@@ -4,7 +4,7 @@
###############################################################################
name = coreutils
-version = 8.30
+version = 8.31
release = 1
groups = System/Base
@@ -23,7 +23,7 @@ sources = %{thisapp}.tar.xz
build
requires
autoconf
- automake
+ automake >= 1.15
e2fsprogs-devel
gettext-devel
gmp-devel
@@ -43,6 +43,10 @@ build
--enable-no-install-program=hostname,kill,uptime \
--with-tty-group
+ prepare_cmds
+ autoreconf -vfi
+ end
+
test
make check RUN_VERY_EXPENSIVE_TESTS=yes || true
end
new file mode 100644
@@ -0,0 +1,565 @@
+diff --git a/src/cut.c b/src/cut.c
+index 7ab6be4..022d0ad 100644
+--- a/src/cut.c
++++ b/src/cut.c
+@@ -28,6 +28,11 @@
+ #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"
+@@ -38,6 +43,18 @@
+
+ #include "set-fields.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"
+
+@@ -54,6 +71,52 @@
+ } \
+ 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 true. Otherwise false. */
++#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 = false; \
++ state_bak = STATE; \
++ MBLENGTH = mbrtowc ((wchar_t *)&WC, BUFPOS, BUFLEN, &STATE); \
++ \
++ switch (MBLENGTH) \
++ { \
++ case (size_t)-1: \
++ case (size_t)-2: \
++ CONVFAIL = true; \
++ STATE = state_bak; \
++ /* Fall througn. */ \
++ \
++ case 0: \
++ MBLENGTH = 1; \
++ break; \
++ } \
++ } \
++ while (0)
++
+
+ /* Pointer inside RP. When checking if a byte or field is selected
+ by a finite range, we check if it is between CURRENT_RP.LO
+@@ -61,6 +124,9 @@
+ CURRENT_RP.HI then we make CURRENT_RP to point to the next range pair. */
+ static struct field_range_pair *current_rp;
+
++/* Length of the delimiter given as argument to -d. */
++size_t delimlen;
++
+ /* This buffer is used to support the semantics of the -s option
+ (or lack of same) when the specified field list includes (does
+ not include) the first field. In both of those cases, the entire
+@@ -77,15 +143,25 @@ enum operating_mode
+ {
+ 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 delimiter-separated fields. */
+ field_mode
+ };
+
+ 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 delimiter characters.
+ Otherwise, all such lines are printed. This option is valid only
+ with field mode. */
+@@ -97,6 +173,9 @@ static bool complement;
+
+ /* The delimiter character for field mode. */
+ static unsigned char delim;
++#if HAVE_WCHAR_H
++static wchar_t wcdelim;
++#endif
+
+ /* The delimiter for each line/record. */
+ static unsigned char line_delim = '\n';
+@@ -164,7 +243,7 @@ Print selected parts of lines from each FILE to standard output.\n\
+ -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\
+@@ -280,6 +359,82 @@ cut_bytes (FILE *stream)
+ }
+ }
+
++#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)
++{
++ uintmax_t 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. */
++ bool convfail = false; /* true, when conversion failed. Otherwise false. */
++ /* Whether to begin printing delimiters between ranges for the current line.
++ Set after we've begun printing data corresponding to the first range. */
++ bool print_delimiter = false;
++
++ idx = 0;
++ buflen = 0;
++ bufpos = buf;
++ memset (&state, '\0', sizeof(mbstate_t));
++
++ current_rp = frp;
++
++ while (1)
++ {
++ REFILL_BUFFER (buf, bufpos, buflen, stream);
++
++ GET_NEXT_WC_FROM_BUFFER (wc, bufpos, buflen, mblength, state, convfail);
++ (void) convfail; /* ignore unused */
++
++ if (wc == WEOF)
++ {
++ if (idx > 0)
++ putchar (line_delim);
++ break;
++ }
++ else if (wc == line_delim)
++ {
++ putchar (line_delim);
++ idx = 0;
++ print_delimiter = false;
++ current_rp = frp;
++ }
++ else
++ {
++ next_item (&idx);
++ if (print_kth (idx))
++ {
++ if (output_delimiter_specified)
++ {
++ if (print_delimiter && is_range_start_index (idx))
++ {
++ fwrite (output_delimiter_string, sizeof (char),
++ output_delimiter_length, stdout);
++ }
++ print_delimiter = true;
++ }
++ fwrite (bufpos, mblength, sizeof(char), stdout);
++ }
++ }
++
++ buflen -= mblength;
++ bufpos += mblength;
++ }
++}
++#endif
++
+ /* Read from stream STREAM, printing to standard output any selected fields. */
+
+ static void
+@@ -425,13 +580,211 @@ cut_fields (FILE *stream)
+ }
+ }
+
++#if HAVE_MBRTOWC
++static void
++cut_fields_mb (FILE *stream)
++{
++ int c;
++ uintmax_t 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. */
++ bool convfail = false; /* true, when conversion failed. Otherwise false. */
++
++ current_rp = frp;
++
++ 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);
++ wc = 0;
++ }
++ 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));
++
++ 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 == line_delim || wc == wcdelim))
++ break;
++ }
++
++ if (len <= 0 && 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 != line_delim))
++ putchar (line_delim);
++ }
++ continue;
++ }
++
++ if (print_kth (1))
++ {
++ /* Print the field, but not the trailing delimiter. */
++ fwrite (field_1_buffer, sizeof (char), len - 1, stdout);
++ found_any_selected_field = 1;
++ }
++ next_item (&field_idx);
++ }
++
++ if (wc != WEOF)
++ {
++ if (print_kth (field_idx))
++ {
++ 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 == line_delim))
++ {
++ buflen -= mblength;
++ bufpos += mblength;
++ break;
++ }
++
++ if (print_kth (field_idx))
++ fwrite (bufpos, mblength, sizeof(char), stdout);
++
++ buflen -= mblength;
++ bufpos += mblength;
++ }
++ }
++
++ if ((!convfail || wc == line_delim) && buflen < 1)
++ wc = WEOF;
++
++ if (!convfail && wc == wcdelim)
++ next_item (&field_idx);
++ else if (wc == WEOF || (!convfail && wc == line_delim))
++ {
++ if (found_any_selected_field
++ || (!empty_input && !(suppress_non_delimited && field_idx == 1)))
++ putchar (line_delim);
++ if (wc == WEOF)
++ break;
++ field_idx = 1;
++ current_rp = frp;
++ 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:
++ if (delimlen == 1)
++ {
++ /* Check if we have utf8 multibyte locale, so we can use this
++ optimization because of uniqueness of characters, which is
++ not true for e.g. SJIS */
++ char * loc = setlocale(LC_CTYPE, NULL);
++ if (loc && (strstr (loc, "UTF-8") || strstr (loc, "utf-8") ||
++ strstr (loc, "UTF8") || strstr (loc, "utf8")))
++ {
++ cut_fields (stream);
++ break;
++ }
++ }
++ 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.
+@@ -483,6 +836,7 @@ main (int argc, char **argv)
+ bool ok;
+ bool delim_specified = false;
+ char *spec_list_string IF_LINT ( = NULL);
++ char mbdelim[MB_LEN_MAX + 1];
+
+ initialize_main (&argc, &argv);
+ set_program_name (argv[0]);
+@@ -505,7 +859,6 @@ main (int argc, char **argv)
+ 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"));
+@@ -513,6 +866,14 @@ main (int argc, char **argv)
+ 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)
+@@ -524,10 +885,38 @@ main (int argc, char **argv)
+ 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);
++ mbdelim[delimlen] = '\0';
++ if (delimlen == 1)
++ delim = *optarg;
++ }
++ }
++
++ 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:
+@@ -540,6 +929,7 @@ main (int argc, char **argv)
+ break;
+
+ case 'n':
++ byte_mode_character_aware = 1;
+ break;
+
+ case 's':
+@@ -579,15 +969,34 @@ main (int argc, char **argv)
+ | (complement ? SETFLD_COMPLEMENT : 0) );
+
+ 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)
new file mode 100644
@@ -0,0 +1,848 @@
+From e87ab5b991b08092a7e07af82b3ec822a8604151 Mon Sep 17 00:00:00 2001
+From: Ondrej Oprala <ooprala@redhat.com>
+Date: Wed, 5 Aug 2015 09:15:09 +0200
+Subject: [PATCH] expand,unexpand: add multibyte support
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+* NEWS: Mention the changes.
+* bootstrap.conf: Add mbfile to the list of modules.
+* configure.ac: Properly initialize mbfile.
+* src/expand.c (expand): Iterate over multibyte characters properly.
+* src/unexpand.c (unexpand): Iterate over multibyte characters
+properly.
+* tests/local.mk: Add new tests.
+* tests/{expand,unexpand}/mb.sh: New tests.
+
+Co-authored-by: Pádraig Brady <pbrady@redhat.com>
+---
+ bootstrap.conf | 1 +
+ configure.ac | 2 +
+ lib/mbfile.c | 3 +
+ lib/mbfile.h | 255 +++++++++++++++++++++++++++++++++++++++++++++++++++
+ m4/mbfile.m4 | 14 +++
+ src/expand.c | 43 +++++----
+ src/unexpand.c | 54 +++++++----
+ tests/expand/mb.sh | 98 ++++++++++++++++++++
+ tests/local.mk | 2 +
+ tests/unexpand/mb.sh | 97 ++++++++++++++++++++
+ 10 files changed, 535 insertions(+), 34 deletions(-)
+ create mode 100644 lib/mbfile.c
+ create mode 100644 lib/mbfile.h
+ create mode 100644 m4/mbfile.m4
+ create mode 100755 tests/expand/mb.sh
+ create mode 100755 tests/unexpand/mb.sh
+
+diff --git a/bootstrap.conf b/bootstrap.conf
+index 8a0ff31..a1c78b2 100644
+--- a/bootstrap.conf
++++ b/bootstrap.conf
+@@ -152,6 +152,7 @@ gnulib_modules="
+ maintainer-makefile
+ malloc-gnu
+ manywarnings
++ mbfile
+ mbrlen
+ mbrtowc
+ mbsalign
+diff --git a/configure.ac b/configure.ac
+index 1e74b36..24c9725 100644
+--- a/configure.ac
++++ b/configure.ac
+@@ -427,6 +427,8 @@ fi
+ # I'm leaving it here for now. This whole thing needs to be modernized...
