update gnulib to 47bf2cf3184027c1eb9c1dfeea5c5b8b2d69710d

* lib/dosname.h, lib/glthread/lock.c, lib/glthread/lock.h,
lib/glthread/threadlib.c, lib/windows-mutex.c, lib/windows-mutex.h,
lib/windows-once.c, lib/windows-once.h, lib/windows-recmutex.c,
lib/windows-recmutex.h, lib/windows-rwlock.c, lib/windows-rwlock.h,
m4/host-cpu-c-abi.m4, m4/lib-ld.m4, m4/lib-link.m4, m4/lib-prefix.m4,
m4/lock.m4, m4/longlong.m4, m4/pthread_rwlock_rdlock.m4,
tools/config.rpath: Delete.
* lib/alloca.in.h, lib/argmatch.c, lib/argmatch.h, lib/arg-nonnull.h,
lib/argp-ba.c, lib/argp-eexst.c, lib/argp-fmtstream.c,
lib/argp-fmtstream.h, lib/argp-fs-xinl.c, lib/argp.h, lib/argp-help.c,
lib/argp-namefrob.h, lib/argp-parse.c, lib/argp-pin.c, lib/argp-pv.c,
lib/argp-pvh.c, lib/argp-xinl.c, lib/asnprintf.c, lib/basename-lgpl.c,
lib/c-ctype.h, lib/c++defs.h, lib/cdefs.h, lib/closeout.c,
lib/closeout.h, lib/close-stream.c, lib/c-strcasecmp.c,
lib/c-strcaseeq.h, lib/c-strcase.h, lib/c-strncasecmp.c,
lib/dirname.h, lib/dirname-lgpl.c, lib/errno.in.h, lib/error.c,
lib/error.h, lib/exitfail.c, lib/exitfail.h, lib/fcntl.in.h,
lib/filename.h, lib/float.c, lib/float+.h, lib/float.in.h,
lib/fpending.c, lib/fpending.h, lib/getopt1.c, lib/getopt.c,
lib/getopt-cdefs.in.h, lib/getopt-core.h, lib/getopt-ext.h,
lib/getopt.in.h, lib/getopt_int.h, lib/getopt-pfx-core.h,
lib/getopt-pfx-ext.h, lib/getprogname.c, lib/getprogname.h,
lib/gettext.h, lib/gettimeofday.c, lib/hard-locale.c,
lib/hard-locale.h, lib/intprops.h, lib/isatty.c, lib/itold.c,
lib/libc-config.h, lib/limits.in.h, lib/localcharset.c,
lib/localcharset.h, lib/localtime-buffer.c, lib/localtime-buffer.h,
lib/lstat.c, lib/Makefile.am, lib/malloca.c, lib/malloca.h,
lib/malloc.c, lib/mbrtowc.c, lib/mbsinit.c, lib/memchr.c,
lib/memchr.valgrind, lib/mempcpy.c, lib/minmax.h, lib/mkdir.c,
lib/mkstemp.c, lib/mkstemps.c, lib/msvc-inval.c, lib/msvc-inval.h,
lib/msvc-nothrow.c, lib/msvc-nothrow.h, lib/_Noreturn.h,
lib/pathmax.h, lib/printf-args.c, lib/printf-args.h,
lib/printf-parse.c, lib/printf-parse.h, lib/progname.c,
lib/progname.h, lib/quotearg.c, lib/quotearg.h, lib/quote.h,
lib/rawmemchr.c, lib/rawmemchr.valgrind, lib/secure_getenv.c,
lib/size_max.h, lib/sleep.c, lib/stat.c, lib/stat-time.h,
lib/stat-w32.c, lib/stat-w32.h, lib/stdalign.in.h, lib/stdbool.in.h,
lib/stddef.in.h, lib/stdint.in.h, lib/stdio-impl.h, lib/stdio.in.h,
lib/stdlib.in.h, lib/stpcpy.c, lib/strcasecmp.c, lib/strchrnul.c,
lib/strchrnul.valgrind, lib/streq.h, lib/strerror.c,
lib/strerror-override.c, lib/strerror-override.h, lib/string.in.h,
lib/strings.in.h, lib/stripslash.c, lib/strncasecmp.c, lib/strndup.c,
lib/strnlen.c, lib/strverscmp.c, lib/sysexits.in.h, lib/sys_stat.in.h,
