also update bootstrap from gnulib

* gnulib-tests/Makefile.am (AM_CFLAGS): Don't augment after including
gnulib.mk.

.gitignore

@@ -52,6 +52,7 @@
 /tests/cspatfile
 /tests/ere.script
 /tests/get-mb-cur-max
+/tests/init.sh
 /tests/khadafy.out
 /tests/patfile
 /tests/spencer1.script

.prev-version

@@ -1 +1 @@
-3.8
+3.12

AUTHORS

@@ -1,4 +1,4 @@
-  Copyright (C) 1992, 1997-2002, 2004-2023 Free Software Foundation, Inc.
+  Copyright (C) 1992, 1997-2002, 2004-2026 Free Software Foundation, Inc.
 
   Copying and distribution of this file, with or without modification,
   are permitted in any medium without royalty provided the copyright

ChangeLog-2009

@@ -1407,7 +1407,7 @@
         is put in different compiled structure patterns[]. The patterns
         are given to dfacomp() and kwsmusts() as is.
         (Ecompile): Likewised.
-        (Fcompile): Reverse to the old behaviour of compiling the enire
+        (Fcompile): Reverse to the old behaviour of compiling the entire
         patterns in one shot.
         (EGexecute): If falling to GNU regex for the matching, loop in the
         array of compile patterns[] to find a match.
@@ -1457,7 +1457,7 @@
         (xrealloc): Removed using lib/xmalloc.c.
         (xmalloc): Removed using lib/xmalloc.c
         (main): Register with atexit() to check for error on stdout.
-        * configure.in: Check for atexit(), call jm_MALLOC, jm_RELLOC and
+        * configure.in: Check for atexit(), call jm_MALLOC, jm_REALLOC and
         jm_PREREQ_ERROR.
         * tests/bre.awk: Removed the hack to drain the buffer since we
         always fclose(stdout) atexit.
@@ -1541,7 +1541,7 @@
         * src/exclude.h: New file.
         * src/grep.c (main): Took the GNU tar code to handle
         the option --include, --exclude, --exclude-from.
-        Files are check for a match, with exlude_filename ().
+        Files are check for a match, with exclude_filename ().
         New option --exclude-from.
         * src/savedir.c: Call exclude_filename() to check for
         file pattern exclusion or inclusion.
@@ -1592,7 +1592,7 @@
 
         * m4/dosfile.m4 (AC_DOSFILE): Move AC_DEFINEs out of AC_CACHE_CHECK.
 
-2001-02-17  Alain Malgoire
+2001-02-17  Alain Magloire
 
         * doc/grep.texi: Document the new options and the new behaviour
         back-references are local.  Use excerpt from Karl Berry regex
@@ -1699,8 +1699,8 @@
         (color): Rename color variable to color_option.
         Removed 'always|never|auto' arguments, not necessary for grep.
         (exclude_pattern): new variable, holder for the file pattern.
-        (include_pattern): new variable, hoder for the file pattern.
-        * src/savedir.c: Signature change, take two new argmuments.
+        (include_pattern): new variable, holder for the file pattern.
+        * src/savedir.c: Signature change, take two new arguments.
         * doc/grep.texi: Document, new options.
         * doc/grep.man: Document, new options.
 
@@ -1712,7 +1712,7 @@
 
 2001-02-09  Alain Magloire
 
-        Patch from Ulrich Drepper to provide hilighting.
+        Patch from Ulrich Drepper to provide highlighting.
 
         * src/grep.c: New option --color.
         (color): New static var.
@@ -1722,7 +1722,7 @@
         to find the offset of the matching string.
         * src/savedir.c: Take advantage of _DIRENT_HAVE_TYPE if supported.
         * src/search.c (EGexecute, Fexecute, Pexecute): Take a new argument
-        when doing exact match for the color hiligting.
+        when doing exact match for the color highlighting.
 
 2000-09-01  Brian Youmans
 
@@ -1792,7 +1792,7 @@
 
 2000-06-02  Paul Eggert
 
-        Problen noted by Gerald Stoller <gerald_stoller@hotmail.com>
+        Problem noted by Gerald Stoller <gerald_stoller@hotmail.com>
 
         * src/grep.c (main): POSIX says that -q overrides -l, which
         in turn overrides the other output options.  Fix grep to
@@ -2208,7 +2208,7 @@
         on pre-OpenVMS 7.x systems; general overhaul.
         * src/getpagesize.h: Reinstate support for different pagesizes on
         VAX and Alpha. Work around problem with DEC C compiler.
-        * src/vms_fab.c: Cast to some assigments; fixed typo argcp vs. argp.
+        * src/vms_fab.c: Cast to some assignments; fixed typo argcp vs. argp.
         * src/vms_fab.h: Added new include files to avoid warnings about
         undefined function prototypes.
         Those patches were provided by Martin P.J. Zinser (zinser@decus.de).
@@ -2670,7 +2670,7 @@
 
 1999-03-16 Volker Borchert
 
-        * configure.in: Use case case ... esac for  checking Visual C++.
+        * configure.in: Use case ... esac for  checking Visual C++.
         When ${CC} contains options it was not recognize.
 
 1999-03-07 Paul Eggert
@@ -2764,7 +2764,7 @@
 
 1999-02-10 Alain Magloire
 
-        * bootstrap/{Makefile{try,am},REAMDE} : skeleton
+        * bootstrap/{Makefile{try,am},README} : skeleton
         provided for system lacking the tools to autoconfigure.
 
         * src/{e,f,}grepmat.c: added guard [HAVE_CONFIG_H]
@@ -2858,7 +2858,7 @@
         * doc/Makefile.am djgpp/Makefile.am m4/Makefile.am vms/Makefile.am:
         New files.
 
-        * m4/progtest.m4: proctect '[]' from m4.
+        * m4/progtest.m4: protect '[]' from m4.
         Noted by Eli Z.
 
         * PATCHES-AC: New file, add the patch for autoconf in the dist.
@@ -3333,7 +3333,7 @@
         Suggested by Harald Hanche-Olsen.
 
         * src/grep.c (main): '-f /dev/null' now specifies no patterns
-        and therfore matches nothing.
+        and therefore matches nothing.
         Reported by Jorge Stolfi.
         Patched by Paul Eggert.
 
@@ -3368,7 +3368,7 @@
         * src/grep.c: reverse back to greping directories,
         One could skip the error message by defining
         SKIP_DIR_ERROR. There is no clear way of doing
-        things, I hope to setle this on the next majore release
+        things, I hope to settle this on the next major release
         Thanks Paul Eggert, Eli Zaretskii and gnits for the
         exchange.
 
@@ -3427,7 +3427,7 @@
         (setmatcher) [HAVE_SETRLIMIT]: Set re_max_failures so that the
         matcher won't ever overflow the stack.
         (main) [__MSDOS__, _WIN32]: Handle backslashes and drive letters
-        in argv[0], remove the .exe suffix, and downcase the prgram name.
+        in argv[0], remove the .exe suffix, and downcase the program name.
         [O_BINARY]: Pass additional DOS-specific options to getopt_long
         and handle them.  Call stat before attempting to open the file, in
         case it is a directory (DOS will fail the open call for
@@ -3497,7 +3497,7 @@
         regex package. Change the way the tests were done to be more
         conformant to automake.
 
-        * configure.in: added --disable-regex for folks with their own fuctions.
+        * configure.in: added --disable-regex for folks with their own functions.
 
         * grep-20d : available for testing
 
@@ -3551,7 +3551,7 @@
 
         * check.sh, scriptgen.awk: fix grep paths.
 
-        * change the directory strucure: grep is now in src to comply with
+        * change the directory structure: grep is now in src to comply with
         gettext.m4.
 
         * grep.c version.c [VERSION]: got rid of version.c,
@@ -3648,6 +3648,6 @@
 
         * Version 2.0 released.
 
-Copyright (C) 1998-2023 Free Software Foundation, Inc.
+Copyright (C) 1998-2026 Free Software Foundation, Inc.
 Copying and distribution of this file, with or without modification,
   are permitted provided the copyright notice and this notice are preserved.

HACKING

@@ -597,7 +597,7 @@ Then just open the index.html file (in the generated lcov-html directory)
 in your favorite web browser.
 
 ========================================================================
-Copyright (C) 2009-2023 Free Software Foundation, Inc.
+Copyright (C) 2009-2026 Free Software Foundation, Inc.
 
 Permission is granted to copy, distribute and/or modify this document
 under the terms of the GNU Free Documentation License, Version 1.3 or

Makefile.am

@@ -1,6 +1,6 @@
 # Process this file with automake to create Makefile.in
 #
-# Copyright 1997-1998, 2005-2023 Free Software Foundation, Inc.
+# Copyright 1997-1998, 2005-2026 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
@@ -66,13 +66,10 @@ gen-ChangeLog:
 # current locale considers to be equal.
 ASSORT = LC_ALL=C sort
 
-# Extract all lines up to the first one starting with "##".
-prologue = perl -ne '/^\#\#/ and exit; print' $(srcdir)/THANKS.in
-
 THANKS: THANKS.in Makefile.am .mailmap thanks-gen
 	$(AM_V_GEN)rm -f $@-t $@;					\
 	{								\
-	  $(prologue); echo;						\
+	  perl -ne '/^\#\#/ and exit; print' $(srcdir)/THANKS.in; echo;	\
 	  { perl -ne '/^$$/.../^$$/ and !/^$$/ and s/  +/\0/ and print'	\
 	      $(srcdir)/THANKS.in;					\
 	    git log --pretty=format:'%aN%x00%aE'			\

NEWS

@@ -1,5 +1,78 @@
 GNU grep NEWS                                    -*- outline -*-
 
+* Noteworthy changes in release ?.? (????-??-??) [?]
+
+
+* Noteworthy changes in release 3.12 (2025-04-10) [stable]
+
+** Bug fixes
+
+  Searching a directory with at least 100,000 entries no longer fails
+  with "Operation not supported" and exit status 2. Now, this prints 1
+  and no diagnostic, as expected:
+    $ mkdir t && cd t && seq 100000|xargs touch && grep -r x .; echo $?
+    1
+  [bug introduced in grep 3.11]
+
+  -mN where 1 < N no longer mistakenly lseeks to end of input merely
+  because standard output is /dev/null.
+
+** Changes in behavior
+
+  The --unix-byte-offsets (-u) option is gone. In grep-3.7 (2021-08-14)
+  it became a warning-only no-op. Before then, it was a Windows-only no-op.
+
+  On Windows platforms and on AIX in 32-bit mode, grep in some cases
+  now supports Unicode characters outside the Basic Multilingual Plane.
+
+
+* Noteworthy changes in release 3.11 (2023-05-13) [stable]
+
+** Bug fixes
+
+  With -P, patterns like [\d] now work again.  Fixing this has caused
+  grep to revert to the behavior of grep 3.8, in that patterns like \w
+  and \b go back to using ASCII rather than Unicode interpretations.
+  However, future versions of GNU grep and/or PCRE2 are likely to fix
+  this and change the behavior of \w and \b back to Unicode again,
+  without breaking [\d] as 3.10 did.
+  [bug introduced in grep 3.10]
+
+  grep no longer fails on files dated after the year 2038,
+  when running on 32-bit x86 and ARM hosts using glibc 2.34+.
+  [bug introduced in grep 3.9]
+
+  grep -P no longer fails to match patterns using negated classes
+  like \D or \W when linked with PCRE2 10.34 or newer.
+  [bug introduced in grep 3.8]
+
+
+** Changes in behavior
+
+  grep --version now prints a line describing the version of PCRE2 it uses.
+  For example, it prints this when built with the very latest from git:
+    grep -P uses PCRE2 10.43-DEV 2023-04-14
+  or this with what's currently available in Fedora 37:
+    grep -P uses PCRE2 10.40 2022-04-14
+
+  previous versions of grep wouldn't respect the user provided settings for
+  PCRE_CFLAGS and PCRE_LIBS when building if a libpcre2-8 pkg-config module
+  was found.
+
+
+* Noteworthy changes in release 3.10 (2023-03-22) [stable]
+
+** Bug fixes
+
+  With -P, \d now matches only ASCII digits, regardless of PCRE
+  options/modes. The changes in grep-3.9 to make \b and \w work
+  properly had the undesirable side effect of making \d also match
+  e.g., the Arabic digits: ٠١٢٣٤٥٦٧٨٩.  With grep-3.9, -P '\d+'
+  would match that ten-digit (20-byte) string. Now, to match such
+  a digit, you would use \p{Nd}. Similarly, \D is now mapped to [^0-9].
+  [bug introduced in grep 3.9]
+
+
 * Noteworthy changes in release 3.9 (2023-03-05) [stable]
 
 ** Bug fixes
@@ -1338,7 +1411,7 @@ necessary to track the evolution of the regex package, and since
 I was changing it anyway I decided to do a general cleanup.
 
 ========================================================================
-Copyright (C) 1992, 1997-2002, 2004-2023 Free Software Foundation, Inc.
+Copyright (C) 1992, 1997-2002, 2004-2026 Free Software Foundation, Inc.
 
   Copying and distribution of this file, with or without modification,
   are permitted in any medium without royalty provided the copyright

README

@@ -1,4 +1,4 @@
-  Copyright (C) 1992, 1997-2002, 2004-2023 Free Software Foundation, Inc.
+  Copyright (C) 1992, 1997-2002, 2004-2026 Free Software Foundation, Inc.
 
   Copying and distribution of this file, with or without modification,
   are permitted in any medium without royalty provided the copyright

README-alpha

@@ -1,4 +1,4 @@
-  Copyright (C) 1992, 1997-2002, 2004-2023 Free Software Foundation, Inc.
+  Copyright (C) 1992, 1997-2002, 2004-2026 Free Software Foundation, Inc.
 
   Copying and distribution of this file, with or without modification,
   are permitted in any medium without royalty provided the copyright

README-hacking

@@ -101,7 +101,7 @@ each program.  One way to do this is to use vc-dwim
 
 -----
 
-Copyright (C) 2002-2023 Free Software Foundation, Inc.
+Copyright (C) 2002-2026 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

THANKS.in

@@ -13,6 +13,7 @@ end of e.g., grep --help).
 Akim Demaille                       akim@epita.fr
 Andreas Schwab                      schwab@suse.de
 Andreas Ley                         andy@rz.uni-karlsruhe.de
+Anton Samokat                       samokat700@gmail.com
 Bastiaan "Darquan" Stougie          darquan@zonnet.nl
 Ben Elliston                        bje@cygnus.com
 Bernd Strieder                      strieder@student.uni-kl.de
@@ -28,6 +29,7 @@ David J MacKenzie                   djm@catapult.va.pubnix.com
 David O'Brien                       obrien@freebsd.org
 'Drake' Daham Wang                  drakewang@gmail.com
 Egmont Koblinger                    egmont@gmail.com
+Emanuele Torre                      torreemanuele6@gmail.com
 Fernando Basso                      fernandobasso.br@gmail.com
 Florian La Roche                    laroche@redhat.com
 François Pinard                     pinard@iro.umontreal.ca
@@ -55,6 +57,7 @@ Karl Pettersson                     karl.pettersson@klpn.se
 Kaveh R. Ghazi                      ghazi@caip.rutgers.edu
 Kazuro Furukawa                     furukawa@apricot.kek.jp
 Keith Bostic                        bostic@bsdi.com
+Koen Claessen                       koen@chalmers.se
 Krishna Sethuraman                  krishna@sgihub.corp.sgi.com
 Kurt D Schwehr                      kdschweh@insci14.ucsd.edu
 Ludovic Courtès                     ludo@gnu.org
@@ -79,6 +82,7 @@ Rainer Orth                         ro@cebitec.uni-bielefeld.de
 Roland Roberts                      rroberts@muller.com
 Ruslan Ermilov                      ru@freebsd.org
 Santiago Vila                       sanvila@unex.es
+Sebastian Carlos                    sebaaa1754@gmail.com
 Shannon Hill                        hill@synnet.com
 Sotiris Vassilopoulos               Sotiris.Vassilopoulos@betatech.gr
 Standish Parsley                    adsspamtrap01@yahoo.com

TODO

@@ -1,6 +1,6 @@
 Things to do for GNU grep
 
-  Copyright (C) 1992, 1997-2002, 2004-2023 Free Software Foundation, Inc.
+  Copyright (C) 1992, 1997-2002, 2004-2026 Free Software Foundation, Inc.
 
   Copying and distribution of this file, with or without modification,
   are permitted in any medium without royalty provided the copyright

bootstrap.conf

@@ -1,6 +1,6 @@
 # Bootstrap configuration.
 
-# Copyright (C) 2006-2023 Free Software Foundation, Inc.
+# Copyright (C) 2006-2026 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
@@ -17,6 +17,9 @@
 
 avoided_gnulib_modules='
   --avoid=lock-tests
+  --avoid=mbuiter
+  --avoid=mbuiterf
+  --avoid=mbrlen-tests
   --avoid=mbrtowc-tests
   --avoid=update-copyright-tests
 '
@@ -28,11 +31,13 @@ argmatch
 assert-h
 c-ctype
 c-stack
-c-strcase
+c-strcasecmp
+c32isalnum
+c32rtomb
 closeout
 configmake
-dirname-lgpl
 dfa
+dirname-lgpl
 do-release-commit-and-tag
 error
 exclude
@@ -52,56 +57,55 @@ hash
 idx
 ignore-value
 intprops
-inttypes
+inttypes-h
 isatty
 isblank
+kwset
 largefile
-locale
+locale-h
 lseek
 maintainer-makefile
 malloc-gnu
 manywarnings
 mbrlen
-mbrtowc
+mbrtoc32-regular
+mbszero
+mcel-prefer
 memchr
 memchr2
 mempcpy
 minmax
+nullptr
 obstack
 openat-safer
 perl
-propername
 rawmemchr
 readme-release
-realloc-gnu
+realloc-posix
 regex
 safe-read
 same-inode
 ssize_t
-stdckdint
-stddef
-stdlib
+stdckdint-h
+stddef-h
+stdlib-h
 stpcpy
 strerror
-string
+string-h
 strstr
-strtoull
-strtoumax
-sys_stat
-unistd
+sys_stat-h
+unistd-h
 unlocked-io
 update-copyright
 useless-if-before-free
 verify
 version-etc-fsf
-wchar
-wcrtomb
-wctob
-wctype-h
+wchar-single
 windows-stat-inodes
 xalloc
 xbinary-io
 xstrtoimax
+year2038
 '
 gnulib_name=libgreputils
 
@@ -129,12 +133,16 @@ gnulib_tool_option_extras="--tests-base=gnulib-tests --with-tests --symlink\
 buildreq="\
 autoconf   2.62
 automake   1.11.1
-autopoint  -
+autopoint  0.19.2
 gettext    -
 git        1.4.4
 gzip       -
+m4         -
 makeinfo   -
 tar        -
+texi2pdf   6.1
+wget       -
+xz         -
 "
 
 bootstrap_post_import_hook ()
@@ -142,22 +150,27 @@ bootstrap_post_import_hook ()
   # Automake requires that ChangeLog exist.
   touch ChangeLog || return 1
 
+  # Copy tests/init.sh from Gnulib.
+  $gnulib_tool --copy-file tests/init.sh
+
   # Copy pkg-config's pkg.m4 so that our downstream users don't need to.
   local ac_dir=`aclocal --print-ac-dir`
   test -s "$ac_dir/dirlist" && ac_dir=$ac_dir:`tr '\n' : < "$ac_dir/dirlist"`
   oIFS=$IFS
   IFS=:
+  local found=false
   for dir in \
     $ACLOCAL_PATH $ac_dir /usr/share/aclocal ''
   do
     IFS=$oIFS
     if test -n "$dir" && test -r "$dir/pkg.m4"; then
       cp "$dir/pkg.m4" m4/pkg.m4
-      return
+      found=:
+      break
     fi
   done
   IFS=$oIFS
-  die 'Cannot find pkg.m4; perhaps you need to install pkg-config'
+  $found || die 'Cannot find pkg.m4; perhaps you need to install pkg-config'
 }
 
 bootstrap_epilogue()

cfg.mk

@@ -1,5 +1,5 @@
 # Customize maint.mk                           -*- makefile -*-
-# Copyright (C) 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2009-2026 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
@@ -45,6 +45,12 @@ announcement_Cc_ = $(translation_project_), $(PACKAGE)-devel@gnu.org
 _gl_TS_unmarked_extern_functions = \
   main usage mb_clen to_uchar dfaerror dfawarn imbrlen
 
+# Write base64-encoded (not hex) checksums into the announcement.
+announce_gen_args = --cksum-checksums
+
+# Add an exemption for sc_makefile_at_at_check.
+_makefile_at_at_check_exceptions = ' && !/MAKEINFO/'
+
 # Now that we have better tests, make this the default.
 export VERBOSE = yes
 
@@ -68,7 +74,7 @@ export VERBOSE = yes
 # 1127556 9e
 export XZ_OPT = -6e
 
-old_NEWS_hash = 5c1796327f2c4c7f40e10e58fd3a8019
+old_NEWS_hash = 3713245f672c3a9d1b455d6cc410c9ec
 
 # We prefer to spell it back-reference, as POSIX does.
 sc_prohibit_backref:
@@ -182,3 +188,5 @@ sc_timeout_prereq:
 	  | $(GREP) .							\
 	  && { echo '$(ME): timeout without use of require_timeout_'	\
 	    1>&2; exit 1; } || :
+
+codespell_ignore_words_list = clen,allo,Nd,abd,alph,debbugs,wee,UE,ois,creche

configure.ac

@@ -1,7 +1,7 @@
 dnl
 dnl autoconf input file for GNU grep
 dnl
-dnl Copyright (C) 1997-2006, 2009-2023 Free Software Foundation, Inc.
+dnl Copyright (C) 1997-2006, 2009-2026 Free Software Foundation, Inc.
 dnl
 dnl This file is part of GNU grep.
 dnl
@@ -51,16 +51,15 @@ AC_PROG_INSTALL
 AC_PROG_CC
 gl_EARLY
 AC_PROG_RANLIB
-PKG_PROG_PKG_CONFIG([0.9.0])
+PKG_PROG_PKG_CONFIG([0.9.0], [PKG_CONFIG=false])
 
 # grep never invokes mbrtowc or mbrlen on empty input,
 # so don't worry about this common bug,
 # as working around it would merely slow grep down.
 gl_cv_func_mbrtowc_empty_input='assume yes'
+gl_cv_func_mbrlen_empty_input='assume yes'
 
 dnl Checks for typedefs, structures, and compiler characteristics.
-AC_TYPE_SIZE_T
-AC_C_CONST
 gl_INIT
 
 # Ensure VLAs are not used.
@@ -134,6 +133,7 @@ if test $gl_gcc_warnings != no; then
   for w in $ws; do
     gl_WARN_ADD([$w])
   done
+  gl_WARN_ADD([-Wtrailing-whitespace]) # This project's coding style
   gl_WARN_ADD([-Wno-missing-field-initializers]) # We need this one
   gl_WARN_ADD([-Wno-sign-compare])     # Too many warnings for now
   gl_WARN_ADD([-Wno-unused-parameter]) # Too many warnings for now
@@ -166,14 +166,19 @@ if test $gl_gcc_warnings != no; then
   AC_SUBST([GNULIB_WARN_CFLAGS])
 
   # For gnulib-tests, the set is slightly smaller still.
-  nw=
-  nw="$nw -Wstrict-prototypes"
   # It's not worth being this picky about test programs.
-  nw="$nw -Wsuggest-attribute=const"
-  nw="$nw -Wsuggest-attribute=pure"
-  nw="$nw -Wsuggest-attribute=format"
+  nw=
   nw="$nw -Wformat-truncation=2"    # False alarm in strerror_r.c
+  nw="$nw -Wmissing-declarations"
+  nw="$nw -Wmissing-prototypes"
+  nw="$nw -Wmissing-variable-declarations"
+  nw="$nw -Wnull-dereference"
   nw="$nw -Wold-style-definition"
+  nw="$nw -Wstrict-prototypes"
+  nw="$nw -Wsuggest-attribute=cold"
+  nw="$nw -Wsuggest-attribute=const"
+  nw="$nw -Wsuggest-attribute=format"
+  nw="$nw -Wsuggest-attribute=pure"
 
   # Disable to avoid warnings in e.g., test-intprops.c and test-limits-h.c
   # due to overlong expansions like this:
@@ -193,13 +198,7 @@ AC_DEFINE([ARGMATCH_DIE], [usage (EXIT_FAILURE)],
 AC_DEFINE([ARGMATCH_DIE_DECL], [void usage (int _e)],
           [Define to the declaration of the xargmatch failure function.])
 