+ gl_WINSIZE_IN_PTEM
+
++gl_MBFILE
++
+ gl_HEADER_TIOCGWINSZ_IN_TERMIOS_H
+
+ if test $gl_cv_sys_tiocgwinsz_needs_termios_h = no && \
+diff --git a/lib/mbfile.c b/lib/mbfile.c
+new file mode 100644
+index 0000000..b0a468e
+--- /dev/null
++++ b/lib/mbfile.c
+@@ -0,0 +1,3 @@
++#include <config.h>
++#define MBFILE_INLINE _GL_EXTERN_INLINE
++#include "mbfile.h"
+diff --git a/lib/mbfile.h b/lib/mbfile.h
+new file mode 100644
+index 0000000..11f1b12
+--- /dev/null
++++ b/lib/mbfile.h
+@@ -0,0 +1,255 @@
++/* Multibyte character I/O: macros for multi-byte encodings.
++ Copyright (C) 2001, 2005, 2009-2015 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 3 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, see <http://www.gnu.org/licenses/>. */
++
++/* Written by Mitsuru Chinen <mchinen@yamato.ibm.com>
++ and Bruno Haible <bruno@clisp.org>. */
++
++/* The macros in this file implement multi-byte character input from a
++ stream.
++
++ mb_file_t
++ is the type for multibyte character input stream, usable for variable
++ declarations.
++
++ mbf_char_t
++ is the type for multibyte character or EOF, usable for variable
++ declarations.
++
++ mbf_init (mbf, stream)
++ initializes the MB_FILE for reading from stream.
++
++ mbf_getc (mbc, mbf)
++ reads the next multibyte character from mbf and stores it in mbc.
++
++ mb_iseof (mbc)
++ returns true if mbc represents the EOF value.
++
++ Here are the function prototypes of the macros.
++
++ extern void mbf_init (mb_file_t mbf, FILE *stream);
++ extern void mbf_getc (mbf_char_t mbc, mb_file_t mbf);
++ extern bool mb_iseof (const mbf_char_t mbc);
++ */
++
++#ifndef _MBFILE_H
++#define _MBFILE_H 1
++
++#include <assert.h>
++#include <stdbool.h>
++#include <stdio.h>
++#include <string.h>
++
++/* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before
++ <wchar.h>.
++ BSD/OS 4.1 has a bug: <stdio.h> and <time.h> must be included before
++ <wchar.h>. */
++#include <stdio.h>
++#include <time.h>
++#include <wchar.h>
++
++#include "mbchar.h"
++
++#ifndef _GL_INLINE_HEADER_BEGIN
++ #error "Please include config.h first."
++#endif
++_GL_INLINE_HEADER_BEGIN
++#ifndef MBFILE_INLINE
++# define MBFILE_INLINE _GL_INLINE
++#endif
++
++struct mbfile_multi {
++ FILE *fp;
++ bool eof_seen;
++ bool have_pushback;
++ mbstate_t state;
++ unsigned int bufcount;
++ char buf[MBCHAR_BUF_SIZE];
++ struct mbchar pushback;
++};
++
++MBFILE_INLINE void
++mbfile_multi_getc (struct mbchar *mbc, struct mbfile_multi *mbf)
++{
++ size_t bytes;
++
++ /* If EOF has already been seen, don't use getc. This matters if
++ mbf->fp is connected to an interactive tty. */
++ if (mbf->eof_seen)
++ goto eof;
++
++ /* Return character pushed back, if there is one. */
++ if (mbf->have_pushback)
++ {
++ mb_copy (mbc, &mbf->pushback);
++ mbf->have_pushback = false;
++ return;
++ }
++
++ /* Before using mbrtowc, we need at least one byte. */
++ if (mbf->bufcount == 0)
++ {
++ int c = getc (mbf->fp);
++ if (c == EOF)
++ {
++ mbf->eof_seen = true;
++ goto eof;
++ }
++ mbf->buf[0] = (unsigned char) c;
++ mbf->bufcount++;
++ }
++
++ /* Handle most ASCII characters quickly, without calling mbrtowc(). */
++ if (mbf->bufcount == 1 && mbsinit (&mbf->state) && is_basic (mbf->buf[0]))
++ {
++ /* These characters are part of the basic character set. ISO C 99
++ guarantees that their wide character code is identical to their
++ char code. */
++ mbc->wc = mbc->buf[0] = mbf->buf[0];
++ mbc->wc_valid = true;
++ mbc->ptr = &mbc->buf[0];
++ mbc->bytes = 1;
++ mbf->bufcount = 0;
++ return;
++ }
++
++ /* Use mbrtowc on an increasing number of bytes. Read only as many bytes
++ from mbf->fp as needed. This is needed to give reasonable interactive
++ behaviour when mbf->fp is connected to an interactive tty. */
++ for (;;)
++ {
++ /* We don't know whether the 'mbrtowc' function updates the state when
++ it returns -2, - this is the ISO C 99 and glibc-2.2 behaviour - or
++ not - amended ANSI C, glibc-2.1 and Solaris 2.7 behaviour. We
++ don't have an autoconf test for this, yet.
++ The new behaviour would allow us to feed the bytes one by one into
++ mbrtowc. But the old behaviour forces us to feed all bytes since
++ the end of the last character into mbrtowc. Since we want to retry
++ with more bytes when mbrtowc returns -2, we must backup the state
++ before calling mbrtowc, because implementations with the new
++ behaviour will clobber it. */
++ mbstate_t backup_state = mbf->state;
++
++ bytes = mbrtowc (&mbc->wc, &mbf->buf[0], mbf->bufcount, &mbf->state);
++
++ if (bytes == (size_t) -1)
++ {
++ /* An invalid multibyte sequence was encountered. */
++ /* Return a single byte. */
++ bytes = 1;
++ mbc->wc_valid = false;
++ break;
++ }
++ else if (bytes == (size_t) -2)
++ {
++ /* An incomplete multibyte character. */
++ mbf->state = backup_state;
++ if (mbf->bufcount == MBCHAR_BUF_SIZE)
++ {
++ /* An overlong incomplete multibyte sequence was encountered. */
++ /* Return a single byte. */
++ bytes = 1;
++ mbc->wc_valid = false;
++ break;
++ }
++ else
++ {
++ /* Read one more byte and retry mbrtowc. */
++ int c = getc (mbf->fp);
++ if (c == EOF)
++ {
++ /* An incomplete multibyte character at the end. */
++ mbf->eof_seen = true;
++ bytes = mbf->bufcount;
++ mbc->wc_valid = false;
++ break;
++ }
++ mbf->buf[mbf->bufcount] = (unsigned char) c;
++ mbf->bufcount++;
++ }
++ }
++ else
++ {
++ if (bytes == 0)
++ {
++ /* A null wide character was encountered. */
++ bytes = 1;
++ assert (mbf->buf[0] == '\0');
++ assert (mbc->wc == 0);
++ }
++ mbc->wc_valid = true;
++ break;
++ }
++ }
++
++ /* Return the multibyte sequence mbf->buf[0..bytes-1]. */
++ mbc->ptr = &mbc->buf[0];
++ memcpy (&mbc->buf[0], &mbf->buf[0], bytes);
++ mbc->bytes = bytes;
++
++ mbf->bufcount -= bytes;
++ if (mbf->bufcount > 0)
++ {
++ /* It's not worth calling memmove() for so few bytes. */
++ unsigned int count = mbf->bufcount;
++ char *p = &mbf->buf[0];
++
++ do
++ {
++ *p = *(p + bytes);
++ p++;
++ }
++ while (--count > 0);
++ }
++ return;
++
++eof:
++ /* An mbchar_t with bytes == 0 is used to indicate EOF. */
++ mbc->ptr = NULL;
++ mbc->bytes = 0;
++ mbc->wc_valid = false;
++ return;
++}
++
++MBFILE_INLINE void
++mbfile_multi_ungetc (const struct mbchar *mbc, struct mbfile_multi *mbf)
++{
++ mb_copy (&mbf->pushback, mbc);
++ mbf->have_pushback = true;
++}
++
++typedef struct mbfile_multi mb_file_t;
++
++typedef mbchar_t mbf_char_t;
++
++#define mbf_init(mbf, stream) \
++ ((mbf).fp = (stream), \
++ (mbf).eof_seen = false, \
++ (mbf).have_pushback = false, \
++ memset (&(mbf).state, '\0', sizeof (mbstate_t)), \
++ (mbf).bufcount = 0)
++
++#define mbf_getc(mbc, mbf) mbfile_multi_getc (&(mbc), &(mbf))
++
++#define mbf_ungetc(mbc, mbf) mbfile_multi_ungetc (&(mbc), &(mbf))
++
++#define mb_iseof(mbc) ((mbc).bytes == 0)
++
++#ifndef _GL_INLINE_HEADER_BEGIN
++ #error "Please include config.h first."
++#endif
++_GL_INLINE_HEADER_BEGIN
++
++#endif /* _MBFILE_H */
+diff --git a/m4/mbfile.m4 b/m4/mbfile.m4
+new file mode 100644
+index 0000000..8589902
+--- /dev/null
++++ b/m4/mbfile.m4
+@@ -0,0 +1,14 @@
++# mbfile.m4 serial 7
++dnl Copyright (C) 2005, 2008-2015 Free Software Foundation, Inc.
++dnl This file is free software; the Free Software Foundation
++dnl gives unlimited permission to copy and/or distribute it,
++dnl with or without modifications, as long as this notice is preserved.
++
++dnl autoconf tests required for use of mbfile.h
++dnl From Bruno Haible.