lib/sys_time.in.h, lib/sys_types.in.h, lib/sys_wait.in.h,
lib/tempname.c, lib/tempname.h, lib/time.in.h, lib/unistd.in.h,
lib/vasnprintf.c, lib/vasnprintf.h, lib/verify.h, lib/vsnprintf.c,
lib/warn-on-use.h, lib/wchar.in.h, lib/wctype.in.h,
lib/windows-initguard.h, lib/xalloc-die.c, lib/xalloc.h,
lib/xalloc-oversized.h, lib/xmalloc.c, lib/xsize.h, m4/00gnulib.m4,
m4/absolute-header.m4, m4/alloca.m4, m4/argp.m4, m4/codeset.m4,
m4/dirname.m4, m4/double-slash-root.m4, m4/eealloc.m4, m4/errno_h.m4,
m4/error.m4, m4/exponentd.m4, m4/extensions.m4, m4/extern-inline.m4,
m4/fcntl_h.m4, m4/fcntl-o.m4, m4/float_h.m4, m4/fpending.m4,
m4/getopt.m4, m4/getprogname.m4, m4/gettimeofday.m4,
m4/gnulib-cache.m4, m4/gnulib-common.m4, m4/gnulib-comp.m4,
m4/gnulib-tool.m4, m4/include_next.m4, m4/__inline.m4, m4/intmax_t.m4,
m4/inttypes_h.m4, m4/isatty.m4, m4/largefile.m4, m4/limits-h.m4,
m4/localcharset.m4, m4/locale-fr.m4, m4/locale-ja.m4, m4/locale-zh.m4,
m4/localtime-buffer.m4, m4/lstat.m4, m4/malloca.m4, m4/malloc.m4,
m4/math_h.m4, m4/mbrtowc.m4, m4/mbsinit.m4, m4/mbstate_t.m4,
m4/memchr.m4, m4/mempcpy.m4, m4/minmax.m4, m4/mkdir.m4, m4/mkstemp.m4,
m4/mkstemps.m4, m4/mmap-anon.m4, m4/msvc-inval.m4, m4/msvc-nothrow.m4,
m4/multiarch.m4, m4/nocrash.m4, m4/off_t.m4, m4/pathmax.m4,
m4/printf.m4, m4/quotearg.m4, m4/quote.m4, m4/rawmemchr.m4,
m4/secure_getenv.m4, m4/size_max.m4, m4/sleep.m4, m4/ssize_t.m4,
m4/stat.m4, m4/stat-time.m4, m4/stdalign.m4, m4/stdbool.m4,
m4/stddef_h.m4, m4/std-gnu11.m4, m4/stdint_h.m4, m4/stdint.m4,
m4/stdio_h.m4, m4/stdlib_h.m4, m4/stpcpy.m4, m4/strcase.m4,
m4/strchrnul.m4, m4/strerror.m4, m4/string_h.m4, m4/strings_h.m4,
m4/strndup.m4, m4/strnlen.m4, m4/strverscmp.m4, m4/sysexits.m4,
m4/sys_socket_h.m4, m4/sys_stat_h.m4, m4/sys_time_h.m4,
m4/sys_types_h.m4, m4/sys_wait_h.m4, m4/tempname.m4, m4/threadlib.m4,
m4/time_h.m4, m4/unistd_h.m4, m4/vasnprintf.m4, m4/vsnprintf.m4,
m4/warn-on-use.m4, m4/wchar_h.m4, m4/wchar_t.m4, m4/wctype_h.m4,
m4/wint_t.m4, m4/xalloc.m4, m4/xsize.m4: Update.
* lib/inttypes.in.h, lib/lc-charset-dispatch.c,
lib/lc-charset-dispatch.h, lib/locale.in.h, lib/mbrtowc-impl.h,
lib/mbrtowc-impl-utf8.h, lib/mbtowc-lock.c, lib/mbtowc-lock.h,
lib/setlocale-lock.c, lib/setlocale_null.c, lib/setlocale_null.h,
m4/inttypes.m4, m4/locale_h.m4, m4/setlocale_null.m4,
m4/visibility.m4, m4/zzgnulib.m4: New files.
This commit is contained in:
Alexandre Duret-Lutz 2020-04-16 10:47:14 +02:00
parent fe642bc9ad
commit 1e86463205
287 changed files with 6133 additions and 6812 deletions

View file

@ -1,5 +1,5 @@
/* Convert multibyte character to wide character.
Copyright (C) 1999-2002, 2005-2019 Free Software Foundation, Inc.
