[poky] [PATCH 2/4] coreutils: Add various bug fixes
Mark Hatle
mark.hatle at windriver.com
Thu Feb 3 19:44:28 PST 2011
Add a number of bug fixes, mostly imported from Fedora and Wind River
Linux.
cp-i-u: fix unnecessary prompting
fix-install: Fix installing to a dangling symlink
i18n: li18nux/lsb compliance
ls-x: Fix incorrect output
overflow: Fix potential overflow in who command
Signed-off-by: Mark Hatle <mark.hatle at windriver.com>
---
.../coreutils-6.9/coreutils-6.9-cp-i-u.patch | 118 +
.../coreutils-6.9/coreutils-fix-install.patch | 99 +
.../coreutils/coreutils-6.9/coreutils-i18n.patch | 4049 ++++++++++++++++++++
.../coreutils/coreutils-6.9/coreutils-ls-x.patch | 115 +
.../coreutils-6.9/coreutils-overflow.patch | 17 +
meta/recipes-core/coreutils/coreutils_6.9.bb | 5 +
6 files changed, 4403 insertions(+), 0 deletions(-)
create mode 100644 meta/recipes-core/coreutils/coreutils-6.9/coreutils-6.9-cp-i-u.patch
create mode 100644 meta/recipes-core/coreutils/coreutils-6.9/coreutils-fix-install.patch
create mode 100644 meta/recipes-core/coreutils/coreutils-6.9/coreutils-i18n.patch
create mode 100644 meta/recipes-core/coreutils/coreutils-6.9/coreutils-ls-x.patch
create mode 100644 meta/recipes-core/coreutils/coreutils-6.9/coreutils-overflow.patch
diff --git a/meta/recipes-core/coreutils/coreutils-6.9/coreutils-6.9-cp-i-u.patch b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-6.9-cp-i-u.patch
new file mode 100644
index 0000000..6fec683
--- /dev/null
+++ b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-6.9-cp-i-u.patch
@@ -0,0 +1,118 @@
+This patch was imported from the Fedora Core 8 coreutils-6.9-9 package.
+
+The package is stated as being Licensed as GPLv2+.
+
+Signed-off-by: Mark Hatle <mark.hatle at windriver.com>
+
+----
+
+When "cp -i --update old new" would do nothing because "new" is
+newer than "old", cp would nonetheless prompt for whether it is
+ok to overwrite "new". Then, regardless of the response (because
+of the --update option), cp would do nothing.
+
+The following patch eliminates the unnecessary prompt in that case.
+
+diff --git a/src/copy.c b/src/copy.c
+index b7bf73b..0e549d2 100644
+--- a/src/copy.c
++++ b/src/copy.c
+@@ -1210,6 +1210,30 @@ copy_internal (char const *src_name, char const *dst_name,
+ return false;
+ }
+
++ if (!S_ISDIR (src_mode) && x->update)
++ {
++ /* When preserving time stamps (but not moving within a file
++ system), don't worry if the destination time stamp is
++ less than the source merely because of time stamp
++ truncation. */
++ int options = ((x->preserve_timestamps
++ && ! (x->move_mode
++ && dst_sb.st_dev == src_sb.st_dev))
++ ? UTIMECMP_TRUNCATE_SOURCE
++ : 0);
++
++ if (0 <= utimecmp (dst_name, &dst_sb, &src_sb, options))
++ {
++ /* We're using --update and the destination is not older
++ than the source, so do not copy or move. Pretend the
++ rename succeeded, so the caller (if it's mv) doesn't
++ end up removing the source file. */
++ if (rename_succeeded)
++ *rename_succeeded = true;
++ return true;
++ }
++ }
++
+ /* When there is an existing destination file, we may end up
+ returning early, and hence not copying/moving the file.
+ This may be due to an interactive `negative' reply to the
+@@ -1302,30 +1326,6 @@ copy_internal (char const *src_name, char const *dst_name,
+ return false;
+ }
+ }
+-
+- if (x->update)
+- {
+- /* When preserving time stamps (but not moving within a file
+- system), don't worry if the destination time stamp is
+- less than the source merely because of time stamp
+- truncation. */
+- int options = ((x->preserve_timestamps
+- && ! (x->move_mode
+- && dst_sb.st_dev == src_sb.st_dev))
+- ? UTIMECMP_TRUNCATE_SOURCE
+- : 0);
+-
+- if (0 <= utimecmp (dst_name, &dst_sb, &src_sb, options))
+- {
+- /* We're using --update and the destination is not older
+- than the source, so do not copy or move. Pretend the
+- rename succeeded, so the caller (if it's mv) doesn't
+- end up removing the source file. */
+- if (rename_succeeded)
+- *rename_succeeded = true;
+- return true;
+- }
+- }
+ }
+
+ if (x->move_mode)
+diff --git a/tests/mv/update b/tests/mv/update
+index 0c06024..6c3d149 100755
+--- a/tests/mv/update
++++ b/tests/mv/update
+@@ -1,7 +1,7 @@
+ #!/bin/sh
+ # make sure --update works as advertised
+
+-# Copyright (C) 2001, 2004, 2006 Free Software Foundation, Inc.
++# Copyright (C) 2001, 2004, 2006-2007 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
+@@ -46,11 +46,16 @@ fi
+
+ fail=0
+
+-for cp_or_mv in cp mv; do
+- # This is a no-op.
+- $cp_or_mv --update old new || fail=1
+- case "`cat new`" in new) ;; *) fail=1 ;; esac
+- case "`cat old`" in old) ;; *) fail=1 ;; esac
++for interactive in '' -i; do
++ for cp_or_mv in cp mv; do
++ # This is a no-op, with no prompt.
++ # With coreutils-6.9 and earlier, using --update with -i would
++ # mistakenly elicit a prompt.
++ $cp_or_mv $interactive --update old new < /dev/null > out 2>&1 || fail=1
++ test -s out && fail=1
++ case "`cat new`" in new) ;; *) fail=1 ;; esac
++ case "`cat old`" in old) ;; *) fail=1 ;; esac
++ done
+ done
+
+ # This will actually perform the rename.
+--
+1.5.3.rc1.16.g9d6f
diff --git a/meta/recipes-core/coreutils/coreutils-6.9/coreutils-fix-install.patch b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-fix-install.patch
new file mode 100644
index 0000000..1b989f5
--- /dev/null
+++ b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-fix-install.patch
@@ -0,0 +1,99 @@
+The install command doesn't over write the dangling symlink, for
+example:
+
+$ install fileA /tmp/fileA
+
+If /tmp/fileA is a dangling symlink, there would be an error:
+
+install: cannot create regular file '/tmp/fileA': File exists
+
+This is because of the following code in copy.c:
+
+ if (!new_dst)
+ {
+ if (XSTAT (x, dst_name, &dst_sb) != 0)
+ {
+ if (errno != ENOENT)
+ {
+ error (0, errno, _("cannot stat %s"), quote (dst_name));
+ return false;
+ }
+ else
+ {
+ new_dst = true;
+ }
+ }
+
+XSTAT() use stat() for dst_name(the dangling symlink /tmp/fileA) when
+install.c invokes it, and stat will set errno to ENOENT, and then
+new_dst will be set to true which means that /tmp/fileA doesn't exist,
+then we will create /tmp/fileA without remove it first, so the error
+comes.
+
+This is fixed in a way which adds the member cmd_install in
+struct cp_options to make sure my change only affected to the install
+command and use lstat to fix the problem.
+
+Signed-off-by: Robert Yang <liezhi.yang at windriver.com>
+Signed-off-by: Mark Hatle <mark.hatle at windriver.com>
+
+---
+ src/copy.c | 10 +++++++++-
+ src/copy.h | 3 +++
+ src/install.c | 1 +
+ 3 files changed, 13 insertions(+), 1 deletions(-)
+
+diff --git a/src/copy.c b/src/copy.c
+--- a/src/copy.c
++++ b/src/copy.c
+@@ -1029,6 +1029,7 @@ copy_internal (char const *src_name, char const *dst_name,
+ bool delayed_ok;
+ bool copied_as_regular = false;
+ bool preserve_metadata;
++ int dst_stat_result;
+
+ if (x->move_mode && rename_succeeded)
+ *rename_succeeded = false;
+@@ -1069,7 +1070,14 @@ copy_internal (char const *src_name, char const *dst_name,
+
+ if (!new_dst)
+ {
+- if (XSTAT (x, dst_name, &dst_sb) != 0)
++ if ( x->cmd_install && ( x->backup_type == no_backups))
++ dst_stat_result = lstat (dst_name, &dst_sb);
++ else
++ {
++ dst_stat_result = XSTAT (x, dst_name, &dst_sb);
++ }
++
++ if (dst_stat_result != 0)
+ {
+ if (errno != ENOENT)
+ {
+diff --git a/src/copy.h b/src/copy.h
+--- a/src/copy.h
++++ b/src/copy.h
+@@ -114,6 +114,9 @@ struct cp_options
+ If that fails, then resort to copying. */
+ bool move_mode;
+
++ /* For the install command */
++ bool cmd_install;
++
+ /* Whether this process has appropriate privileges to chown a file
+ whose owner is not the effective user ID. */
+ bool chown_privileges;
+diff --git a/src/install.c b/src/install.c
+--- a/src/install.c
++++ b/src/install.c
+@@ -149,6 +149,7 @@ cp_option_init (struct cp_options *x)
+ x->hard_link = false;
+ x->interactive = I_UNSPECIFIED;
+ x->move_mode = false;
++ x->cmd_install = true;
+ x->chown_privileges = chown_privileges ();
+ x->one_file_system = false;
+ x->preserve_ownership = false;
+--
+1.7.0.1
+
diff --git a/meta/recipes-core/coreutils/coreutils-6.9/coreutils-i18n.patch b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-i18n.patch
new file mode 100644
index 0000000..3587186
--- /dev/null
+++ b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-i18n.patch
@@ -0,0 +1,4049 @@
+This patch was imported from the Fedora Core 8 coreutils-6.9-9 package.
+
+The package is stated as being Licensed as GPLv2+.
+
+The comment indicates that the purpose is lin18nux/lsb compliance.
+
+Signed-off-by: Mark Hatle <mark.hatle at windriver.com>
+
+--- /dev/null 2007-03-01 09:16:39.219409909 +0000
++++ coreutils-6.8+/tests/sort/sort-mb-tests 2007-03-01 15:08:24.000000000 +0000
+@@ -0,0 +1,58 @@
++#! /bin/sh
++case $# in
++ 0) xx='../../src/sort';;
++ *) xx="$1";;
++esac
++test "$VERBOSE" && echo=echo || echo=:
++$echo testing program: $xx
++errors=0
++test "$srcdir" || srcdir=.
++test "$VERBOSE" && $xx --version 2> /dev/null
++
++export LC_ALL=en_US.UTF-8
++locale -k LC_CTYPE 2>&1 | grep -q charmap.*UTF-8 || exit 77
++errors=0
++
++$xx -t ＠ -k2 -n mb1.I > mb1.O
++code=$?
++if test $code != 0; then
++ $echo "Test mb1 failed: $xx return code $code differs from expected value 0" 1>&2
++ errors=`expr $errors + 1`
++else
++ cmp mb1.O $srcdir/mb1.X > /dev/null 2>&1
++ case $? in
++ 0) if test "$VERBOSE"; then $echo "passed mb1"; fi;;
++ 1) $echo "Test mb1 failed: files mb1.O and $srcdir/mb1.X differ" 1>&2
++ (diff -c mb1.O $srcdir/mb1.X) 2> /dev/null
++ errors=`expr $errors + 1`;;
++ 2) $echo "Test mb1 may have failed." 1>&2
++ $echo The command "cmp mb1.O $srcdir/mb1.X" failed. 1>&2
++ errors=`expr $errors + 1`;;
++ esac
++fi
++
++$xx -t ＠ -k4 -n mb2.I > mb2.O
++code=$?
++if test $code != 0; then
++ $echo "Test mb2 failed: $xx return code $code differs from expected value 0" 1>&2
++ errors=`expr $errors + 1`
++else
++ cmp mb2.O $srcdir/mb2.X > /dev/null 2>&1
++ case $? in
++ 0) if test "$VERBOSE"; then $echo "passed mb2"; fi;;
++ 1) $echo "Test mb2 failed: files mb2.O and $srcdir/mb2.X differ" 1>&2
++ (diff -c mb2.O $srcdir/mb2.X) 2> /dev/null
++ errors=`expr $errors + 1`;;
++ 2) $echo "Test mb2 may have failed." 1>&2
++ $echo The command "cmp mb2.O $srcdir/mb2.X" failed. 1>&2
++ errors=`expr $errors + 1`;;
++ esac
++fi
++
++if test $errors = 0; then
++ $echo Passed all 113 tests. 1>&2
++else
++ $echo Failed $errors tests. 1>&2
++fi
++test $errors = 0 || errors=1
++exit $errors
+--- /dev/null 2007-03-01 09:16:39.219409909 +0000
++++ coreutils-6.8+/tests/sort/mb2.I 2007-03-01 15:08:24.000000000 +0000
+@@ -0,0 +1,4 @@
++Apple＠ＡＡ10＠＠20
++Banana＠ＡＡ5＠＠30
++Citrus＠ＡＡ20＠＠5
++Cherry＠ＡＡ30＠＠10
+--- /dev/null 2007-03-01 09:16:39.219409909 +0000
++++ coreutils-6.8+/tests/sort/mb2.X 2007-03-01 15:08:24.000000000 +0000
+@@ -0,0 +1,4 @@
++Citrus＠ＡＡ20＠＠5
++Cherry＠ＡＡ30＠＠10
++Apple＠ＡＡ10＠＠20
++Banana＠ＡＡ5＠＠30
+--- /dev/null 2007-03-01 09:16:39.219409909 +0000
++++ coreutils-6.8+/tests/sort/mb1.I 2007-03-01 15:08:24.000000000 +0000
+@@ -0,0 +1,4 @@
++Apple＠10
++Banana＠5
++Citrus＠20
++Cherry＠30
+--- /dev/null 2007-03-01 09:16:39.219409909 +0000
++++ coreutils-6.8+/tests/sort/mb1.X 2007-03-01 15:08:24.000000000 +0000
+@@ -0,0 +1,4 @@
++Banana＠5
++Apple＠10
++Citrus＠20
++Cherry＠30
+--- coreutils-6.8+/tests/sort/Makefile.am.i18n 2007-01-24 07:47:37.000000000 +0000
++++ coreutils-6.8+/tests/sort/Makefile.am 2007-03-01 15:09:59.000000000 +0000
+@@ -66,15 +66,17 @@
+ bigfield.O bigfield.E
+ ##test-files-end
+
+-EXTRA_DIST = Test.pm $x-tests $(explicit) $(maint_gen)
+-noinst_SCRIPTS = $x-tests
++run_gen += mb1.0 mb2.0
++
++EXTRA_DIST = Test.pm $x-tests $(explicit) $(maint_gen) mb1.I mb1.X mb2.I mb2.X
++noinst_SCRIPTS = $x-tests # $x-mb-tests
+ TESTS_ENVIRONMENT = \
+ CU_TEST_NAME=`basename $(abs_srcdir)`,$$tst \
+ PATH="$(VG_PATH_PREFIX)`pwd`/../../src$(PATH_SEPARATOR)$$PATH"
+
+ editpl = sed -e 's,@''PERL''@,$(PERL),g' -e 's,@''srcdir''@,$(srcdir),g'
+
+-TESTS = $x-tests
++TESTS = $x-tests $x-mb-tests
+
+ mk_script = $(srcdir)/../mk-script
+ $(srcdir)/$x-tests: $(mk_script) Test.pm Makefile.am
+--- coreutils-6.8+/lib/linebuffer.h.i18n 2005-05-14 07:44:24.000000000 +0100
++++ coreutils-6.8+/lib/linebuffer.h 2007-03-01 15:08:24.000000000 +0000
+@@ -22,6 +22,11 @@
+
+ # include <stdio.h>
+
++/* Get mbstate_t. */
++# if HAVE_WCHAR_H
++# include <wchar.h>
++# endif
++
+ /* A `struct linebuffer' holds a line of text. */
+
+ struct linebuffer
+@@ -29,6 +34,9 @@
+ size_t size; /* Allocated. */
+ size_t length; /* Used. */
+ char *buffer;
++# if HAVE_WCHAR_H
++ mbstate_t state;
++# endif
+ };
+
+ /* Initialize linebuffer LINEBUFFER for use. */
+--- coreutils-6.8+/src/expand.c.i18n 2007-01-14 15:41:28.000000000 +0000
++++ coreutils-6.8+/src/expand.c 2007-03-01 15:08:24.000000000 +0000