-dnl Checks for header files.
-AC_HEADER_DIRENT
-
-dnl Checks for functions.
-AC_FUNC_CLOSEDIR_VOID
-
-AC_CHECK_FUNCS_ONCE([isascii setlocale])
+AC_CHECK_FUNCS_ONCE([setlocale])
 
 dnl I18N feature
 AM_GNU_GETTEXT_VERSION([0.18.2])
@@ -212,6 +211,8 @@ dnl then the installer should configure --with-included-regex.
 AM_CONDITIONAL([USE_INCLUDED_REGEX], [test "$ac_use_included_regex" = yes])
 if test "$ac_use_included_regex" = no; then
   AC_MSG_WARN([Included lib/regex.c not used])
+else
+  AC_DEFINE([USE_INCLUDED_REGEX], 1, [building with included regex code])
 fi
 
 gl_FUNC_PCRE

doc/Makefile.am

@@ -1,7 +1,7 @@
 # Process this file with automake to create Makefile.in
 # Makefile.am for grep/doc.
 #
-# Copyright 2008-2023 Free Software Foundation, Inc.
+# Copyright 2008-2026 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
@@ -16,6 +16,10 @@
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <https://www.gnu.org/licenses/>.
 
+# The customization variable CHECK_NORMAL_MENU_STRUCTURE is necessary with
+# makeinfo versions ≥ 6.8.
+MAKEINFO = @MAKEINFO@ -c CHECK_NORMAL_MENU_STRUCTURE=1
+
 info_TEXINFOS = grep.texi
 grep_TEXINFOS = fdl.texi
 

doc/fdl.texi

@@ -5,8 +5,8 @@
 @c hence no sectioning command or @node.
 
 @display
-Copyright @copyright{} 2000--2002, 2007--2008, 2023 Free Software Foundation,
-Inc.
+Copyright @copyright{} 2000--2002, 2007--2008, 2023--2026 Free Software
+Foundation, Inc.
 @uref{https://fsf.org/}
 
 Everyone is permitted to copy and distribute verbatim copies

doc/grep.in.1

@@ -2,7 +2,7 @@
 .de dT
 .ds Dt \\$2
 ..
-.dT Time-stamp: "2019-12-29"
+.dT Time-stamp: "2025-03-21"
 .\" Update the above date whenever a change to either this file or
 .\" grep.c's 'usage' function results in a nontrivial change to the man page.
 .\" In Emacs, you can update the date by running 'M-x time-stamp'
@@ -11,8 +11,10 @@
 .
 .TH GREP 1 \*(Dt "GNU grep @VERSION@" "User Commands"
 .
-.if !\w|\*(lq| \{\
-.\" groff an-old.tmac does not seem to be in use, so define lq and rq.
+.ie \n(.g .ds ' \(aq
+.el .ds ' '
+.if !\w@\*(lq@ \{\
+.\" The implementation lacks \*(lq and presumably \*(rq.
 .	ie \n(.g \{\
 .		ds lq \(lq\"
 .		ds rq \(rq\"
@@ -23,114 +25,110 @@
 .	\}
 .\}
 .
-.if !\w|\*(la| \{\
+.as mC
+.if !\w@\*(mC@ \{\
 .\" groff an-ext.tmac does not seem to be in use, so define the parts of
-.\" it that are used below.  For a copy of groff an-ext.tmac, please see:
-.\" https://git.savannah.gnu.org/cgit/groff.git/plain/tmac/an-ext.tmac
-.\" --- Start of lines taken from groff an-ext.tmac
+.\" it that are used below, taken from groff 1.23.0.  For a copy, please see:
+.\" https://git.savannah.gnu.org/cgit/groff.git/plain/tmac/an-ext.tmac?id=1.23.0
+.nr mG \n(.g-1
+.\" --- Start of lines taken from groff an-ext.tmac,
+.\" except with "nr mH 14" replaced by "nr mH 0"
+.\" and with mS, SY, YS definitions omitted.
 .
-.\" Check whether we are using grohtml.
-.nr mH 0
-.if \n(.g \
-.  if '\*(.T'html' \
-.    nr mH 1
+.\" Define this to your implementation's constant-width typeface.
+.ds mC CW
+.if n .ds mC R
+.
+.\" Save the automatic hyphenation mode.
+.\"
+.\" In AT&T troff, there was no register exposing the hyphenation mode,
+.\" and no way to save and restore it.  Set `mH` to a reasonable value
+.\" for your implementation and preference.
+.de mY
+.  ie !\\n(.g \
+.    nr mH 0
+.  el \
+.    do nr mH \\n[.hy] \" groff extension register
+..
+.
+.nr mE 0 \" in an example (EX/EE)?
+.
+.\" Prepare link text for mail/web hyperlinks.  `MT` and `UR` call this.
+.de mV
+.  ds m1 \\$1\"
+..
 .
 .
-.\" Map mono-width fonts to standard fonts for groff's TTY device.
-.if n \{\
-.  do ftr CR R
-.  do ftr CI I
-.  do ftr CB B
-.\}
+.\" Emit hyperlink.  The optional argument supplies trailing punctuation
+.\" after link text.  `ME` and `UE` call this.
+.de mQ
+.  mY
+.  nh
+<\\*(m1>\\$1
+.  hy \\n(mH
+..
 .
-.\" groff has glyph entities for angle brackets.
-.ie \n(.g \{\
-.  ds la \(la\"
-.  ds ra \(ra\"
-.\}
-.el \{\
-.  ds la <\"
-.  ds ra >\"
-.  \" groff's man macros control hyphenation with this register.
-.  nr HY 1
-.\}
 .
 .\" Start URL.
+.if \n(.g-\n(mG \{\
 .de UR
-.  ds m1 \\$1\"
-.  nh
-.  if \\n(mH \{\
-.    \" Start diversion in a new environment.
-.    do ev URL-div
-.    do di URL-div
-.  \}
+.  mV \\$1
 ..
+.\}
 .
 .
 .\" End URL.
+.if \n(.g-\n(mG \{\
 .de UE
-.  ie \\n(mH \{\
-.    br
-.    di
-.    ev
-.
-.    \" Has there been one or more input lines for the link text?
-.    ie \\n(dn \{\
-.      do HTML-NS "<a href=""\\*(m1"">"
-.      \" Yes, strip off final newline of diversion and emit it.
-.      do chop URL-div
-.      do URL-div
-\c
-.      do HTML-NS </a>
-.    \}
-.    el \
-.      do HTML-NS "<a href=""\\*(m1"">\\*(m1</a>"
-\&\\$*\"
-.  \}
-.  el \
-\\*(la\\*(m1\\*(ra\\$*\"
-.
-.  hy \\n(HY
+.  mQ \\$1
 ..
+.\}
 .
 .
 .\" Start email address.
+.if \n(.g-\n(mG \{\
 .de MT
-.  ds m1 \\$1\"
-.  nh
-.  if \\n(mH \{\
-.    \" Start diversion in a new environment.
-.    do ev URL-div
-.    do di URL-div
-.  \}
+.  mV \\$1
 ..
+.\}
 .
 .
 .\" End email address.
+.if \n(.g-\n(mG \{\
 .de ME
-.  ie \\n(mH \{\
-.    br
-.    di
-.    ev
-.
-.    \" Has there been one or more input lines for the link text?
-.    ie \\n(dn \{\
-.      do HTML-NS "<a href=""mailto:\\*(m1"">"
-.      \" Yes, strip off final newline of diversion and emit it.
-.      do chop URL-div
-.      do URL-div
-\c
-.      do HTML-NS </a>
-.    \}
-.    el \
-.      do HTML-NS "<a href=""mailto:\\*(m1"">\\*(m1</a>"
-\&\\$*\"
-.  \}
-.  el \
-\\*(la\\*(m1\\*(ra\\$*\"
-.
-.  hy \\n(HY
+.  mQ \\$1
 ..
+.\}
+.
+.
+.\" Start example.
+.if \n(.g-\n(mG \{\
+.de EX
+.  br
+.  if !\\n(mE \{\
+.    nr mF \\n(.f
+.    nr mP \\n(PD
+.    nr PD 1v
+.    nf
+.    ft \\*(mC
+.    nr mE 1
+.  \}
+..
+.\}
+.
+.
+.\" End example.
+.if \n(.g-\n(mG \{\
+.de EE
+.  br
+.  if \\n(mE \{\
+.    ft \\n(mF
+.    nr PD \\n(mP
+.    fi
+.    nr mE 0
+.  \}
+..
+.\}
 .\" --- End of lines taken from groff an-ext.tmac
 .\}
 .
@@ -141,30 +139,33 @@ grep \- print lines that match patterns
 .
 .SH SYNOPSIS
 .B grep
-.RI [ OPTION .\|.\|.]\&
+.RI [ OPTION ].\|.\|.\&
 .I PATTERNS
-.RI [ FILE .\|.\|.]
+.RI [ FILE ].\|.\|.
 .br
 .B grep
-.RI [ OPTION .\|.\|.]\&
+.RI [ OPTION ].\|.\|.\&
 .B \-e
 .I PATTERNS
 \&.\|.\|.\&
-.RI [ FILE .\|.\|.]
+.RI [ FILE ].\|.\|.
 .br
 .B grep
-.RI [ OPTION .\|.\|.]\&
+.RI [ OPTION ].\|.\|.\&
 .B \-f
 .I PATTERN_FILE
 \&.\|.\|.\&
-.RI [ FILE .\|.\|.]
+.RI [ FILE ].\|.\|.
 .
 .SH DESCRIPTION
 .B grep
-searches for
-.I PATTERNS
-in each
+searches for patterns in each
 .IR FILE .
+In the synopsis's first form, which is used if no
+.B \-e
+or
+.B \-f
+options are present, the first operand
 .I PATTERNS
 is one or more patterns separated by newline characters, and
 .B grep
@@ -244,6 +245,11 @@ If this option is used multiple times or is combined with the
 .RB ( \-\^\-regexp )
 option, search for all patterns given.
 The empty file contains zero patterns, and therefore matches nothing.
+If
+.I FILE
+is
+.B \-
+, read patterns from standard input.
 .TP
 .BR \-i ", " \-\^\-ignore\-case
 Ignore case distinctions in patterns and input data,
@@ -399,7 +405,7 @@ Display input actually coming from standard input as input coming from file
 This can be useful for commands that transform a file's contents
 before searching,
 e.g.,
-.BR "gzip \-cd foo.gz | grep \-\^\-label=foo \-H 'some pattern'" .
+.BR "gzip \-cd foo.gz | grep \-\^\-label=foo \-H \*'some pattern\*'" .
 See also the
 .B \-H
 option.
@@ -590,7 +596,7 @@ On the other hand, when reading files whose text encodings are
 unknown, it can be helpful to use
 .B \-a
 or to set
-.B LC_ALL='C'
+.B LC_ALL=\*'C\*'
 in the environment, in order to find more matches even if the matches
 are unsafe for direct display.
 .TP
@@ -669,7 +675,7 @@ whose base name matches
 Ignore any redundant trailing slashes in
 .IR GLOB .
 .TP
-.BR \-I
+.B \-I
 Process a binary file as if it did not contain matching data; this is
 equivalent to the
 .B \-\^\-binary\-files=without\-match
@@ -715,12 +721,12 @@ This can cause a performance penalty.
 .BR \-U ", " \-\^\-binary
 Treat the file(s) as binary.
 By default, under MS-DOS and MS-Windows,
-.BR grep
+.B grep
 guesses whether a file is text or binary as described for the
 .B \-\^\-binary\-files
 option.
 If
-.BR grep
+.B grep
 decides the file is a text file, it strips the CR characters from the
 original file contents (to make regular expressions with
 .B ^
@@ -744,7 +750,7 @@ Like the
 or
 .B \-\^\-null
 option, this option can be used with commands like
-.B sort -z
+.B "sort \-z"
 to process arbitrary file names.
 .
 .SH "REGULAR EXPRESSIONS"
@@ -756,13 +762,14 @@ expressions, by using various operators to combine smaller expressions.
 understands three different versions of regular expression syntax:
 \*(lqbasic\*(rq (BRE), \*(lqextended\*(rq (ERE) and \*(lqperl\*(rq (PCRE).
 In GNU
-.B grep
-there is no difference in available functionality between basic and
-extended syntax.
-In other implementations, basic regular expressions are less powerful.
+.BR grep ,
+basic and extended regular expressions are merely different notations
+for the same pattern-matching functionality.
+In other implementations, basic regular expressions are ordinarily
+less powerful than extended, though occasionally it is the other way around.
 The following description applies to extended regular expressions;
 differences for basic regular expressions are summarized afterwards.
-Perl-compatible regular expressions give additional functionality, and are
+Perl-compatible regular expressions have different functionality, and are
 documented in
 .BR pcre2syntax (3)
 and
@@ -802,19 +809,21 @@ matches any single digit.
 Within a bracket expression, a
 .I "range expression"
 consists of two characters separated by a hyphen.
-It matches any single character that sorts between the two characters,
-inclusive, using the locale's collating sequence and character set.
-For example, in the default C locale,
+In the default C locale, it matches any single character that appears
+between the two characters in ASCII order, inclusive.
+For example,
 .B [a\-d]
 is equivalent to
 .BR [abcd] .
-Many locales sort characters in dictionary order, and in these locales
+In other locales the behavior is unspecified:
 .B [a\-d]
-is typically not equivalent to
-.BR [abcd] ;
-it might be equivalent to
-.BR [aBbCcDd] ,
-for example.
+might be equivalent to
+.B [abcd]
+or
+.B [aBbCcDd]
+or some other bracket expression,
+or it might fail to match any character, or the set of
+characters that it matches might be erratic, or it might be invalid.
 To obtain the traditional interpretation of bracket expressions,
 you can use the C locale by setting the
 .B LC_ALL
@@ -954,7 +963,7 @@ In basic regular expressions the meta-characters
 .BR | ,
 .BR ( ,
 and
-.BR )
+.B )
 lose their special meaning; instead use the backslashed
 versions
 .BR \e? ,
@@ -1243,38 +1252,13 @@ front of the operand list and are treated as options.
 Also, POSIX requires that unrecognized options be diagnosed as
 \*(lqillegal\*(rq, but since they are not really against the law the default
 is to diagnose them as \*(lqinvalid\*(rq.
-.B POSIXLY_CORRECT
-also disables \fB_\fP\fIN\fP\fB_GNU_nonoption_argv_flags_\fP,
-described below.
-.TP
-\fB_\fP\fIN\fP\fB_GNU_nonoption_argv_flags_\fP
-(Here
-.I N
-is
-.BR grep 's
-numeric process ID.)  If the
-.IR i th
-character of this environment variable's value is
-.BR 1 ,
-do not consider the
-.IR i th
-operand of
-.B grep
-to be an option, even if it appears to be one.
-A shell can put this variable in the environment for each command it runs,
-specifying which operands are the results of file name wildcard
-expansion and therefore should not be treated as options.
-This behavior is available only with the GNU C library, and only
-when
-.B POSIXLY_CORRECT
-is not set.
 .
 .SH NOTES
 This man page is maintained only fitfully;
 the full documentation is often more up-to-date.
 .
 .SH COPYRIGHT
-Copyright 1998-2000, 2002, 2005-2023 Free Software Foundation, Inc.
+Copyright 1998\(en2000, 2002, 2005\(en2026 Free Software Foundation, Inc.
 .PP
 This is free software;
 see the source for copying conditions.
@@ -1318,7 +1302,7 @@ contain \*(lqg\*(rq and end in \*(lq.h\*(rq.
 The
 .B \-n
 option outputs line numbers, the
-.B \-\-
+.B \-\^\-
 argument treats expansions of \*(lq*g*.h\*(rq starting with \*(lq\-\*(rq
 as file names not options,
 and the empty file /dev/null causes file names to be output
@@ -1326,7 +1310,7 @@ even if only one file name happens to be of the form \*(lq*g*.h\*(rq.
 .PP
 .in +2n
 .EX
-$ \fBgrep\fP \-n \-\- 'f.*\e.c$' *g*.h /dev/null
+$ \fBgrep\fP \-n \-\^\- \*'f.*\e.c$\*' *g*.h /dev/null
 argmatch.h:1:/* definitions and prototypes for argmatch.c
 .EE
 .in

doc/grep.texi

@@ -30,7 +30,7 @@
 @copying
 This manual is for @command{grep}, a pattern matching engine.
 
-Copyright @copyright{} 1999--2002, 2005, 2008--2023 Free Software Foundation,
+Copyright @copyright{} 1999--2002, 2005, 2008--2026 Free Software Foundation,
 Inc.
 
 @quotation
@@ -109,7 +109,7 @@ there is no way to match newline characters in a text.
 The general synopsis of the @command{grep} command line is
 
 @example
-grep [@var{option}...] [@var{patterns}] [@var{file}...]
+grep [@var{option}]@dots{} [@var{patterns}] [@var{file}]@dots{}
 @end example
 
 @noindent
@@ -202,6 +202,7 @@ in a shell command.
 Obtain patterns from @var{file}, one per line.
 If this option is used multiple times or is combined with the
 @option{-e} (@option{--regexp}) option, search for all patterns given.
+When @var{file} is @samp{-}, read patterns from standard input.
 The empty file contains zero patterns, and therefore matches nothing.
 (@option{-f} is specified by POSIX.)
 
@@ -217,13 +218,25 @@ so that characters that differ only in case
 match each other.  Although this is straightforward when letters
 differ in case only via lowercase-uppercase pairs, the behavior is
 unspecified in other situations.  For example, uppercase ``S'' has an
-unusual lowercase counterpart ``ſ'' (Unicode character U+017F, LATIN
+unusual lowercase counterpart
+@c This does not display correctly in PDF with texinfo 7.1
+@c and pdfTeX 3.141592653-2.6-1.40.25 (TeX Live 2023/Fedora 40).
+@ifnottex
+``ſ''
+@end ifnottex
+(Unicode character U+017F LATIN
 SMALL LETTER LONG S) in many locales, and it is unspecified whether
 this unusual character matches ``S'' or ``s'' even though uppercasing
 it yields ``S''.  Another example: the lowercase German letter ``ß''
-(U+00DF, LATIN SMALL LETTER SHARP S) is normally capitalized as the
+(U+00DF LATIN SMALL LETTER SHARP S) is normally capitalized as the
 two-character string ``SS'' but it does not match ``SS'', and it might
-not match the uppercase letter ``ẞ'' (U+1E9E, LATIN CAPITAL LETTER
+not match the uppercase letter
+@c This does not display correctly in PDF with texinfo 7.1
+@c and pdfTeX 3.141592653-2.6-1.40.25 (TeX Live 2023/Fedora 40).
+@ifnottex
+``ẞ''
+@end ifnottex
+(U+1E9E LATIN CAPITAL LETTER
 SHARP S) even though lowercasing the latter yields the former.
 
 @option{-y} is an obsolete synonym that is provided for compatibility.
@@ -548,7 +561,7 @@ Print @var{num} lines of leading and trailing output context.
 @opindex --group-separator
 @cindex group separator
 When @option{-A}, @option{-B} or @option{-C} are in use,
-print @var{string} instead of @option{--} between groups of lines.
+print @var{string} instead of @samp{--} between groups of lines.
 
 @item --no-group-separator
 @opindex --group-separator
@@ -624,9 +637,11 @@ When some output is suppressed, @command{grep} follows any output
 with a message to standard error saying that a binary file matches.
 
 If @var{type} is @samp{without-match},
-when @command{grep} discovers null input binary data
-it assumes that the rest of the file does not match;
+when @command{grep} discovers null binary data in an input file
+it assumes that any unprocessed input does not match;
 this is equivalent to the @option{-I} option.
+In this case the region of unprocessed input starts no later than the
+null binary data, and continues to end of file.
 
 If @var{type} is @samp{text},
 @command{grep} processes binary data as if it were text;
@@ -642,6 +657,16 @@ is not matched when @var{type} is @samp{text}.  Conversely, when
 @var{type} is @samp{binary} the pattern @samp{.} (period) might not
 match a null byte.
 
+The heuristic that @command{grep} uses to intuit whether input is
+binary is specific to @command{grep} and may well be unsuitable for
+other applications, as it depends on command-line options, on locale,
+and on hardware and operating system characteristics such as system
+page size and input buffering.  For example, if the input consists of
+a matching text line followed by nonmatching data that contains a null
+byte, @command{grep} might either output the matching line or treat
+the file as binary, depending on whether the unprocessed input happens
+to include the matching text line.
+
 @emph{Warning:} The @option{-a} (@option{--binary-files=text}) option
 might output binary garbage, which can have nasty side effects if the
 output is a terminal and if the terminal driver interprets some of it
@@ -695,7 +720,7 @@ immediately after a slash (@samp{/}) in the name.
 When searching recursively, skip any subfile whose base
 name matches @var{glob}; the base name is the part after the last
 slash.  A pattern can use
-@samp{*}, @samp{?}, and @samp{[}...@samp{]} as wildcards,
+@samp{*}, @samp{?}, and @samp{[}@dots{}@samp{]} as wildcards,
 and @code{\} to quote a wildcard or backslash character literally.
 
 @item --exclude-from=@var{file}
@@ -763,10 +788,10 @@ directory, recursively, following all symbolic links.
 @item --
 @opindex --
 @cindex option delimiter
-Delimit the option list.  Later arguments, if any, are treated as
-operands even if they begin with @samp{-}.  For example, @samp{grep PAT --
--file1 file2} searches for the pattern PAT in the files named @file{-file1}
-and @file{file2}.
+Delimit the option list.  Any later argument is not treated as an
+option even if it begins with @samp{-}.  For example,
+@samp{grep -- -PAT -file1 file2} searches for the pattern @samp{-PAT}
+in the files named @file{-file1} and @file{file2}.
 
 @item --line-buffered
 @opindex --line-buffered
@@ -884,7 +909,7 @@ for permitted values and their meanings as character attributes.)
 These substring values are integers in decimal representation
 and can be concatenated with semicolons.
 @command{grep} takes care of assembling the result
-into a complete SGR sequence (@samp{\33[}...@samp{m}).
+into a complete SGR sequence (@samp{\33[}@dots{}@samp{m}).
 Common values to concatenate include
 @samp{1} for bold,
 @samp{4} for underline,
@@ -997,7 +1022,7 @@ or when EL is too slow or causes too much flicker.
 The default is false (i.e., the capability is omitted).
 @end table
 
-Note that boolean capabilities have no @samp{=}... part.
+Note that boolean capabilities have no @samp{=}@dots{} part.
 They are omitted (i.e., false) by default and become true when specified.
 
 
@@ -1060,18 +1085,6 @@ and are treated as options.
 This variable specifies the output terminal type, which can affect
 what the @option{--color} option does.  @xref{General Output Control}.
 