++
++AC_DEFUN([gl_MBFILE],
++[
++ AC_REQUIRE([AC_TYPE_MBSTATE_T])
++ :
++])
+diff --git a/src/expand.c b/src/expand.c
+index 9fa2e10..380e020 100644
+--- a/src/expand.c
++++ b/src/expand.c
+@@ -37,6 +37,9 @@
+ #include <stdio.h>
+ #include <getopt.h>
+ #include <sys/types.h>
++
++#include <mbfile.h>
++
+ #include "system.h"
+ #include "die.h"
+ #include "xstrndup.h"
+@@ -100,19 +103,19 @@ expand (void)
+ {
+ /* Input stream. */
+ FILE *fp = next_file (NULL);
++ mb_file_t mbf;
++ mbf_char_t c;
+
+ if (!fp)
+ return;
+
++ mbf_init (mbf, fp);
++
+ while (true)
+ {
+- /* Input character, or EOF. */
+- int c;
+-
+ /* If true, perform translations. */
+ bool convert = true;
+
+-
+ /* The following variables have valid values only when CONVERT
+ is true: */
+
+@@ -122,17 +125,23 @@ expand (void)
+ /* Index in TAB_LIST of next tab stop to examine. */
+ size_t tab_index = 0;
+
+-
+ /* Convert a line of text. */
+
+ do
+ {
+- while ((c = getc (fp)) < 0 && (fp = next_file (fp)))
+- continue;
++ do {
++ mbf_getc (c, mbf);
++ if (mb_iseof (c))
++ {
++ mbf_init (mbf, fp = next_file (fp));
++ continue;
++ }
++ }
++ while (false);
+
+ if (convert)
+ {
+- if (c == '\t')
++ if (mb_iseq (c, '\t'))
+ {
+ /* Column the next input tab stop is on. */
+ uintmax_t next_tab_column;
+@@ -151,32 +160,34 @@ expand (void)
+ if (putchar (' ') < 0)
+ die (EXIT_FAILURE, errno, _("write error"));
+
+- c = ' ';
++ mb_setascii (&c, ' ');
+ }
+- else if (c == '\b')
++ else if (mb_iseq (c, '\b'))
+ {
+ /* Go back one column, and force recalculation of the
+ next tab stop. */
+ column -= !!column;
+ tab_index -= !!tab_index;
+ }
+- else
++ /* A leading control character could make us trip over. */
++ else if (!mb_iscntrl (c))
+ {
+- column++;
++ column += mb_width (c);
+ if (!column)
+ die (EXIT_FAILURE, 0, _("input line is too long"));
+ }
+
+- convert &= convert_entire_line || !! isblank (c);
++ convert &= convert_entire_line || mb_isblank (c);
+ }
+
+- if (c < 0)
++ if (mb_iseof (c))
+ return;
+
+- if (putchar (c) < 0)
++ mb_putc (c, stdout);
++ if (ferror (stdout))
+ die (EXIT_FAILURE, errno, _("write error"));
+ }
+- while (c != '\n');
++ while (!mb_iseq (c, '\n'));
+ }
+ }
+
+diff --git a/src/unexpand.c b/src/unexpand.c
+index 7801274..569a7ee 100644
+--- a/src/unexpand.c
++++ b/src/unexpand.c
+@@ -38,6 +38,9 @@
+ #include <stdio.h>
+ #include <getopt.h>
+ #include <sys/types.h>
++
++#include <mbfile.h>
++
+ #include "system.h"
+ #include "die.h"
+ #include "xstrndup.h"
+@@ -107,11 +110,12 @@ unexpand (void)
+ {
+ /* Input stream. */
+ FILE *fp = next_file (NULL);
++ mb_file_t mbf;
+
+ /* The array of pending blanks. In non-POSIX locales, blanks can
+ include characters other than spaces, so the blanks must be
+ stored, not merely counted. */
+- char *pending_blank;
++ mbf_char_t *pending_blank;
+
+ if (!fp)
+ return;
+@@ -119,12 +123,14 @@ unexpand (void)
+ /* The worst case is a non-blank character, then one blank, then a
+ tab stop, then MAX_COLUMN_WIDTH - 1 blanks, then a non-blank; so
+ allocate MAX_COLUMN_WIDTH bytes to store the blanks. */
+- pending_blank = xmalloc (max_column_width);
++ pending_blank = xmalloc (max_column_width * sizeof (mbf_char_t));
++
++ mbf_init (mbf, fp);
+
+ while (true)
+ {
+ /* Input character, or EOF. */
+- int c;
++ mbf_char_t c;
+
+ /* If true, perform translations. */
+ bool convert = true;
+@@ -158,12 +164,19 @@ unexpand (void)
+
+ do
+ {
+- while ((c = getc (fp)) < 0 && (fp = next_file (fp)))
+- continue;
++ do {
++ mbf_getc (c, mbf);
++ if (mb_iseof (c))
++ {
++ mbf_init (mbf, fp = next_file (fp));
++ continue;
++ }
++ }
++ while (false);
+
+ if (convert)
+ {
+- bool blank = !! isblank (c);
++ bool blank = mb_isblank (c);
+
+ if (blank)
+ {
+@@ -180,16 +193,16 @@ unexpand (void)
+ if (next_tab_column < column)
+ die (EXIT_FAILURE, 0, _("input line is too long"));
+
+- if (c == '\t')
++ if (mb_iseq (c, '\t'))
+ {
+ column = next_tab_column;
+
+ if (pending)
+- pending_blank[0] = '\t';
++ mb_setascii (&pending_blank[0], '\t');
+ }
+ else
+ {
+- column++;
++ column += mb_width (c);
+
+ if (! (prev_blank && column == next_tab_column))
+ {
+@@ -197,13 +210,14 @@ unexpand (void)
+ will be replaced by tabs. */
+ if (column == next_tab_column)
+ one_blank_before_tab_stop = true;
+- pending_blank[pending++] = c;
++ mb_copy (&pending_blank[pending++], &c);
+ prev_blank = true;
+ continue;
+ }
+
+ /* Replace the pending blanks by a tab or two. */
+- pending_blank[0] = c = '\t';
++ mb_setascii (&c, '\t');
++ mb_setascii (&pending_blank[0], '\t');
+ }
+
+ /* Discard pending blanks, unless it was a single
+@@ -211,7 +225,7 @@ unexpand (void)
+ pending = one_blank_before_tab_stop;
+ }
+ }
+- else if (c == '\b')
++ else if (mb_iseq (c, '\b'))
+ {
+ /* Go back one column, and force recalculation of the
+ next tab stop. */
+@@ -221,7 +235,7 @@ unexpand (void)
+ }
+ else
+ {
+- column++;
++ column += mb_width (c);
+ if (!column)
+ die (EXIT_FAILURE, 0, _("input line is too long"));
+ }
+@@ -229,8 +243,11 @@ unexpand (void)
+ if (pending)
+ {
+ if (pending > 1 && one_blank_before_tab_stop)
+- pending_blank[0] = '\t';
+- if (fwrite (pending_blank, 1, pending, stdout) != pending)
++ mb_setascii (&pending_blank[0], '\t');
++
++ for (int n = 0; n < pending; ++n)
++ mb_putc (pending_blank[n], stdout);
++ if (ferror (stdout))
+ die (EXIT_FAILURE, errno, _("write error"));
+ pending = 0;
+ one_blank_before_tab_stop = false;
+@@ -240,16 +257,17 @@ unexpand (void)
+ convert &= convert_entire_line || blank;
+ }
+
+- if (c < 0)
++ if (mb_iseof (c))
+ {
+ free (pending_blank);
+ return;
+ }
+
+- if (putchar (c) < 0)
++ mb_putc (c, stdout);
++ if (ferror (stdout))
+ die (EXIT_FAILURE, errno, _("write error"));
+ }
+- while (c != '\n');
++ while (!mb_iseq (c, '\n'));
+ }
+ }
+
+diff --git a/tests/expand/mb.sh b/tests/expand/mb.sh
+new file mode 100755
+index 0000000..7971e18
+--- /dev/null
++++ b/tests/expand/mb.sh
+@@ -0,0 +1,98 @@
++#!/bin/sh
++
++# Copyright (C) 2012-2015 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 3 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, see <http://www.gnu.org/licenses/>.
++
++. "${srcdir=.}/tests/init.sh"; path_prepend_ ./src
++print_ver_ expand
++
++export LC_ALL=en_US.UTF-8
++
++#input containing multibyte characters
++cat <<\EOF > in || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++EOF
++env printf ' äöü\t. öüä. \tä xx\n' >> in || framework_failure_
++
++cat <<\EOF > exp || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++expand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++#test characters with display widths != 1
++env printf '12345678
++e\t|ascii(1)
++\u00E9\t|composed(1)
++e\u0301\t|decomposed(1)
++\u3000\t|ideo-space(2)
++\uFF0D\t|full-hypen(2)
++' > in || framework_failure_
++
++env printf '12345678
++e |ascii(1)
++\u00E9 |composed(1)
++e\u0301 |decomposed(1)
++\u3000 |ideo-space(2)
++\uFF0D |full-hypen(2)
++' > exp || framework_failure_
++
++expand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++#shouldn't fail with "input line too long"
++#when a line starts with a control character
++env printf '\n' > in || framework_failure_
++
++expand < in > out || fail=1
++compare in out > /dev/null 2>&1 || fail=1
++
++#non-Unicode characters interspersed between Unicode ones
++env printf '12345678
++\t\xFF|
++\xFF\t|
++\t\xFFä|
++ä\xFF\t|
++\tä\xFF|
++\xFF\tä|
++äbcdef\xFF\t|
++' > in || framework_failure_
++
++env printf '12345678
++ \xFF|
++\xFF |
++ \xFFä|
++ä\xFF |
++ ä\xFF|
++\xFF ä|
++äbcdef\xFF |
++' > exp || framework_failure_
++
++expand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++exit $fail
+diff --git a/tests/local.mk b/tests/local.mk
+index 192f776..8053397 100644
+--- a/tests/local.mk
++++ b/tests/local.mk
+@@ -544,6 +544,7 @@ all_tests = \
+ tests/du/threshold.sh \
+ tests/du/trailing-slash.sh \
+ tests/du/two-args.sh \
++ tests/expand/mb.sh \
+ tests/id/gnu-zero-uids.sh \
+ tests/id/no-context.sh \
+ tests/id/context.sh \
+@@ -684,6 +685,7 @@ all_tests = \
+ tests/touch/read-only.sh \
+ tests/touch/relative.sh \
+ tests/touch/trailing-slash.sh \
++ tests/unexpand/mb.sh \
+ $(all_root_tests)
+
+ # See tests/factor/create-test.sh.
