root/maint/gnulib/lib/c32is-impl.h

/* */

INCLUDED FROM

DEFINITIONS

1. FUNC

   1 /* Test whether a 32-bit wide character belongs to a specific character class.
   2    Copyright (C) 2020-2021 Free Software Foundation, Inc.
   3 
   4    This file is free software.
   5    It is dual-licensed under "the GNU LGPLv3+ or the GNU GPLv2+".
   6    You can redistribute it and/or modify it under either
   7      - the terms of the GNU Lesser General Public License as published
   8        by the Free Software Foundation; either version 3, or (at your
   9        option) any later version, or
  10      - the terms of the GNU General Public License as published by the
  11        Free Software Foundation; either version 2, or (at your option)
  12        any later version, or
  13      - the same dual license "the GNU LGPLv3+ or the GNU GPLv2+".
  14 
  15    This file is distributed in the hope that it will be useful,
  16    but WITHOUT ANY WARRANTY; without even the implied warranty of
  17    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  18    Lesser General Public License and the GNU General Public License
  19    for more details.
  20 
  21    You should have received a copy of the GNU Lesser General Public
  22    License and of the GNU General Public License along with this
  23    program.  If not, see <https://www.gnu.org/licenses/>.  */
  24 
  25 /* Written by Bruno Haible <bruno@clisp.org>, 2020.  */
  26 
  27 #include <wchar.h>
  28 #include <wctype.h>
  29 
  30 #ifdef __CYGWIN__
  31 # include <cygwin/version.h>
  32 #endif
  33 
  34 #if GNULIB_defined_mbstate_t
  35 # include "localcharset.h"
  36 # include "streq.h"
  37 #endif
  38 
  39 #include "unictype.h"
  40 #include "verify.h"
  41 
  42 int
  43 FUNC (wint_t wc)
     /*  */
  44 {
  45   /* The char32_t encoding of a multibyte character is defined by the way
  46      mbrtoc32() is defined.  */
  47 
  48 #if GNULIB_defined_mbstate_t            /* AIX, IRIX */
  49   /* mbrtoc32() is defined on top of mbtowc() for the non-UTF-8 locales
  50      and directly for the UTF-8 locales.  */
  51   if (wc != WEOF)
  52     {
  53       const char *encoding = locale_charset ();
  54       if (STREQ_OPT (encoding, "UTF-8", 'U', 'T', 'F', '-', '8', 0, 0, 0, 0))
  55         return UCS_FUNC (wc);
  56       else
  57         return WCHAR_FUNC (wc);
  58     }
  59   else
  60     return 0;
  61 
  62 #elif HAVE_WORKING_MBRTOC32             /* glibc */
  63   /* mbrtoc32() is essentially defined by the system libc.  */
  64 
  65 # if defined __GLIBC__
  66   /* The char32_t encoding of a multibyte character is known to be the same as
  67      the wchar_t encoding.  */
  68   return WCHAR_FUNC (wc);
  69 # else
  70   /* The char32_t encoding of a multibyte character is known to be UCS-4,
  71      different from the the wchar_t encoding.  */
  72   if (wc != WEOF)
  73     return UCS_FUNC (wc);
  74   else
  75     return 0;
  76 # endif
  77 
  78 #elif _GL_LARGE_CHAR32_T                /* Cygwin, mingw, MSVC */
  79   /* The wchar_t encoding is UTF-16.
  80      The char32_t encoding is UCS-4.  */
  81 
  82 # if defined __CYGWIN__ && CYGWIN_VERSION_DLL_MAJOR >= 1007
  83   /* As an extension to POSIX, the iswalnum() function of Cygwin >= 1.7
  84      supports also wc arguments outside the Unicode BMP, that is, outside
  85      the 'wchar_t' range.  See
  86      <https://lists.gnu.org/archive/html/bug-gnulib/2011-02/msg00019.html>
  87      = <https://cygwin.com/ml/cygwin/2011-02/msg00044.html>.  */
  88   return WCHAR_FUNC (wc);
  89 # else
  90   if (wc == WEOF || wc == (wchar_t) wc)
  91     /* wc is in the range for the isw* functions.  */
  92     return WCHAR_FUNC (wc);
  93   else
  94     return UCS_FUNC (wc);
  95 # endif
  96 
  97 #else /* macOS, FreeBSD, NetBSD, OpenBSD, HP-UX, Solaris, Minix, Android */
  98   /* char32_t and wchar_t are equivalent.  */
  99   verify (sizeof (char32_t) == sizeof (wchar_t));
 100 
 101   return WCHAR_FUNC (wc);
 102 #endif
 103 }

/* */