int *)pstate = 0;
 170 #  elif defined __CYGWIN__
 171       /* Cygwin defines 'mbstate_t' as an 8-byte struct; the first 4 bytes
 172          matter.  */
 173       ps->__count = 0;
 174 #  else
 175       pstate[0] = 0;
 176 #  endif
 177       return res;
 178
 179      incomplete:
 180       {
 181         size_t k = nstate;
 182         /* Here 0 <= k < m < 4.  */
 183         pstate[++k] = s[0];
 184         if (k < m)
 185           {
 186             pstate[++k] = s[1];
 187             if (k < m)
 188               pstate[++k] = s[2];
 189           }
 190         if (k != m)
 191           abort ();
 192       }
 193       pstate[0] = m;
 194       return (size_t)(-2);
 195
 196      invalid:
 197       errno = EILSEQ;
 198       /* The conversion state is undefined, says POSIX.  */
 199       return (size_t)(-1);
 200     }
 201   else
 202     {
 203       wchar_t wc;
 204       size_t ret = mbrtowc (&wc, s, n, ps);
 205       if (ret < (size_t) -2 && pwc != NULL)
 206         *pwc = wc;
 207       return ret;
 208     }
 209
 210 # else
 211
 212   /* char32_t and wchar_t are equivalent.
 213      Two implementations are possible:
 214        - We can call the original mbrtoc32 (if it exists) and handle
 215          MBRTOC32_IN_C_LOCALE_MAYBE_EILSEQ.
 216        - We can call mbrtowc.
 217      The latter is simpler.   */
 218   wchar_t wc;
 219   size_t ret = mbrtowc (&wc, s, n, ps);
 220   if (ret < (size_t) -2 && pwc != NULL)
 221     *pwc = wc;
 222   return ret;
 223
 224 # endif
 225 }
 226
 227 #endif