-@item _@var{N}_GNU_nonoption_argv_flags_
-@vindex _@var{N}_GNU_nonoption_argv_flags_ @r{environment variable}
-(Here @code{@var{N}} is @command{grep}'s numeric process ID.)
-If the @var{i}th character of this environment variable's value is @samp{1},
-do not consider the @var{i}th operand of @command{grep} to be an option,
-even if it appears to be one.
-A shell can put this variable in the environment for each command it runs,
-specifying which operands are the results of file name wildcard expansion
-and therefore should not be treated as options.
-This behavior is available only with the GNU C library,
-and only when @env{POSIXLY_CORRECT} is not set.
-
 @end table
 
 The @env{GREP_OPTIONS} environment variable of @command{grep} 2.20 and
@@ -1148,10 +1161,74 @@ Interpret patterns as fixed strings, not regular expressions.
 @opindex --perl-regexp
 @cindex matching Perl-compatible regular expressions
 Interpret patterns as Perl-compatible regular expressions (PCREs).
-PCRE support is here to stay, but consider this option experimental when
-combined with the @option{-z} (@option{--null-data}) option, and note that
-@samp{grep@ -P} may warn of unimplemented features.
-@xref{Other Options}.
+
+For documentation, refer to @url{https://www.pcre.org/}, with these caveats:
+@itemize
+@item
+In a UTF-8 locale, Perl treats data as UTF-8 only under certain
+conditions, e.g., if @command{perl} is invoked with the @option{-C}
+option or the @env{PERL_UNICODE} environment variable set appropriately.
+Similarly, @command{pcre2grep} treats data as UTF-8 only if
+invoked with @option{-u} or @option{-U}.
+In contrast, in a UTF-8 locale @command{grep} and @command{git grep}
+always treat data as UTF-8.
+
+@item
+In Perl and @command{git grep -P}, @samp{\d} matches all Unicode digits,
+even if they are not ASCII.
+For example, @samp{\d} matches
+@ifnottex
+``٣''
+@end ifnottex
+(U+0663 ARABIC-INDIC DIGIT THREE).
+In contrast, in @samp{grep -P}, @samp{\d} matches only
+the ten ASCII digits, regardless of locale.
+In @command{pcre2grep}, @samp{\d} ordinarily behaves like Perl and
+@command{git grep -P}, but when given the @option{--posix-digit} option
+it behaves like @samp{grep -P}.
+(On all platforms, @samp{\D} matches the complement of @samp{\d}.)
+
+@item
+The pattern @samp{[[:digit:]]} matches all Unicode digits
+in Perl, @samp{grep -P}, @command{git grep -P}, and @command{pcre2grep},
+so you can use it
+to get the effect of Perl's @samp{\d} on all these platforms.
+In other words, in Perl and @command{git grep -P},
+@samp{\d} is equivalent to @samp{[[:digit:]]},
+whereas in @samp{grep -P}, @samp{\d} is equivalent to @samp{[0-9]},
+and @command{pcre2grep} ordinarily follows Perl but
+when given @option{--posix-digit} it follows @samp{grep -P}.
+
+(On all these platforms, @samp{[[:digit:]]} is equivalent to @samp{\p@{Nd@}}
+and to @samp{\p@{General_Category: Decimal_Number@}}.)
+
+@item
+If @command{grep} is built with PCRE2 version 10.43 (2024) or later,
+@samp{(?aD)} causes @samp{\d} to behave like @samp{[0-9]} and
+@samp{(?-aD)} causes it to behave like @samp{[[:digit:]]}.
+
+@item
+Although PCRE tracks the syntax and semantics of Perl's regular
+expressions, the match is not always exact.  Perl
+evolves and a Perl installation may predate or postdate the PCRE2
+installation on the same host, or their Unicode versions may differ,
+or Perl and PCRE2 may disagree about an obscure construct.
+
+@item
+By default, @command{grep} applies each regexp to a line at a time,
+so the @samp{(?s)} directive (making @samp{.} match line breaks)
+is generally ineffective.
+However, with @option{-z} (@option{--null-data}) it can work:
+@example
+$ printf 'a\nb\n' |grep -zP '(?s)a.b'
+a
+b
+@end example
+But beware: with the @option{-z} (@option{--null-data}) and a file
+containing no NUL byte, grep must read the entire file into memory
+before processing any of it.
+Thus, it will exhaust memory and fail for some large files.
+@end itemize
 
 @end table
 
@@ -1167,12 +1244,13 @@ by using various operators to combine smaller expressions.
 three different versions of regular expression syntax:
 basic (BRE), extended (ERE), and Perl-compatible (PCRE).
 In GNU @command{grep},
-there is no difference in available functionality between basic and
-extended syntax.
-In other implementations, basic regular expressions are less powerful.
+basic and extended regular expressions are merely different notations
+for the same pattern-matching functionality.
+In other implementations, basic regular expressions are ordinarily
+less powerful than extended, though occasionally it is the other way around.
 The following description applies to extended regular expressions;
 differences for basic regular expressions are summarized afterwards.
-Perl-compatible regular expressions give additional functionality, and
+Perl-compatible regular expressions have different functionality, and
 are documented in the @i{pcre2syntax}(3) and @i{pcre2pattern}(3) manual
 pages, but work only if PCRE is available in the system.
 
@@ -1297,13 +1375,13 @@ encoding error.
 @cindex range expression
 Within a bracket expression, a @dfn{range expression} consists of two
 characters separated by a hyphen.
-It matches any single character that
-sorts between the two characters, inclusive.
-In the default C locale, the sorting sequence is the native character
-order; for example, @samp{[a-d]} is equivalent to @samp{[abcd]}.
-In other locales, the sorting sequence is not specified, and
-@samp{[a-d]} might be equivalent to @samp{[abcd]} or to
-@samp{[aBbCcDd]}, or it might fail to match any character, or the set of
+In the default C locale, it matches any single character that appears
+between the two characters in ASCII order, inclusive.
+For example, @samp{[a-d]} is equivalent to @samp{[abcd]}.
+In other locales the behavior is unspecified:
+@samp{[a-d]} might be equivalent to
+@samp{[abcd]} or @samp{[aBbCcDd]} or some other bracket expression,
+or it might fail to match any character, or the set of
 characters that it matches might be erratic, or it might be invalid.
 To obtain the traditional interpretation
 of bracket expressions, you can use the @samp{C} locale by setting the
@@ -1706,8 +1784,9 @@ A range expression outside the POSIX locale.  For example, in some
 locales @samp{[a-z]} might match some characters that are not
 lowercase letters, or might not match some lowercase letters, or might
 be invalid.  With GNU @command{grep} it is not documented whether
-these range expressions use native code points, or use the collating
-sequence specified by the @env{LC_COLLATE} category, or have some
+these range expressions use native code points, or use the collation
+sequence specified by the @env{LC_COLLATE} category, or use the
+collation ordering used by @command{sort} and @code{strcoll}, or have some
 other interpretation.  Outside the POSIX locale, it is portable to use
 @samp{[[:lower:]]} to match a lower-case letter, or
 @samp{[abcdefghijklmnopqrstuvwxyz]} to match an ASCII lower-case
@@ -1975,7 +2054,7 @@ Note that on some platforms,
 except the available memory.
 
 @item
-Why does @command{grep} report ``Binary file matches''?
+Why does @command{grep} report ``binary file matches''?
 
 If @command{grep} listed all matching ``lines'' from a binary file, it
 would probably generate output that is not useful, and it might even
@@ -2109,7 +2188,7 @@ removed from POSIX in 2001, deprecated by GNU Grep 2.5.3 in 2007,
 and changed to issue obsolescence warnings by GNU Grep 3.8 in 2022;
 eventually, they are planned to be removed entirely.
 
-If you prefer the old names, you can use use your own substitutes,
+If you prefer the old names, you can use your own substitutes,
 such as a shell script named @command{egrep} with the following
 contents:
 

gnulib

@@ -1 +1 @@
-Subproject commit 2ba7c75ed134b73c857ca5575d8803313a50db60
+Subproject commit 4f6ac2c3c689cd7312b5f9da97791b14bbc2ee53

gnulib-tests/Makefile.am

@@ -1,4 +1,3 @@
-AM_CFLAGS =
-include gnulib.mk
+AM_CFLAGS = $(GNULIB_TEST_WARN_CFLAGS) $(WERROR_CFLAGS)
 
-AM_CFLAGS += $(GNULIB_TEST_WARN_CFLAGS) $(WERROR_CFLAGS)
+include gnulib.mk

lib/Makefile.am

@@ -1,4 +1,4 @@
-# Copyright 1997-1998, 2005-2023 Free Software Foundation, Inc.
+# Copyright 1997-1998, 2005-2026 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

lib/colorize-posix.c

@@ -1,5 +1,5 @@
 /* Output colorization.
-   Copyright 2011-2023 Free Software Foundation, Inc.
+   Copyright 2011-2026 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
@@ -12,9 +12,7 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 /* Without this pragma, gcc 4.7.0 20120102 suggests that the
    init_colorize function might be candidate for attribute 'const'  */

lib/colorize-w32.c

@@ -1,5 +1,5 @@
 /* Output colorization on MS-Windows.
-   Copyright 2011-2023 Free Software Foundation, Inc.
+   Copyright 2011-2026 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
@@ -12,9 +12,7 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 /* Written by Eli Zaretskii.  */
 
@@ -96,7 +94,7 @@ w32_sgr2attr (const char *sgr_seq)
     {
       if (*p == ';' || *p == '\0')
         {
-          code = strtol (s, NULL, 10);
+          code = strtol (s, nullptr, 10);
           s = p + (*p != '\0');
 
           switch (code)

lib/colorize.h

@@ -1,6 +1,6 @@
 /* Output colorization.
 
-   Copyright 2011-2023 Free Software Foundation, Inc.
+   Copyright 2011-2026 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, or (at your option)
@@ -12,9 +12,7 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 extern int should_colorize (void);
 extern void init_colorize (void);

m4/pcre.m4

@@ -1,6 +1,6 @@
 # pcre.m4 - check for PCRE library support
 
-# Copyright (C) 2010-2023 Free Software Foundation, Inc.
+# Copyright (C) 2010-2026 Free Software Foundation, Inc.
 # This file is free software; the Free Software Foundation
 # gives unlimited permission to copy and/or distribute it,
 # with or without modifications, as long as this notice is preserved.
@@ -21,21 +21,38 @@ AC_DEFUN([gl_FUNC_PCRE],
   use_pcre=no
 
   if test $test_pcre != no; then
-    PKG_CHECK_MODULES([PCRE], [libpcre2-8], [], [: ${PCRE_LIBS=-lpcre2-8}])
+
+    AS_CASE([${PCRE_CFLAGS+set}@${PCRE_LIBS+set}@$PKG_CONFIG],
+      [@@false], [],
+      [@@*], [PKG_CHECK_MODULES([PCRE], [libpcre2-8], [], [:])])
 
     AC_CACHE_CHECK([for pcre2_compile], [pcre_cv_have_pcre2_compile],
       [pcre_saved_CFLAGS=$CFLAGS
        pcre_saved_LIBS=$LIBS
-       CFLAGS="$CFLAGS $PCRE_CFLAGS"
-       LIBS="$PCRE_LIBS $LIBS"
-       AC_LINK_IFELSE(
-         [AC_LANG_PROGRAM([[#define PCRE2_CODE_UNIT_WIDTH 8
-                            #include <pcre2.h>
-                          ]],
-            [[pcre2_code *p = pcre2_compile (0, 0, 0, 0, 0, 0);
-              return !p;]])],
-         [pcre_cv_have_pcre2_compile=yes],
-         [pcre_cv_have_pcre2_compile=no])
+       pcre_cv_have_pcre2_compile=no
+
+       while
+         CFLAGS="$pcre_saved_CFLAGS $PCRE_CFLAGS"
+         LIBS="$pcre_saved_LIBS $PCRE_LIBS"
+         AC_LINK_IFELSE(
+           [AC_LANG_PROGRAM([[#define PCRE2_CODE_UNIT_WIDTH 8
+                              #include <pcre2.h>
+                            ]],
+              [[pcre2_code *p = pcre2_compile (0, 0, 0, 0, 0, 0);
+                return !p;]])],
+           [pcre_cv_have_pcre2_compile=yes])
+         test $pcre_cv_have_pcre2_compile = no
+       do
+         AS_CASE([$PCRE_CFLAGS@$PCRE_LIBS],
+           [@-lpcre2-8],
+             [# Even the fallback setting fails; give up.
+              PCRE_LIBS=
+              break])
+         # Fallback setting.
+         PCRE_CFLAGS=
+         PCRE_LIBS=-lpcre2-8
+       done
+
        CFLAGS=$pcre_saved_CFLAGS
        LIBS=$pcre_saved_LIBS])
 

po/POTFILES.in

@@ -1,6 +1,6 @@
 # List of files which containing translatable strings.
 #
-# Copyright 1997-1998, 2005-2023 Free Software Foundation, Inc.
+# Copyright 1997-1998, 2005-2026 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

src/Makefile.am

@@ -1,5 +1,5 @@
 ## Process this file with automake to create Makefile.in
-# Copyright 1997-1998, 2005-2023 Free Software Foundation, Inc.
+# Copyright 1997-1998, 2005-2026 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
@@ -28,13 +28,12 @@ grep_SOURCES =					\
   die.h						\
   grep.c					\
   kwsearch.c					\
-  kwset.c					\
   searchutils.c
 if USE_PCRE
 grep_SOURCES += pcresearch.c
 endif
 
-noinst_HEADERS = grep.h kwset.h search.h system.h
+noinst_HEADERS = grep.h search.h system.h
 
 # Sometimes, the expansion of $(LIBINTL) includes -lc which may
 # include modules defining variables like 'optind', so libgreputils.a
@@ -42,7 +41,9 @@ noinst_HEADERS = grep.h kwset.h search.h system.h
 # But libgreputils.a must also follow $(LIBINTL), since libintl uses
 # replacement functions defined in libgreputils.a.
 LDADD = \
-  ../lib/libgreputils.a $(LIBINTL) ../lib/libgreputils.a $(LIBICONV) \
+  ../lib/libgreputils.a $(LIBINTL) ../lib/libgreputils.a \
+  $(HARD_LOCALE_LIB) $(LIBC32CONV) \
+  $(LIBSIGSEGV) $(LIBUNISTRING) $(MBRTOWC_LIB) $(SETLOCALE_NULL_LIB) \
   $(LIBTHREAD)
 
 grep_LDADD = $(LDADD) $(PCRE_LIBS) $(LIBCSTACK)

src/dfasearch.c

@@ -1,5 +1,5 @@
 /* dfasearch.c - searching subroutines using dfa and regex for grep.
-   Copyright 1992, 1998, 2000, 2007, 2009-2023 Free Software Foundation, Inc.
+   Copyright 1992, 1998, 2000, 2007, 2009-2026 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
@@ -12,15 +12,13 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 /* Written August 1992 by Mike Haertel. */
 
 #include <config.h>
 #include "intprops.h"
-#include "search.h"
+#include <search.h>
 #include "die.h"
 #include <error.h>
 
@@ -145,15 +143,15 @@ regex_compile (struct dfa_comp *dc, char const *p, idx_t len,
                bool syntax_only)
 {
   struct re_pattern_buffer pat;
-  pat.buffer = NULL;
+  pat.buffer = nullptr;
   pat.allocated = 0;
 
   /* Do not use a fastmap with -i, to work around glibc Bug#20381.  */
   static_assert (UCHAR_MAX < IDX_MAX);
   idx_t uchar_max = UCHAR_MAX;
-  pat.fastmap = (syntax_only | match_icase) ? NULL : ximalloc (uchar_max + 1);
+  pat.fastmap = syntax_only | match_icase ? nullptr : ximalloc (uchar_max + 1);
 
-  pat.translate = NULL;
+  pat.translate = nullptr;
 
   if (syntax_only)
     re_set_syntax (syntax_bits | RE_NO_SUB);
@@ -227,7 +225,7 @@ GEAcompile (char *pattern, idx_t size, reg_syntax_t syntax_bits,
   char const *prev = pattern;
 
   /* Buffer containing back-reference-free patterns.  */
-  char *buf = NULL;
+  char *buf = nullptr;
   idx_t buflen = 0;
   idx_t bufalloc = 0;
 
@@ -324,13 +322,13 @@ GEAcompile (char *pattern, idx_t size, reg_syntax_t syntax_bits,
       size = total;
     }
   else
-    motif = NULL;
+    motif = nullptr;
 
   dfaparse (pattern, size, dc->dfa);
   kwsmusts (dc);
-  dfacomp (NULL, 0, dc->dfa, 1);
+  dfacomp (nullptr, 0, dc->dfa, 1);
 
-  if (buf != NULL)
+  if (buf)
     {
       if (exact || !dfasupported (dc->dfa))
         {
@@ -424,7 +422,7 @@ EGexecute (void *vdc, char const *buf, idx_t size, idx_t *match_size,
                     goto success;
                   if (mb_start < beg)
                     mb_start = beg;
-                  if (mb_goback (&mb_start, NULL, match, buflim) == 0)
+                  if (mb_goback (&mb_start, nullptr, match, buflim) == 0)
                     goto success;
                   /* The matched line starts in the middle of a multibyte
                      character.  Perform the DFA search starting from the
@@ -440,8 +438,8 @@ EGexecute (void *vdc, char const *buf, idx_t size, idx_t *match_size,
                  potential matches; this is more likely to be fast
                  than falling back to KWset would be.  */
               next_beg = dfaexec (superset, dfa_beg, (char *) end, 0,
-                                  &count, NULL);
-              if (next_beg == NULL || next_beg == end)
+                                  &count, nullptr);
+              if (!next_beg || next_beg == end)
                 continue;
 
               /* Narrow down to the line we've found.  */
@@ -463,7 +461,7 @@ EGexecute (void *vdc, char const *buf, idx_t size, idx_t *match_size,
 
           /* If there's no match, or if we've matched the sentinel,
              we're done.  */
-          if (next_beg == NULL || next_beg == end)
+          if (!next_beg || next_beg == end)
             continue;
 
           /* Narrow down to the line we've found.  */

src/die.h

@@ -1,5 +1,5 @@
 /* Report an error and exit.
-   Copyright 2016-2023 Free Software Foundation, Inc.
+   Copyright 2016-2026 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
@@ -12,9 +12,7 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 #ifndef DIE_H
 #define DIE_H

src/grep.c

@@ -1,5 +1,5 @@
 /* grep.c - main driver file for grep.
-   Copyright (C) 1992, 1997-2002, 2004-2023 Free Software Foundation, Inc.
+   Copyright (C) 1992, 1997-2002, 2004-2026 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
@@ -12,15 +12,14 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 /* Written July 1992 by Mike Haertel.  */
 
 #include <config.h>
 #include <sys/types.h>
 #include <sys/stat.h>
+#include <uchar.h>
 #include <wchar.h>
 #include <inttypes.h>
 #include <stdarg.h>
@@ -35,18 +34,17 @@
 #include "closeout.h"
 #include "colorize.h"
 #include "die.h"
-#include "error.h"
+#include <error.h>
 #include "exclude.h"
 #include "exitfail.h"
 #include "fcntl-safer.h"
 #include "fts_.h"
-#include "getopt.h"
+#include <getopt.h>
 #include "grep.h"
 #include "hash.h"
 #include "intprops.h"
-#include "propername.h"
 #include "safe-read.h"
-#include "search.h"
+#include <search.h>
 #include "c-strcase.h"
 #include "version-etc.h"
 #include "xalloc.h"
@@ -178,7 +176,7 @@ update_patterns (char *keys, idx_t dupfree_size, idx_t size,
 
       intptr_t dst_offset_1 = dst - keys + 1;
       int inserted = hash_insert_if_absent (pattern_table,
-                                            (void *) dst_offset_1, NULL);
+                                            (void *) dst_offset_1, nullptr);
       if (inserted)
         {
           if (inserted < 0)
@@ -425,20 +423,24 @@ color_cap_ne_fct (void)
 static const struct color_cap color_dict[] =
   {
     { "mt", &selected_match_color, color_cap_mt_fct }, /* both ms/mc */
-    { "ms", &selected_match_color, NULL }, /* selected matched text */
-    { "mc", &context_match_color,  NULL }, /* context matched text */
-    { "fn", &filename_color,       NULL }, /* filename */
-    { "ln", &line_num_color,       NULL }, /* line number */
-    { "bn", &byte_num_color,       NULL }, /* byte (sic) offset */
-    { "se", &sep_color,            NULL }, /* separator */
-    { "sl", &selected_line_color,  NULL }, /* selected lines */
-    { "cx", &context_line_color,   NULL }, /* context lines */
-    { "rv", NULL,                  color_cap_rv_fct }, /* -v reverses sl/cx */
-    { "ne", NULL,                  color_cap_ne_fct }, /* no EL on SGR_* */
-    { NULL, NULL,                  NULL }
+    { "ms", &selected_match_color, nullptr }, /* selected matched text */
+    { "mc", &context_match_color,  nullptr }, /* context matched text */
+    { "fn", &filename_color,       nullptr }, /* filename */
+    { "ln", &line_num_color,       nullptr }, /* line number */
+    { "bn", &byte_num_color,       nullptr }, /* byte (sic) offset */
+    { "se", &sep_color,            nullptr }, /* separator */
+    { "sl", &selected_line_color,  nullptr }, /* selected lines */
+    { "cx", &context_line_color,   nullptr }, /* context lines */
+    { "rv", nullptr,               color_cap_rv_fct }, /* -v reverses sl/cx */
+    { "ne", nullptr,               color_cap_ne_fct }, /* no EL on SGR_* */
+    { nullptr, nullptr,            nullptr }
   };
 
-/* Saved errno value from failed output functions on stdout.  */
+/* Saved errno value from failed output functions on stdout.
+   prline polls this to decide whether to die.
+   Setting it to nonzero just before exiting can prevent clean_up_stdout
+   from misbehaving on a buggy OS where 'close (STDOUT_FILENO)' fails
+   with EACCES.  */
 static int stdout_errno;
 
 static void
@@ -503,57 +505,56 @@ enum
 /* Long options equivalences. */
 static struct option const long_options[] =
 {
-  {"basic-regexp",    no_argument, NULL, 'G'},
-  {"extended-regexp", no_argument, NULL, 'E'},
-  {"fixed-regexp",    no_argument, NULL, 'F'},
-  {"fixed-strings",   no_argument, NULL, 'F'},
-  {"perl-regexp",     no_argument, NULL, 'P'},
-  {"after-context", required_argument, NULL, 'A'},
-  {"before-context", required_argument, NULL, 'B'},
-  {"binary-files", required_argument, NULL, BINARY_FILES_OPTION},
-  {"byte-offset", no_argument, NULL, 'b'},
-  {"context", required_argument, NULL, 'C'},
-  {"color", optional_argument, NULL, COLOR_OPTION},
-  {"colour", optional_argument, NULL, COLOR_OPTION},
-  {"count", no_argument, NULL, 'c'},
-  {"devices", required_argument, NULL, 'D'},
-  {"directories", required_argument, NULL, 'd'},
-  {"exclude", required_argument, NULL, EXCLUDE_OPTION},
-  {"exclude-from", required_argument, NULL, EXCLUDE_FROM_OPTION},
-  {"exclude-dir", required_argument, NULL, EXCLUDE_DIRECTORY_OPTION},
-  {"file", required_argument, NULL, 'f'},
-  {"files-with-matches", no_argument, NULL, 'l'},
-  {"files-without-match", no_argument, NULL, 'L'},
-  {"group-separator", required_argument, NULL, GROUP_SEPARATOR_OPTION},
+  {"basic-regexp",    no_argument, nullptr, 'G'},
+  {"extended-regexp", no_argument, nullptr, 'E'},
+  {"fixed-regexp",    no_argument, nullptr, 'F'},
+  {"fixed-strings",   no_argument, nullptr, 'F'},
+  {"perl-regexp",     no_argument, nullptr, 'P'},
+  {"after-context", required_argument, nullptr, 'A'},
+  {"before-context", required_argument, nullptr, 'B'},
+  {"binary-files", required_argument, nullptr, BINARY_FILES_OPTION},
+  {"byte-offset", no_argument, nullptr, 'b'},
+  {"context", required_argument, nullptr, 'C'},
+  {"color", optional_argument, nullptr, COLOR_OPTION},
+  {"colour", optional_argument, nullptr, COLOR_OPTION},
+  {"count", no_argument, nullptr, 'c'},
+  {"devices", required_argument, nullptr, 'D'},
+  {"directories", required_argument, nullptr, 'd'},
+  {"exclude", required_argument, nullptr, EXCLUDE_OPTION},
+  {"exclude-from", required_argument, nullptr, EXCLUDE_FROM_OPTION},
+  {"exclude-dir", required_argument, nullptr, EXCLUDE_DIRECTORY_OPTION},
+  {"file", required_argument, nullptr, 'f'},
+  {"files-with-matches", no_argument, nullptr, 'l'},
+  {"files-without-match", no_argument, nullptr, 'L'},
+  {"group-separator", required_argument, nullptr, GROUP_SEPARATOR_OPTION},
   {"help", no_argument, &show_help, 1},
-  {"include", required_argument, NULL, INCLUDE_OPTION},
-  {"ignore-case", no_argument, NULL, 'i'},
-  {"no-ignore-case", no_argument, NULL, NO_IGNORE_CASE_OPTION},
-  {"initial-tab", no_argument, NULL, 'T'},
-  {"label", required_argument, NULL, LABEL_OPTION},
-  {"line-buffered", no_argument, NULL, LINE_BUFFERED_OPTION},
-  {"line-number", no_argument, NULL, 'n'},
-  {"line-regexp", no_argument, NULL, 'x'},
-  {"max-count", required_argument, NULL, 'm'},
+  {"include", required_argument, nullptr, INCLUDE_OPTION},
+  {"ignore-case", no_argument, nullptr, 'i'},
+  {"no-ignore-case", no_argument, nullptr, NO_IGNORE_CASE_OPTION},
+  {"initial-tab", no_argument, nullptr, 'T'},
+  {"label", required_argument, nullptr, LABEL_OPTION},
+  {"line-buffered", no_argument, nullptr, LINE_BUFFERED_OPTION},
+  {"line-number", no_argument, nullptr, 'n'},
+  {"line-regexp", no_argument, nullptr, 'x'},
+  {"max-count", required_argument, nullptr, 'm'},
 
