summaryrefslogtreecommitdiffstats
path: root/gl/lib/fnmatch.c
diff options
context:
space:
mode:
Diffstat (limited to 'gl/lib/fnmatch.c')
-rw-r--r--gl/lib/fnmatch.c103
1 files changed, 67 insertions, 36 deletions
diff --git a/gl/lib/fnmatch.c b/gl/lib/fnmatch.c
index 32cfb48..832b5a1 100644
--- a/gl/lib/fnmatch.c
+++ b/gl/lib/fnmatch.c
@@ -1,4 +1,4 @@
-/* Copyright (C) 1991-2023 Free Software Foundation, Inc.
+/* Copyright (C) 1991-2024 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
@@ -35,9 +35,40 @@
#if defined _LIBC || HAVE_ALLOCA
# include <alloca.h>
#endif
-#include <wchar.h>
-#include <wctype.h>
#include <stddef.h>
+#include <uchar.h>
+#if defined _LIBC || !_GL_SMALL_WCHAR_T
+/* It's OK to use wchar_t, since it's wide enough. */
+# include <wchar.h>
+# include <wctype.h>
+# define WCHAR_T wchar_t
+# define WINT_T wint_t
+# define BTOWC btowc
+# define MBSRTOWCS mbsrtowcs
+# define WCSLEN wcslen
+# define WCSCAT wcscat
+# define WMEMPCPY wmempcpy
+# define WMEMCHR wmemchr
+# define TOWLOWER towlower
+# define WCTYPE_T wctype_t
+# define WCTYPE wctype
+# define ISWCTYPE iswctype
+#else
+/* wchar_t is too small, use char32_t instead. */
+# include "unistr.h"
+# define WCHAR_T char32_t
+# define WINT_T char32_t
+# define BTOWC btoc32
+# define MBSRTOWCS mbsrtoc32s
+# define WCSLEN u32_strlen
+# define WCSCAT u32_strcat
+# define WMEMPCPY u32_pcpy
+# define WMEMCHR(S, C, N) u32_chr (S, N, C)
+# define TOWLOWER c32tolower
+# define WCTYPE_T c32_type_test_t
+# define WCTYPE c32_get_type_test
+# define ISWCTYPE c32_apply_type_test
+#endif
/* We need some of the locale data (the collation sequence information)
but there is no interface to get this information in general. Therefore
@@ -111,7 +142,7 @@ typedef ptrdiff_t idx_t;
# define CHAR_CLASS_MAX_LENGTH 256
#endif
-#define IS_CHAR_CLASS(string) wctype (string)
+#define IS_CHAR_CLASS(string) WCTYPE (string)
/* Avoid depending on library functions or files
whose names are inconsistent. */
@@ -129,7 +160,7 @@ static int posixly_correct;
#define END end_pattern
#define STRUCT fnmatch_struct
#define L_(CS) CS
-#define BTOWC(C) btowc (C)
+#define UCHAR_TO_WCHAR(C) BTOWC (C)
#define STRLEN(S) strlen (S)
#define STRCAT(D, S) strcat (D, S)
#define MEMPCPY(D, S, N) mempcpy (D, S, N)
@@ -142,19 +173,19 @@ static int posixly_correct;
#include "fnmatch_loop.c"
-#define FOLD(c) ((flags & FNM_CASEFOLD) ? towlower (c) : (c))
-#define CHAR wchar_t
-#define UCHAR wint_t
-#define INT wint_t
+#define FOLD(c) ((flags & FNM_CASEFOLD) ? TOWLOWER (c) : (c))
+#define CHAR WCHAR_T
+#define UCHAR WINT_T
+#define INT WINT_T
#define FCT internal_fnwmatch
#define EXT ext_wmatch
#define END end_wpattern
#define L_(CS) L##CS
-#define BTOWC(C) (C)
-#define STRLEN(S) wcslen (S)
-#define STRCAT(D, S) wcscat (D, S)
-#define MEMPCPY(D, S, N) wmempcpy (D, S, N)
-#define MEMCHR(S, C, N) wmemchr (S, C, N)
+#define UCHAR_TO_WCHAR(C) (C)
+#define STRLEN(S) WCSLEN (S)
+#define STRCAT(D, S) WCSCAT (D, S)
+#define MEMPCPY(D, S, N) WMEMPCPY (D, S, N)
+#define MEMCHR(S, C, N) WMEMCHR (S, C, N)
#define WIDE_CHAR_VERSION 1
#ifdef _LIBC
/* Change the name the header defines so it doesn't conflict with
@@ -172,8 +203,8 @@ static int posixly_correct;
for a member of the portable character set is the same code point as
its single-byte encoding, we can use a simplified method to convert the
string to a multibyte character string. */
-static wctype_t
-is_char_class (const wchar_t *wcs)
+static WCTYPE_T
+is_char_class (const WCHAR_T *wcs)
{
char s[CHAR_CLASS_MAX_LENGTH + 1];
char *cp = s;
@@ -184,7 +215,7 @@ is_char_class (const wchar_t *wcs)
#ifdef _LIBC
if (*wcs < 0x20 || *wcs > 0x7e
|| *wcs == 0x24 || *wcs == 0x40 || *wcs == 0x60)