+diff --git a/tests/unexpand/mb.sh b/tests/unexpand/mb.sh
+new file mode 100755
+index 0000000..60d4c1a
+--- /dev/null
++++ b/tests/unexpand/mb.sh
+@@ -0,0 +1,97 @@
++#!/bin/sh
++
++# Copyright (C) 2012-2015 Free Software Foundation, Inc.
++
++# This program is free software: you can redistribute it and/or modify
++# it under the terms of the GNU General Public License as published by
++# the Free Software Foundation, either version 3 of the License, or
++# (at your option) any later version.
++
++# This program is distributed in the hope that it will be useful,
++# but WITHOUT ANY WARRANTY; without even the implied warranty of
++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
++# GNU General Public License for more details.
++
++# You should have received a copy of the GNU General Public License
++# along with this program. If not, see <http://www.gnu.org/licenses/>.
++
++. "${srcdir=.}/tests/init.sh"; path_prepend_ ./src
++print_ver_ unexpand
++
++export LC_ALL=en_US.UTF-8
++
++#input containing multibyte characters
++cat > in <<\EOF
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++cat > exp <<\EOF
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++unexpand -a < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++#test characters with a display width larger than 1
++
++env printf '12345678
++e |ascii(1)
++\u00E9 |composed(1)
++e\u0301 |decomposed(1)
++\u3000 |ideo-space(2)
++\uFF0D |full-hypen(2)
++' > in || framework_failure_
++
++env printf '12345678
++e\t|ascii(1)
++\u00E9\t|composed(1)
++e\u0301\t|decomposed(1)
++\u3000\t|ideo-space(2)
++\uFF0D\t|full-hypen(2)
++' > exp || framework_failure_
++
++unexpand -a < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++#test input where a blank of width > 1 is not being substituted
++in="$(LC_ALL=en_US.UTF-8 printf ' \u3000 ö ü ß')"
++exp=' ö ü ß'
++
++unexpand -a < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++#non-Unicode characters interspersed between Unicode ones
++env printf '12345678
++ \xFF|
++\xFF |
++ \xFFä|
++ä\xFF |
++ ä\xFF|
++\xFF ä|
++äbcdef\xFF |
++' > in || framework_failure_
++
++env printf '12345678
++\t\xFF|
++\xFF\t|
++\t\xFFä|
++ä\xFF\t|
++\tä\xFF|
++\xFF\tä|
++äbcdef\xFF\t|
++' > exp || framework_failure_
++
++unexpand -a < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
+--
+2.7.4
+
new file mode 100644
@@ -0,0 +1,28 @@
+From 02424bfcd719bbaa695f4e1c3ef17ad91b0d23c0 Mon Sep 17 00:00:00 2001
+From: Lubomir Rintel <lkundrak@v3.sk>
+Date: Thu, 28 Jan 2016 20:57:22 +0100
+Subject: [PATCH] unexpand: fix blank line handling
+
+ echo '' |./src/unexpand -a
+
+Really?
+---
+ src/unexpand.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/src/unexpand.c b/src/unexpand.c
+index 569a7ee..3bbbd66 100644
+--- a/src/unexpand.c
++++ b/src/unexpand.c
+@@ -233,7 +233,7 @@ unexpand (void)
+ next_tab_column = column;
+ tab_index -= !!tab_index;
+ }
+- else
++ else if (!mb_iseq (c, '\n'))
+ {
+ column += mb_width (c);
+ if (!column)
+--
+2.7.4
+
new file mode 100644
@@ -0,0 +1,108 @@
+diff --git a/src/expand.c b/src/expand.c
+index 380e020..310b349 100644
+--- a/src/expand.c
++++ b/src/expand.c
+@@ -129,15 +129,19 @@ expand (void)
+
+ do
+ {
+- do {
++ while (true) {
+ mbf_getc (c, mbf);
+- if (mb_iseof (c))
++ if ((mb_iseof (c)) && (fp = next_file (fp)))
+ {
+- mbf_init (mbf, fp = next_file (fp));
++ mbf_init (mbf, fp);
+ continue;
+ }
++ else
++ {
++ break;
++ }
+ }
+- while (false);
++
+
+ if (convert)
+ {
+diff --git a/src/unexpand.c b/src/unexpand.c
+index 3bbbd66..863a90a 100644
+--- a/src/unexpand.c
++++ b/src/unexpand.c
+@@ -164,15 +164,19 @@ unexpand (void)
+
+ do
+ {
+- do {
++ while (true) {
+ mbf_getc (c, mbf);
+- if (mb_iseof (c))
++ if ((mb_iseof (c)) && (fp = next_file (fp)))
+ {
+- mbf_init (mbf, fp = next_file (fp));
++ mbf_init (mbf, fp);
+ continue;
+ }
++ else
++ {
++ break;
++ }
+ }
+- while (false);
++
+
+ if (convert)
+ {
+diff --git a/tests/expand/mb.sh b/tests/expand/mb.sh
+index 7971e18..031be7a 100755
+--- a/tests/expand/mb.sh
++++ b/tests/expand/mb.sh
+@@ -44,6 +44,20 @@ EOF
+ expand < in > out || fail=1
+ compare exp out > /dev/null 2>&1 || fail=1
+
++#multiple files as an input
++cat <<\EOF >> exp || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++expand ./in ./in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
+ #test characters with display widths != 1
+ env printf '12345678
+ e\t|ascii(1)
+diff --git a/tests/unexpand/mb.sh b/tests/unexpand/mb.sh
+index 60d4c1a..8d75652 100755
+--- a/tests/unexpand/mb.sh
++++ b/tests/unexpand/mb.sh
+@@ -44,6 +44,22 @@ EOF
+ unexpand -a < in > out || fail=1
+ compare exp out > /dev/null 2>&1 || fail=1
+
++
++#multiple files as an input
++cat >> exp <<\EOF
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++
++unexpand -a ./in ./in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
+ #test characters with a display width larger than 1
+
+ env printf '12345678
new file mode 100644
@@ -0,0 +1,80 @@
+From ff424639fe863cbd6963add1a79b97290c1606c6 Mon Sep 17 00:00:00 2001
+From: rpm-build <rpm-build>
+Date: Fri, 3 Feb 2017 12:26:53 +0100
+Subject: [PATCH] fold.c: preserve new-lines in mutlibyte text
+
+---
+ src/fold.c | 49 ++++++++++++++++++++++++-------------------------
+ 1 file changed, 24 insertions(+), 25 deletions(-)
+
+diff --git a/src/fold.c b/src/fold.c
+index d23edd5..8c232a7 100644
+--- a/src/fold.c
++++ b/src/fold.c
+@@ -342,39 +342,38 @@ fold_multibyte_text (FILE *istream, size_t width, int *saved_errno)
+ }
+
+ rescan:
+- if (operating_mode == byte_mode) /* byte mode */
++ if (convfail)
++ increment = 1;
++ else if (wc == L'\n')
++ {
++ /* preserve newline */
++ fwrite (line_out, sizeof(char), offset_out, stdout);
++ START_NEW_LINE;
++ continue;
++ }
++ else if (operating_mode == byte_mode) /* byte mode */
+ increment = mblength;
+ else if (operating_mode == character_mode) /* character mode */
+ increment = 1;
+- else /* column mode */
++ else /* column mode */
+ {
+- if (convfail)
+- increment = 1;
+- else
++ switch (wc)
+ {
+- 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'\b':
+- increment = (column > 0) ? -1 : 0;
+- break;
++ case L'\r':
++ increment = -1 * column;
++ break;
+
+- case L'\r':
+- increment = -1 * column;
+- break;
++ case L'\t':
++ increment = 8 - column % 8;
++ break;
+
+- case L'\t':
+- increment = 8 - column % 8;
+- break;
+-
+- default:
+- increment = wcwidth (wc);
+- increment = (increment < 0) ? 0 : increment;
+- }
++ default:
++ increment = wcwidth (wc);
++ increment = (increment < 0) ? 0 : increment;
+ }
+ }
+
+--
+2.7.4
+
new file mode 100644
@@ -0,0 +1,35 @@
+From 3976ef5a20369d8b490907ab2cba2d617305a5e0 Mon Sep 17 00:00:00 2001
+From: Kamil Dudka <kdudka@redhat.com>
+Date: Mon, 30 May 2016 16:19:20 +0200
+Subject: [PATCH] sort: do not use static array 'blanks' in human_numcompare()
+
+... because the array is not initialized with MB locales. Note this is
+rather a conservative fix. I plan to do more cleanup of the i18n patch
+in Fedora to prevent mistakes like this in future updates of coreutils.