Copyright (C) 1999-2002, 2005-2020 Free Software Foundation, Inc.
Written by Bruno Haible <bruno@clisp.org>, 2008.
This program is free software: you can redistribute it and/or modify
@ -20,21 +20,39 @@
/* Specification. */
#include <wchar.h>
#if C_LOCALE_MAYBE_EILSEQ
# include "hard-locale.h"
# include <locale.h>
#endif
#if GNULIB_defined_mbstate_t
/* Implement mbrtowc() on top of mbtowc(). */
/* Implement mbrtowc() on top of mbtowc() for the non-UTF-8 locales
and directly for the UTF-8 locales. */
# include <errno.h>
# include <stdint.h>
# include <stdlib.h>
# include "localcharset.h"
# include "streq.h"
# if defined _WIN32 && !defined __CYGWIN__
# define WIN32_LEAN_AND_MEAN /* avoid including junk */
# include <windows.h>
# elif HAVE_PTHREAD_API
# include <pthread.h>
# if HAVE_THREADS_H && HAVE_WEAK_SYMBOLS
# include <threads.h>
# pragma weak thrd_exit
# define c11_threads_in_use() (thrd_exit != NULL)
# else
# define c11_threads_in_use() 0
# endif
# elif HAVE_THREADS_H
# include <threads.h>
# endif
# include "verify.h"
# include "glthread/lock.h"
# include "lc-charset-dispatch.h"
# include "mbtowc-lock.h"
# ifndef FALLTHROUGH
# if __GNUC__ < 7
@ -44,417 +62,24 @@
# endif
# endif
/* Returns a classification of special values of the encoding of the current
locale. */
typedef enum {
enc_other, /* other */
enc_utf8, /* UTF-8 */
enc_eucjp, /* EUC-JP */
enc_94, /* EUC-KR, GB2312, BIG5 */
enc_euctw, /* EUC-TW */
enc_gb18030, /* GB18030 */
enc_sjis /* SJIS */
} enc_t;
static inline enc_t
locale_enc (void)
{
const char *encoding = locale_charset ();
if (STREQ_OPT (encoding, "UTF-8", 'U', 'T', 'F', '-', '8', 0, 0, 0, 0))
return enc_utf8;
if (STREQ_OPT (encoding, "EUC-JP", 'E', 'U', 'C', '-', 'J', 'P', 0, 0, 0))
return enc_eucjp;
if (STREQ_OPT (encoding, "EUC-KR", 'E', 'U', 'C', '-', 'K', 'R', 0, 0, 0)
|| STREQ_OPT (encoding, "GB2312", 'G', 'B', '2', '3', '1', '2', 0, 0, 0)
|| STREQ_OPT (encoding, "BIG5", 'B', 'I', 'G', '5', 0, 0, 0, 0, 0))
return enc_94;
if (STREQ_OPT (encoding, "EUC-TW", 'E', 'U', 'C', '-', 'T', 'W', 0, 0, 0))
return enc_euctw;
if (STREQ_OPT (encoding, "GB18030", 'G', 'B', '1', '8', '0', '3', '0', 0, 0))
return enc_gb18030;
if (STREQ_OPT (encoding, "SJIS", 'S', 'J', 'I', 'S', 0, 0, 0, 0, 0))
return enc_sjis;
return enc_other;
}
# if GNULIB_WCHAR_SINGLE
/* When we know that the locale does not change, provide a speedup by
caching the value of locale_enc. */
static int cached_locale_enc = -1;
static inline enc_t
locale_enc_cached (void)
{
if (cached_locale_enc < 0)
cached_locale_enc = locale_enc ();
return cached_locale_enc;
}
# else
/* By default, don't make assumptions, hence no caching. */
# define locale_enc_cached locale_enc
# endif
/* This lock protects the internal state of mbtowc against multiple simultaneous
calls of mbrtowc. */
gl_lock_define_initialized(static, mbtowc_lock)
verify (sizeof (mbstate_t) >= 4);
static char internal_state[4];
size_t
mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
{
char *pstate = (char *)ps;
if (s == NULL)
{
pwc = NULL;
s = "";
n = 1;
}
if (n == 0)
return (size_t)(-2);
/* Here n > 0. */
if (pstate == NULL)