-  {"no-filename", no_argument, NULL, 'h'},
-  {"no-group-separator", no_argument, NULL, GROUP_SEPARATOR_OPTION},
-  {"no-messages", no_argument, NULL, 's'},
-  {"null", no_argument, NULL, 'Z'},
-  {"null-data", no_argument, NULL, 'z'},
-  {"only-matching", no_argument, NULL, 'o'},
-  {"quiet", no_argument, NULL, 'q'},
-  {"recursive", no_argument, NULL, 'r'},
-  {"dereference-recursive", no_argument, NULL, 'R'},
-  {"regexp", required_argument, NULL, 'e'},
-  {"invert-match", no_argument, NULL, 'v'},
-  {"silent", no_argument, NULL, 'q'},
-  {"text", no_argument, NULL, 'a'},
-  {"binary", no_argument, NULL, 'U'},
-  {"unix-byte-offsets", no_argument, NULL, 'u'},
-  {"version", no_argument, NULL, 'V'},
-  {"with-filename", no_argument, NULL, 'H'},
-  {"word-regexp", no_argument, NULL, 'w'},
+  {"no-filename", no_argument, nullptr, 'h'},
+  {"no-group-separator", no_argument, nullptr, GROUP_SEPARATOR_OPTION},
+  {"no-messages", no_argument, nullptr, 's'},
+  {"null", no_argument, nullptr, 'Z'},
+  {"null-data", no_argument, nullptr, 'z'},
+  {"only-matching", no_argument, nullptr, 'o'},
+  {"quiet", no_argument, nullptr, 'q'},
+  {"recursive", no_argument, nullptr, 'r'},
+  {"dereference-recursive", no_argument, nullptr, 'R'},
+  {"regexp", required_argument, nullptr, 'e'},
+  {"invert-match", no_argument, nullptr, 'v'},
+  {"silent", no_argument, nullptr, 'q'},
+  {"text", no_argument, nullptr, 'a'},
+  {"binary", no_argument, nullptr, 'U'},
+  {"version", no_argument, nullptr, 'V'},
+  {"with-filename", no_argument, nullptr, 'H'},
+  {"word-regexp", no_argument, nullptr, 'w'},
   {0, 0, 0, 0}
 };
 
@@ -584,7 +585,7 @@ enum directories_type
 /* How to handle directories.  */
 static char const *const directories_args[] =
 {
-  "read", "recurse", "skip", NULL
+  "read", "recurse", "skip", nullptr
 };
 static enum directories_type const directories_types[] =
 {
@@ -761,7 +762,7 @@ buf_has_encoding_errors (char *buf, idx_t size)
   if (! unibyte_mask)
     return false;
 
-  mbstate_t mbs = { 0 };
+  mbstate_t mbs; mbszero (&mbs);
   ptrdiff_t clen;
 
   buf[size] = -1;
@@ -874,6 +875,7 @@ static int bufdesc;		/* File descriptor. */
 static char *bufbeg;		/* Beginning of user-visible stuff. */
 static char *buflim;		/* Limit of user-visible stuff. */
 static idx_t pagesize;		/* alignment of memory pages */
+static idx_t good_readsize;	/* good size to pass to 'read' */
 static off_t bufoffset;		/* Read offset.  */
 static off_t after_last_match;	/* Pointer after last matching line that
                                    would have been output if we were
@@ -882,8 +884,15 @@ static bool skip_nuls;		/* Skip '\0' in data.  */
 static bool skip_empty_lines;	/* Skip empty lines in data.  */
 static intmax_t totalnl;	/* Total newline count before lastnl. */
 
-/* Initial buffer size, not counting slop. */
-enum { INITIAL_BUFSIZE = 96 * 1024 };
+/* Minimum value for good_readsize.
+   If it's too small, there are more syscalls;
+   if too large, it wastes memory and likely cache.
+   Use 96 KiB as it gave good results in a benchmark in 2018
+   (see 2018-09-06 commit labeled "grep: triple initial buffer size: 32k->96k")
+   even though the same benchmark in 2024 found no significant
+   difference for values from 32 KiB to 1024 KiB on Ubuntu 24.04.1 LTS
+   with an Intel Xeon W-1350.  */
+enum { GOOD_READSIZE_MIN = 96 * 1024 };
 
 /* Return VAL aligned to the next multiple of ALIGNMENT.  VAL can be
    an integer or a pointer.  Both args must be free of side effects.  */
@@ -948,9 +957,9 @@ fillbuf (idx_t save, struct stat const *st)
 {
   char *readbuf;
 
-  /* After BUFLIM, we need room for at least a page of data plus a
+  /* After BUFLIM, we need room for a good-sized read plus a
      trailing uword.  */
-  idx_t min_after_buflim = pagesize + uword_size;
+  idx_t min_after_buflim = good_readsize + uword_size;
 
   if (min_after_buflim <= buffer + bufalloc - buflim)
     readbuf = buflim;
@@ -959,8 +968,8 @@ fillbuf (idx_t save, struct stat const *st)
       char *newbuf;
 
       /* For data to be searched we need room for the saved bytes,
-         plus at least a page of data to read.  */
-      idx_t minsize = save + pagesize;
+         plus at least a good-sized read.  */
+      idx_t minsize = save + good_readsize;
 
       /* Add enough room so that the buffer is aligned and has room
          for byte sentinels fore and aft, and so that a uword can
@@ -986,7 +995,7 @@ fillbuf (idx_t save, struct stat const *st)
                 alloc_max = MAX (a, bufalloc + incr_min);
             }
 
-          newbuf = xpalloc (NULL, &bufalloc, incr_min, alloc_max, 1);
+          newbuf = xpalloc (nullptr, &bufalloc, incr_min, alloc_max, 1);
         }
 
       readbuf = ALIGN_TO (newbuf + 1 + save, pagesize);
@@ -1003,16 +1012,13 @@ fillbuf (idx_t save, struct stat const *st)
 
   clear_asan_poison ();
 
-  idx_t readsize = buffer + bufalloc - uword_size - readbuf;
-  readsize -= readsize % pagesize;
-
-  idx_t fillsize;
+  ptrdiff_t fillsize;
   bool cc = true;
 
   while (true)
     {
-      fillsize = safe_read (bufdesc, readbuf, readsize);
-      if (fillsize == SAFE_READ_ERROR)
+      fillsize = safe_read (bufdesc, readbuf, good_readsize);
+      if (fillsize < 0)
         {
           fillsize = 0;
           cc = false;
@@ -1087,14 +1093,14 @@ static bool count_matches;	/* Count matching lines.  */
 static intmax_t max_count;	/* Max number of selected
                                    lines from an input file.  */
 static bool line_buffered;	/* Use line buffering.  */
-static char *label = NULL;      /* Fake filename for stdin */
+static char *label;		/* Fake filename for stdin */
 
 
 /* Internal variables to keep track of byte count, context, etc. */
 static intmax_t totalcc;	/* Total character count before bufbeg. */
 static char const *lastnl;	/* Pointer after last newline counted. */
 static char *lastout;		/* Pointer after last character output;
-                                   NULL if no character has been output
+                                   null if no character has been output
                                    or if it's conceptually before bufbeg. */
 static intmax_t outleft;	/* Maximum number of selected lines.  */
 static intmax_t pending;	/* Pending lines of output.
@@ -1211,7 +1217,7 @@ print_line_middle (char *beg, char *lim,
   idx_t match_size;
   ptrdiff_t match_offset;
   char *cur;
-  char *mid = NULL;
+  char *mid = nullptr;
   char *b;
 
   for (cur = beg;
@@ -1243,7 +1249,7 @@ print_line_middle (char *beg, char *lim,
             {
               char sep = out_invert ? SEP_CHAR_REJECTED : SEP_CHAR_SELECTED;
               if (! print_line_head (b, match_size, lim, sep))
-                return NULL;
+                return nullptr;
             }
           else
             {
@@ -1251,7 +1257,7 @@ print_line_middle (char *beg, char *lim,
               if (mid)
                 {
                   cur = mid;
-                  mid = NULL;
+                  mid = nullptr;
                 }
               fwrite_errno (cur, 1, b - cur);
             }
@@ -1315,7 +1321,7 @@ prline (char *beg, char *lim, char sep)
                      ? selected_match_color : context_match_color);
     }
   else
-    line_color = match_color = NULL; /* Shouldn't be used.  */
+    line_color = match_color = nullptr; /* Shouldn't be used.  */
 
   if ((only_matching && matching)
       || (color_option && (*line_color || *match_color)))
@@ -1466,7 +1472,7 @@ grepbuf (char *beg, char const *lim)
     {
       idx_t match_size;
       ptrdiff_t match_offset = execute (compiled_pattern, p, lim - p,
-                                        &match_size, NULL);
+                                        &match_size, nullptr);
       if (match_offset < 0)
         {
           if (!out_invert)
@@ -1481,13 +1487,18 @@ grepbuf (char *beg, char const *lim)
         break;
       if (!out_invert || p < b)
         {
+          if (list_files != LISTFILES_NONE)
+            return 1;
           char *prbeg = out_invert ? p : b;
           char *prend = out_invert ? b : endp;
           prtext (prbeg, prend);
           if (!outleft || done_on_match)
             {
               if (exit_on_match)
-                exit (errseen ? exit_failure : EXIT_SUCCESS);
+                {
+                  stdout_errno = -1;
+                  exit (EXIT_SUCCESS);
+                }
               break;
             }
         }
@@ -1503,9 +1514,6 @@ grep (int fd, struct stat const *st, bool *ineof)
 {
   intmax_t nlines, i;
   idx_t residue, save;
-  char oldc;
-  char *beg;
-  char *lim;
   char eol = eolbyte;
   char nul_zapper = '\0';
   bool done_on_match_0 = done_on_match;
@@ -1558,7 +1566,11 @@ grep (int fd, struct stat const *st, bool *ineof)
           if (binary_files == WITHOUT_MATCH_BINARY_FILES)
             return 0;
           if (!count_matches)
-            done_on_match = out_quiet = true;
+            {
+              out_quiet = true;
+              if (max_count == INTMAX_MAX)
+                done_on_match = true;
+            }
           nlines_first_null = nlines;
           nul_zapper = eol;
           skip_nuls = skip_empty_lines;
@@ -1568,7 +1580,7 @@ grep (int fd, struct stat const *st, bool *ineof)
       if (lastout)
         lastout = bufbeg;
 
-      beg = bufbeg + save;
+      char *beg = bufbeg + save;
 
       /* no more data to scan (eof) except for maybe a residue -> break */
       if (beg == buflim)
@@ -1581,16 +1593,9 @@ grep (int fd, struct stat const *st, bool *ineof)
 
       /* Determine new residue (the length of an incomplete line at the end of
          the buffer, 0 means there is no incomplete last line).  */
-      oldc = beg[-1];
-      beg[-1] = eol;
-      /* If rawmemrchr existed it could be used here, since we have ensured
-         that this use of memrchr is guaranteed never to return NULL.  */
-      lim = memrchr (beg - 1, eol, buflim - beg + 1);
-      ++lim;
-      beg[-1] = oldc;
-      if (lim == beg)
-        lim = beg - residue;
+      char *last_eol = memrchr (beg, eol, buflim - beg);
       beg -= residue;
+      char *lim = last_eol ? last_eol + 1 : beg;
       residue = buflim - lim;
 
       if (beg < lim)
@@ -1776,19 +1781,19 @@ drain_input (int fd, struct stat const *st)
     {
 #ifdef SPLICE_F_MOVE
       /* Should be faster, since it need not copy data to user space.  */
-      nbytes = splice (fd, NULL, STDOUT_FILENO, NULL,
-                       INITIAL_BUFSIZE, SPLICE_F_MOVE);
+      nbytes = splice (fd, nullptr, STDOUT_FILENO, nullptr,
+                       good_readsize, SPLICE_F_MOVE);
       if (0 <= nbytes || errno != EINVAL)
         {
           while (0 < nbytes)
-            nbytes = splice (fd, NULL, STDOUT_FILENO, NULL,
-                             INITIAL_BUFSIZE, SPLICE_F_MOVE);
+            nbytes = splice (fd, nullptr, STDOUT_FILENO, nullptr,
+                             good_readsize, SPLICE_F_MOVE);
           return nbytes == 0;
         }
 #endif
     }
   while ((nbytes = safe_read (fd, buffer, bufalloc)))
-    if (nbytes == SAFE_READ_ERROR)
+    if (nbytes < 0)
       return false;
   return true;
 }
@@ -1864,8 +1869,8 @@ grepdesc (int desc, bool command_line)
         suppressible_error (errno);
 
       fts_arg[0] = (char *) filename;
-      fts_arg[1] = NULL;
-      fts = fts_open (fts_arg, opts, NULL);
+      fts_arg[1] = nullptr;
+      fts = fts_open (fts_arg, opts, nullptr);
 
       if (!fts)
         xalloc_die ();
@@ -2060,10 +2065,11 @@ Context control:\n\
   -U, --binary              do not strip CR characters at EOL (MSDOS/Windows)\n\
 \n"));
       printf (_("\
-When FILE is '-', read standard input.  With no FILE, read '.' if\n\
-recursive, '-' otherwise.  With fewer than two FILEs, assume -h.\n\
-Exit status is 0 if any line is selected, 1 otherwise;\n\
-if any error occurs and -q is not given, the exit status is 2.\n"));
+When FILE is '-', read standard input.  If no FILE is given, read standard\n\
+input, but with -r, recursively search the working directory instead.  With\n\
+fewer than two FILEs, assume -h.  Exit status is 0 if any line is selected,\n\
+1 otherwise; if any error occurs and -q is not given, the exit status is 2.\n"
+                ));
       emit_bug_reporting_address ();
     }
   exit (status);
@@ -2132,7 +2138,7 @@ get_nondigit_option (int argc, char *const *argv, intmax_t *default_context)
   while (true)
     {
       opt = getopt_long (argc, (char **) argv, short_options,
-                         long_options, NULL);
+                         long_options, nullptr);
       if (! c_isdigit (opt))
         break;
 
@@ -2185,14 +2191,14 @@ parse_grep_colors (void)
   char *val;
 
   p = getenv ("GREP_COLORS"); /* Plural! */
-  if (p == NULL || *p == '\0')
+  if (!p || *p == '\0')
     return;
 
   /* Work off a writable copy.  */
   q = xstrdup (p);
 
   name = q;
-  val = NULL;
+  val = nullptr;
   /* From now on, be well-formed or you're gone.  */
   for (;;)
     if (*q == ':' || *q == '\0')
@@ -2214,7 +2220,7 @@ parse_grep_colors (void)
         if (c == '\0')
           return;
         name = q;
-        val = NULL;
+        val = nullptr;
       }
     else if (*q == '=')
       {
@@ -2223,7 +2229,7 @@ parse_grep_colors (void)
         *q++ = '\0'; /* Terminate name.  */
         val = q; /* Can be the empty string.  */
       }
-    else if (val == NULL)
+    else if (!val)
       q++; /* Accumulate name.  */
     else if (*q == ';' || c_isdigit (*q))
       q++; /* Accumulate val.  Protect the terminal from being sent crap.  */
@@ -2235,7 +2241,7 @@ parse_grep_colors (void)
 static bool
 contains_encoding_error (char const *pat, idx_t patlen)
 {
-  mbstate_t mbs = { 0 };
+  mbstate_t mbs; mbszero (&mbs);
   ptrdiff_t charlen;
 
   for (idx_t i = 0; i < patlen; i += charlen)
@@ -2261,12 +2267,12 @@ setup_ok_fold (void)
         continue;
 
       int ok = 1;
-      wchar_t folded[CASE_FOLDED_BUFSIZE];
+      char32_t folded[CASE_FOLDED_BUFSIZE];
       for (int n = case_folded_counterparts (wi, folded); 0 <= --n; )
         {
           char buf[MB_LEN_MAX];
-          mbstate_t s = { 0 };
-          if (wcrtomb (buf, folded[n], &s) != 1)
+          mbstate_t s; mbszero (&s);
+          if (c32rtomb (buf, folded[n], &s) != 1)
             {
               ok = -1;
               break;
@@ -2291,8 +2297,8 @@ fgrep_icase_charlen (char const *pat, idx_t patlen, mbstate_t *mbs)
   if (localeinfo.sbctowc[pat0] != WEOF)
     return ok_fold[pat0];
 
-  wchar_t wc;
-  size_t wn = mbrtowc (&wc, pat, patlen, mbs);
+  char32_t wc;
+  size_t wn = mbrtoc32 (&wc, pat, patlen, mbs);
 
   /* If PAT starts with an encoding error, Fcompile does not work.  */
   if (MB_LEN_MAX < wn)
@@ -2301,7 +2307,7 @@ fgrep_icase_charlen (char const *pat, idx_t patlen, mbstate_t *mbs)
   /* PAT starts with a multibyte character.  Fcompile works if the
      character has no case folded counterparts and toupper translates
      none of its encoding's bytes.  */
-  wchar_t folded[CASE_FOLDED_BUFSIZE];
+  char32_t folded[CASE_FOLDED_BUFSIZE];
   if (case_folded_counterparts (wc, folded))
     return -1;
   for (idx_t i = wn; 0 < --i; )
@@ -2321,7 +2327,7 @@ fgrep_icase_charlen (char const *pat, idx_t patlen, mbstate_t *mbs)
 static bool
 fgrep_icase_available (char const *pat, idx_t patlen)
 {
-  mbstate_t mbs = {0,};
+  mbstate_t mbs; mbszero (&mbs);
 
   for (idx_t i = 0; i < patlen; )
     {
@@ -2341,7 +2347,7 @@ fgrep_to_grep_pattern (char **keys_p, idx_t *len_p)
 {
   idx_t len = *len_p;
   char *keys = *keys_p;
-  mbstate_t mb_state = { 0 };
+  mbstate_t mb_state; mbszero (&mb_state);
   char *new_keys = xnmalloc (len + 1, 2);
   char *p = new_keys;
 
@@ -2393,7 +2399,7 @@ try_fgrep_pattern (int matcher, char *keys, idx_t *len_p)
   char *new_keys = ximalloc (len + 1);
   char *p = new_keys;
   char const *q = keys;
-  mbstate_t mb_state = { 0 };
+  mbstate_t mb_state; mbszero (&mb_state);
 
   while (len != 0)
     {
@@ -2461,7 +2467,7 @@ try_fgrep_pattern (int matcher, char *keys, idx_t *len_p)
 int
 main (int argc, char **argv)
 {
-  char *keys = NULL;
+  char *keys = nullptr;
   idx_t keycc = 0, keyalloc = 0;
   int matcher = -1;
   int opt;
@@ -2493,13 +2499,14 @@ main (int argc, char **argv)
 #endif
 #if defined ENABLE_NLS
   bindtextdomain (PACKAGE, LOCALEDIR);
+  bindtextdomain ("gnulib", GNULIB_LOCALEDIR);
   textdomain (PACKAGE);
 #endif
 
   init_localeinfo (&localeinfo);
 
   atexit (clean_up_stdout);
-  c_stack_action (NULL);
+  c_stack_action (nullptr);
 
   last_recursive = 0;
 
@@ -2571,11 +2578,6 @@ main (int argc, char **argv)
           binary = true;
         break;
 
-      case 'u':
-        /* Obsolete option; it had no effect; FIXME: remove in 2023  */
-        error (0, 0, _("warning: --unix-byte-offsets (-u) is obsolete"));
-        break;
-
       case 'V':
         show_version = true;
         break;
@@ -2699,7 +2701,6 @@ main (int argc, char **argv)
 
       case 'q':
         exit_on_match = true;
-        exit_failure = 0;
         break;
 
       case 'R':
@@ -2827,9 +2828,12 @@ main (int argc, char **argv)
   if (show_version)
     {
       version_etc (stdout, getprogname (), PACKAGE_NAME, VERSION,
-                   (char *) NULL);
+                   (char *) nullptr);
       puts (_("Written by Mike Haertel and others; see\n"
               "<https://git.savannah.gnu.org/cgit/grep.git/tree/AUTHORS>."));
+#if HAVE_LIBPCRE
+      Pprint_version ();
+#endif
       return EXIT_SUCCESS;
     }
 
@@ -2894,9 +2898,10 @@ main (int argc, char **argv)
   if ((exit_on_match | dev_null_output) || list_files != LISTFILES_NONE)
     {
       count_matches = false;
-      done_on_match = true;
+      if (max_count == INTMAX_MAX)
+        done_on_match = true;
     }
-  out_quiet = count_matches | done_on_match;
+  out_quiet = count_matches | done_on_match | exit_on_match;
 
   if (out_after < 0)
     out_after = default_context;
@@ -2918,7 +2923,7 @@ main (int argc, char **argv)
     {
       /* Legacy.  */
       char *userval = getenv ("GREP_COLOR");
-      if (userval != NULL && *userval != '\0')
+      if (userval && *userval)
         for (char *q = userval; *q == ';' || c_isdigit (*q); q++)
           if (!q[1])
             {
@@ -2977,8 +2982,8 @@ main (int argc, char **argv)
   /* We need one byte prior and one after.  */
   char eolbytes[3] = { 0, eolbyte, 0 };
   idx_t match_size;
-  skip_empty_lines = ((execute (compiled_pattern, eolbytes + 1, 1,
-                                &match_size, NULL) == 0)
+  skip_empty_lines = (!execute (compiled_pattern, eolbytes + 1, 1,
+                                &match_size, nullptr)
                       == out_invert);
 
   int num_operands = argc - optind;
@@ -2998,7 +3003,8 @@ main (int argc, char **argv)
   if (! (0 < psize && psize <= (IDX_MAX - uword_size) / 2))
     abort ();
   pagesize = psize;
-  bufalloc = ALIGN_TO (INITIAL_BUFSIZE, pagesize) + pagesize + uword_size;
+  good_readsize = ALIGN_TO (GOOD_READSIZE_MIN, pagesize);
+  bufalloc = good_readsize + pagesize + uword_size;
   buffer = ximalloc (bufalloc);
 
   if (fts_options & FTS_LOGICAL && devices == READ_COMMAND_LINE_DEVICES)
@@ -3011,20 +3017,20 @@ main (int argc, char **argv)
     }
   else if (directories == RECURSE_DIRECTORIES && 0 < last_recursive)
     {
-      static char *const cwd_only[] = { (char *) ".", NULL };
+      static char *const cwd_only[] = { (char *) ".", nullptr };
       files = cwd_only;
       omit_dot_slash = true;
     }
   else
     {
-      static char *const stdin_only[] = { (char *) "-", NULL };
+      static char *const stdin_only[] = { (char *) "-", nullptr };
       files = stdin_only;
     }
 
   bool status = true;
   do
     status &= grep_command_line_arg (*files++);
-  while (*files != NULL);
+  while (*files);
 
   return errseen ? EXIT_TROUBLE : status;
 }

src/grep.h

@@ -1,5 +1,5 @@
 /* grep.h - interface to grep driver for searching subroutines.
-   Copyright (C) 1992, 1998, 2001, 2007, 2009-2023 Free Software Foundation,
+   Copyright (C) 1992, 1998, 2001, 2007, 2009-2026 Free Software Foundation,
    Inc.
 