+@@ -38,11 +38,28 @@
+ #include <stdio.h>
+ #include <getopt.h>
+ #include <sys/types.h>
++
++/* Get mbstate_t, mbrtowc(), wcwidth(). */
++#if HAVE_WCHAR_H
++# include <wchar.h>
++#endif
++
+ #include "system.h"
+ #include "error.h"
+ #include "quote.h"
+ #include "xstrndup.h"
+
++/* MB_LEN_MAX is incorrectly defined to be 1 in at least one GCC
++ installation; work around this configuration error. */
++#if !defined MB_LEN_MAX || MB_LEN_MAX < 2
++# define MB_LEN_MAX 16
++#endif
++
++/* Some systems, like BeOS, have multibyte encodings but lack mbstate_t. */
++#if HAVE_MBRTOWC && defined mbstate_t
++# define mbrtowc(pwc, s, n, ps) (mbrtowc) (pwc, s, n, 0)
++#endif
++
+ /* The official name of this program (e.g., no `g' prefix). */
+ #define PROGRAM_NAME "expand"
+
+@@ -183,6 +200,7 @@
+ stops = num_start + len - 1;
+ }
+ }
++
+ else
+ {
+ error (0, 0, _("tab size contains invalid character(s): %s"),
+@@ -365,6 +383,142 @@
+ }
+ }
+
++#if HAVE_MBRTOWC
++static void
++expand_multibyte (void)
++{
++ FILE *fp; /* Input strem. */
++ mbstate_t i_state; /* Current shift state of the input stream. */
++ mbstate_t i_state_bak; /* Back up the I_STATE. */
++ mbstate_t o_state; /* Current shift state of the output stream. */
++ char buf[MB_LEN_MAX + BUFSIZ]; /* For spooling a read byte sequence. */
++ char *bufpos; /* Next read position of BUF. */
++ size_t buflen = 0; /* The length of the byte sequence in buf. */
++ wchar_t wc; /* A gotten wide character. */
++ size_t mblength; /* The byte size of a multibyte character
++ which shows as same character as WC. */
++ int tab_index = 0; /* Index in `tab_list' of next tabstop. */
++ int column = 0; /* Column on screen of the next char. */
++ int next_tab_column; /* Column the next tab stop is on. */
++ int convert = 1; /* If nonzero, perform translations. */
++
++ fp = next_file ((FILE *) NULL);
++ if (fp == NULL)
++ return;
++
++ memset (&o_state, '\0', sizeof(mbstate_t));
++ memset (&i_state, '\0', sizeof(mbstate_t));
++
++ for (;;)
++ {
++ /* Refill the buffer BUF. */
++ if (buflen < MB_LEN_MAX && !feof(fp) && !ferror(fp))
++ {
++ memmove (buf, bufpos, buflen);
++ buflen += fread (buf + buflen, sizeof(char), BUFSIZ, fp);
++ bufpos = buf;
++ }
++
++ /* No character is left in BUF. */
++ if (buflen < 1)
++ {
++ fp = next_file (fp);
++
++ if (fp == NULL)
++ break; /* No more files. */
++ else
++ {
++ memset (&i_state, '\0', sizeof(mbstate_t));
++ continue;
++ }
++ }
++
++ /* Get a wide character. */
++ i_state_bak = i_state;
++ mblength = mbrtowc (&wc, bufpos, buflen, &i_state);
++
++ switch (mblength)
++ {
++ case (size_t)-1: /* illegal byte sequence. */
++ case (size_t)-2:
++ mblength = 1;
++ i_state = i_state_bak;
++ if (convert)
++ {
++ ++column;
++ if (convert_entire_line == 0)
++ convert = 0;
++ }
++ putchar (*bufpos);
++ break;
++
++ case 0: /* null. */
++ mblength = 1;
++ if (convert && convert_entire_line == 0)
++ convert = 0;
++ putchar ('\0');
++ break;
++
++ default:
++ if (wc == L'\n') /* LF. */
++ {
++ tab_index = 0;
++ column = 0;
++ convert = 1;
++ putchar ('\n');
++ }
++ else if (wc == L'\t' && convert) /* Tab. */
++ {
++ if (tab_size == 0)
++ {
++ /* Do not let tab_index == first_free_tab;
++ stop when it is 1 less. */
++ while (tab_index < first_free_tab - 1
++ && column >= tab_list[tab_index])
++ tab_index++;
++ next_tab_column = tab_list[tab_index];
++ if (tab_index < first_free_tab - 1)
++ tab_index++;
++ if (column >= next_tab_column)
++ next_tab_column = column + 1;
++ }
++ else
++ next_tab_column = column + tab_size - column % tab_size;
++
++ while (column < next_tab_column)
++ {
++ putchar (' ');
++ ++column;
++ }
++ }
++ else /* Others. */
++ {
++ if (convert)
++ {
++ if (wc == L'\b')
++ {
++ if (column > 0)
++ --column;
++ }
++ else
++ {
++ int width; /* The width of WC. */
++
++ width = wcwidth (wc);
++ column += (width > 0) ? width : 0;
++ if (convert_entire_line == 0)
++ convert = 0;
++ }
++ }
++ fwrite (bufpos, sizeof(char), mblength, stdout);
++ }
++ }
++ buflen -= mblength;
++ bufpos += mblength;
++ }
++}
++#endif
++
+ int
+ main (int argc, char **argv)
+ {
+@@ -429,7 +583,12 @@
+
+ file_list = (optind < argc ? &argv[optind] : stdin_argv);
+
+- expand ();
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ expand_multibyte ();
++ else
++#endif
++ expand ();
+
+ if (have_read_stdin && fclose (stdin) != 0)
+ error (EXIT_FAILURE, errno, "-");
+--- coreutils-6.8+/src/join.c.i18n 2007-01-14 15:41:28.000000000 +0000
++++ coreutils-6.8+/src/join.c 2007-03-01 15:08:24.000000000 +0000
+@@ -23,16 +23,30 @@
+ #include <sys/types.h>
+ #include <getopt.h>
+
++/* Get mbstate_t, mbrtowc(), mbrtowc(), wcwidth(). */
++#if HAVE_WCHAR_H
++# include <wchar.h>
++#endif
++
++/* Get iswblank(), towupper. */
++#if HAVE_WCTYPE_H
++# include <wctype.h>
++#endif
++
+ #include "system.h"
+ #include "error.h"
+ #include "hard-locale.h"
+ #include "linebuffer.h"
+-#include "memcasecmp.h"
+ #include "quote.h"
+ #include "stdio--.h"
+ #include "xmemcoll.h"
+ #include "xstrtol.h"
+
++/* Some systems, like BeOS, have multibyte encodings but lack mbstate_t. */
++#if HAVE_MBRTOWC && defined mbstate_t
++# define mbrtowc(pwc, s, n, ps) (mbrtowc) (pwc, s, n, 0)
++#endif
++
+ /* The official name of this program (e.g., no `g' prefix). */
+ #define PROGRAM_NAME "join"
+
+@@ -104,10 +118,12 @@
+ /* Last element in `outlist', where a new element can be added. */
+ static struct outlist *outlist_end = &outlist_head;
+
+-/* Tab character separating fields. If negative, fields are separated
+- by any nonempty string of blanks, otherwise by exactly one
+- tab character whose value (when cast to unsigned char) equals TAB. */
+-static int tab = -1;
++/* Tab character separating fields. If NULL, fields are separated
++ by any nonempty string of blanks. */
++static char *tab = NULL;
++
++/* The number of bytes used for tab. */
++static size_t tablen = 0;
+
+ static struct option const longopts[] =
+ {
+@@ -190,6 +206,8 @@
+
+ /* Fill in the `fields' structure in LINE. */
+
++/* Fill in the `fields' structure in LINE. */
++
+ static void
+ xfields (struct line *line)
+ {
+@@ -199,10 +217,11 @@
+ if (ptr == lim)
+ return;
+
+- if (0 <= tab)
++ if (tab != NULL)
+ {
++ unsigned char t = tab[0];
+ char *sep;
+- for (; (sep = memchr (ptr, tab, lim - ptr)) != NULL; ptr = sep + 1)
++ for (; (sep = memchr (ptr, t, lim - ptr)) != NULL; ptr = sep + 1)
+ extract_field (line, ptr, sep - ptr);
+ }
+ else
+@@ -229,6 +248,148 @@
+ extract_field (line, ptr, lim - ptr);
+ }
+
++#if HAVE_MBRTOWC
++static void
++xfields_multibyte (struct line *line)
++{
++ char *ptr = line->buf.buffer;
++ char const *lim = ptr + line->buf.length - 1;
++ wchar_t wc = 0;
++ size_t mblength = 1;
++ mbstate_t state, state_bak;
++
++ memset (&state, 0, sizeof (mbstate_t));
++
++ if (ptr == lim)
++ return;
++
++ if (tab != NULL)
++ {
++ unsigned char t = tab[0];
++ char *sep = ptr;
++ for (; ptr < lim; ptr = sep + mblength)
++ {
++ sep = ptr;
++ while (sep < lim)
++ {
++ state_bak = state;
++ mblength = mbrtowc (&wc, sep, lim - sep + 1, &state);
++
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ mblength = 1;
++ state = state_bak;
++ }
++ mblength = (mblength < 1) ? 1 : mblength;
++
++ if (mblength == tablen && !memcmp (sep, tab, mblength))
++ break;
++ else
++ {
++ sep += mblength;
++ continue;
++ }
++ }
++
++ if (sep == lim)
++ break;
++
++ extract_field (line, ptr, sep - ptr);
++ }
++ }
++ else
++ {
++ /* Skip leading blanks before the first field. */
++ while(ptr < lim)
++ {
++ state_bak = state;
++ mblength = mbrtowc (&wc, ptr, lim - ptr + 1, &state);
++
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ mblength = 1;
++ state = state_bak;
++ break;
++ }
++ mblength = (mblength < 1) ? 1 : mblength;
++
++ if (!iswblank(wc))
++ break;
++ ptr += mblength;
++ }
++
++ do
++ {
++ char *sep;
++ state_bak = state;
++ mblength = mbrtowc (&wc, ptr, lim - ptr + 1, &state);
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ mblength = 1;
++ state = state_bak;
++ break;
++ }
++ mblength = (mblength < 1) ? 1 : mblength;
++
++ sep = ptr + mblength;
++ while (sep != lim)
++ {
++ state_bak = state;
++ mblength = mbrtowc (&wc, sep, lim - sep + 1, &state);
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ mblength = 1;
++ state = state_bak;
++ break;
++ }
++ mblength = (mblength < 1) ? 1 : mblength;
++
++ if (iswblank (wc))
++ break;
++
++ sep += mblength;
++ }
++
++ extract_field (line, ptr, sep - ptr);
++ if (sep == lim)
++ return;
++
++ state_bak = state;
++ mblength = mbrtowc (&wc, sep, lim - sep + 1, &state);
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ mblength = 1;
++ state = state_bak;
++ break;
++ }
++ mblength = (mblength < 1) ? 1 : mblength;
++
++ ptr = sep + mblength;
++ while (ptr != lim)
++ {
++ state_bak = state;
++ mblength = mbrtowc (&wc, ptr, lim - ptr + 1, &state);
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ mblength = 1;
++ state = state_bak;
++ break;
++ }
++ mblength = (mblength < 1) ? 1 : mblength;
++
++ if (!iswblank (wc))
++ break;
++
++ ptr += mblength;
++ }
++ }
++ while (ptr != lim);
++ }
++
++ extract_field (line, ptr, lim - ptr);
++}
++#endif
++
+ /* Read a line from FP into LINE and split it into fields.
+ Return true if successful. */
+
+@@ -249,6 +410,11 @@
+ line->nfields_allocated = 0;
+ line->nfields = 0;
+ line->fields = NULL;
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ xfields_multibyte (line);
++ else
++#endif
+ xfields (line);
+ return true;
+ }
+@@ -303,56 +469,114 @@
+ keycmp (struct line const *line1, struct line const *line2)
+ {
+ /* Start of field to compare in each file. */
+- char *beg1;
+- char *beg2;
+-
+- size_t len1;
+- size_t len2; /* Length of fields to compare. */
++ char *beg[2];
++ char *copy[2];
++ size_t len[2]; /* Length of fields to compare. */
+ int diff;
++ int i, j;
+
+ if (join_field_1 < line1->nfields)
+ {
+- beg1 = line1->fields[join_field_1].beg;
+- len1 = line1->fields[join_field_1].len;
++ beg[0] = line1->fields[join_field_1].beg;
++ len[0] = line1->fields[join_field_1].len;
+ }
+ else
+ {
+- beg1 = NULL;
+- len1 = 0;
++ beg[0] = NULL;
++ len[0] = 0;
+ }
+
+ if (join_field_2 < line2->nfields)
+ {
+- beg2 = line2->fields[join_field_2].beg;
+- len2 = line2->fields[join_field_2].len;
++ beg[1] = line2->fields[join_field_2].beg;
++ len[1] = line2->fields[join_field_2].len;
+ }
+ else
+ {
+- beg2 = NULL;
+- len2 = 0;
++ beg[1] = NULL;
++ len[1] = 0;
+ }
+
+- if (len1 == 0)
+- return len2 == 0 ? 0 : -1;
+- if (len2 == 0)
++ if (len[0] == 0)
++ return len[1] == 0 ? 0 : -1;
++ if (len[1] == 0)
+ return 1;
+
+ if (ignore_case)
+ {
+- /* FIXME: ignore_case does not work with NLS (in particular,
+- with multibyte chars). */
+- diff = memcasecmp (beg1, beg2, MIN (len1, len2));
++#ifdef HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ size_t mblength;
++ wchar_t wc, uwc;
++ mbstate_t state, state_bak;
++
++ memset (&state, '\0', sizeof (mbstate_t));
++
++ for (i = 0; i < 2; i++)
++ {
++ copy[i] = alloca (len[i] + 1);
++
++ for (j = 0; j < MIN (len[0], len[1]);)
++ {
++ state_bak = state;
++ mblength = mbrtowc (&wc, beg[i] + j, len[i] - j, &state);
++
++ switch (mblength)
++ {
++ case (size_t) -1:
++ case (size_t) -2:
++ state = state_bak;
++ /* Fall through */
++ case 0:
++ mblength = 1;
++ break;
++
++ default:
++ uwc = towupper (wc);
++
++ if (uwc != wc)
++ {
++ mbstate_t state_wc;
++
++ memset (&state_wc, '\0', sizeof (mbstate_t));
++ wcrtomb (copy[i] + j, uwc, &state_wc);
++ }
++ else
++ memcpy (copy[i] + j, beg[i] + j, mblength);
++ }
++ j += mblength;
++ }
++ copy[i][j] = '\0';
++ }
++ }
++ else
++#endif
++ {
++ for (i = 0; i < 2; i++)
++ {
++ copy[i] = alloca (len[i] + 1);
++
++ for (j = 0; j < MIN (len[0], len[1]); j++)
++ copy[i][j] = toupper (beg[i][j]);
++
++ copy[i][j] = '\0';
++ }
++ }
+ }
+ else
+ {
+- if (hard_LC_COLLATE)
+- return xmemcoll (beg1, len1, beg2, len2);
+- diff = memcmp (beg1, beg2, MIN (len1, len2));
++ copy[0] = (unsigned char *) beg[0];
++ copy[1] = (unsigned char *) beg[1];
+ }
+
++ if (hard_LC_COLLATE)
++ return xmemcoll ((char *) copy[0], len[0], (char *) copy[1], len[1]);
++ diff = memcmp (copy[0], copy[1], MIN (len[0], len[1]));
++
+ if (diff)
+ return diff;
+- return len1 < len2 ? -1 : len1 != len2;
++ return len[0] - len[1];
+ }
+
+ /* Print field N of LINE if it exists and is nonempty, otherwise
+@@ -377,11 +601,18 @@
+
+ /* Print the join of LINE1 and LINE2. */
+
++#define PUT_TAB_CHAR \
++ do \
++ { \
++ (tab != NULL) ? \
++ fwrite(tab, sizeof(char), tablen, stdout) : putchar (' '); \
++ } \
++ while (0)
++
+ static void
+ prjoin (struct line const *line1, struct line const *line2)
+ {
+ const struct outlist *outlist;
+- char output_separator = tab < 0 ? ' ' : tab;
+
+ outlist = outlist_head.next;
+ if (outlist)
+@@ -397,12 +628,12 @@
+ if (o->file == 0)