pstate = internal_state;
{
size_t nstate = pstate[0];
char buf[4];
const char *p;
size_t m;
enc_t enc;
int res;
switch (nstate)
{
case 0:
p = s;
m = n;
break;
case 3:
buf[2] = pstate[3];
FALLTHROUGH;
case 2:
buf[1] = pstate[2];
FALLTHROUGH;
case 1:
buf[0] = pstate[1];
p = buf;
m = nstate;
buf[m++] = s[0];
if (n >= 2 && m < 4)
{
buf[m++] = s[1];
if (n >= 3 && m < 4)
buf[m++] = s[2];
}
break;
default:
errno = EINVAL;
return (size_t)(-1);
}
/* Here m > 0. */
enc = locale_enc_cached ();
if (enc == enc_utf8) /* UTF-8 */
{
/* Achieve multi-thread safety by not calling mbtowc() at all. */
/* Cf. unistr/u8-mbtouc.c. */
unsigned char c = (unsigned char) p[0];
if (c < 0x80)
{
if (pwc != NULL)
*pwc = c;
res = (c == 0 ? 0 : 1);
goto success;
}
if (c >= 0xc2)
{
if (c < 0xe0)
{
if (m == 1)
goto incomplete;
else /* m >= 2 */
{
unsigned char c2 = (unsigned char) p[1];
if ((c2 ^ 0x80) < 0x40)
{
if (pwc != NULL)
*pwc = ((unsigned int) (c & 0x1f) << 6)
| (unsigned int) (c2 ^ 0x80);
res = 2;
goto success;
}
}
}
else if (c < 0xf0)
{
if (m == 1)
goto incomplete;
else
{
unsigned char c2 = (unsigned char) p[1];
if ((c2 ^ 0x80) < 0x40
&& (c >= 0xe1 || c2 >= 0xa0)
&& (c != 0xed || c2 < 0xa0))
{
if (m == 2)
goto incomplete;
else /* m >= 3 */
{
unsigned char c3 = (unsigned char) p[2];
if ((c3 ^ 0x80) < 0x40)
{
if (pwc != NULL)
*pwc = ((unsigned int) (c & 0x0f) << 12)
| ((unsigned int) (c2 ^ 0x80) << 6)
| (unsigned int) (c3 ^ 0x80);
res = 3;
goto success;
}
}
}
}
}
else if (c <= 0xf4)
{
if (m == 1)
goto incomplete;
else
{
unsigned char c2 = (unsigned char) p[1];
if ((c2 ^ 0x80) < 0x40
&& (c >= 0xf1 || c2 >= 0x90)
&& (c < 0xf4 || (c == 0xf4 && c2 < 0x90)))
{
if (m == 2)
goto incomplete;
else
{
unsigned char c3 = (unsigned char) p[2];
if ((c3 ^ 0x80) < 0x40)
{
if (m == 3)
goto incomplete;
else /* m >= 4 */
{
unsigned char c4 = (unsigned char) p[3];
if ((c4 ^ 0x80) < 0x40)
{
if (pwc != NULL)
*pwc = ((unsigned int) (c & 0x07) << 18)
| ((unsigned int) (c2 ^ 0x80) << 12)
| ((unsigned int) (c3 ^ 0x80) << 6)
| (unsigned int) (c4 ^ 0x80);
res = 4;
goto success;
}
}
}
}
}
}
}
}
goto invalid;
}
else
{
/* The hidden internal state of mbtowc would make this function not
multi-thread safe. Achieve multi-thread safety through a lock. */
gl_lock_lock (mbtowc_lock);
/* Put the hidden internal state of mbtowc into its initial state.
This is needed at least with glibc, uClibc, and MSVC CRT.
See <https://sourceware.org/bugzilla/show_bug.cgi?id=9674>. */
mbtowc (NULL, NULL, 0);
res = mbtowc (pwc, p, m);
gl_lock_unlock (mbtowc_lock);
if (res >= 0)
{
if (pwc != NULL && ((*pwc == 0) != (res == 0)))
abort ();
goto success;
}
/* mbtowc does not distinguish between invalid and incomplete multibyte
sequences. But mbrtowc needs to make this distinction.
There are two possible approaches:
- Use iconv() and its return value.
- Use built-in knowledge about the possible encodings.
Given the low quality of implementation of iconv() on the systems
that lack mbrtowc(), we use the second approach.
The possible encodings are:
- 8-bit encodings,
- EUC-JP, EUC-KR, GB2312, EUC-TW, BIG5, GB18030, SJIS,
- UTF-8 (already handled above).