    This program is free software; you can redistribute it and/or modify
@@ -13,9 +13,7 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 #ifndef GREP_GREP_H
 #define GREP_GREP_H 1

src/kwsearch.c

@@ -1,5 +1,5 @@
 /* kwsearch.c - searching subroutines using kwset for grep.
-   Copyright 1992, 1998, 2000, 2007, 2009-2023 Free Software Foundation, Inc.
+   Copyright 1992, 1998, 2000, 2007, 2009-2026 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
@@ -12,14 +12,12 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 /* Written August 1992 by Mike Haertel. */
 
 #include <config.h>
-#include "search.h"
+#include <search.h>
 
 /* A compiled -F pattern list.  */
 
@@ -50,7 +48,7 @@ void *
 Fcompile (char *pattern, idx_t size, reg_syntax_t ignored, bool exact)
 {
   kwset_t kwset;
-  char *buf = NULL;
+  char *buf = nullptr;
   idx_t bufalloc = 0;
 
   kwset = kwsinit (true);
@@ -71,7 +69,7 @@ Fcompile (char *pattern, idx_t size, reg_syntax_t ignored, bool exact)
                 {
                   free (buf);
                   bufalloc = len;
-                  buf = xpalloc (NULL, &bufalloc, 2, -1, 1);
+                  buf = xpalloc (nullptr, &bufalloc, 2, -1, 1);
                   buf[0] = eolbyte;
                 }
               memcpy (buf + 1, p, len);
@@ -96,7 +94,7 @@ Fcompile (char *pattern, idx_t size, reg_syntax_t ignored, bool exact)
   kwsearch->words = words;
   kwsearch->pattern = pattern;
   kwsearch->size = size;
-  kwsearch->re = NULL;
+  kwsearch->re = nullptr;
   return kwsearch;
 }
 
@@ -151,7 +149,7 @@ Fexecute (void *vcp, char const *buf, idx_t size, idx_t *match_size,
         goto success_in_beg_and_len;
       if (match_lines)
         {
-          len += start_ptr == NULL;
+          len += !start_ptr;
           goto success_in_beg_and_len;
         }
       if (! match_words)
@@ -199,7 +197,7 @@ Fexecute (void *vcp, char const *buf, idx_t size, idx_t *match_size,
                   end = buf + size;
 
                 if (0 <= EGexecute (kwsearch->re, beg, end - beg,
-                                    match_size, NULL))
+                                    match_size, nullptr))
                   goto success_match_words;
                 beg = end - 1;
                 break;

src/kwset.c

@@ -1,912 +0,0 @@
-/* kwset.c - search for any of a set of keywords.
-   Copyright (C) 1989, 1998, 2000, 2005, 2007, 2009-2023 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, 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.  */
-
-/* Written August 1989 by Mike Haertel.  */
-
-/* For more on the Aho-Corasick and Boyer-Moore algorithms,
-   as well as other algorithms that might help improve performance,
-   see the grep manual's "Performance" chapter.  */
-
-#include <config.h>
-
-#include "kwset.h"
-
-#include <stdckdint.h>
-#include <stdint.h>
-#include <sys/types.h>
-
-#include "system.h"
-#include "memchr2.h"
-#include "obstack.h"
-#include "xalloc.h"
-#include "verify.h"
-
-#define obstack_chunk_alloc xmalloc
-#define obstack_chunk_free free
-
-static unsigned char
-U (char ch)
-{
-  return to_uchar (ch);
-}
-
-/* Balanced tree of edges and labels leaving a given trie node.  */
-struct tree
-{
-  struct tree *llink;		/* Left link; MUST be first field.  */
-  struct tree *rlink;		/* Right link (to larger labels).  */
-  struct trie *trie;		/* Trie node pointed to by this edge.  */
-  unsigned char label;		/* Label on this edge.  */
-  char balance;			/* Difference in depths of subtrees.  */
-};
-
-/* Node of a trie representing a set of keywords.  */
-struct trie
-{
-  /* If an accepting node, this is either 2*W + 1 where W is the word
-     index, or is -1 if Aho-Corasick is in use and FAIL
-     specifies where to look for more info.  If not an accepting node,
-     this is zero.  */
-  ptrdiff_t accepting;
-
-  struct tree *links;		/* Tree of edges leaving this node.  */
-  struct trie *parent;		/* Parent of this node.  */
-  struct trie *next;		/* List of all trie nodes in level order.  */
-  struct trie *fail;		/* Aho-Corasick failure function.  */
-  idx_t depth;			/* Depth of this node from the root.  */
-  idx_t shift;			/* Shift function for search failures.  */
-  idx_t maxshift;		/* Max shift of self and descendants.  */
-};
-
-/* Structure returned opaquely to the caller, containing everything.  */
-struct kwset
-{
-  struct obstack obstack;	/* Obstack for node allocation.  */
-  idx_t words;			/* Number of words in the trie.  */
-  struct trie *trie;		/* The trie itself.  */
-  idx_t mind;			/* Minimum depth of an accepting node.  */
-  unsigned char delta[NCHAR];	/* Delta table for rapid search.  */
-  struct trie *next[NCHAR];	/* Table of children of the root.  */
-  char *target;			/* Target string if there's only one.  */
-  idx_t *shift;			/* Used in Boyer-Moore search for one
-                                   string.  */
-  char const *trans;		/* Character translation table.  */
-
-  /* This helps to match a terminal byte, which is the first byte
-     for Aho-Corasick, and the last byte for Boyer-More.  If all the
-     patterns have the same terminal byte (after translation via TRANS
-     if TRANS is nonnull), then this is that byte as an unsigned char.
-     Otherwise this is -1 if there is disagreement among the strings
-     about terminal bytes, and -2 if there are no terminal bytes and
-     no disagreement because all the patterns are empty.  */
-  int gc1;
-
-  /* This helps to match a terminal byte.  If 0 <= GC1HELP, B is
-     terminal when B == GC1 || B == GC1HELP (note that GC1 == GCHELP
-     is common here).  This is typically faster than evaluating
-     to_uchar (TRANS[B]) == GC1.  */
-  int gc1help;
-
-  /* If the string has two or more bytes, this is the penultimate byte,
-     after translation via TRANS if TRANS is nonnull.  This variable
-     is used only by Boyer-Moore.  */
-  char gc2;
-
-  /* kwsexec implementation.  */
-  ptrdiff_t (*kwsexec) (kwset_t, char const *, idx_t, struct kwsmatch *, bool);
-};
-
-/* Use TRANS to transliterate C.  A null TRANS does no transliteration.  */
-static inline char
-tr (char const *trans, char c)
-{
-  return trans ? trans[U(c)] : c;
-}
-
-static ptrdiff_t acexec (kwset_t, char const *, idx_t,
-                         struct kwsmatch *, bool);
-static ptrdiff_t bmexec (kwset_t, char const *, idx_t,
-                         struct kwsmatch *, bool);
-
-/* Return a newly allocated keyword set.  A nonnull TRANS specifies a
-   table of character translations to be applied to all pattern and
-   search text.  */
-kwset_t
-kwsalloc (char const *trans)
-{
-  struct kwset *kwset = xmalloc (sizeof *kwset);
-
-  obstack_init (&kwset->obstack);
-  kwset->words = 0;
-  kwset->trie = obstack_alloc (&kwset->obstack, sizeof *kwset->trie);
-  kwset->trie->accepting = 0;
-  kwset->trie->links = NULL;
-  kwset->trie->parent = NULL;
-  kwset->trie->next = NULL;
-  kwset->trie->fail = NULL;
-  kwset->trie->depth = 0;
-  kwset->trie->shift = 0;
-  kwset->mind = IDX_MAX;
-  kwset->target = NULL;
-  kwset->trans = trans;
-  kwset->kwsexec = acexec;
-
-  return kwset;
-}
-
-/* This upper bound is valid for CHAR_BIT >= 4 and
-   exact for CHAR_BIT in { 4..11, 13, 15, 17, 19 }.  */
-enum { DEPTH_SIZE = CHAR_BIT + CHAR_BIT / 2 };
-
-/* Add the given string to the contents of the keyword set.  */
-void
-kwsincr (kwset_t kwset, char const *text, idx_t len)
-{
-  assume (0 <= len);
-  struct trie *trie = kwset->trie;
-  char const *trans = kwset->trans;
-  bool reverse = kwset->kwsexec == bmexec;
-
-  if (reverse)
-    text += len;
-
-  /* Descend the trie (built of keywords) character-by-character,
-     installing new nodes when necessary.  */
-  while (len--)
-    {
-      unsigned char uc = reverse ? *--text : *text++;
-      unsigned char label = trans ? trans[uc] : uc;
-
-      /* Descend the tree of outgoing links for this trie node,
-         looking for the current character and keeping track
-         of the path followed.  */
-      struct tree *cur = trie->links;
-      struct tree *links[DEPTH_SIZE];
-      enum { L, R } dirs[DEPTH_SIZE];
-      links[0] = (struct tree *) &trie->links;
-      dirs[0] = L;
-      idx_t depth = 1;
-
-      while (cur && label != cur->label)
-        {
-          links[depth] = cur;
-          if (label < cur->label)
-            dirs[depth++] = L, cur = cur->llink;
-          else
-            dirs[depth++] = R, cur = cur->rlink;
-        }
-
-      /* The current character doesn't have an outgoing link at
-         this trie node, so build a new trie node and install
-         a link in the current trie node's tree.  */
-      if (!cur)
-        {
-          cur = obstack_alloc (&kwset->obstack, sizeof *cur);
-          cur->llink = NULL;
-          cur->rlink = NULL;
-          cur->trie = obstack_alloc (&kwset->obstack, sizeof *cur->trie);
-          cur->trie->accepting = 0;
-          cur->trie->links = NULL;
-          cur->trie->parent = trie;
-          cur->trie->next = NULL;
-          cur->trie->fail = NULL;
-          cur->trie->depth = trie->depth + 1;
-          cur->trie->shift = 0;
-          cur->label = label;
-          cur->balance = 0;
-
-          /* Install the new tree node in its parent.  */
-          if (dirs[--depth] == L)
-            links[depth]->llink = cur;
-          else
-            links[depth]->rlink = cur;
-
-          /* Back up the tree fixing the balance flags.  */
-          while (depth && !links[depth]->balance)
-            {
-              if (dirs[depth] == L)
-                --links[depth]->balance;
-              else
-                ++links[depth]->balance;
-              --depth;
-            }
-
-          /* Rebalance the tree by pointer rotations if necessary.  */
-          if (depth && ((dirs[depth] == L && --links[depth]->balance)
-                        || (dirs[depth] == R && ++links[depth]->balance)))
-            {
-              struct tree *t, *r, *l, *rl, *lr;
-
-              switch (links[depth]->balance)
-                {
-                case (char) -2:
-                  switch (dirs[depth + 1])
-                    {
-                    case L:
-                      r = links[depth], t = r->llink, rl = t->rlink;
-                      t->rlink = r, r->llink = rl;
-                      t->balance = r->balance = 0;
-                      break;
-                    case R:
-                      r = links[depth], l = r->llink, t = l->rlink;
-                      rl = t->rlink, lr = t->llink;
-                      t->llink = l, l->rlink = lr, t->rlink = r, r->llink = rl;
-                      l->balance = t->balance != 1 ? 0 : -1;
-                      r->balance = t->balance != (char) -1 ? 0 : 1;
-                      t->balance = 0;
-                      break;
-                    default:
-                      abort ();
-                    }
-                  break;
-                case 2:
-                  switch (dirs[depth + 1])
-                    {
-                    case R:
-                      l = links[depth], t = l->rlink, lr = t->llink;
-                      t->llink = l, l->rlink = lr;
-                      t->balance = l->balance = 0;
-                      break;
-                    case L:
-                      l = links[depth], r = l->rlink, t = r->llink;
-                      lr = t->llink, rl = t->rlink;
-                      t->llink = l, l->rlink = lr, t->rlink = r, r->llink = rl;
-                      l->balance = t->balance != 1 ? 0 : -1;
-                      r->balance = t->balance != (char) -1 ? 0 : 1;
-                      t->balance = 0;
-                      break;
-                    default:
-                      abort ();
-                    }
-                  break;
-                default:
-                  abort ();
-                }
-
-              if (dirs[depth - 1] == L)
-                links[depth - 1]->llink = t;
-              else
-                links[depth - 1]->rlink = t;
-            }
-        }
-
-      trie = cur->trie;
-    }
-
-  /* Mark the node finally reached as accepting, encoding the
-     index number of this word in the keyword set so far.  */
-  if (!trie->accepting)
-    trie->accepting = 2 * kwset->words + 1;
-  ++kwset->words;
-
-  /* Keep track of the longest and shortest string of the keyword set.  */
-  if (trie->depth < kwset->mind)
-    kwset->mind = trie->depth;
-}
-
-idx_t
-kwswords (kwset_t kwset)
-{
-  return kwset->words;
-}
-
-/* Enqueue the trie nodes referenced from the given tree in the
-   given queue.  */
-static void
-enqueue (struct tree *tree, struct trie **last)
-{
-  if (!tree)
-    return;
-  enqueue (tree->llink, last);
-  enqueue (tree->rlink, last);
-  (*last) = (*last)->next = tree->trie;
-}
-
-/* Compute the Aho-Corasick failure function for the trie nodes referenced
-   from the given tree, given the failure function for their parent as
-   well as a last resort failure node.  */
-static void
-treefails (struct tree const *tree, struct trie const *fail,
-           struct trie *recourse, bool reverse)
-{
-  struct tree *cur;
-
-  if (!tree)
-    return;
-
-  treefails (tree->llink, fail, recourse, reverse);
-  treefails (tree->rlink, fail, recourse, reverse);
-
-  /* Find, in the chain of fails going back to the root, the first
-     node that has a descendant on the current label.  */
-  while (fail)
-    {
-      cur = fail->links;
-      while (cur && tree->label != cur->label)
-        if (tree->label < cur->label)
-          cur = cur->llink;
-        else
-          cur = cur->rlink;
-      if (cur)
-        {
-          tree->trie->fail = cur->trie;
-          if (!reverse && cur->trie->accepting && !tree->trie->accepting)
-            tree->trie->accepting = -1;
-          return;
-        }
-      fail = fail->fail;
-    }
-
-  tree->trie->fail = recourse;
-}
-
-/* Set delta entries for the links of the given tree such that
-   the preexisting delta value is larger than the current depth.  */
-static void
-treedelta (struct tree const *tree, idx_t depth, unsigned char delta[])
-{
-  if (!tree)
-    return;
-  treedelta (tree->llink, depth, delta);
-  treedelta (tree->rlink, depth, delta);
-  if (depth < delta[tree->label])
-    delta[tree->label] = depth;
-}
-
-/* Return true if A has every label in B.  */
-static bool _GL_ATTRIBUTE_PURE
-hasevery (struct tree const *a, struct tree const *b)
-{
-  if (!b)
-    return true;
-  if (!hasevery (a, b->llink))
-    return false;
-  if (!hasevery (a, b->rlink))
-    return false;
-  while (a && b->label != a->label)
-    if (b->label < a->label)
-      a = a->llink;
-    else
-      a = a->rlink;
-  return !!a;
-}
-
-/* Compute a vector, indexed by character code, of the trie nodes
-   referenced from the given tree.  */
-static void
-treenext (struct tree const *tree, struct trie *next[])
-{
-  if (!tree)
-    return;
-  treenext (tree->llink, next);
-  treenext (tree->rlink, next);
-  next[tree->label] = tree->trie;
-}
-
-/* Prepare a built keyword set for use.  */
-void
-kwsprep (kwset_t kwset)
-{
-  char const *trans = kwset->trans;
-  unsigned char deltabuf[NCHAR];
-  unsigned char *delta = trans ? deltabuf : kwset->delta;
-  struct trie *curr, *last;
-
-  /* Use Boyer-Moore if just one pattern, Aho-Corasick otherwise.  */
-  bool reverse = kwset->words == 1;
-
-  if (reverse)
-    {
-      kwset_t new_kwset;
-
-      /* Enqueue the immediate descendants in the level order queue.  */
-      for (curr = last = kwset->trie; curr; curr = curr->next)
-        enqueue (curr->links, &last);
-
-      /* Looking for just one string.  Extract it from the trie.  */
-      kwset->target = obstack_alloc (&kwset->obstack, kwset->mind);
-      curr = kwset->trie;
-      for (idx_t i = 0; i < kwset->mind; i++)
-        {
-          kwset->target[i] = curr->links->label;
-          curr = curr->next;
-        }
-
-      new_kwset = kwsalloc (kwset->trans);
-      new_kwset->kwsexec = bmexec;
-      kwsincr (new_kwset, kwset->target, kwset->mind);
-      obstack_free (&kwset->obstack, NULL);
-      *kwset = *new_kwset;
-      free (new_kwset);
-    }
-
-  /* Initial values for the delta table; will be changed later.  The
-     delta entry for a given character is the smallest depth of any
-     node at which an outgoing edge is labeled by that character.  */
-  memset (delta, MIN (kwset->mind, UCHAR_MAX), sizeof deltabuf);
-
-  /* Traverse the nodes of the trie in level order, simultaneously
-     computing the delta table, failure function, and shift function.  */
-  for (curr = last = kwset->trie; curr; curr = curr->next)
-    {
-      /* Enqueue the immediate descendants in the level order queue.  */
-      enqueue (curr->links, &last);
-
-      /* Update the delta table for the descendants of this node.  */
-      treedelta (curr->links, curr->depth, delta);
-
-      /* Compute the failure function for the descendants of this node.  */
-      treefails (curr->links, curr->fail, kwset->trie, reverse);
-
-      if (reverse)
-        {
-          curr->shift = kwset->mind;
-          curr->maxshift = kwset->mind;
-
-          /* Update the shifts at each node in the current node's chain
-             of fails back to the root.  */
-          struct trie *fail;
-          for (fail = curr->fail; fail; fail = fail->fail)
-            {
-              /* If the current node has some outgoing edge that the fail
-                 doesn't, then the shift at the fail should be no larger
-                 than the difference of their depths.  */
-              if (!hasevery (fail->links, curr->links))
-                if (curr->depth - fail->depth < fail->shift)
-                  fail->shift = curr->depth - fail->depth;
-
-              /* If the current node is accepting then the shift at the
-                 fail and its descendants should be no larger than the
-                 difference of their depths.  */
-              if (curr->accepting && fail->maxshift > curr->depth - fail->depth)
-                fail->maxshift = curr->depth - fail->depth;
-            }
-        }
-    }
-
-  if (reverse)
-    {
-      /* Traverse the trie in level order again, fixing up all nodes whose
-         shift exceeds their inherited maxshift.  */
-      for (curr = kwset->trie->next; curr; curr = curr->next)
-        {
-          if (curr->maxshift > curr->parent->maxshift)
-            curr->maxshift = curr->parent->maxshift;
-          if (curr->shift > curr->maxshift)
-            curr->shift = curr->maxshift;
-        }
-    }
-
-  /* Create a vector, indexed by character code, of the outgoing links
-     from the root node.  Accumulate GC1 and GC1HELP.  */
-  struct trie *nextbuf[NCHAR];
-  struct trie **next = trans ? nextbuf : kwset->next;
-  memset (next, 0, sizeof nextbuf);
-  treenext (kwset->trie->links, next);
-  int gc1 = -2;
-  int gc1help = -1;
-  for (int i = 0; i < NCHAR; i++)
-    {
-      int ti = i;
-      if (trans)
-        {
-          ti = U(trans[i]);
-          kwset->next[i] = next[ti];
-        }
-      if (kwset->next[i])
-        {
-          if (gc1 < -1)
-            {
-              gc1 = ti;
-              gc1help = i;
-            }
-          else if (gc1 == ti)
-            gc1help = gc1help == ti ? i : -1;
-          else if (i == ti && gc1 == gc1help)
-            gc1help = i;
-          else
-            gc1 = -1;
-        }
-    }
-  kwset->gc1 = gc1;
-  kwset->gc1help = gc1help;
-
-  if (reverse)
-    {
-      /* Looking for just one string.  Extract it from the trie.  */
-      kwset->target = obstack_alloc (&kwset->obstack, kwset->mind);
-      curr = kwset->trie;
-      for (idx_t i = kwset->mind; 0 < i; i--)
-        {
-          kwset->target[i - 1] = curr->links->label;
-          curr = curr->next;
-        }
-
-      if (kwset->mind > 1)
-        {
-          /* Looking for the delta2 shift that might be made after a
-             backwards match has failed.  Extract it from the trie.  */
-          kwset->shift
-            = obstack_alloc (&kwset->obstack,
-                             sizeof *kwset->shift * (kwset->mind - 1));
-          curr = kwset->trie->next;
-          for (idx_t i = 0; i < kwset->mind - 1; i++)
-            {
-              kwset->shift[i] = curr->shift;
-              curr = curr->next;
-            }
-
-          /* The penultimate byte.  */
-          kwset->gc2 = tr (trans, kwset->target[kwset->mind - 2]);
-        }
-    }
-
-  /* Fix things up for any translation table.  */
-  if (trans)
-    for (int i = 0; i < NCHAR; ++i)
-      kwset->delta[i] = delta[U(trans[i])];
-}
-
-/* Delta2 portion of a Boyer-Moore search.  *TP is the string text
-   pointer; it is updated in place.  EP is the end of the string text,
-   and SP the end of the pattern.  LEN is the pattern length; it must
-   be at least 2.  TRANS, if nonnull, is the input translation table.
-   GC1 and GC2 are the last and second-from last bytes of the pattern,
-   transliterated by TRANS; the caller precomputes them for
-   efficiency.  If D1 is nonnull, it is a delta1 table for shifting *TP
-   when failing.  KWSET->shift says how much to shift.  */
-static inline bool
-bm_delta2_search (char const **tpp, char const *ep, char const *sp,
-                  idx_t len,
-                  char const *trans, char gc1, char gc2,
-                  unsigned char const *d1, kwset_t kwset)
-{
-  char const *tp = *tpp;
-  idx_t d = len, skip = 0;
-
-  while (true)
-    {
-      idx_t i = 2;
-      if (tr (trans, tp[-2]) == gc2)
-        {
-          while (++i <= d)
-            if (tr (trans, tp[-i]) != tr (trans, sp[-i]))
-              break;
-          if (i > d)
-            {
-              for (i = d + skip + 1; i <= len; ++i)
-                if (tr (trans, tp[-i]) != tr (trans, sp[-i]))
-                  break;
-              if (i > len)
-                {
-                  *tpp = tp - len;
-                  return true;
-                }
-            }
-        }
-
-      tp += d = kwset->shift[i - 2];
-      if (tp > ep)
-        break;
-      if (tr (trans, tp[-1]) != gc1)
-        {
-          if (d1)
-            tp += d1[U(tp[-1])];
-          break;
-        }
-      skip = i - 1;
-    }
-
-  *tpp = tp;
-  return false;
-}
-
-/* Return the address of the first byte in the buffer S (of size N)
-   that matches the terminal byte specified by KWSET, or NULL if there
-   is no match.  KWSET->gc1 should be nonnegative.  */
-static char const *
-memchr_kwset (char const *s, idx_t n, kwset_t kwset)
-{
-  char const *slim = s + n;
-  if (kwset->gc1help < 0)
-    {
-      for (; s < slim; s++)
-        if (kwset->next[U(*s)])
-          return s;
-    }
-  else
-    {
-      int small_heuristic = 2;
-      idx_t small_bytes = small_heuristic * sizeof (unsigned long int);
-      while (s < slim)
-        {
-          if (kwset->next[U(*s)])
-            return s;
-          s++;
-          if ((uintptr_t) s % small_bytes == 0)
-            return memchr2 (s, kwset->gc1, kwset->gc1help, slim - s);
-        }
-    }
-  return NULL;
-}
-
-/* Fast Boyer-Moore search (inlinable version).  */
-static inline ptrdiff_t _GL_ATTRIBUTE_PURE
-bmexec_trans (kwset_t kwset, char const *text, idx_t size)
-{
-  assume (0 <= size);
-  unsigned char const *d1;
-  char const *ep, *sp, *tp;
-  int d;
-  idx_t len = kwset->mind;
-  char const *trans = kwset->trans;
-
-  if (len == 0)
-    return 0;
-  if (len > size)
-    return -1;
-  if (len == 1)
-    {
-      tp = memchr_kwset (text, size, kwset);
-      return tp ? tp - text : -1;
-    }
-
-  d1 = kwset->delta;
-  sp = kwset->target + len;
-  tp = text + len;
-  char gc1 = kwset->gc1;
-  char gc2 = kwset->gc2;
-
-  /* Significance of 12: 1 (initial offset) + 10 (skip loop) + 1 (md2).  */
-  idx_t len12;
-  if (!ckd_mul (&len12, len, 12) && len12 < size)
-    /* 11 is not a bug, the initial offset happens only once.  */
-    for (ep = text + size - 11 * len; tp <= ep; )
-      {
-        char const *tp0 = tp;
-        d = d1[U(tp[-1])], tp += d;
-        d = d1[U(tp[-1])], tp += d;
-        if (d != 0)
-          {
-            d = d1[U(tp[-1])], tp += d;
-            d = d1[U(tp[-1])], tp += d;
-            d = d1[U(tp[-1])], tp += d;
-            if (d != 0)
-              {
-                d = d1[U(tp[-1])], tp += d;
-                d = d1[U(tp[-1])], tp += d;
-                d = d1[U(tp[-1])], tp += d;
-                if (d != 0)
-                  {
-                    d = d1[U(tp[-1])], tp += d;
-                    d = d1[U(tp[-1])], tp += d;
-
-                    /* As a heuristic, prefer memchr to seeking by
-                       delta1 when the latter doesn't advance much.  */
-                    int advance_heuristic = 16 * sizeof (long);
-                    if (advance_heuristic <= tp - tp0)
-                      continue;
-                    tp--;
-                    tp = memchr_kwset (tp, text + size - tp, kwset);
-                    if (! tp)
-                      return -1;
-                    tp++;
-                    if (ep <= tp)
-                      break;
-                  }
-              }
-          }
-        if (bm_delta2_search (&tp, ep, sp, len, trans, gc1, gc2, d1, kwset))
-          return tp - text;
-      }
-
-  /* Now only a few characters are left to search.  Carefully avoid
-     ever producing an out-of-bounds pointer.  */
-  ep = text + size;
-  d = d1[U(tp[-1])];
-  while (d <= ep - tp)
-    {
-      d = d1[U((tp += d)[-1])];
-      if (d != 0)
-        continue;
-      if (bm_delta2_search (&tp, ep, sp, len, trans, gc1, gc2, NULL, kwset))
-        return tp - text;
-    }
-
-  return -1;
-}
-
-/* Fast Boyer-Moore search.  */
-static ptrdiff_t
-bmexec (kwset_t kwset, char const *text, idx_t size,
-        struct kwsmatch *kwsmatch, bool longest)
-{
-  /* Help the compiler inline in two ways, depending on whether
-     kwset->trans is null.  */
-  ptrdiff_t ret = (IGNORE_DUPLICATE_BRANCH_WARNING
-                   (kwset->trans
-                    ? bmexec_trans (kwset, text, size)
-                    : bmexec_trans (kwset, text, size)));
-  kwsmatch->index = 0;
-  kwsmatch->offset = ret;
-  kwsmatch->size = kwset->mind;
-  return ret;
-}
-
-/* Hairy multiple string search with the Aho-Corasick algorithm.
-   (inlinable version)  */
-static inline ptrdiff_t
-acexec_trans (kwset_t kwset, char const *text, idx_t len,
-              struct kwsmatch *kwsmatch, bool longest)
-{
-  struct trie const *trie, *accept;
-  char const *tp, *left, *lim;
-  struct tree const *tree;
-  char const *trans;
-
-  /* Initialize register copies and look for easy ways out.  */
-  if (len < kwset->mind)
-    return -1;
-  trans = kwset->trans;
-  trie = kwset->trie;
-  lim = text + len;
-  tp = text;
-
-  if (!trie->accepting)
-    {
-      unsigned char c;
-      int gc1 = kwset->gc1;
-
-      while (true)
-        {
-          if (gc1 < 0)
-            {
-              while (! (trie = kwset->next[c = tr (trans, *tp++)]))
-                if (tp >= lim)
-                  return -1;
-            }
-          else
-            {
-              tp = memchr_kwset (tp, lim - tp, kwset);
-              if (!tp)
-                return -1;
-              c = tr (trans, *tp++);
-              trie = kwset->next[c];
-            }
-
-          while (true)
-            {
-              if (trie->accepting)
-                goto match;
-              if (tp >= lim)
-                return -1;
-              c = tr (trans, *tp++);
-
-              for (tree = trie->links; c != tree->label; )
-                {
-                  tree = c < tree->label ? tree->llink : tree->rlink;
-                  if (! tree)
-                    {
-                      trie = trie->fail;
-                      if (!trie)
-                        {
-                          trie = kwset->next[c];
-                          if (trie)
-                            goto have_trie;
-                          if (tp >= lim)
-                            return -1;
-                          goto next_c;
-                        }
-                      if (trie->accepting)
-                        {
-                          --tp;
-                          goto match;
-                        }
-                      tree = trie->links;
-                    }
-                }
-              trie = tree->trie;
-            have_trie:;
-            }
-        next_c:;
-        }
-    }
-
- match:
-  accept = trie;
-  while (accept->accepting < 0)
-    accept = accept->fail;
-  left = tp - accept->depth;
-
-  /* Try left-most longest match.  */
-  if (longest)
-    {
-      while (tp < lim)
-        {
-          struct trie const *accept1;
-          char const *left1;
-          unsigned char c = tr (trans, *tp++);
-
-          do
-            {
-              tree = trie->links;
-              while (tree && c != tree->label)
-                tree = c < tree->label ? tree->llink : tree->rlink;
-            }
-          while (!tree && (trie = trie->fail) && accept->depth <= trie->depth);
-
-          if (!tree)
-            break;
-          trie = tree->trie;
-          if (trie->accepting)
-            {
-              accept1 = trie;
-              while (accept1->accepting < 0)
-                accept1 = accept1->fail;
-              left1 = tp - accept1->depth;
-              if (left1 <= left)
-                {
-                  left = left1;
-                  accept = accept1;
-                }
-            }
-        }
-    }
-
-  kwsmatch->index = accept->accepting >> 1;
-  kwsmatch->offset = left - text;
-  kwsmatch->size = accept->depth;
-
-  return left - text;
-}
-
-/* Hairy multiple string search with Aho-Corasick algorithm.  */
-static ptrdiff_t
-acexec (kwset_t kwset, char const *text, idx_t size,
-        struct kwsmatch *kwsmatch, bool longest)
-{
-  assume (0 <= size);
-  /* Help the compiler inline in two ways, depending on whether
-     kwset->trans is null.  */
-  return (IGNORE_DUPLICATE_BRANCH_WARNING
-          (kwset->trans
-           ? acexec_trans (kwset, text, size, kwsmatch, longest)
-           : acexec_trans (kwset, text, size, kwsmatch, longest)));
-}
-
-/* Find the first instance of a KWSET member in TEXT, which has SIZE bytes.
-   Return the offset (into TEXT) of the first byte of the matching substring,
-   or -1 if no match is found.  Upon a match, store details in
-   *KWSMATCH: index of matched keyword, start offset (same as the return
-   value), and length.  If LONGEST, find the longest match; otherwise
-   any match will do.  */
-ptrdiff_t
-kwsexec (kwset_t kwset, char const *text, idx_t size,
-         struct kwsmatch *kwsmatch, bool longest)
-{
-  return kwset->kwsexec (kwset, text, size, kwsmatch, longest);
-}
-
-/* Free the components of the given keyword set.  */
-void
-kwsfree (kwset_t kwset)
-{
-  obstack_free (&kwset->obstack, NULL);
-  free (kwset);
-}

