> > Thus, wouldn't it be wise to fix strcasecmp itself, and include it
> > unconditionally in GnuLib ?
>
> Yes. Here's a patch to that effect. (Completely untested so far. Also
> quite inefficient. One can do better with mbrtowc().)
Well, that code was broken. And a similar code I wrote based on mbrtowc()
was broken as well: It did not fulfill the transitivity rule
cmp (s1, s2) < 0 && cmp (s2, s3) < 0 ==> cmp (s1, s3) < 0
when some of s1, s2, s3 were strings containing invalid multibyte characters.
The lesson to be learned from this is: you cannot write good multibyte
support if you don't have a "multibyte character" abstraction. If you use
the naked C99 functions, the code will be so complicated that it will
never behave well and consistently.
So I dug out an earlier 'mbchar' abstraction, and rewrite strcasecmp()
as follows. With this patch, a simple test program shows that it works
fine in a Turkish locale:
========================================================
#include <locale.h>
#include <stdio.h>
int main (int argc, char * argv[])
{
setlocale (LC_ALL, "");
printf ("cmp %s %s -> %d\n", argv[1], argv[2],
rpl_strcasecmp (argv[1], argv[2]));
return 0;
}
=========================================================
$ ./a.out abi ABI
cmp abi ABI -> 0
$ LC_ALL=tr_TR-UTF-8 ./a.out abi ABI
cmp abi ABI -> -148
Bruno
Index: modules/strcase
===================================================================
RCS file: /cvsroot/gnulib/gnulib/modules/strcase,v
retrieving revision 1.5
diff -c -3 -r1.5 strcase
*** modules/strcase 22 Sep 2004 15:11:04 -0000 1.5
--- modules/strcase 16 Aug 2005 12:13:35 -0000
***************
*** 6,13 ****
--- 6,16 ----
lib/strcasecmp.c
lib/strncasecmp.c
m4/strcase.m4
+ m4/mbrtowc.m4
Depends-on:
+ strnlen1
+ mbchar
configure.ac:
gl_STRCASE
Index: lib/strcase.h
===================================================================
RCS file: /cvsroot/gnulib/gnulib/lib/strcase.h,v
retrieving revision 1.4
diff -c -3 -r1.4 strcase.h
*** lib/strcase.h 14 May 2005 06:03:58 -0000 1.4
--- lib/strcase.h 16 Aug 2005 12:13:35 -0000
***************
*** 1,5 ****
/* Case-insensitive string comparison functions.
! Copyright (C) 1995-1996, 2001, 2003 Free Software Foundation, Inc.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
--- 1,5 ----
/* Case-insensitive string comparison functions.
! Copyright (C) 1995-1996, 2001, 2003, 2005 Free Software Foundation, Inc.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
***************
*** 29,35 ****
/* Compare strings S1 and S2, ignoring case, returning less than, equal to or
greater than zero if S1 is lexicographically less than, equal to or greater
than S2.
! Note: This function does not work correctly in multibyte locales. */
extern int strcasecmp (const char *s1, const char *s2);
/* Compare no more than N characters of strings S1 and S2, ignoring case,
--- 29,36 ----
/* Compare strings S1 and S2, ignoring case, returning less than, equal to or
greater than zero if S1 is lexicographically less than, equal to or greater
than S2.
! Note: This function may, in multibyte locales, return 0 for strings of
! different lengths! */
extern int strcasecmp (const char *s1, const char *s2);
/* Compare no more than N characters of strings S1 and S2, ignoring case,
Index: lib/strncasecmp.c
===================================================================
RCS file: /cvsroot/gnulib/gnulib/lib/strncasecmp.c,v
retrieving revision 1.4
diff -c -3 -r1.4 strncasecmp.c
*** lib/strncasecmp.c 2 Jan 1999 15:55:44 -0000 1.4
--- lib/strncasecmp.c 16 Aug 2005 12:13:35 -0000
***************
*** 1,2 ****
! #define LENGTH_LIMIT
! #include "strcasecmp.c"
--- 1,58 ----
! /* strncasecmp.c -- case insensitive string comparator
! Copyright (C) 1998, 1999 Free Software Foundation, Inc.
!
! This program is free software; you can redistribute it and/or modify
! it under the terms of the GNU General Public License as published by
! the Free Software Foundation; either version 2, or (at your option)
! any later version.
!