+---
+ src/sort.c | 8 +++-----
+ 1 file changed, 3 insertions(+), 5 deletions(-)
+
+diff --git a/src/sort.c b/src/sort.c
+index 9e07ad8..e47b039 100644
+--- a/src/sort.c
++++ b/src/sort.c
+@@ -2304,12 +2304,10 @@ find_unit_order (char const *number)
+ <none/unknown> < K/k < M < G < T < P < E < Z < Y */
+
+ static int
+-human_numcompare (char const *a, char const *b)
++human_numcompare (char *a, char *b)
+ {
+- while (blanks[to_uchar (*a)])
+- a++;
+- while (blanks[to_uchar (*b)])
+- b++;
++ skipblanks(&a, a + strlen(a));
++ skipblanks(&b, b + strlen(b));
+
+ int diff = find_unit_order (a) - find_unit_order (b);
+ return (diff ? diff : strnumcmp (a, b, decimal_point, thousands_sep));
+--
+2.5.5
+
new file mode 100644
@@ -0,0 +1,456 @@
+From 7a7c776a4e228d180e74614fd8c8afcad5d4bdf7 Mon Sep 17 00:00:00 2001
+From: Jakub Martisko <jamartis@redhat.com>
+Date: Thu, 7 Jul 2016 12:53:26 +0200
+Subject: [PATCH] coreutils-i18n-un-expand-BOM.patch
+
+---
+ src/expand-common.c | 114 +++++++++++++++++++++++++++++++++++++++++++++++++++
+ src/expand-common.h | 12 ++++++
+ src/expand.c | 45 +++++++++++++++++++-
+ src/unexpand.c | 43 ++++++++++++++++++-
+ tests/expand/mb.sh | 71 ++++++++++++++++++++++++++++++++
+ tests/unexpand/mb.sh | 59 ++++++++++++++++++++++++++
+ 6 files changed, 342 insertions(+), 2 deletions(-)
+
+diff --git a/src/expand-common.c b/src/expand-common.c
+index 4657e46..97cbb09 100644
+--- a/src/expand-common.c
++++ b/src/expand-common.c
+@@ -19,6 +19,7 @@
+ #include <assert.h>
+ #include <stdio.h>
+ #include <sys/types.h>
++#include <mbfile.h>
+ #include "system.h"
+ #include "die.h"
+ #include "error.h"
+@@ -126,6 +127,119 @@ set_increment_size (uintmax_t tabval)
+ return ok;
+ }
+
++extern int
++set_utf_locale (void)
++{
++ /*try using some predefined locale */
++ const char* predef_locales[] = {"C.UTF8","en_US.UTF8","en_GB.UTF8"};
++
++ const int predef_locales_count=3;
++ for (int i=0;i<predef_locales_count;i++)
++ {
++ if (setlocale(LC_ALL,predef_locales[i])!=NULL)
++ {
++ break;
++ }
++ else if (i==predef_locales_count-1)
++ {
++ return 1;
++ error (EXIT_FAILURE, errno, _("cannot set UTF-8 locale"));
++ }
++ }
++ return 0;
++}
++
++extern bool
++check_utf_locale(void)
++{
++ char* locale = setlocale (LC_CTYPE , NULL);
++ if (locale == NULL)
++ {
++ return false;
++ }
++ else if (strcasestr(locale, "utf8") == NULL && strcasestr(locale, "utf-8") == NULL)
++ {
++ return false;
++ }
++ return true;
++}
++
++extern bool
++check_bom(FILE* fp, mb_file_t *mbf)
++{
++ int c;
++
++
++ c=fgetc(fp);
++
++ /*test BOM header of the first file */
++ mbf->bufcount=0;
++ if (c == 0xEF)
++ {
++ c=fgetc(fp);
++ }
++ else
++ {
++ if (c != EOF)
++ {
++ ungetc(c,fp);
++ }
++ return false;
++ }
++
++ if (c == 0xBB)
++ {
++ c=fgetc(fp);
++ }
++ else
++ {
++ if ( c!= EOF )
++ {
++ mbf->buf[0]=(unsigned char) 0xEF;
++ mbf->bufcount=1;
++ ungetc(c,fp);
++ return false;
++ }
++ else
++ {
++ ungetc(0xEF,fp);
++ return false;
++ }
++ }
++ if (c == 0xBF)
++ {
++ mbf->bufcount=0;
++ return true;
++ }
++ else
++ {
++ if (c != EOF)
++ {
++ mbf->buf[0]=(unsigned char) 0xEF;
++ mbf->buf[1]=(unsigned char) 0xBB;
++ mbf->bufcount=2;
++ ungetc(c,fp);
++ return false;
++ }
++ else
++ {
++ mbf->buf[0]=(unsigned char) 0xEF;
++ mbf->bufcount=1;
++ ungetc(0xBB,fp);
++ return false;
++ }
++ }
++ return false;
++}
++
++extern void
++print_bom(void)
++{
++ putc (0xEF, stdout);
++ putc (0xBB, stdout);
++ putc (0xBF, stdout);
++}
++
+ /* Add the comma or blank separated list of tab stops STOPS
+ to the list of tab stops. */
+ extern void
+diff --git a/src/expand-common.h b/src/expand-common.h
+index 8cb2079..763bfda 100644
+--- a/src/expand-common.h
++++ b/src/expand-common.h
+@@ -34,6 +34,18 @@ extern size_t max_column_width;
+ /* The desired exit status. */
+ extern int exit_status;
+
++extern int
++set_utf_locale (void);
++
++extern bool
++check_utf_locale(void);
++
++extern bool
++check_bom(FILE* fp, mb_file_t *mbf);
++
++extern void
++print_bom(void);
++
+ /* Add tab stop TABVAL to the end of 'tab_list'. */
+ extern void
+ add_tab_stop (uintmax_t tabval);
+diff --git a/src/expand.c b/src/expand.c
+index 310b349..4136824 100644
+--- a/src/expand.c
++++ b/src/expand.c
+@@ -103,11 +103,33 @@ expand (void)
+ FILE *fp = next_file (NULL);
+ mb_file_t mbf;
+ mbf_char_t c;
++ /* True if the starting locale is utf8. */
++ bool using_utf_locale;
++
++ /* True if the first file contains BOM header. */
++ bool found_bom;
++ using_utf_locale=check_utf_locale();
+
+ if (!fp)
+ return;
+-
+ mbf_init (mbf, fp);
++ found_bom=check_bom(fp,&mbf);
++
++ if (using_utf_locale == false && found_bom == true)
++ {
++ /*try using some predefined locale */
++
++ if (set_utf_locale () != 0)
++ {
++ error (EXIT_FAILURE, errno, _("cannot set UTF-8 locale"));
++ }
++ }
++
++
++ if (found_bom == true)
++ {
++ print_bom();
++ }
+
+ while (true)
+ {
+@@ -132,6 +154,27 @@ expand (void)
+ if ((mb_iseof (c)) && (fp = next_file (fp)))
+ {
+ mbf_init (mbf, fp);
++ if (fp!=NULL)
++ {
++ if (check_bom(fp,&mbf)==true)
++ {
++ /*Not the first file - check BOM header*/
++ if (using_utf_locale==false && found_bom==false)
++ {
++ /*BOM header in subsequent file but not in the first one. */
++ error (EXIT_FAILURE, errno, _("combination of files with and without BOM header"));
++ }
++ }
++ else
++ {
++ if(using_utf_locale==false && found_bom==true)
++ {
++ /*First file conatined BOM header - locale was switched to UTF
++ *all subsequent files should contain BOM. */
++ error (EXIT_FAILURE, errno, _("combination of files with and without BOM header"));
++ }
++ }
++ }
+ continue;
+ }
+ else
+diff --git a/src/unexpand.c b/src/unexpand.c
+index 863a90a..5681b58 100644
+--- a/src/unexpand.c
++++ b/src/unexpand.c
+@@ -116,16 +116,36 @@ unexpand (void)
+ include characters other than spaces, so the blanks must be
+ stored, not merely counted. */
+ mbf_char_t *pending_blank;
++ /* True if the starting locale is utf8. */
++ bool using_utf_locale;
++
++ /* True if the first file contains BOM header. */
++ bool found_bom;
++ using_utf_locale=check_utf_locale();
+
+ if (!fp)
+ return;
++ mbf_init (mbf, fp);
++ found_bom=check_bom(fp,&mbf);
++
++ if (using_utf_locale == false && found_bom == true)
++ {
++ /*try using some predefined locale */
+
++ if (set_utf_locale () != 0)
++ {
++ error (EXIT_FAILURE, errno, _("cannot set UTF-8 locale"));
++ }
++ }
+ /* The worst case is a non-blank character, then one blank, then a
+ tab stop, then MAX_COLUMN_WIDTH - 1 blanks, then a non-blank; so
+ allocate MAX_COLUMN_WIDTH bytes to store the blanks. */
+ pending_blank = xmalloc (max_column_width * sizeof (mbf_char_t));
+
+- mbf_init (mbf, fp);
++ if (found_bom == true)
++ {
++ print_bom();
++ }
+
+ while (true)
+ {
+@@ -169,6 +189,27 @@ unexpand (void)
+ if ((mb_iseof (c)) && (fp = next_file (fp)))
+ {
+ mbf_init (mbf, fp);
++ if (fp!=NULL)
++ {
++ if (check_bom(fp,&mbf)==true)
++ {
++ /*Not the first file - check BOM header*/
++ if (using_utf_locale==false && found_bom==false)
++ {
++ /*BOM header in subsequent file but not in the first one. */
++ error (EXIT_FAILURE, errno, _("combination of files with and without BOM header"));
++ }
++ }
++ else
++ {
++ if(using_utf_locale==false && found_bom==true)
++ {
++ /*First file conatined BOM header - locale was switched to UTF
++ *all subsequent files should contain BOM. */
++ error (EXIT_FAILURE, errno, _("combination of files with and without BOM header"));
++ }
++ }
++ }
+ continue;
+ }
+ else
+diff --git a/tests/expand/mb.sh b/tests/expand/mb.sh
+index 031be7a..1621c84 100755
+--- a/tests/expand/mb.sh
++++ b/tests/expand/mb.sh
+@@ -109,4 +109,75 @@ env printf '12345678
+ expand < in > out || fail=1
+ compare exp out > /dev/null 2>&1 || fail=1
+
++
++
++#BOM header test 1
++printf "\xEF\xBB\xBF" > in; cat <<\EOF >> in || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++EOF
++env printf ' äöü\t. öüä. \tä xx\n' >> in || framework_failure_
++
++printf "\xEF\xBB\xBF" > exp; cat <<\EOF >> exp || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++
++expand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++LANG=C expand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++LC_ALL=C expand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++
++printf '\xEF\xBB\xBF' > in1; cat <<\EOF >> in1 || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++EOF
++env printf ' äöü\t. öüä. \tä xx\n' >> in1 || framework_failure_
++
++
++printf '\xEF\xBB\xBF' > exp; cat <<\EOF >> exp || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++expand in1 in1 > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++LANG=C expand in1 in1 > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++LC_ALL=C expand in1 in1 > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
+ exit $fail
+diff --git a/tests/unexpand/mb.sh b/tests/unexpand/mb.sh
+index 8d75652..9d4ee3e 100755
+--- a/tests/unexpand/mb.sh
++++ b/tests/unexpand/mb.sh
+@@ -111,3 +111,62 @@ env printf '12345678
+
+ unexpand -a < in > out || fail=1
+ compare exp out > /dev/null 2>&1 || fail=1
++
++#BOM header test 1
++printf "\xEF\xBB\xBF" > in; cat <<\EOF >> in || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++env printf ' äöü\t. öüä. \tä xx\n' >> in || framework_failure_
++
++printf "\xEF\xBB\xBF" > exp; cat <<\EOF >> exp || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++unexpand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++LANG=C unexpand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++LC_ALL=C unexpand < in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++
++printf "\xEF\xBB\xBF" > exp; cat <<\EOF >> exp || framework_failure_
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++1234567812345678123456781
++. . . .