src/kwset.h

@@ -1,46 +0,0 @@
-/* kwset.h - header declaring the keyword set library.
-   Copyright (C) 1989, 1998, 2005, 2007, 2009-2023 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, 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.  */
-
-/* Written August 1989 by Mike Haertel.  */
-
-#include <stddef.h>
-
-#include <idx.h>
-
-struct kwsmatch
-{
-  idx_t index;	/* Index number of matching keyword.  */
-  idx_t offset;	/* Offset of match.  */
-  idx_t size;	/* Length of match.  */
-};
-
-#include <arg-nonnull.h>
-#include <idx.h>
-
-struct kwset;
-typedef struct kwset *kwset_t;
-
-extern kwset_t kwsalloc (char const *);
-extern void kwsincr (kwset_t, char const *, idx_t);
-extern idx_t kwswords (kwset_t) _GL_ATTRIBUTE_PURE;
-extern void kwsprep (kwset_t);
-extern ptrdiff_t kwsexec (kwset_t, char const *, idx_t,
-                          struct kwsmatch *, bool)
-  _GL_ARG_NONNULL ((4));
-extern void kwsfree (kwset_t);

src/pcresearch.c

@@ -1,5 +1,5 @@
 /* pcresearch.c - searching subroutines using PCRE for grep.
-   Copyright 2000, 2007, 2009-2023 Free Software Foundation, Inc.
+   Copyright 2000, 2007, 2009-2026 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
@@ -12,13 +12,11 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 #include <config.h>
 
-#include "search.h"
+#include <search.h>
 #include "die.h"
 
 #include <stdckdint.h>
@@ -35,6 +33,18 @@
 # define PCRE2_ERROR_DEPTHLIMIT PCRE2_ERROR_RECURSIONLIMIT
 # define pcre2_set_depth_limit pcre2_set_recursion_limit
 #endif
+#ifndef PCRE2_EXTRA_ASCII_BSD
+# define PCRE2_EXTRA_ASCII_BSD 0
+#endif
+
+/* Use PCRE2_MATCH_INVALID_UTF if supported and not buggy;
+   see <https://github.com/PCRE2Project/pcre2/issues/224>.
+   Assume the bug will be fixed after PCRE2 10.42.  */
+#if defined PCRE2_MATCH_INVALID_UTF && 10 < PCRE2_MAJOR + (42 < PCRE2_MINOR)
+enum { MATCH_INVALID_UTF = PCRE2_MATCH_INVALID_UTF };
+#else
+enum { MATCH_INVALID_UTF = 0 };
+#endif
 
 struct pcre_comp
 {
@@ -71,6 +81,15 @@ private_free (void *ptr, _GL_UNUSED void *unused)
   free (ptr);
 }
 
+void
+Pprint_version (void)
+{
+  char *buf = ximalloc (pcre2_config (PCRE2_CONFIG_VERSION, nullptr));
+  pcre2_config (PCRE2_CONFIG_VERSION, buf);
+  printf (_("\ngrep -P uses PCRE2 %s\n"), buf);
+  free (buf);
+}
+
 /* Match the already-compiled PCRE pattern against the data in SUBJECT,
    of size SEARCH_BYTES and starting with offset SEARCH_OFFSET, with
    options OPTIONS.
@@ -101,7 +120,7 @@ jit_exec (struct pcre_comp *pc, char const *subject, idx_t search_bytes,
             xalloc_die ();
           if (!pc->mcontext)
             pc->mcontext = pcre2_match_context_create (pc->gcontext);
-          pcre2_jit_stack_assign (pc->mcontext, NULL, pc->jit_stack);
+          pcre2_jit_stack_assign (pc->mcontext, nullptr, pc->jit_stack);
         }
       else if (e == PCRE2_ERROR_DEPTHLIMIT)
         {
@@ -118,16 +137,11 @@ jit_exec (struct pcre_comp *pc, char const *subject, idx_t search_bytes,
     }
 }
 
-/* Return true if E is an error code for bad UTF-8, and if pcre2_match
-   could return E because PCRE lacks PCRE2_MATCH_INVALID_UTF.  */
+/* Return true if E is an error code for bad UTF-8.  */
 static bool
 bad_utf8_from_pcre2 (int e)
 {
-#ifdef PCRE2_MATCH_INVALID_UTF
-  return false;
-#else
   return PCRE2_ERROR_UTF8_ERR21 <= e && e <= PCRE2_ERROR_UTF8_ERR1;
-#endif
 }
 
 /* Compile the -P style PATTERN, containing SIZE bytes that are
@@ -142,7 +156,7 @@ Pcompile (char *pattern, idx_t size, reg_syntax_t ignored, bool exact)
   char *patlim = pattern + size;
   struct pcre_comp *pc = ximalloc (sizeof *pc);
   pcre2_general_context *gcontext = pc->gcontext
-    = pcre2_general_context_create (private_malloc, private_free, NULL);
+    = pcre2_general_context_create (private_malloc, private_free, nullptr);
   pcre2_compile_context *ccontext = pcre2_compile_context_create (gcontext);
 
   if (localeinfo.multibyte)
@@ -153,14 +167,23 @@ Pcompile (char *pattern, idx_t size, reg_syntax_t ignored, bool exact)
              _("-P supports only unibyte locales on this platform"));
       if (! localeinfo.using_utf8)
         die (EXIT_TROUBLE, 0, _("-P supports only unibyte and UTF-8 locales"));
-      flags |= (PCRE2_UTF | PCRE2_UCP);
+
+      flags |= PCRE2_UTF;
+
+      /* If supported, consider invalid UTF-8 as a barrier not an error.  */
+      flags |= MATCH_INVALID_UTF;
+
+      /* If PCRE2_EXTRA_ASCII_BSD is available, use PCRE2_UCP
+         so that \d does not have the undesirable effect of matching
+         non-ASCII digits.  Otherwise (i.e., with PCRE2 10.42 and earlier),
+         escapes like \w have only their ASCII interpretations,
+         but that's better than the confusion that would ensue if \d
+         matched non-ASCII digits.  */
+      flags |= PCRE2_EXTRA_ASCII_BSD ? PCRE2_UCP : 0;
+
 #if 0
       /* Do not match individual code units but only UTF-8.  */
       flags |= PCRE2_NEVER_BACKSLASH_C;
-#endif
-#ifdef PCRE2_MATCH_INVALID_UTF
-      /* Consider invalid UTF-8 as a barrier, instead of error.  */
-      flags |= PCRE2_MATCH_INVALID_UTF;
 #endif
     }
 
@@ -168,19 +191,23 @@ Pcompile (char *pattern, idx_t size, reg_syntax_t ignored, bool exact)
   if (rawmemchr (pattern, '\n') != patlim)
     die (EXIT_TROUBLE, 0, _("the -P option only supports a single pattern"));
 
-  void *re_storage = NULL;
+#ifdef PCRE2_EXTRA_MATCH_LINE
+  uint32_t extra_options = (PCRE2_EXTRA_ASCII_BSD
+                            | (match_lines ? PCRE2_EXTRA_MATCH_LINE : 0));
+  pcre2_set_compile_extra_options (ccontext, extra_options);
+#endif
+
+  void *re_storage = nullptr;
   if (match_lines)
     {
-#ifdef PCRE2_EXTRA_MATCH_LINE
-      pcre2_set_compile_extra_options (ccontext, PCRE2_EXTRA_MATCH_LINE);
-#else
-      static char const /* These sizes omit trailing NUL.  */
-        xprefix[4] = "^(?:", xsuffix[2] = ")$";
-      idx_t re_size = size + sizeof xprefix + sizeof xsuffix;
+#ifndef PCRE2_EXTRA_MATCH_LINE
+      static char const *const xprefix = "^(?:";
+      static char const *const xsuffix = ")$";
+      idx_t re_size = size + strlen (xprefix) + strlen (xsuffix);
       char *re = re_storage = ximalloc (re_size);
-      char *rez = mempcpy (re, xprefix, sizeof xprefix);
+      char *rez = mempcpy (re, xprefix, strlen (xprefix));
       rez = mempcpy (rez, pattern, size);
-      memcpy (rez, xsuffix, sizeof xsuffix);
+      memcpy (rez, xsuffix, strlen (xsuffix));
       pattern = re;
       size = re_size;
 #endif
@@ -189,18 +216,20 @@ Pcompile (char *pattern, idx_t size, reg_syntax_t ignored, bool exact)
     {
       /* PCRE2_EXTRA_MATCH_WORD is incompatible with grep -w;
          do things the grep way.  */
-      static char const /* These sizes omit trailing NUL.  */
-        wprefix[10] = "(?<!\\w)(?:", wsuffix[7] = ")(?!\\w)";
-      idx_t re_size = size + sizeof wprefix + sizeof wsuffix;
+      static char const *const wprefix = "(?<!\\w)(?:";
+      static char const *const wsuffix = ")(?!\\w)";
+      idx_t re_size = size + strlen (wprefix) + strlen (wsuffix);
       char *re = re_storage = ximalloc (re_size);
-      char *rez = mempcpy (re, wprefix, sizeof wprefix);
+      char *rez = mempcpy (re, wprefix, strlen (wprefix));
       rez = mempcpy (rez, pattern, size);
-      memcpy (rez, wsuffix, sizeof wsuffix);
+      memcpy (rez, wsuffix, strlen (wsuffix));
       pattern = re;
       size = re_size;
     }
 
-  pcre2_set_character_tables (ccontext, pcre2_maketables (gcontext));
+  if (!localeinfo.multibyte)
+    pcre2_set_character_tables (ccontext, pcre2_maketables (gcontext));
+
   pc->cre = pcre2_compile ((PCRE2_SPTR) pattern, size, flags,
                            &ec, &e, ccontext);
   if (!pc->cre)
@@ -214,15 +243,15 @@ Pcompile (char *pattern, idx_t size, reg_syntax_t ignored, bool exact)
   free (re_storage);
   pcre2_compile_context_free (ccontext);
 
-  pc->mcontext = NULL;
+  pc->mcontext = nullptr;
   pc->data = pcre2_match_data_create_from_pattern (pc->cre, gcontext);
 
-  ec = pcre2_jit_compile (pc->cre, PCRE2_JIT_COMPLETE);
-  if (ec && ec != PCRE2_ERROR_JIT_BADOPTION && ec != PCRE2_ERROR_NOMEMORY)
-    die (EXIT_TROUBLE, 0, _("JIT internal error: %d"), ec);
+  /* Ignore any failure return from pcre2_jit_compile, as that merely
+     means JIT won't be used during matching.  */
+  pcre2_jit_compile (pc->cre, PCRE2_JIT_COMPLETE);
 
   /* The PCRE documentation says that a 32 KiB stack is the default.  */
-  pc->jit_stack = NULL;
+  pc->jit_stack = nullptr;
   pc->jit_stack_size = 32 << 10;
 
   pc->empty_match[false] = jit_exec (pc, "", 0, 0, PCRE2_NOTBOL);
@@ -287,7 +316,7 @@ Pexecute (void *vcp, char const *buf, idx_t size, idx_t *match_size,
 
           e = jit_exec (pc, subject, line_end - subject,
                         search_offset, options);
-          if (!bad_utf8_from_pcre2 (e))
+          if (MATCH_INVALID_UTF || !bad_utf8_from_pcre2 (e))
             break;
 
           idx_t valid_bytes = pcre2_get_startchar (pc->data);

src/search.h

@@ -1,5 +1,5 @@
 /* search.c - searching subroutines using dfa, kwset and regex for grep.
-   Copyright 1992, 1998, 2000, 2007, 2009-2023 Free Software Foundation, Inc.
+   Copyright 1992, 1998, 2000, 2007, 2009-2026 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
@@ -12,9 +12,7 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 #ifndef GREP_SEARCH_H
 #define GREP_SEARCH_H 1
@@ -24,7 +22,6 @@
 #include <sys/types.h>
 #include <stdint.h>
 #include <wchar.h>
-#include <wctype.h>
 #include <regex.h>
 
 #include "system.h"
@@ -65,6 +62,7 @@ extern ptrdiff_t Fexecute (void *, char const *, idx_t, idx_t *, char const *);
 /* pcresearch.c */
 extern void *Pcompile (char *, idx_t, reg_syntax_t, bool);
 extern ptrdiff_t Pexecute (void *, char const *, idx_t, idx_t *, char const *);
+extern void Pprint_version (void);
 
 /* grep.c */
 extern struct localeinfo localeinfo;

src/searchutils.c

@@ -1,5 +1,5 @@
 /* searchutils.c - helper subroutines for grep's matchers.
-   Copyright 1992, 1998, 2000, 2007, 2009-2023 Free Software Foundation, Inc.
+   Copyright 1992, 1998, 2000, 2007, 2009-2026 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
@@ -12,15 +12,15 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 #include <config.h>
 
 #define SEARCH_INLINE _GL_EXTERN_INLINE
 #define SYSTEM_INLINE _GL_EXTERN_INLINE
-#include "search.h"
+#include <search.h>
+
+#include <uchar.h>
 
 /* For each byte B, sbwordchar[B] is true if B is a single-byte
    character that is a word constituent, and is false otherwise.  */
@@ -30,7 +30,7 @@ static bool sbwordchar[NCHAR];
 static bool
 wordchar (wint_t wc)
 {
-  return wc == L'_' || iswalnum (wc);
+  return wc == L'_' || c32isalnum (wc);
 }
 
 void
@@ -43,7 +43,7 @@ wordinit (void)
 kwset_t
 kwsinit (bool mb_trans)
 {
-  char *trans = NULL;
+  char *trans = nullptr;
 
   if (match_icase && (MB_CUR_MAX == 1 || mb_trans))
     {
@@ -113,7 +113,7 @@ mb_goback (char const **mb_start, idx_t *mbclen, char const *cur,
 
               if (long_enough)
                 {
-                  mbstate_t mbs = { 0 };
+                  mbstate_t mbs; mbszero (&mbs);
                   ptrdiff_t clen = imbrlen (cur - i, end - (cur - i), &mbs);
                   if (0 <= clen)
                     {
@@ -129,7 +129,7 @@ mb_goback (char const **mb_start, idx_t *mbclen, char const *cur,
     {
       /* In non-UTF-8 encodings, to find character boundaries one must
          in general scan forward from the start of the buffer.  */
-      mbstate_t mbs = { 0 };
+      mbstate_t mbs; mbszero (&mbs);
       ptrdiff_t clen;
 
       do
@@ -162,27 +162,27 @@ mb_goback (char const **mb_start, idx_t *mbclen, char const *cur,
 static idx_t
 wordchars_count (char const *buf, char const *end, bool countall)
 {
-  idx_t n = 0;
-  mbstate_t mbs = { 0 };
-  while (n < end - buf)
+  mbstate_t mbs; mbszero (&mbs);
+  char const *p = buf;
+  while (p < end)
     {
-      unsigned char b = buf[n];
+      unsigned char b = *p;
       if (sbwordchar[b])
-        n++;
+        p++;
       else if (localeinfo.sbclen[b] != -2)
         break;
       else
         {
-          wchar_t wc = 0;
-          size_t wcbytes = mbrtowc (&wc, buf + n, end - buf - n, &mbs);
+          char32_t wc = 0;
+          size_t wcbytes = mbrtoc32 (&wc, p, end - p, &mbs);
           if (!wordchar (wc))
             break;
-          n += wcbytes + !wcbytes;
+          p += wcbytes + !wcbytes;
         }
       if (!countall)
         break;
     }
-  return n;
+  return p - buf;
 }
 
 /* Examine the start of BUF for the longest prefix containing just
@@ -214,6 +214,6 @@ wordchar_prev (char const *buf, char const *cur, char const *end)
   if (! localeinfo.multibyte || localeinfo.using_utf8 & ~(b >> 7))
     return sbwordchar[b];
   char const *p = buf;
-  cur -= mb_goback (&p, NULL, cur, end);
+  cur -= mb_goback (&p, nullptr, cur, end);
   return wordchar_next (cur, end);
 }

src/system.h

@@ -1,5 +1,5 @@
 /* Portability cruft.  Include after config.h and sys/types.h.
-   Copyright 1996, 1998-2000, 2007, 2009-2023 Free Software Foundation, Inc.
+   Copyright 1996, 1998-2000, 2007, 2009-2026 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
@@ -12,9 +12,7 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 #ifndef GREP_SYSTEM_H
 #define GREP_SYSTEM_H 1

tests/100k-entries

@@ -0,0 +1,15 @@
+#!/bin/sh
+# This would make grep-3.11 fail with ENOTSUP and exit 2.
+. "${srcdir=.}/init.sh"; path_prepend_ ../src
+expensive_
+
+fail=0
+
+mkdir t || framework_failure_
+(cd t && seq 100000|xargs touch) || framework_failure_
+
+returns_ 1 grep -r x t > out 2> err
+compare /dev/null out || fail=1
+compare /dev/null err || fail=1
+
+Exit $fail

tests/Coreutils.pm

@@ -1,7 +1,7 @@
 package Coreutils;
 # This is a testing framework.
 