! This program is distributed in the hope that it will be useful,
! but WITHOUT ANY WARRANTY; without even the implied warranty of
! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
! GNU General Public License for more details.
!
! You should have received a copy of the GNU General Public License
! along with this program; if not, write to the Free Software Foundation,
! Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
!
! #if HAVE_CONFIG_H
! # include <config.h>
! #endif
!
! /* Specification. */
! #include "strcase.h"
!
! #include <ctype.h>
!
! #define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch))
!
! /* Compare no more than N bytes of strings S1 and S2,
! ignoring case, returning less than, equal to or
! greater than zero if S1 is lexicographically less
! than, equal to or greater than S2. */
!
! int
! strncasecmp (const char *s1, const char *s2, size_t n)
! {
! register const unsigned char *p1 = (const unsigned char *) s1;
! register const unsigned char *p2 = (const unsigned char *) s2;
! unsigned char c1, c2;
!
! if (p1 == p2 || n == 0)
! return 0;
!
! do
! {
! c1 = TOLOWER (*p1);
! c2 = TOLOWER (*p2);
!
! if (--n == 0 || c1 == '\0')
! break;
!
! ++p1;
! ++p2;
! }
! while (c1 == c2);
!
! return c1 - c2;
! }
Index: lib/strcasecmp.c
===================================================================
RCS file: /cvsroot/gnulib/gnulib/lib/strcasecmp.c,v
retrieving revision 1.6
diff -c -3 -r1.6 strcasecmp.c
*** lib/strcasecmp.c 14 May 2005 06:03:58 -0000 1.6
--- lib/strcasecmp.c 16 Aug 2005 12:13:35 -0000
***************
*** 1,5 ****
! /* strcasecmp.c -- case insensitive string comparator
! Copyright (C) 1998, 1999 Free Software Foundation, Inc.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
--- 1,7 ----
! /* Case-insensitive string comparison function.
! Copyright (C) 1998, 1999, 2005 Free Software Foundation, Inc.
! Written by Bruno Haible <[EMAIL PROTECTED]>, 2005,
! based on earlier glibc code.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
***************
*** 15,66 ****
along with this program; if not, write to the Free Software Foundation,
Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
! #if HAVE_CONFIG_H
# include <config.h>
#endif
! #ifdef LENGTH_LIMIT
! # define STRXCASECMP_FUNCTION strncasecmp
! # define STRXCASECMP_DECLARE_N , size_t n
! # define LENGTH_LIMIT_EXPR(Expr) Expr
! #else
! # define STRXCASECMP_FUNCTION strcasecmp
! # define STRXCASECMP_DECLARE_N /* empty */
! # define LENGTH_LIMIT_EXPR(Expr) 0
! #endif
- #include <stddef.h>
#include <ctype.h>
! #define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch))
! /* Compare {{no more than N characters of }}strings S1 and S2,
! ignoring case, returning less than, equal to or
! greater than zero if S1 is lexicographically less
! than, equal to or greater than S2. */
! int
! STRXCASECMP_FUNCTION (const char *s1, const char *s2 STRXCASECMP_DECLARE_N)
{
! register const unsigned char *p1 = (const unsigned char *) s1;
! register const unsigned char *p2 = (const unsigned char *) s2;
! unsigned char c1, c2;
! if (p1 == p2 || LENGTH_LIMIT_EXPR (n == 0))
return 0;
! do
{
! c1 = TOLOWER (*p1);
! c2 = TOLOWER (*p2);
! if (LENGTH_LIMIT_EXPR (--n == 0) || c1 == '\0')
! break;
! ++p1;
! ++p2;
}
! while (c1 == c2);
! return c1 - c2;
}
--- 17,241 ----
along with this program; if not, write to the Free Software Foundation,
Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
! #ifdef HAVE_CONFIG_H
# include <config.h>
#endif
! /* Specification. */
! #include "strcase.h"
#include <ctype.h>
! #if HAVE_MBRTOWC
! #include "strnlen1.h"
! /* Like mbiter.h, except it doesn't look at the entire string. */
!
! #include "mbchar.h"
!
! #include <assert.h>
! #include <stdbool.h>
! #include <stdlib.h>
! #include <wchar.h>
! #include <wctype.h>
!