Use specialized code for each. */
if (m >= 4 || m >= MB_CUR_MAX)
goto invalid;
/* Here MB_CUR_MAX > 1 and 0 < m < 4. */
switch (enc)
{
/* As a reference for this code, you can use the GNU libiconv
implementation. Look for uses of the RET_TOOFEW macro. */
case enc_eucjp: /* EUC-JP */
{
if (m == 1)
{
unsigned char c = (unsigned char) p[0];
if ((c >= 0xa1 && c < 0xff) || c == 0x8e || c == 0x8f)
goto incomplete;
}
if (m == 2)
{
unsigned char c = (unsigned char) p[0];
if (c == 0x8f)
{
unsigned char c2 = (unsigned char) p[1];
if (c2 >= 0xa1 && c2 < 0xff)
goto incomplete;
}
}
goto invalid;
}
case enc_94: /* EUC-KR, GB2312, BIG5 */
{
if (m == 1)
{
unsigned char c = (unsigned char) p[0];
if (c >= 0xa1 && c < 0xff)
goto incomplete;
}
goto invalid;
}
case enc_euctw: /* EUC-TW */
{
if (m == 1)
{
unsigned char c = (unsigned char) p[0];
if ((c >= 0xa1 && c < 0xff) || c == 0x8e)
goto incomplete;
}
else /* m == 2 || m == 3 */
{
unsigned char c = (unsigned char) p[0];
if (c == 0x8e)
goto incomplete;
}
goto invalid;
}
case enc_gb18030: /* GB18030 */
{
if (m == 1)
{
unsigned char c = (unsigned char) p[0];
if ((c >= 0x90 && c <= 0xe3) || (c >= 0xf8 && c <= 0xfe))
goto incomplete;
}
else /* m == 2 || m == 3 */
{
unsigned char c = (unsigned char) p[0];
if (c >= 0x90 && c <= 0xe3)
{
unsigned char c2 = (unsigned char) p[1];
if (c2 >= 0x30 && c2 <= 0x39)
{
if (m == 2)
goto incomplete;
else /* m == 3 */
{
unsigned char c3 = (unsigned char) p[2];
if (c3 >= 0x81 && c3 <= 0xfe)
goto incomplete;
}
}
}
}
goto invalid;
}
case enc_sjis: /* SJIS */
{
if (m == 1)
{
unsigned char c = (unsigned char) p[0];
if ((c >= 0x81 && c <= 0x9f) || (c >= 0xe0 && c <= 0xea)
|| (c >= 0xf0 && c <= 0xf9))
goto incomplete;
}
goto invalid;
}
default:
/* An unknown multibyte encoding. */
goto incomplete;
}
}
success:
/* res >= 0 is the corrected return value of mbtowc (pwc, p, m). */
if (nstate >= (res > 0 ? res : 1))
abort ();
res -= nstate;
pstate[0] = 0;
return res;
incomplete:
{
size_t k = nstate;
/* Here 0 <= k < m < 4. */
pstate[++k] = s[0];
if (k < m)
{
pstate[++k] = s[1];
if (k < m)
pstate[++k] = s[2];
}
if (k != m)
abort ();
}
pstate[0] = m;
return (size_t)(-2);
invalid:
errno = EILSEQ;
/* The conversion state is undefined, says POSIX. */
return (size_t)(-1);
}
# define FITS_IN_CHAR_TYPE(wc) ((wc) <= WCHAR_MAX)
# include "mbrtowc-impl.h"
}
#else
/* Override the system's mbrtowc() function. */
# if MBRTOWC_IN_C_LOCALE_MAYBE_EILSEQ
# include "hard-locale.h"
# include <locale.h>
# endif
# undef mbrtowc
size_t
@ -512,14 +137,20 @@ rpl_mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
}
# endif
# if MBRTOWC_STORES_INCOMPLETE_BUG
ret = mbrtowc (&wc, s, n, ps);
if (ret < (size_t) -2 && pwc != NULL)
*pwc = wc;
# else
ret = mbrtowc (pwc, s, n, ps);
# endif
# if MBRTOWC_NUL_RETVAL_BUG
if (ret < (size_t) -2 && !*pwc)
return 0;
# endif
# if C_LOCALE_MAYBE_EILSEQ
# if MBRTOWC_IN_C_LOCALE_MAYBE_EILSEQ
if ((size_t) -2 <= ret && n != 0 && ! hard_locale (LC_CTYPE))
{
unsigned char uc = *s;