-# Copyright (C) 1998-2015, 2017-2023 Free Software Foundation, Inc.
+# Copyright (C) 1998-2015, 2017-2026 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

tests/CuSkip.pm

@@ -1,7 +1,7 @@
 package CuSkip;
 # Skip a test: emit diag to log and to stderr, and exit 77
 
-# Copyright (C) 2011-2015, 2017-2023 Free Software Foundation, Inc.
+# Copyright (C) 2011-2015, 2017-2026 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

tests/CuTmpdir.pm

@@ -1,7 +1,7 @@
 package CuTmpdir;
 # create, then chdir into a temporary sub-directory
 
-# Copyright (C) 2007-2015, 2017-2023 Free Software Foundation, Inc.
+# Copyright (C) 2007-2015, 2017-2026 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

tests/Makefile.am

@@ -1,5 +1,5 @@
 ## Process this file with automake to create Makefile.in
-# Copyright 1997-1998, 2005-2023 Free Software Foundation, Inc.
+# Copyright 1997-1998, 2005-2026 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
@@ -41,18 +41,27 @@ AM_CFLAGS = $(WARN_CFLAGS) $(WERROR_CFLAGS)
 
 # Tell the linker to omit references to unused shared libraries.
 AM_LDFLAGS = $(IGNORE_UNUSED_LIBRARIES_CFLAGS)
-LDADD = ../lib/libgreputils.a $(LIBINTL) ../lib/libgreputils.a
+LDADD = ../lib/libgreputils.a $(LIBINTL) ../lib/libgreputils.a \
+  $(HARD_LOCALE_LIB) $(LIBC32CONV) $(LIBCSTACK) \
+  $(LIBSIGSEGV) $(LIBUNISTRING) $(MBRTOWC_LIB) $(SETLOCALE_NULL_LIB) \
+  $(LIBTHREAD)
 
 # The triple-backref test is expected to fail with both the system
 # matcher (i.e., with glibc) and with the included matcher.
 # Both matchers need to be fixed.
-# FIXME-2015: Remove this once the glibc and gnulib bugs are fixed.
+# FIXME-2025: Remove this once the glibc and gnulib bugs are fixed.
 XFAIL_TESTS = triple-backref
 
+# The glibc-infloop test is expected to fail with both the system
+# matcher (i.e., with glibc) and with the included matcher.
+# Both matchers need to be fixed.
+# FIXME-2025: Remove this once the glibc and gnulib bugs are fixed.
+XFAIL_TESTS += glibc-infloop
+
 # Equivalence classes are only supported when using the system
 # matcher (which means only with glibc).
 # The included matcher needs to be fixed.
-# FIXME-2015: Remove this once the gnulib bug is fixed.
+# FIXME-2025: Remove this once the gnulib bug is fixed.
 if USE_INCLUDED_REGEX
 XFAIL_TESTS += equiv-classes
 else
@@ -62,6 +71,7 @@ else
 endif
 
 TESTS =						\
+  100k-entries					\
   backref					\
   backref-alt					\
   backref-multibyte-slow			\
@@ -108,6 +118,7 @@ TESTS =						\
   fillbuf-long-line				\
   fmbtest					\
   foad1						\
+  glibc-infloop					\
   grep-dev-null					\
   grep-dev-null-out				\
   grep-dir					\
@@ -139,6 +150,7 @@ TESTS =						\
   options					\
   pcre						\
   pcre-abort					\
+  pcre-ascii-digits				\
   pcre-context					\
   pcre-count					\
   pcre-infloop					\
@@ -147,6 +159,7 @@ TESTS =						\
   pcre-jitstack					\
   pcre-o					\
   pcre-utf8					\
+  pcre-utf8-bug224				\
   pcre-utf8-w					\
   pcre-w					\
   pcre-wx-backref				\
@@ -177,11 +190,13 @@ TESTS =						\
   unibyte-bracket-expr				\
   unibyte-negated-circumflex			\
   utf8-bracket					\
+  version-pcre					\
   warn-char-classes				\
   word-delim-multibyte				\
   word-multi-file				\
   word-multibyte				\
   write-error-msg				\
+  y2038-vs-32-bit				\
   yesno						\
   z-anchor-newline
 
@@ -252,6 +267,7 @@ TESTS_ENVIRONMENT =				\
   srcdir='$(srcdir)'				\
   top_srcdir='$(top_srcdir)'			\
   CC='$(CC)'					\
+  CONFIG_HEADER='$(abs_top_builddir)/$(CONFIG_INCLUDE)' \
   GREP_TEST_NAME=`echo $$tst|sed 's,^\./,,;s,/,-,g'` \
   MAKE=$(MAKE)					\
   MALLOC_PERTURB_=$(MALLOC_PERTURB_)		\

tests/backref

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test for back-references and other things.
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/backref-alt

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test for a bug in glibc's regex code as of 2015-09-19.
 #
-# Copyright 2015-2023 Free Software Foundation, Inc.
+# Copyright 2015-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/backslash-dot

@@ -1,7 +1,7 @@
 #! /bin/sh
 # This once failed to match: echo . | grep '\.'
 #
-# Copyright (C) 2020-2023 Free Software Foundation, Inc.
+# Copyright (C) 2020-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/backslash-s-and-repetition-operators

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Ensure that \s and \S work with repetition operators.
 #
-# Copyright (C) 2013-2023 Free Software Foundation, Inc.
+# Copyright (C) 2013-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/backslash-s-vs-invalid-multibyte

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Ensure that neither \s nor \S matches an invalid multibyte character.
 #
-# Copyright (C) 2013-2023 Free Software Foundation, Inc.
+# Copyright (C) 2013-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/binary-file-matches

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test for the "binary file ... matches" diagnostic.
 #
-# Copyright (C) 2020-2023 Free Software Foundation, Inc.
+# Copyright (C) 2020-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/bre

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Regression test for GNU grep.
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/bre.awk

@@ -1,4 +1,4 @@
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/c-locale

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Regression test for GNU grep.
 #
-# Copyright 2016-2023 Free Software Foundation, Inc.
+# Copyright 2016-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/case-fold-titlecase

@@ -1,7 +1,7 @@
 #!/bin/sh
 # Check that case folding works even with titlecase and similarly odd chars.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/color-colors

@@ -1,7 +1,7 @@
 #!/bin/sh
 # Check that GREP_COLOR elicits a warning.
 
-# Copyright 2022-2023 Free Software Foundation, Inc.
+# Copyright 2022-2026 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

tests/count-newline

@@ -2,7 +2,7 @@
 # Test that newline is counted correctly even when the transition
 # table is rebuilt.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/dfa-coverage

@@ -1,7 +1,7 @@
 #!/bin/sh
 # Exercise the final reachable code in dfa.c's match_mb_charset.
 
-# Copyright (C) 2012-2023 Free Software Foundation, Inc.
+# Copyright (C) 2012-2026 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

tests/dfa-heap-overrun

@@ -1,7 +1,7 @@
 #!/bin/sh
 # Trigger a heap overrun in grep-2.6..grep-2.8.
 
-# Copyright (C) 2011-2023 Free Software Foundation, Inc.
+# Copyright (C) 2011-2026 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

tests/dfa-invalid-utf8

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test whether "grep '.'" matches invalid UTF-8 byte sequences.
 #
-# Copyright 2019-2023 Free Software Foundation, Inc.
+# Copyright 2019-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/empty

@@ -2,7 +2,7 @@
 # test that the empty file means no pattern
 # and an empty pattern means match all.
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/empty-line-mb

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Exercise bugs in grep-2.13 with -i, -n and an RE of ^$ in a multi-byte locale.
 #
-# Copyright (C) 2012-2023 Free Software Foundation, Inc.
+# Copyright (C) 2012-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/encoding-error

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test grep's behavior on encoding errors.
 #
-# Copyright 2015-2023 Free Software Foundation, Inc.
+# Copyright 2015-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/envvar-check

@@ -1,7 +1,7 @@
 # -*- sh -*-
 # Check environment variables for sane values while testing.
 
-# Copyright (C) 2000-2023 Free Software Foundation, Inc.
+# Copyright (C) 2000-2026 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

tests/ere

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Regression test for GNU grep.
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/ere.awk

@@ -1,4 +1,4 @@
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/false-match-mb-non-utf8

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test for false matches in grep 2.19..2.26 in multibyte, non-UTF8 locales
 #
-# Copyright (C) 2016-2023 Free Software Foundation, Inc.
+# Copyright (C) 2016-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/fgrep-longest

@@ -2,7 +2,7 @@
 # With multiple matches, grep -Fo could print a shorter one.
 # This bug affected grep versions 2.26 through 2.27.
 #
-# Copyright (C) 2017-2023 Free Software Foundation, Inc.
+# Copyright (C) 2017-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/file

@@ -4,7 +4,7 @@
 # grep -F -f pattern_file file
 # grep -G -f pattern_file file
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/filename-lineno.pl

@@ -4,7 +4,7 @@
 # file or line number from which the offending regular expression came.
 # With 2.26, now, each such diagnostic has a "FILENAME:LINENO: " prefix.
 
-# Copyright (C) 2016-2023 Free Software Foundation, Inc.
+# Copyright (C) 2016-2026 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

tests/fmbtest

@@ -1,5 +1,5 @@
 #! /bin/sh
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright
@@ -10,7 +10,7 @@
 cz=cs_CZ.UTF-8
 
 # If cs_CZ.UTF-8 locale doesn't work, skip this test.
-LC_ALL=$cz locale -k LC_CTYPE 2>/dev/null | grep -q charmap.*UTF-8 \
+test "`LC_ALL=$cz locale charmap 2>/dev/null`" = UTF-8 \
   || skip_ this system lacks the $cz locale
 
 # If matching is done in single-byte mode, skip this test too

tests/foad1

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test various combinations of command-line options.
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright
@@ -150,7 +150,7 @@ Exit $failures
 # The rest of this file is meant to be executed under this locale.
 LC_ALL=cs_CZ.UTF-8; export LC_ALL
 # If the UTF-8 locale doesn't work, skip these tests silently.
-locale -k LC_CTYPE 2>/dev/null | grep -q "charmap.*UTF-8" || Exit $failures
+test "`locale charmap 2>/dev/null`" = UTF-8 || Exit $failures
 
 # Test character class erroneously matching a '[' character.
 grep_test "[/" "" "[[:alpha:]]" -E

tests/get-mb-cur-max.c

@@ -1,5 +1,5 @@
 /* Auxiliary program to detect support for a locale.
-   Copyright 2010-2023 Free Software Foundation, Inc.
+   Copyright 2010-2026 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
@@ -12,9 +12,7 @@
    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.  */
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
 
 #include <config.h>
 #include <locale.h>

tests/glibc-infloop

@@ -0,0 +1,30 @@
+#!/bin/sh
+# This would infloop when using glibc's regex at least until glibc-2.36.
+. "${srcdir=.}/init.sh"; path_prepend_ ../src
+
+require_timeout_
+require_en_utf8_locale_
+
+fail=0
+
+cat <<\EOF > glibc-check.c
+#include <features.h>
+#ifdef __GLIBC__
+int ok;
+#else
+# error "not glibc"
+#endif
+EOF
+$CC -c glibc-check.c && glibc=1 || glibc=0
+
+grep '^#define USE_INCLUDED_REGEX 1' "$CONFIG_HEADER" \
+  && included_regex=1 || included_regex=0
+
+case $glibc:$included_regex in
+  0:0) skip_ 'runs only with glibc or when built with the included regex'
+esac
+
+echo a > in || framework_failure_
+timeout 2 env LC_ALL=en_US.UTF-8 grep -E -w '((()|a)|())*' in || fail=1
+
+Exit $fail

tests/hash-collision-perf

@@ -2,7 +2,7 @@
 # Test for this performance regression:
 # grep-3.5 and 3.6 would take O(N^2) time for some sets of input regexps.
 
-# Copyright 2020-2023 Free Software Foundation, Inc.
+# Copyright 2020-2026 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
@@ -34,6 +34,9 @@ while :; do
   small_ms=$(LC_ALL=C user_time_ 1 grep --file=in empty) || fail=1
   test $small_ms -ge 200 && break
   n_pat=$(expr $n_pat '*' 2)
+  case $n_pat:$small_ms in
+    640000:0) skip_ 'user_time_ appears always to report 0 elapsed ms';;
+  esac
 done
 
 # Now, search for those same digits mapped to A-J.

tests/help-version

@@ -2,7 +2,7 @@
 # Make sure all of these programs work properly
 # when invoked with --help or --version.
 
-# Copyright (C) 2000-2023 Free Software Foundation, Inc.
+# Copyright (C) 2000-2026 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

tests/high-bit-range

@@ -1,7 +1,7 @@
 #!/bin/sh
 # Exercise high-bit-set unibyte-in-[...]-range bug.
 
-# Copyright (C) 2011-2023 Free Software Foundation, Inc.
+# Copyright (C) 2011-2026 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