+ {
+ if (line1 == &uni_blank)
+- {
++ {
+ line = line2;
+ field = join_field_2;
+ }
+ else
+- {
++ {
+ line = line1;
+ field = join_field_1;
+ }
+@@ -416,7 +647,7 @@
+ o = o->next;
+ if (o == NULL)
+ break;
+- putchar (output_separator);
++ PUT_TAB_CHAR;
+ }
+ putchar ('\n');
+ }
+@@ -434,23 +665,23 @@
+ prfield (join_field_1, line1);
+ for (i = 0; i < join_field_1 && i < line1->nfields; ++i)
+ {
+- putchar (output_separator);
++ PUT_TAB_CHAR;
+ prfield (i, line1);
+ }
+ for (i = join_field_1 + 1; i < line1->nfields; ++i)
+ {
+- putchar (output_separator);
++ PUT_TAB_CHAR;
+ prfield (i, line1);
+ }
+
+ for (i = 0; i < join_field_2 && i < line2->nfields; ++i)
+ {
+- putchar (output_separator);
++ PUT_TAB_CHAR;
+ prfield (i, line2);
+ }
+ for (i = join_field_2 + 1; i < line2->nfields; ++i)
+ {
+- putchar (output_separator);
++ PUT_TAB_CHAR;
+ prfield (i, line2);
+ }
+ putchar ('\n');
+@@ -859,20 +1090,41 @@
+
+ case 't':
+ {
+- unsigned char newtab = optarg[0];
+- if (! newtab)
++ char *newtab;
++ size_t newtablen;
++ if (! optarg[0])
+ error (EXIT_FAILURE, 0, _("empty tab"));
+- if (optarg[1])
++ newtab = xstrdup (optarg);
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ mbstate_t state;
++
++ memset (&state, 0, sizeof (mbstate_t));
++ newtablen = mbrtowc (NULL, newtab,
++ strnlen (newtab, MB_LEN_MAX),
++ &state);
++ if (newtablen == (size_t) 0
++ || newtablen == (size_t) -1
++ || newtablen == (size_t) -2)
++ newtablen = 1;
++ }
++ else
++#endif
++ newtablen = 1;
++
++ if (newtablen == 1 && newtab[1])
++ {
++ if (STREQ (newtab, "\\0"))
++ newtab[0] = '\0';
++ }
++ if (tab != NULL && strcmp (tab, newtab))
+ {
+- if (STREQ (optarg, "\\0"))
+- newtab = '\0';
+- else
+- error (EXIT_FAILURE, 0, _("multi-character tab %s"),
+- quote (optarg));
++ free (newtab);
++ error (EXIT_FAILURE, 0, _("incompatible tabs"));
+ }
+- if (0 <= tab && tab != newtab)
+- error (EXIT_FAILURE, 0, _("incompatible tabs"));
+ tab = newtab;
++ tablen = newtablen;
+ }
+ break;
+
+--- coreutils-6.8+/src/uniq.c.i18n 2007-01-14 15:41:28.000000000 +0000
++++ coreutils-6.8+/src/uniq.c 2007-03-01 15:08:24.000000000 +0000
+@@ -23,6 +23,16 @@
+ #include <getopt.h>
+ #include <sys/types.h>
+
++/* Get mbstate_t, mbrtowc(). */
++#if HAVE_WCHAR_H
++# include <wchar.h>
++#endif
++
++/* Get isw* functions. */
++#if HAVE_WCTYPE_H
++# include <wctype.h>
++#endif
++
+ #include "system.h"
+ #include "argmatch.h"
+ #include "linebuffer.h"
+@@ -32,7 +42,19 @@
+ #include "quote.h"
+ #include "xmemcoll.h"
+ #include "xstrtol.h"
+-#include "memcasecmp.h"
++#include "xmemcoll.h"
++
++/* MB_LEN_MAX is incorrectly defined to be 1 in at least one GCC
++ installation; work around this configuration error. */
++#if !defined MB_LEN_MAX || MB_LEN_MAX < 2
++# define MB_LEN_MAX 16
++#endif
++
++/* Some systems, like BeOS, have multibyte encodings but lack mbstate_t. */
++#if HAVE_MBRTOWC && defined mbstate_t
++# define mbrtowc(pwc, s, n, ps) (mbrtowc) (pwc, s, n, 0)
++#endif
++
+
+ /* The official name of this program (e.g., no `g' prefix). */
+ #define PROGRAM_NAME "uniq"
+@@ -109,6 +131,10 @@
+ /* Select whether/how to delimit groups of duplicate lines. */
+ static enum delimit_method delimit_groups;
+
++/* Function pointers. */
++static char *
++(*find_field) (struct linebuffer *line);
++
+ static struct option const longopts[] =
+ {
+ {"count", no_argument, NULL, 'c'},
+@@ -198,7 +224,7 @@
+ return a pointer to the beginning of the line's field to be compared. */
+
+ static char *
+-find_field (const struct linebuffer *line)
++find_field_uni (struct linebuffer *line)
+ {
+ size_t count;
+ char *lp = line->buffer;
+@@ -219,6 +245,83 @@
+ return lp + i;
+ }
+
++#if HAVE_MBRTOWC
++
++# define MBCHAR_TO_WCHAR(WC, MBLENGTH, LP, POS, SIZE, STATEP, CONVFAIL) \
++ do \
++ { \
++ mbstate_t state_bak; \
++ \
++ CONVFAIL = 0; \
++ state_bak = *STATEP; \
++ \
++ MBLENGTH = mbrtowc (&WC, LP + POS, SIZE - POS, STATEP); \
++ \
++ switch (MBLENGTH) \
++ { \
++ case (size_t)-2: \
++ case (size_t)-1: \
++ *STATEP = state_bak; \
++ CONVFAIL++; \
++ /* Fall through */ \
++ case 0: \
++ MBLENGTH = 1; \
++ } \
++ } \
++ while (0)
++
++static char *
++find_field_multi (struct linebuffer *line)
++{
++ size_t count;
++ char *lp = line->buffer;
++ size_t size = line->length - 1;
++ size_t pos;
++ size_t mblength;
++ wchar_t wc;
++ mbstate_t *statep;
++ int convfail;
++
++ pos = 0;
++ statep = &(line->state);
++
++ /* skip fields. */
++ for (count = 0; count < skip_fields && pos < size; count++)
++ {
++ while (pos < size)
++ {
++ MBCHAR_TO_WCHAR (wc, mblength, lp, pos, size, statep, convfail);
++
++ if (convfail || !iswblank (wc))
++ {
++ pos += mblength;
++ break;
++ }
++ pos += mblength;
++ }
++
++ while (pos < size)
++ {
++ MBCHAR_TO_WCHAR (wc, mblength, lp, pos, size, statep, convfail);
++
++ if (!convfail && iswblank (wc))
++ break;
++
++ pos += mblength;
++ }
++ }
++
++ /* skip fields. */
++ for (count = 0; count < skip_chars && pos < size; count++)
++ {
++ MBCHAR_TO_WCHAR (wc, mblength, lp, pos, size, statep, convfail);
++ pos += mblength;
++ }
++
++ return lp + pos;
++}
++#endif
++
+ /* Return false if two strings OLD and NEW match, true if not.
+ OLD and NEW point not to the beginnings of the lines
+ but rather to the beginnings of the fields to compare.
+@@ -227,6 +330,8 @@
+ static bool
+ different (char *old, char *new, size_t oldlen, size_t newlen)
+ {
++ char *copy_old, *copy_new;
++
+ if (check_chars < oldlen)
+ oldlen = check_chars;
+ if (check_chars < newlen)
+@@ -234,14 +339,92 @@
+
+ if (ignore_case)
+ {
+- /* FIXME: This should invoke strcoll somehow. */
+- return oldlen != newlen || memcasecmp (old, new, oldlen);
++ size_t i;
++
++ copy_old = alloca (oldlen + 1);
++ copy_new = alloca (oldlen + 1);
++
++ for (i = 0; i < oldlen; i++)
++ {
++ copy_old[i] = toupper (old[i]);
++ copy_new[i] = toupper (new[i]);
++ }
+ }
+- else if (hard_LC_COLLATE)
+- return xmemcoll (old, oldlen, new, newlen) != 0;
+ else
+- return oldlen != newlen || memcmp (old, new, oldlen);
++ {
++ copy_old = (char *)old;
++ copy_new = (char *)new;
++ }
++
++ return xmemcoll (copy_old, oldlen, copy_new, newlen);
++}
++
++#if HAVE_MBRTOWC
++static int
++different_multi (const char *old, const char *new, size_t oldlen, size_t newlen, mbstate_t oldstate, mbstate_t newstate)
++{
++ size_t i, j, chars;
++ const char *str[2];
++ char *copy[2];
++ size_t len[2];
++ mbstate_t state[2];
++ size_t mblength;
++ wchar_t wc, uwc;
++ mbstate_t state_bak;
++
++ str[0] = old;
++ str[1] = new;
++ len[0] = oldlen;
++ len[1] = newlen;
++ state[0] = oldstate;
++ state[1] = newstate;
++
++ for (i = 0; i < 2; i++)
++ {
++ copy[i] = alloca (len[i] + 1);
++
++ for (j = 0, chars = 0; j < len[i] && chars < check_chars; chars++)
++ {
++ state_bak = state[i];
++ mblength = mbrtowc (&wc, str[i] + j, len[i] - j, &(state[i]));
++
++ switch (mblength)
++ {
++ case (size_t)-1:
++ case (size_t)-2:
++ state[i] = state_bak;
++ /* Fall through */
++ case 0:
++ mblength = 1;
++ break;
++
++ default:
++ if (ignore_case)
++ {
++ uwc = towupper (wc);
++
++ if (uwc != wc)
++ {
++ mbstate_t state_wc;
++
++ memset (&state_wc, '\0', sizeof(mbstate_t));
++ wcrtomb (copy[i] + j, uwc, &state_wc);
++ }
++ else
++ memcpy (copy[i] + j, str[i] + j, mblength);
++ }
++ else
++ memcpy (copy[i] + j, str[i] + j, mblength);
++ }
++ j += mblength;
++ }
++ copy[i][j] = '\0';
++ len[i] = j;
++ }
++
++ return xmemcoll (copy[0], len[0], copy[1], len[1]);
+ }
++#endif
+
+ /* Output the line in linebuffer LINE to standard output
+ provided that the switches say it should be output.
+@@ -295,15 +478,43 @@
+ {
+ char *prevfield IF_LINT (= NULL);
+ size_t prevlen IF_LINT (= 0);
++#if HAVE_MBRTOWC
++ mbstate_t prevstate;
++
++ memset (&prevstate, '\0', sizeof (mbstate_t));
++#endif
+
+ while (!feof (stdin))
+ {
+ char *thisfield;
+ size_t thislen;
++#if HAVE_MBRTOWC
++ mbstate_t thisstate;
++#endif
++
+ if (readlinebuffer (thisline, stdin) == 0)
+ break;
+ thisfield = find_field (thisline);
+ thislen = thisline->length - 1 - (thisfield - thisline->buffer);
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ thisstate = thisline->state;
++
++ if (prevline->length == 0 || different_multi
++ (thisfield, prevfield, thislen, prevlen, thisstate, prevstate))
++ {
++ fwrite (thisline->buffer, sizeof (char),
++ thisline->length, stdout);
++
++ SWAP_LINES (prevline, thisline);
++ prevfield = thisfield;
++ prevlen = thislen;
++ prevstate = thisstate;
++ }
++ }
++ else
++#endif
+ if (prevline->length == 0
+ || different (thisfield, prevfield, thislen, prevlen))
+ {
+@@ -322,17 +533,26 @@
+ size_t prevlen;
+ uintmax_t match_count = 0;
+ bool first_delimiter = true;
++#if HAVE_MBRTOWC
++ mbstate_t prevstate;
++#endif
+
+ if (readlinebuffer (prevline, stdin) == 0)
+ goto closefiles;
+ prevfield = find_field (prevline);
+ prevlen = prevline->length - 1 - (prevfield - prevline->buffer);
++#if HAVE_MBRTOWC
++ prevstate = prevline->state;
++#endif
+
+ while (!feof (stdin))
+ {
+ bool match;
+ char *thisfield;
+ size_t thislen;
++#if HAVE_MBRTOWC
++ mbstate_t thisstate;
++#endif
+ if (readlinebuffer (thisline, stdin) == 0)
+ {
+ if (ferror (stdin))
+@@ -341,6 +561,15 @@
+ }
+ thisfield = find_field (thisline);
+ thislen = thisline->length - 1 - (thisfield - thisline->buffer);
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ thisstate = thisline->state;
++ match = !different_multi (thisfield, prevfield,
++ thislen, prevlen, thisstate, prevstate);
++ }
++ else
++#endif
+ match = !different (thisfield, prevfield, thislen, prevlen);
+ match_count += match;
+
+@@ -373,6 +602,9 @@
+ SWAP_LINES (prevline, thisline);
+ prevfield = thisfield;
+ prevlen = thislen;
++#if HAVE_MBRTOWC
++ prevstate = thisstate;
++#endif
+ if (!match)
+ match_count = 0;
+ }
+@@ -417,6 +649,19 @@
+
+ atexit (close_stdout);
+
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ find_field = find_field_multi;
++ }
++ else
++#endif
++ {
++ find_field = find_field_uni;
++ }
++
++
++
+ skip_chars = 0;
+ skip_fields = 0;
+ check_chars = SIZE_MAX;
+--- coreutils-6.8+/src/fold.c.i18n 2007-02-23 12:01:47.000000000 +0000
++++ coreutils-6.8+/src/fold.c 2007-03-01 15:08:24.000000000 +0000
+@@ -23,11 +23,33 @@
+ #include <getopt.h>
+ #include <sys/types.h>
+
++/* Get mbstate_t, mbrtowc(), wcwidth(). */
++#if HAVE_WCHAR_H
++# include <wchar.h>
++#endif
++
++/* Get iswprint(), iswblank(), wcwidth(). */
++#if HAVE_WCTYPE_H
++# include <wctype.h>
++#endif
++
+ #include "system.h"
+ #include "error.h"
+ #include "quote.h"
+ #include "xstrtol.h"
+
++/* MB_LEN_MAX is incorrectly defined to be 1 in at least one GCC
++ installation; work around this configuration error. */
++#if !defined MB_LEN_MAX || MB_LEN_MAX < 2
++# undef MB_LEN_MAX
++# define MB_LEN_MAX 16
++#endif
++
++/* Some systems, like BeOS, have multibyte encodings but lack mbstate_t. */
++#if HAVE_MBRTOWC && defined mbstate_t
++# define mbrtowc(pwc, s, n, ps) (mbrtowc) (pwc, s, n, 0)
++#endif
++
+ #define TAB_WIDTH 8
+
+ /* The official name of this program (e.g., no `g' prefix). */
+@@ -35,23 +57,44 @@
+
+ #define AUTHORS "David MacKenzie"
+
++#define FATAL_ERROR(Message) \
++ do \
++ { \
++ error (0, 0, (Message)); \
++ usage (2); \
++ } \
++ while (0)
++
++enum operating_mode
++{
++ /* Fold texts by columns that are at the given positions. */
++ column_mode,
++
++ /* Fold texts by bytes that are at the given positions. */
++ byte_mode,
++
++ /* Fold texts by characters that are at the given positions. */
++ character_mode,
++};
++
+ /* The name this program was run with. */
+ char *program_name;
+
++/* The argument shows current mode. (Default: column_mode) */
++static enum operating_mode operating_mode;
++
+ /* If nonzero, try to break on whitespace. */
+ static bool break_spaces;
+
+-/* If nonzero, count bytes, not column positions. */
+-static bool count_bytes;
+-
+ /* If nonzero, at least one of the files we read was standard input. */
+ static bool have_read_stdin;
+
+-static char const shortopts[] = "bsw:0::1::2::3::4::5::6::7::8::9::";
++static char const shortopts[] = "bcsw:0::1::2::3::4::5::6::7::8::9::";
+
+ static struct option const longopts[] =
+ {
+ {"bytes", no_argument, NULL, 'b'},
++ {"characters", no_argument, NULL, 'c'},
+ {"spaces", no_argument, NULL, 's'},
+ {"width", required_argument, NULL, 'w'},
+ {GETOPT_HELP_OPTION_DECL},
+@@ -81,6 +124,7 @@
+ "), stdout);
+ fputs (_("\
+ -b, --bytes count bytes rather than columns\n\
++ -c, --characters count characters rather than columns\n\
+ -s, --spaces break at spaces\n\
+ -w, --width=WIDTH use WIDTH columns instead of 80\n\
+ "), stdout);
+@@ -98,7 +142,7 @@
+ static size_t
+ adjust_column (size_t column, char c)
+ {
+- if (!count_bytes)
++ if (operating_mode != byte_mode)
+ {
+ if (c == '\b')
+ {
+@@ -121,30 +165,14 @@
+ to stdout, with maximum line length WIDTH.