++a b c d
++. . . .
++ä ö ü ß
++. . . .
++ äöü . öüä. ä xx
++EOF
++
++
++unexpand in in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++LANG=C unexpand in in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
++
++LC_ALL=C unexpand in in > out || fail=1
++compare exp out > /dev/null 2>&1 || fail=1
+--
+2.9.3
+
@@ -13,7 +13,6 @@ TODO: merge upstream
src/uniq.c | 265 ++++++++++++++-
tests/i18n/sort.sh | 29 ++
tests/local.mk | 2 +
- tests/misc/cut.pl | 7 +-
tests/misc/expand.pl | 42 +++
tests/misc/fold.pl | 50 ++-
tests/misc/join.pl | 50 +++
@@ -23,9 +22,9 @@ TODO: merge upstream
tests/misc/unexpand.pl | 39 +++
tests/misc/uniq.pl | 55 ++++
tests/pr/pr-tests.pl | 49 +++
- 18 files changed, 2435 insertions(+), 162 deletions(-)
- create mode 100644 tests/i18n/sort.sh
- create mode 100644 tests/misc/sort-mb-tests.sh
+ 17 files changed, 2430 insertions(+), 160 deletions(-)
+ create mode 100755 tests/i18n/sort.sh
+ create mode 100755 tests/misc/sort-mb-tests.sh
diff --git a/lib/linebuffer.h b/lib/linebuffer.h
index 64181af..9b8fe5a 100644
@@ -315,7 +314,7 @@ index 8cd0d6b..d23edd5 100644
+ fwrite (line_out, sizeof(char), offset_out, stdout);
+ START_NEW_LINE;
+ continue;
-+
++
+ case L'\b':
+ increment = (column > 0) ? -1 : 0;
+ break;
@@ -672,7 +671,7 @@ index 98b461c..9990f38 100644
+ extract_field (line, ptr, lim - ptr);
+}
+#endif
-+
++
static void
freeline (struct line *line)
{
@@ -845,7 +844,7 @@ index 98b461c..9990f38 100644
xfields (line);
if (prevline[which - 1])
-@@ -567,21 +807,28 @@ prfield (size_t n, struct line const *line)
+@@ -563,21 +803,28 @@ prfield (size_t n, struct line const *line)
/* Output all the fields in line, other than the join field. */
@@ -877,7 +876,7 @@ index 98b461c..9990f38 100644
prfield (i, line);
}
}
-@@ -592,7 +839,6 @@ static void
+@@ -588,7 +835,6 @@ static void
prjoin (struct line const *line1, struct line const *line2)
{
const struct outlist *outlist;
@@ -885,7 +884,7 @@ index 98b461c..9990f38 100644
size_t field;
struct line const *line;
-@@ -626,7 +872,7 @@ prjoin (struct line const *line1, struct line const *line2)
+@@ -622,7 +868,7 @@ prjoin (struct line const *line1, struct line const *line2)
o = o->next;
if (o == NULL)
break;
@@ -894,7 +893,7 @@ index 98b461c..9990f38 100644
}
putchar (eolchar);
}
-@@ -1104,20 +1350,43 @@ main (int argc, char **argv)
+@@ -1099,20 +1345,43 @@ main (int argc, char **argv)
case 't':
{
@@ -1171,7 +1170,7 @@ index 26f221f..633f50e 100644
use_col_separator = true;
if (optarg)
separator_string (optarg);
-@@ -1166,10 +1250,45 @@ getoptnum (const char *n_str, int min, int *num, const char *err)
+@@ -1165,10 +1249,45 @@ getoptnum (const char *n_str, int min, int *num, const char *err)
a number. */
static void
@@ -1219,7 +1218,7 @@ index 26f221f..633f50e 100644
if (*arg)
{
long int tmp_long;
-@@ -1191,6 +1310,11 @@ static void
+@@ -1190,6 +1309,11 @@ static void
init_parameters (int number_of_files)
{
int chars_used_by_number = 0;
@@ -1231,7 +1230,7 @@ index 26f221f..633f50e 100644
lines_per_body = lines_per_page - lines_per_header - lines_per_footer;
if (lines_per_body <= 0)
-@@ -1228,7 +1352,7 @@ init_parameters (int number_of_files)
+@@ -1227,7 +1351,7 @@ init_parameters (int number_of_files)
else
col_sep_string = column_separator;
@@ -1240,7 +1239,7 @@ index 26f221f..633f50e 100644
use_col_separator = true;
}
/* It's rather pointless to define a TAB separator with column
-@@ -1258,11 +1382,11 @@ init_parameters (int number_of_files)
+@@ -1257,11 +1381,11 @@ init_parameters (int number_of_files)
+ TAB_WIDTH (chars_per_input_tab, chars_per_number); */
/* Estimate chars_per_text without any margin and keep it constant. */
@@ -1254,7 +1253,7 @@ index 26f221f..633f50e 100644
/* The number is part of the column width unless we are
printing files in parallel. */
-@@ -1271,7 +1395,7 @@ init_parameters (int number_of_files)
+@@ -1270,7 +1394,7 @@ init_parameters (int number_of_files)
}
int sep_chars, useful_chars;
@@ -1263,7 +1262,7 @@ index 26f221f..633f50e 100644
sep_chars = INT_MAX;
if (INT_SUBTRACT_WRAPV (chars_per_line - chars_used_by_number, sep_chars,
&useful_chars))
-@@ -1294,7 +1418,7 @@ init_parameters (int number_of_files)
+@@ -1293,7 +1417,7 @@ init_parameters (int number_of_files)
We've to use 8 as the lower limit, if we use chars_per_default_tab = 8
to expand a tab which is not an input_tab-char. */
free (clump_buff);
@@ -1272,7 +1271,7 @@ index 26f221f..633f50e 100644
}
/* Open the necessary files,
-@@ -1402,7 +1526,7 @@ init_funcs (void)
+@@ -1399,7 +1523,7 @@ init_funcs (void)
/* Enlarge p->start_position of first column to use the same form of
padding_not_printed with all columns. */
@@ -1281,7 +1280,7 @@ index 26f221f..633f50e 100644
/* This loop takes care of all but the rightmost column. */
-@@ -1436,7 +1560,7 @@ init_funcs (void)
+@@ -1433,7 +1557,7 @@ init_funcs (void)
}
else
{
@@ -1290,7 +1289,7 @@ index 26f221f..633f50e 100644
h_next = h + chars_per_column;
}
}
-@@ -1727,9 +1851,9 @@ static void
+@@ -1724,9 +1848,9 @@ static void
align_column (COLUMN *p)
{
padding_not_printed = p->start_position;
@@ -1302,7 +1301,7 @@ index 26f221f..633f50e 100644
padding_not_printed = ANYWHERE;
}
-@@ -2004,13 +2128,13 @@ store_char (char c)
+@@ -2001,13 +2125,13 @@ store_char (char c)
/* May be too generous. */
buff = X2REALLOC (buff, &buff_allocated);
}
@@ -1318,7 +1317,7 @@ index 26f221f..633f50e 100644
char *s;
int num_width;
-@@ -2027,22 +2151,24 @@ add_line_number (COLUMN *p)
+@@ -2024,22 +2148,24 @@ add_line_number (COLUMN *p)
/* 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. */
@@ -1347,7 +1346,7 @@ index 26f221f..633f50e 100644
output_position = POS_AFTER_TAB (chars_per_output_tab,
output_position);
}
-@@ -2203,7 +2329,7 @@ print_white_space (void)
+@@ -2198,7 +2324,7 @@ print_white_space (void)
while (goal - h_old > 1
&& (h_new = POS_AFTER_TAB (chars_per_output_tab, h_old)) <= goal)
{
@@ -1356,7 +1355,7 @@ index 26f221f..633f50e 100644
h_old = h_new;
}
while (++h_old <= goal)
-@@ -2223,6 +2349,7 @@ print_sep_string (void)
+@@ -2218,6 +2344,7 @@ print_sep_string (void)
{
char const *s = col_sep_string;
int l = col_sep_length;
@@ -1364,7 +1363,7 @@ index 26f221f..633f50e 100644
if (separators_not_printed <= 0)
{
-@@ -2234,6 +2361,7 @@ print_sep_string (void)
+@@ -2229,6 +2356,7 @@ print_sep_string (void)
{
for (; separators_not_printed > 0; --separators_not_printed)
{
@@ -1372,7 +1371,7 @@ index 26f221f..633f50e 100644
while (l-- > 0)
{
/* 3 types of sep_strings: spaces only, spaces and chars,
-@@ -2247,12 +2375,15 @@ print_sep_string (void)
+@@ -2242,12 +2370,15 @@ print_sep_string (void)
}
else
{
@@ -1389,7 +1388,7 @@ index 26f221f..633f50e 100644
/* sep_string ends with some spaces */
if (spaces_not_printed > 0)
print_white_space ();
-@@ -2280,7 +2411,7 @@ print_clump (COLUMN *p, int n, char *clump)
+@@ -2275,7 +2406,7 @@ print_clump (COLUMN *p, int n, char *clump)
required number of tabs and spaces. */
static void
@@ -1398,7 +1397,7 @@ index 26f221f..633f50e 100644
{
if (tabify_output)
{
-@@ -2304,6 +2435,74 @@ print_char (char c)
+@@ -2299,6 +2430,74 @@ print_char (char c)
putchar (c);
}
@@ -1473,7 +1472,7 @@ index 26f221f..633f50e 100644
/* Skip to page PAGE before printing.