- return (wctype_t) 0;
+ return (WCTYPE_T) 0;
#else
switch (*wcs)
{
@@ -210,13 +241,13 @@ is_char_class (const wchar_t *wcs)
case L'z': case L'{': case L'|': case L'}': case L'~':
break;
default:
- return (wctype_t) 0;
+ return (WCTYPE_T) 0;
}
#endif
/* Avoid overrunning the buffer. */
if (cp == s + CHAR_CLASS_MAX_LENGTH)
- return (wctype_t) 0;
+ return (WCTYPE_T) 0;
*cp++ = (char) *wcs++;
}
@@ -224,7 +255,7 @@ is_char_class (const wchar_t *wcs)
*cp = '\0';
- return wctype (s);
+ return WCTYPE (s);
}
#define IS_CHAR_CLASS(string) is_char_class (string)
@@ -239,10 +270,10 @@ fnmatch (const char *pattern, const char *string, int flags)
mbstate_t ps;
size_t n;
const char *p;
- wchar_t *wpattern_malloc = NULL;
- wchar_t *wpattern;
- wchar_t *wstring_malloc = NULL;
- wchar_t *wstring;
+ WCHAR_T *wpattern_malloc = NULL;
+ WCHAR_T *wpattern;
+ WCHAR_T *wstring_malloc = NULL;
+ WCHAR_T *wstring;
size_t alloca_used = 0;
/* Convert the strings into wide characters. */
@@ -251,9 +282,9 @@ fnmatch (const char *pattern, const char *string, int flags)
n = strnlen (pattern, 1024);
if (__glibc_likely (n < 1024))
{
- wpattern = (wchar_t *) alloca_account ((n + 1) * sizeof (wchar_t),
+ wpattern = (WCHAR_T *) alloca_account ((n + 1) * sizeof (WCHAR_T),
alloca_used);
- n = mbsrtowcs (wpattern, &p, n + 1, &ps);
+ n = MBSRTOWCS (wpattern, &p, n + 1, &ps);
if (__glibc_unlikely (n == (size_t) -1))
/* Something wrong.
XXX Do we have to set 'errno' to something which mbsrtows hasn't
@@ -268,23 +299,23 @@ fnmatch (const char *pattern, const char *string, int flags)
else
{
prepare_wpattern:
- n = mbsrtowcs (NULL, &pattern, 0, &ps);
+ n = MBSRTOWCS (NULL, &pattern, 0, &ps);
if (__glibc_unlikely (n == (size_t) -1))
/* Something wrong.
XXX Do we have to set 'errno' to something which mbsrtows hasn't
already done? */
return -1;
- if (__glibc_unlikely (n >= (size_t) -1 / sizeof (wchar_t)))
+ if (__glibc_unlikely (n >= (size_t) -1 / sizeof (WCHAR_T)))
{
__set_errno (ENOMEM);
return -2;
}
wpattern_malloc = wpattern
- = (wchar_t *) malloc ((n + 1) * sizeof (wchar_t));
+ = (WCHAR_T *) malloc ((n + 1) * sizeof (WCHAR_T));
assert (mbsinit (&ps));
if (wpattern == NULL)
return -2;
- (void) mbsrtowcs (wpattern, &pattern, n + 1, &ps);
+ (void) MBSRTOWCS (wpattern, &pattern, n + 1, &ps);
}
assert (mbsinit (&ps));
@@ -292,9 +323,9 @@ fnmatch (const char *pattern, const char *string, int flags)
p = string;
if (__glibc_likely (n < 1024))
{
- wstring = (wchar_t *) alloca_account ((n + 1) * sizeof (wchar_t),
+ wstring = (WCHAR_T *) alloca_account ((n + 1) * sizeof (WCHAR_T),
alloca_used);
- n = mbsrtowcs (wstring, &p, n + 1, &ps);
+ n = MBSRTOWCS (wstring, &p, n + 1, &ps);
if (__glibc_unlikely (n == (size_t) -1))
{
/* Something wrong.
@@ -313,13 +344,13 @@ fnmatch (const char *pattern, const char *string, int flags)
else
{
prepare_wstring:
- n = mbsrtowcs (NULL, &string, 0, &ps);
+ n = MBSRTOWCS (NULL, &string, 0, &ps);
if (__glibc_unlikely (n == (size_t) -1))
/* Something wrong.
XXX Do we have to set 'errno' to something which mbsrtows hasn't
already done? */
goto free_return;
- if (__glibc_unlikely (n >= (size_t) -1 / sizeof (wchar_t)))
+ if (__glibc_unlikely (n >= (size_t) -1 / sizeof (WCHAR_T)))
{
free (wpattern_malloc);
__set_errno (ENOMEM);
@@ -327,14 +358,14 @@ fnmatch (const char *pattern, const char *string, int flags)
}
wstring_malloc = wstring
- = (wchar_t *) malloc ((n + 1) * sizeof (wchar_t));
+ = (WCHAR_T *) malloc ((n + 1) * sizeof (WCHAR_T));
if (wstring == NULL)
{
free (wpattern_malloc);
return -2;
}
assert (mbsinit (&ps));
- (void) mbsrtowcs (wstring, &string, n + 1, &ps);
+ (void) MBSRTOWCS (wstring, &string, n + 1, &ps);
}
int res = internal_fnwmatch (wpattern, wstring, wstring + n,