+ Return true if successful. */
+
+-static bool
+-fold_file (char const *filename, size_t width)
++static void
++fold_text (FILE *istream, size_t width, int *saved_errno)
+ {
+- FILE *istream;
+ int c;
+ size_t column = 0; /* Screen column where next char will go. */
+ size_t offset_out = 0; /* Index in `line_out' for next char. */
+ static char *line_out = NULL;
+ static size_t allocated_out = 0;
+- int saved_errno;
+-
+- if (STREQ (filename, "-"))
+- {
+- istream = stdin;
+- have_read_stdin = true;
+- }
+- else
+- istream = fopen (filename, "r");
+-
+- if (istream == NULL)
+- {
+- error (0, errno, "%s", filename);
+- return false;
+- }
+
+ while ((c = getc (istream)) != EOF)
+ {
+@@ -172,6 +200,15 @@
+ bool found_blank = false;
+ size_t logical_end = offset_out;
+
++ /* If LINE_OUT has no wide character,
++ put a new wide character in LINE_OUT
++ if column is bigger than width. */
++ if (offset_out == 0)
++ {
++ line_out[offset_out++] = c;
++ continue;
++ }
++
+ /* Look for the last blank. */
+ while (logical_end)
+ {
+@@ -218,11 +255,225 @@
+ line_out[offset_out++] = c;
+ }
+
+- saved_errno = errno;
++ *saved_errno = errno;
++
++ if (offset_out)
++ fwrite (line_out, sizeof (char), (size_t) offset_out, stdout);
++
++ free(line_out);
++}
++
++#if HAVE_MBRTOWC
++static void
++fold_multibyte_text (FILE *istream, size_t width, int *saved_errno)
++{
++ char buf[MB_LEN_MAX + BUFSIZ]; /* For spooling a read byte sequence. */
++ size_t buflen = 0; /* The length of the byte sequence in buf. */
++ char *bufpos; /* Next read position of BUF. */
++ wint_t wc; /* A gotten wide character. */
++ size_t mblength; /* The byte size of a multibyte character which shows
++ as same character as WC. */
++ mbstate_t state, state_bak; /* State of the stream. */
++ int convfail; /* 1, when conversion is failed. Otherwise 0. */
++
++ char *line_out = NULL;
++ size_t offset_out = 0; /* Index in `line_out' for next char. */
++ size_t allocated_out = 0;
++
++ int increment;
++ size_t column = 0;
++
++ size_t last_blank_pos;
++ size_t last_blank_column;
++ int is_blank_seen;
++ int last_blank_increment;
++ int is_bs_following_last_blank;
++ size_t bs_following_last_blank_num;
++ int is_cr_after_last_blank;
++
++#define CLEAR_FLAGS \
++ do \
++ { \
++ last_blank_pos = 0; \
++ last_blank_column = 0; \
++ is_blank_seen = 0; \
++ is_bs_following_last_blank = 0; \
++ bs_following_last_blank_num = 0; \
++ is_cr_after_last_blank = 0; \
++ } \
++ while (0)
++
++#define START_NEW_LINE \
++ do \
++ { \
++ putchar ('\n'); \
++ column = 0; \
++ offset_out = 0; \
++ CLEAR_FLAGS; \
++ } \
++ while (0)
++
++ CLEAR_FLAGS;
++ memset (&state, '\0', sizeof(mbstate_t));
++
++ for (;; bufpos += mblength, buflen -= mblength)
++ {
++ if (buflen < MB_LEN_MAX && !feof (istream) && !ferror (istream))
++ {
++ memmove (buf, bufpos, buflen);
++ buflen += fread (buf + buflen, sizeof(char), BUFSIZ, istream);
++ bufpos = buf;
++ }
++
++ if (buflen < 1)
++ break;
++
++ /* Get a wide character. */
++ convfail = 0;
++ state_bak = state;
++ mblength = mbrtowc ((wchar_t *)&wc, bufpos, buflen, &state);
++
++ switch (mblength)
++ {
++ case (size_t)-1:
++ case (size_t)-2:
++ convfail++;
++ state = state_bak;
++ /* Fall through. */
++
++ case 0:
++ mblength = 1;
++ break;
++ }
++
++rescan:
++ if (operating_mode == byte_mode) /* byte mode */
++ increment = mblength;
++ else if (operating_mode == character_mode) /* character mode */
++ increment = 1;
++ else /* column mode */
++ {
++ if (convfail)
++ increment = 1;
++ else
++ {
++ switch (wc)
++ {
++ case L'\n':
++ fwrite (line_out, sizeof(char), offset_out, stdout);
++ START_NEW_LINE;
++ continue;
++
++ case L'\b':
++ increment = (column > 0) ? -1 : 0;
++ break;
++
++ case L'\r':
++ increment = -1 * column;
++ break;
++
++ case L'\t':
++ increment = 8 - column % 8;
++ break;
++
++ default:
++ increment = wcwidth (wc);
++ increment = (increment < 0) ? 0 : increment;
++ }
++ }
++ }
++
++ if (column + increment > width && break_spaces && last_blank_pos)
++ {
++ fwrite (line_out, sizeof(char), last_blank_pos, stdout);
++ putchar ('\n');
++
++ offset_out = offset_out - last_blank_pos;
++ column = column - last_blank_column + ((is_cr_after_last_blank)
++ ? last_blank_increment : bs_following_last_blank_num);
++ memmove (line_out, line_out + last_blank_pos, offset_out);
++ CLEAR_FLAGS;
++ goto rescan;
++ }
++
++ if (column + increment > width && column != 0)
++ {
++ fwrite (line_out, sizeof(char), offset_out, stdout);
++ START_NEW_LINE;
++ goto rescan;
++ }
++
++ if (allocated_out < offset_out + mblength)
++ {
++ allocated_out += 1024;
++ line_out = xrealloc (line_out, allocated_out);
++ }
++
++ memcpy (line_out + offset_out, bufpos, mblength);
++ offset_out += mblength;
++ column += increment;
++
++ if (is_blank_seen && !convfail && wc == L'\r')
++ is_cr_after_last_blank = 1;
++
++ if (is_bs_following_last_blank && !convfail && wc == L'\b')
++ ++bs_following_last_blank_num;
++ else
++ is_bs_following_last_blank = 0;
++
++ if (break_spaces && !convfail && iswblank (wc))
++ {
++ last_blank_pos = offset_out;
++ last_blank_column = column;
++ is_blank_seen = 1;
++ last_blank_increment = increment;
++ is_bs_following_last_blank = 1;
++ bs_following_last_blank_num = 0;
++ is_cr_after_last_blank = 0;
++ }
++ }
++
++ *saved_errno = errno;
+
+ if (offset_out)
+ fwrite (line_out, sizeof (char), (size_t) offset_out, stdout);
+
++ free(line_out);
++}
++#endif
++
++/* Fold file FILENAME, or standard input if FILENAME is "-",
++ to stdout, with maximum line length WIDTH.
++ Return 0 if successful, 1 if an error occurs. */
++
++static bool
++fold_file (char *filename, size_t width)
++{
++ FILE *istream;
++ int saved_errno;
++
++ if (STREQ (filename, "-"))
++ {
++ istream = stdin;
++ have_read_stdin = 1;
++ }
++ else
++ istream = fopen (filename, "r");
++
++ if (istream == NULL)
++ {
++ error (0, errno, "%s", filename);
++ return 1;
++ }
++
++ /* Define how ISTREAM is being folded. */
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ fold_multibyte_text (istream, width, &saved_errno);
++ else
++#endif
++ fold_text (istream, width, &saved_errno);
++
+ if (ferror (istream))
+ {
+ error (0, saved_errno, "%s", filename);
+@@ -255,7 +506,8 @@
+
+ atexit (close_stdout);
+
+- break_spaces = count_bytes = have_read_stdin = false;
++ operating_mode = column_mode;
++ break_spaces = have_read_stdin = false;
+
+ while ((optc = getopt_long (argc, argv, shortopts, longopts, NULL)) != -1)
+ {
+@@ -264,7 +516,15 @@
+ switch (optc)
+ {
+ case 'b': /* Count bytes rather than columns. */
+- count_bytes = true;
++ if (operating_mode != column_mode)
++ FATAL_ERROR (_("only one way of folding may be specified"));
++ operating_mode = byte_mode;
++ break;
++
++ case 'c':
++ if (operating_mode != column_mode)
++ FATAL_ERROR (_("only one way of folding may be specified"));
++ operating_mode = character_mode;
+ break;
+
+ case 's': /* Break at word boundaries. */
+--- coreutils-6.8+/src/sort.c.i18n 2007-02-24 11:23:23.000000000 +0000
++++ coreutils-6.8+/src/sort.c 2007-03-01 15:10:57.000000000 +0000
+@@ -23,10 +23,19 @@
+
+ #include <config.h>
+
++#include <assert.h>
+ #include <getopt.h>
+ #include <sys/types.h>
+ #include <sys/wait.h>
+ #include <signal.h>
++#if HAVE_WCHAR_H
++# include <wchar.h>
++#endif
++/* Get isw* functions. */
++#if HAVE_WCTYPE_H
++# include <wctype.h>
++#endif
++
+ #include "system.h"
+ #include "argmatch.h"
+ #include "error.h"
+@@ -116,14 +125,38 @@
+ /* Thousands separator; if -1, then there isn't one. */
+ static int thousands_sep;
+
++static int force_general_numcompare = 0;
++
+ /* Nonzero if the corresponding locales are hard. */
+ static bool hard_LC_COLLATE;
+-#if HAVE_NL_LANGINFO
++#if HAVE_LANGINFO_CODESET
+ static bool hard_LC_TIME;
+ #endif
+
+ #define NONZERO(x) ((x) != 0)
+
++/* get a multibyte character's byte length. */
++#define GET_BYTELEN_OF_CHAR(LIM, PTR, MBLENGTH, STATE) \
++ do \
++ { \
++ wchar_t wc; \
++ mbstate_t state_bak; \
++ \
++ state_bak = STATE; \
++ mblength = mbrtowc (&wc, PTR, LIM - PTR, &STATE); \
++ \
++ switch (MBLENGTH) \
++ { \
++ case (size_t)-1: \
++ case (size_t)-2: \
++ STATE = state_bak; \
++ /* Fall through. */ \
++ case 0: \
++ MBLENGTH = 1; \
++ } \
++ } \
++ while (0)
++
+ /* The kind of blanks for '-b' to skip in various options. */
+ enum blanktype { bl_start, bl_end, bl_both };
+
+@@ -261,13 +294,11 @@
+ they were read if all keys compare equal. */
+ static bool stable;
+
+-/* If TAB has this value, blanks separate fields. */
+-enum { TAB_DEFAULT = CHAR_MAX + 1 };
+-
+-/* Tab character separating fields. If TAB_DEFAULT, then fields are
++/* Tab character separating fields. If tab_length is 0, then fields are
+ separated by the empty string between a non-blank character and a blank
+ character. */
+-static int tab = TAB_DEFAULT;
++static char tab[MB_LEN_MAX + 1];
++static size_t tab_length = 0;
+
+ /* Flag to remove consecutive duplicate lines from the output.
+ Only the last of a sequence of equal lines will be output. */
+@@ -639,6 +670,44 @@
+ update_proc (pid);
+ }
+
++/* Function pointers. */
++static void
++(*inittables) (void);
++static char *
++(*begfield) (const struct line*, const struct keyfield *);
++static char *
++(*limfield) (const struct line*, const struct keyfield *);
++static int
++(*getmonth) (char const *, size_t);
++static int
++(*keycompare) (const struct line *, const struct line *);
++static int
++(*numcompare) (const char *, const char *);
++
++/* Test for white space multibyte character.
++ Set LENGTH the byte length of investigated multibyte character. */
++#if HAVE_MBRTOWC
++static int
++ismbblank (const char *str, size_t len, size_t *length)
++{
++ size_t mblength;
++ wchar_t wc;
++ mbstate_t state;
++
++ memset (&state, '\0', sizeof(mbstate_t));
++ mblength = mbrtowc (&wc, str, len, &state);
++
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ *length = 1;
++ return 0;
++ }
++
++ *length = (mblength < 1) ? 1 : mblength;
++ return iswblank (wc);
++}
++#endif
++
+ /* Clean up any remaining temporary files. */
+
+ static void
+@@ -978,7 +1047,7 @@
+ free (node);
+ }
+
+-#if HAVE_NL_LANGINFO
++#if HAVE_LANGINFO_CODESET
+
+ static int
+ struct_month_cmp (const void *m1, const void *m2)
+@@ -993,7 +1062,7 @@
+ /* Initialize the character class tables. */
+
+ static void
+-inittables (void)
++inittables_uni (void)
+ {
+ size_t i;
+
+@@ -1005,7 +1074,7 @@
+ fold_toupper[i] = toupper (i);
+ }
+
+-#if HAVE_NL_LANGINFO
++#if HAVE_LANGINFO_CODESET
+ /* If we're not in the "C" locale, read different names for months. */
+ if (hard_LC_TIME)
+ {
+@@ -1031,6 +1100,64 @@
+ #endif
+ }
+
++#if HAVE_MBRTOWC
++static void
++inittables_mb (void)
++{
++ int i, j, k, l;
++ char *name, *s;
++ size_t s_len, mblength;
++ char mbc[MB_LEN_MAX];
++ wchar_t wc, pwc;
++ mbstate_t state_mb, state_wc;
++
++ for (i = 0; i < MONTHS_PER_YEAR; i++)
++ {
++ s = (char *) nl_langinfo (ABMON_1 + i);
++ s_len = strlen (s);
++ monthtab[i].name = name = (char *) xmalloc (s_len + 1);
++ monthtab[i].val = i + 1;
++
++ memset (&state_mb, '\0', sizeof (mbstate_t));
++ memset (&state_wc, '\0', sizeof (mbstate_t));
++
++ for (j = 0; j < s_len;)
++ {
++ if (!ismbblank (s + j, s_len - j, &mblength))
++ break;
++ j += mblength;
++ }
++
++ for (k = 0; j < s_len;)
++ {
++ mblength = mbrtowc (&wc, (s + j), (s_len - j), &state_mb);
++ assert (mblength != (size_t)-1 && mblength != (size_t)-2);
++ if (mblength == 0)
++ break;
++
++ pwc = towupper (wc);
++ if (pwc == wc)
++ {
++ memcpy (mbc, s + j, mblength);
++ j += mblength;
++ }
++ else
++ {
++ j += mblength;
++ mblength = wcrtomb (mbc, pwc, &state_wc);
++ assert (mblength != (size_t)0 && mblength != (size_t)-1);
++ }
++
++ for (l = 0; l < mblength; l++)
++ name[k++] = mbc[l];
++ }
++ name[k] = '\0';
++ }
++ qsort ((void *) monthtab, MONTHS_PER_YEAR,
++ sizeof (struct month), struct_month_cmp);
++}
++#endif
++
+ /* Specify the amount of main memory to use when sorting. */
+ static void
+ specify_sort_size (char const *s)
+@@ -1241,7 +1368,7 @@
+ by KEY in LINE. */
+
+ static char *
+-begfield (const struct line *line, const struct keyfield *key)
++begfield_uni (const struct line *line, const struct keyfield *key)
+ {
+ char *ptr = line->text, *lim = ptr + line->length - 1;
+ size_t sword = key->sword;
+@@ -1251,10 +1378,10 @@
+ /* The leading field separator itself is included in a field when -t
+ is absent. */
+
+- if (tab != TAB_DEFAULT)
++ if (tab_length)
+ while (ptr < lim && sword--)
+ {
+- while (ptr < lim && *ptr != tab)
++ while (ptr < lim && *ptr != tab[0])
+ ++ptr;
+ if (ptr < lim)
+ ++ptr;
+@@ -1282,11 +1409,70 @@
+ return ptr;
+ }
+
++#if HAVE_MBRTOWC
++static char *
++begfield_mb (const struct line *line, const struct keyfield *key)
++{
++ int i;
++ char *ptr = line->text, *lim = ptr + line->length - 1;
++ size_t sword = key->sword;
++ size_t schar = key->schar;
++ size_t mblength;
++ mbstate_t state;
++
++ memset (&state, '\0', sizeof(mbstate_t));
++
++ if (tab_length)
++ while (ptr < lim && sword--)
++ {
++ while (ptr < lim && memcmp (ptr, tab, tab_length) != 0)
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++ ptr += mblength;
++ }
++ if (ptr < lim)
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++ ptr += mblength;
++ }
++ }
++ else
++ while (ptr < lim && sword--)
++ {
++ while (ptr < lim && ismbblank (ptr, lim - ptr, &mblength))
++ ptr += mblength;
++ if (ptr < lim)
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++ ptr += mblength;
++ }
++ while (ptr < lim && !ismbblank (ptr, lim - ptr, &mblength))
++ ptr += mblength;
++ }
++
++ if (key->skipsblanks)
++ while (ptr < lim && ismbblank (ptr, lim - ptr, &mblength))
++ ptr += mblength;
++
++ for (i = 0; i < schar; i++)
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++
++ if (ptr + mblength > lim)
++ break;
++ else
++ ptr += mblength;
++ }
++
++ return ptr;
++}
++#endif
++
+ /* Return the limit of (a pointer to the first character after) the field
+ in LINE specified by KEY. */
+
+ static char *
+-limfield (const struct line *line, const struct keyfield *key)
++limfield_uni (const struct line *line, const struct keyfield *key)
+ {
+ char *ptr = line->text, *lim = ptr + line->length - 1;
+ size_t eword = key->eword, echar = key->echar;
+@@ -1299,10 +1485,10 @@
+ `beginning' is the first character following the delimiting TAB.