PAGE may be larger than total number of pages. */
-@@ -2483,9 +2682,9 @@ read_line (COLUMN *p)
+@@ -2476,9 +2675,9 @@ read_line (COLUMN *p)
align_empty_cols = false;
}
@@ -1485,8 +1484,8 @@ index 26f221f..633f50e 100644
padding_not_printed = ANYWHERE;
}
-@@ -2555,7 +2754,7 @@ print_stored (COLUMN *p)
- int i;
+@@ -2547,7 +2746,7 @@ print_stored (COLUMN *p)
+ COLUMN *q;
int line = p->current_line++;
- char *first = &buff[line_vector[line]];
@@ -1494,7 +1493,7 @@ index 26f221f..633f50e 100644
/* FIXME
UMR: Uninitialized memory read:
* This is occurring while in:
-@@ -2567,7 +2766,7 @@ print_stored (COLUMN *p)
+@@ -2559,7 +2758,7 @@ print_stored (COLUMN *p)
xmalloc [xmalloc.c:94]
init_store_cols [pr.c:1648]
*/
@@ -1503,7 +1502,7 @@ index 26f221f..633f50e 100644
pad_vertically = true;
-@@ -2586,9 +2785,9 @@ print_stored (COLUMN *p)
+@@ -2579,9 +2778,9 @@ print_stored (COLUMN *p)
}
}
@@ -1515,7 +1514,7 @@ index 26f221f..633f50e 100644
padding_not_printed = ANYWHERE;
}
-@@ -2601,8 +2800,8 @@ print_stored (COLUMN *p)
+@@ -2594,8 +2793,8 @@ print_stored (COLUMN *p)
if (spaces_not_printed == 0)
{
output_position = p->start_position + end_vector[line];
@@ -1526,7 +1525,7 @@ index 26f221f..633f50e 100644
}
return true;
-@@ -2621,7 +2820,7 @@ print_stored (COLUMN *p)
+@@ -2614,7 +2813,7 @@ print_stored (COLUMN *p)
number of characters is 1.) */
static int
@@ -1535,7 +1534,7 @@ index 26f221f..633f50e 100644
{
unsigned char uc = c;
char *s = clump_buff;
-@@ -2631,10 +2830,10 @@ char_to_clump (char c)
+@@ -2624,10 +2823,10 @@ char_to_clump (char c)
int chars;
int chars_per_c = 8;
@@ -1548,7 +1547,7 @@ index 26f221f..633f50e 100644
{
width = TAB_WIDTH (chars_per_c, input_position);
-@@ -2715,6 +2914,164 @@ char_to_clump (char c)
+@@ -2708,6 +2907,164 @@ char_to_clump (char c)
return chars;
}
@@ -1732,7 +1731,7 @@ index 6d2eec5..f189a0d 100644
#include "system.h"
#include "argmatch.h"
#include "die.h"
-@@ -165,14 +173,39 @@ static int decimal_point;
+@@ -161,14 +169,39 @@ static int decimal_point;
/* Thousands separator; if -1, then there isn't one. */
static int thousands_sep;
@@ -1773,7 +1772,7 @@ index 6d2eec5..f189a0d 100644
/* The kind of blanks for '-b' to skip in various options. */
enum blanktype { bl_start, bl_end, bl_both };
-@@ -346,13 +379,11 @@ static bool reverse;
+@@ -342,13 +375,11 @@ static bool reverse;
they were read if all keys compare equal. */
static bool stable;
@@ -1790,7 +1789,7 @@ index 6d2eec5..f189a0d 100644
/* Flag to remove consecutive duplicate lines from the output.
Only the last of a sequence of equal lines will be output. */
-@@ -811,6 +842,46 @@ reap_all (void)
+@@ -806,6 +837,46 @@ reap_all (void)
reap (-1);
}
@@ -1837,7 +1836,7 @@ index 6d2eec5..f189a0d 100644
/* Clean up any remaining temporary files. */
static void
-@@ -1255,7 +1326,7 @@ zaptemp (char const *name)
+@@ -1274,7 +1345,7 @@ zaptemp (char const *name)
free (node);
}
@@ -1846,7 +1845,7 @@ index 6d2eec5..f189a0d 100644
static int
struct_month_cmp (void const *m1, void const *m2)
-@@ -1270,7 +1341,7 @@ struct_month_cmp (void const *m1, void const *m2)
+@@ -1289,7 +1360,7 @@ struct_month_cmp (void const *m1, void const *m2)
/* Initialize the character class tables. */
static void
@@ -1855,7 +1854,7 @@ index 6d2eec5..f189a0d 100644
{
size_t i;
-@@ -1282,7 +1353,7 @@ inittables (void)
+@@ -1301,7 +1372,7 @@ inittables (void)
fold_toupper[i] = toupper (i);
}
@@ -1864,7 +1863,7 @@ index 6d2eec5..f189a0d 100644
/* If we're not in the "C" locale, read different names for months. */
if (hard_LC_TIME)
{
-@@ -1364,6 +1435,84 @@ specify_nmerge (int oi, char c, char const *s)
+@@ -1383,6 +1454,84 @@ specify_nmerge (int oi, char c, char const *s)
xstrtol_fatal (e, oi, c, long_options, s);
}
@@ -1949,7 +1948,7 @@ index 6d2eec5..f189a0d 100644
/* Specify the amount of main memory to use when sorting. */
static void
specify_sort_size (int oi, char c, char const *s)
-@@ -1597,7 +1746,7 @@ buffer_linelim (struct buffer const *buf)
+@@ -1614,7 +1763,7 @@ buffer_linelim (struct buffer const *buf)
by KEY in LINE. */
static char *
@@ -1958,7 +1957,7 @@ index 6d2eec5..f189a0d 100644
{
char *ptr = line->text, *lim = ptr + line->length - 1;
size_t sword = key->sword;
-@@ -1606,10 +1755,10 @@ begfield (struct line const *line, struct keyfield const *key)
+@@ -1623,10 +1772,10 @@ begfield (struct line const *line, struct keyfield const *key)
/* The leading field separator itself is included in a field when -t
is absent. */
@@ -1971,7 +1970,7 @@ index 6d2eec5..f189a0d 100644
++ptr;
if (ptr < lim)
++ptr;
-@@ -1635,11 +1784,70 @@ begfield (struct line const *line, struct keyfield const *key)
+@@ -1652,11 +1801,70 @@ begfield (struct line const *line, struct keyfield const *key)
return ptr;
}
@@ -2043,7 +2042,7 @@ index 6d2eec5..f189a0d 100644
{
char *ptr = line->text, *lim = ptr + line->length - 1;
size_t eword = key->eword, echar = key->echar;
-@@ -1654,10 +1862,10 @@ limfield (struct line const *line, struct keyfield const *key)
+@@ -1671,10 +1879,10 @@ limfield (struct line const *line, struct keyfield const *key)
'beginning' is the first character following the delimiting TAB.
Otherwise, leave PTR pointing at the first 'blank' character after
the preceding field. */
@@ -2056,7 +2055,7 @@ index 6d2eec5..f189a0d 100644
++ptr;
if (ptr < lim && (eword || echar))
++ptr;
-@@ -1703,10 +1911,10 @@ limfield (struct line const *line, struct keyfield const *key)
+@@ -1720,10 +1928,10 @@ limfield (struct line const *line, struct keyfield const *key)
*/
/* Make LIM point to the end of (one byte past) the current field. */
@@ -2069,7 +2068,7 @@ index 6d2eec5..f189a0d 100644
if (newlim)
lim = newlim;
}
-@@ -1737,6 +1945,130 @@ limfield (struct line const *line, struct keyfield const *key)
+@@ -1754,6 +1962,130 @@ limfield (struct line const *line, struct keyfield const *key)
return ptr;
}
@@ -2200,7 +2199,7 @@ index 6d2eec5..f189a0d 100644
/* 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
-@@ -1823,8 +2155,22 @@ fillbuf (struct buffer *buf, FILE *fp, char const *file)
+@@ -1840,8 +2172,22 @@ fillbuf (struct buffer *buf, FILE *fp, char const *file)
else
{
if (key->skipsblanks)
@@ -2225,7 +2224,7 @@ index 6d2eec5..f189a0d 100644
line->keybeg = line_start;
}
}
-@@ -1974,7 +2320,7 @@ human_numcompare (char const *a, char const *b)
+@@ -1991,7 +2337,7 @@ human_numcompare (char const *a, char const *b)
hideously fast. */
static int
@@ -2234,7 +2233,7 @@ index 6d2eec5..f189a0d 100644
{
while (blanks[to_uchar (*a)])
a++;
-@@ -1984,6 +2330,25 @@ numcompare (char const *a, char const *b)
+@@ -2001,6 +2347,25 @@ numcompare (char const *a, char const *b)
return strnumcmp (a, b, decimal_point, thousands_sep);
}
@@ -2259,8 +2258,8 @@ index 6d2eec5..f189a0d 100644
+
/* Work around a problem whereby the long double value returned by glibc's
strtold ("NaN", ...) contains uninitialized bits: clear all bytes of
- A and B before calling strtold. FIXME: remove this function once
-@@ -2034,7 +2399,7 @@ general_numcompare (char const *sa, char const *sb)
+ A and B before calling strtold. FIXME: remove this function if
+@@ -2051,7 +2416,7 @@ general_numcompare (char const *sa, char const *sb)
Return 0 if the name in S is not recognized. */
static int
@@ -2269,7 +2268,7 @@ index 6d2eec5..f189a0d 100644
{
size_t lo = 0;
size_t hi = MONTHS_PER_YEAR;
-@@ -2310,15 +2675,14 @@ debug_key (struct line const *line, struct keyfield const *key)
+@@ -2327,15 +2692,14 @@ debug_key (struct line const *line, struct keyfield const *key)
char saved = *lim;
*lim = '\0';
@@ -2287,7 +2286,7 @@ index 6d2eec5..f189a0d 100644
else if (key->general_numeric)
ignore_value (strtold (beg, &tighter_lim));
else if (key->numeric || key->human_numeric)
-@@ -2452,7 +2816,7 @@ key_warnings (struct keyfield const *gkey, bool gkey_only)
+@@ -2469,7 +2833,7 @@ key_warnings (struct keyfield const *gkey, bool gkey_only)