tests/init.sh

@@ -1,706 +0,0 @@
-# source this file; set up for tests
-
-# Copyright (C) 2009-2023 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 <https://www.gnu.org/licenses/>.
-
-# Using this file in a test
-# =========================
-#
-# The typical skeleton of a test looks like this:
-#
-#   #!/bin/sh
-#   . "${srcdir=.}/init.sh"; path_prepend_ .
-#   Execute some commands.
-#   Note that these commands are executed in a subdirectory, therefore you
-#   need to prepend "../" to relative filenames in the build directory.
-#   Note that the "path_prepend_ ." is useful only if the body of your
-#   test invokes programs residing in the initial directory.
-#   For example, if the programs you want to test are in src/, and this test
-#   script is named tests/test-1, then you would use "path_prepend_ ../src",
-#   or perhaps export PATH='$(abs_top_builddir)/src$(PATH_SEPARATOR)'"$$PATH"
-#   to all tests via automake's TESTS_ENVIRONMENT.
-#   Set the exit code 0 for success, 77 for skipped, or 1 or other for failure.
-#   Use the skip_ and fail_ functions to print a diagnostic and then exit
-#   with the corresponding exit code.
-#   Exit $?
-
-# Executing a test that uses this file
-# ====================================
-#
-# Running a single test:
-#   $ make check TESTS=test-foo.sh
-#
-# Running a single test, with verbose output:
-#   $ make check TESTS=test-foo.sh VERBOSE=yes
-#
-# Running a single test, keeping the temporary directory:
-#   $ make check TESTS=test-foo.sh KEEP=yes
-#
-# Running a single test, with single-stepping:
-#   1. Go into a sub-shell:
-#   $ bash
-#   2. Set relevant environment variables from TESTS_ENVIRONMENT in the
-#      Makefile:
-#   $ export srcdir=../../tests # this is an example
-#   3. Execute the commands from the test, copy&pasting them one by one:
-#   $ . "$srcdir/init.sh"; path_prepend_ .
-#   ...
-#   4. Finally
-#   $ exit
-
-# =============================================================================
-# Elementary diagnostics
-
-ME_=`expr "./$0" : '.*/\(.*\)$'`
-
-# Prepare PATH_SEPARATOR.
-# The user is always right.
-if test "${PATH_SEPARATOR+set}" != set; then
-  # Determine PATH_SEPARATOR by trying to find /bin/sh in a PATH which
-  # contains only /bin. Note that ksh looks also at the FPATH variable,
-  # so we have to set that as well for the test.
-  PATH_SEPARATOR=:
-  (PATH='/bin;/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 \
-    && { (PATH='/bin:/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 \
-           || PATH_SEPARATOR=';'
-       }
-fi
-
-# We use a trap below for cleanup.  This requires us to go through
-# hoops to get the right exit status transported through the handler.
-# So use 'Exit STATUS' instead of 'exit STATUS' inside of the tests.
-# Turn off errexit here so that we don't trip the bug with OSF1/Tru64
-# sh inside this function.
-Exit () { set +e; (exit $1); exit $1; }
-
-# Print warnings (e.g., about skipped and failed tests) to this file number.
-# Override by defining to say, 9, in init.cfg, and putting say,
-#   export ...ENVVAR_SETTINGS...; $(SHELL) 9>&2
-# in the definition of TESTS_ENVIRONMENT in your tests/Makefile.am file.
-# This is useful when using automake's parallel tests mode, to print
-# the reason for skip/failure to console, rather than to the .log files.
-: ${stderr_fileno_=2}
-
-# Note that correct expansion of "$*" depends on IFS starting with ' '.
-# Always write the full diagnostic to stderr.
-# When stderr_fileno_ is not 2, also emit the first line of the
-# diagnostic to that file descriptor.
-warn_ ()
-{
-  # If IFS does not start with ' ', set it and emit the warning in a subshell.
-  case $IFS in
-    ' '*) printf '%s\n' "$*" >&2
-          test $stderr_fileno_ = 2 \
-            || { printf '%s\n' "$*" | sed 1q >&$stderr_fileno_ ; } ;;
-    *) (IFS=' '; warn_ "$@");;
-  esac
-}
-fail_ () { warn_ "$ME_: failed test: $@"; Exit 1; }
-skip_ () { warn_ "$ME_: skipped test: $@"; Exit 77; }
-fatal_ () { warn_ "$ME_: hard error: $@"; Exit 99; }
-framework_failure_ () { warn_ "$ME_: set-up failure: $@"; Exit 99; }
-
-# =============================================================================
-# Ensure the shell supports modern syntax.
-
-# Sanitize this shell to POSIX mode, if possible.
-DUALCASE=1; export DUALCASE
-if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then
-  emulate sh
-  NULLCMD=:
-  alias -g '${1+"$@"}'='"$@"'
-  setopt NO_GLOB_SUBST
-else
-  case `(set -o) 2>/dev/null` in
-    *posix*) set -o posix ;;
-  esac
-fi
-
-# We require $(...) support unconditionally.
-# We require that the printf built-in work correctly regarding octal escapes;
-# this eliminates /bin/sh on AIX 7.2.
-# We require non-surprising "local" semantics (this eliminates dash).
-# This takes the admittedly draconian step of eliminating dash, because the
-# assignment tab=$(printf '\t') works fine, yet preceding it with "local "
-# transforms it into an assignment that sets the variable to the empty string.
-# That is too counter-intuitive, and can lead to subtle run-time malfunction.
-# The example below is less subtle in that with dash, it evokes the run-time
-# exception "dash: 1: local: 1: bad variable name".
-# We require a few additional shell features only when $EXEEXT is nonempty,
-# in order to support automatic $EXEEXT emulation:
-# - hyphen-containing alias names
-# - we prefer to use ${var#...} substitution, rather than having
-#   to work around lack of support for that feature.
-# The following code attempts to find a shell with support for these features.
-# If the current shell passes the test, we're done.  Otherwise, test other
-# shells until we find one that passes.  If one is found, re-exec it.
-# If no acceptable shell is found, skip the current test.
-#
-# The "...set -x; P=1 true 2>err..." test is to disqualify any shell that
-# emits "P=1" into err, as /bin/sh from SunOS 5.11 and OpenBSD 4.7 do.
-#
-# Use "9" to indicate success (rather than 0), in case some shell acts
-# like Solaris 10's /bin/sh but exits successfully instead of with status 2.
-
-# Eval this code in a subshell to determine a shell's suitability.
-# 10 - passes all tests; ok to use
-#  9 - ok, but enabling "set -x" corrupts app stderr; prefer higher score
-#  ? - not ok
-gl_shell_test_script_='
-test $(echo y) = y || exit 1
-LC_ALL=en_US.UTF-8 printf "\\351" 2>/dev/null \
-  | LC_ALL=C tr "\\351" x | LC_ALL=C grep "^x$" > /dev/null \
-  || exit 1
-printf "\\351" 2>/dev/null \
-  | LC_ALL=C tr "\\351" x | LC_ALL=C grep "^x$" > /dev/null \
-  || exit 1
-f_local_() { local v=1; }; f_local_ || exit 1
-f_dash_local_fail_() { local t=$(printf " 1"); }; f_dash_local_fail_
-score_=10
-if test "$VERBOSE" = yes; then
-  test -n "$( (exec 3>&1; set -x; P=1 true 2>&3) 2> /dev/null)" && score_=9
-fi
-test -z "$EXEEXT" && exit $score_
-shopt -s expand_aliases
-alias a-b="echo zoo"
-v=abx
-     test ${v%x} = ab \
-  && test ${v#a} = bx \
-  && test $(a-b) = zoo \
-  && exit $score_
-'
-
-if test "x$1" = "x--no-reexec"; then
-  shift
-else
-  # Assume a working shell.  Export to subshells (setup_ needs this).
-  gl_set_x_corrupts_stderr_=false
-  export gl_set_x_corrupts_stderr_
-
-  # Record the first marginally acceptable shell.
-  marginal_=
-
-  # Search for a shell that meets our requirements.
-  for re_shell_ in __current__ "${CONFIG_SHELL:-no_shell}" \
-      /bin/sh bash dash zsh pdksh fail
-  do
-    test "$re_shell_" = no_shell && continue
-
-    # If we've made it all the way to the sentinel, "fail" without
-    # finding even a marginal shell, skip this test.
-    if test "$re_shell_" = fail; then
-      test -z "$marginal_" && skip_ failed to find an adequate shell
-      re_shell_=$marginal_
-      break
-    fi
-
-    # When testing the current shell, simply "eval" the test code.
-    # Otherwise, run it via $re_shell_ -c ...
-    if test "$re_shell_" = __current__; then
-      # 'eval'ing this code makes Solaris 10's /bin/sh exit with
-      # $? set to 2.  It does not evaluate any of the code after the
-      # "unexpected" first '('.  Thus, we must run it in a subshell.
-      ( eval "$gl_shell_test_script_" ) > /dev/null 2>&1
-    else
-      "$re_shell_" -c "$gl_shell_test_script_" 2>/dev/null
-    fi
-
-    st_=$?
-
-    # $re_shell_ works just fine.  Use it.
-    if test $st_ = 10; then
-      gl_set_x_corrupts_stderr_=false
-      break
-    fi
-
-    # If this is our first marginally acceptable shell, remember it.
-    if test "$st_:$marginal_" = 9: ; then
-      marginal_="$re_shell_"
-      gl_set_x_corrupts_stderr_=true
-    fi
-  done
-
-  if test "$re_shell_" != __current__; then
-    # Found a usable shell.  Preserve -v and -x.
-    case $- in
-      *v*x* | *x*v*) opts_=-vx ;;
-      *v*) opts_=-v ;;
-      *x*) opts_=-x ;;
-      *) opts_= ;;
-    esac
-    re_shell=$re_shell_
-    export re_shell
-    exec "$re_shell_" $opts_ "$0" --no-reexec "$@"
-    echo "$ME_: exec failed" 1>&2
-    exit 127
-  fi
-fi
-
-# =============================================================================
-# Ensure the shell behaves reasonably.
-
-# If this is bash, turn off all aliases.
-test -n "$BASH_VERSION" && unalias -a
-
-# Note that when supporting $EXEEXT (transparently mapping from PROG_NAME to
-# PROG_NAME.exe), we want to support hyphen-containing names like test-acos.
-# That is part of the shell-selection test above.  Why use aliases rather
-# than functions?  Because support for hyphen-containing aliases is more
-# widespread than that for hyphen-containing function names.
-test -n "$EXEEXT" && test -n "$BASH_VERSION" && shopt -s expand_aliases
-
-# =============================================================================
-# Creating a temporary directory (needed by the core test framework)
-
-# Create a temporary directory, much like mktemp -d does.
-# Written by Jim Meyering.
-#
-# Usage: mktempd_ /tmp phoey.XXXXXXXXXX
-#
-# First, try to use the mktemp program.
-# Failing that, we'll roll our own mktemp-like function:
-#  - try to get random bytes from /dev/urandom, mapping them to file-name bytes
-#  - failing that, generate output from a combination of quickly-varying
-#      sources and awk.
-#  - try to create the desired directory.
-#  - make only $MAX_TRIES_ attempts
-
-# Helper function.  Print $N pseudo-random bytes from a-zA-Z0-9.
-rand_bytes_ ()
-{
-  n_=$1
-
-  # Maybe try openssl rand -base64 $n_prime_|tr '+/=\012' abcd first?
-  # But if they have openssl, they probably have mktemp, too.
-
-  chars_=abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789
-  dev_rand_=/dev/urandom
-  if test -r "$dev_rand_"; then
-    # Note: 256-length($chars_) == 194; 3 copies of $chars_ is 186 + 8 = 194.
-    dd ibs=$n_ count=1 if=$dev_rand_ 2>/dev/null \
-      | LC_ALL=C tr -c $chars_ 01234567$chars_$chars_$chars_
-    return
-  fi
-
-  # Fall back on quickly-varying sources + awk.
-  # Limit awk program to 7th Edition Unix so that it works even on Solaris 10.
-
-  (date; date +%N; free; who -a; w; ps auxww; ps -ef) 2>&1 | awk '
-     BEGIN {
-       n = '"$n_"'
-       for (i = 0; i < 256; i++)
-         ordinal[sprintf ("%c", i)] = i
-     }
-     {
-       for (i = 1; i <= length; i++)
-         a[ai++ % n] += ordinal[substr ($0, i, 1)]
-     }
-     END {
-       chars = "'"$chars_"'"
-       charslen = length (chars)
-       for (i = 0; i < n; i++)
-         printf "%s", substr (chars, a[i] % charslen + 1, 1)
-       printf "\n"
-     }
-  '
-}
-
-mktempd_ ()
-{
-  case $# in
-  2);;
-  *) fail_ "Usage: mktempd_ DIR TEMPLATE";;
-  esac
-
-  destdir_=$1
-  template_=$2
-
-  MAX_TRIES_=4
-
-  # Disallow any trailing slash on specified destdir:
-  # it would subvert the post-mktemp "case"-based destdir test.
-  case $destdir_ in
-  / | //) destdir_slash_=$destdir;;
-  */) fail_ "invalid destination dir: remove trailing slash(es)";;
-  *) destdir_slash_=$destdir_/;;
-  esac
-
-  case $template_ in
-  *XXXX) ;;
-  *) fail_ \
-       "invalid template: $template_ (must have a suffix of at least 4 X's)";;
-  esac
-
-  # First, try to use mktemp.
-  d=`unset TMPDIR; { mktemp -d -t -p "$destdir_" "$template_"; } 2>/dev/null` &&
-
-  # The resulting name must be in the specified directory.
-  case $d in "$destdir_slash_"*) :;; *) false;; esac &&
-
-  # It must have created the directory.
-  test -d "$d" &&
-
-  # It must have 0700 permissions.  Handle sticky "S" bits.
-  perms=`ls -dgo "$d" 2>/dev/null` &&
-  case $perms in drwx--[-S]---*) :;; *) false;; esac && {
-    echo "$d"
-    return
-  }
-
-  # If we reach this point, we'll have to create a directory manually.
-
-  # Get a copy of the template without its suffix of X's.
-  base_template_=`echo "$template_"|sed 's/XX*$//'`
-
-  # Calculate how many X's we've just removed.
-  template_length_=`echo "$template_" | wc -c`
-  nx_=`echo "$base_template_" | wc -c`
-  nx_=`expr $template_length_ - $nx_`
-
-  err_=
-  i_=1
-  while :; do
-    X_=`rand_bytes_ $nx_`
-    candidate_dir_="$destdir_slash_$base_template_$X_"
-    err_=`mkdir -m 0700 "$candidate_dir_" 2>&1` \
-      && { echo "$candidate_dir_"; return; }
-    test $MAX_TRIES_ -le $i_ && break;
-    i_=`expr $i_ + 1`
-  done
-  fail_ "$err_"
-}
-
-# =============================================================================
-# Core test framework
-
-# An arbitrary prefix to help distinguish test directories.
-testdir_prefix_ () { printf gt; }
-
-# Set up the environment for the test to run in.
-setup_ ()
-{
-  if test "$VERBOSE" = yes; then
-    # Test whether set -x may cause the selected shell to corrupt an
-    # application's stderr.  Many do, including zsh-4.3.10 and the /bin/sh
-    # from SunOS 5.11, OpenBSD 4.7 and Irix 6.5.
-    # If enabling verbose output this way would cause trouble, simply
-    # issue a warning and refrain.
-    if $gl_set_x_corrupts_stderr_; then
-      warn_ "using SHELL=$SHELL with 'set -x' corrupts stderr"
-    else
-      set -x
-    fi
-  fi
-
-  initial_cwd_=$PWD
-
-  # Create and enter the temporary directory.
-  pfx_=`testdir_prefix_`
-  test_dir_=`mktempd_ "$initial_cwd_" "$pfx_-$ME_.XXXX"` \
-    || fail_ "failed to create temporary directory in $initial_cwd_"
-  cd "$test_dir_" || fail_ "failed to cd to temporary directory"
-  # Set variables srcdir, builddir, for the convenience of the test.
-  case $srcdir in
-    /* | ?:*) ;;
-    *) srcdir="../$srcdir" ;;
-  esac
-  builddir=".."
-  export srcdir builddir
-
-  # As autoconf-generated configure scripts do, ensure that IFS
-  # is defined initially, so that saving and restoring $IFS works.
-  gl_init_sh_nl_='
-'
-  IFS=" ""	$gl_init_sh_nl_"
-
-  # This trap statement, along with a trap on 0 below, ensure that the
-  # temporary directory, $test_dir_, is removed upon exit as well as
-  # upon receipt of any of the listed signals.
-  for sig_ in 1 2 3 13 15; do
-    eval "trap 'Exit $(expr $sig_ + 128)' $sig_"
-  done
-
-  # Remove relative and non-accessible directories from PATH, including '.'
-  # and Zero-length entries.
-  saved_IFS="$IFS"
-  IFS=:
-  new_PATH=
-  sep_=
-  for dir in $PATH; do
-    case "$dir" in
-      /*) test -d "$dir/." || continue
-          new_PATH="${new_PATH}${sep_}${dir}"
-          sep_=':';;
-    esac
-  done
-  IFS="$saved_IFS"
-  PATH="$new_PATH"
-  export PATH
-}
-
-# This is a stub function that is run upon trap (upon regular exit and
-# interrupt).  Override it with a per-test function, e.g., to unmount
-# a partition, or to undo any other global state changes.
-cleanup_ () { :; }
-
-# Run the user-overridable cleanup_ function, remove the temporary
-# directory and exit with the incoming value of $?.
-remove_tmp_ ()
-{
-  __st=$?
-  cleanup_
-  if test "$KEEP" = yes; then
-    echo "Not removing temporary directory $test_dir_"
-  else
-    # cd out of the directory we're about to remove
-    cd "$initial_cwd_" || cd / || cd /tmp
-    chmod -R u+rwx "$test_dir_"
-    # If removal fails and exit status was to be 0, then change it to 1.
-    rm -rf "$test_dir_" || { test $__st = 0 && __st=1; }
-  fi
-  exit $__st
-}
-
-# =============================================================================
-# Prepending directories to PATH
-
-# Given a directory name, DIR, if every entry in it that matches *.exe
-# contains only the specified bytes (see the case stmt below), then print
-# a space-separated list of those names and return 0.  Otherwise, don't
-# print anything and return 1.  Naming constraints apply also to DIR.
-find_exe_basenames_ ()
-{
-  feb_dir_=$1
-  feb_fail_=0
-  feb_result_=
-  feb_sp_=
-  for feb_file_ in $feb_dir_/*.exe; do
-    # If there was no *.exe file, or there existed a file named "*.exe" that
-    # was deleted between the above glob expansion and the existence test
-    # below, just skip it.
-    test "x$feb_file_" = "x$feb_dir_/*.exe" && test ! -f "$feb_file_" \
-      && continue
-    # Exempt [.exe, since we can't create a function by that name, yet
-    # we can't invoke [ by PATH search anyways due to shell builtins.
-    test "x$feb_file_" = "x$feb_dir_/[.exe" && continue
-    case $feb_file_ in
-      *[!-a-zA-Z/0-9_.+]*) feb_fail_=1; break;;
-      *) # Remove leading file name components as well as the .exe suffix.
-         feb_file_=${feb_file_##*/}
-         feb_file_=${feb_file_%.exe}
-         feb_result_="$feb_result_$feb_sp_$feb_file_";;
-    esac
-    feb_sp_=' '
-  done
-  test $feb_fail_ = 0 && printf %s "$feb_result_"
-  return $feb_fail_
-}
-
-# Consider the files in directory, $1.
-# For each file name of the form PROG.exe, create an alias named
-# PROG that simply invokes PROG.exe, then return 0.  If any selected
-# file name or the directory name, $1, contains an unexpected character,
-# define no alias and return 1.
-create_exe_shims_ ()
-{
-  case $EXEEXT in
-    '') return 0 ;;
-    .exe) ;;
-    *) echo "$0: unexpected \$EXEEXT value: $EXEEXT" 1>&2; return 1 ;;
-  esac
-
-  base_names_=`find_exe_basenames_ $1` \
-    || { echo "$0 (exe_shim): skipping directory: $1" 1>&2; return 0; }
-
-  if test -n "$base_names_"; then
-    for base_ in $base_names_; do
-      alias "$base_"="$base_$EXEEXT"
-    done
-  fi
-
-  return 0
-}
-
-# Use this function to prepend to PATH an absolute name for each
-# specified, possibly-$initial_cwd_-relative, directory.
-path_prepend_ ()
-{
-  while test $# != 0; do
-    path_dir_=$1
-    case $path_dir_ in
-      '') fail_ "invalid path dir: '$1'";;
-      /* | ?:*) abs_path_dir_=$path_dir_;;
-      *) abs_path_dir_=$initial_cwd_/$path_dir_;;
-    esac
-    case $abs_path_dir_ in
-      *$PATH_SEPARATOR*) fail_ "invalid path dir: '$abs_path_dir_'";;
-    esac
-    PATH="$abs_path_dir_$PATH_SEPARATOR$PATH"
-
-    # Create an alias, FOO, for each FOO.exe in this directory.
-    create_exe_shims_ "$abs_path_dir_" \
-      || fail_ "something failed (above): $abs_path_dir_"
-    shift
-  done
-  export PATH
-}
-
-# =============================================================================
-# Convenience environment variables for the tests
-
-# -----------------------------------------------------------------------------
-
-# Enable glibc's malloc-perturbing option.
-# This is useful for exposing code that depends on the fact that
-# malloc-related functions often return memory that is mostly zeroed.
-# If you have the time and cycles, use valgrind to do an even better job.
-: ${MALLOC_PERTURB_=87}
-export MALLOC_PERTURB_
-
-# -----------------------------------------------------------------------------
-
-# The interpreter for Bourne-shell scripts.
-# No special standards compatibility requirements.
-# Some environments, such as Android, don't have /bin/sh.
-if test -f /bin/sh$EXEEXT; then
-  BOURNE_SHELL=/bin/sh
-else
-  BOURNE_SHELL=sh
-fi
-
-# =============================================================================
-# Convenience functions for the tests
-
-# -----------------------------------------------------------------------------
-# Return value checking
-
-# This is used to simplify checking of the return value
-# which is useful when ensuring a command fails as desired.
-# I.e., just doing `command ... &&fail=1` will not catch
-# a segfault in command for example.  With this helper you
-# instead check an explicit exit code like
-#   returns_ 1 command ... || fail
-returns_ () {
-  # Disable tracing so it doesn't interfere with stderr of the wrapped command
-  { set +x; } 2>/dev/null
-
-  local exp_exit="$1"
-  shift
-  "$@"
-  test $? -eq $exp_exit && ret_=0 || ret_=1
-
-  if test "$VERBOSE" = yes && test "$gl_set_x_corrupts_stderr_" = false; then
-    set -x
-  fi
-  { return $ret_; } 2>/dev/null
-}
-
-# -----------------------------------------------------------------------------
-# Text file comparison
-
-# Emit a header similar to that from diff -u;  Print the simulated "diff"
-# command so that the order of arguments is clear.  Don't bother with @@ lines.
-emit_diff_u_header_ ()
-{
-  printf '%s\n' "diff -u $*" \
-    "--- $1	1970-01-01" \
-    "+++ $2	1970-01-01"
-}
-
-# Arrange not to let diff or cmp operate on /dev/null,
-# since on some systems (at least OSF/1 5.1), that doesn't work.
-# When there are not two arguments, or no argument is /dev/null, return 2.
-# When one argument is /dev/null and the other is not empty,
-# cat the nonempty file to stderr and return 1.
-# Otherwise, return 0.
-compare_dev_null_ ()
-{
-  test $# = 2 || return 2
-
-  if test "x$1" = x/dev/null; then
-    test -s "$2" || return 0
-    emit_diff_u_header_ "$@"; sed 's/^/+/' "$2"
-    return 1
-  fi
-
-  if test "x$2" = x/dev/null; then
-    test -s "$1" || return 0
-    emit_diff_u_header_ "$@"; sed 's/^/-/' "$1"
-    return 1
-  fi
-
-  return 2
-}
-
-for diff_opt_ in -u -U3 -c '' no; do
-  test "$diff_opt_" != no &&
-    diff_out_=`exec 2>/dev/null
-      LC_ALL=C diff $diff_opt_ "$0" "$0" < /dev/null` &&
-    break
-done
-if test "$diff_opt_" != no; then
-  if test -z "$diff_out_"; then
-    compare_ () { LC_ALL=C diff $diff_opt_ "$@"; }
-  else
-    compare_ ()
-    {
-      # If no differences were found, AIX and HP-UX 'diff' produce output
-      # like "No differences encountered".  Hide this output.
-      LC_ALL=C diff $diff_opt_ "$@" > diff.out
-      diff_status_=$?
-      test $diff_status_ -eq 0 || cat diff.out || diff_status_=2
-      rm -f diff.out || diff_status_=2
-      return $diff_status_
-    }
-  fi
-elif cmp -s /dev/null /dev/null 2>/dev/null; then
-  compare_ () { cmp -s "$@"; }
-else
-  compare_ () { cmp "$@"; }
-fi
-
-# Usage: compare EXPECTED ACTUAL
-#
-# Given compare_dev_null_'s preprocessing, defer to compare_ if 2 or more.
-# Otherwise, propagate $? to caller: any diffs have already been printed.
-compare ()
-{
-  # This looks like it can be factored to use a simple "case $?"
-  # after unchecked compare_dev_null_ invocation, but that would
-  # fail in a "set -e" environment.
-  if compare_dev_null_ "$@"; then
-    return 0
-  else
-    case $? in
-      1) return 1;;
-      *) compare_ "$@";;
-    esac
-  fi
-}
-
-# -----------------------------------------------------------------------------
-
-# If you want to override the testdir_prefix_ function,
-# or to add more utility functions, use this file.
-test -f "$srcdir/init.cfg" \
-  && . "$srcdir/init.cfg"
-
-# =============================================================================
-# Set up the environment for the test to run in.
-
-setup_ "$@"
-# This trap is here, rather than in the setup_ function, because some
-# shells run the exit trap at shell function exit, rather than script exit.
-trap remove_tmp_ EXIT

tests/initial-tab

@@ -1,7 +1,7 @@
 #!/bin/sh
 # Exercise -T.
 
-# Copyright 2016-2023 Free Software Foundation, Inc.
+# Copyright 2016-2026 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

tests/khadafy

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Regression test for GNU grep.
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/kwset-abuse

@@ -2,7 +2,7 @@
 # Evoke a segfault in a hard-to-reach code path of kwset.c.
 # This bug affected grep versions 2.19 through 2.21.
 #
-# Copyright (C) 2015-2023 Free Software Foundation, Inc.
+# Copyright (C) 2015-2026 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

tests/long-pattern-perf

@@ -1,7 +1,7 @@
 #!/bin/sh
 # grep-2.21 would incur a 100x penalty for 10x increase in regexp length
 
-# Copyright 2015-2023 Free Software Foundation, Inc.
+# Copyright 2015-2026 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

tests/many-regex-performance

@@ -3,7 +3,7 @@
 # grep-3.4 would require O(N^2) RSS for N regexps
 # grep-3.5 requires O(N) in the most common cases.
 
-# Copyright 2020-2023 Free Software Foundation, Inc.
+# Copyright 2020-2026 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

tests/match-lines

@@ -3,7 +3,7 @@
 # grep -F -x -o PAT print an extra newline for each match.
 # This would fail for grep-2.19 and grep-2.20.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/max-count-overread

@@ -12,4 +12,20 @@ echo x > exp || framework_failure_
 yes x | timeout 10 grep -m1 x > out || fail=1
 compare exp out || fail=1
 
+# Make sure -m2 stops reading even when output is /dev/null.
+# In grep 3.11, it would continue reading.
+printf 'x\nx\nx\n' >in || framework_failure
+(grep -m2 x >/dev/null && head -n1) <in >out || fail=1
+compare exp out || fail=1
+
+# The following two tests would fail before v3.11-70
+echo x > in || framework_failure_
+echo in > exp || framework_failure_
+grep -l -m1 . in > out || fail=1
+compare exp out || fail=1
+
+# Ensure that this prints nothing and exits successfully.
+grep -q -m1 . in > out || fail=1
+compare /dev/null out || fail=1
+
 Exit $fail

tests/mb-dot-newline

@@ -2,7 +2,7 @@
 # Trigger a bug in the DFA matcher.
 # This would fail for grep-2.20.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/mb-non-UTF8-overrun

@@ -2,7 +2,7 @@
 # grep would sometimes read beyond end of input, when using a non-UTF8
 # multibyte locale.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/mb-non-UTF8-perf-Fw

@@ -1,7 +1,7 @@
 #!/bin/sh
 # Test for a performance regression with -Fw and a non-UTF8 multibyte locale.
 
-# Copyright 2019-2023 Free Software Foundation, Inc.
+# Copyright 2019-2026 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

tests/mb-non-UTF8-performance

@@ -3,7 +3,7 @@
 # The 200x is on an Intel i7-based system.
 # On an AMD FX-4100, it would take up to 2500x longer.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/mb-non-UTF8-word-boundary

@@ -1,7 +1,7 @@
 #!/bin/sh
 # grep -Fw could false-match when using a non-UTF8 multibyte locale.
 
-# Copyright 2019-2023 Free Software Foundation, Inc.
+# Copyright 2019-2026 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

tests/multibyte-white-space

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test whether \s matches SP and UTF-8 multi-byte white space characters.
 #
-# Copyright (C) 2013-2023 Free Software Foundation, Inc.
+# Copyright (C) 2013-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/multiple-begin-or-end-line

@@ -2,7 +2,7 @@
 # Test a pattern of multiple begin or end line constraints.
 # This would mistakenly print a line when using grep-2.19.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/null-byte

@@ -1,7 +1,7 @@
 #!/bin/sh
 # Test NUL bytes in patterns and data.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/options

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Test for POSIX options for grep
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/pcre

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Simple PCRE tests.
 #
-# Copyright (C) 2001, 2006, 2009-2023 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2006, 2009-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/pcre-abort

@@ -2,7 +2,7 @@
 # Show that grep handles PCRE2_ERROR_MATCHLIMIT.
 # In grep-2.8, it would abort.
 #
-# Copyright (C) 2011-2023 Free Software Foundation, Inc.
+# Copyright (C) 2011-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/pcre-ascii-digits

@@ -0,0 +1,55 @@
+#!/bin/sh
+# Ensure that grep -P's \d matches only the 10 ASCII digits.
+# With, grep-3.9, \d would match e.g., the multibyte Arabic digits.
+# The same applied to \D.
+#
+# Copyright (C) 2023-2026 Free Software Foundation, Inc.
+#
+# Copying and distribution of this file, with or without modification,
+# are permitted in any medium without royalty provided the copyright
+# notice and this notice are preserved.
+
+. "${srcdir=.}/init.sh"; path_prepend_ ../src
+require_en_utf8_locale_
+LC_ALL=en_US.UTF-8
+export LC_ALL
+require_pcre_
+
+echo . | grep -qP '(*UTF).' 2>/dev/null \
+  || skip_ 'PCRE unicode support is compiled out'
+echo 0 | grep -qP '(?aD)\d' \
+  || skip_ 'PCRE 10.42 and older lack (?aD)'
+
+fail=0
+
+# $ printf %s ٠١٢٣٤٥٦٧٨٩|od -An -to1 -w10 |sed 's/ /\\/g'; : arabic digits
+# \331\240\331\241\331\242\331\243\331\244
+# \331\245\331\246\331\247\331\250\331\251
+printf '\331\240\331\241\331\242\331\243\331\244' > in || framework_failure_
+printf '\331\245\331\246\331\247\331\250\331\251' >> in || framework_failure_
+printf '\n' >> in || framework_failure_
+
+# Ensure that \d matches no Arabic-Indic digits.
+returns_ 1 grep -P '\d' in > out || fail=1
+compare /dev/null out || fail=1
+
+# Ensure that ^\D+$ matches all the Arabic-Indic digits.
+grep -P '^\D+$' in > out || fail=1
+compare in out || fail=1
+
+# When built with PCRE2 10.43 and newer, one may use (?aD) and (?-aD)
+# to toggle between modes.  (?aD) is the default (making \d == [0-9]).
+# (?-aD) relaxes \d, making it match "all" digits.
+# Use mixed digits as input: Arabic-Indic digit zero and ASCII 4.
+printf '\331\2404\n' > in2 || framework_failure_
+
+returns_ 1 grep -P '\d\d' in2 > out || fail=1
+compare /dev/null out || fail=1
+
+grep -P '(?-aD)\d(?aD)\d' in2 > out || fail=1
+compare in2 out || fail=1
+
+returns_ 1 grep -P '\d(?-aD)\d' in2 > out || fail=1
+compare /dev/null out || fail=1
+
+Exit $fail

tests/pcre-count

@@ -2,7 +2,7 @@
 # grep -P / grep -Pc are inconsistent results
 # This bug affected grep versions 2.21 through 2.22.
 #
-# Copyright (C) 2015-2023 Free Software Foundation, Inc.
+# Copyright (C) 2015-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright

tests/pcre-infloop

@@ -2,7 +2,7 @@
 # With some versions of libpcre, apparently including 8.35,
 # the following would trigger an infinite loop in its match function.
 
-# Copyright 2014-2023 Free Software Foundation, Inc.
+# Copyright 2014-2026 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

tests/pcre-invalid-utf8-infloop

@@ -1,7 +1,7 @@
 #! /bin/sh
 # Ensure that grep -oaP doesn't infloop for invalid multi-byte input
 #
-# Copyright (C) 2015-2023 Free Software Foundation, Inc.
+# Copyright (C) 2015-2026 Free Software Foundation, Inc.
 #
 # Copying and distribution of this file, with or without modification,
 # are permitted in any medium without royalty provided the copyright