+ Otherwise, leave PTR pointing at the first `blank' character after
+ the preceding field. */
+- if (tab != TAB_DEFAULT)
++ if (tab_length)
+ while (ptr < lim && eword--)
+ {
+- while (ptr < lim && *ptr != tab)
++ while (ptr < lim && *ptr != tab[0])
+ ++ptr;
+ if (ptr < lim && (eword | echar))
+ ++ptr;
+@@ -1348,10 +1534,10 @@
+ */
+
+ /* Make LIM point to the end of (one byte past) the current field. */
+- if (tab != TAB_DEFAULT)
++ if (tab_length)
+ {
+ char *newlim;
+- newlim = memchr (ptr, tab, lim - ptr);
++ newlim = memchr (ptr, tab[0], lim - ptr);
+ if (newlim)
+ lim = newlim;
+ }
+@@ -1384,6 +1570,107 @@
+ return ptr;
+ }
+
++#if HAVE_MBRTOWC
++static char *
++limfield_mb (const struct line *line, const struct keyfield *key)
++{
++ char *ptr = line->text, *lim = ptr + line->length - 1;
++ size_t eword = key->eword, echar = key->echar;
++ int i;
++ size_t mblength;
++ mbstate_t state;
++
++ memset (&state, '\0', sizeof(mbstate_t));
++
++ if (tab_length)
++ while (ptr < lim && eword--)
++ {
++ while (ptr < lim && memcmp (ptr, tab, tab_length) != 0)
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++ ptr += mblength;
++ }
++ if (ptr < lim && (eword | echar))
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++ ptr += mblength;
++ }
++ }
++ else
++ while (ptr < lim && eword--)
++ {
++ while (ptr < lim && ismbblank (ptr, lim - ptr, &mblength))
++ ptr += mblength;
++ if (ptr < lim)
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++ ptr += mblength;
++ }
++ while (ptr < lim && !ismbblank (ptr, lim - ptr, &mblength))
++ ptr += mblength;
++ }
++
++
++# ifdef POSIX_UNSPECIFIED
++ /* Make LIM point to the end of (one byte past) the current field. */
++ if (tab_length)
++ {
++ char *newlim, *p;
++
++ newlim = NULL;
++ for (p = ptr; p < lim;)
++ {
++ if (memcmp (p, tab, tab_length) == 0)
++ {
++ newlim = p;
++ break;
++ }
++
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++ p += mblength;
++ }
++ }
++ else
++ {
++ char *newlim;
++ newlim = ptr;
++
++ while (newlim < lim && ismbblank (newlim, lim - newlim, &mblength))
++ newlim += mblength;
++ if (ptr < lim)
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++ ptr += mblength;
++ }
++ while (newlim < lim && !ismbblank (newlim, lim - newlim, &mblength))
++ newlim += mblength;
++ lim = newlim;
++ }
++# endif
++
++ /* If we're skipping leading blanks, don't start counting characters
++ * until after skipping past any leading blanks. */
++ if (key->skipsblanks)
++ while (ptr < lim && ismbblank (ptr, lim - ptr, &mblength))
++ ptr += mblength;
++
++ memset (&state, '\0', sizeof(mbstate_t));
++
++ /* Advance PTR by ECHAR (if possible), but no further than LIM. */
++ for (i = 0; i < echar; i++)
++ {
++ GET_BYTELEN_OF_CHAR (lim, ptr, mblength, state);
++
++ if (ptr + mblength > lim)
++ break;
++ else
++ ptr += mblength;
++ }
++
++ return ptr;
++}
++#endif
++
+ /* Fill BUF reading from FP, moving buf->left bytes from the end
+ of buf->buf to the beginning first. If EOF is reached and the
+ file wasn't terminated by a newline, supply one. Set up BUF's line
+@@ -1466,8 +1753,24 @@
+ else
+ {
+ if (key->skipsblanks)
+- while (blanks[to_uchar (*line_start)])
+- line_start++;
++ {
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ size_t mblength;
++ mbstate_t state;
++ memset (&state, '\0', sizeof(mbstate_t));
++ while (line_start < line->keylim &&
++ ismbblank (line_start,
++ line->keylim - line_start,
++ &mblength))
++ line_start += mblength;
++ }
++ else
++#endif
++ while (blanks[to_uchar (*line_start)])
++ line_start++;
++ }
+ line->keybeg = line_start;
+ }
+ }
+@@ -1500,7 +1803,7 @@
+ hideously fast. */
+
+ static int
+-numcompare (const char *a, const char *b)
++numcompare_uni (const char *a, const char *b)
+ {
+ while (blanks[to_uchar (*a)])
+ a++;
+@@ -1510,6 +1813,25 @@
+ return strnumcmp (a, b, decimal_point, thousands_sep);
+ }
+
++#if HAVE_MBRTOWC
++static int
++numcompare_mb (const char *a, const char *b)
++{
++ size_t mblength, len;
++ len = strlen (a); /* okay for UTF-8 */
++ while (*a && ismbblank (a, len > MB_CUR_MAX ? MB_CUR_MAX : len, &mblength))
++ {
++ a += mblength;
++ len -= mblength;
++ }
++ len = strlen (b); /* okay for UTF-8 */
++ while (*b && ismbblank (b, len > MB_CUR_MAX ? MB_CUR_MAX : len, &mblength))
++ b += mblength;
++
++ return strnumcmp (a, b, decimal_point, thousands_sep);
++}
++#endif /* HAV_EMBRTOWC */
++
+ static int
+ general_numcompare (const char *sa, const char *sb)
+ {
+@@ -1543,7 +1865,7 @@
+ Return 0 if the name in S is not recognized. */
+
+ static int
+-getmonth (char const *month, size_t len)
++getmonth_uni (char const *month, size_t len)
+ {
+ size_t lo = 0;
+ size_t hi = MONTHS_PER_YEAR;
+@@ -1698,11 +2020,79 @@
+ return diff;
+ }
+
++#if HAVE_MBRTOWC
++static int
++getmonth_mb (const char *s, size_t len)
++{
++ char *month;
++ register size_t i;
++ register int lo = 0, hi = MONTHS_PER_YEAR, result;
++ char *tmp;
++ size_t wclength, mblength;
++ const char **pp;
++ const wchar_t **wpp;
++ wchar_t *month_wcs;
++ mbstate_t state;
++
++ while (len > 0 && ismbblank (s, len, &mblength))
++ {
++ s += mblength;
++ len -= mblength;
++ }
++
++ if (len == 0)
++ return 0;
++
++ month = (char *) alloca (len + 1);
++
++ tmp = (char *) alloca (len + 1);
++ memcpy (tmp, s, len);
++ tmp[len] = '\0';
++ pp = (const char **)&tmp;
++ month_wcs = (wchar_t *) alloca ((len + 1) * sizeof (wchar_t));
++ memset (&state, '\0', sizeof(mbstate_t));
++
++ wclength = mbsrtowcs (month_wcs, pp, len + 1, &state);
++ assert (wclength != (size_t)-1 && *pp == NULL);
++
++ for (i = 0; i < wclength; i++)
++ {
++ month_wcs[i] = towupper(month_wcs[i]);
++ if (iswblank (month_wcs[i]))
++ {
++ month_wcs[i] = L'\0';
++ break;
++ }
++ }
++
++ wpp = (const wchar_t **)&month_wcs;
++
++ mblength = wcsrtombs (month, wpp, len + 1, &state);
++ assert (mblength != (-1) && *wpp == NULL);
++
++ do
++ {
++ int ix = (lo + hi) / 2;
++
++ if (strncmp (month, monthtab[ix].name, strlen (monthtab[ix].name)) < 0)
++ hi = ix;
++ else
++ lo = ix;
++ }
++ while (hi - lo > 1);
++
++ result = (!strncmp (month, monthtab[lo].name, strlen (monthtab[lo].name))
++ ? monthtab[lo].val : 0);
++
++ return result;
++}
++#endif
++
+ /* Compare two lines A and B trying every key in sequence until there
+ are no more keys or a difference is found. */
+
+ static int
+-keycompare (const struct line *a, const struct line *b)
++keycompare_uni (const struct line *a, const struct line *b)
+ {
+ struct keyfield const *key = keylist;
+
+@@ -1875,6 +2265,179 @@
+ return key->reverse ? -diff : diff;
+ }
+
++#if HAVE_MBRTOWC
++static int
++keycompare_mb (const struct line *a, const struct line *b)
++{
++ struct keyfield *key = keylist;
++
++ /* For the first iteration only, the key positions have been
++ precomputed for us. */
++ char *texta = a->keybeg;
++ char *textb = b->keybeg;
++ char *lima = a->keylim;
++ char *limb = b->keylim;
++
++ size_t mblength_a, mblength_b;
++ wchar_t wc_a, wc_b;
++ mbstate_t state_a, state_b;
++
++ int diff;
++
++ memset (&state_a, '\0', sizeof(mbstate_t));
++ memset (&state_b, '\0', sizeof(mbstate_t));
++
++ for (;;)
++ {
++ unsigned char *translate = (unsigned char *) key->translate;
++ bool const *ignore = key->ignore;
++
++ /* Find the lengths. */
++ size_t lena = lima <= texta ? 0 : lima - texta;
++ size_t lenb = limb <= textb ? 0 : limb - textb;
++
++ /* Actually compare the fields. */
++ if (key->random)
++ diff = compare_random (texta, lena, textb, lenb);
++ else if (key->numeric | key->general_numeric)
++ {
++ char savea = *lima, saveb = *limb;
++
++ *lima = *limb = '\0';
++ if (force_general_numcompare)
++ diff = general_numcompare (texta, textb);
++ else
++ diff = ((key->numeric ? numcompare : general_numcompare)
++ (texta, textb));
++ *lima = savea, *limb = saveb;
++ }
++ else if (key->month)
++ diff = getmonth (texta, lena) - getmonth (textb, lenb);
++ else
++ {
++ if (ignore || translate)
++ {
++ char *copy_a = (char *) alloca (lena + 1 + lenb + 1);
++ char *copy_b = copy_a + lena + 1;
++ size_t new_len_a, new_len_b;
++ size_t i, j;
++
++ /* Ignore and/or translate chars before comparing. */
++# define IGNORE_CHARS(NEW_LEN, LEN, TEXT, COPY, WC, MBLENGTH, STATE) \
++ do \
++ { \
++ wchar_t uwc; \
++ char mbc[MB_LEN_MAX]; \
++ mbstate_t state_wc; \
++ \
++ for (NEW_LEN = i = 0; i < LEN;) \
++ { \
++ mbstate_t state_bak; \
++ \
++ state_bak = STATE; \
++ MBLENGTH = mbrtowc (&WC, TEXT + i, LEN - i, &STATE); \
++ \
++ if (MBLENGTH == (size_t)-2 || MBLENGTH == (size_t)-1 \
++ || MBLENGTH == 0) \
++ { \
++ if (MBLENGTH == (size_t)-2 || MBLENGTH == (size_t)-1) \
++ STATE = state_bak; \
++ if (!ignore) \
++ COPY[NEW_LEN++] = TEXT[i++]; \
++ continue; \
++ } \
++ \
++ if (ignore) \
++ { \
++ if ((ignore == nonprinting && !iswprint (WC)) \
++ || (ignore == nondictionary \
++ && !iswalnum (WC) && !iswblank (WC))) \
++ { \
++ i += MBLENGTH; \
++ continue; \
++ } \
++ } \
++ \
++ if (translate) \
++ { \
++ \
++ uwc = towupper(WC); \
++ if (WC == uwc) \
++ { \
++ memcpy (mbc, TEXT + i, MBLENGTH); \
++ i += MBLENGTH; \
++ } \
++ else \
++ { \
++ i += MBLENGTH; \
++ WC = uwc; \
++ memset (&state_wc, '\0', sizeof (mbstate_t)); \
++ \
++ MBLENGTH = wcrtomb (mbc, WC, &state_wc); \
++ assert (MBLENGTH != (size_t)-1 && MBLENGTH != 0); \
++ } \
++ \
++ for (j = 0; j < MBLENGTH; j++) \
++ COPY[NEW_LEN++] = mbc[j]; \
++ } \
++ else \
++ for (j = 0; j < MBLENGTH; j++) \
++ COPY[NEW_LEN++] = TEXT[i++]; \
++ } \
++ COPY[NEW_LEN] = '\0'; \
++ } \
++ while (0)
++ IGNORE_CHARS (new_len_a, lena, texta, copy_a,
++ wc_a, mblength_a, state_a);
++ IGNORE_CHARS (new_len_b, lenb, textb, copy_b,
++ wc_b, mblength_b, state_b);
++ diff = xmemcoll (copy_a, new_len_a, copy_b, new_len_b);
++ }
++ else if (lena == 0)
++ diff = - NONZERO (lenb);
++ else if (lenb == 0)
++ goto greater;
++ else
++ diff = xmemcoll (texta, lena, textb, lenb);
++ }
++
++ if (diff)
++ goto not_equal;
++
++ key = key->next;
++ if (! key)
++ break;
++
++ /* Find the beginning and limit of the next field. */
++ if (key->eword != -1)
++ lima = limfield (a, key), limb = limfield (b, key);
++ else
++ lima = a->text + a->length - 1, limb = b->text + b->length - 1;
++
++ if (key->sword != -1)
++ texta = begfield (a, key), textb = begfield (b, key);
++ else
++ {
++ texta = a->text, textb = b->text;
++ if (key->skipsblanks)
++ {
++ while (texta < lima && ismbblank (texta, lima - texta, &mblength_a))
++ texta += mblength_a;
++ while (textb < limb && ismbblank (textb, limb - textb, &mblength_b))
++ textb += mblength_b;
++ }
++ }
++ }
++
++ return 0;
++
++greater:
++ diff = 1;
++not_equal:
++ return key->reverse ? -diff : diff;
++}
++#endif
++
+ /* Compare two lines A and B, returning negative, zero, or positive
+ depending on whether A compares less than, equal to, or greater than B. */
+
+@@ -2744,7 +3305,7 @@
+ initialize_exit_failure (SORT_FAILURE);
+
+ hard_LC_COLLATE = hard_locale (LC_COLLATE);
+-#if HAVE_NL_LANGINFO
++#if HAVE_LANGINFO_CODESET
+ hard_LC_TIME = hard_locale (LC_TIME);
+ #endif
+
+@@ -2765,6 +3326,27 @@
+ thousands_sep = -1;
+ }
+
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ inittables = inittables_mb;
++ begfield = begfield_mb;
++ limfield = limfield_mb;
++ getmonth = getmonth_mb;
++ keycompare = keycompare_mb;
++ numcompare = numcompare_mb;
++ }
++ else
++#endif
++ {
++ inittables = inittables_uni;
++ begfield = begfield_uni;
++ limfield = limfield_uni;
++ getmonth = getmonth_uni;
++ keycompare = keycompare_uni;
++ numcompare = numcompare_uni;
++ }
++
+ have_read_stdin = false;
+ inittables ();
+
+@@ -3015,13 +3597,35 @@
+
+ case 't':
+ {
+- char newtab = optarg[0];
+- if (! newtab)
++ char newtab[MB_LEN_MAX + 1];
++ size_t newtab_length = 1;
++ strncpy (newtab, optarg, MB_LEN_MAX);
++ if (! newtab[0])
+ error (SORT_FAILURE, 0, _("empty tab"));
+- if (optarg[1])
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ wchar_t wc;
++ mbstate_t state;
++ size_t i;
++
++ memset (&state, '\0', sizeof (mbstate_t));
++ newtab_length = mbrtowc (&wc, newtab, strnlen (newtab,
++ MB_LEN_MAX),
++ &state);
++ switch (newtab_length)
++ {
++ case (size_t) -1:
++ case (size_t) -2:
++ case 0:
++ newtab_length = 1;
++ }
++ }
++#endif
++ if (newtab_length == 1 && optarg[1])
+ {
+ if (STREQ (optarg, "\\0"))
+- newtab = '\0';
++ newtab[0] = '\0';
+ else
+ {
+ /* Provoke with `sort -txx'. Complain about
+@@ -3032,9 +3636,12 @@
+ quote (optarg));
+ }
+ }
+- if (tab != TAB_DEFAULT && tab != newtab)
++ if (tab_length
++ && (tab_length != newtab_length
++ || memcmp (tab, newtab, tab_length) != 0))
+ error (SORT_FAILURE, 0, _("incompatible tabs"));
+- tab = newtab;
++ memcpy (tab, newtab, newtab_length);
++ tab_length = newtab_length;
+ }
+ break;
+
+--- coreutils-6.8+/src/unexpand.c.i18n 2007-01-14 15:41:28.000000000 +0000
++++ coreutils-6.8+/src/unexpand.c 2007-03-01 15:08:24.000000000 +0000
+@@ -39,11 +39,28 @@
+ #include <stdio.h>
+ #include <getopt.h>
+ #include <sys/types.h>
++
++/* Get mbstate_t, mbrtowc(), wcwidth(). */
++#if HAVE_WCHAR_H
++# include <wchar.h>
++#endif
++
+ #include "system.h"
+ #include "error.h"
+ #include "quote.h"
+ #include "xstrndup.h"
+
++/* MB_LEN_MAX is incorrectly defined to be 1 in at least one GCC
++ installation; work around this configuration error. */
++#if !defined MB_LEN_MAX || MB_LEN_MAX < 2
++# define MB_LEN_MAX 16
++#endif
++
++/* Some systems, like BeOS, have multibyte encodings but lack mbstate_t. */
++#if HAVE_MBRTOWC && defined mbstate_t
++# define mbrtowc(pwc, s, n, ps) (mbrtowc) (pwc, s, n, 0)
++#endif
++
+ /* The official name of this program (e.g., no `g' prefix). */
+ #define PROGRAM_NAME "unexpand"
+
+@@ -110,6 +127,208 @@
+ {NULL, 0, NULL, 0}
+ };
+
++static FILE *next_file (FILE *fp);
++
++#if HAVE_MBRTOWC
++static void
++unexpand_multibyte (void)
++{
++ FILE *fp; /* Input stream. */
++ mbstate_t i_state; /* Current shift state of the input stream. */
++ mbstate_t i_state_bak; /* Back up the I_STATE. */
++ mbstate_t o_state; /* Current shift state of the output stream. */
++ char buf[MB_LEN_MAX + BUFSIZ]; /* For spooling a read byte sequence. */
++ char *bufpos; /* Next read position of BUF. */
++ size_t buflen = 0; /* The length of the byte sequence in buf. */
++ wint_t wc; /* A gotten wide character. */
++ size_t mblength; /* The byte size of a multibyte character
++ which shows as same character as WC. */
++
++ /* Index in `tab_list' of next tabstop: */
++ int tab_index = 0; /* For calculating width of pending tabs. */
++ int print_tab_index = 0; /* For printing as many tabs as possible. */
++ unsigned int column = 0; /* Column on screen of next char. */
++ int next_tab_column; /* Column the next tab stop is on. */
++ int convert = 1; /* If nonzero, perform translations. */
++ unsigned int pending = 0; /* Pending columns of blanks. */
++
++ fp = next_file ((FILE *) NULL);
++ if (fp == NULL)
++ return;
++
++ memset (&o_state, '\0', sizeof(mbstate_t));
++ memset (&i_state, '\0', sizeof(mbstate_t));
++
++ for (;;)
++ {
++ if (buflen < MB_LEN_MAX && !feof(fp) && !ferror(fp))
++ {
++ memmove (buf, bufpos, buflen);
++ buflen += fread (buf + buflen, sizeof(char), BUFSIZ, fp);
++ bufpos = buf;
++ }
++
++ /* Get a wide character. */
++ if (buflen < 1)
++ {
++ mblength = 1;
++ wc = WEOF;
++ }
++ else
++ {
++ i_state_bak = i_state;
++ mblength = mbrtowc ((wchar_t *)&wc, bufpos, buflen, &i_state);
++ }
++
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ i_state = i_state_bak;
++ wc = L'\0';
++ }
++
++ if (wc == L' ' && convert && column < INT_MAX)
++ {
++ ++pending;
++ ++column;
++ }
++ else if (wc == L'\t' && convert)
++ {
++ if (tab_size == 0)
++ {
++ /* Do not let tab_index == first_free_tab;
++ stop when it is 1 less. */
++ while (tab_index < first_free_tab - 1
++ && column >= tab_list[tab_index])
++ tab_index++;
++ next_tab_column = tab_list[tab_index];