/* Warn about significant leading blanks. */
bool implicit_skip = key_numeric (key) || key->month;
bool line_offset = key->eword == 0 && key->echar != 0; /* -k1.x,1.y */
@@ -2296,7 +2295,7 @@ index 6d2eec5..f189a0d 100644
&& ((!key->skipsblanks && !implicit_skip)
|| (!key->skipsblanks && key->schar)
|| (!key->skipeblanks && key->echar)))
-@@ -2510,11 +2874,87 @@ key_warnings (struct keyfield const *gkey, bool gkey_only)
+@@ -2527,11 +2891,87 @@ key_warnings (struct keyfield const *gkey, bool gkey_only)
error (0, 0, _("option '-r' only applies to last-resort comparison"));
}
@@ -2385,7 +2384,7 @@ index 6d2eec5..f189a0d 100644
{
struct keyfield *key = keylist;
-@@ -2599,7 +3039,7 @@ keycompare (struct line const *a, struct line const *b)
+@@ -2616,7 +3056,7 @@ keycompare (struct line const *a, struct line const *b)
else if (key->human_numeric)
diff = human_numcompare (ta, tb);
else if (key->month)
@@ -2394,7 +2393,7 @@ index 6d2eec5..f189a0d 100644
else if (key->random)
diff = compare_random (ta, tlena, tb, tlenb);
else if (key->version)
-@@ -2715,6 +3155,211 @@ keycompare (struct line const *a, struct line const *b)
+@@ -2732,6 +3172,211 @@ keycompare (struct line const *a, struct line const *b)
return key->reverse ? -diff : diff;
}
@@ -2606,16 +2605,16 @@ index 6d2eec5..f189a0d 100644
/* Compare two lines A and B, returning negative, zero, or positive
depending on whether A compares less than, equal to, or greater than B. */
-@@ -2742,7 +3387,7 @@ compare (struct line const *a, struct line const *b)
+@@ -2759,7 +3404,7 @@ compare (struct line const *a, struct line const *b)
diff = - NONZERO (blen);
else if (blen == 0)
diff = 1;
- else if (hard_LC_COLLATE)
+ else if (hard_LC_COLLATE && !folding)
{
- /* Note xmemcoll0 is a performance enhancement as
+ /* xmemcoll0 is a performance enhancement as
it will not unconditionally write '\0' after the
-@@ -4139,6 +4784,7 @@ set_ordering (char const *s, struct keyfield *key, enum blanktype blanktype)
+@@ -4149,6 +4794,7 @@ set_ordering (char const *s, struct keyfield *key, enum blanktype blanktype)
break;
case 'f':
key->translate = fold_toupper;
@@ -2623,7 +2622,7 @@ index 6d2eec5..f189a0d 100644
break;
case 'g':
key->general_numeric = true;
-@@ -4218,7 +4864,7 @@ main (int argc, char **argv)
+@@ -4228,7 +4874,7 @@ main (int argc, char **argv)
initialize_exit_failure (SORT_FAILURE);
hard_LC_COLLATE = hard_locale (LC_COLLATE);
@@ -2632,7 +2631,7 @@ index 6d2eec5..f189a0d 100644
hard_LC_TIME = hard_locale (LC_TIME);
#endif
-@@ -4239,6 +4885,29 @@ main (int argc, char **argv)
+@@ -4249,6 +4895,29 @@ main (int argc, char **argv)
thousands_sep = -1;
}
@@ -2662,7 +2661,7 @@ index 6d2eec5..f189a0d 100644
have_read_stdin = false;
inittables ();
-@@ -4513,13 +5182,34 @@ main (int argc, char **argv)
+@@ -4523,13 +5192,34 @@ main (int argc, char **argv)
case 't':
{
@@ -2701,7 +2700,7 @@ index 6d2eec5..f189a0d 100644
else
{
/* Provoke with 'sort -txx'. Complain about
-@@ -4530,9 +5220,11 @@ main (int argc, char **argv)
+@@ -4540,9 +5230,11 @@ main (int argc, char **argv)
quote (optarg));
}
}
@@ -2715,7 +2714,7 @@ index 6d2eec5..f189a0d 100644
}
break;
-@@ -4770,12 +5462,10 @@ main (int argc, char **argv)
+@@ -4771,12 +5463,10 @@ main (int argc, char **argv)
sort (files, nfiles, outfile, nthreads);
}
@@ -2843,7 +2842,7 @@ index 87a0c93..9f755d9 100644
+ while (pos < size)
+ {
+ MBCHAR_TO_WCHAR (wc, mblength, lp, pos, size, statep, convfail);
-+
++
+ if (convfail || !(iswblank (wc) || wc == '\n'))
+ {
+ pos += mblength;
@@ -2886,7 +2885,7 @@ index 87a0c93..9f755d9 100644
if (check_chars < oldlen)
oldlen = check_chars;
if (check_chars < newlen)
-@@ -295,14 +401,103 @@ different (char *old, char *new, size_t oldlen, size_t newlen)
+@@ -295,15 +401,104 @@ different (char *old, char *new, size_t oldlen, size_t newlen)
if (ignore_case)
{
@@ -2918,8 +2917,8 @@ index 87a0c93..9f755d9 100644
+
+ 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)
@@ -2990,11 +2989,12 @@ index 87a0c93..9f755d9 100644
+ free (copy[1]);
+ return rc;
+
- }
++}
+#endif
-
++
/* Output the line in linebuffer LINE to standard output
provided that the switches say it should be output.
+ MATCH is true if the line matches the previous line.
@@ -367,19 +562,38 @@ check_file (const char *infile, const char *outfile, char delimiter)
char *prevfield IF_LINT ( = NULL);
size_t prevlen IF_LINT ( = 0);
@@ -3118,7 +3118,7 @@ index 87a0c93..9f755d9 100644
skip_fields = 0;
check_chars = SIZE_MAX;
diff --git a/tests/i18n/sort.sh b/tests/i18n/sort.sh
-new file mode 100644
+new file mode 100755
index 0000000..26c95de
--- /dev/null
+++ b/tests/i18n/sort.sh
@@ -3156,7 +3156,7 @@ diff --git a/tests/local.mk b/tests/local.mk
index 568944e..192f776 100644
--- a/tests/local.mk
+++ b/tests/local.mk
-@@ -350,6 +350,8 @@ all_tests = \
+@@ -368,6 +368,8 @@ all_tests = \
tests/misc/sort-discrim.sh \
tests/misc/sort-files0-from.pl \
tests/misc/sort-float.sh \
@@ -3165,32 +3165,6 @@ index 568944e..192f776 100644
tests/misc/sort-h-thousands-sep.sh \
tests/misc/sort-merge.pl \
tests/misc/sort-merge-fdlimit.sh \
-diff --git a/tests/misc/cut.pl b/tests/misc/cut.pl
-index f6f8a56..b426a80 100755
---- a/tests/misc/cut.pl
-+++ b/tests/misc/cut.pl
-@@ -23,9 +23,11 @@ use strict;
- # Turn off localization of executable's output.
- @ENV{qw(LANGUAGE LANG LC_ALL)} = ('C') x 3;
-
--my $mb_locale = $ENV{LOCALE_FR_UTF8};
-+my $mb_locale;
-+# uncommented enable multibyte paths
-+$mb_locale = $ENV{LOCALE_FR_UTF8};
- ! defined $mb_locale || $mb_locale eq 'none'
-- and $mb_locale = 'C';
-+ and $mb_locale = 'C';
-
- my $prog = 'cut';
- my $try = "Try '$prog --help' for more information.\n";
-@@ -240,6 +242,7 @@ if ($mb_locale ne 'C')
- my @new_t = @$t;
- my $test_name = shift @new_t;
-
-+ next if ($test_name =~ "newline-[12][0-9]");
- push @new, ["$test_name-mb", @new_t, {ENV => "LC_ALL=$mb_locale"}];
- }
- push @Tests, @new;
diff --git a/tests/misc/expand.pl b/tests/misc/expand.pl
index 8a9cad1..9293e39 100755
--- a/tests/misc/expand.pl
@@ -3211,7 +3185,7 @@ index 8a9cad1..9293e39 100755
my @Tests =
(
['t1', '--tabs=3', {IN=>"a\tb"}, {OUT=>"a b"}],
-@@ -140,6 +149,8 @@ my @Tests =
+@@ -168,6 +177,8 @@ my @Tests =
# Test errors
@@ -3220,8 +3194,8 @@ index 8a9cad1..9293e39 100755
['e1', '--tabs="a"', {IN=>''}, {OUT=>''}, {EXIT=>1},
{ERR => "$prog: tab size contains invalid character(s): 'a'\n"}],
['e2', "-t $UINTMAX_OFLOW", {IN=>''}, {OUT=>''}, {EXIT=>1},
-@@ -150,6 +161,37 @@ my @Tests =
- {ERR => "$prog: tab sizes must be ascending\n"}],
+@@ -184,6 +195,37 @@ my @Tests =
+ {ERR => "$prog: '/' specifier not at start of number: '/'\n"}],
);
+if ($mb_locale ne 'C')
@@ -3351,7 +3325,7 @@ index 4d399d8..07f2823 100755
my $delim = chr 0247;
sub t_subst ($)
{
-@@ -329,8 +338,49 @@ foreach my $t (@tv)
+@@ -333,8 +342,49 @@ foreach my $t (@tv)
push @Tests, $new_ent;
}
@@ -3402,7 +3376,7 @@ index 4d399d8..07f2823 100755
my $verbose = $ENV{VERBOSE};
diff --git a/tests/misc/sort-mb-tests.sh b/tests/misc/sort-mb-tests.sh
-new file mode 100644
+new file mode 100755
index 0000000..11836ba
--- /dev/null
+++ b/tests/misc/sort-mb-tests.sh
@@ -3533,7 +3507,7 @@ index c3e7f8e..6ecd3ff 100755
# Since each test is run with a file name and with redirected stdin,
# the name in the diagnostic is either the file name or "-".
# Normalize each diagnostic to use '-'.
-@@ -424,6 +429,38 @@ foreach my $t (@Tests)
+@@ -423,6 +428,38 @@ foreach my $t (@Tests)
}
}
@@ -3572,7 +3546,7 @@ index c3e7f8e..6ecd3ff 100755
@Tests = triple_test \@Tests;
# Remember that triple_test creates from each test with exactly one "IN"
-@@ -433,6 +470,7 @@ foreach my $t (@Tests)
+@@ -432,6 +469,7 @@ foreach my $t (@Tests)
# Remove the IN_PIPE version of the "output-is-input" test above.
# The others aren't susceptible because they have three inputs each.
@Tests = grep {$_->[0] ne 'output-is-input.p'} @Tests;