! struct mbiter_multi
! {
! bool at_end; /* true if the end of the string has been
reached */
! bool in_shift; /* true if next byte may not be interpreted as ASCII */
! mbstate_t state; /* if in_shift: current shift state */
! bool next_done; /* true if mbi_avail has already filled the following */
! struct mbchar cur; /* the current character:
! const char *cur.ptr pointer to current character
! The following are only valid after mbi_avail.
! size_t cur.bytes number of bytes of current character
! bool cur.wc_valid true if wc is a valid wide character
! wchar_t cur.wc if wc_valid: the current character
! */
! };
!
! static inline void
! mbiter_multi_next (struct mbiter_multi *iter)
! {
! if (iter->next_done)
! return;
! if (iter->in_shift)
! goto with_shift;
! /* Handle most ASCII characters quickly, without calling mbrtowc(). */
! if (is_basic (*iter->cur.ptr))
! {
! /* These characters are part of the basic character set. ISO C 99
! guarantees that their wide character code is identical to their
! char code. */
! iter->cur.bytes = 1;
! iter->cur.wc = *iter->cur.ptr;
! iter->cur.wc_valid = true;
! }
! else
! {
! assert (mbsinit (&iter->state));
! iter->in_shift = true;
! with_shift:
! iter->cur.bytes = mbrtowc (&iter->cur.wc, iter->cur.ptr,
! strnlen1 (iter->cur.ptr, MB_CUR_MAX),
! &iter->state);
! if (iter->cur.bytes == (size_t) -1)
! {
! /* An invalid multibyte sequence was encountered. */
! iter->cur.bytes = 1;
! iter->cur.wc_valid = false;
! /* Whether to set iter->in_shift = false and reset iter->state
! or not is not very important; the string is bogus anyway. */
! }
! else if (iter->cur.bytes == (size_t) -2)
! {
! /* An incomplete multibyte character at the end. */
! iter->cur.bytes = strlen (iter->cur.ptr) + 1;
! iter->cur.wc_valid = false;
! /* Whether to set iter->in_shift = false and reset iter->state
! or not is not important; the string end is reached anyway. */
! }
! else
! {
! if (iter->cur.bytes == 0)
! {
! /* A null wide character was encountered. */
! iter->cur.bytes = 1;
! assert (*iter->cur.ptr == '\0');
! assert (iter->cur.wc == 0);
! }
! iter->cur.wc_valid = true;
!
! /* When in the initial state, we can go back treating ASCII
! characters more quickly. */
! if (mbsinit (&iter->state))
! iter->in_shift = false;
! }
! }
! iter->next_done = true;
! }
!
! static inline void
! mbiter_multi_reloc (struct mbiter_multi *iter, ptrdiff_t ptrdiff)
{
! iter->cur.ptr += ptrdiff;
! }
!
! /* Iteration macros. */
! typedef struct mbiter_multi mbi_iterator_t;
! #define mbi_init(iter, startptr) \
! ((iter).cur.ptr = (startptr), (iter).at_end = false, \
! (iter).in_shift = false, memset (&(iter).state, '\0', sizeof (mbstate_t)),
\
! (iter).next_done = false)
! #define mbi_avail(iter) \
! (!(iter).at_end && (mbiter_multi_next (&(iter)), true))
! #define mbi_advance(iter) \
! ((mb_isnul ((iter).cur) ? ((iter).at_end = true) : 0), \
! (iter).cur.ptr += (iter).cur.bytes, (iter).next_done = false)
!
! /* Access to the current character. */
! #define mbi_cur(iter) (iter).cur
! #define mbi_cur_ptr(iter) (iter).cur.ptr
!
! #endif
!
! #define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch))
! /* Compare strings S1 and S2, ignoring case, returning less than, equal to or
! greater than zero if S1 is lexicographically less than, equal to or greater
! than S2.
! Note: This function may, in multibyte locales, return 0 for strings of
! different lengths! */
! int
! strcasecmp (const char *s1, const char *s2)
! {
! if (s1 == s2)
return 0;
! /* Be careful not to look at the entire extent of s1 or s2 until needed.
! This is useful because when two strings differ, the difference is
! most often already in the very few first characters. */
! #if HAVE_MBRTOWC
! if (MB_CUR_MAX > 1)
{
! mbi_iterator_t iter1;
! mbi_iterator_t iter2;
! mbi_init (iter1, s1);
! mbi_init (iter2, s2);
! while (mbi_avail (iter1) && mbi_avail (iter2))
! {
! /* Sort invalid characters after all valid ones. */
! if (!mbi_cur (iter1).wc_valid)
! {
! if (!mbi_cur (iter2).wc_valid)
! {
! /* Compare two invalid characters. */
! int cmp;
!