++ if (tab_index < first_free_tab - 1)
++ tab_index++;
++ if (column >= next_tab_column)
++ {
++ convert = 0; /* Ran out of tab stops. */
++ goto flush_pend_mb;
++ }
++ }
++ else
++ {
++ next_tab_column = column + tab_size - column % tab_size;
++ }
++ pending += next_tab_column - column;
++ column = next_tab_column;
++ }
++ else
++ {
++flush_pend_mb:
++ /* Flush pending spaces. Print as many tabs as possible,
++ then print the rest as spaces. */
++ if (pending == 1)
++ {
++ putchar (' ');
++ pending = 0;
++ }
++ column -= pending;
++ while (pending > 0)
++ {
++ if (tab_size == 0)
++ {
++ /* Do not let print_tab_index == first_free_tab;
++ stop when it is 1 less. */
++ while (print_tab_index < first_free_tab - 1
++ && column >= tab_list[print_tab_index])
++ print_tab_index++;
++ next_tab_column = tab_list[print_tab_index];
++ if (print_tab_index < first_free_tab - 1)
++ print_tab_index++;
++ }
++ else
++ {
++ next_tab_column =
++ column + tab_size - column % tab_size;
++ }
++ if (next_tab_column - column <= pending)
++ {
++ putchar ('\t');
++ pending -= next_tab_column - column;
++ column = next_tab_column;
++ }
++ else
++ {
++ --print_tab_index;
++ column += pending;
++ while (pending != 0)
++ {
++ putchar (' ');
++ pending--;
++ }
++ }
++ }
++
++ if (wc == WEOF)
++ {
++ fp = next_file (fp);
++ if (fp == NULL)
++ break; /* No more files. */
++ else
++ {
++ memset (&i_state, '\0', sizeof(mbstate_t));
++ continue;
++ }
++ }
++
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ if (convert)
++ {
++ ++column;
++ if (convert_entire_line == 0)
++ convert = 0;
++ }
++ mblength = 1;
++ putchar (buf[0]);
++ }
++ else if (mblength == 0)
++ {
++ if (convert && convert_entire_line == 0)
++ convert = 0;
++ mblength = 1;
++ putchar ('\0');
++ }
++ else
++ {
++ if (convert)
++ {
++ if (wc == L'\b')
++ {
++ if (column > 0)
++ --column;
++ }
++ else
++ {
++ int width; /* The width of WC. */
++
++ width = wcwidth (wc);
++ column += (width > 0) ? width : 0;
++ if (convert_entire_line == 0)
++ convert = 0;
++ }
++ }
++
++ if (wc == L'\n')
++ {
++ tab_index = print_tab_index = 0;
++ column = pending = 0;
++ convert = 1;
++ }
++ fwrite (bufpos, sizeof(char), mblength, stdout);
++ }
++ }
++ buflen -= mblength;
++ bufpos += mblength;
++ }
++}
++#endif
++
++
+ void
+ usage (int status)
+ {
+@@ -531,7 +750,12 @@
+
+ file_list = (optind < argc ? &argv[optind] : stdin_argv);
+
+- unexpand ();
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ unexpand_multibyte ();
++ else
++#endif
++ unexpand ();
+
+ if (have_read_stdin && fclose (stdin) != 0)
+ error (EXIT_FAILURE, errno, "-");
+--- coreutils-6.8+/src/pr.c.i18n 2007-01-14 15:41:28.000000000 +0000
++++ coreutils-6.8+/src/pr.c 2007-03-01 15:08:24.000000000 +0000
+@@ -313,6 +313,32 @@
+
+ #include <getopt.h>
+ #include <sys/types.h>
++
++/* Get MB_LEN_MAX. */
++#include <limits.h>
++/* MB_LEN_MAX is incorrectly defined to be 1 in at least one GCC
++ installation; work around this configuration error. */
++#if !defined MB_LEN_MAX || MB_LEN_MAX == 1
++# define MB_LEN_MAX 16
++#endif
++
++/* Get MB_CUR_MAX. */
++#include <stdlib.h>
++
++/* Solaris 2.5 has a bug: <wchar.h> must be included before <wctype.h>. */
++/* Get mbstate_t, mbrtowc(), wcwidth(). */
++#if HAVE_WCHAR_H
++# include <wchar.h>
++#endif
++
++/* Get iswprint(). -- for wcwidth(). */
++#if HAVE_WCTYPE_H
++# include <wctype.h>
++#endif
++#if !defined iswprint && !HAVE_ISWPRINT
++# define iswprint(wc) 1
++#endif
++
+ #include "system.h"
+ #include "error.h"
+ #include "hard-locale.h"
+@@ -324,6 +350,18 @@
+ #include "strftime.h"
+ #include "xstrtol.h"
+
++/* Some systems, like BeOS, have multibyte encodings but lack mbstate_t. */
++#if HAVE_MBRTOWC && defined mbstate_t
++# define mbrtowc(pwc, s, n, ps) (mbrtowc) (pwc, s, n, 0)
++#endif
++
++#ifndef HAVE_DECL_WCWIDTH
++"this configure-time declaration test was not run"
++#endif
++#if !HAVE_DECL_WCWIDTH
++extern int wcwidth ();
++#endif
++
+ /* The official name of this program (e.g., no `g' prefix). */
+ #define PROGRAM_NAME "pr"
+
+@@ -416,7 +454,20 @@
+
+ #define NULLCOL (COLUMN *)0
+
+-static int char_to_clump (char c);
++/* Funtion pointers to switch functions for single byte locale or for
++ multibyte locale. If multibyte functions do not exist in your sysytem,
++ these pointers always point the function for single byte locale. */
++static void (*print_char) (char c);
++static int (*char_to_clump) (char c);
++
++/* Functions for single byte locale. */
++static void print_char_single (char c);
++static int char_to_clump_single (char c);
++
++/* Functions for multibyte locale. */
++static void print_char_multi (char c);
++static int char_to_clump_multi (char c);
++
+ static bool read_line (COLUMN *p);
+ static bool print_page (void);
+ static bool print_stored (COLUMN *p);
+@@ -426,6 +477,7 @@
+ static void pad_across_to (int position);
+ static void add_line_number (COLUMN *p);
+ static void getoptarg (char *arg, char switch_char, char *character,
++ int *character_length, int *character_width,
+ int *number);
+ void usage (int status);
+ static void print_files (int number_of_files, char **av);
+@@ -440,7 +492,6 @@
+ static void pad_down (int lines);
+ static void read_rest_of_line (COLUMN *p);
+ static void skip_read (COLUMN *p, int column_number);
+-static void print_char (char c);
+ static void cleanup (void);
+ static void print_sep_string (void);
+ static void separator_string (const char *optarg_S);
+@@ -455,7 +506,7 @@
+ we store the leftmost columns contiguously in buff.
+ To print a line from buff, get the index of the first character
+ from line_vector[i], and print up to line_vector[i + 1]. */
+-static char *buff;
++static unsigned char *buff;
+
+ /* Index of the position in buff where the next character
+ will be stored. */
+@@ -559,7 +610,7 @@
+ static bool untabify_input = false;
+
+ /* (-e) The input tab character. */
+-static char input_tab_char = '\t';
++static char input_tab_char[MB_LEN_MAX] = "\t";
+
+ /* (-e) Tabstops are at chars_per_tab, 2*chars_per_tab, 3*chars_per_tab, ...
+ where the leftmost column is 1. */
+@@ -569,7 +620,10 @@
+ static bool tabify_output = false;
+
+ /* (-i) The output tab character. */
+-static char output_tab_char = '\t';
++static char output_tab_char[MB_LEN_MAX] = "\t";
++
++/* (-i) The byte length of output tab character. */
++static int output_tab_char_length = 1;
+
+ /* (-i) The width of the output tab. */
+ static int chars_per_output_tab = 8;
+@@ -643,7 +697,13 @@
+ static bool numbered_lines = false;
+
+ /* (-n) Character which follows each line number. */
+-static char number_separator = '\t';
++static char number_separator[MB_LEN_MAX] = "\t";
++
++/* (-n) The byte length of the character which follows each line number. */
++static int number_separator_length = 1;
++
++/* (-n) The character width of the character which follows each line number. */
++static int number_separator_width = 0;
+
+ /* (-n) line counting starts with 1st line of input file (not with 1st
+ line of 1st page printed). */
+@@ -696,6 +756,7 @@
+ -a|COLUMN|-m is a `space' and with the -J option a `tab'. */
+ static char *col_sep_string = "";
+ static int col_sep_length = 0;
++static int col_sep_width = 0;
+ static char *column_separator = " ";
+ static char *line_separator = "\t";
+
+@@ -852,6 +913,13 @@
+ col_sep_length = (int) strlen (optarg_S);
+ col_sep_string = xmalloc (col_sep_length + 1);
+ strcpy (col_sep_string, optarg_S);
++
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ col_sep_width = mbswidth (col_sep_string, 0);
++ else
++#endif
++ col_sep_width = col_sep_length;
+ }
+
+ int
+@@ -877,6 +945,21 @@
+
+ atexit (close_stdout);
+
++/* Define which functions are used, the ones for single byte locale or the ones
++ for multibyte locale. */
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1)
++ {
++ print_char = print_char_multi;
++ char_to_clump = char_to_clump_multi;
++ }
++ else
++#endif
++ {
++ print_char = print_char_single;
++ char_to_clump = char_to_clump_single;
++ }
++
+ n_files = 0;
+ file_names = (argc > 1
+ ? xmalloc ((argc - 1) * sizeof (char *))
+@@ -949,8 +1032,12 @@
+ break;
+ case 'e':
+ if (optarg)
+- getoptarg (optarg, 'e', &input_tab_char,
+- &chars_per_input_tab);
++ {
++ int dummy_length, dummy_width;
++
++ getoptarg (optarg, 'e', input_tab_char, &dummy_length,
++ &dummy_width, &chars_per_input_tab);
++ }
+ /* Could check tab width > 0. */
+ untabify_input = true;
+ break;
+@@ -963,8 +1050,12 @@
+ break;
+ case 'i':
+ if (optarg)
+- getoptarg (optarg, 'i', &output_tab_char,
+- &chars_per_output_tab);
++ {
++ int dummy_width;
++
++ getoptarg (optarg, 'i', output_tab_char, &output_tab_char_length,
++ &dummy_width, &chars_per_output_tab);
++ }
+ /* Could check tab width > 0. */
+ tabify_output = true;
+ break;
+@@ -991,8 +1082,8 @@
+ case 'n':
+ numbered_lines = true;
+ if (optarg)
+- getoptarg (optarg, 'n', &number_separator,
+- &chars_per_number);
++ getoptarg (optarg, 'n', number_separator, &number_separator_length,
++ &number_separator_width, &chars_per_number);
+ break;
+ case 'N':
+ skip_count = false;
+@@ -1031,7 +1122,7 @@
+ old_s = false;
+ /* Reset an additional input of -s, -S dominates -s */
+ col_sep_string = "";
+- col_sep_length = 0;
++ col_sep_length = col_sep_width = 0;
+ use_col_separator = true;
+ if (optarg)
+ separator_string (optarg);
+@@ -1188,10 +1279,45 @@
+ a number. */
+
+ static void
+-getoptarg (char *arg, char switch_char, char *character, int *number)
++getoptarg (char *arg, char switch_char, char *character, int *character_length,
++ int *character_width, int *number)
+ {
+ if (!ISDIGIT (*arg))
+- *character = *arg++;
++ {
++#ifdef HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1) /* for multibyte locale. */
++ {
++ wchar_t wc;
++ size_t mblength;
++ int width;
++ mbstate_t state = {'\0'};
++
++ mblength = mbrtowc (&wc, arg, strnlen(arg, MB_LEN_MAX), &state);
++
++ if (mblength == (size_t)-1 || mblength == (size_t)-2)
++ {
++ *character_length = 1;
++ *character_width = 1;
++ }
++ else
++ {
++ *character_length = (mblength < 1) ? 1 : mblength;
++ width = wcwidth (wc);
++ *character_width = (width < 0) ? 0 : width;
++ }
++
++ strncpy (character, arg, *character_length);
++ arg += *character_length;
++ }
++ else /* for single byte locale. */
++#endif
++ {
++ *character = *arg++;
++ *character_length = 1;
++ *character_width = 1;
++ }
++ }
++
+ if (*arg)
+ {
+ long int tmp_long;
+@@ -1256,7 +1382,7 @@
+ else
+ col_sep_string = column_separator;
+
+- col_sep_length = 1;
++ col_sep_length = col_sep_width = 1;
+ use_col_separator = true;
+ }
+ /* It's rather pointless to define a TAB separator with column
+@@ -1288,11 +1414,11 @@
+ TAB_WIDTH (chars_per_input_tab, chars_per_number); */
+
+ /* Estimate chars_per_text without any margin and keep it constant. */
+- if (number_separator == '\t')
++ if (number_separator[0] == '\t')
+ number_width = chars_per_number +
+ TAB_WIDTH (chars_per_default_tab, chars_per_number);
+ else
+- number_width = chars_per_number + 1;
++ number_width = chars_per_number + number_separator_width;
+
+ /* The number is part of the column width unless we are
+ printing files in parallel. */
+@@ -1307,7 +1433,7 @@
+ }
+
+ chars_per_column = (chars_per_line - chars_used_by_number -
+- (columns - 1) * col_sep_length) / columns;
++ (columns - 1) * col_sep_width) / columns;
+
+ if (chars_per_column < 1)
+ error (EXIT_FAILURE, 0, _("page width too narrow"));
+@@ -1432,7 +1558,7 @@
+
+ /* Enlarge p->start_position of first column to use the same form of
+ padding_not_printed with all columns. */
+- h = h + col_sep_length;
++ h = h + col_sep_width;
+
+ /* This loop takes care of all but the rightmost column. */
+
+@@ -1466,7 +1592,7 @@
+ }
+ else
+ {
+- h = h_next + col_sep_length;
++ h = h_next + col_sep_width;
+ h_next = h + chars_per_column;
+ }
+ }
+@@ -1756,9 +1882,9 @@
+ align_column (COLUMN *p)
+ {
+ padding_not_printed = p->start_position;
+- if (padding_not_printed - col_sep_length > 0)
++ if (padding_not_printed - col_sep_width > 0)
+ {
+- pad_across_to (padding_not_printed - col_sep_length);
++ pad_across_to (padding_not_printed - col_sep_width);
+ padding_not_printed = ANYWHERE;
+ }
+
+@@ -2029,13 +2155,13 @@
+ /* May be too generous. */
+ buff = X2REALLOC (buff, &buff_allocated);
+ }
+- buff[buff_current++] = c;
++ buff[buff_current++] = (unsigned char) c;
+ }
+
+ static void
+ add_line_number (COLUMN *p)
+ {
+- int i;
++ int i, j;
+ char *s;
+ int left_cut;
+
+@@ -2058,22 +2184,24 @@
+ /* Tabification is assumed for multiple columns, also for n-separators,
+ but `default n-separator = TAB' hasn't been given priority over
+ equal column_width also specified by POSIX. */
+- if (number_separator == '\t')
++ if (number_separator[0] == '\t')
+ {
+ i = number_width - chars_per_number;
+ while (i-- > 0)
+ (p->char_func) (' ');
+ }
+ else
+- (p->char_func) (number_separator);
++ for (j = 0; j < number_separator_length; j++)
++ (p->char_func) (number_separator[j]);
+ }
+ else
+ /* To comply with POSIX, we avoid any expansion of default TAB
+ separator with a single column output. No column_width requirement
+ has to be considered. */
+ {
+- (p->char_func) (number_separator);
+- if (number_separator == '\t')
++ for (j = 0; j < number_separator_length; j++)
++ (p->char_func) (number_separator[j]);
++ if (number_separator[0] == '\t')
+ output_position = POS_AFTER_TAB (chars_per_output_tab,
+ output_position);
+ }
+@@ -2234,7 +2362,7 @@
+ while (goal - h_old > 1
+ && (h_new = POS_AFTER_TAB (chars_per_output_tab, h_old)) <= goal)
+ {
+- putchar (output_tab_char);
++ fwrite (output_tab_char, sizeof(char), output_tab_char_length, stdout);
+ h_old = h_new;
+ }
+ while (++h_old <= goal)
+@@ -2254,6 +2382,7 @@
+ {
+ char *s;
+ int l = col_sep_length;
++ int not_space_flag;
+
+ s = col_sep_string;
+
+@@ -2267,6 +2396,7 @@
+ {
+ for (; separators_not_printed > 0; --separators_not_printed)
+ {
++ not_space_flag = 0;
+ while (l-- > 0)
+ {
+ /* 3 types of sep_strings: spaces only, spaces and chars,
+@@ -2280,12 +2410,15 @@
+ }
+ else
+ {
++ not_space_flag = 1;
+ if (spaces_not_printed > 0)
+ print_white_space ();
+ putchar (*s++);
+- ++output_position;
+ }
+ }
++ if (not_space_flag)
++ output_position += col_sep_width;
++
+ /* sep_string ends with some spaces */
+ if (spaces_not_printed > 0)
+ print_white_space ();
+@@ -2313,7 +2446,7 @@
+ required number of tabs and spaces. */
+
+ static void
+-print_char (char c)
++print_char_single (char c)
+ {
+ if (tabify_output)
+ {
+@@ -2337,6 +2470,74 @@
+ putchar (c);
+ }
+
++#ifdef HAVE_MBRTOWC
++static void
++print_char_multi (char c)
++{
++ static size_t mbc_pos = 0;
++ static char mbc[MB_LEN_MAX] = {'\0'};
++ static mbstate_t state = {'\0'};
++ mbstate_t state_bak;
++ wchar_t wc;
++ size_t mblength;
++ int width;
++
++ if (tabify_output)
++ {
++ state_bak = state;
++ mbc[mbc_pos++] = c;
++ mblength = mbrtowc (&wc, mbc, mbc_pos, &state);
++
++ while (mbc_pos > 0)
++ {
++ switch (mblength)
++ {
++ case (size_t)-2:
++ state = state_bak;
++ return;
++
++ case (size_t)-1:
++ state = state_bak;
++ ++output_position;
++ putchar (mbc[0]);
++ memmove (mbc, mbc + 1, MB_CUR_MAX - 1);
++ --mbc_pos;
++ break;
++
++ case 0:
++ mblength = 1;
++
++ default:
++ if (wc == L' ')
++ {
++ memmove (mbc, mbc + mblength, MB_CUR_MAX - mblength);
++ --mbc_pos;
++ ++spaces_not_printed;
++ return;
++ }
++ else if (spaces_not_printed > 0)
++ print_white_space ();
++
++ /* Nonprintables are assumed to have width 0, except L'\b'. */
++ if ((width = wcwidth (wc)) < 1)
++ {
++ if (wc == L'\b')
++ --output_position;
++ }
++ else
++ output_position += width;
++
++ fwrite (mbc, sizeof(char), mblength, stdout);
++ memmove (mbc, mbc + mblength, MB_CUR_MAX - mblength);
++ mbc_pos -= mblength;
++ }
++ }
++ return;
++ }
++ putchar (c);
++}
++#endif
++
+ /* Skip to page PAGE before printing.
+ PAGE may be larger than total number of pages. */
+
+@@ -2517,9 +2718,9 @@
+ align_empty_cols = false;
+ }
+
+- if (padding_not_printed - col_sep_length > 0)
++ if (padding_not_printed - col_sep_width > 0)
+ {
+- pad_across_to (padding_not_printed - col_sep_length);
++ pad_across_to (padding_not_printed - col_sep_width);
+ padding_not_printed = ANYWHERE;
+ }
+
+@@ -2620,9 +2821,9 @@
+ }
+ }
+
+- if (padding_not_printed - col_sep_length > 0)
++ if (padding_not_printed - col_sep_width > 0)
+ {
+- pad_across_to (padding_not_printed - col_sep_length);
++ pad_across_to (padding_not_printed - col_sep_width);
+ padding_not_printed = ANYWHERE;
+ }
+
+@@ -2635,8 +2836,8 @@
+ if (spaces_not_printed == 0)
+ {
+ output_position = p->start_position + end_vector[line];
+- if (p->start_position - col_sep_length == chars_per_margin)
+- output_position -= col_sep_length;
++ if (p->start_position - col_sep_width == chars_per_margin)
++ output_position -= col_sep_width;
+ }
+
+ return true;
+@@ -2655,7 +2856,7 @@
+ number of characters is 1.) */
+
+ static int
+-char_to_clump (char c)
++char_to_clump_single (char c)
+ {
+ unsigned char uc = c;
+ char *s = clump_buff;
+@@ -2665,10 +2866,10 @@
+ int chars;
+ int chars_per_c = 8;
+
+- if (c == input_tab_char)
++ if (c == input_tab_char[0])
+ chars_per_c = chars_per_input_tab;
+
+- if (c == input_tab_char || c == '\t')
++ if (c == input_tab_char[0] || c == '\t')
+ {
+ width = TAB_WIDTH (chars_per_c, input_position);
+
+@@ -2739,6 +2940,154 @@
+ return chars;
+ }
+
++#ifdef HAVE_MBRTOWC
++static int
++char_to_clump_multi (char c)
++{
++ static size_t mbc_pos = 0;
++ static char mbc[MB_LEN_MAX] = {'\0'};
++ static mbstate_t state = {'\0'};
++ mbstate_t state_bak;
++ wchar_t wc;
++ size_t mblength;
++ int wc_width;
++ register char *s = clump_buff;
++ register int i, j;
++ char esc_buff[4];
++ int width;
++ int chars;
++ int chars_per_c = 8;
++
++ state_bak = state;
++ mbc[mbc_pos++] = c;
++ mblength = mbrtowc (&wc, mbc, mbc_pos, &state);
++
++ width = 0;
++ chars = 0;
++ while (mbc_pos > 0)
++ {
++ switch (mblength)
++ {
++ case (size_t)-2:
++ state = state_bak;
++ return 0;
++
++ case (size_t)-1:
++ state = state_bak;
++ mblength = 1;
++
++ if (use_esc_sequence || use_cntrl_prefix)
++ {
++ width = +4;
++ chars = +4;
++ *s++ = '\\';
++ sprintf (esc_buff, "%03o", mbc[0]);
++ for (i = 0; i <= 2; ++i)
++ *s++ = (int) esc_buff[i];
++ }
++ else
++ {
++ width += 1;
++ chars += 1;
++ *s++ = mbc[0];
++ }
++ break;
++
++ case 0:
++ mblength = 1;
++ /* Fall through */
++
++ default:
++ if (memcmp (mbc, input_tab_char, mblength) == 0)
++ chars_per_c = chars_per_input_tab;
++
++ if (memcmp (mbc, input_tab_char, mblength) == 0 || c == '\t')
++ {
++ int width_inc;
++
++ width_inc = TAB_WIDTH (chars_per_c, input_position);
++ width += width_inc;
++
++ if (untabify_input)
++ {
++ for (i = width_inc; i; --i)
++ *s++ = ' ';
++ chars += width_inc;
++ }
++ else
++ {
++ for (i = 0; i < mblength; i++)
++ *s++ = mbc[i];
++ chars += mblength;
++ }
++ }
++ else if ((wc_width = wcwidth (wc)) < 1)
++ {
++ if (use_esc_sequence)
++ {
++ for (i = 0; i < mblength; i++)
++ {
++ width += 4;
++ chars += 4;
++ *s++ = '\\';
++ sprintf (esc_buff, "%03o", c);
++ for (j = 0; j <= 2; ++j)
++ *s++ = (int) esc_buff[j];
++ }
++ }
++ else if (use_cntrl_prefix)
++ {
++ if (wc < 0200)
++ {
++ width += 2;
++ chars += 2;
++ *s++ = '^';
++ *s++ = wc ^ 0100;
++ }
++ else
++ {
++ for (i = 0; i < mblength; i++)
++ {
++ width += 4;
++ chars += 4;
++ *s++ = '\\';
++ sprintf (esc_buff, "%03o", c);
++ for (j = 0; j <= 2; ++j)
++ *s++ = (int) esc_buff[j];
++ }
++ }
++ }
++ else if (wc == L'\b')
++ {
++ width += -1;
++ chars += 1;
++ *s++ = c;
++ }
++ else
++ {
++ width += 0;
++ chars += mblength;
++ for (i = 0; i < mblength; i++)
++ *s++ = mbc[i];
++ }
++ }
++ else
++ {
++ width += wc_width;
++ chars += mblength;
++ for (i = 0; i < mblength; i++)
++ *s++ = mbc[i];
++ }
++ }
++ memmove (mbc, mbc + mblength, MB_CUR_MAX - mblength);
++ mbc_pos -= mblength;
++ }
++
++ input_position += width;
++ return chars;
++}
++#endif
++
+ /* We've just printed some files and need to clean up things before
+ looking for more options and printing the next batch of files.
+
+--- coreutils-6.8+/src/cut.c.i18n 2007-01-14 15:41:28.000000000 +0000
++++ coreutils-6.8+/src/cut.c 2007-03-01 15:08:24.000000000 +0000
+@@ -29,6 +29,11 @@
+ #include <assert.h>
+ #include <getopt.h>
+ #include <sys/types.h>
++
++/* Get mbstate_t, mbrtowc(). */
++#if HAVE_WCHAR_H
++# include <wchar.h>
++#endif
+ #include "system.h"
+
+ #include "error.h"
+@@ -37,6 +42,18 @@
+ #include "quote.h"
+ #include "xstrndup.h"
+
++/* MB_LEN_MAX is incorrectly defined to be 1 in at least one GCC
++ installation; work around this configuration error. */
++#if !defined MB_LEN_MAX || MB_LEN_MAX < 2
++# undef MB_LEN_MAX
++# define MB_LEN_MAX 16
++#endif
++
++/* Some systems, like BeOS, have multibyte encodings but lack mbstate_t. */
++#if HAVE_MBRTOWC && defined mbstate_t
++# define mbrtowc(pwc, s, n, ps) (mbrtowc) (pwc, s, n, 0)
++#endif
++
+ /* The official name of this program (e.g., no `g' prefix). */
+ #define PROGRAM_NAME "cut"
+
+@@ -67,6 +84,52 @@
+ } \
+ while (0)
+
++/* Refill the buffer BUF to get a multibyte character. */
++#define REFILL_BUFFER(BUF, BUFPOS, BUFLEN, STREAM) \
++ do \
++ { \
++ if (BUFLEN < MB_LEN_MAX && !feof (STREAM) && !ferror (STREAM)) \
++ { \
++ memmove (BUF, BUFPOS, BUFLEN); \
++ BUFLEN += fread (BUF + BUFLEN, sizeof(char), BUFSIZ, STREAM); \
++ BUFPOS = BUF; \
++ } \
++ } \
++ while (0)
++
++/* Get wide character on BUFPOS. BUFPOS is not included after that.
++ If byte sequence is not valid as a character, CONVFAIL is 1. Otherwise 0. */
++#define GET_NEXT_WC_FROM_BUFFER(WC, BUFPOS, BUFLEN, MBLENGTH, STATE, CONVFAIL) \
++ do \
++ { \
++ mbstate_t state_bak; \
++ \
++ if (BUFLEN < 1) \
++ { \
++ WC = WEOF; \
++ break; \
++ } \
++ \
++ /* Get a wide character. */ \
++ CONVFAIL = 0; \
++ state_bak = STATE; \
++ MBLENGTH = mbrtowc ((wchar_t *)&WC, BUFPOS, BUFLEN, &STATE); \
++ \
++ switch (MBLENGTH) \
++ { \
++ case (size_t)-1: \
++ case (size_t)-2: \
++ CONVFAIL++; \
++ STATE = state_bak; \
++ /* Fall througn. */ \
++ \
++ case 0: \
++ MBLENGTH = 1; \
++ break; \
++ } \
++ } \
++ while (0)
++
+ struct range_pair
+ {
+ size_t lo;
+@@ -85,7 +148,7 @@
+ /* The number of bytes allocated for FIELD_1_BUFFER. */
+ static size_t field_1_bufsize;
+
+-/* The largest field or byte index used as an endpoint of a closed
++/* The largest byte, character or field index used as an endpoint of a closed
+ or degenerate range specification; this doesn't include the starting
+ index of right-open-ended ranges. For example, with either range spec
+ `2-5,9-', `2-3,5,9-' this variable would be set to 5. */
+@@ -97,10 +160,11 @@
+
+ /* This is a bit vector.
+ In byte mode, which bytes to output.
++ In character mode, which characters to output.
+ In field mode, which DELIM-separated fields to output.
+- Both bytes and fields are numbered starting with 1,
++ Bytes, characters and fields are numbered starting with 1,
+ so the zeroth bit of this array is unused.
+- A field or byte K has been selected if
++ A byte, character or field K has been selected if
+ (K <= MAX_RANGE_ENDPOINT and is_printable_field(K))
+ || (EOL_RANGE_START > 0 && K >= EOL_RANGE_START). */
+ static unsigned char *printable_field;
+@@ -109,9 +173,12 @@
+ {
+ undefined_mode,
+
+- /* Output characters that are in the given bytes. */
++ /* Output bytes that are at the given positions. */
+ byte_mode,
+
++ /* Output characters that are at the given positions. */
++ character_mode,
++
+ /* Output the given delimeter-separated fields. */
+ field_mode
+ };
+@@ -121,6 +188,13 @@
+
+ static enum operating_mode operating_mode;
+
++/* If nonzero, when in byte mode, don't split multibyte characters. */
++static int byte_mode_character_aware;
++
++/* If nonzero, the function for single byte locale is work
++ if this program runs on multibyte locale. */
++static int force_singlebyte_mode;
++
+ /* If true do not output lines containing no delimeter characters.
+ Otherwise, all such lines are printed. This option is valid only
+ with field mode. */
+@@ -132,6 +206,9 @@
+
+ /* The delimeter character for field mode. */
+ static unsigned char delim;
++#if HAVE_WCHAR_H
++static wchar_t wcdelim;
++#endif
+
+ /* True if the --output-delimiter=STRING option was specified. */
+ static bool output_delimiter_specified;
+@@ -205,7 +282,7 @@
+ -f, --fields=LIST select only these fields; also print any line\n\
+ that contains no delimiter character, unless\n\
+ the -s option is specified\n\
+- -n (ignored)\n\
++ -n with -b: don't split multibyte characters\n\
+ "), stdout);
+ fputs (_("\
+ --complement complement the set of selected bytes, characters\n\
+@@ -362,7 +439,7 @@
+ in_digits = false;
+ /* Starting a range. */
+ if (dash_found)
+- FATAL_ERROR (_("invalid byte or field list"));
++ FATAL_ERROR (_("invalid byte, character or field list"));
+ dash_found = true;
+ fieldstr++;
+
+@@ -387,14 +464,16 @@
+ if (value == 0)
+ {
+ /* `n-'. From `initial' to end of line. */
+- eol_range_start = initial;
++ if (eol_range_start == 0 ||
++ (eol_range_start != 0 && eol_range_start > initial))
++ eol_range_start = initial;
+ field_found = true;
+ }
+ else
+ {
+ /* `m-n' or `-n' (1-n). */
+ if (value < initial)
+- FATAL_ERROR (_("invalid byte or field list"));
++ FATAL_ERROR (_("invalid byte, character or field list"));
+
+ /* Is there already a range going to end of line? */
+ if (eol_range_start != 0)
+@@ -467,6 +546,9 @@
+ if (operating_mode == byte_mode)
+ error (0, 0,
+ _("byte offset %s is too large"), quote (bad_num));
++ else if (operating_mode == character_mode)
++ error (0, 0,
++ _("character offset %s is too large"), quote (bad_num));
+ else
+ error (0, 0,
+ _("field number %s is too large"), quote (bad_num));
+@@ -477,7 +559,7 @@
+ fieldstr++;
+ }
+ else
+- FATAL_ERROR (_("invalid byte or field list"));
++ FATAL_ERROR (_("invalid byte, character or field list"));
+ }
+
+ max_range_endpoint = 0;
+@@ -570,6 +652,63 @@
+ }
+ }
+
++#if HAVE_MBRTOWC
++/* This function is in use for the following case.
++
++ 1. Read from the stream STREAM, printing to standard output any selected
++ characters.
++
++ 2. Read from stream STREAM, printing to standard output any selected bytes,
++ without splitting multibyte characters. */
++
++static void
++cut_characters_or_cut_bytes_no_split (FILE *stream)
++{
++ int idx; /* number of bytes or characters in the line so far. */
++ char buf[MB_LEN_MAX + BUFSIZ]; /* For spooling a read byte sequence. */
++ char *bufpos; /* Next read position of BUF. */
++ size_t buflen; /* The length of the byte sequence in buf. */
++ wint_t wc; /* A gotten wide character. */
++ size_t mblength; /* The byte size of a multibyte character which shows
++ as same character as WC. */
++ mbstate_t state; /* State of the stream. */
++ int convfail; /* 1, when conversion is failed. Otherwise 0. */
++
++ idx = 0;
++ buflen = 0;
++ bufpos = buf;
++ memset (&state, '\0', sizeof(mbstate_t));
++
++ while (1)
++ {
++ REFILL_BUFFER (buf, bufpos, buflen, stream);
++
++ GET_NEXT_WC_FROM_BUFFER (wc, bufpos, buflen, mblength, state, convfail);
++
++ if (wc == WEOF)
++ {
++ if (idx > 0)
++ putchar ('\n');
++ break;
++ }
++ else if (wc == L'\n')
++ {
++ putchar ('\n');
++ idx = 0;
++ }
++ else
++ {
++ idx += (operating_mode == byte_mode) ? mblength : 1;
++ if (print_kth (idx, NULL))
++ fwrite (bufpos, mblength, sizeof(char), stdout);
++ }
++
++ buflen -= mblength;
++ bufpos += mblength;
++ }
++}
++#endif
++
+ /* Read from stream STREAM, printing to standard output any selected fields. */
+
+ static void
+@@ -692,13 +831,192 @@
+ }
+ }
+
++#if HAVE_MBRTOWC
++static void
++cut_fields_mb (FILE *stream)
++{
++ int c;
++ unsigned int field_idx;
++ int found_any_selected_field;
++ int buffer_first_field;
++ int empty_input;
++ char buf[MB_LEN_MAX + BUFSIZ]; /* For spooling a read byte sequence. */
++ char *bufpos; /* Next read position of BUF. */
++ size_t buflen; /* The length of the byte sequence in buf. */
++ wint_t wc = 0; /* A gotten wide character. */
++ size_t mblength; /* The byte size of a multibyte character which shows
++ as same character as WC. */
++ mbstate_t state; /* State of the stream. */
++ int convfail; /* 1, when conversion is failed. Otherwise 0. */
++
++ found_any_selected_field = 0;
++ field_idx = 1;
++ bufpos = buf;
++ buflen = 0;
++ memset (&state, '\0', sizeof(mbstate_t));
++
++ c = getc (stream);
++ empty_input = (c == EOF);
++ if (c != EOF)
++ ungetc (c, stream);
++ else
++ wc = WEOF;
++
++ /* To support the semantics of the -s flag, we may have to buffer
++ all of the first field to determine whether it is `delimited.'
++ But that is unnecessary if all non-delimited lines must be printed
++ and the first field has been selected, or if non-delimited lines
++ must be suppressed and the first field has *not* been selected.
++ That is because a non-delimited line has exactly one field. */
++ buffer_first_field = (suppress_non_delimited ^ !print_kth (1, NULL));
++
++ while (1)
++ {
++ if (field_idx == 1 && buffer_first_field)
++ {
++ int len = 0;
++
++ while (1)
++ {
++ REFILL_BUFFER (buf, bufpos, buflen, stream);
++
++ GET_NEXT_WC_FROM_BUFFER
++ (wc, bufpos, buflen, mblength, state, convfail);
++
++ if (wc == WEOF)
++ break;
++
++ field_1_buffer = xrealloc (field_1_buffer, len + mblength);
++ memcpy (field_1_buffer + len, bufpos, mblength);
++ len += mblength;
++ buflen -= mblength;
++ bufpos += mblength;
++
++ if (!convfail && (wc == L'\n' || wc == wcdelim))
++ break;
++ }
++
++ if (wc == WEOF)
++ break;
++
++ /* If the first field extends to the end of line (it is not
++ delimited) and we are printing all non-delimited lines,
++ print this one. */
++ if (convfail || (!convfail && wc != wcdelim))
++ {
++ if (suppress_non_delimited)
++ {
++ /* Empty. */
++ }
++ else
++ {
++ fwrite (field_1_buffer, sizeof (char), len, stdout);
++ /* Make sure the output line is newline terminated. */
++ if (convfail || (!convfail && wc != L'\n'))
++ putchar ('\n');
++ }
++ continue;
++ }
++
++ if (print_kth (1, NULL))
++ {
++ /* Print the field, but not the trailing delimiter. */
++ fwrite (field_1_buffer, sizeof (char), len - 1, stdout);
++ found_any_selected_field = 1;
++ }
++ ++field_idx;
++ }
++
++ if (wc != WEOF)
++ {
++ if (print_kth (field_idx, NULL))
++ {
++ if (found_any_selected_field)
++ {
++ fwrite (output_delimiter_string, sizeof (char),
++ output_delimiter_length, stdout);
++ }
++ found_any_selected_field = 1;
++ }
++
++ while (1)
++ {
++ REFILL_BUFFER (buf, bufpos, buflen, stream);
++
++ GET_NEXT_WC_FROM_BUFFER
++ (wc, bufpos, buflen, mblength, state, convfail);
++
++ if (wc == WEOF)
++ break;
++ else if (!convfail && (wc == wcdelim || wc == L'\n'))
++ {
++ buflen -= mblength;
++ bufpos += mblength;
++ break;
++ }
++
++ if (print_kth (field_idx, NULL))
++ fwrite (bufpos, mblength, sizeof(char), stdout);
++
++ buflen -= mblength;
++ bufpos += mblength;
++ }
++ }
++
++ if ((!convfail || wc == L'\n') && buflen < 1)
++ wc = WEOF;
++
++ if (!convfail && wc == wcdelim)
++ ++field_idx;
++ else if (wc == WEOF || (!convfail && wc == L'\n'))
++ {
++ if (found_any_selected_field
++ || (!empty_input && !(suppress_non_delimited && field_idx == 1)))
++ putchar ('\n');
++ if (wc == WEOF)
++ break;
++ field_idx = 1;
++ found_any_selected_field = 0;
++ }
++ }
++}
++#endif
++
+ static void
+ cut_stream (FILE *stream)
+ {
+- if (operating_mode == byte_mode)
+- cut_bytes (stream);
++#if HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1 && !force_singlebyte_mode)
++ {
++ switch (operating_mode)
++ {
++ case byte_mode:
++ if (byte_mode_character_aware)
++ cut_characters_or_cut_bytes_no_split (stream);
++ else
++ cut_bytes (stream);
++ break;
++
++ case character_mode:
++ cut_characters_or_cut_bytes_no_split (stream);
++ break;
++
++ case field_mode:
++ cut_fields_mb (stream);
++ break;
++
++ default:
++ abort ();
++ }
++ }
+ else
+- cut_fields (stream);
++#endif
++ {
++ if (operating_mode == field_mode)
++ cut_fields (stream);
++ else
++ cut_bytes (stream);
++ }
+ }
+
+ /* Process file FILE to standard output.
+@@ -748,6 +1066,8 @@
+ bool ok;
+ bool delim_specified = false;
+ char *spec_list_string IF_LINT(= NULL);
++ char mbdelim[MB_LEN_MAX + 1];
++ size_t delimlen = 0;
+
+ initialize_main (&argc, &argv);
+ program_name = argv[0];
+@@ -770,7 +1090,6 @@
+ switch (optc)
+ {
+ case 'b':
+- case 'c':
+ /* Build the byte list. */
+ if (operating_mode != undefined_mode)
+ FATAL_ERROR (_("only one type of list may be specified"));
+@@ -778,6 +1097,14 @@
+ spec_list_string = optarg;
+ break;
+
++ case 'c':
++ /* Build the character list. */
++ if (operating_mode != undefined_mode)
++ FATAL_ERROR (_("only one type of list may be specified"));
++ operating_mode = character_mode;
++ spec_list_string = optarg;
++ break;
++
+ case 'f':
+ /* Build the field list. */
+ if (operating_mode != undefined_mode)
+@@ -789,10 +1116,35 @@
+ case 'd':
+ /* New delimiter. */
+ /* Interpret -d '' to mean `use the NUL byte as the delimiter.' */
+- if (optarg[0] != '\0' && optarg[1] != '\0')
+- FATAL_ERROR (_("the delimiter must be a single character"));
+- delim = optarg[0];
+- delim_specified = true;
++#if HAVE_MBRTOWC
++ {
++ if(MB_CUR_MAX > 1)
++ {
++ mbstate_t state;
++
++ memset (&state, '\0', sizeof(mbstate_t));
++ delimlen = mbrtowc (&wcdelim, optarg, strnlen(optarg, MB_LEN_MAX), &state);
++
++ if (delimlen == (size_t)-1 || delimlen == (size_t)-2)
++ ++force_singlebyte_mode;
++ else
++ {
++ delimlen = (delimlen < 1) ? 1 : delimlen;
++ if (wcdelim != L'\0' && *(optarg + delimlen) != '\0')
++ FATAL_ERROR (_("the delimiter must be a single character"));
++ memcpy (mbdelim, optarg, delimlen);
++ }
++ }
++
++ if (MB_CUR_MAX <= 1 || force_singlebyte_mode)
++#endif
++ {
++ if (optarg[0] != '\0' && optarg[1] != '\0')
++ FATAL_ERROR (_("the delimiter must be a single character"));
++ delim = (unsigned char) optarg[0];
++ }
++ delim_specified = true;
++ }
+ break;
+
+ case OUTPUT_DELIMITER_OPTION:
+@@ -805,6 +1157,7 @@
+ break;
+
+ case 'n':
++ byte_mode_character_aware = 1;
+ break;
+
+ case 's':
+@@ -827,7 +1180,7 @@
+ if (operating_mode == undefined_mode)
+ FATAL_ERROR (_("you must specify a list of bytes, characters, or fields"));
+
+- if (delim != '\0' && operating_mode != field_mode)
++ if (delim_specified && operating_mode != field_mode)
+ FATAL_ERROR (_("an input delimiter may be specified only\
+ when operating on fields"));
+
+@@ -854,15 +1207,34 @@
+ }
+
+ if (!delim_specified)
+- delim = '\t';
++ {
++ delim = '\t';
++#ifdef HAVE_MBRTOWC
++ wcdelim = L'\t';
++ mbdelim[0] = '\t';
++ mbdelim[1] = '\0';
++ delimlen = 1;
++#endif
++ }
+
+ if (output_delimiter_string == NULL)
+ {
+- static char dummy[2];
+- dummy[0] = delim;
+- dummy[1] = '\0';
+- output_delimiter_string = dummy;
+- output_delimiter_length = 1;
++#ifdef HAVE_MBRTOWC
++ if (MB_CUR_MAX > 1 && !force_singlebyte_mode)
++ {
++ output_delimiter_string = xstrdup(mbdelim);
++ output_delimiter_length = delimlen;
++ }
++
++ if (MB_CUR_MAX <= 1 || force_singlebyte_mode)
++#endif
++ {
++ static char dummy[2];
++ dummy[0] = delim;
++ dummy[1] = '\0';
++ output_delimiter_string = dummy;
++ output_delimiter_length = 1;
++ }
+ }
+
+ if (optind == argc)
diff --git a/meta/recipes-core/coreutils/coreutils-6.9/coreutils-ls-x.patch b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-ls-x.patch
new file mode 100644
index 0000000..2d61438
--- /dev/null
+++ b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-ls-x.patch
@@ -0,0 +1,115 @@
+This patch was imported from the Fedora Core 8 coreutils-6.9-9 package.
+
+The package is stated as being Licensed as GPLv2+.
+
+Signed-off-by: Mark Hatle <mark.hatle at windriver.com>
+
+--- coreutils-6.9/src/ls.c.ls-x 2007-06-13 14:27:36.000000000 +0100
++++ coreutils-6.9/src/ls.c 2007-06-13 14:28:42.000000000 +0100
+@@ -4151,16 +4151,16 @@
+ size_t pos = 0;
+ size_t cols = calculate_columns (false);
+ struct column_info const *line_fmt = &column_info[cols - 1];
+- size_t name_length = length_of_file_name_and_frills (cwd_file);
++ struct fileinfo const *f = sorted_file[0];
++ size_t name_length = length_of_file_name_and_frills (f);
+ size_t max_name_length = line_fmt->col_arr[0];
+
+ /* Print first entry. */
+- print_file_name_and_frills (cwd_file);
++ print_file_name_and_frills (f);
+
+ /* Now the rest. */
+ for (filesno = 1; filesno < cwd_n_used; ++filesno)
+ {
+- struct fileinfo const *f;
+ size_t col = filesno % cols;
+
+ if (col == 0)
+--- coreutils-6.9/tests/ls/Makefile.am.ls-x 2007-03-18 21:36:43.000000000 +0000
++++ coreutils-6.9/tests/ls/Makefile.am 2007-06-13 14:28:42.000000000 +0100
+@@ -24,7 +24,7 @@
+ stat-dtype \
+ inode dangle file-type recursive dired infloop \
+ rt-1 time-1 symlink-slash follow-slink no-arg m-option \
+- stat-vs-dirent
++ stat-vs-dirent x-option
+
+ EXTRA_DIST = $(TESTS)
+ TESTS_ENVIRONMENT = \
+--- /dev/null 2007-06-13 08:43:51.993263382 +0100
++++ coreutils-6.9/tests/ls/x-option 2007-06-13 14:28:42.000000000 +0100
+@@ -0,0 +1,59 @@
++#!/bin/sh
++# Exercise the -x option.
++
++# Copyright (C) 2007 Free Software Foundation, Inc.
++
++# This program is free software; you can redistribute it and/or modify
++# it under the terms of the GNU General Public License as published by
++# the Free Software Foundation; either version 2 of the License, or
++# (at your option) any later version.
++
++# This program is distributed in the hope that it will be useful,
++# but WITHOUT ANY WARRANTY; without even the implied warranty of
++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++# GNU General Public License for more details.
++
++# You should have received a copy of the GNU General Public License
++# along with this program; if not, write to the Free Software
++# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
++# 02110-1301, USA.
++
++if test "$VERBOSE" = yes; then
++ set -x
++ ls --version
++fi
++
++. $srcdir/../envvar-check
++. $srcdir/../lang-default
++
++pwd=`pwd`
++t0=`echo "$0"|sed 's,.*/,,'`.tmp; tmp=$t0/$$
++trap 'status=$?; cd "$pwd" && chmod -R u+rwx $t0 && rm -rf $t0 && exit $status' 0
++trap '(exit $?); exit $?' 1 2 13 15
++
++framework_failure=0
++mkdir -p $tmp || framework_failure=1
++cd $tmp || framework_failure=1
++mkdir subdir || framework_failure=1
++touch subdir/b || framework_failure=1
++touch subdir/a || framework_failure=1
++
++if test $framework_failure = 1; then
++ echo "$0: failure in testing framework" 1>&2
++ (exit 1); exit 1
++fi
++
++fail=0
++
++# Coreutils 6.8 and 6.9 would output this in the wrong order.
++ls -x subdir > out || fail=1
++ls -rx subdir >> out || fail=1
++cat <<\EOF > exp || fail=1
++a b
++b a
++EOF
++
++cmp out exp || fail=1
++test $fail = 1 && diff out exp 2> /dev/null
++
++(exit $fail); exit $fail
+--- coreutils-6.9/NEWS.ls-x 2007-03-22 21:19:45.000000000 +0000
++++ coreutils-6.9/NEWS 2007-06-13 14:28:42.000000000 +0100
+@@ -13,6 +13,11 @@
+ Using pr -m -s (i.e. merging files, with TAB as the output separator)
+ no longer inserts extraneous spaces between output columns.
+
++** Bug fixes
++
++ ls -x DIR would sometimes output the wrong string in place of the
++ first entry. [introduced in coreutils-6.8]
++
+
+ * Noteworthy changes in release 6.8 (2007-02-24) [not-unstable]
+
diff --git a/meta/recipes-core/coreutils/coreutils-6.9/coreutils-overflow.patch b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-overflow.patch
new file mode 100644
index 0000000..d702757
--- /dev/null
+++ b/meta/recipes-core/coreutils/coreutils-6.9/coreutils-overflow.patch
@@ -0,0 +1,17 @@
+This patch was imported from the Fedora Core 8 coreutils-6.9-9 package.
+
+The package is stated as being Licensed as GPLv2+.
+
+Signed-off-by: Mark Hatle <mark.hatle at windriver.com>
+
+--- coreutils-5.2.1/src/who.c.overflow 2005-05-25 09:59:06.000000000 +0100
++++ coreutils-5.2.1/src/who.c 2005-05-25 10:00:31.000000000 +0100
+@@ -75,7 +75,7 @@
+ # define NEW_TIME 0
+ #endif
+
+-#define IDLESTR_LEN 6
++#define IDLESTR_LEN 10
+
+ #if HAVE_STRUCT_XTMP_UT_PID
+ # define PIDSTR_DECL_AND_INIT(Var, Utmp_ent) \
diff --git a/meta/recipes-core/coreutils/coreutils_6.9.bb b/meta/recipes-core/coreutils/coreutils_6.9.bb
index 89f420a..60ea903 100644
--- a/meta/recipes-core/coreutils/coreutils_6.9.bb
+++ b/meta/recipes-core/coreutils/coreutils_6.9.bb
@@ -16,6 +16,11 @@ inherit autotools gettext
SRC_URI = "http://ftp.gnu.org/gnu/coreutils/${BP}.tar.bz2 \
file://gnulib_m4.patch \
file://futimens.patch \
+ file://coreutils-ls-x.patch \
+ file://coreutils-6.9-cp-i-u.patch \
+ file://coreutils-i18n.patch \
+ file://coreutils-overflow.patch \
+ file://coreutils-fix-install.patch \
file://man-touch.patch"
# [ gets a special treatment and is not included in this
--
1.7.3.4
More information about the poky
mailing list