! if (mbi_cur (iter1).bytes > mbi_cur (iter2).bytes)
! return 1;
! if (mbi_cur (iter1).bytes < mbi_cur (iter2).bytes)
! return -1;
! cmp = memcmp (mbi_cur_ptr (iter1), mbi_cur_ptr (iter2),
! mbi_cur (iter1).bytes);
! if (cmp != 0)
! return cmp;
! }
! else
! /* mbi_cur (iter1) invalid, mbi_cur (iter2) valid. */
! return 1;
! }
! else
! {
! if (!mbi_cur (iter2).wc_valid)
! /* mbi_cur (iter1) valid, mbi_cur (iter2) invalid. */
! return -1;
! else
! {
! /* Compare two valid characters. */
! wchar_t c1 = towlower (mbi_cur (iter1).wc);
! wchar_t c2 = towlower (mbi_cur (iter2).wc);
!
! if (c1 > c2)
! return 1;
! if (c1 < c2)
! return -1;
! }
! }
! mbi_advance (iter1);
! mbi_advance (iter2);
! }
! if (mbi_avail (iter1))
! /* s2 terminated before s1. */
! return 1;
! if (mbi_avail (iter2))
! /* s1 terminated before s2. */
! return -1;
! return 0;
}
! else
! #endif
! {
! const unsigned char *p1 = (const unsigned char *) s1;
! const unsigned char *p2 = (const unsigned char *) s2;
! unsigned char c1, c2;
!
! do
! {
! c1 = TOLOWER (*p1);
! c2 = TOLOWER (*p2);
!
! if (c1 == '\0')
! break;
!
! ++p1;
! ++p2;
! }
! while (c1 == c2);
! return c1 - c2;
! }
}
Index: m4/strcase.m4
===================================================================
RCS file: /cvsroot/gnulib/gnulib/m4/strcase.m4,v
retrieving revision 1.2
diff -c -3 -r1.2 strcase.m4
*** m4/strcase.m4 18 Jan 2005 13:07:56 -0000 1.2
--- m4/strcase.m4 16 Aug 2005 12:13:35 -0000
***************
*** 1,5 ****
! # strcase.m4 serial 1
! dnl Copyright (C) 2002 Free Software Foundation, Inc.
dnl This file is free software; the Free Software Foundation
dnl gives unlimited permission to copy and/or distribute it,
dnl with or without modifications, as long as this notice is preserved.
--- 1,5 ----
! # strcase.m4 serial 2
! dnl Copyright (C) 2002, 2005 Free Software Foundation, Inc.
dnl This file is free software; the Free Software Foundation
dnl gives unlimited permission to copy and/or distribute it,
dnl with or without modifications, as long as this notice is preserved.
***************
*** 12,21 ****
AC_DEFUN([gl_FUNC_STRCASECMP],
[
! AC_REPLACE_FUNCS(strcasecmp)
! if test $ac_cv_func_strcasecmp = no; then
! gl_PREREQ_STRCASECMP
! fi
])
AC_DEFUN([gl_FUNC_STRNCASECMP],
--- 12,22 ----
AC_DEFUN([gl_FUNC_STRCASECMP],
[
! dnl No known system has a strcasecmp() function that works correctly in
! dnl multibyte locales. Therefore we use our version always.
! AC_LIBOBJ(strcasecmp)
! AC_DEFINE(strcasecmp, rpl_strcasecmp, [Define to rpl_strcasecmp always.])
! gl_PREREQ_STRCASECMP
])
AC_DEFUN([gl_FUNC_STRNCASECMP],
***************
*** 28,34 ****
# Prerequisites of lib/strcasecmp.c.
AC_DEFUN([gl_PREREQ_STRCASECMP], [
! :
])
# Prerequisites of lib/strncasecmp.c.
--- 29,35 ----
# Prerequisites of lib/strcasecmp.c.
AC_DEFUN([gl_PREREQ_STRCASECMP], [
! gl_FUNC_MBRTOWC
])
# Prerequisites of lib/strncasecmp.c.
_______________________________________________
bug-gnulib mailing list
bug-gnulib@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnulib
