Adding upstream version 1.10.upstream/1.10upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

41 files changed, 8565 insertions, 0 deletions

diff --git a/AUTHORS b/AUTHORS
new file mode 100644
index 0000000..f36c5a2
--- /dev/null
+++ b/AUTHORS
@@ -0,0 +1 @@
+The zutils were written by Antonio Diaz Diaz.
diff --git a/COPYING b/COPYING
new file mode 100644
index 0000000..4ad17ae
--- /dev/null
+++ b/COPYING
@@ -0,0 +1,338 @@
+                    GNU GENERAL PUBLIC LICENSE
+                       Version 2, June 1991
+
+ Copyright (C) 1989, 1991 Free Software Foundation, Inc.,
+ 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+                            Preamble
+
+  The licenses for most software are designed to take away your
+freedom to share and change it.  By contrast, the GNU General Public
+License is intended to guarantee your freedom to share and change free
+software--to make sure the software is free for all its users.  This
+General Public License applies to most of the Free Software
+Foundation's software and to any other program whose authors commit to
+using it.  (Some other Free Software Foundation software is covered by
+the GNU Lesser General Public License instead.)  You can apply it to
+your programs, too.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+this service if you wish), that you receive source code or can get it
+if you want it, that you can change the software or use pieces of it
+in new free programs; and that you know you can do these things.
+
+  To protect your rights, we need to make restrictions that forbid
+anyone to deny you these rights or to ask you to surrender the rights.
+These restrictions translate to certain responsibilities for you if you
+distribute copies of the software, or if you modify it.
+
+  For example, if you distribute copies of such a program, whether
+gratis or for a fee, you must give the recipients all the rights that
+you have.  You must make sure that they, too, receive or can get the
+source code.  And you must show them these terms so they know their
+rights.
+
+  We protect your rights with two steps: (1) copyright the software, and
+(2) offer you this license which gives you legal permission to copy,
+distribute and/or modify the software.
+
+  Also, for each author's protection and ours, we want to make certain
+that everyone understands that there is no warranty for this free
+software.  If the software is modified by someone else and passed on, we
+want its recipients to know that what they have is not the original, so
+that any problems introduced by others will not reflect on the original
+authors' reputations.
+
+  Finally, any free program is threatened constantly by software
+patents.  We wish to avoid the danger that redistributors of a free
+program will individually obtain patent licenses, in effect making the
+program proprietary.  To prevent this, we have made it clear that any
+patent must be licensed for everyone's free use or not licensed at all.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+                    GNU GENERAL PUBLIC LICENSE
+   TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+
+  0. This License applies to any program or other work which contains
+a notice placed by the copyright holder saying it may be distributed
+under the terms of this General Public License.  The "Program", below,
+refers to any such program or work, and a "work based on the Program"
+means either the Program or any derivative work under copyright law:
+that is to say, a work containing the Program or a portion of it,
+either verbatim or with modifications and/or translated into another
+language.  (Hereinafter, translation is included without limitation in
+the term "modification".)  Each licensee is addressed as "you".
+
+Activities other than copying, distribution and modification are not
+covered by this License; they are outside its scope.  The act of
+running the Program is not restricted, and the output from the Program
+is covered only if its contents constitute a work based on the
+Program (independent of having been made by running the Program).
+Whether that is true depends on what the Program does.
+
+  1. You may copy and distribute verbatim copies of the Program's
+source code as you receive it, in any medium, provided that you
+conspicuously and appropriately publish on each copy an appropriate
+copyright notice and disclaimer of warranty; keep intact all the
+notices that refer to this License and to the absence of any warranty;
+and give any other recipients of the Program a copy of this License
+along with the Program.
+
+You may charge a fee for the physical act of transferring a copy, and
+you may at your option offer warranty protection in exchange for a fee.
+
+  2. You may modify your copy or copies of the Program or any portion
+of it, thus forming a work based on the Program, and copy and
+distribute such modifications or work under the terms of Section 1
+above, provided that you also meet all of these conditions:
+
+    a) You must cause the modified files to carry prominent notices
+    stating that you changed the files and the date of any change.
+
+    b) You must cause any work that you distribute or publish, that in
+    whole or in part contains or is derived from the Program or any
+    part thereof, to be licensed as a whole at no charge to all third
+    parties under the terms of this License.
+
+    c) If the modified program normally reads commands interactively
+    when run, you must cause it, when started running for such
+    interactive use in the most ordinary way, to print or display an
+    announcement including an appropriate copyright notice and a
+    notice that there is no warranty (or else, saying that you provide
+    a warranty) and that users may redistribute the program under
+    these conditions, and telling the user how to view a copy of this
+    License.  (Exception: if the Program itself is interactive but
+    does not normally print such an announcement, your work based on
+    the Program is not required to print an announcement.)
+
+These requirements apply to the modified work as a whole.  If
+identifiable sections of that work are not derived from the Program,
+and can be reasonably considered independent and separate works in
+themselves, then this License, and its terms, do not apply to those
+sections when you distribute them as separate works.  But when you
+distribute the same sections as part of a whole which is a work based
+on the Program, the distribution of the whole must be on the terms of
+this License, whose permissions for other licensees extend to the
+entire whole, and thus to each and every part regardless of who wrote it.
+
+Thus, it is not the intent of this section to claim rights or contest
+your rights to work written entirely by you; rather, the intent is to
+exercise the right to control the distribution of derivative or
+collective works based on the Program.
+
+In addition, mere aggregation of another work not based on the Program
+with the Program (or with a work based on the Program) on a volume of
+a storage or distribution medium does not bring the other work under
+the scope of this License.
+
+  3. You may copy and distribute the Program (or a work based on it,
+under Section 2) in object code or executable form under the terms of
+Sections 1 and 2 above provided that you also do one of the following:
+
+    a) Accompany it with the complete corresponding machine-readable
+    source code, which must be distributed under the terms of Sections
+    1 and 2 above on a medium customarily used for software interchange; or,
+
+    b) Accompany it with a written offer, valid for at least three
+    years, to give any third party, for a charge no more than your
+    cost of physically performing source distribution, a complete
+    machine-readable copy of the corresponding source code, to be
+    distributed under the terms of Sections 1 and 2 above on a medium
+    customarily used for software interchange; or,
+
+    c) Accompany it with the information you received as to the offer
+    to distribute corresponding source code.  (This alternative is
+    allowed only for noncommercial distribution and only if you
+    received the program in object code or executable form with such
+    an offer, in accord with Subsection b above.)
+
+The source code for a work means the preferred form of the work for
+making modifications to it.  For an executable work, complete source
+code means all the source code for all modules it contains, plus any
+associated interface definition files, plus the scripts used to
+control compilation and installation of the executable.  However, as a
+special exception, the source code distributed need not include
+anything that is normally distributed (in either source or binary
+form) with the major components (compiler, kernel, and so on) of the
+operating system on which the executable runs, unless that component
+itself accompanies the executable.
+
+If distribution of executable or object code is made by offering
+access to copy from a designated place, then offering equivalent
+access to copy the source code from the same place counts as
+distribution of the source code, even though third parties are not
+compelled to copy the source along with the object code.
+
+  4. You may not copy, modify, sublicense, or distribute the Program
+except as expressly provided under this License.  Any attempt
+otherwise to copy, modify, sublicense or distribute the Program is
+void, and will automatically terminate your rights under this License.
+However, parties who have received copies, or rights, from you under
+this License will not have their licenses terminated so long as such
+parties remain in full compliance.
+
+  5. You are not required to accept this License, since you have not
+signed it.  However, nothing else grants you permission to modify or
+distribute the Program or its derivative works.  These actions are
+prohibited by law if you do not accept this License.  Therefore, by
+modifying or distributing the Program (or any work based on the
+Program), you indicate your acceptance of this License to do so, and
+all its terms and conditions for copying, distributing or modifying
+the Program or works based on it.
+
+  6. Each time you redistribute the Program (or any work based on the
+Program), the recipient automatically receives a license from the
+original licensor to copy, distribute or modify the Program subject to
+these terms and conditions.  You may not impose any further
+restrictions on the recipients' exercise of the rights granted herein.
+You are not responsible for enforcing compliance by third parties to
+this License.
+
+  7. If, as a consequence of a court judgment or allegation of patent
+infringement or for any other reason (not limited to patent issues),
+conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot
+distribute so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you
+may not distribute the Program at all.  For example, if a patent
+license would not permit royalty-free redistribution of the Program by
+all those who receive copies directly or indirectly through you, then
+the only way you could satisfy both it and this License would be to
+refrain entirely from distribution of the Program.
+
+If any portion of this section is held invalid or unenforceable under
+any particular circumstance, the balance of the section is intended to
+apply and the section as a whole is intended to apply in other
+circumstances.
+
+It is not the purpose of this section to induce you to infringe any
+patents or other property right claims or to contest validity of any
+such claims; this section has the sole purpose of protecting the
+integrity of the free software distribution system, which is
+implemented by public license practices.  Many people have made
+generous contributions to the wide range of software distributed
+through that system in reliance on consistent application of that
+system; it is up to the author/donor to decide if he or she is willing
+to distribute software through any other system and a licensee cannot
+impose that choice.
+
+This section is intended to make thoroughly clear what is believed to
+be a consequence of the rest of this License.
+
+  8. If the distribution and/or use of the Program is restricted in
+certain countries either by patents or by copyrighted interfaces, the
+original copyright holder who places the Program under this License
+may add an explicit geographical distribution limitation excluding
+those countries, so that distribution is permitted only in or among
+countries not thus excluded.  In such case, this License incorporates
+the limitation as if written in the body of this License.
+
+  9. The Free Software Foundation may publish revised and/or new versions
+of the General Public License from time to time.  Such new versions will
+be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+Each version is given a distinguishing version number.  If the Program
+specifies a version number of this License which applies to it and "any
+later version", you have the option of following the terms and conditions
+either of that version or of any later version published by the Free
+Software Foundation.  If the Program does not specify a version number of
+this License, you may choose any version ever published by the Free Software
+Foundation.
+
+  10. If you wish to incorporate parts of the Program into other free
+programs whose distribution conditions are different, write to the author
+to ask for permission.  For software which is copyrighted by the Free
+Software Foundation, write to the Free Software Foundation; we sometimes
+make exceptions for this.  Our decision will be guided by the two goals
+of preserving the free status of all derivatives of our free software and
+of promoting the sharing and reuse of software generally.
+
+                            NO WARRANTY
+
+  11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
+FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
+OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
+PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
+OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS
+TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
+PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
+REPAIR OR CORRECTION.
+
+  12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
+REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
+INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
+OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
+TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
+YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
+PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGES.
+
+                     END OF TERMS AND CONDITIONS
+
+            How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+convey the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the program's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation, either version 2 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+Also add information on how to contact you by electronic and paper mail.
+
+If the program is interactive, make it output a short notice like this
+when it starts in an interactive mode:
+
+    Gnomovision version 69, Copyright (C) <year>  <name of author>
+    Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
+    This is free software, and you are welcome to redistribute it
+    under certain conditions; type `show c' for details.
+
+The hypothetical commands `show w' and `show c' should show the appropriate
+parts of the General Public License.  Of course, the commands you use may
+be called something other than `show w' and `show c'; they could even be
+mouse-clicks or menu items--whatever suits your program.
+
+You should also get your employer (if you work as a programmer) or your
+school, if any, to sign a "copyright disclaimer" for the program, if
+necessary.  Here is a sample; alter the names:
+
+  Yoyodyne, Inc., hereby disclaims all copyright interest in the program
+  `Gnomovision' (which makes passes at compilers) written by James Hacker.
+
+  <signature of Ty Coon>, 1 April 1989
+  Ty Coon, President of Vice
+
+This General Public License does not permit incorporating your program into
+proprietary programs.  If your program is a subroutine library, you may
+consider it more useful to permit linking proprietary applications with the
+library.  If this is what you want to do, use the GNU Lesser General
+Public License instead of this License.
diff --git a/ChangeLog b/ChangeLog
new file mode 100644
index 0000000..cdbc3da
--- /dev/null
+++ b/ChangeLog
@@ -0,0 +1,187 @@
+2021-01-05  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.10 released.
+	* zdiff.cc (set_fifonames): Encode pid in little endian order.
+	* zupdate.cc (zupdate_file): Fix a portability issue with Solaris 10.
+	* zutils.texi: Document that 'zgrep -L' fails with GNU grep 3.2 to 3.4.
+	* check.sh: Test empty input files with all tools except zupdate.
+
+2020-06-27  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.9 released.
+	* zcmp.cc, zdiff.cc: Read standard input only if requested.
+	* zdiff.cc (main): Pass options '-W' and '-y' to diff.
+	* zutils.cc (test_format): Detect bzip2 and lzip files better.
+	* ztest.cc (main): Continue testing if any input file is a terminal.
+	  If verbosity >= 1, print number of files that failed the test.
+	* zcat.cc, zgrep.cc, ztest.cc (main): Check return value of close(infd).
+	* zutils.cc (good_status): Ignore trailing data remaining in feeder.
+	* zupdate.cc (zupdate_file): Support new and old lzip option '-o'.
+	  Keep combined extensions: tgz, tbz, tbz2, txz --> tlz.
+	  Quote file names in zcmp_command to allow file names with spaces.
+	* *.cc (main): Set a valid invocation_name even if argc == 0.
+	* zutils.texi: Improve descriptions of zcat, zcmp, and zdiff.
+
+2019-01-01  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.8 released.
+	* zcat.cc: Fix a buffer overflow on outbuf when '-v' is used.
+	* zcat.cc (cat): A canary byte has been added to outbuf.
+	* New option '-R, --dereference-recursive'.
+	* Option '-r, --recursive' now skips symlinks.
+	* If no files and recursive, examine current working directory.
+	* recursive.cc (test_full_name): Detect directory loops.
+	* recursive.cc: Ignore directories if not --recursive.
+	* recursive.cc: Remove extra trailing slashes from directory args.
+	* zcatgrep.cc (open_instream): Show correct errno.
+	* zutils.cc (good_status): Wait for killed child.
+	* Test and document continuation or exit of zcat, zgrep, ztest,
+	  and zupdate in case of error.
+	* configure: Accept appending to CXXFLAGS, 'CXXFLAGS+=OPTIONS'.
+
+2018-02-13  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.7 released.
+	* zgrep.cc (main): Pass option '--color' to grep.
+	* check.sh: Add new tests for zgrep.
+
+2017-04-05  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.6 released.
+	* zcmp.cc: Accept 'B' suffix in '--ignore-initial=1kB:1234B'.
+	* zutils.cc (feed_data): Show input file name in error messages.
+
+2016-05-15  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.5 released.
+	* zupdate.cc (zupdate_file): Pass '-q' to zcmp if verbosity < 0.
+	* zcat.cc, zgrep.cc, ztest.cc (main): Don't use stdin more than once.
+	* zdiff.cc (set_fifonames): Use '_' if both names are different.
+	* configure: Avoid warning on some shells when testing for g++.
+	* Makefile.in: Detect the existence of install-info.
+	* check.sh: A POSIX shell is required to run the tests.
+
+2015-05-29  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.4 released.
+	* Option '--format' has been renamed to '-O, --force-format'.
+	* Add new option '-M, --format=<list>' to all utilities.
+	* zgrep.cc (main): Pass '-e' to grep if pattern begins with '-'.
+	* Makefile.in: New targets 'install*-compress'.
+
+2014-08-30  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.3 released.
+	* check.sh: Fix two values of expected exit status.
+	* zutils.texi: Document that '--format' does not verify format.
+	* Add two missing #includes.
+	* Change license to GPL version 2 or later.
+
+2014-02-01  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.2 released.
+	* New utility; zupdate.
+	* Remove zutils executable. Utils are now independent executables.
+	* zgrep.cc: Fix the exit status returned on error.
+	* zutils.texinfo: Rename to zutils.texi.
+
+2013-08-02  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.1 released.
+	* Add options '--bz2', '--gz', '--lz', and '--xz' to all utilities.
+	* Add runtime configuration file 'zutilsrc'.
+	* New function 'good_status' checks exit status of all children.
+	* Fix all uses of decompressed/uncompressed in the documentation.
+
+2013-05-31  Antonio Diaz Diaz  <antonio@gnu.org>
+
+	* Version 1.0 released.
+	* Add new option '--format' to all utilities.
+	* main.cc (main): Make 'grep_show_name' tri-state so that file
+	  name is no prefixed to output by default when searching one
+	  file and '--recursive' has not been selected.
+	* zgrep.cc: Fix output of option '-L' (it behaved like '-l').
+	* zcmp.cc: Fix deadlock when option '-n' is used.
+	* zdiff.cc (set_data_feeder): Call compressor with option '-q'
+	  only if verbosity < 0.
+	* zutils.cc (set_data_feeder): Likewise.
+	* Change quote characters in messages as advised by GNU Standards.
+	* configure: Options now accept a separate argument.
+	  Rename 'datadir' to 'datarootdir'. Ignore environment variables.
+	* Makefile.in: New target 'install-bin'.
+	* Use 'setmode' instead of '_setmode' on Windows and OS/2.
+	* zcat.cc (Line_number): Fix a portability issue with Solaris 9.
+	* INSTALL: Document installing zutils along with GNU gzip.
+
+2011-01-11  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.9 released.
+	* configure: New options 'DIFF' and 'GREP'.
+	* zcmp.cc: Fix deadlock when files differ.
+	* zgrep.cc: Fix deadlock when binary file matches.
+
+2010-11-15  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.8 released.
+	* main.cc: New options '--zcat', '--zgrep', and '--ztest'.
+	* zcat.cc: New file implementing zcat+cat functionality in C++.
+	* zcmp.cc: New file implementing zcmp+cmp functionality in C++.
+	* doc/zcmp.1: New file.
+	* Remove files zcmp.in, zdiff.in.
+	* zdiff.cc: New file implementing zdiff functionality in C++.
+	* zgrep.cc: New file implementing zgrep functionality in C++.
+	* All mentions to zegrep and zfgrep have been removed from the
+	  documentation because egrep and fgrep are deprecated.
+	* ztest.cc: New file implementing ztest functionality in C++.
+	* Makefile.in: Add quotes to directory names.
+	* check.sh: Use 'test.txt' instead of 'COPYING' for testing.
+	* Remove environment safeguards from configure as requested by
+	  Richard Stallman. Now environment variables affect configure.
+
+2009-10-21  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.7 released.
+	* New utility; ztest.
+	* zcat.in: New option '-r, --recursive'.
+
+2009-10-05  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.6 released.
+	* zcat.in, zgrep.in: Remove again default compressor. Format of
+	  data read from stdin is now automatically detected.
+	* Makefile.in: Add option '--name' to help2man invocation.
+
+2009-10-01  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.5 released.
+	* zcat.in, zgrep.in: Read again data from stdin.
+	* Add again default compressor for stdin only.
+
+2009-09-17  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.4 released.
+	* Add two new utilities; zegrep and zfgrep.
+	* Add zutils executable which recognizes file formats.
+
+2009-08-28  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.3 released.
+	* Remove default compressor.
+	* zcat.in, zgrep.in: Don't read data from stdin.
+	* Update home page and mailing list addresses.
+
+2009-08-13  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.2 released.
+	* Add support for xz.
+
+2009-08-07  Antonio Diaz Diaz  <ant_diaz@teleline.es>
+
+	* Version 0.1 released.
+
+
+Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+This file is a collection of facts, and thus it is not copyrightable,
+but just in case, you have unlimited permission to copy, distribute, and
+modify it.
diff --git a/INSTALL b/INSTALL
new file mode 100644
index 0000000..3c9c429
--- /dev/null
+++ b/INSTALL
@@ -0,0 +1,81 @@
+Requirements
+------------
+You will need a C++11 compiler. (gcc 3.3.6 or newer is recommended).
+I use gcc 6.1.0 and 4.1.2, but the code should compile with any standards
+compliant compiler.
+Gcc is available at http://gcc.gnu.org.
+
+POSIX compliant versions of diff and grep are required for zdiff and zgrep.
+
+(Option -L of zgrep fails (prints wrong results, returns wrong status, and
+even hangs) when using GNU grep versions 3.2 to 3.4 inclusive because of a
+wrong change in the exit status of grep, which was reverted in GNU grep 3.5).
+
+Compressors for bzip2, gzip and lzip formats are required to run the tests.
+
+If you are installing zutils along with GNU gzip and want to keep the
+gzip scripts, the recommended method is to configure gzip as follows:
+
+   ./configure --program-transform-name='s/^z/gz/'
+
+This renames, at installation time, the gzip scripts and man pages to
+'gzcat', 'gzcat.1', etc, avoiding the name clashing with the programs
+and man pages from zutils.
+
+
+Procedure
+---------
+1. Unpack the archive if you have not done so already:
+
+	tar -xf zutils[version].tar.lz
+or
+	lzip -cd zutils[version].tar.lz | tar -xf -
+
+This creates the directory ./zutils[version] containing the source from
+the main archive.
+
+2. Change to zutils directory and run configure.
+   (Try 'configure --help' for usage instructions).
+
+	cd zutils[version]
+	./configure
+
+3. Run make.
+
+	make
+
+4. Optionally, type 'make check' to run the tests that come with zutils.
+
+5. Type 'make install' to install the programs and any data files and
+   documentation.
+
+   Or type 'make install-compress', which additionally compresses the
+   info manual and the man pages after installation.
+   (Installing compressed docs may become the default in the future).
+
+   You can install only the programs, the info manual, or the man pages by
+   typing 'make install-bin', 'make install-info', or 'make install-man'
+   respectively.
+
+
+Another way
+-----------
+You can also compile zutils into a separate directory.
+To do this, you must use a version of 'make' that supports the variable
+'VPATH', such as GNU 'make'. 'cd' to the directory where you want the
+object files and executables to go and run the 'configure' script.
+'configure' automatically checks for the source code in '.', in '..', and
+in the directory that 'configure' is in.
+
+'configure' recognizes the option '--srcdir=DIR' to control where to
+look for the sources. Usually 'configure' can determine that directory
+automatically.
+
+After running 'configure', you can run 'make' and 'make install' as
+explained above.
+
+
+Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+This file is free documentation: you have unlimited permission to copy,
+distribute, and modify it.
diff --git a/Makefile.in b/Makefile.in
new file mode 100644
index 0000000..55a974e
--- /dev/null
+++ b/Makefile.in
@@ -0,0 +1,238 @@
+
+DISTNAME = $(pkgname)-$(pkgversion)
+INSTALL = install
+INSTALL_PROGRAM = $(INSTALL) -m 755
+INSTALL_SCRIPT = $(INSTALL) -m 755
+INSTALL_DATA = $(INSTALL) -m 644
+INSTALL_DIR = $(INSTALL) -d -m 755
+SHELL = /bin/sh
+CAN_RUN_INSTALLINFO = $(SHELL) -c "install-info --version" > /dev/null 2>&1
+
+objs = arg_parser.o rc.o zutils.o \
+       zcat.o zcmp.o zdiff.o zgrep.o ztest.o zupdate.o
+zcat_objs = arg_parser.o rc.o zutils.o zcat.o
+zcmp_objs = arg_parser.o rc.o zutils.o zcmp.o
+zdiff_objs = arg_parser.o rc.o zutils.o zdiff.o
+zgrep_objs = arg_parser.o rc.o zutils.o zgrep.o
+ztest_objs = arg_parser.o rc.o zutils.o ztest.o
+zupdate_objs = arg_parser.o rc.o zupdate.o
+programs = zcat zcmp zdiff zgrep ztest zupdate
+scripts = zegrep zfgrep
+
+
+.PHONY : all install install-bin install-info install-man \
+         install-strip install-compress install-strip-compress \
+         install-bin-strip install-info-compress install-man-compress \
+         uninstall uninstall-bin uninstall-info uninstall-man \
+         doc info man check dist clean distclean
+
+all : $(programs) $(scripts)
+
+zcat : $(zcat_objs)
+	$(CXX) $(LDFLAGS) $(CXXFLAGS) -o $@ $(zcat_objs)
+
+zcmp : $(zcmp_objs)
+	$(CXX) $(LDFLAGS) $(CXXFLAGS) -o $@ $(zcmp_objs)
+
+zdiff : $(zdiff_objs)
+	$(CXX) $(LDFLAGS) $(CXXFLAGS) -o $@ $(zdiff_objs)
+
+zegrep : zegrep.in
+	cat $(VPATH)/zegrep.in > $@
+	chmod a+x zegrep
+
+zfgrep : zfgrep.in
+	cat $(VPATH)/zfgrep.in > $@
+	chmod a+x zfgrep
+
+zgrep : $(zgrep_objs)
+	$(CXX) $(LDFLAGS) $(CXXFLAGS) -o $@ $(zgrep_objs)
+
+ztest : $(ztest_objs)
+	$(CXX) $(LDFLAGS) $(CXXFLAGS) -o $@ $(ztest_objs)
+
+zupdate : $(zupdate_objs)
+	$(CXX) $(LDFLAGS) $(CXXFLAGS) -o $@ $(zupdate_objs)
+
+rc.o : rc.cc
+	$(CXX) $(CPPFLAGS) $(CXXFLAGS) -DPROGVERSION=\"$(pkgversion)\" -DSYSCONFDIR=\"$(sysconfdir)\" -c -o $@ $<
+
+zdiff.o : zdiff.cc
+	$(CXX) $(CPPFLAGS) $(CXXFLAGS) -DDIFF=\"$(DIFF)\" -c -o $@ $<
+
+zgrep.o : zgrep.cc
+	$(CXX) $(CPPFLAGS) $(CXXFLAGS) -DGREP=\"$(GREP)\" -c -o $@ $<
+
+%.o : %.cc
+	$(CXX) $(CPPFLAGS) $(CXXFLAGS) -c -o $@ $<
+
+$(objs)      : Makefile
+$(scripts)   : Makefile
+arg_parser.o : arg_parser.h
+rc.o         : arg_parser.h rc.h
+zcat.o       : arg_parser.h rc.h zutils.h recursive.cc zcatgrep.cc
+zcmp.o       : arg_parser.h rc.h zutils.h zcmpdiff.cc
+zdiff.o      : arg_parser.h rc.h zutils.h zcmpdiff.cc
+zgrep.o      : arg_parser.h rc.h zutils.h recursive.cc zcatgrep.cc
+ztest.o      : arg_parser.h rc.h zutils.h recursive.cc
+zupdate.o    : arg_parser.h rc.h recursive.cc
+zutils.o     : rc.h zutils.h
+
+
+doc : info man
+
+info : $(VPATH)/doc/$(pkgname).info
+
+$(VPATH)/doc/$(pkgname).info : $(VPATH)/doc/$(pkgname).texi
+	cd $(VPATH)/doc && makeinfo $(pkgname).texi
+
+man : $(VPATH)/doc/zcat.1 $(VPATH)/doc/zcmp.1 $(VPATH)/doc/zdiff.1 \
+      $(VPATH)/doc/zgrep.1 $(VPATH)/doc/ztest.1 $(VPATH)/doc/zupdate.1
+
+$(VPATH)/doc/zcat.1 : zcat
+	help2man -n 'decompress and concatenate files to standard output' \
+	  -o $@ --no-info ./zcat
+
+$(VPATH)/doc/zcmp.1 : zcmp
+	help2man -n 'decompress and compare two files byte by byte' \
+	  -o $@ --no-info ./zcmp
+
+$(VPATH)/doc/zdiff.1 : zdiff
+	help2man -n 'decompress and compare two files line by line' \
+	  -o $@ --no-info ./zdiff
+
+$(VPATH)/doc/zgrep.1 : zgrep
+	help2man -n 'search compressed files for a regular expression' \
+	  -o $@ --no-info ./zgrep
+
+$(VPATH)/doc/ztest.1 : ztest
+	help2man -n 'verify the integrity of compressed files' \
+	  -o $@ --no-info ./ztest
+
+$(VPATH)/doc/zupdate.1 : zupdate
+	help2man -n 'recompress bzip2, gzip, xz files to lzip format' \
+	  -o $@ --no-info ./zupdate
+
+Makefile : $(VPATH)/configure $(VPATH)/Makefile.in
+	./config.status
+
+check : all
+	@$(VPATH)/testsuite/check.sh $(VPATH)/testsuite $(pkgversion)
+
+install : install-bin install-info install-man
+install-strip : install-bin-strip install-info install-man
+install-compress : install-bin install-info-compress install-man-compress
+install-strip-compress : install-bin-strip install-info-compress install-man-compress
+
+install-bin : all
+	if [ ! -d "$(DESTDIR)$(bindir)" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(bindir)" ; fi
+	$(INSTALL_PROGRAM) ./zcat "$(DESTDIR)$(bindir)/zcat"
+	$(INSTALL_PROGRAM) ./zcmp "$(DESTDIR)$(bindir)/zcmp"
+	$(INSTALL_PROGRAM) ./zdiff "$(DESTDIR)$(bindir)/zdiff"
+	$(INSTALL_SCRIPT) ./zegrep "$(DESTDIR)$(bindir)/zegrep"
+	$(INSTALL_SCRIPT) ./zfgrep "$(DESTDIR)$(bindir)/zfgrep"
+	$(INSTALL_PROGRAM) ./zgrep "$(DESTDIR)$(bindir)/zgrep"
+	$(INSTALL_PROGRAM) ./ztest "$(DESTDIR)$(bindir)/ztest"
+	$(INSTALL_PROGRAM) ./zupdate "$(DESTDIR)$(bindir)/zupdate"
+	if [ ! -e "$(DESTDIR)$(sysconfdir)/$(pkgname)rc" ] ; then \
+	  if [ ! -d "$(DESTDIR)$(sysconfdir)" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(sysconfdir)" ; fi ; \
+	  $(INSTALL_DATA) $(VPATH)/$(pkgname)rc "$(DESTDIR)$(sysconfdir)/$(pkgname)rc" ; \
+	fi
+
+install-bin-strip : all
+	$(MAKE) INSTALL_PROGRAM='$(INSTALL_PROGRAM) -s' install-bin
+
+install-info :
+	if [ ! -d "$(DESTDIR)$(infodir)" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(infodir)" ; fi
+	-rm -f "$(DESTDIR)$(infodir)/$(pkgname).info"*
+	$(INSTALL_DATA) $(VPATH)/doc/$(pkgname).info "$(DESTDIR)$(infodir)/$(pkgname).info"
+	-if $(CAN_RUN_INSTALLINFO) ; then \
+	  install-info --info-dir="$(DESTDIR)$(infodir)" "$(DESTDIR)$(infodir)/$(pkgname).info" ; \
+	fi
+
+install-info-compress : install-info
+	lzip -v -9 "$(DESTDIR)$(infodir)/$(pkgname).info"
+
+install-man :
+	if [ ! -d "$(DESTDIR)$(mandir)/man1" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(mandir)/man1" ; fi
+	-rm -f "$(DESTDIR)$(mandir)/man1/zcat.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/zcmp.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/zdiff.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/zgrep.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/ztest.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/zupdate.1"*
+	$(INSTALL_DATA) $(VPATH)/doc/zcat.1 "$(DESTDIR)$(mandir)/man1/zcat.1"
+	$(INSTALL_DATA) $(VPATH)/doc/zcmp.1 "$(DESTDIR)$(mandir)/man1/zcmp.1"
+	$(INSTALL_DATA) $(VPATH)/doc/zdiff.1 "$(DESTDIR)$(mandir)/man1/zdiff.1"
+	$(INSTALL_DATA) $(VPATH)/doc/zgrep.1 "$(DESTDIR)$(mandir)/man1/zgrep.1"
+	$(INSTALL_DATA) $(VPATH)/doc/ztest.1 "$(DESTDIR)$(mandir)/man1/ztest.1"
+	$(INSTALL_DATA) $(VPATH)/doc/zupdate.1 "$(DESTDIR)$(mandir)/man1/zupdate.1"
+
+install-man-compress : install-man
+	lzip -v -9 "$(DESTDIR)$(mandir)/man1/zcat.1"
+	lzip -v -9 "$(DESTDIR)$(mandir)/man1/zcmp.1"
+	lzip -v -9 "$(DESTDIR)$(mandir)/man1/zdiff.1"
+	lzip -v -9 "$(DESTDIR)$(mandir)/man1/zgrep.1"
+	lzip -v -9 "$(DESTDIR)$(mandir)/man1/ztest.1"
+	lzip -v -9 "$(DESTDIR)$(mandir)/man1/zupdate.1"
+
+uninstall : uninstall-man uninstall-info uninstall-bin
+
+uninstall-bin :
+	-rm -f "$(DESTDIR)$(bindir)/zcat"
+	-rm -f "$(DESTDIR)$(bindir)/zcmp"
+	-rm -f "$(DESTDIR)$(bindir)/zdiff"
+	-rm -f "$(DESTDIR)$(bindir)/zegrep"
+	-rm -f "$(DESTDIR)$(bindir)/zfgrep"
+	-rm -f "$(DESTDIR)$(bindir)/zgrep"
+	-rm -f "$(DESTDIR)$(bindir)/ztest"
+	-rm -f "$(DESTDIR)$(bindir)/zupdate"
+	-rm -f "$(DESTDIR)$(sysconfdir)/$(pkgname)rc"
+
+uninstall-info :
+	-if $(CAN_RUN_INSTALLINFO) ; then \
+	  install-info --info-dir="$(DESTDIR)$(infodir)" --remove "$(DESTDIR)$(infodir)/$(pkgname).info" ; \
+	fi
+	-rm -f "$(DESTDIR)$(infodir)/$(pkgname).info"*
+
+uninstall-man :
+	-rm -f "$(DESTDIR)$(mandir)/man1/zcat.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/zcmp.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/zdiff.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/zgrep.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/ztest.1"*
+	-rm -f "$(DESTDIR)$(mandir)/man1/zupdate.1"*
+
+dist : doc
+	ln -sf $(VPATH) $(DISTNAME)
+	tar -Hustar --owner=root --group=root -cvf $(DISTNAME).tar \
+	  $(DISTNAME)/AUTHORS \
+	  $(DISTNAME)/COPYING \
+	  $(DISTNAME)/ChangeLog \
+	  $(DISTNAME)/INSTALL \
+	  $(DISTNAME)/Makefile.in \
+	  $(DISTNAME)/NEWS \
+	  $(DISTNAME)/README \
+	  $(DISTNAME)/configure \
+	  $(DISTNAME)/doc/*.1 \
+	  $(DISTNAME)/doc/$(pkgname).info \
+	  $(DISTNAME)/doc/$(pkgname).texi \
+	  $(DISTNAME)/$(pkgname)rc \
+	  $(DISTNAME)/*.h \
+	  $(DISTNAME)/*.cc \
+	  $(DISTNAME)/z*.in \
+	  $(DISTNAME)/testsuite/check.sh \
+	  $(DISTNAME)/testsuite/test.txt \
+	  $(DISTNAME)/testsuite/test.txt.tar \
+	  $(DISTNAME)/testsuite/zcat_vs.dat \
+	  $(DISTNAME)/testsuite/test_bad_crc.lz \
+	  $(DISTNAME)/testsuite/zero_bad_crc.lz \
+	  $(DISTNAME)/testsuite/zero_bad_crc.gz
+	rm -f $(DISTNAME)
+	lzip -v -9 $(DISTNAME).tar
+
+clean :
+	-rm -f $(programs) $(scripts) $(objs)
+
+distclean : clean
+	-rm -f Makefile config.status *.tar *.tar.lz
diff --git a/NEWS b/NEWS
new file mode 100644
index 0000000..4b62179
--- /dev/null
+++ b/NEWS
@@ -0,0 +1,9 @@
+Changes in version 1.10:
+
+A portability issue with Solaris 10 has been fixed.
+
+It has been documented in the manual that 'zgrep -L' fails with GNU grep
+versions 3.2 to 3.4 inclusive because of a wrong change reverted in GNU grep
+3.5.
+
+'make check' now tests empty input files with all tools except zupdate.
diff --git a/README b/README
new file mode 100644
index 0000000..8725fe3
--- /dev/null
+++ b/README
@@ -0,0 +1,47 @@
+Description
+
+Zutils is a collection of utilities able to process any combination of
+compressed and uncompressed files transparently. If any file given,
+including standard input, is compressed, its decompressed content is used.
+Compressed files are decompressed on the fly; no temporary files are
+created.
+
+These utilities are not wrapper scripts but safer and more efficient C++
+programs. In particular the option '--recursive' is very efficient in
+those utilities supporting it.
+
+The utilities provided are zcat, zcmp, zdiff, zgrep, ztest, and zupdate.
+The formats supported are bzip2, gzip, lzip, and xz.
+Zutils uses external compressors. The compressor to be used for each format
+is configurable at runtime.
+
+zcat, zcmp, zdiff, and zgrep are improved replacements for the shell scripts
+provided by GNU gzip. ztest is unique to zutils. zupdate is similar to
+gzip's znew.
+
+NOTE: Bzip2 and lzip provide well-defined values of exit status, which makes
+them safe to use with zutils. Gzip and xz may return ambiguous warning
+values, making them less reliable back ends for zutils.
+
+FORMAT NOTE 1: The option '--format' allows the processing of a subset
+of formats in recursive mode and when trying compressed file names:
+'zgrep foo -r --format=bz2,lz somedir somefile.tar'.
+
+FORMAT NOTE 2: If the option '--force-format' is given, the files are
+passed to the corresponding decompressor without verifying their format,
+allowing for example the processing of compress'd (.Z) files with gzip:
+'zcmp --force-format=gz file.Z file.lz'.
+
+LANGUAGE NOTE: Uncompressed = not compressed = plain data; it may never have
+been compressed. Decompressed is used to refer to data which have undergone
+the process of decompression.
+
+
+Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+This file is free documentation: you have unlimited permission to copy,
+distribute, and modify it.
+
+The file Makefile.in is a data file used by configure to produce the
+Makefile. It has the same copyright owner and permissions that configure
+itself.
diff --git a/arg_parser.cc b/arg_parser.cc
new file mode 100644
index 0000000..2e40a13
--- /dev/null
+++ b/arg_parser.cc
@@ -0,0 +1,196 @@
+/* Arg_parser - POSIX/GNU command line argument parser. (C++ version)
+   Copyright (C) 2006-2021 Antonio Diaz Diaz.
+
+   This library is free software. Redistribution and use in source and
+   binary forms, with or without modification, are permitted provided
+   that the following conditions are met:
+
+   1. Redistributions of source code must retain the above copyright
+   notice, this list of conditions, and the following disclaimer.
+
+   2. Redistributions in binary form must reproduce the above copyright
+   notice, this list of conditions, and the following disclaimer in the
+   documentation and/or other materials provided with the distribution.
+
+   This library 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.
+*/
+
+#include <cstring>
+#include <string>
+#include <vector>
+
+#include "arg_parser.h"
+
+
+bool Arg_parser::parse_long_option( const char * const opt, const char * const arg,
+                                    const Option options[], int & argind )
+  {
+  unsigned len;
+  int index = -1;
+  bool exact = false, ambig = false;
+
+  for( len = 0; opt[len+2] && opt[len+2] != '='; ++len ) ;
+
+  // Test all long options for either exact match or abbreviated matches.
+  for( int i = 0; options[i].code != 0; ++i )
+    if( options[i].name && std::strncmp( options[i].name, &opt[2], len ) == 0 )
+      {
+      if( std::strlen( options[i].name ) == len )	// Exact match found
+        { index = i; exact = true; break; }
+      else if( index < 0 ) index = i;		// First nonexact match found
+      else if( options[index].code != options[i].code ||
+               options[index].has_arg != options[i].has_arg )
+        ambig = true;		// Second or later nonexact match found
+      }
+
+  if( ambig && !exact )
+    {
+    error_ = "option '"; error_ += opt; error_ += "' is ambiguous";
+    return false;
+    }
+
+  if( index < 0 )		// nothing found
+    {
+    error_ = "unrecognized option '"; error_ += opt; error_ += '\'';
+    return false;
+    }
+
+  ++argind;
+  data.push_back( Record( options[index].code ) );
+
+  if( opt[len+2] )		// '--<long_option>=<argument>' syntax
+    {
+    if( options[index].has_arg == no )
+      {
+      error_ = "option '--"; error_ += options[index].name;
+      error_ += "' doesn't allow an argument";
+      return false;
+      }
+    if( options[index].has_arg == yes && !opt[len+3] )
+      {
+      error_ = "option '--"; error_ += options[index].name;
+      error_ += "' requires an argument";
+      return false;
+      }
+    data.back().argument = &opt[len+3];
+    return true;
+    }
+
+  if( options[index].has_arg == yes )
+    {
+    if( !arg || !arg[0] )
+      {
+      error_ = "option '--"; error_ += options[index].name;
+      error_ += "' requires an argument";
+      return false;
+      }
+    ++argind; data.back().argument = arg;
+    return true;
+    }
+
+  return true;
+  }
+
+
+bool Arg_parser::parse_short_option( const char * const opt, const char * const arg,
+                                     const Option options[], int & argind )
+  {
+  int cind = 1;			// character index in opt
+
+  while( cind > 0 )
+    {
+    int index = -1;
+    const unsigned char c = opt[cind];
+
+    if( c != 0 )
+      for( int i = 0; options[i].code; ++i )
+        if( c == options[i].code )
+          { index = i; break; }
+
+    if( index < 0 )
+      {
+      error_ = "invalid option -- '"; error_ += c; error_ += '\'';
+      return false;
+      }
+
+    data.push_back( Record( c ) );
+    if( opt[++cind] == 0 ) { ++argind; cind = 0; }	// opt finished
+
+    if( options[index].has_arg != no && cind > 0 && opt[cind] )
+      {
+      data.back().argument = &opt[cind]; ++argind; cind = 0;
+      }
+    else if( options[index].has_arg == yes )
+      {
+      if( !arg || !arg[0] )
+        {
+        error_ = "option requires an argument -- '"; error_ += c;
+        error_ += '\'';
+        return false;
+        }
+      data.back().argument = arg; ++argind; cind = 0;
+      }
+    }
+  return true;
+  }
+
+
+Arg_parser::Arg_parser( const int argc, const char * const argv[],
+                        const Option options[], const bool in_order )
+  {
+  if( argc < 2 || !argv || !options ) return;
+
+  std::vector< const char * > non_options;	// skipped non-options
+  int argind = 1;				// index in argv
+
+  while( argind < argc )
+    {
+    const unsigned char ch1 = argv[argind][0];
+    const unsigned char ch2 = ch1 ? argv[argind][1] : 0;
+
+    if( ch1 == '-' && ch2 )		// we found an option
+      {
+      const char * const opt = argv[argind];
+      const char * const arg = ( argind + 1 < argc ) ? argv[argind+1] : 0;
+      if( ch2 == '-' )
+        {
+        if( !argv[argind][2] ) { ++argind; break; }	// we found "--"
+        else if( !parse_long_option( opt, arg, options, argind ) ) break;
+        }
+      else if( !parse_short_option( opt, arg, options, argind ) ) break;
+      }
+    else
+      {
+      if( in_order ) data.push_back( Record( argv[argind++] ) );
+      else non_options.push_back( argv[argind++] );
+      }
+    }
+  if( !error_.empty() ) data.clear();
+  else
+    {
+    for( unsigned i = 0; i < non_options.size(); ++i )
+      data.push_back( Record( non_options[i] ) );
+    while( argind < argc )
+      data.push_back( Record( argv[argind++] ) );
+    }
+  }
+
+
+Arg_parser::Arg_parser( const char * const opt, const char * const arg,
+                        const Option options[] )
+  {
+  if( !opt || !opt[0] || !options ) return;
+
+  if( opt[0] == '-' && opt[1] )		// we found an option
+    {
+    int argind = 1;			// dummy
+    if( opt[1] == '-' )
+      { if( opt[2] ) parse_long_option( opt, arg, options, argind ); }
+    else
+      parse_short_option( opt, arg, options, argind );
+    if( !error_.empty() ) data.clear();
+    }
+  else data.push_back( Record( opt ) );
+  }
diff --git a/arg_parser.h b/arg_parser.h
new file mode 100644
index 0000000..5629b90
--- /dev/null
+++ b/arg_parser.h
@@ -0,0 +1,99 @@
+/* Arg_parser - POSIX/GNU command line argument parser. (C++ version)
+   Copyright (C) 2006-2021 Antonio Diaz Diaz.
+
+   This library is free software. Redistribution and use in source and
+   binary forms, with or without modification, are permitted provided
+   that the following conditions are met:
+
+   1. Redistributions of source code must retain the above copyright
+   notice, this list of conditions, and the following disclaimer.
+
+   2. Redistributions in binary form must reproduce the above copyright
+   notice, this list of conditions, and the following disclaimer in the
+   documentation and/or other materials provided with the distribution.
+
+   This library 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.
+*/
+
+/* Arg_parser reads the arguments in 'argv' and creates a number of
+   option codes, option arguments, and non-option arguments.
+
+   In case of error, 'error' returns a non-empty error message.
+
+   'options' is an array of 'struct Option' terminated by an element
+   containing a code which is zero. A null name means a short-only
+   option. A code value outside the unsigned char range means a
+   long-only option.
+
+   Arg_parser normally makes it appear as if all the option arguments
+   were specified before all the non-option arguments for the purposes
+   of parsing, even if the user of your program intermixed option and
+   non-option arguments. If you want the arguments in the exact order
+   the user typed them, call 'Arg_parser' with 'in_order' = true.
+
+   The argument '--' terminates all options; any following arguments are
+   treated as non-option arguments, even if they begin with a hyphen.
+
+   The syntax for optional option arguments is '-<short_option><argument>'
+   (without whitespace), or '--<long_option>=<argument>'.
+*/
+
+class Arg_parser
+  {
+public:
+  enum Has_arg { no, yes, maybe };
+
+  struct Option
+    {
+    int code;			// Short option letter or code ( code != 0 )
+    const char * name;		// Long option name (maybe null)
+    Has_arg has_arg;
+    };
+
+private:
+  struct Record
+    {
+    int code;
+    std::string argument;
+    explicit Record( const int c ) : code( c ) {}
+    explicit Record( const char * const arg ) : code( 0 ), argument( arg ) {}
+    };
+
+  const std::string empty_arg;
+  std::string error_;
+  std::vector< Record > data;
+
+  bool parse_long_option( const char * const opt, const char * const arg,
+                          const Option options[], int & argind );
+  bool parse_short_option( const char * const opt, const char * const arg,
+                           const Option options[], int & argind );
+
+public:
+  Arg_parser( const int argc, const char * const argv[],
+              const Option options[], const bool in_order = false );
+
+  // Restricted constructor. Parses a single token and argument (if any).
+  Arg_parser( const char * const opt, const char * const arg,
+              const Option options[] );
+
+  const std::string & error() const { return error_; }
+
+  // The number of arguments parsed. May be different from argc.
+  int arguments() const { return data.size(); }
+
+  /* If code( i ) is 0, argument( i ) is a non-option.
+     Else argument( i ) is the option's argument (or empty). */
+  int code( const int i ) const
+    {
+    if( i >= 0 && i < arguments() ) return data[i].code;
+    else return 0;
+    }
+
+  const std::string & argument( const int i ) const
+    {
+    if( i >= 0 && i < arguments() ) return data[i].argument;
+    else return empty_arg;
+    }
+  };
diff --git a/configure b/configure
new file mode 100755
index 0000000..48c7f81
--- /dev/null
+++ b/configure
@@ -0,0 +1,207 @@
+#! /bin/sh
+# configure script for Zutils - Utilities dealing with compressed files
+# Copyright (C) 2009-2021 Antonio Diaz Diaz.
+#
+# This configure script is free software: you have unlimited permission
+# to copy, distribute, and modify it.
+
+pkgname=zutils
+pkgversion=1.10
+srctrigger=doc/${pkgname}.texi
+
+# clear some things potentially inherited from environment.
+LC_ALL=C
+export LC_ALL
+srcdir=
+prefix=/usr/local
+exec_prefix='$(prefix)'
+bindir='$(exec_prefix)/bin'
+datarootdir='$(prefix)/share'
+infodir='$(datarootdir)/info'
+mandir='$(datarootdir)/man'
+sysconfdir='$(prefix)/etc'
+CXX=g++
+CPPFLAGS=
+CXXFLAGS='-Wall -W -O2'
+LDFLAGS=
+DIFF=diff
+GREP=grep
+
+# checking whether we are using GNU C++.
+/bin/sh -c "${CXX} --version" > /dev/null 2>&1 || { CXX=c++ ; CXXFLAGS=-O2 ; }
+
+# Loop over all args
+args=
+no_create=
+while [ $# != 0 ] ; do
+
+	# Get the first arg, and shuffle
+	option=$1 ; arg2=no
+	shift
+
+	# Add the argument quoted to args
+	if [ -z "${args}" ] ; then args="\"${option}\""
+	else args="${args} \"${option}\"" ; fi
+
+	# Split out the argument for options that take them
+	case ${option} in
+	*=*) optarg=`echo "${option}" | sed -e 's,^[^=]*=,,;s,/$,,'` ;;
+	esac
+
+	# Process the options
+	case ${option} in
+	--help | -h)
+		echo "Usage: $0 [OPTION]... [VAR=VALUE]..."
+		echo
+		echo "To assign makefile variables (e.g., CXX, CXXFLAGS...), specify them as"
+		echo "arguments to configure in the form VAR=VALUE."
+		echo
+		echo "Options and variables: [defaults in brackets]"
+		echo "  -h, --help            display this help and exit"
+		echo "  -V, --version         output version information and exit"
+		echo "  --srcdir=DIR          find the sources in DIR [. or ..]"
+		echo "  --prefix=DIR          install into DIR [${prefix}]"
+		echo "  --exec-prefix=DIR     base directory for arch-dependent files [${exec_prefix}]"
+		echo "  --bindir=DIR          user executables directory [${bindir}]"
+		echo "  --datarootdir=DIR     base directory for doc and data [${datarootdir}]"
+		echo "  --infodir=DIR         info files directory [${infodir}]"
+		echo "  --mandir=DIR          man pages directory [${mandir}]"
+		echo "  --sysconfdir=DIR      read-only single-machine data directory [${sysconfdir}]"
+		echo "  CXX=COMPILER          C++ compiler to use [${CXX}]"
+		echo "  CPPFLAGS=OPTIONS      command line options for the preprocessor [${CPPFLAGS}]"
+		echo "  CXXFLAGS=OPTIONS      command line options for the C++ compiler [${CXXFLAGS}]"
+		echo "  CXXFLAGS+=OPTIONS     append options to the current value of CXXFLAGS"
+		echo "  LDFLAGS=OPTIONS       command line options for the linker [${LDFLAGS}]"
+		echo "  DIFF=NAME             diff program to use with zdiff [${DIFF}]"
+		echo "  GREP=NAME             grep program to use with zgrep [${GREP}]"
+		echo
+		exit 0 ;;
+	--version | -V)
+		echo "Configure script for ${pkgname} version ${pkgversion}"
+		exit 0 ;;
+	--srcdir)            srcdir=$1 ; arg2=yes ;;
+	--prefix)            prefix=$1 ; arg2=yes ;;
+	--exec-prefix)  exec_prefix=$1 ; arg2=yes ;;
+	--bindir)            bindir=$1 ; arg2=yes ;;
+	--datarootdir)  datarootdir=$1 ; arg2=yes ;;
+	--infodir)          infodir=$1 ; arg2=yes ;;
+	--mandir)            mandir=$1 ; arg2=yes ;;
+	--sysconfdir)    sysconfdir=$1 ; arg2=yes ;;
+
+	--srcdir=*)            srcdir=${optarg} ;;
+	--prefix=*)            prefix=${optarg} ;;
+	--exec-prefix=*)  exec_prefix=${optarg} ;;
+	--bindir=*)            bindir=${optarg} ;;
+	--datarootdir=*)  datarootdir=${optarg} ;;
+	--infodir=*)          infodir=${optarg} ;;
+	--mandir=*)            mandir=${optarg} ;;
+	--sysconfdir=*)    sysconfdir=${optarg} ;;
+	--no-create)              no_create=yes ;;
+
+	CXX=*)            CXX=${optarg} ;;
+	CPPFLAGS=*)  CPPFLAGS=${optarg} ;;
+	CXXFLAGS=*)  CXXFLAGS=${optarg} ;;
+	CXXFLAGS+=*) CXXFLAGS="${CXXFLAGS} ${optarg}" ;;
+	LDFLAGS=*)    LDFLAGS=${optarg} ;;
+	DIFF=*)          DIFF=${optarg} ;;
+	GREP=*)          GREP=${optarg} ;;
+
+	--*)
+		echo "configure: WARNING: unrecognized option: '${option}'" 1>&2 ;;
+	*=* | *-*-*) ;;
+	*)
+		echo "configure: unrecognized option: '${option}'" 1>&2
+		echo "Try 'configure --help' for more information." 1>&2
+		exit 1 ;;
+	esac
+
+	# Check if the option took a separate argument
+	if [ "${arg2}" = yes ] ; then
+		if [ $# != 0 ] ; then args="${args} \"$1\"" ; shift
+		else echo "configure: Missing argument to '${option}'" 1>&2
+			exit 1
+		fi
+	fi
+done
+
+# Find the source files, if location was not specified.
+srcdirtext=
+if [ -z "${srcdir}" ] ; then
+	srcdirtext="or . or .." ; srcdir=.
+	if [ ! -r "${srcdir}/${srctrigger}" ] ; then srcdir=.. ; fi
+	if [ ! -r "${srcdir}/${srctrigger}" ] ; then
+		## the sed command below emulates the dirname command
+		srcdir=`echo "$0" | sed -e 's,[^/]*$,,;s,/$,,;s,^$,.,'`
+	fi
+fi
+
+if [ ! -r "${srcdir}/${srctrigger}" ] ; then
+	echo "configure: Can't find sources in ${srcdir} ${srcdirtext}" 1>&2
+	echo "configure: (At least ${srctrigger} is missing)." 1>&2
+	exit 1
+fi
+
+# Set srcdir to . if that's what it is.
+if [ "`pwd`" = "`cd "${srcdir}" ; pwd`" ] ; then srcdir=. ; fi
+
+echo
+if [ -z "${no_create}" ] ; then
+	echo "creating config.status"
+	rm -f config.status
+	cat > config.status << EOF
+#! /bin/sh
+# This file was generated automatically by configure. Don't edit.
+# Run this file to recreate the current configuration.
+#
+# This script is free software: you have unlimited permission
+# to copy, distribute, and modify it.
+
+exec /bin/sh $0 ${args} --no-create
+EOF
+	chmod +x config.status
+fi
+
+echo "creating Makefile"
+echo "VPATH = ${srcdir}"
+echo "prefix = ${prefix}"
+echo "exec_prefix = ${exec_prefix}"
+echo "bindir = ${bindir}"
+echo "datarootdir = ${datarootdir}"
+echo "infodir = ${infodir}"
+echo "mandir = ${mandir}"
+echo "sysconfdir = ${sysconfdir}"
+echo "CXX = ${CXX}"
+echo "CPPFLAGS = ${CPPFLAGS}"
+echo "CXXFLAGS = ${CXXFLAGS}"
+echo "LDFLAGS = ${LDFLAGS}"
+echo "DIFF = ${DIFF}"
+echo "GREP = ${GREP}"
+rm -f Makefile
+cat > Makefile << EOF
+# Makefile for Zutils - Utilities dealing with compressed files
+# Copyright (C) 2009-2021 Antonio Diaz Diaz.
+# This file was generated automatically by configure. Don't edit.
+#
+# This Makefile is free software: you have unlimited permission
+# to copy, distribute, and modify it.
+
+pkgname = ${pkgname}
+pkgversion = ${pkgversion}
+VPATH = ${srcdir}
+prefix = ${prefix}
+exec_prefix = ${exec_prefix}
+bindir = ${bindir}
+datarootdir = ${datarootdir}
+infodir = ${infodir}
+mandir = ${mandir}
+sysconfdir = ${sysconfdir}
+CXX = ${CXX}
+CPPFLAGS = ${CPPFLAGS}
+CXXFLAGS = ${CXXFLAGS}
+LDFLAGS = ${LDFLAGS}
+DIFF = ${DIFF}
+GREP = ${GREP}
+EOF
+cat "${srcdir}/Makefile.in" >> Makefile
+
+echo "OK. Now you can run make."
diff --git a/doc/zcat.1 b/doc/zcat.1
new file mode 100644
index 0000000..fbaf821
--- /dev/null
+++ b/doc/zcat.1
@@ -0,0 +1,103 @@
+.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.47.16.
+.TH ZCAT "1" "January 2021" "zutils 1.10" "User Commands"
+.SH NAME
+zcat \- decompress and concatenate files to standard output
+.SH SYNOPSIS
+.B zcat
+[\fI\,options\/\fR] [\fI\,files\/\fR]
+.SH DESCRIPTION
+zcat copies each file argument to standard output in sequence. If any
+file given is compressed, its decompressed content is copied. If a file
+given does not exist, and its name does not end with one of the known
+extensions, zcat tries the compressed file names corresponding to the
+formats supported. If a file fails to decompress, zcat continues copying the
+rest of the files.
+.PP
+If a file is specified as '\-', data are read from standard input,
+decompressed if needed, and sent to standard output. Data read from
+standard input must be of the same type; all uncompressed or all in the
+same compressed format.
+.PP
+If no files are specified, recursive searches examine the current
+working directory, and nonrecursive searches read standard input.
+.PP
+The formats supported are bzip2, gzip, lzip, and xz.
+.PP
+Exit status is 0 if no errors occurred, 1 otherwise.
+.SH OPTIONS
+.TP
+\fB\-h\fR, \fB\-\-help\fR
+display this help and exit
+.TP
+\fB\-V\fR, \fB\-\-version\fR
+output version information and exit
+.TP
+\fB\-A\fR, \fB\-\-show\-all\fR
+equivalent to '\-vET'
+.TP
+\fB\-b\fR, \fB\-\-number\-nonblank\fR
+number nonblank output lines
+.TP
+\fB\-e\fR
+equivalent to '\-vE'
+.TP
+\fB\-E\fR, \fB\-\-show\-ends\fR
+display '$' at end of each line
+.TP
+\fB\-M\fR, \fB\-\-format=\fR<list>
+process only the formats in <list>
+.TP
+\fB\-n\fR, \fB\-\-number\fR
+number all output lines
+.TP
+\fB\-N\fR, \fB\-\-no\-rcfile\fR
+don't read runtime configuration file
+.TP
+\fB\-O\fR, \fB\-\-force\-format=\fR<fmt>
+force the format given (bz2, gz, lz, xz)
+.TP
+\fB\-q\fR, \fB\-\-quiet\fR
+suppress all messages
+.TP
+\fB\-r\fR, \fB\-\-recursive\fR
+operate recursively on directories
+.TP
+\fB\-R\fR, \fB\-\-dereference\-recursive\fR
+recursively follow symbolic links
+.TP
+\fB\-s\fR, \fB\-\-squeeze\-blank\fR
+never more than one single blank line
+.TP
+\fB\-t\fR
+equivalent to '\-vT'
+.TP
+\fB\-T\fR, \fB\-\-show\-tabs\fR
+display TAB characters as '^I'
+.TP
+\fB\-v\fR, \fB\-\-show\-nonprinting\fR
+use '^' and 'M\-' notation, except for LF and TAB
+.TP
+\fB\-\-verbose\fR
+verbose mode (show error messages)
+.TP
+\fB\-\-bz2=\fR<command>
+set compressor and options for bzip2 format
+.TP
+\fB\-\-gz=\fR<command>
+set compressor and options for gzip format
+.TP
+\fB\-\-lz=\fR<command>
+set compressor and options for lzip format
+.TP
+\fB\-\-xz=\fR<command>
+set compressor and options for xz format
+.SH "REPORTING BUGS"
+Report bugs to zutils\-bug@nongnu.org
+.br
+Zutils home page: http://www.nongnu.org/zutils/zutils.html
+.SH COPYRIGHT
+Copyright \(co 2021 Antonio Diaz Diaz.
+License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl.html>
+.br
+This is free software: you are free to change and redistribute it.
+There is NO WARRANTY, to the extent permitted by law.
diff --git a/doc/zcmp.1 b/doc/zcmp.1
new file mode 100644
index 0000000..5ae5231
--- /dev/null
+++ b/doc/zcmp.1
@@ -0,0 +1,90 @@
+.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.47.16.
+.TH ZCMP "1" "January 2021" "zutils 1.10" "User Commands"
+.SH NAME
+zcmp \- decompress and compare two files byte by byte
+.SH SYNOPSIS
+.B zcmp
+[\fI\,options\/\fR] \fI\,file1 \/\fR[\fI\,file2\/\fR]
+.SH DESCRIPTION
+zcmp compares two files and, if they differ, writes to standard output the
+first byte and line number where they differ. Bytes and lines are numbered
+starting with 1. A hyphen '\-' used as a file argument means standard input.
+If any file given is compressed, its decompressed content is used. Compressed
+files are decompressed on the fly; no temporary files are created.
+.PP
+The formats supported are bzip2, gzip, lzip, and xz.
+.PP
+zcmp compares file1 to file2. The standard input is used only if file1 or
+file2 refers to standard input. If file2 is omitted zcmp tries the
+following:
+.IP
+\- If file1 is compressed, compares its decompressed contents with
+the corresponding uncompressed file (the name of file1 with the
+extension removed).
+.IP
+\- If file1 is uncompressed, compares it with the decompressed
+contents of file1.[lz|bz2|gz|xz] (the first one that is found).
+.PP
+Exit status is 0 if inputs are identical, 1 if different, 2 if trouble.
+.SH OPTIONS
+.TP
+\fB\-h\fR, \fB\-\-help\fR
+display this help and exit
+.TP
+\fB\-V\fR, \fB\-\-version\fR
+output version information and exit
+.TP
+\fB\-b\fR, \fB\-\-print\-bytes\fR
+print differing bytes
+.TP
+\fB\-i\fR, \fB\-\-ignore\-initial=\fR<n>[:<n2>]
+ignore differences in the first <n> bytes
+.TP
+\fB\-l\fR, \fB\-\-list\fR
+list position, value of all differing bytes
+.TP
+\fB\-M\fR, \fB\-\-format=\fR<list>
+process only the formats in <list>
+.TP
+\fB\-n\fR, \fB\-\-bytes=\fR<n>
+compare at most <n> bytes
+.TP
+\fB\-N\fR, \fB\-\-no\-rcfile\fR
+don't read runtime configuration file
+.TP
+\fB\-O\fR, \fB\-\-force\-format\fR=\fI\,[\/\fR<f1>][,<f2>]
+force the formats given (bz2, gz, lz, xz)
+.TP
+\fB\-q\fR, \fB\-\-quiet\fR
+suppress all messages
+.TP
+\fB\-s\fR, \fB\-\-silent\fR
+(same as \fB\-\-quiet\fR)
+.TP
+\fB\-v\fR, \fB\-\-verbose\fR
+verbose mode (same as \fB\-\-list\fR)
+.TP
+\fB\-\-bz2=\fR<command>
+set compressor and options for bzip2 format
+.TP
+\fB\-\-gz=\fR<command>
+set compressor and options for gzip format
+.TP
+\fB\-\-lz=\fR<command>
+set compressor and options for lzip format
+.TP
+\fB\-\-xz=\fR<command>
+set compressor and options for xz format
+.PP
+Numbers may be followed by a multiplier: k = kB = 10^3 = 1000,
+Ki = KiB = 2^10 = 1024, M = 10^6, Mi = 2^20, G = 10^9, Gi = 2^30, etc...
+.SH "REPORTING BUGS"
+Report bugs to zutils\-bug@nongnu.org
+.br
+Zutils home page: http://www.nongnu.org/zutils/zutils.html
+.SH COPYRIGHT
+Copyright \(co 2021 Antonio Diaz Diaz.
+License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl.html>
+.br
+This is free software: you are free to change and redistribute it.
+There is NO WARRANTY, to the extent permitted by law.
diff --git a/doc/zdiff.1 b/doc/zdiff.1
new file mode 100644
index 0000000..65a34b7
--- /dev/null
+++ b/doc/zdiff.1
@@ -0,0 +1,121 @@
+.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.47.16.
+.TH ZDIFF "1" "January 2021" "zutils 1.10" "User Commands"
+.SH NAME
+zdiff \- decompress and compare two files line by line
+.SH SYNOPSIS
+.B zdiff
+[\fI\,options\/\fR] \fI\,file1 \/\fR[\fI\,file2\/\fR]
+.SH DESCRIPTION
+zdiff compares two files and, if they differ, writes to standard output the
+differences line by line. A hyphen '\-' used as a file argument means standard
+input. If any file given is compressed, its decompressed content is used.
+zdiff is a front end to the program diff and has the limitation that messages
+from diff refer to temporary file names instead of those specified.
+.PP
+The formats supported are bzip2, gzip, lzip, and xz.
+.PP
+zdiff compares file1 to file2. The standard input is used only if file1 or
+file2 refers to standard input. If file2 is omitted zdiff tries the
+following:
+.IP
+\- If file1 is compressed, compares its decompressed contents with
+the corresponding uncompressed file (the name of file1 with the
+extension removed).
+.IP
+\- If file1 is uncompressed, compares it with the decompressed
+contents of file1.[lz|bz2|gz|xz] (the first one that is found).
+.PP
+Exit status is 0 if inputs are identical, 1 if different, 2 if trouble.
+Some options only work if the diff program used supports them.
+.SH OPTIONS
+.TP
+\fB\-h\fR, \fB\-\-help\fR
+display this help and exit
+.TP
+\fB\-V\fR, \fB\-\-version\fR
+output version information and exit
+.TP
+\fB\-a\fR, \fB\-\-text\fR
+treat all files as text
+.TP
+\fB\-b\fR, \fB\-\-ignore\-space\-change\fR
+ignore changes in the amount of white space
+.TP
+\fB\-B\fR, \fB\-\-ignore\-blank\-lines\fR
+ignore changes whose lines are all blank
+.TP
+\fB\-c\fR
+use the context output format
+.TP
+\fB\-C\fR, \fB\-\-context=\fR<n>
+same as \fB\-c\fR but use <n> lines of context
+.TP
+\fB\-d\fR, \fB\-\-minimal\fR
+try hard to find a smaller set of changes
+.TP
+\fB\-E\fR, \fB\-\-ignore\-tab\-expansion\fR
+ignore changes due to tab expansion
+.TP
+\fB\-i\fR, \fB\-\-ignore\-case\fR
+ignore case differences in file contents
+.TP
+\fB\-M\fR, \fB\-\-format=\fR<list>
+process only the formats in <list>
+.TP
+\fB\-N\fR, \fB\-\-no\-rcfile\fR
+don't read runtime configuration file
+.TP
+\fB\-O\fR, \fB\-\-force\-format\fR=\fI\,[\/\fR<f1>][,<f2>]
+force the formats given (bz2, gz, lz, xz)
+.TP
+\fB\-p\fR, \fB\-\-show\-c\-function\fR
+show which C function each change is in
+.TP
+\fB\-q\fR, \fB\-\-brief\fR
+output only whether files differ
+.TP
+\fB\-s\fR, \fB\-\-report\-identical\-files\fR
+report when two files are identical
+.TP
+\fB\-t\fR, \fB\-\-expand\-tabs\fR
+expand tabs to spaces in output
+.TP
+\fB\-T\fR, \fB\-\-initial\-tab\fR
+make tabs line up by prepending a tab
+.TP
+\fB\-u\fR
+use the unified output format
+.TP
+\fB\-U\fR, \fB\-\-unified=\fR<n>
+same as \fB\-u\fR but use <n> lines of context
+.TP
+\fB\-w\fR, \fB\-\-ignore\-all\-space\fR
+ignore all white space
+.TP
+\fB\-W\fR, \fB\-\-width=\fR<n>
+output at most <n> print columns
+.TP
+\fB\-y\fR, \fB\-\-side\-by\-side\fR
+output in two columns
+.TP
+\fB\-\-bz2=\fR<command>
+set compressor and options for bzip2 format
+.TP
+\fB\-\-gz=\fR<command>
+set compressor and options for gzip format
+.TP
+\fB\-\-lz=\fR<command>
+set compressor and options for lzip format
+.TP
+\fB\-\-xz=\fR<command>
+set compressor and options for xz format
+.SH "REPORTING BUGS"
+Report bugs to zutils\-bug@nongnu.org
+.br
+Zutils home page: http://www.nongnu.org/zutils/zutils.html
+.SH COPYRIGHT
+Copyright \(co 2021 Antonio Diaz Diaz.
+License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl.html>
+.br
+This is free software: you are free to change and redistribute it.
+There is NO WARRANTY, to the extent permitted by law.
diff --git a/doc/zgrep.1 b/doc/zgrep.1
new file mode 100644
index 0000000..69ed0cd
--- /dev/null
+++ b/doc/zgrep.1
@@ -0,0 +1,153 @@
+.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.47.16.
+.TH ZGREP "1" "January 2021" "zutils 1.10" "User Commands"
+.SH NAME
+zgrep \- search compressed files for a regular expression
+.SH SYNOPSIS
+.B zgrep
+[\fI\,options\/\fR] \fI\,<pattern> \/\fR[\fI\,files\/\fR]
+.SH DESCRIPTION
+zgrep is a front end to the program grep that allows transparent search
+on any combination of compressed and uncompressed files. If any file
+given is compressed, its decompressed content is used. If a file given
+does not exist, and its name does not end with one of the known
+extensions, zgrep tries the compressed file names corresponding to the
+formats supported. If a file fails to decompress, zgrep continues
+searching the rest of the files.
+.PP
+If a file is specified as '\-', data are read from standard input,
+decompressed if needed, and fed to grep. Data read from standard input
+must be of the same type; all uncompressed or all in the same
+compressed format.
+.PP
+If no files are specified, recursive searches examine the current
+working directory, and nonrecursive searches read standard input.
+.PP
+The formats supported are bzip2, gzip, lzip, and xz.
+.PP
+Exit status is 0 if match, 1 if no match, 2 if trouble.
+Some options only work if the grep program used supports them.
+.SH OPTIONS
+.TP
+\fB\-\-help\fR
+display this help and exit
+.TP
+\fB\-V\fR, \fB\-\-version\fR
+output version information and exit
+.TP
+\fB\-a\fR, \fB\-\-text\fR
+treat all files as text
+.TP
+\fB\-A\fR, \fB\-\-after\-context=\fR<n>
+print <n> lines of trailing context
+.TP
+\fB\-b\fR, \fB\-\-byte\-offset\fR
+print the byte offset of each line
+.TP
+\fB\-B\fR, \fB\-\-before\-context=\fR<n>
+print <n> lines of leading context
+.TP
+\fB\-c\fR, \fB\-\-count\fR
+only print a count of matching lines per file
+.TP
+\fB\-C\fR, \fB\-\-context=\fR<n>
+print <n> lines of output context
+.TP
+\fB\-\-color[=\fR<when>]
+show matched strings in color
+.TP
+\fB\-e\fR, \fB\-\-regexp=\fR<pattern>
+use <pattern> as the pattern to match
+.TP
+\fB\-E\fR, \fB\-\-extended\-regexp\fR
+<pattern> is an extended regular expression
+.TP
+\fB\-f\fR, \fB\-\-file=\fR<file>
+obtain patterns from <file>
+.TP
+\fB\-F\fR, \fB\-\-fixed\-strings\fR
+<pattern> is a set of newline\-separated strings
+.TP
+\fB\-h\fR, \fB\-\-no\-filename\fR
+suppress the prefixing filename on output
+.TP
+\fB\-H\fR, \fB\-\-with\-filename\fR
+print the filename for each match
+.TP
+\fB\-i\fR, \fB\-\-ignore\-case\fR
+ignore case distinctions
+.TP
+\fB\-I\fR
+ignore binary files
+.TP
+\fB\-l\fR, \fB\-\-files\-with\-matches\fR
+only print names of files containing matches
+.TP
+\fB\-L\fR, \fB\-\-files\-without\-match\fR
+only print names of files containing no matches
+.TP
+\fB\-m\fR, \fB\-\-max\-count=\fR<n>
+stop after <n> matches
+.TP
+\fB\-M\fR, \fB\-\-format=\fR<list>
+process only the formats in <list>
+.TP
+\fB\-n\fR, \fB\-\-line\-number\fR
+print the line number of each line
+.TP
+\fB\-N\fR, \fB\-\-no\-rcfile\fR
+don't read runtime configuration file
+.TP
+\fB\-o\fR, \fB\-\-only\-matching\fR
+show only the part of a line matching <pattern>
+.TP
+\fB\-O\fR, \fB\-\-force\-format=\fR<fmt>
+force the format given (bz2, gz, lz, xz)
+.TP
+\fB\-q\fR, \fB\-\-quiet\fR
+suppress all messages
+.TP
+\fB\-r\fR, \fB\-\-recursive\fR
+operate recursively on directories
+.TP
+\fB\-R\fR, \fB\-\-dereference\-recursive\fR
+recursively follow symbolic links
+.TP
+\fB\-s\fR, \fB\-\-no\-messages\fR
+suppress error messages
+.TP
+\fB\-v\fR, \fB\-\-invert\-match\fR
+select non\-matching lines
+.TP
+\fB\-\-verbose\fR
+verbose mode (show error messages)
+.TP
+\fB\-w\fR, \fB\-\-word\-regexp\fR
+match only whole words
+.TP
+\fB\-x\fR, \fB\-\-line\-regexp\fR
+match only whole lines
+.TP
+\fB\-\-bz2=\fR<command>
+set compressor and options for bzip2 format
+.TP
+\fB\-\-gz=\fR<command>
+set compressor and options for gzip format
+.TP
+\fB\-\-lz=\fR<command>
+set compressor and options for lzip format
+.TP
+\fB\-\-xz=\fR<command>
+set compressor and options for xz format
+.PP
+Numbers may be followed by a multiplier: k = kB = 10^3 = 1000,
+Ki = KiB = 2^10 = 1024, M = 10^6, Mi = 2^20, G = 10^9, Gi = 2^30, etc...
+.SH "REPORTING BUGS"
+Report bugs to zutils\-bug@nongnu.org
+.br
+Zutils home page: http://www.nongnu.org/zutils/zutils.html
+.SH COPYRIGHT
+Copyright \(co 2021 Antonio Diaz Diaz.
+License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl.html>
+.br
+This is free software: you are free to change and redistribute it.
+There is NO WARRANTY, to the extent permitted by law.
diff --git a/doc/ztest.1 b/doc/ztest.1
new file mode 100644
index 0000000..45cda27
--- /dev/null
+++ b/doc/ztest.1
@@ -0,0 +1,82 @@
+.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.47.16.
+.TH ZTEST "1" "January 2021" "zutils 1.10" "User Commands"
+.SH NAME
+ztest \- verify the integrity of compressed files
+.SH SYNOPSIS
+.B ztest
+[\fI\,options\/\fR] [\fI\,files\/\fR]
+.SH DESCRIPTION
+ztest verifies the integrity of the compressed files specified.
+Uncompressed files are ignored. If a file is specified as '\-', the
+integrity of compressed data read from standard input is verified. Data
+read from standard input must be all in the same compressed format. If
+a file fails to decompress, does not exist, can't be opened, or is a
+terminal, ztest continues verifying the rest of the files. A final
+diagnostic is shown at verbosity level 1 or higher if any file fails the
+test when testing multiple files.
+.PP
+If no files are specified, recursive searches examine the current
+working directory, and nonrecursive searches read standard input.
+.PP
+The formats supported are bzip2, gzip, lzip, and xz.
+.PP
+Note that error detection in the xz format is broken. First, some xz
+files lack integrity information. Second, not all xz decompressors can
+verify the integrity of all xz files. Third, section 2.1.1.2 'Stream
+Flags' of the xz format specification allows xz decompressors to produce
+garbage output without issuing any warning. Therefore, xz files can't
+always be verified as reliably as files in the other formats can.
+.PP
+Exit status is 0 if all compressed files verify OK, 1 if environmental
+problems (file not found, invalid flags, I/O errors, etc), 2 if any
+compressed file is corrupt or invalid.
+.SH OPTIONS
+.TP
+\fB\-h\fR, \fB\-\-help\fR
+display this help and exit
+.TP
+\fB\-V\fR, \fB\-\-version\fR
+output version information and exit
+.TP
+\fB\-M\fR, \fB\-\-format=\fR<list>
+process only the formats in <list>
+.TP
+\fB\-N\fR, \fB\-\-no\-rcfile\fR
+don't read runtime configuration file
+.TP
+\fB\-O\fR, \fB\-\-force\-format=\fR<fmt>
+force the format given (bz2, gz, lz, xz)
+.TP
+\fB\-q\fR, \fB\-\-quiet\fR
+suppress all messages
+.TP
+\fB\-r\fR, \fB\-\-recursive\fR
+operate recursively on directories
+.TP
+\fB\-R\fR, \fB\-\-dereference\-recursive\fR
+recursively follow symbolic links
+.TP
+\fB\-v\fR, \fB\-\-verbose\fR
+be verbose (a 2nd \fB\-v\fR gives more)
+.TP
+\fB\-\-bz2=\fR<command>
+set compressor and options for bzip2 format
+.TP
+\fB\-\-gz=\fR<command>
+set compressor and options for gzip format
+.TP
+\fB\-\-lz=\fR<command>
+set compressor and options for lzip format
+.TP
+\fB\-\-xz=\fR<command>
+set compressor and options for xz format
+.SH "REPORTING BUGS"
+Report bugs to zutils\-bug@nongnu.org
+.br
+Zutils home page: http://www.nongnu.org/zutils/zutils.html
+.SH COPYRIGHT
+Copyright \(co 2021 Antonio Diaz Diaz.
+License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl.html>
+.br
+This is free software: you are free to change and redistribute it.
+There is NO WARRANTY, to the extent permitted by law.
diff --git a/doc/zupdate.1 b/doc/zupdate.1
new file mode 100644
index 0000000..dcd3d24
--- /dev/null
+++ b/doc/zupdate.1
@@ -0,0 +1,91 @@
+.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.47.16.
+.TH ZUPDATE "1" "January 2021" "zutils 1.10" "User Commands"
+.SH NAME
+zupdate \- recompress bzip2, gzip, xz files to lzip format
+.SH SYNOPSIS
+.B zupdate
+[\fI\,options\/\fR] [\fI\,files\/\fR]
+.SH DESCRIPTION
+zupdate recompresses files from bzip2, gzip, and xz formats to lzip
+format. Each original is compared with the new file and then deleted.
+Only regular files with standard file name extensions are recompressed,
+other files are ignored. Compressed files are decompressed and then
+recompressed on the fly; no temporary files are created. The lzip format
+is chosen as destination because it is the most appropriate for
+long\-term data archiving.
+.PP
+If no files are specified, recursive searches examine the current
+working directory, and nonrecursive searches do nothing.
+.PP
+If the lzip compressed version of a file already exists, the file is
+skipped unless the option '\-\-force' is given. In this case, if the
+comparison with the existing lzip version fails, an error is returned
+and the original file is not deleted. The operation of zupdate is meant
+to be safe and not cause any data loss. Therefore, existing lzip
+compressed files are never overwritten nor deleted.
+.PP
+The names of the original files must have one of the following extensions:
+\&'.bz2', '.gz', or '.xz', which are recompressed to '.lz';
+\&'.tbz', '.tbz2', '.tgz', or '.txz', which are recompressed to '.tlz'.
+.PP
+Exit status is 0 if all the compressed files were successfully recompressed
+(if needed), compared, and deleted (if requested). Non\-zero otherwise.
+.SH OPTIONS
+.TP
+\fB\-h\fR, \fB\-\-help\fR
+display this help and exit
+.TP
+\fB\-V\fR, \fB\-\-version\fR
+output version information and exit
+.TP
+\fB\-f\fR, \fB\-\-force\fR
+don't skip a file even if the .lz exists
+.TP
+\fB\-k\fR, \fB\-\-keep\fR
+keep (don't delete) input files
+.TP
+\fB\-l\fR, \fB\-\-lzip\-verbose\fR
+pass one option \fB\-v\fR to the lzip compressor
+.TP
+\fB\-M\fR, \fB\-\-format=\fR<list>
+process only the formats in <list>
+.TP
+\fB\-N\fR, \fB\-\-no\-rcfile\fR
+don't read runtime configuration file
+.TP
+\fB\-q\fR, \fB\-\-quiet\fR
+suppress all messages
+.TP
+\fB\-r\fR, \fB\-\-recursive\fR
+operate recursively on directories
+.TP
+\fB\-R\fR, \fB\-\-dereference\-recursive\fR
+recursively follow symbolic links
+.TP
+\fB\-v\fR, \fB\-\-verbose\fR
+be verbose (a 2nd \fB\-v\fR gives more)
+.TP
+\fB\-0\fR .. \fB\-9\fR
+set compression level [default 9]
+.TP
+\fB\-\-bz2=\fR<command>
+set compressor and options for bzip2 format
+.TP
+\fB\-\-gz=\fR<command>
+set compressor and options for gzip format
+.TP
+\fB\-\-lz=\fR<command>
+set compressor and options for lzip format
+.TP
+\fB\-\-xz=\fR<command>
+set compressor and options for xz format
+.SH "REPORTING BUGS"
+Report bugs to zutils\-bug@nongnu.org
+.br
+Zutils home page: http://www.nongnu.org/zutils/zutils.html
+.SH COPYRIGHT
+Copyright \(co 2021 Antonio Diaz Diaz.
+License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl.html>
+.br
+This is free software: you are free to change and redistribute it.
+There is NO WARRANTY, to the extent permitted by law.
diff --git a/doc/zutils.info b/doc/zutils.info
new file mode 100644
index 0000000..854100f
--- /dev/null
+++ b/doc/zutils.info
@@ -0,0 +1,838 @@
+This is zutils.info, produced by makeinfo version 4.13+ from zutils.texi.
+
+INFO-DIR-SECTION Data Compression
+START-INFO-DIR-ENTRY
+* Zutils: (zutils).             Utilities dealing with compressed files
+END-INFO-DIR-ENTRY
+
+
+File: zutils.info,  Node: Top,  Next: Introduction,  Up: (dir)
+
+Zutils Manual
+*************
+
+This manual is for Zutils (version 1.10, 5 January 2021).
+
+* Menu:
+
+* Introduction::        Purpose and features of zutils
+* Common options::      Options common to all utilities
+* The zutilsrc file::   The zutils configuration file
+* Zcat::                Concatenating compressed files
+* Zcmp::                Comparing compressed files byte by byte
+* Zdiff::               Comparing compressed files line by line
+* Zgrep::               Searching inside compressed files
+* Ztest::               Testing the integrity of compressed files
+* Zupdate::             Recompressing files to lzip format
+* Problems::            Reporting bugs
+* Concept index::       Index of concepts
+
+
+   Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+   This manual is free documentation: you have unlimited permission to copy,
+distribute, and modify it.
+
+
+File: zutils.info,  Node: Introduction,  Next: Common options,  Prev: Top,  Up: Top
+
+1 Introduction
+**************
+
+Zutils is a collection of utilities able to process any combination of
+compressed and uncompressed files transparently. If any file given,
+including standard input, is compressed, its decompressed content is used.
+Compressed files are decompressed on the fly; no temporary files are
+created.
+
+   These utilities are not wrapper scripts but safer and more efficient C++
+programs. In particular the option '--recursive' is very efficient in those
+utilities supporting it.
+
+The utilities provided are zcat, zcmp, zdiff, zgrep, ztest, and zupdate.
+The formats supported are bzip2, gzip, lzip, and xz.
+Zutils uses external compressors. The compressor to be used for each format
+is configurable at runtime.
+
+   zcat, zcmp, zdiff, and zgrep are improved replacements for the shell
+scripts provided by GNU gzip. ztest is unique to zutils. zupdate is similar
+to gzip's znew.
+
+   NOTE: Bzip2 and lzip provide well-defined values of exit status, which
+makes them safe to use with zutils. Gzip and xz may return ambiguous warning
+values, making them less reliable back ends for zutils. *Note
+compressor-requirements::.
+
+   FORMAT NOTE 1: The option '--format' allows the processing of a subset
+of formats in recursive mode and when trying compressed file names:
+'zgrep foo -r --format=bz2,lz somedir somefile.tar'.
+
+   FORMAT NOTE 2: If the option '--force-format' is given, the files are
+passed to the corresponding decompressor without verifying their format,
+allowing for example the processing of compress'd (.Z) files with gzip:
+'zcmp --force-format=gz file.Z file.lz'.
+
+   LANGUAGE NOTE: Uncompressed = not compressed = plain data; it may never
+have been compressed. Decompressed is used to refer to data which have
+undergone the process of decompression.
+
+
+   Numbers given as arguments to options (positions, sizes) may be followed
+by a multiplier and an optional 'B' for "byte".
+
+   Table of SI and binary prefixes (unit multipliers):
+
+Prefix   Value                     |   Prefix   Value
+k        kilobyte  (10^3 = 1000)   |   Ki       kibibyte (2^10 = 1024)
+M        megabyte  (10^6)          |   Mi       mebibyte (2^20)
+G        gigabyte  (10^9)          |   Gi       gibibyte (2^30)
+T        terabyte  (10^12)         |   Ti       tebibyte (2^40)
+P        petabyte  (10^15)         |   Pi       pebibyte (2^50)
+E        exabyte   (10^18)         |   Ei       exbibyte (2^60)
+Z        zettabyte (10^21)         |   Zi       zebibyte (2^70)
+Y        yottabyte (10^24)         |   Yi       yobibyte (2^80)
+
+
+File: zutils.info,  Node: Common options,  Next: The zutilsrc file,  Prev: Introduction,  Up: Top
+
+2 Common options
+****************
+
+The following options: are available in all the utilities. Rather than
+writing identical descriptions for each of the programs, they are described
+here. *Note Argument syntax: (arg_parser)Argument syntax.
+
+'-h'
+'--help'
+     Print an informative help message describing the options and exit.
+     zgrep only supports the '--help' form of this option.
+
+'-V'
+'--version'
+     Print the version number on the standard output and exit. This version
+     number should be included in all bug reports.
+
+'-M FORMAT_LIST'
+'--format=FORMAT_LIST'
+     Process only the formats listed in the comma-separated FORMAT_LIST.
+     Valid formats are 'bz2', 'gz', 'lz', 'xz', and 'un' for
+     'uncompressed', meaning "any file name without a known extension".
+     This option excludes files based on extension, instead of format,
+     because it is more efficient. The exclusion only applies to names
+     generated automatically (for example when adding extensions to a file
+     name or when operating recursively on directories). Files given in the
+     command line are always processed.
+
+     Each format in FORMAT_LIST enables file names with the following
+     extensions:
+
+     bz2   enables   .bz2 .tbz .tbz2
+     gz    enables   .gz .tgz
+     lz    enables   .lz .tlz
+     xz    enables   .xz .txz
+     un    enables   any other file name
+
+'-N'
+'--no-rcfile'
+     Don't read the runtime configuration file 'zutilsrc'.
+
+'--bz2=COMMAND'
+'--gz=COMMAND'
+'--lz=COMMAND'
+'--xz=COMMAND'
+     Set program to be used as (de)compressor for the corresponding format.
+     COMMAND may include arguments. For example '--lz='plzip --threads=2''.
+     The program set with '--lz' is used for both compression and
+     decompression. The other three are used only for decompression. The
+     name of the program can't begin with '-'. These options override the
+     values set in 'zutilsrc'. The compression program used must meet three
+     requirements:
+
+       1. When called with the option '-d', it must read compressed data
+          from the standard input and produce decompressed data on the
+          standard output.
+
+       2. If the option '-q' is passed to zutils, the compression program
+          must also accept it.
+
+       3. It must return 0 if no errors occurred, and a non-zero value
+          otherwise.
+
+
+
+File: zutils.info,  Node: The zutilsrc file,  Next: Zcat,  Prev: Common options,  Up: Top
+
+3 The zutils configuration file 'zutilsrc'
+******************************************
+
+'zutilsrc' is the runtime configuration file for zutils. In it you may
+define the compressor name and options to be used for each format.
+'zutilsrc' is optional; you don't need to install it in order to run zutils.
+
+   The compressors specified in the command line override those specified
+in 'zutilsrc'.
+
+   You may copy the system 'zutilsrc' file '${sysconfdir}/zutilsrc' to
+'$HOME/.zutilsrc' and customize these options as you like. The file syntax
+is fairly obvious (and there are further instructions in it):
+
+  1. Any line beginning with '#' is a comment line.
+
+  2. Each non-comment line defines the command to be used for the
+     corresponding format, with the syntax:
+          <format> = <compressor> [options]
+     where <format> is one of 'bz2', 'gz', 'lz', or 'xz'.
+
+
+File: zutils.info,  Node: Zcat,  Next: Zcmp,  Prev: The zutilsrc file,  Up: Top
+
+4 Zcat
+******
+
+zcat copies each FILE argument to standard output in sequence. If any file
+given is compressed, its decompressed content is copied. If a file given
+does not exist, and its name does not end with one of the known extensions,
+zcat tries the compressed file names corresponding to the formats
+supported. If a file fails to decompress, zcat continues copying the rest
+of the files.
+
+   If a file is specified as '-', data are read from standard input,
+decompressed if needed, and sent to standard output. Data read from
+standard input must be of the same type; all uncompressed or all in the
+same compressed format.
+
+   If no files are specified, recursive searches examine the current working
+directory, and nonrecursive searches read standard input.
+
+   The format for running zcat is:
+
+     zcat [OPTIONS] [FILES]
+
+Exit status is 0 if no errors occurred, 1 otherwise.
+
+   zcat supports the following options:
+
+'-A'
+'--show-all'
+     Equivalent to '-vET'.
+
+'-b'
+'--number-nonblank'
+     Number all nonblank output lines, starting with 1. The line count is
+     unlimited.
+
+'-e'
+     Equivalent to '-vE'.
+
+'-E'
+'--show-ends'
+     Print a '$' after the end of each line.
+
+'-n'
+'--number'
+     Number all output lines, starting with 1. The line count is unlimited.
+
+'-O FORMAT'
+'--force-format=FORMAT'
+     Force the compressed format given. Valid values for FORMAT are 'bz2',
+     'gz', 'lz', and 'xz'. If this option is used, the files are passed to
+     the corresponding decompressor without verifying their format, and the
+     exact file name must be given. Other names won't be tried.
+
+'-q'
+'--quiet'
+     Quiet operation. Suppress all messages.
+
+'-r'
+'--recursive'
+     For each directory operand, read and process all files in that
+     directory, recursively. Follow symbolic links given in the command
+     line, but skip symbolic links that are encountered recursively.
+
+'-R'
+'--dereference-recursive'
+     For each directory operand, read and process all files in that
+     directory, recursively, following all symbolic links.
+
+'-s'
+'--squeeze-blank'
+     Replace multiple adjacent blank lines with a single blank line.
+
+'-t'
+     Equivalent to '-vT'.
+
+'-T'
+'--show-tabs'
+     Print TAB characters as '^I'.
+
+'-v'
+'--show-nonprinting'
+     Print control characters except for LF (newline) and TAB using '^'
+     notation and precede characters larger than 127 with 'M-' (which
+     stands for "meta").
+
+'--verbose'
+     Verbose mode. Show error messages.
+
+
+
+File: zutils.info,  Node: Zcmp,  Next: Zdiff,  Prev: Zcat,  Up: Top
+
+5 Zcmp
+******
+
+zcmp compares two files and, if they differ, writes to standard output the
+first byte and line number where they differ. Bytes and lines are numbered
+starting with 1. A hyphen '-' used as a FILE argument means standard input.
+If any file given is compressed, its decompressed content is used.
+Compressed files are decompressed on the fly; no temporary files are
+created.
+
+   The format for running zcmp is:
+
+     zcmp [OPTIONS] FILE1 [FILE2]
+
+This compares FILE1 to FILE2. The standard input is used only if FILE1 or
+FILE2 refers to standard input. If FILE2 is omitted zcmp tries the
+following:
+
+   - If FILE1 is compressed, compares its decompressed contents with the
+     corresponding uncompressed file (the name of FILE1 with the extension
+     removed).
+
+   - If FILE1 is uncompressed, compares it with the decompressed contents
+     of FILE1.[lz|bz2|gz|xz] (the first one that is found).
+
+An exit status of 0 means no differences were found, 1 means some
+differences were found, and 2 means trouble.
+
+   zcmp supports the following options:
+
+'-b'
+'--print-bytes'
+     Print the differing bytes. Print control bytes as a '^' followed by a
+     letter, and precede bytes larger than 127 with 'M-' (which stands for
+     "meta").
+
+'-i SIZE'
+'--ignore-initial=SIZE'
+     Ignore any differences in the first SIZE bytes of the input files.
+     Treat files with fewer than SIZE bytes as if they were empty. If SIZE
+     is in the form 'SIZE1:SIZE2', ignore the first SIZE1 bytes of the
+     first input file and the first SIZE2 bytes of the second input file.
+
+'-l'
+'-v'
+'--list'
+'--verbose'
+     Print the byte numbers (in decimal) and values (in octal) of all
+     differing bytes.
+
+'-n COUNT'
+'--bytes=COUNT'
+     Compare at most COUNT input bytes.
+
+'-O [FORMAT1][,FORMAT2]'
+'--force-format=[FORMAT1][,FORMAT2]'
+     Force the compressed formats given. Any of FORMAT1 or FORMAT2 may be
+     omitted and the corresponding format will be automatically detected.
+     Valid values for FORMAT are 'bz2', 'gz', 'lz', and 'xz'. If at least
+     one format is specified with this option, the file is passed to the
+     corresponding decompressor without verifying its format, and the exact
+     file names of both FILE1 and FILE2 must be given. Other names won't be
+     tried.
+
+'-q'
+'-s'
+'--quiet'
+'--silent'
+     Don't print anything; only return an exit status indicating whether the
+     files differ.
+
+
+
+File: zutils.info,  Node: Zdiff,  Next: Zgrep,  Prev: Zcmp,  Up: Top
+
+6 Zdiff
+*******
+
+zdiff compares two files and, if they differ, writes to standard output the
+differences line by line. A hyphen '-' used as a FILE argument means
+standard input. If any file given is compressed, its decompressed content
+is used. zdiff is a front end to the program diff and has the limitation
+that messages from diff refer to temporary file names instead of those
+specified.
+
+   The format for running zdiff is:
+
+     zdiff [OPTIONS] FILE1 [FILE2]
+
+This compares FILE1 to FILE2. The standard input is used only if FILE1 or
+FILE2 refers to standard input. If FILE2 is omitted zdiff tries the
+following:
+
+   - If FILE1 is compressed, compares its decompressed contents with the
+     corresponding uncompressed file (the name of FILE1 with the extension
+     removed).
+
+   - If FILE1 is uncompressed, compares it with the decompressed contents
+     of FILE1.[lz|bz2|gz|xz] (the first one that is found).
+
+An exit status of 0 means no differences were found, 1 means some
+differences were found, and 2 means trouble.
+
+   zdiff supports the following options (some options only work if the diff
+program used supports them):
+
+'-a'
+'--text'
+     Treat all files as text.
+
+'-b'
+'--ignore-space-change'
+     Ignore changes in the amount of white space.
+
+'-B'
+'--ignore-blank-lines'
+     Ignore changes whose lines are all blank.
+
+'-c'
+     Use the context output format.
+
+'-C N'
+'--context=N'
+     Same as -c but use N lines of context.
+
+'-d'
+'--minimal'
+     Try hard to find a smaller set of changes.
+
+'-E'
+'--ignore-tab-expansion'
+     Ignore changes due to tab expansion.
+
+'-i'
+'--ignore-case'
+     Ignore case differences in file contents.
+
+'-O [FORMAT1][,FORMAT2]'
+'--force-format=[FORMAT1][,FORMAT2]'
+     Force the compressed formats given. Any of FORMAT1 or FORMAT2 may be
+     omitted and the corresponding format will be automatically detected.
+     Valid values for FORMAT are 'bz2', 'gz', 'lz', and 'xz'. If at least
+     one format is specified with this option, the file is passed to the
+     corresponding decompressor without verifying its format, and the exact
+     file names of both FILE1 and FILE2 must be given. Other names won't be
+     tried.
+
+'-p'
+'--show-c-function'
+     Show which C function each change is in.
+
+'-q'
+'--brief'
+     Output only whether files differ.
+
+'-s'
+'--report-identical-files'
+     Report when two files are identical.
+
+'-t'
+'--expand-tabs'
+     Expand tabs to spaces in output.
+
+'-T'
+'--initial-tab'
+     Make tabs line up by prepending a tab.
+
+'-u'
+     Use the unified output format.
+
+'-U N'
+'--unified=N'
+     Same as -u but use N lines of context.
+
+'-w'
+'--ignore-all-space'
+     Ignore all white space.
+
+
+
+File: zutils.info,  Node: Zgrep,  Next: Ztest,  Prev: Zdiff,  Up: Top
+
+7 Zgrep
+*******
+
+zgrep is a front end to the program grep that allows transparent search on
+any combination of compressed and uncompressed files. If any file given is
+compressed, its decompressed content is used. If a file given does not
+exist, and its name does not end with one of the known extensions, zgrep
+tries the compressed file names corresponding to the formats supported. If
+a file fails to decompress, zgrep continues searching the rest of the files.
+
+   If a file is specified as '-', data are read from standard input,
+decompressed if needed, and fed to grep. Data read from standard input must
+be of the same type; all uncompressed or all in the same compressed format.
+
+   If no files are specified, recursive searches examine the current working
+directory, and nonrecursive searches read standard input.
+
+   The format for running zgrep is:
+
+     zgrep [OPTIONS] PATTERN [FILES]
+
+An exit status of 0 means at least one match was found, 1 means no matches
+were found, and 2 means trouble.
+
+   zgrep supports the following options (some options only work if the grep
+program used supports them):
+
+'-a'
+'--text'
+     Treat all files as text.
+
+'-A N'
+'--after-context=N'
+     Print N lines of trailing context.
+
+'-b'
+'--byte-offset'
+     Print the byte offset of each line.
+
+'-B N'
+'--before-context=N'
+     Print N lines of leading context.
+
+'-c'
+'--count'
+     Only print a count of matching lines per file.
+
+'-C N'
+'--context=N'
+     Print N lines of output context.
+
+'--color[=WHEN]'
+     Show matched strings in color. WHEN is 'never', 'always', or 'auto'.
+
+'-e PATTERN'
+'--regexp=PATTERN'
+     Use PATTERN as the pattern to match.
+
+'-E'
+'--extended-regexp'
+     Treat PATTERN as an extended regular expression.
+
+'-f FILE'
+'--file=FILE'
+     Obtain patterns from FILE, one per line.
+     When searching in several files at once, command substitution can be
+     used with '-e' to read FILE only once, for example if FILE is not a
+     regular file: 'zgrep -e "$(cat FILE)" file1.lz file2.gz'
+
+'-F'
+'--fixed-strings'
+     Treat PATTERN as a set of newline-separated strings.
+
+'-h'
+'--no-filename'
+     Suppress the prefixing of file names on output when multiple files are
+     searched.
+
+'-H'
+'--with-filename'
+     Print the file name for each match.
+
+'-i'
+'--ignore-case'
+     Ignore case distinctions.
+
+'-I'
+     Ignore binary files.
+
+'-l'
+'--files-with-matches'
+     Only print names of files containing at least one match.
+
+'-L'
+'--files-without-match'
+     Only print names of files not containing any matches.
+     Note: option -L fails (prints wrong results, returns wrong status, and
+     even hangs) when using GNU grep versions 3.2 to 3.4 inclusive because
+     of a wrong change in the exit status of grep, which was reverted in
+     GNU grep 3.5.
+
+'-m N'
+'--max-count=N'
+     Stop after N matches.
+
+'-n'
+'--line-number'
+     Prefix each matched line with its line number in the input file.
+
+'-o'
+'--only-matching'
+     Show only the part of matching lines that actually matches PATTERN.
+
+'-O FORMAT'
+'--force-format=FORMAT'
+     Force the compressed format given. Valid values for FORMAT are 'bz2',
+     'gz', 'lz', and 'xz'. If this option is used, the files are passed to
+     the corresponding decompressor without verifying their format, and the
+     exact file name must be given. Other names won't be tried.
+
+'-q'
+'--quiet'
+     Suppress all messages. Exit immediately with zero status if any match
+     is found, even if an error was detected.
+
+'-r'
+'--recursive'
+     For each directory operand, read and process all files in that
+     directory, recursively. Follow symbolic links given in the command
+     line, but skip symbolic links that are encountered recursively.
+
+'-R'
+'--dereference-recursive'
+     For each directory operand, read and process all files in that
+     directory, recursively, following all symbolic links.
+
+'-s'
+'--no-messages'
+     Suppress error messages about nonexistent or unreadable files.
+
+'-v'
+'--invert-match'
+     Select non-matching lines.
+
+'--verbose'
+     Verbose mode. Show error messages.
+
+'-w'
+'--word-regexp'
+     Match only whole words.
+
+'-x'
+'--line-regexp'
+     Match only whole lines.
+
+
+
+File: zutils.info,  Node: Ztest,  Next: Zupdate,  Prev: Zgrep,  Up: Top
+
+8 Ztest
+*******
+
+ztest verifies the integrity of the compressed files specified.
+Uncompressed files are ignored. If a file is specified as '-', the
+integrity of compressed data read from standard input is verified. Data
+read from standard input must be all in the same compressed format. If a
+file fails to decompress, does not exist, can't be opened, or is a
+terminal, ztest continues verifying the rest of the files. A final
+diagnostic is shown at verbosity level 1 or higher if any file fails the
+test when testing multiple files.
+
+   If no files are specified, recursive searches examine the current working
+directory, and nonrecursive searches read standard input.
+
+   Note that error detection in the xz format is broken. First, some xz
+files lack integrity information. Second, not all xz decompressors can
+verify the integrity of all xz files. Third, section 2.1.1.2 'Stream Flags'
+of the xz format specification allows xz decompressors to produce garbage
+output without issuing any warning. Therefore, xz files can't always be
+verified as reliably as files in the other formats can.
+
+   The format for running ztest is:
+
+     ztest [OPTIONS] [FILES]
+
+The exit status is 0 if all compressed files verify OK, 1 if environmental
+problems (file not found, invalid flags, I/O errors, etc), 2 if any
+compressed file is corrupt or invalid.
+
+   ztest supports the following options:
+
+'-O FORMAT'
+'--force-format=FORMAT'
+     Force the compressed format given. Valid values for FORMAT are 'bz2',
+     'gz', 'lz', and 'xz'. If this option is used, the files are passed to
+     the corresponding decompressor without verifying their format, and any
+     files in a format that the decompressor can't understand will fail.
+     For example, '--force-format=gz' can test gzipped (.gz) and compress'd
+     (.Z) files if the compressor used is GNU gzip.
+
+'-q'
+'--quiet'
+     Quiet operation. Suppress all messages.
+
+'-r'
+'--recursive'
+     For each directory operand, read and process all files in that
+     directory, recursively. Follow symbolic links given in the command
+     line, but skip symbolic links that are encountered recursively.
+
+'-R'
+'--dereference-recursive'
+     For each directory operand, read and process all files in that
+     directory, recursively, following all symbolic links.
+
+'-v'
+'--verbose'
+     Verbose mode. Show the verify status for each file processed.
+     Further -v's increase the verbosity level.
+
+
+
+File: zutils.info,  Node: Zupdate,  Next: Problems,  Prev: Ztest,  Up: Top
+
+9 Zupdate
+*********
+
+zupdate recompresses files from bzip2, gzip, and xz formats to lzip format.
+Each original is compared with the new file and then deleted. Only regular
+files with standard file name extensions are recompressed, other files are
+ignored. Compressed files are decompressed and then recompressed on the fly;
+no temporary files are created. If an error happens while recompressing a
+file, zupdate exits immediately without recompressing the rest of the files.
+The lzip format is chosen as destination because it is the most appropriate
+for long-term data archiving.
+
+   If no files are specified, recursive searches examine the current working
+directory, and nonrecursive searches do nothing.
+
+   If the lzip compressed version of a file already exists, the file is
+skipped unless the option '--force' is given. In this case, if the
+comparison with the existing lzip version fails, an error is returned and
+the original file is not deleted. The operation of zupdate is meant to be
+safe and not cause any data loss. Therefore, existing lzip compressed files
+are never overwritten nor deleted.
+
+   Combining the options '--force' and '--keep', as in
+'zupdate -f -k *.gz', verifies that there are no differences between each
+pair of files in a multiformat set of files.
+
+   The names of the original files must have one of the following
+extensions:
+'.bz2', '.gz', or '.xz', which are recompressed to '.lz';
+'.tbz', '.tbz2', '.tgz', or '.txz', which are recompressed to '.tlz'.
+Keeping the combined extensions ('.tgz' -> '.tlz') may be useful when
+recompressing Slackware packages, for example.
+
+   Recompressing a file is much like copying or moving it; therefore zupdate
+preserves the access and modification dates, permissions, and, when
+possible, ownership of the file just as 'cp -p' does. (If the user ID or
+the group ID can't be duplicated, the file permission bits S_ISUID and
+S_ISGID are cleared).
+
+   The format for running zupdate is:
+
+     zupdate [OPTIONS] [FILES]
+
+Exit status is 0 if all the compressed files were successfully recompressed
+(if needed), compared, and deleted (if requested). Non-zero otherwise.
+
+   zupdate supports the following options:
+
+'-f'
+'--force'
+     Don't skip a file for which a lzip compressed version already exists.
+     '--force' compares the content of the input file with the content of
+     the existing lzip file and deletes the input file if both contents are
+     identical.
+
+'-k'
+'--keep'
+     Keep (don't delete) the input file after comparing it with the lzip
+     file.
+
+'-l'
+'--lzip-verbose'
+     Pass one option '-v' to the lzip compressor so that it shows the
+     compression ratio for each file processed. Using lzip 1.15 or newer, a
+     second '-l' shows the progress of compression. Use it together with
+     '-v' to see the name of the file.
+
+'-q'
+'--quiet'
+     Quiet operation. Suppress all messages.
+
+'-r'
+'--recursive'
+     For each directory operand, read and process all files in that
+     directory, recursively. Follow symbolic links given in the command
+     line, but skip symbolic links that are encountered recursively.
+
+'-R'
+'--dereference-recursive'
+     For each directory operand, read and process all files in that
+     directory, recursively, following all symbolic links.
+
+'-v'
+'--verbose'
+     Verbose mode. Show the files being processed. A second '-v' also shows
+     the files being ignored.
+
+'-0 .. -9'
+     Set the compression level of lzip. By default zupdate passes '-9' to
+     lzip. Custom compression options can be passed to lzip with the option
+     '--lz'. For example '--lz='lzip -9 -s64MiB''.
+
+
+
+File: zutils.info,  Node: Problems,  Next: Concept index,  Prev: Zupdate,  Up: Top
+
+10 Reporting bugs
+*****************
+
+There are probably bugs in zutils. There are certainly errors and omissions
+in this manual. If you report them, they will get fixed. If you don't, no
+one will ever know about them and they will remain unfixed for all
+eternity, if not longer.
+
+   If you find a bug in zutils, please send electronic mail to
+<zutils-bug@nongnu.org>. Include the version number, which you can find by
+running 'zupdate --version'.
+
+
+File: zutils.info,  Node: Concept index,  Prev: Problems,  Up: Top
+
+Concept index
+*************
+
+
+* Menu:
+
+* bugs:                                  Problems.                  (line 6)
+* common options:                        Common options.            (line 6)
+* getting help:                          Problems.                  (line 6)
+* introduction:                          Introduction.              (line 6)
+* zcat:                                  Zcat.                      (line 6)
+* zcmp:                                  Zcmp.                      (line 6)
+* zdiff:                                 Zdiff.                     (line 6)
+* zgrep:                                 Zgrep.                     (line 6)
+* ztest:                                 Ztest.                     (line 6)
+* zupdate:                               Zupdate.                   (line 6)
+* zutilsrc:                              The zutilsrc file.         (line 6)
+
+
+
+Tag Table:
+Node: Top222
+Node: Introduction1151
+Node: Common options3776
+Ref: compressor-requirements5847
+Node: The zutilsrc file6219
+Node: Zcat7180
+Node: Zcmp9743
+Node: Zdiff12233
+Node: Zgrep14973
+Node: Ztest19218
+Node: Zupdate21725
+Node: Problems25409
+Node: Concept index25943
+
+End Tag Table
+
+
+Local Variables:
+coding: iso-8859-15
+End:
diff --git a/doc/zutils.texi b/doc/zutils.texi
new file mode 100644
index 0000000..c494185
--- /dev/null
+++ b/doc/zutils.texi
@@ -0,0 +1,882 @@
+\input texinfo @c -*-texinfo-*-
+@c %**start of header
+@setfilename zutils.info
+@documentencoding ISO-8859-15
+@settitle Zutils Manual
+@finalout
+@c %**end of header
+
+@set UPDATED 5 January 2021
+@set VERSION 1.10
+
+@dircategory Data Compression
+@direntry
+* Zutils: (zutils).             Utilities dealing with compressed files
+@end direntry
+
+
+@ifnothtml
+@titlepage
+@title Zutils
+@subtitle Utilities dealing with compressed files
+@subtitle for Zutils version @value{VERSION}, @value{UPDATED}
+@author by Antonio Diaz Diaz
+
+@page
+@vskip 0pt plus 1filll
+@end titlepage
+
+@contents
+@end ifnothtml
+
+@ifnottex
+@node Top
+@top
+
+This manual is for Zutils (version @value{VERSION}, @value{UPDATED}).
+
+@menu
+* Introduction::        Purpose and features of zutils
+* Common options::      Options common to all utilities
+* The zutilsrc file::   The zutils configuration file
+* Zcat::                Concatenating compressed files
+* Zcmp::                Comparing compressed files byte by byte
+* Zdiff::               Comparing compressed files line by line
+* Zgrep::               Searching inside compressed files
+* Ztest::               Testing the integrity of compressed files
+* Zupdate::             Recompressing files to lzip format
+* Problems::            Reporting bugs
+* Concept index::       Index of concepts
+@end menu
+
+@sp 1
+Copyright @copyright{} 2009-2021 Antonio Diaz Diaz.
+
+This manual is free documentation: you have unlimited permission to copy,
+distribute, and modify it.
+@end ifnottex
+
+
+@node Introduction
+@chapter Introduction
+@cindex introduction
+
+@uref{http://www.nongnu.org/zutils/zutils.html,,Zutils}
+is a collection of utilities able to process any combination of
+compressed and uncompressed files transparently. If any file given,
+including standard input, is compressed, its decompressed content is used.
+Compressed files are decompressed on the fly; no temporary files are
+created.
+
+These utilities are not wrapper scripts but safer and more efficient C++
+programs. In particular the option @samp{--recursive} is very efficient in
+those utilities supporting it.
+
+@noindent
+The utilities provided are zcat, zcmp, zdiff, zgrep, ztest, and zupdate.@*
+The formats supported are bzip2, gzip, lzip, and xz.@*
+Zutils uses external compressors. The compressor to be used for each format
+is configurable at runtime.
+
+zcat, zcmp, zdiff, and zgrep are improved replacements for the shell scripts
+provided by GNU gzip. ztest is unique to zutils. zupdate is similar to
+gzip's znew.
+
+NOTE: Bzip2 and lzip provide well-defined values of exit status, which makes
+them safe to use with zutils. Gzip and xz may return ambiguous warning
+values, making them less reliable back ends for zutils.
+@xref{compressor-requirements}.
+
+FORMAT NOTE 1: The option @samp{--format} allows the processing of a subset
+of formats in recursive mode and when trying compressed file names:
+@w{@samp{zgrep foo -r --format=bz2,lz somedir somefile.tar}}.
+
+FORMAT NOTE 2: If the option @samp{--force-format} is given, the files are
+passed to the corresponding decompressor without verifying their format,
+allowing for example the processing of compress'd (.Z) files with gzip:
+@w{@samp{zcmp --force-format=gz file.Z file.lz}}.
+
+LANGUAGE NOTE: Uncompressed = not compressed = plain data; it may never have
+been compressed. Decompressed is used to refer to data which have undergone
+the process of decompression.
+
+@sp 1
+Numbers given as arguments to options (positions, sizes) may be followed
+by a multiplier and an optional @samp{B} for "byte".
+
+Table of SI and binary prefixes (unit multipliers):
+
+@multitable {Prefix} {kilobyte  (10^3 = 1000)} {|} {Prefix} {kibibyte (2^10 = 1024)}
+@item Prefix @tab Value               @tab | @tab Prefix @tab Value
+@item k @tab kilobyte  (10^3 = 1000)  @tab | @tab Ki @tab kibibyte (2^10 = 1024)
+@item M @tab megabyte  (10^6)         @tab | @tab Mi @tab mebibyte (2^20)
+@item G @tab gigabyte  (10^9)         @tab | @tab Gi @tab gibibyte (2^30)
+@item T @tab terabyte  (10^12)        @tab | @tab Ti @tab tebibyte (2^40)
+@item P @tab petabyte  (10^15)        @tab | @tab Pi @tab pebibyte (2^50)
+@item E @tab exabyte   (10^18)        @tab | @tab Ei @tab exbibyte (2^60)
+@item Z @tab zettabyte (10^21)        @tab | @tab Zi @tab zebibyte (2^70)
+@item Y @tab yottabyte (10^24)        @tab | @tab Yi @tab yobibyte (2^80)
+@end multitable
+
+
+@node Common options
+@chapter Common options
+@cindex common options
+
+The following
+@uref{http://www.nongnu.org/arg-parser/manual/arg_parser_manual.html#Argument-syntax,,options}:
+are available in all the utilities. Rather than writing identical
+descriptions for each of the programs, they are described here.
+@ifnothtml
+@xref{Argument syntax,,,arg_parser}.
+@end ifnothtml
+
+@table @code
+@item -h
+@itemx --help
+Print an informative help message describing the options and exit. zgrep
+only supports the @samp{--help} form of this option.
+
+@item -V
+@itemx --version
+Print the version number on the standard output and exit.
+This version number should be included in all bug reports.
+
+@item -M @var{format_list}
+@itemx --format=@var{format_list}
+Process only the formats listed in the comma-separated
+@var{format_list}. Valid formats are @samp{bz2}, @samp{gz}, @samp{lz},
+@samp{xz}, and @samp{un} for @samp{uncompressed}, meaning "any file name
+without a known extension". This option excludes files based on
+extension, instead of format, because it is more efficient. The
+exclusion only applies to names generated automatically (for example
+when adding extensions to a file name or when operating recursively on
+directories). Files given in the command line are always processed.
+
+Each format in @var{format_list} enables file names with the following
+extensions:
+
+@multitable {bz2} {enables} {any other file name}
+@item bz2 @tab enables @tab .bz2 .tbz .tbz2
+@item gz  @tab enables @tab .gz .tgz
+@item lz  @tab enables @tab .lz .tlz
+@item xz  @tab enables @tab .xz .txz
+@item un  @tab enables @tab any other file name
+@end multitable
+
+@item -N
+@itemx --no-rcfile
+Don't read the runtime configuration file @samp{zutilsrc}.
+
+@item --bz2=@var{command}
+@itemx --gz=@var{command}
+@itemx --lz=@var{command}
+@itemx --xz=@var{command}
+Set program to be used as (de)compressor for the corresponding format.
+@var{command} may include arguments. For example
+@w{@samp{--lz='plzip --threads=2'}}. The program set with @samp{--lz} is
+used for both compression and decompression. The other three are used only
+for decompression. The name of the program can't begin with @samp{-}. These
+options override the values set in @file{zutilsrc}. The compression program
+used must meet three requirements:
+
+@anchor{compressor-requirements}
+@enumerate
+@item
+When called with the option @samp{-d}, it must read compressed data from
+the standard input and produce decompressed data on the standard output.
+@item
+If the option @samp{-q} is passed to zutils, the compression program must
+also accept it.
+@item
+It must return 0 if no errors occurred, and a non-zero value otherwise.
+@end enumerate
+
+@end table
+
+
+@node The zutilsrc file
+@chapter The zutils configuration file 'zutilsrc'
+@cindex zutilsrc
+
+@file{zutilsrc} is the runtime configuration file for zutils. In it you
+may define the compressor name and options to be used for each format.
+@file{zutilsrc} is optional; you don't need to install it in order to run
+zutils.
+
+The compressors specified in the command line override those specified
+in @file{zutilsrc}.
+
+You may copy the system @file{zutilsrc} file @file{$@{sysconfdir@}/zutilsrc}
+to @file{$HOME/.zutilsrc} and customize these options as you like. The file
+syntax is fairly obvious (and there are further instructions in it):
+
+@enumerate
+@item
+Any line beginning with @samp{#} is a comment line.
+@item
+Each non-comment line defines the command to be used for the corresponding
+format, with the syntax:
+@example
+<format> = <compressor> [options]
+@end example
+where <format> is one of @samp{bz2}, @samp{gz}, @samp{lz}, or @samp{xz}.
+@end enumerate
+
+
+@node Zcat
+@chapter Zcat
+@cindex zcat
+
+zcat copies each @var{file} argument to standard output in sequence. If any
+file given is compressed, its decompressed content is copied. If a file
+given does not exist, and its name does not end with one of the known
+extensions, zcat tries the compressed file names corresponding to the
+formats supported. If a file fails to decompress, zcat continues copying the
+rest of the files.
+
+If a file is specified as @samp{-}, data are read from standard input,
+decompressed if needed, and sent to standard output. Data read from
+standard input must be of the same type; all uncompressed or all in the
+same compressed format.
+
+If no files are specified, recursive searches examine the current working
+directory, and nonrecursive searches read standard input.
+
+The format for running zcat is:
+
+@example
+zcat [@var{options}] [@var{files}]
+@end example
+
+@noindent
+Exit status is 0 if no errors occurred, 1 otherwise.
+
+zcat supports the following options:
+
+@table @code
+@item -A
+@itemx --show-all
+Equivalent to @samp{-vET}.
+
+@item -b
+@itemx --number-nonblank
+Number all nonblank output lines, starting with 1. The line count is
+unlimited.
+
+@item -e
+Equivalent to @samp{-vE}.
+
+@item -E
+@itemx --show-ends
+Print a @samp{$} after the end of each line.
+
+@item -n
+@itemx --number
+Number all output lines, starting with 1. The line count is unlimited.
+
+@item -O @var{format}
+@itemx --force-format=@var{format}
+Force the compressed format given. Valid values for @var{format} are
+@samp{bz2}, @samp{gz}, @samp{lz}, and @samp{xz}. If this option is used,
+the files are passed to the corresponding decompressor without verifying
+their format, and the exact file name must be given. Other names won't
+be tried.
+
+@item -q
+@itemx --quiet
+Quiet operation. Suppress all messages.
+
+@item -r
+@itemx --recursive
+For each directory operand, read and process all files in that directory,
+recursively. Follow symbolic links given in the command line, but skip
+symbolic links that are encountered recursively.
+
+@item -R
+@itemx --dereference-recursive
+For each directory operand, read and process all files in that directory,
+recursively, following all symbolic links.
+
+@item -s
+@itemx --squeeze-blank
+Replace multiple adjacent blank lines with a single blank line.
+
+@item -t
+Equivalent to @samp{-vT}.
+
+@item -T
+@itemx --show-tabs
+Print TAB characters as @samp{^I}.
+
+@item -v
+@itemx --show-nonprinting
+Print control characters except for LF (newline) and TAB using @samp{^}
+notation and precede characters larger than 127 with @samp{M-} (which
+stands for "meta").
+
+@item --verbose
+Verbose mode. Show error messages.
+
+@end table
+
+
+@node Zcmp
+@chapter Zcmp
+@cindex zcmp
+
+zcmp compares two files and, if they differ, writes to standard output the
+first byte and line number where they differ. Bytes and lines are numbered
+starting with 1. A hyphen @samp{-} used as a @var{file} argument means
+standard input. If any file given is compressed, its decompressed content is
+used. Compressed files are decompressed on the fly; no temporary files are
+created.
+
+The format for running zcmp is:
+
+@example
+zcmp [@var{options}] @var{file1} [@var{file2}]
+@end example
+
+@noindent
+This compares @var{file1} to @var{file2}. The standard input is used only if
+@var{file1} or @var{file2} refers to standard input. If @var{file2} is
+omitted zcmp tries the following:
+
+@itemize -
+@item
+If @var{file1} is compressed, compares its decompressed contents with
+the corresponding uncompressed file (the name of @var{file1} with the
+extension removed).
+@item
+If @var{file1} is uncompressed, compares it with the decompressed
+contents of @var{file1}.[lz|bz2|gz|xz] (the first one that is found).
+@end itemize
+
+@noindent
+An exit status of 0 means no differences were found, 1 means some
+differences were found, and 2 means trouble.
+
+zcmp supports the following options:
+
+@table @code
+@item -b
+@itemx --print-bytes
+Print the differing bytes. Print control bytes as a @samp{^} followed by
+a letter, and precede bytes larger than 127 with @samp{M-} (which stands
+for "meta").
+
+@item -i @var{size}
+@itemx --ignore-initial=@var{size}
+Ignore any differences in the first @var{size} bytes of the input files.
+Treat files with fewer than @var{size} bytes as if they were empty. If
+@var{size} is in the form @samp{@var{size1}:@var{size2}}, ignore the
+first @var{size1} bytes of the first input file and the first
+@var{size2} bytes of the second input file.
+
+@item -l
+@itemx -v
+@itemx --list
+@itemx --verbose
+Print the byte numbers (in decimal) and values (in octal) of all
+differing bytes.
+
+@item -n @var{count}
+@itemx --bytes=@var{count}
+Compare at most @var{count} input bytes.
+
+@item -O [@var{format1}][,@var{format2}]
+@itemx --force-format=[@var{format1}][,@var{format2}]
+Force the compressed formats given. Any of @var{format1} or
+@var{format2} may be omitted and the corresponding format will be
+automatically detected. Valid values for @var{format} are @samp{bz2},
+@samp{gz}, @samp{lz}, and @samp{xz}. If at least one format is specified
+with this option, the file is passed to the corresponding decompressor
+without verifying its format, and the exact file names of both
+@var{file1} and @var{file2} must be given. Other names won't be tried.
+
+@item -q
+@itemx -s
+@itemx --quiet
+@itemx --silent
+Don't print anything; only return an exit status indicating whether the
+files differ.
+
+@end table
+
+
+@node Zdiff
+@chapter Zdiff
+@cindex zdiff
+
+zdiff compares two files and, if they differ, writes to standard output the
+differences line by line. A hyphen @samp{-} used as a @var{file} argument
+means standard input. If any file given is compressed, its decompressed
+content is used. zdiff is a front end to the program diff and has the
+limitation that messages from diff refer to temporary file names instead of
+those specified.
+
+The format for running zdiff is:
+
+@example
+zdiff [@var{options}] @var{file1} [@var{file2}]
+@end example
+
+@noindent
+This compares @var{file1} to @var{file2}. The standard input is used only if
+@var{file1} or @var{file2} refers to standard input. If @var{file2} is
+omitted zdiff tries the following:
+
+@itemize -
+@item
+If @var{file1} is compressed, compares its decompressed contents with
+the corresponding uncompressed file (the name of @var{file1} with the
+extension removed).
+@item
+If @var{file1} is uncompressed, compares it with the decompressed
+contents of @var{file1}.[lz|bz2|gz|xz] (the first one that is found).
+@end itemize
+
+@noindent
+An exit status of 0 means no differences were found, 1 means some
+differences were found, and 2 means trouble.
+
+zdiff supports the following options (some options only work if the diff
+program used supports them):
+
+@table @code
+@item -a
+@itemx --text
+Treat all files as text.
+
+@item -b
+@itemx --ignore-space-change
+Ignore changes in the amount of white space.
+
+@item -B
+@itemx --ignore-blank-lines
+Ignore changes whose lines are all blank.
+
+@itemx -c
+Use the context output format.
+
+@item -C @var{n}
+@itemx --context=@var{n}
+Same as -c but use @var{n} lines of context.
+
+@item -d
+@itemx --minimal
+Try hard to find a smaller set of changes.
+
+@item -E
+@itemx --ignore-tab-expansion
+Ignore changes due to tab expansion.
+
+@item -i
+@itemx --ignore-case
+Ignore case differences in file contents.
+
+@item -O [@var{format1}][,@var{format2}]
+@itemx --force-format=[@var{format1}][,@var{format2}]
+Force the compressed formats given. Any of @var{format1} or
+@var{format2} may be omitted and the corresponding format will be
+automatically detected. Valid values for @var{format} are @samp{bz2},
+@samp{gz}, @samp{lz}, and @samp{xz}. If at least one format is specified
+with this option, the file is passed to the corresponding decompressor
+without verifying its format, and the exact file names of both
+@var{file1} and @var{file2} must be given. Other names won't be tried.
+
+@item -p
+@itemx --show-c-function
+Show which C function each change is in.
+
+@item -q
+@itemx --brief
+Output only whether files differ.
+
+@item -s
+@itemx --report-identical-files
+Report when two files are identical.
+
+@item -t
+@itemx --expand-tabs
+Expand tabs to spaces in output.
+
+@item -T
+@itemx --initial-tab
+Make tabs line up by prepending a tab.
+
+@item -u
+Use the unified output format.
+
+@item -U @var{n}
+@itemx --unified=@var{n}
+Same as -u but use @var{n} lines of context.
+
+@item -w
+@itemx --ignore-all-space
+Ignore all white space.
+
+@end table
+
+
+@node Zgrep
+@chapter Zgrep
+@cindex zgrep
+
+zgrep is a front end to the program grep that allows transparent search
+on any combination of compressed and uncompressed files. If any file
+given is compressed, its decompressed content is used. If a file given
+does not exist, and its name does not end with one of the known
+extensions, zgrep tries the compressed file names corresponding to the
+formats supported. If a file fails to decompress, zgrep continues
+searching the rest of the files.
+
+If a file is specified as @samp{-}, data are read from standard input,
+decompressed if needed, and fed to grep. Data read from standard input
+must be of the same type; all uncompressed or all in the same
+compressed format.
+
+If no files are specified, recursive searches examine the current working
+directory, and nonrecursive searches read standard input.
+
+The format for running zgrep is:
+
+@example
+zgrep [@var{options}] @var{pattern} [@var{files}]
+@end example
+
+@noindent
+An exit status of 0 means at least one match was found, 1 means no
+matches were found, and 2 means trouble.
+
+zgrep supports the following options (some options only work if the grep
+program used supports them):
+
+@table @code
+@item -a
+@itemx --text
+Treat all files as text.
+
+@item -A @var{n}
+@itemx --after-context=@var{n}
+Print @var{n} lines of trailing context.
+
+@item -b
+@itemx --byte-offset
+Print the byte offset of each line.
+
+@item -B @var{n}
+@itemx --before-context=@var{n}
+Print @var{n} lines of leading context.
+
+@item -c
+@itemx --count
+Only print a count of matching lines per file.
+
+@item -C @var{n}
+@itemx --context=@var{n}
+Print @var{n} lines of output context.
+
+@item --color[=@var{when}]
+Show matched strings in color. @var{when} is @samp{never}, @samp{always},
+or @samp{auto}.
+
+@item -e @var{pattern}
+@itemx --regexp=@var{pattern}
+Use @var{pattern} as the pattern to match.
+
+@item -E
+@itemx --extended-regexp
+Treat @var{pattern} as an extended regular expression.
+
+@item -f @var{file}
+@itemx --file=@var{file}
+Obtain patterns from @var{file}, one per line.@*
+When searching in several files at once, command substitution can be
+used with @samp{-e} to read @var{file} only once, for example if
+@var{file} is not a regular file:
+@w{@samp{zgrep -e "$(cat @var{file})" file1.lz file2.gz}}
+
+@item -F
+@itemx --fixed-strings
+Treat @var{pattern} as a set of newline-separated strings.
+
+@item -h
+@itemx --no-filename
+Suppress the prefixing of file names on output when multiple files are
+searched.
+
+@item -H
+@itemx --with-filename
+Print the file name for each match.
+
+@item -i
+@itemx --ignore-case
+Ignore case distinctions.
+
+@item -I
+Ignore binary files.
+
+@item -l
+@itemx --files-with-matches
+Only print names of files containing at least one match.
+
+@item -L
+@itemx --files-without-match
+Only print names of files not containing any matches.@*
+Note: option -L fails (prints wrong results, returns wrong status, and even
+hangs) when using GNU grep versions 3.2 to 3.4 inclusive because of a wrong
+change in the exit status of grep, which was reverted in GNU grep 3.5.
+
+@item -m @var{n}
+@itemx --max-count=@var{n}
+Stop after @var{n} matches.
+
+@item -n
+@itemx --line-number
+Prefix each matched line with its line number in the input file.
+
+@item -o
+@itemx --only-matching
+Show only the part of matching lines that actually matches @var{pattern}.
+
+@item -O @var{format}
+@itemx --force-format=@var{format}
+Force the compressed format given. Valid values for @var{format} are
+@samp{bz2}, @samp{gz}, @samp{lz}, and @samp{xz}. If this option is used,
+the files are passed to the corresponding decompressor without verifying
+their format, and the exact file name must be given. Other names won't
+be tried.
+
+@item -q
+@itemx --quiet
+Suppress all messages. Exit immediately with zero status if any match is
+found, even if an error was detected.
+
+@item -r
+@itemx --recursive
+For each directory operand, read and process all files in that directory,
+recursively. Follow symbolic links given in the command line, but skip
+symbolic links that are encountered recursively.
+
+@item -R
+@itemx --dereference-recursive
+For each directory operand, read and process all files in that directory,
+recursively, following all symbolic links.
+
+@item -s
+@itemx --no-messages
+Suppress error messages about nonexistent or unreadable files.
+
+@item -v
+@itemx --invert-match
+Select non-matching lines.
+
+@item --verbose
+Verbose mode. Show error messages.
+
+@item -w
+@itemx --word-regexp
+Match only whole words.
+
+@item -x
+@itemx --line-regexp
+Match only whole lines.
+
+@end table
+
+
+@node Ztest
+@chapter Ztest
+@cindex ztest
+
+ztest verifies the integrity of the compressed files specified.
+Uncompressed files are ignored. If a file is specified as @samp{-}, the
+integrity of compressed data read from standard input is verified. Data
+read from standard input must be all in the same compressed format. If
+a file fails to decompress, does not exist, can't be opened, or is a
+terminal, ztest continues verifying the rest of the files. A final
+diagnostic is shown at verbosity level 1 or higher if any file fails the
+test when testing multiple files.
+
+If no files are specified, recursive searches examine the current working
+directory, and nonrecursive searches read standard input.
+
+Note that error detection in the xz format is broken. First, some xz
+files lack integrity information. Second, not all xz decompressors can
+@uref{http://www.nongnu.org/lzip/xz_inadequate.html#fragmented,,verify the integrity}
+of all xz files. Third, section 2.1.1.2 'Stream Flags' of the
+@uref{http://tukaani.org/xz/xz-file-format.txt,,xz format specification}
+allows xz decompressors to produce garbage output without issuing any
+warning. Therefore, xz files can't always be verified as reliably as
+files in the other formats can.
+@c We can only hope that xz is soon abandoned.
+
+The format for running ztest is:
+
+@example
+ztest [@var{options}] [@var{files}]
+@end example
+
+@noindent
+The exit status is 0 if all compressed files verify OK, 1 if
+environmental problems (file not found, invalid flags, I/O errors, etc),
+2 if any compressed file is corrupt or invalid.
+
+ztest supports the following options:
+
+@table @code
+@item -O @var{format}
+@itemx --force-format=@var{format}
+Force the compressed format given. Valid values for @var{format} are
+@samp{bz2}, @samp{gz}, @samp{lz}, and @samp{xz}. If this option is used, the
+files are passed to the corresponding decompressor without verifying their
+format, and any files in a format that the decompressor can't understand
+will fail. For example, @samp{--force-format=gz} can test gzipped (.gz) and
+compress'd (.Z) files if the compressor used is GNU gzip.
+
+@item -q
+@itemx --quiet
+Quiet operation. Suppress all messages.
+
+@item -r
+@itemx --recursive
+For each directory operand, read and process all files in that directory,
+recursively. Follow symbolic links given in the command line, but skip
+symbolic links that are encountered recursively.
+
+@item -R
+@itemx --dereference-recursive
+For each directory operand, read and process all files in that directory,
+recursively, following all symbolic links.
+
+@item -v
+@itemx --verbose
+Verbose mode. Show the verify status for each file processed.@*
+Further -v's increase the verbosity level.
+
+@end table
+
+
+@node Zupdate
+@chapter Zupdate
+@cindex zupdate
+
+zupdate recompresses files from bzip2, gzip, and xz formats to lzip format.
+Each original is compared with the new file and then deleted. Only regular
+files with standard file name extensions are recompressed, other files are
+ignored. Compressed files are decompressed and then recompressed on the fly;
+no temporary files are created. If an error happens while recompressing a
+file, zupdate exits immediately without recompressing the rest of the files.
+The lzip format is chosen as destination because it is the most appropriate
+for long-term data archiving.
+
+If no files are specified, recursive searches examine the current working
+directory, and nonrecursive searches do nothing.
+
+If the lzip compressed version of a file already exists, the file is
+skipped unless the option @samp{--force} is given. In this case, if the
+comparison with the existing lzip version fails, an error is returned
+and the original file is not deleted. The operation of zupdate is meant
+to be safe and not cause any data loss. Therefore, existing lzip
+compressed files are never overwritten nor deleted.
+
+Combining the options @samp{--force} and @samp{--keep}, as in
+@w{@samp{zupdate -f -k *.gz}}, verifies that there are no differences
+between each pair of files in a multiformat set of files.
+
+The names of the original files must have one of the following extensions:@*
+@samp{.bz2}, @samp{.gz}, or @samp{.xz}, which are recompressed to
+@samp{.lz};@*
+@samp{.tbz}, @samp{.tbz2}, @samp{.tgz}, or @samp{.txz}, which are
+recompressed to @samp{.tlz}.@*
+Keeping the combined extensions (@samp{.tgz} --> @samp{.tlz}) may be useful
+when recompressing Slackware packages, for example.
+
+Recompressing a file is much like copying or moving it; therefore zupdate
+preserves the access and modification dates, permissions, and, when
+possible, ownership of the file just as @samp{cp -p} does. (If the user ID or
+the group ID can't be duplicated, the file permission bits S_ISUID and
+S_ISGID are cleared).
+
+The format for running zupdate is:
+
+@example
+zupdate [@var{options}] [@var{files}]
+@end example
+
+@noindent
+Exit status is 0 if all the compressed files were successfully recompressed
+(if needed), compared, and deleted (if requested). Non-zero otherwise.
+
+zupdate supports the following options:
+
+@table @code
+@item -f
+@itemx --force
+Don't skip a file for which a lzip compressed version already exists.
+@samp{--force} compares the content of the input file with the content
+of the existing lzip file and deletes the input file if both contents
+are identical.
+
+@item -k
+@itemx --keep
+Keep (don't delete) the input file after comparing it with the lzip file.
+
+@item -l
+@itemx --lzip-verbose
+Pass one option @samp{-v} to the lzip compressor so that it shows the
+compression ratio for each file processed. Using lzip 1.15 or newer, a
+second @samp{-l} shows the progress of compression. Use it together with
+@samp{-v} to see the name of the file.
+
+@item -q
+@itemx --quiet
+Quiet operation. Suppress all messages.
+
+@item -r
+@itemx --recursive
+For each directory operand, read and process all files in that directory,
+recursively. Follow symbolic links given in the command line, but skip
+symbolic links that are encountered recursively.
+
+@item -R
+@itemx --dereference-recursive
+For each directory operand, read and process all files in that directory,
+recursively, following all symbolic links.
+
+@item -v
+@itemx --verbose
+Verbose mode. Show the files being processed. A second @samp{-v} also
+shows the files being ignored.
+
+@item -0 .. -9
+Set the compression level of lzip. By default zupdate passes @samp{-9} to
+lzip. Custom compression options can be passed to lzip with the option
+@samp{--lz}. For example @w{@samp{--lz='lzip -9 -s64MiB'}}.
+
+@end table
+
+
+@node Problems
+@chapter Reporting bugs
+@cindex bugs
+@cindex getting help
+
+There are probably bugs in zutils. There are certainly errors and
+omissions in this manual. If you report them, they will get fixed. If
+you don't, no one will ever know about them and they will remain unfixed
+for all eternity, if not longer.
+
+If you find a bug in zutils, please send electronic mail to
+@email{zutils-bug@@nongnu.org}. Include the version number, which you can
+find by running @w{@samp{zupdate --version}}.
+
+
+@node Concept index
+@unnumbered Concept index
+
+@printindex cp
+
+@bye
diff --git a/rc.cc b/rc.cc
new file mode 100644
index 0000000..0162510
--- /dev/null
+++ b/rc.cc
@@ -0,0 +1,411 @@
+/* Zutils - Utilities dealing with compressed files
+   Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#define _FILE_OFFSET_BITS 64
+
+#include <cerrno>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <string>
+#include <vector>
+#include <unistd.h>
+#include <sys/wait.h>
+
+#include "arg_parser.h"
+#include "rc.h"
+
+
+const char * invocation_name = 0;
+const char * program_name = 0;
+int verbosity = 0;
+
+namespace {
+
+const char * const config_file_name = "zutilsrc";
+const char * const     program_year = "2021";
+
+std::string compressor_names[num_formats] =
+  { "bzip2", "gzip", "lzip", "xz" };		// default compressor names
+
+// args to compressors read from rc or from options --[bglx]z, maybe empty
+std::vector< std::string > compressor_args[num_formats];
+
+// vector of enabled formats plus [num_formats] for uncompressed.
+// empty means all enabled.
+std::vector< bool > enabled_formats;
+
+const struct { const char * from; const char * to; int format_index; }
+  known_extensions[] = {
+  { ".bz2",  "",     fmt_bz2 },
+  { ".tbz",  ".tar", fmt_bz2 },
+  { ".tbz2", ".tar", fmt_bz2 },
+  { ".gz",   "",     fmt_gz },
+  { ".tgz",  ".tar", fmt_gz },
+  { ".lz",   "",     fmt_lz },
+  { ".tlz",  ".tar", fmt_lz },
+  { ".xz",   "",     fmt_xz },
+  { ".txz",  ".tar", fmt_xz },
+  { 0,       0,      -1 } };
+
+
+int my_fgetc( FILE * const f )
+  {
+  int ch;
+  bool comment = false;
+
+  do {
+    ch = std::fgetc( f );
+    if( ch == '#' ) comment = true;
+    else if( ch == '\n' || ch == EOF ) comment = false;
+    else if( ch == '\\' && comment )
+      {
+      const int c = std::fgetc( f );
+      if( c == '\n' ) { std::ungetc( c, f ); comment = false; }
+      }
+    }
+  while( comment );
+  return ch;
+  }
+
+
+// Returns the parity of escapes (backslashes) at the end of a string.
+bool trailing_escape( const std::string & s )
+  {
+  unsigned len = s.size();
+  bool odd_escape = false;
+  while( len > 0 && s[--len] == '\\' ) odd_escape = !odd_escape;
+  return odd_escape;
+  }
+
+
+/* Read a line discarding comments, leading whitespace, and blank lines.
+   Escaped newlines are discarded.
+   Returns the empty string if at EOF.
+*/
+const std::string & my_fgets( FILE * const f, int & linenum )
+  {
+  static std::string s;
+  bool strip = true;			// strip leading whitespace
+  s.clear();
+
+  while( true )
+    {
+    int ch = my_fgetc( f );
+    if( strip )
+      {
+      strip = false;
+      while( std::isspace( ch ) )
+        { if( ch == '\n' ) { ++linenum; } ch = my_fgetc( f ); }
+      }
+    if( ch == EOF ) { if( s.size() ) { ++linenum; } break; }
+    else if( ch == '\n' )
+      {
+      ++linenum; strip = true;
+      if( trailing_escape( s ) ) s.erase( s.size() - 1 );
+      else if( s.size() ) break;
+      }
+    else s += ch;
+    }
+  return s;
+  }
+
+
+bool parse_compressor_command( const std::string & s, int i,
+                               const int format_index )
+  {
+  const int len = s.size();
+  while( i < len && std::isspace( s[i] ) ) ++i;		// strip spaces
+  int l = i;
+  while( i < len && !std::isspace( s[i] ) ) ++i;
+  if( l >= i || s[l] == '-' ) return false;
+  compressor_names[format_index].assign( s, l, i - l );
+
+  compressor_args[format_index].clear();
+  while( i < len )
+    {
+    while( i < len && std::isspace( s[i] ) ) ++i;	// strip spaces
+    l = i;
+    while( i < len && !std::isspace( s[i] ) ) ++i;
+    if( l < i )
+      compressor_args[format_index].push_back( std::string( s, l, i - l ) );
+    }
+  return true;
+  }
+
+
+bool parse_rc_line( const std::string & line,
+                    const char * const filename, const int linenum )
+  {
+  const int len = line.size();
+  int i = 0;
+  while( i < len && std::isspace( line[i] ) ) ++i;	// strip spaces
+  int l = i;
+  while( i < len && line[i] != '=' && !std::isspace( line[i] ) ) ++i;
+  if( l >= i )
+    { if( verbosity >= 0 )
+        std::fprintf( stderr, "%s %d: missing format name.\n", filename, linenum );
+      return false; }
+  const std::string name( line, l, i - l );
+  int format_index = -1;
+  for( int j = 0; j < num_formats; ++j )
+    if( name == format_names[j] ) { format_index = j; break; }
+  if( format_index < 0 )
+    { if( verbosity >= 0 )
+        std::fprintf( stderr, "%s %d: bad format name '%s'\n",
+                      filename, linenum, name.c_str() );
+      return false; }
+
+  while( i < len && std::isspace( line[i] ) ) ++i;	// strip spaces
+  if( i <= 0 || i >= len || line[i] != '=' )
+    { if( verbosity >= 0 )
+        std::fprintf( stderr, "%s %d: missing '='\n", filename, linenum );
+      return false; }
+  ++i;							// skip the '='
+  if( !parse_compressor_command( line, i, format_index ) )
+    {
+    if( verbosity >= 0 )
+      std::fprintf( stderr, "%s %d: missing compressor name.\n", filename, linenum );
+    return false;
+    }
+  return true;
+  }
+
+
+    // Returns 0 for success, 1 for file not found, 2 for syntax error.
+int process_rcfile( const std::string & name )
+  {
+  FILE * const f = std::fopen( name.c_str(), "r" );
+  if( !f ) return 1;
+
+  int linenum = 0;
+  int retval = 0;
+
+  while( true )
+    {
+    const std::string & line = my_fgets( f, linenum );
+    if( line.empty() ) break;				// EOF
+    if( !parse_rc_line( line, name.c_str(), linenum ) )
+      { retval = 2; break; }
+    }
+  std::fclose( f );
+  return retval;
+  }
+
+} // end namespace
+
+
+bool enabled_format( const int format_index )
+  {
+  if( enabled_formats.size() <= num_formats ) return true;	// all enabled
+  if( format_index < 0 ) return enabled_formats[num_formats];	// uncompressed
+  return enabled_formats[format_index];
+  }
+
+
+void parse_format_list( const std::string & arg )
+  {
+  const std::string un( "uncompressed" );
+  bool error = arg.empty();
+  enabled_formats.assign( num_formats + 1, false );
+
+  for( unsigned l = 0, r; l < arg.size(); l = r + 1 )
+    {
+    r = std::min( arg.find( ',', l ), arg.size() );
+    if( l >= r ) { error = true; break; }		// empty format
+    int format_index = num_formats;
+    const std::string s( arg, l, r - l );
+    for( int i = 0; i < num_formats; ++i )
+      if( s == format_names[i] )
+        { format_index = i; break; }
+    if( format_index == num_formats && un.find( s ) != 0 )
+      { error = true; break; }
+    enabled_formats[format_index] = true;
+    }
+  if( error )
+    { show_error( "Bad argument for option '--format'." ); std::exit( 1 ); }
+  }
+
+
+int parse_format_type( const std::string & arg )
+  {
+  for( int i = 0; i < num_formats; ++i )
+    if( arg == format_names[i] )
+      return i;
+  show_error( "Bad argument for option '--force-format'." );
+  std::exit( 1 );
+  }
+
+
+int extension_index( const std::string & name )
+  {
+  for( int eindex = 0; known_extensions[eindex].from; ++eindex )
+    {
+    const std::string ext( known_extensions[eindex].from );
+    if( name.size() > ext.size() &&
+        name.compare( name.size() - ext.size(), ext.size(), ext ) == 0 )
+      return eindex;
+    }
+  return -1;
+  }
+
+int extension_format( const int eindex )
+  { return ( eindex >= 0 ) ? known_extensions[eindex].format_index : -1; }
+
+const char * extension_from( const int eindex )
+  { return known_extensions[eindex].from; }
+
+const char * extension_to( const int eindex )
+  { return known_extensions[eindex].to; }
+
+
+void maybe_process_config_file( const Arg_parser & parser )
+  {
+  for( int i = 0; i < parser.arguments(); ++i )
+    if( parser.code( i ) == 'N' ) return;
+  std::string name;
+  const char * p = std::getenv( "HOME" ); if( p ) name = p;
+  if( name.size() )
+    {
+    name += "/."; name += config_file_name;
+    const int retval = process_rcfile( name );
+    if( retval == 0 ) return;
+    if( retval == 2 ) std::exit( 2 );
+    }
+  name = SYSCONFDIR; name += '/'; name += config_file_name;
+  const int retval = process_rcfile( name );
+  if( retval == 2 ) std::exit( 2 );
+  }
+
+
+void parse_compressor( const std::string & arg, const int format_index,
+                       const int eretval )
+  {
+  if( !parse_compressor_command( arg, 0, format_index ) )
+    { show_error( "Missing compressor name." ); std::exit( eretval ); }
+  }
+
+
+const char * get_compressor_name( const int format_index )
+  {
+  if( format_index >= 0 && format_index < num_formats &&
+      compressor_names[format_index].size() )
+    return compressor_names[format_index].c_str();
+  return 0;
+  }
+
+
+const std::vector< std::string > & get_compressor_args( const int format_index )
+  {
+  return compressor_args[format_index];
+  }
+
+
+void show_help_addr()
+  {
+  std::printf( "\nReport bugs to zutils-bug@nongnu.org\n"
+               "Zutils home page: http://www.nongnu.org/zutils/zutils.html\n" );
+  }
+
+
+void show_version()
+  {
+  std::printf( "%s (zutils) %s\n", program_name, PROGVERSION );
+  std::printf( "Copyright (C) %s Antonio Diaz Diaz.\n", program_year );
+  std::printf( "License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl.html>\n"
+               "This is free software: you are free to change and redistribute it.\n"
+               "There is NO WARRANTY, to the extent permitted by law.\n" );
+  }
+
+
+void show_error( const char * const msg, const int errcode, const bool help )
+  {
+  if( verbosity < 0 ) return;
+  if( msg && msg[0] )
+    std::fprintf( stderr, "%s: %s%s%s\n", program_name, msg,
+                  ( errcode > 0 ) ? ": " : "",
+                  ( errcode > 0 ) ? std::strerror( errcode ) : "" );
+  if( help )
+    std::fprintf( stderr, "Try '%s --help' for more information.\n",
+                  invocation_name );
+  }
+
+
+void show_file_error( const char * const filename, const char * const msg,
+                      const int errcode )
+  {
+  if( verbosity >= 0 )
+    std::fprintf( stderr, "%s: %s: %s%s%s\n", program_name, filename, msg,
+                  ( errcode > 0 ) ? ": " : "",
+                  ( errcode > 0 ) ? std::strerror( errcode ) : "" );
+  }
+
+
+void internal_error( const char * const msg )
+  {
+  if( verbosity >= 0 )
+    std::fprintf( stderr, "%s: internal error: %s\n", program_name, msg );
+  std::exit( 3 );
+  }
+
+
+void show_close_error( const char * const prog_name )
+  {
+  if( verbosity >= 0 )
+    std::fprintf( stderr, "%s: Error closing output of %s: %s\n",
+                  program_name, prog_name, std::strerror( errno ) );
+  }
+
+
+void show_exec_error( const char * const prog_name )
+  {
+  if( verbosity >= 0 )
+    std::fprintf( stderr, "%s: Can't exec '%s': %s\n",
+                  program_name, prog_name, std::strerror( errno ) );
+  }
+
+
+void show_fork_error( const char * const prog_name )
+  {
+  if( verbosity >= 0 )
+    std::fprintf( stderr, "%s: Can't fork '%s': %s\n",
+                  program_name, prog_name, std::strerror( errno ) );
+  }
+
+
+int wait_for_child( const pid_t pid, const char * const name,
+                    const int eretval, const bool isgzxz )
+  {
+  int status;
+  while( waitpid( pid, &status, 0 ) == -1 )
+    {
+    if( errno != EINTR )
+      {
+      if( verbosity >= 0 )
+        std::fprintf( stderr, "%s: Error waiting termination of '%s': %s\n",
+                      program_name, name, std::strerror( errno ) );
+      _exit( eretval );
+      }
+    }
+  if( WIFEXITED( status ) )
+    {
+    const int tmp = WEXITSTATUS( status );
+    if( isgzxz && eretval == 1 && tmp == 1 ) return 2;		// for ztest
+    return tmp;
+    }
+  return eretval;
+  }
diff --git a/rc.h b/rc.h
new file mode 100644
index 0000000..38a737e
--- /dev/null
+++ b/rc.h
@@ -0,0 +1,62 @@
+/* Zutils - Utilities dealing with compressed files
+   Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+enum { fmt_bz2, fmt_gz, fmt_lz, fmt_xz, num_formats };	// format_index
+const char * const format_names[num_formats] = { "bz2", "gz", "lz", "xz" };
+const char * const simple_extensions[num_formats] =
+  { ".bz2", ".gz", ".lz", ".xz" };
+const int format_order[num_formats] =
+  { fmt_lz, fmt_bz2, fmt_gz, fmt_xz };			// search order
+
+bool enabled_format( const int format_index );
+void parse_format_list( const std::string & arg );
+int parse_format_type( const std::string & arg );
+
+int extension_index( const std::string & name );	// -1 if unknown
+int extension_format( const int eindex );		// -1 if uncompressed
+const char * extension_from( const int eindex );
+const char * extension_to( const int eindex );
+
+extern const char * invocation_name;
+extern const char * program_name;
+extern int verbosity;
+
+class Arg_parser;
+
+void maybe_process_config_file( const Arg_parser & parser );
+
+void parse_compressor( const std::string & arg, const int format_index,
+                       const int eretval = 2 );
+
+const char * get_compressor_name( const int format_index );
+const std::vector< std::string > & get_compressor_args( const int format_index );
+
+void show_help_addr();
+void show_version();
+void show_error( const char * const msg, const int errcode = 0,
+                 const bool help = false );
+void show_file_error( const char * const filename, const char * const msg,
+                      const int errcode = 0 );
+void internal_error( const char * const msg );
+void show_close_error( const char * const prog_name = "data feeder" );
+void show_exec_error( const char * const prog_name );
+void show_fork_error( const char * const prog_name );
+
+// Returns exit status of child process 'pid', or 'eretval' in case of error.
+//
+int wait_for_child( const pid_t pid, const char * const name,
+                    const int eretval = 2, const bool isgzxz = false );
diff --git a/recursive.cc b/recursive.cc
new file mode 100644
index 0000000..a69e117
--- /dev/null
+++ b/recursive.cc
@@ -0,0 +1,109 @@
+/* Zutils - Utilities dealing with compressed files
+   Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+/* Returns true if full_name is a regular file with an enabled extension
+   or (a link to) a directory. */
+bool test_full_name( const std::string & full_name, const struct stat * stp,
+                     const bool follow )
+  {
+  struct stat st, st2;
+  if( follow && stat( full_name.c_str(), &st ) != 0 ) return false;
+  if( !follow && lstat( full_name.c_str(), &st ) != 0 ) return false;
+  if( S_ISREG( st.st_mode ) )			// regular file
+    return enabled_format( extension_format( extension_index( full_name ) ) );
+  if( !S_ISDIR( st.st_mode ) ) return false;
+
+  std::string prev_dir( full_name );
+  bool loop = ( stp && st.st_ino == stp->st_ino && st.st_dev == stp->st_dev );
+  if( !loop )
+    for( unsigned i = prev_dir.size(); i > 1; )
+      {
+      while( i > 0 && prev_dir[i-1] != '/' ) --i;
+      if( i == 0 ) break;
+      if( i > 1 ) --i;		// remove trailing slash except at root dir
+      prev_dir.resize( i );
+      if( stat( prev_dir.c_str(), &st2 ) != 0 || !S_ISDIR( st2.st_mode ) ||
+          ( st.st_ino == st2.st_ino && st.st_dev == st2.st_dev ) )
+        { loop = true; break; }
+      }
+  if( loop )			// full_name already visited or above tree
+    show_file_error( full_name.c_str(), "warning: Recursive directory loop." );
+  return !loop;			// (link to) directory
+  }
+
+
+/* Returns in input_filename the next filename, or "." for stdin.
+   ("." was chosen because it is not a valid filename).
+   Sets 'error' to true if a directory fails to open. */
+bool next_filename( std::list< std::string > & filenames,
+                    std::string & input_filename, bool & error,
+                    const int recursive, const bool ignore_stdin = false,
+                    const bool no_messages = false )
+  {
+  while( !filenames.empty() )
+    {
+    input_filename = filenames.front();
+    filenames.pop_front();
+    if( input_filename == "-" )
+      {
+      if( ignore_stdin ) continue;
+      input_filename = "."; return true;
+      }
+    struct stat st;
+    if( stat( input_filename.c_str(), &st ) == 0 && S_ISDIR( st.st_mode ) )
+      {
+      if( recursive )
+        {
+        DIR * const dirp = opendir( input_filename.c_str() );
+        if( !dirp )
+          {
+          if( !no_messages )
+            show_file_error( input_filename.c_str(), "Can't open directory", errno );
+          error = true; continue;
+          }
+        for( unsigned i = input_filename.size();
+             i > 1 && input_filename[i-1] == '/'; --i )
+          input_filename.resize( i - 1 );	// remove trailing slashes
+        struct stat stdot, *stdotp = 0;
+        if( input_filename[0] != '/' )		// relative path
+          {
+          if( input_filename == "." ) input_filename.clear();
+          if( stat( ".", &stdot ) == 0 && S_ISDIR( stdot.st_mode ) )
+            stdotp = &stdot;
+          }
+        if( input_filename.size() && input_filename != "/" )
+          input_filename += '/';
+        std::list< std::string > tmp_list;
+        while( true )
+          {
+          const struct dirent * const entryp = readdir( dirp );
+          if( !entryp ) { closedir( dirp ); break; }
+          const std::string tmp_name( entryp->d_name );
+          if( tmp_name == "." || tmp_name == ".." ) continue;
+          const std::string full_name( input_filename + tmp_name );
+          if( test_full_name( full_name, stdotp, recursive == 2 ) )
+            tmp_list.push_back( full_name );
+          }
+        filenames.splice( filenames.begin(), tmp_list );
+        }
+      continue;
+      }
+    return true;
+    }
+  input_filename.clear();
+  return false;
+  }
diff --git a/testsuite/check.sh b/testsuite/check.sh
new file mode 100755
index 0000000..6a6ef38
--- /dev/null
+++ b/testsuite/check.sh
@@ -0,0 +1,560 @@
+#! /bin/sh
+# check script for Zutils - Utilities dealing with compressed files
+# Copyright (C) 2009-2021 Antonio Diaz Diaz.
+#
+# This script is free software: you have unlimited permission
+# to copy, distribute, and modify it.
+
+LC_ALL=C
+export LC_ALL
+objdir=`pwd`
+testdir=`cd "$1" ; pwd`
+ZCAT="${objdir}"/zcat
+ZCMP="${objdir}"/zcmp
+ZDIFF="${objdir}"/zdiff
+ZGREP="${objdir}"/zgrep
+ZEGREP="${objdir}"/zegrep
+ZFGREP="${objdir}"/zfgrep
+ZTEST="${objdir}"/ztest
+ZUPDATE="${objdir}"/zupdate
+compressors="bzip2 gzip lzip"
+extensions="bz2 gz lz"
+compressor_needed() { echo "${compressors} are needed to run tests" ; exit 1 ; }
+framework_failure() { echo "failure in testing framework" ; exit 1 ; }
+
+if [ ! -f "${ZUPDATE}" ] || [ ! -x "${ZUPDATE}" ] ; then
+	echo "${ZUPDATE}: cannot execute"
+	exit 1
+fi
+
+[ -e "${ZUPDATE}" ] 2> /dev/null ||
+	{
+	echo "$0: a POSIX shell is required to run the tests"
+	echo "Try bash -c \"$0 $1 $2\""
+	exit 1
+	}
+
+if [ -d tmp ] ; then rm -rf tmp ; fi
+mkdir tmp
+cd "${objdir}"/tmp || framework_failure
+
+for i in ${compressors}; do
+	cat "${testdir}"/test.txt > in || framework_failure
+	$i in || compressor_needed
+	printf "Hello World!\n" > hello || framework_failure
+	$i hello || compressor_needed
+done
+
+cat "${testdir}"/test.txt > in || framework_failure
+cat "${testdir}"/test.txt.tar > in.tar || framework_failure
+printf "01234567890" > pin.tar || framework_failure
+cat in.tar in.tar in.tar in.tar >> pin.tar || framework_failure
+cat in > -in- || framework_failure
+cat in.lz > -in-.lz || framework_failure
+cat in.lz > lz_only.lz || framework_failure
+cat in in in in in in > in6 || framework_failure
+bad0_lz="${testdir}"/zero_bad_crc.lz
+bad0_gz="${testdir}"/zero_bad_crc.gz
+bad1_lz="${testdir}"/test_bad_crc.lz
+touch empty empty.bz2 empty.gz empty.lz
+fail=0
+test_failed() { fail=1 ; printf " $1" ; [ -z "$2" ] || printf "($2)" ; }
+
+printf "testing    zcat-%s..." "$2"
+
+for i in ${extensions}; do
+	"${ZCAT}" -N in.$i > copy || test_failed $LINENO $i
+	cmp in copy || test_failed $LINENO $i
+	"${ZCAT}" -N empty.$i in.$i > copy || test_failed $LINENO $i
+	cmp in copy || test_failed $LINENO $i
+	"${ZCAT}" -N --format=un in.$i > copy || test_failed $LINENO $i
+	cmp in copy || test_failed $LINENO $i
+	"${ZCAT}" -N --force-format=$i in.$i > copy || test_failed $LINENO $i
+	cmp in copy || test_failed $LINENO $i
+	"${ZCAT}" -N in.$i | dd bs=1000 count=1 > copy 2> /dev/null ||
+		test_failed $LINENO $i
+	dd if=in bs=1000 count=1 2> /dev/null | cmp - copy ||
+		test_failed $LINENO $i
+done
+
+printf "LZIP\001-.............................." | "${ZCAT}" -N > /dev/null 2>&1
+[ $? = 1 ] || test_failed $LINENO
+printf "LZIPxxxxxx" | "${ZCAT}" -N > /dev/null || test_failed $LINENO
+printf "BZh9xxxxxx" | "${ZCAT}" -N > /dev/null || test_failed $LINENO
+"${ZCAT}" -N -v -s "${testdir}"/zcat_vs.dat > /dev/null || test_failed $LINENO
+"${ZCAT}" -N < in > copy || test_failed $LINENO
+cmp in copy || test_failed $LINENO
+"${ZCAT}" -N < in.gz > copy || test_failed $LINENO
+cmp in copy || test_failed $LINENO
+"${ZCAT}" -N < in.bz2 > copy || test_failed $LINENO
+cmp in copy || test_failed $LINENO
+"${ZCAT}" -N < in.lz > copy || test_failed $LINENO
+cmp in copy || test_failed $LINENO
+"${ZCAT}" -N -O lz - - < in.lz > copy || test_failed $LINENO
+cmp in copy || test_failed $LINENO
+"${ZCAT}" -N --lz='lzip -q' < in.lz > copy || test_failed $LINENO
+cmp in copy || test_failed $LINENO
+"${ZCAT}" -N in > copy || test_failed $LINENO
+cmp in copy || test_failed $LINENO
+"${ZCAT}" -N lz_only > copy || test_failed $LINENO
+cmp in copy || test_failed $LINENO
+cat in.lz in in in in | "${ZCAT}" -N > copy || test_failed $LINENO	# tdata
+cmp in copy || test_failed $LINENO
+"${ZCAT}" -N in in.gz in.bz2 in.lz -- -in- -in-.lz > copy || test_failed $LINENO
+cmp in6 copy || test_failed $LINENO
+"${ZCAT}" -Nq in in.gz in.bz2 in.lz "${bad0_lz}" -- -in- -in-.lz > copy
+[ $? = 1 ] || test_failed $LINENO
+cmp in6 copy || test_failed $LINENO
+"${ZCAT}" -Nq "${bad1_lz}" -- -in-.lz in in.gz in.bz2 in.lz > copy
+[ $? = 1 ] || test_failed $LINENO
+cmp in6 copy || test_failed $LINENO
+"${ZCAT}" -N . || test_failed $LINENO
+"${ZCAT}" -N -r . > /dev/null || test_failed $LINENO
+"${ZCAT}" -N -r > /dev/null || test_failed $LINENO
+"${ZCAT}" -N -R . > /dev/null || test_failed $LINENO
+"${ZCAT}" -N -R > /dev/null || test_failed $LINENO
+
+"${ZCAT}" -Nq "" < in.lz > /dev/null
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -Nq --format=, in.lz
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -Nq --format=,lz in.lz
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -Nq --format=gz,,lz in.lz
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -Nq --format=lz,, in.lz
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -Nq --format=nc in.lz
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -Nq --lz='-lzip -q' in.lz
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -N --force-format=gz in.bz2 2> /dev/null
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -N --force-format=bz2 in.lz 2> /dev/null
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -Nq --force-format=lz in.gz
+[ $? = 1 ] || test_failed $LINENO
+"${ZCAT}" -N --bad-option 2> /dev/null
+[ $? = 1 ] || test_failed $LINENO
+
+
+printf "\ntesting    zcmp-%s..." "$2"
+
+for i in ${extensions}; do
+	"${ZCMP}" -N in.$i || test_failed $LINENO $i
+	"${ZCMP}" -N in in.$i || test_failed $LINENO $i
+	"${ZCMP}" -N in in.$i --force-format=,$i || test_failed $LINENO $i
+	"${ZCMP}" -N in.$i in || test_failed $LINENO $i
+	"${ZCMP}" -N in.$i in --force-format=$i || test_failed $LINENO $i
+	"${ZCMP}" -N -i 1kB:1000 -n 500 in6 in.$i || test_failed $LINENO $i
+	"${ZCMP}" -N -i 1KiB:1024 -n 50 in.$i in6 || test_failed $LINENO $i
+	"${ZCMP}" -N empty empty.$i || test_failed $LINENO $i
+done
+
+"${ZCMP}" -Nq in in6
+[ $? = 1 ] || test_failed $LINENO
+"${ZCMP}" -N -n 0 in in6 || test_failed $LINENO
+"${ZCMP}" -N -n 100B in in6 || test_failed $LINENO
+"${ZCMP}" -N -n 1k in in6 || test_failed $LINENO
+"${ZCMP}" -N -n 10kB in in6 || test_failed $LINENO
+"${ZCMP}" -Nq in.tar pin.tar
+[ $? = 1 ] || test_failed $LINENO
+"${ZCMP}" -Nq -i 0B:11B in.tar pin.tar
+[ $? = 1 ] || test_failed $LINENO
+"${ZCMP}" -N -i 0:11 -n 0 in.tar pin.tar || test_failed $LINENO
+"${ZCMP}" -N -i 0:11 -n 100 in.tar pin.tar || test_failed $LINENO
+"${ZCMP}" -N -i 0:11 -n 1Ki in.tar pin.tar || test_failed $LINENO
+"${ZCMP}" -N -i 0:11 -n 10KiB in.tar pin.tar || test_failed $LINENO
+"${ZCMP}" -N - - || test_failed $LINENO
+"${ZCMP}" -Nq -
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -N in in || test_failed $LINENO
+"${ZCMP}" -N in || test_failed $LINENO
+"${ZCMP}" -N --format=gz,bz2 in || test_failed $LINENO
+"${ZCMP}" -N --format=gz in || test_failed $LINENO
+"${ZCMP}" -N in.lz in.gz || test_failed $LINENO
+"${ZCMP}" -N --lz='lzip -q' in.lz in.gz || test_failed $LINENO
+"${ZCMP}" -N in.gz -- -in-.lz || test_failed $LINENO
+"${ZCMP}" -N -- -in-.lz in.gz || test_failed $LINENO
+"${ZCMP}" -N in -- -in-.lz || test_failed $LINENO
+"${ZCMP}" -N -- -in- in.lz || test_failed $LINENO
+"${ZCMP}" -N in.lz -- -in- || test_failed $LINENO
+"${ZCMP}" -N -- -in-.lz in || test_failed $LINENO
+"${ZCMP}" -N -- -in- in || test_failed $LINENO
+"${ZCMP}" -N in -- -in- || test_failed $LINENO
+"${ZCMP}" -N in.lz - < in || test_failed $LINENO
+"${ZCMP}" -N - in.lz < in || test_failed $LINENO
+"${ZCMP}" -N in - < in.lz || test_failed $LINENO
+"${ZCMP}" -N - in < in.lz || test_failed $LINENO
+"${ZCMP}" -N lz_only.lz - < in || test_failed $LINENO
+"${ZCMP}" -Nq lz_only.lz
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -Nq "" in
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -Nq --force-format=lz in.lz
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -Nq --force-format=lz in.gz in.lz
+r=$?
+{ [ $r = 1 ] || [ $r = 2 ] ; } || test_failed $LINENO
+"${ZCMP}" -Nq -i 100BB in in
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -Nq -i 100BB:100 in in
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -Nq -i 100: in in
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -Nq -n -1 in in
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -N -q -n 100BB in in
+[ $? = 2 ] || test_failed $LINENO
+"${ZCMP}" -N --bad-option in in 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+
+
+printf "\ntesting   zdiff-%s..." "$2"
+
+for i in ${extensions}; do
+	"${ZDIFF}" -N in.$i > /dev/null || test_failed $LINENO $i
+	"${ZDIFF}" -N in in.$i > /dev/null || test_failed $LINENO $i
+	"${ZDIFF}" -N in in.$i --force-format=,$i > /dev/null ||
+		test_failed $LINENO $i
+	"${ZDIFF}" -N in.$i in > /dev/null || test_failed $LINENO $i
+	"${ZDIFF}" -N in.$i in --force-format=$i, > /dev/null ||
+		test_failed $LINENO $i
+	"${ZDIFF}" -N empty empty.$i > /dev/null || test_failed $LINENO $i
+done
+
+"${ZDIFF}" -N in in6 > /dev/null
+[ $? = 1 ] || test_failed $LINENO
+# GNU diff 3.0 returns 2 when binary files differ
+"${ZDIFF}" -N in.tar pin.tar > /dev/null && test_failed $LINENO
+"${ZDIFF}" -N - - || test_failed $LINENO
+"${ZDIFF}" -N - 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZDIFF}" -N in in || test_failed $LINENO
+"${ZDIFF}" -N in || test_failed $LINENO
+"${ZDIFF}" -N --format=gz,bz2 in || test_failed $LINENO
+"${ZDIFF}" -N --format=gz in || test_failed $LINENO
+"${ZDIFF}" -N in.lz in.gz > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N --lz='lzip -q' in.lz in.gz > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N in.gz -- -in-.lz > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N -- -in-.lz in.gz > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N in -- -in-.lz > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N -- -in- in.lz > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N in.lz -- -in- > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N -- -in-.lz in > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N -- -in- in > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N in -- -in- > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N in.lz - < in > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N - in.lz < in > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N in - < in.lz > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N - in < in.lz > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N lz_only.lz - < in > /dev/null || test_failed $LINENO
+"${ZDIFF}" -N lz_only.lz 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZDIFF}" -N "" in 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZDIFF}" -N --bz2='-bzip2' in.bz2 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZDIFF}" -Nq --force-format=bz2 in.bz2 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZDIFF}" -N -q --force-format=,lz in.lz in.bz2 > /dev/null 2>&1
+r=$?
+{ [ $r = 1 ] || [ $r = 2 ] ; } || test_failed $LINENO
+"${ZDIFF}" -N --bad-option 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+
+mkdir tmp2
+cat in > tmp2/a || framework_failure
+cat in.lz > tmp2/a.lz || framework_failure
+"${ZDIFF}" -N --format=bz2 tmp2/a 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZDIFF}" -N --format=gz tmp2/a 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZDIFF}" -N --format=lz tmp2/a.lz 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZDIFF}" -N --format=lz tmp2/a || test_failed $LINENO
+"${ZDIFF}" -N --format=un tmp2/a.lz || test_failed $LINENO
+rm -r tmp2 || framework_failure
+
+
+printf "\ntesting   zgrep-%s..." "$2"
+
+for i in ${extensions}; do
+	"${ZGREP}" -N "GNU" in.$i > /dev/null || test_failed $LINENO $i
+	"${ZGREP}" -N "GNU" in.$i hello.$i > /dev/null || test_failed $LINENO $i
+	"${ZGREP}" -N "GNU" hello.$i in.$i > /dev/null || test_failed $LINENO $i
+	"${ZGREP}" -N -q "GNU" in.$i hello.$i || test_failed $LINENO $i
+	"${ZGREP}" -N -q "GNU" hello.$i in.$i || test_failed $LINENO $i
+	"${ZGREP}" -N "GNU" < in.$i > /dev/null || test_failed $LINENO $i
+	"${ZGREP}" -N -l "GNU" in.$i > /dev/null || test_failed $LINENO $i
+	"${ZGREP}" -N -L "GNU" in.$i || test_failed $LINENO $i
+	"${ZGREP}" -N --force-format=$i "GNU" in.$i > /dev/null ||
+		test_failed $LINENO $i
+	"${ZGREP}" -N -v "nx_pattern" in.$i > /dev/null ||
+		test_failed $LINENO $i
+	"${ZGREP}" -N "nx_pattern" in.$i && test_failed $LINENO $i
+	"${ZGREP}" -N -l "nx_pattern" in.$i && test_failed $LINENO $i
+	"${ZGREP}" -N -L "nx_pattern" in.$i > /dev/null &&
+		test_failed $LINENO $i
+	"${ZGREP}" -N --force-format=$i "GNU" in 2> /dev/null
+	[ $? = 2 ] || test_failed $LINENO $i
+	"${ZGREP}" -N "nx_pattern" empty.$i && test_failed $LINENO $i
+done
+
+"${ZGREP}" -N pin.tar -e "GNU" > /dev/null || test_failed $LINENO
+"${ZGREP}" -N "GNU" < pin.tar > /dev/null || test_failed $LINENO
+"${ZGREP}" -N -r "GNU" . > /dev/null || test_failed $LINENO
+"${ZGREP}" -N -r "GNU" > /dev/null || test_failed $LINENO
+"${ZGREP}" -N -R "GNU" . > /dev/null || test_failed $LINENO
+"${ZGREP}" -N -R "GNU" > /dev/null || test_failed $LINENO
+"${ZGREP}" -N "nx_pattern" -r . in > /dev/null && test_failed $LINENO
+"${ZGREP}" -N -e "GNU" in > /dev/null || test_failed $LINENO
+"${ZGREP}" -N "GNU" < in > /dev/null || test_failed $LINENO
+"${ZGREP}" -N -O lz "nx_pattern" - - < in.lz > /dev/null && test_failed $LINENO
+"${ZGREP}" -N -e "-free" --lz='lzip -q' < in.lz > /dev/null ||
+	test_failed $LINENO
+"${ZGREP}" -N -- "-free" -in- > /dev/null || test_failed $LINENO
+"${ZGREP}" -N -q -- "-free" nx_file -in-.lz || test_failed $LINENO
+"${ZGREP}" -N "GNU" in in.gz in.bz2 in.lz -- -in- > /dev/null ||
+	test_failed $LINENO
+"${ZGREP}" -N -l "GNU" in in.gz in.bz2 in.lz -- -in- > /dev/null ||
+	test_failed $LINENO
+"${ZGREP}" -N -L "GNU" in in.gz in.bz2 in.lz -- -in- || test_failed $LINENO
+"${ZGREP}" -N -l "nx_pattern" in in.gz in.bz2 in.lz -- -in- &&
+	test_failed $LINENO
+"${ZGREP}" -N -L "nx_pattern" in in.gz in.bz2 in.lz -- -in- > /dev/null &&
+	test_failed $LINENO
+"${ZGREP}" -Nq -l "01234567890" in "${bad1_lz}" in.lz && test_failed $LINENO
+"${ZGREP}" -Nq -l "01234567890" in "${bad1_lz}" in.lz pin.tar > /dev/null ||
+	test_failed $LINENO
+
+"${ZGREP}" -N "GNU" .
+[ $? = 1 ] || test_failed $LINENO
+"${ZGREP}" -N "GNU" "" < in.lz 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZGREP}" -N --bad-option 2> /dev/null
+[ $? = 2 ] || test_failed $LINENO
+"${ZGREP}" -N "GNU" -s nx_file
+[ $? = 2 ] || test_failed $LINENO
+"${ZGREP}" -N -q
+[ $? = 2 ] || test_failed $LINENO
+
+"${ZEGREP}" -N "GNU" in > /dev/null || test_failed $LINENO
+"${ZFGREP}" -N "GNU" in > /dev/null || test_failed $LINENO
+
+
+printf "\ntesting   ztest-%s..." "$2"
+
+for i in ${extensions}; do
+	"${ZTEST}" -N --force-format=$i < in.$i || test_failed $LINENO $i
+	"${ZTEST}" -N --force-format=$i < in 2> /dev/null
+	[ $? = 2 ] || test_failed $LINENO $i
+	"${ZTEST}" -N --force-format=$i in 2> /dev/null
+	[ $? = 2 ] || test_failed $LINENO $i
+done
+
+"${ZTEST}" -N in in.gz in.bz2 in.lz -- -in- || test_failed $LINENO
+"${ZTEST}" -N < in.gz || test_failed $LINENO
+"${ZTEST}" -N < in.bz2 || test_failed $LINENO
+"${ZTEST}" -N < in.lz || test_failed $LINENO
+"${ZTEST}" -N - < in.lz || test_failed $LINENO
+"${ZTEST}" -N - in.gz - < in.lz || test_failed $LINENO
+"${ZTEST}" -N --lz='lzip -q' < in.lz || test_failed $LINENO
+"${ZTEST}" -N -r . || test_failed $LINENO
+"${ZTEST}" -N -r || test_failed $LINENO
+"${ZTEST}" -N -R . || test_failed $LINENO
+"${ZTEST}" -N -R || test_failed $LINENO
+"${ZTEST}" -N empty empty.bz2 empty.gz empty.lz || test_failed $LINENO
+
+"${ZTEST}" -Nq in.gz "${bad0_lz}" in.bz2 "${bad1_lz}" in.lz
+[ $? = 2 ] || test_failed $LINENO
+lines=`"${ZTEST}" -N in.gz "${bad0_lz}" in.bz2 "${bad1_lz}" in.lz 2>&1 | wc -l`
+[ "${lines}" -eq 2 ] || test_failed $LINENO "${lines}"
+lines=`"${ZTEST}" -Nv in.gz "${bad0_lz}" in.bz2 "${bad1_lz}" in.lz 2>&1 | wc -l`
+[ "${lines}" -eq 6 ] || test_failed $LINENO "${lines}"
+"${ZTEST}" -Nq < in
+[ $? = 2 ] || test_failed $LINENO
+"${ZTEST}" -Nq "" < in.lz
+[ $? = 1 ] || test_failed $LINENO
+dd if=in.lz bs=1000 count=1 2> /dev/null | "${ZTEST}" -N -q
+[ $? = 2 ] || test_failed $LINENO
+"${ZTEST}" -Nq --force-format=lz in.bz2
+[ $? = 2 ] || test_failed $LINENO
+"${ZTEST}" -N --lz='lzip --bad-option' in.lz 2> /dev/null
+[ $? = 1 ] || test_failed $LINENO
+"${ZTEST}" -N --bad-option 2> /dev/null
+[ $? = 1 ] || test_failed $LINENO
+
+
+printf "\ntesting zupdate-%s..." "$2"
+
+"${ZUPDATE}" -N "" || test_failed $LINENO
+cat in.bz2 > a.bz2 || framework_failure
+cat in.gz > a.gz || framework_failure
+"${ZUPDATE}" -Nq --bz2=bad_command a.bz2
+[ $? = 1 ] || test_failed $LINENO
+"${ZUPDATE}" -N --bz2='bzip2 --bad-option' a.bz2 > /dev/null 2>&1
+[ $? = 1 ] || test_failed $LINENO
+"${ZUPDATE}" -Nq --gz=bad_command a.gz
+[ $? = 1 ] || test_failed $LINENO
+"${ZUPDATE}" -N --gz='gzip --bad-option' a.gz 2> /dev/null
+[ $? = 1 ] || test_failed $LINENO
+"${ZUPDATE}" -Nq --lz=bad_command a.gz
+[ $? = 1 ] || test_failed $LINENO
+"${ZUPDATE}" -N --lz='lzip --bad-option' a.gz 2> /dev/null
+[ $? = 1 ] || test_failed $LINENO
+"${ZUPDATE}" -N --bad-option 2> /dev/null
+[ $? = 1 ] || test_failed $LINENO
+
+cat in.lz in.lz > a.lz || framework_failure
+"${ZUPDATE}" -Nq -f a.bz2 a.gz
+[ $? = 1 ] || test_failed $LINENO
+[ -e a.bz2 ] || test_failed $LINENO
+[ -e a.gz ] || test_failed $LINENO
+[ -e a.lz ] || test_failed $LINENO
+rm -f a.lz || framework_failure
+
+"${ZUPDATE}" -N a.bz2 || test_failed $LINENO
+[ ! -e a.bz2 ] || test_failed $LINENO
+[ -e a.gz ] || test_failed $LINENO
+[ -e a.lz ] || test_failed $LINENO
+rm -f a.lz || framework_failure
+"${ZUPDATE}" -N a.gz || test_failed $LINENO
+[ ! -e a.bz2 ] || test_failed $LINENO
+[ ! -e a.gz ] || test_failed $LINENO
+[ -e a.lz ] || test_failed $LINENO
+rm -f a.lz || framework_failure
+
+cat in.bz2 > a.bz2 || framework_failure
+cat in.gz > a.gz || framework_failure
+"${ZUPDATE}" -Nq a.bz2 a.gz
+[ $? = 1 ] || test_failed $LINENO
+[ ! -e a.bz2 ] || test_failed $LINENO
+[ -e a.gz ] || test_failed $LINENO
+[ -e a.lz ] || test_failed $LINENO
+rm -f a.lz || framework_failure
+
+cat in.bz2 > a.bz2 || framework_failure
+cat in.gz > a.gz || framework_failure
+"${ZUPDATE}" -N -f -k a.bz2 a.gz || test_failed $LINENO
+[ -e a.bz2 ] || test_failed $LINENO
+[ -e a.gz ] || test_failed $LINENO
+[ -e a.lz ] || test_failed $LINENO
+rm -f a.lz || framework_failure
+
+cat in.bz2 > a.bz2 || framework_failure
+cat in.gz > a.gz || framework_failure
+"${ZUPDATE}" -N -f a.bz2 a.gz || test_failed $LINENO
+[ ! -e a.bz2 ] || test_failed $LINENO
+[ ! -e a.gz ] || test_failed $LINENO
+[ ! -e a ] || test_failed $LINENO
+[ -e a.lz ] || test_failed $LINENO
+rm -f a.lz || framework_failure
+
+cat in.bz2 > a.tbz || framework_failure		# keep combined extensions
+cat in.bz2 > b.tbz2 || framework_failure
+cat in.gz > c.tgz || framework_failure
+"${ZUPDATE}" -N a.tbz b.tbz2 c.tgz || test_failed $LINENO
+[ ! -e a.tbz ] || test_failed $LINENO
+[ ! -e b.tbz2 ] || test_failed $LINENO
+[ ! -e c.tgz ] || test_failed $LINENO
+[ ! -e a ] || test_failed $LINENO
+[ ! -e b ] || test_failed $LINENO
+[ ! -e c ] || test_failed $LINENO
+[ ! -e a.lz ] || test_failed $LINENO
+[ ! -e b.lz ] || test_failed $LINENO
+[ ! -e c.lz ] || test_failed $LINENO
+[ -e a.tlz ] || test_failed $LINENO
+[ -e b.tlz ] || test_failed $LINENO
+[ -e c.tlz ] || test_failed $LINENO
+rm -f a.tlz b.tlz c.tlz || framework_failure
+
+cat in.bz2 > a.bz2 || framework_failure
+cat "${bad0_gz}" > b.gz || framework_failure
+cat in.gz > c.gz || framework_failure
+"${ZUPDATE}" -N -f a.bz2 b.gz c.gz 2> /dev/null && test_failed $LINENO
+[ ! -e a.bz2 ] || test_failed $LINENO
+[ -e b.gz ] || test_failed $LINENO
+[ -e c.gz ] || test_failed $LINENO
+[ ! -e a ] || test_failed $LINENO
+[ ! -e b ] || test_failed $LINENO
+[ ! -e c ] || test_failed $LINENO
+[ -e a.lz ] || test_failed $LINENO
+rm -f a.lz b.gz c.gz || framework_failure
+
+cat in.bz2 > a.bz2 || framework_failure
+"${ZUPDATE}" -N -1 -q a.bz2 || test_failed $LINENO
+[ ! -e a.bz2 ] || test_failed $LINENO
+[ -e a.lz ] || test_failed $LINENO
+rm -f a.lz || framework_failure
+
+cat in.gz > 'name with spaces.gz' || framework_failure
+"${ZUPDATE}" -N -1 -q 'name with spaces.gz' || test_failed $LINENO
+[ ! -e 'name with spaces.gz' ] || test_failed $LINENO
+[ -e 'name with spaces.lz' ] || test_failed $LINENO
+"${ZCMP}" -N in 'name with spaces.lz' || test_failed $LINENO
+rm -f 'name with spaces.lz' || framework_failure
+
+mkdir tmp2
+mkdir tmp2/tmp3
+cat in.bz2 > tmp2/tmp3/a.bz2 || framework_failure
+cat in.gz > tmp2/tmp3/a.gz || framework_failure
+"${ZUPDATE}" -N -r --format=gz tmp2 || test_failed $LINENO
+[ -e tmp2/tmp3/a.bz2 ] || test_failed $LINENO
+[ ! -e tmp2/tmp3/a.gz ] || test_failed $LINENO
+[ -e tmp2/tmp3/a.lz ] || test_failed $LINENO
+rm -f tmp2/tmp3/a.lz || framework_failure
+"${ZUPDATE}" -N -r --format=bz2 tmp2 || test_failed $LINENO
+[ ! -e tmp2/tmp3/a.bz2 ] || test_failed $LINENO
+[ ! -e tmp2/tmp3/a.gz ] || test_failed $LINENO
+[ -e tmp2/tmp3/a.lz ] || test_failed $LINENO
+rm -f tmp2/tmp3/a.lz || framework_failure
+cat in.bz2 > tmp2/tmp3/a.bz2 || framework_failure
+cat in.gz > tmp2/tmp3/a.gz || framework_failure
+cd tmp2 || framework_failure
+"${ZUPDATE}" -N -r -k -f . || test_failed $LINENO
+[ -e tmp3/a.bz2 ] || test_failed $LINENO
+[ -e tmp3/a.gz ] || test_failed $LINENO
+[ -e tmp3/a.lz ] || test_failed $LINENO
+rm -f tmp3/a.lz || framework_failure
+"${ZUPDATE}" -N -r -k -f || test_failed $LINENO
+[ -e tmp3/a.bz2 ] || test_failed $LINENO
+[ -e tmp3/a.gz ] || test_failed $LINENO
+[ -e tmp3/a.lz ] || test_failed $LINENO
+rm -f tmp3/a.lz || framework_failure
+"${ZUPDATE}" -N -R -k -f . || test_failed $LINENO
+[ -e tmp3/a.bz2 ] || test_failed $LINENO
+[ -e tmp3/a.gz ] || test_failed $LINENO
+[ -e tmp3/a.lz ] || test_failed $LINENO
+rm -f tmp3/a.lz || framework_failure
+"${ZUPDATE}" -N -R -k -f || test_failed $LINENO
+[ -e tmp3/a.bz2 ] || test_failed $LINENO
+[ -e tmp3/a.gz ] || test_failed $LINENO
+[ -e tmp3/a.lz ] || test_failed $LINENO
+rm -f tmp3/a.lz || framework_failure
+"${ZUPDATE}" -N -r -f . || test_failed $LINENO
+[ ! -e tmp3/a.bz2 ] || test_failed $LINENO
+[ ! -e tmp3/a.gz ] || test_failed $LINENO
+[ -e tmp3/a.lz ] || test_failed $LINENO
+cd .. || framework_failure
+rm -r tmp2 || framework_failure
+
+rm -f empty empty.bz2 empty.gz empty.lz || framework_failure
+if ln -s '.' slink 2> /dev/null ; then
+	"${ZCAT}" -N -r slink > /dev/null || test_failed $LINENO
+	"${ZGREP}" -N -r "GNU" slink > /dev/null || test_failed $LINENO
+	"${ZTEST}" -N -r slink || test_failed $LINENO
+	"${ZUPDATE}" -N -r -f slink || test_failed $LINENO
+else
+	printf "\nwarning: skipping link test: 'ln' does not work on your system."
+fi
+rm -f slink || framework_failure
+
+echo
+if [ ${fail} = 0 ] ; then
+	echo "tests completed successfully."
+	cd "${objdir}" && rm -r tmp
+else
+	echo "tests failed."
+fi
+exit ${fail}
diff --git a/testsuite/test.txt b/testsuite/test.txt
new file mode 100644
index 0000000..9196a3a
--- /dev/null
+++ b/testsuite/test.txt
@@ -0,0 +1,676 @@
+                    GNU GENERAL PUBLIC LICENSE
+                       Version 2, June 1991
+
+ Copyright (C) 1989, 1991 Free Software Foundation, Inc.,
+ 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+                            Preamble
+
+  The licenses for most software are designed to take away your
+freedom to share and change it.  By contrast, the GNU General Public
+License is intended to guarantee your freedom to share and change free
+software--to make sure the software is free for all its users.  This
+General Public License applies to most of the Free Software
+Foundation's software and to any other program whose authors commit to
+using it.  (Some other Free Software Foundation software is covered by
+the GNU Lesser General Public License instead.)  You can apply it to
+your programs, too.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+this service if you wish), that you receive source code or can get it
+if you want it, that you can change the software or use pieces of it
+in new free programs; and that you know you can do these things.
+
+  To protect your rights, we need to make restrictions that forbid
+anyone to deny you these rights or to ask you to surrender the rights.
+These restrictions translate to certain responsibilities for you if you
+distribute copies of the software, or if you modify it.
+
+  For example, if you distribute copies of such a program, whether
+gratis or for a fee, you must give the recipients all the rights that
+you have.  You must make sure that they, too, receive or can get the
+source code.  And you must show them these terms so they know their
+rights.
+
+  We protect your rights with two steps: (1) copyright the software, and
+(2) offer you this license which gives you legal permission to copy,
+distribute and/or modify the software.
+
+  Also, for each author's protection and ours, we want to make certain
+that everyone understands that there is no warranty for this free
+software.  If the software is modified by someone else and passed on, we
+want its recipients to know that what they have is not the original, so
+that any problems introduced by others will not reflect on the original
+authors' reputations.
+
+  Finally, any free program is threatened constantly by software
+patents.  We wish to avoid the danger that redistributors of a free
+program will individually obtain patent licenses, in effect making the
+program proprietary.  To prevent this, we have made it clear that any
+patent must be licensed for everyone's free use or not licensed at all.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+                    GNU GENERAL PUBLIC LICENSE
+   TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+
+  0. This License applies to any program or other work which contains
+a notice placed by the copyright holder saying it may be distributed
+under the terms of this General Public License.  The "Program", below,
+refers to any such program or work, and a "work based on the Program"
+means either the Program or any derivative work under copyright law:
+that is to say, a work containing the Program or a portion of it,
+either verbatim or with modifications and/or translated into another
+language.  (Hereinafter, translation is included without limitation in
+the term "modification".)  Each licensee is addressed as "you".
+
+Activities other than copying, distribution and modification are not
+covered by this License; they are outside its scope.  The act of
+running the Program is not restricted, and the output from the Program
+is covered only if its contents constitute a work based on the
+Program (independent of having been made by running the Program).
+Whether that is true depends on what the Program does.
+
+  1. You may copy and distribute verbatim copies of the Program's
+source code as you receive it, in any medium, provided that you
+conspicuously and appropriately publish on each copy an appropriate
+copyright notice and disclaimer of warranty; keep intact all the
+notices that refer to this License and to the absence of any warranty;
+and give any other recipients of the Program a copy of this License
+along with the Program.
+
+You may charge a fee for the physical act of transferring a copy, and
+you may at your option offer warranty protection in exchange for a fee.
+
+  2. You may modify your copy or copies of the Program or any portion
+of it, thus forming a work based on the Program, and copy and
+distribute such modifications or work under the terms of Section 1
+above, provided that you also meet all of these conditions:
+
+    a) You must cause the modified files to carry prominent notices
+    stating that you changed the files and the date of any change.
+
+    b) You must cause any work that you distribute or publish, that in
+    whole or in part contains or is derived from the Program or any
+    part thereof, to be licensed as a whole at no charge to all third
+    parties under the terms of this License.
+
+    c) If the modified program normally reads commands interactively
+    when run, you must cause it, when started running for such
+    interactive use in the most ordinary way, to print or display an
+    announcement including an appropriate copyright notice and a
+    notice that there is no warranty (or else, saying that you provide
+    a warranty) and that users may redistribute the program under
+    these conditions, and telling the user how to view a copy of this
+    License.  (Exception: if the Program itself is interactive but
+    does not normally print such an announcement, your work based on
+    the Program is not required to print an announcement.)
+
+These requirements apply to the modified work as a whole.  If
+identifiable sections of that work are not derived from the Program,
+and can be reasonably considered independent and separate works in
+themselves, then this License, and its terms, do not apply to those
+sections when you distribute them as separate works.  But when you
+distribute the same sections as part of a whole which is a work based
+on the Program, the distribution of the whole must be on the terms of
+this License, whose permissions for other licensees extend to the
+entire whole, and thus to each and every part regardless of who wrote it.
+
+Thus, it is not the intent of this section to claim rights or contest
+your rights to work written entirely by you; rather, the intent is to
+exercise the right to control the distribution of derivative or
+collective works based on the Program.
+
+In addition, mere aggregation of another work not based on the Program
+with the Program (or with a work based on the Program) on a volume of
+a storage or distribution medium does not bring the other work under
+the scope of this License.
+
+  3. You may copy and distribute the Program (or a work based on it,
+under Section 2) in object code or executable form under the terms of
+Sections 1 and 2 above provided that you also do one of the following:
+
+    a) Accompany it with the complete corresponding machine-readable
+    source code, which must be distributed under the terms of Sections
+    1 and 2 above on a medium customarily used for software interchange; or,
+
+    b) Accompany it with a written offer, valid for at least three
+    years, to give any third party, for a charge no more than your
+    cost of physically performing source distribution, a complete
+    machine-readable copy of the corresponding source code, to be
+    distributed under the terms of Sections 1 and 2 above on a medium
+    customarily used for software interchange; or,
+
+    c) Accompany it with the information you received as to the offer
+    to distribute corresponding source code.  (This alternative is
+    allowed only for noncommercial distribution and only if you
+    received the program in object code or executable form with such
+    an offer, in accord with Subsection b above.)
+
+The source code for a work means the preferred form of the work for
+making modifications to it.  For an executable work, complete source
+code means all the source code for all modules it contains, plus any
+associated interface definition files, plus the scripts used to
+control compilation and installation of the executable.  However, as a
+special exception, the source code distributed need not include
+anything that is normally distributed (in either source or binary
+form) with the major components (compiler, kernel, and so on) of the
+operating system on which the executable runs, unless that component
+itself accompanies the executable.
+
+If distribution of executable or object code is made by offering
+access to copy from a designated place, then offering equivalent
+access to copy the source code from the same place counts as
+distribution of the source code, even though third parties are not
+compelled to copy the source along with the object code.
+
+  4. You may not copy, modify, sublicense, or distribute the Program
+except as expressly provided under this License.  Any attempt
+otherwise to copy, modify, sublicense or distribute the Program is
+void, and will automatically terminate your rights under this License.
+However, parties who have received copies, or rights, from you under
+this License will not have their licenses terminated so long as such
+parties remain in full compliance.
+
+  5. You are not required to accept this License, since you have not
+signed it.  However, nothing else grants you permission to modify or
+distribute the Program or its derivative works.  These actions are
+prohibited by law if you do not accept this License.  Therefore, by
+modifying or distributing the Program (or any work based on the
+Program), you indicate your acceptance of this License to do so, and
+all its terms and conditions for copying, distributing or modifying
+the Program or works based on it.
+
+  6. Each time you redistribute the Program (or any work based on the
+Program), the recipient automatically receives a license from the
+original licensor to copy, distribute or modify the Program subject to
+these terms and conditions.  You may not impose any further
+restrictions on the recipients' exercise of the rights granted herein.
+You are not responsible for enforcing compliance by third parties to
+this License.
+
+  7. If, as a consequence of a court judgment or allegation of patent
+infringement or for any other reason (not limited to patent issues),
+conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot
+distribute so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you
+may not distribute the Program at all.  For example, if a patent
+license would not permit royalty-free redistribution of the Program by
+all those who receive copies directly or indirectly through you, then
+the only way you could satisfy both it and this License would be to
+refrain entirely from distribution of the Program.
+
+If any portion of this section is held invalid or unenforceable under
+any particular circumstance, the balance of the section is intended to
+apply and the section as a whole is intended to apply in other
+circumstances.
+
+It is not the purpose of this section to induce you to infringe any
+patents or other property right claims or to contest validity of any
+such claims; this section has the sole purpose of protecting the
+integrity of the free software distribution system, which is
+implemented by public license practices.  Many people have made
+generous contributions to the wide range of software distributed
+through that system in reliance on consistent application of that
+system; it is up to the author/donor to decide if he or she is willing
+to distribute software through any other system and a licensee cannot
+impose that choice.
+
+This section is intended to make thoroughly clear what is believed to
+be a consequence of the rest of this License.
+
+  8. If the distribution and/or use of the Program is restricted in
+certain countries either by patents or by copyrighted interfaces, the
+original copyright holder who places the Program under this License
+may add an explicit geographical distribution limitation excluding
+those countries, so that distribution is permitted only in or among
+countries not thus excluded.  In such case, this License incorporates
+the limitation as if written in the body of this License.
+
+  9. The Free Software Foundation may publish revised and/or new versions
+of the General Public License from time to time.  Such new versions will
+be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+Each version is given a distinguishing version number.  If the Program
+specifies a version number of this License which applies to it and "any
+later version", you have the option of following the terms and conditions
+either of that version or of any later version published by the Free
+Software Foundation.  If the Program does not specify a version number of
+this License, you may choose any version ever published by the Free Software
+Foundation.
+
+  10. If you wish to incorporate parts of the Program into other free
+programs whose distribution conditions are different, write to the author
+to ask for permission.  For software which is copyrighted by the Free
+Software Foundation, write to the Free Software Foundation; we sometimes
+make exceptions for this.  Our decision will be guided by the two goals
+of preserving the free status of all derivatives of our free software and
+of promoting the sharing and reuse of software generally.
+
+                            NO WARRANTY
+
+  11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
+FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
+OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
+PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
+OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS
+TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
+PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
+REPAIR OR CORRECTION.
+
+  12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
+REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
+INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
+OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
+TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
+YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
+PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGES.
+
+                     END OF TERMS AND CONDITIONS
+
+            How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+convey the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the program's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation, either version 2 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+Also add information on how to contact you by electronic and paper mail.
+
+If the program is interactive, make it output a short notice like this
+when it starts in an interactive mode:
+
+    Gnomovision version 69, Copyright (C) <year>  <name of author>
+    Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
+    This is free software, and you are welcome to redistribute it
+    under certain conditions; type `show c' for details.
+
+The hypothetical commands `show w' and `show c' should show the appropriate
+parts of the General Public License.  Of course, the commands you use may
+be called something other than `show w' and `show c'; they could even be
+mouse-clicks or menu items--whatever suits your program.
+
+You should also get your employer (if you work as a programmer) or your
+school, if any, to sign a "copyright disclaimer" for the program, if
+necessary.  Here is a sample; alter the names:
+
+  Yoyodyne, Inc., hereby disclaims all copyright interest in the program
+  `Gnomovision' (which makes passes at compilers) written by James Hacker.
+
+  <signature of Ty Coon>, 1 April 1989
+  Ty Coon, President of Vice
+
+This General Public License does not permit incorporating your program into
+proprietary programs.  If your program is a subroutine library, you may
+consider it more useful to permit linking proprietary applications with the
+library.  If this is what you want to do, use the GNU Lesser General
+Public License instead of this License.
+                    GNU GENERAL PUBLIC LICENSE

+                       Version 2, June 1991

+

+ Copyright (C) 1989, 1991 Free Software Foundation, Inc.,

+ 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

+ Everyone is permitted to copy and distribute verbatim copies

+ of this license document, but changing it is not allowed.

+

+                            Preamble

+

+  The licenses for most software are designed to take away your

+freedom to share and change it.  By contrast, the GNU General Public

+License is intended to guarantee your freedom to share and change free

+software--to make sure the software is free for all its users.  This

+General Public License applies to most of the Free Software

+Foundation's software and to any other program whose authors commit to

+using it.  (Some other Free Software Foundation software is covered by

+the GNU Lesser General Public License instead.)  You can apply it to

+your programs, too.

+

+  When we speak of free software, we are referring to freedom, not

+price.  Our General Public Licenses are designed to make sure that you

+have the freedom to distribute copies of free software (and charge for

+this service if you wish), that you receive source code or can get it

+if you want it, that you can change the software or use pieces of it

+in new free programs; and that you know you can do these things.

+

+  To protect your rights, we need to make restrictions that forbid

+anyone to deny you these rights or to ask you to surrender the rights.

+These restrictions translate to certain responsibilities for you if you

+distribute copies of the software, or if you modify it.

+

+  For example, if you distribute copies of such a program, whether

+gratis or for a fee, you must give the recipients all the rights that

+you have.  You must make sure that they, too, receive or can get the

+source code.  And you must show them these terms so they know their

+rights.

+

+  We protect your rights with two steps: (1) copyright the software, and

+(2) offer you this license which gives you legal permission to copy,

+distribute and/or modify the software.

+

+  Also, for each author's protection and ours, we want to make certain

+that everyone understands that there is no warranty for this free

+software.  If the software is modified by someone else and passed on, we

+want its recipients to know that what they have is not the original, so

+that any problems introduced by others will not reflect on the original

+authors' reputations.

+

+  Finally, any free program is threatened constantly by software

+patents.  We wish to avoid the danger that redistributors of a free

+program will individually obtain patent licenses, in effect making the

+program proprietary.  To prevent this, we have made it clear that any

+patent must be licensed for everyone's free use or not licensed at all.

+

+  The precise terms and conditions for copying, distribution and

+modification follow.

+

+                    GNU GENERAL PUBLIC LICENSE

+   TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION

+

+  0. This License applies to any program or other work which contains

+a notice placed by the copyright holder saying it may be distributed

+under the terms of this General Public License.  The "Program", below,

+refers to any such program or work, and a "work based on the Program"

+means either the Program or any derivative work under copyright law:

+that is to say, a work containing the Program or a portion of it,

+either verbatim or with modifications and/or translated into another

+language.  (Hereinafter, translation is included without limitation in

+the term "modification".)  Each licensee is addressed as "you".

+

+Activities other than copying, distribution and modification are not

+covered by this License; they are outside its scope.  The act of

+running the Program is not restricted, and the output from the Program

+is covered only if its contents constitute a work based on the

+Program (independent of having been made by running the Program).

+Whether that is true depends on what the Program does.

+

+  1. You may copy and distribute verbatim copies of the Program's

+source code as you receive it, in any medium, provided that you

+conspicuously and appropriately publish on each copy an appropriate

+copyright notice and disclaimer of warranty; keep intact all the

+notices that refer to this License and to the absence of any warranty;

+and give any other recipients of the Program a copy of this License

+along with the Program.

+

+You may charge a fee for the physical act of transferring a copy, and

+you may at your option offer warranty protection in exchange for a fee.

+

+  2. You may modify your copy or copies of the Program or any portion

+of it, thus forming a work based on the Program, and copy and

+distribute such modifications or work under the terms of Section 1

+above, provided that you also meet all of these conditions:

+

+    a) You must cause the modified files to carry prominent notices

+    stating that you changed the files and the date of any change.

+

+    b) You must cause any work that you distribute or publish, that in

+    whole or in part contains or is derived from the Program or any

+    part thereof, to be licensed as a whole at no charge to all third

+    parties under the terms of this License.

+

+    c) If the modified program normally reads commands interactively

+    when run, you must cause it, when started running for such

+    interactive use in the most ordinary way, to print or display an

+    announcement including an appropriate copyright notice and a

+    notice that there is no warranty (or else, saying that you provide

+    a warranty) and that users may redistribute the program under

+    these conditions, and telling the user how to view a copy of this

+    License.  (Exception: if the Program itself is interactive but

+    does not normally print such an announcement, your work based on

+    the Program is not required to print an announcement.)

+

+These requirements apply to the modified work as a whole.  If

+identifiable sections of that work are not derived from the Program,

+and can be reasonably considered independent and separate works in

+themselves, then this License, and its terms, do not apply to those

+sections when you distribute them as separate works.  But when you

+distribute the same sections as part of a whole which is a work based

+on the Program, the distribution of the whole must be on the terms of

+this License, whose permissions for other licensees extend to the

+entire whole, and thus to each and every part regardless of who wrote it.

+

+Thus, it is not the intent of this section to claim rights or contest

+your rights to work written entirely by you; rather, the intent is to

+exercise the right to control the distribution of derivative or

+collective works based on the Program.

+

+In addition, mere aggregation of another work not based on the Program

+with the Program (or with a work based on the Program) on a volume of

+a storage or distribution medium does not bring the other work under

+the scope of this License.

+

+  3. You may copy and distribute the Program (or a work based on it,

+under Section 2) in object code or executable form under the terms of

+Sections 1 and 2 above provided that you also do one of the following:

+

+    a) Accompany it with the complete corresponding machine-readable

+    source code, which must be distributed under the terms of Sections

+    1 and 2 above on a medium customarily used for software interchange; or,

+

+    b) Accompany it with a written offer, valid for at least three

+    years, to give any third party, for a charge no more than your

+    cost of physically performing source distribution, a complete

+    machine-readable copy of the corresponding source code, to be

+    distributed under the terms of Sections 1 and 2 above on a medium

+    customarily used for software interchange; or,

+

+    c) Accompany it with the information you received as to the offer

+    to distribute corresponding source code.  (This alternative is

+    allowed only for noncommercial distribution and only if you

+    received the program in object code or executable form with such

+    an offer, in accord with Subsection b above.)

+

+The source code for a work means the preferred form of the work for

+making modifications to it.  For an executable work, complete source

+code means all the source code for all modules it contains, plus any

+associated interface definition files, plus the scripts used to

+control compilation and installation of the executable.  However, as a

+special exception, the source code distributed need not include

+anything that is normally distributed (in either source or binary

+form) with the major components (compiler, kernel, and so on) of the

+operating system on which the executable runs, unless that component

+itself accompanies the executable.

+

+If distribution of executable or object code is made by offering

+access to copy from a designated place, then offering equivalent

+access to copy the source code from the same place counts as

+distribution of the source code, even though third parties are not

+compelled to copy the source along with the object code.

+

+  4. You may not copy, modify, sublicense, or distribute the Program

+except as expressly provided under this License.  Any attempt

+otherwise to copy, modify, sublicense or distribute the Program is

+void, and will automatically terminate your rights under this License.

+However, parties who have received copies, or rights, from you under

+this License will not have their licenses terminated so long as such

+parties remain in full compliance.

+

+  5. You are not required to accept this License, since you have not

+signed it.  However, nothing else grants you permission to modify or

+distribute the Program or its derivative works.  These actions are

+prohibited by law if you do not accept this License.  Therefore, by

+modifying or distributing the Program (or any work based on the

+Program), you indicate your acceptance of this License to do so, and

+all its terms and conditions for copying, distributing or modifying

+the Program or works based on it.

+

+  6. Each time you redistribute the Program (or any work based on the

+Program), the recipient automatically receives a license from the

+original licensor to copy, distribute or modify the Program subject to

+these terms and conditions.  You may not impose any further

+restrictions on the recipients' exercise of the rights granted herein.

+You are not responsible for enforcing compliance by third parties to

+this License.

+

+  7. If, as a consequence of a court judgment or allegation of patent

+infringement or for any other reason (not limited to patent issues),

+conditions are imposed on you (whether by court order, agreement or

+otherwise) that contradict the conditions of this License, they do not

+excuse you from the conditions of this License.  If you cannot

+distribute so as to satisfy simultaneously your obligations under this

+License and any other pertinent obligations, then as a consequence you

+may not distribute the Program at all.  For example, if a patent

+license would not permit royalty-free redistribution of the Program by

+all those who receive copies directly or indirectly through you, then

+the only way you could satisfy both it and this License would be to

+refrain entirely from distribution of the Program.

+

+If any portion of this section is held invalid or unenforceable under

+any particular circumstance, the balance of the section is intended to

+apply and the section as a whole is intended to apply in other

+circumstances.

+

+It is not the purpose of this section to induce you to infringe any

+patents or other property right claims or to contest validity of any

+such claims; this section has the sole purpose of protecting the

+integrity of the free software distribution system, which is

+implemented by public license practices.  Many people have made

+generous contributions to the wide range of software distributed

+through that system in reliance on consistent application of that

+system; it is up to the author/donor to decide if he or she is willing

+to distribute software through any other system and a licensee cannot

+impose that choice.

+

+This section is intended to make thoroughly clear what is believed to

+be a consequence of the rest of this License.

+

+  8. If the distribution and/or use of the Program is restricted in

+certain countries either by patents or by copyrighted interfaces, the

+original copyright holder who places the Program under this License

+may add an explicit geographical distribution limitation excluding

+those countries, so that distribution is permitted only in or among

+countries not thus excluded.  In such case, this License incorporates

+the limitation as if written in the body of this License.

+

+  9. The Free Software Foundation may publish revised and/or new versions

+of the General Public License from time to time.  Such new versions will

+be similar in spirit to the present version, but may differ in detail to

+address new problems or concerns.

+

+Each version is given a distinguishing version number.  If the Program

+specifies a version number of this License which applies to it and "any

+later version", you have the option of following the terms and conditions

+either of that version or of any later version published by the Free

+Software Foundation.  If the Program does not specify a version number of

+this License, you may choose any version ever published by the Free Software

+Foundation.

+

+  10. If you wish to incorporate parts of the Program into other free

+programs whose distribution conditions are different, write to the author

+to ask for permission.  For software which is copyrighted by the Free

+Software Foundation, write to the Free Software Foundation; we sometimes

+make exceptions for this.  Our decision will be guided by the two goals

+of preserving the free status of all derivatives of our free software and

+of promoting the sharing and reuse of software generally.

+

+                            NO WARRANTY

+

+  11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY

+FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN

+OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES

+PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED

+OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF

+MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS

+TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE

+PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,

+REPAIR OR CORRECTION.

+

+  12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING

+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR

+REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,

+INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING

+OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED

+TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY

+YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER

+PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE

+POSSIBILITY OF SUCH DAMAGES.

+

+                     END OF TERMS AND CONDITIONS

+

+            How to Apply These Terms to Your New Programs

+

+  If you develop a new program, and you want it to be of the greatest

+possible use to the public, the best way to achieve this is to make it

+free software which everyone can redistribute and change under these terms.

+

+  To do so, attach the following notices to the program.  It is safest

+to attach them to the start of each source file to most effectively

+convey the exclusion of warranty; and each file should have at least

+the "copyright" line and a pointer to where the full notice is found.

+

+    <one line to give the program's name and a brief idea of what it does.>

+    Copyright (C) <year>  <name of author>

+

+    This program is free software: you can redistribute it and/or modify

+    it under the terms of the GNU General Public License as published by

+    the Free Software Foundation, either version 2 of the License, or

+    (at your option) any later version.

+

+    This program is distributed in the hope that it will be useful,

+    but WITHOUT ANY WARRANTY; without even the implied warranty of

+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

+    GNU General Public License for more details.

+

+    You should have received a copy of the GNU General Public License

+    along with this program.  If not, see <http://www.gnu.org/licenses/>.

+

+Also add information on how to contact you by electronic and paper mail.

+

+If the program is interactive, make it output a short notice like this

+when it starts in an interactive mode:

+

+    Gnomovision version 69, Copyright (C) <year>  <name of author>

+    Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.

+    This is free software, and you are welcome to redistribute it

+    under certain conditions; type `show c' for details.

+

+The hypothetical commands `show w' and `show c' should show the appropriate

+parts of the General Public License.  Of course, the commands you use may

+be called something other than `show w' and `show c'; they could even be

+mouse-clicks or menu items--whatever suits your program.

+

+You should also get your employer (if you work as a programmer) or your

+school, if any, to sign a "copyright disclaimer" for the program, if

+necessary.  Here is a sample; alter the names:

+

+  Yoyodyne, Inc., hereby disclaims all copyright interest in the program

+  `Gnomovision' (which makes passes at compilers) written by James Hacker.

+

+  <signature of Ty Coon>, 1 April 1989

+  Ty Coon, President of Vice

+

+This General Public License does not permit incorporating your program into

+proprietary programs.  If your program is a subroutine library, you may

+consider it more useful to permit linking proprietary applications with the

+library.  If this is what you want to do, use the GNU Lesser General

+Public License instead of this License.

diff --git a/testsuite/test.txt.tar b/testsuite/test.txt.tar
new file mode 100644
index 0000000..92d6f1b
--- /dev/null
+++ b/testsuite/test.txt.tar
diff --git a/testsuite/test_bad_crc.lz b/testsuite/test_bad_crc.lz
new file mode 100644
index 0000000..c7d5bc9
--- /dev/null
+++ b/testsuite/test_bad_crc.lz
diff --git a/testsuite/zcat_vs.dat b/testsuite/zcat_vs.dat
new file mode 100644
index 0000000..29978fd
--- /dev/null
+++ b/testsuite/zcat_vs.dat
@@ -0,0 +1,68 @@
+Worst case test file for zcat -vs.
+First 4096 input bytes produce 4095 output bytes because of -s.
+Next 4096 input bytes produce 16384 output bytes, accumulating a total
+of 20479 bytes in the output buffer.
+----------------------------------------------
+
+
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+...............................................................
+€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€
\ No newline at end of file
diff --git a/testsuite/zero_bad_crc.gz b/testsuite/zero_bad_crc.gz
new file mode 100644
index 0000000..a2a9991
--- /dev/null
+++ b/testsuite/zero_bad_crc.gz
diff --git a/testsuite/zero_bad_crc.lz b/testsuite/zero_bad_crc.lz
new file mode 100644
index 0000000..0d3cc93
--- /dev/null
+++ b/testsuite/zero_bad_crc.lz
diff --git a/zcat.cc b/zcat.cc
new file mode 100644
index 0000000..62b93aa
--- /dev/null
+++ b/zcat.cc
@@ -0,0 +1,386 @@
+/* Zcat - decompress and concatenate files to standard output
+   Copyright (C) 2010-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#define _FILE_OFFSET_BITS 64
+
+#include <cerrno>
+#include <climits>
+#include <csignal>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <list>
+#include <string>
+#include <vector>
+#include <dirent.h>
+#include <fcntl.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <sys/stat.h>
+#if defined(__MSVCRT__) || defined(__OS2__)
+#include <io.h>
+#endif
+
+#include "arg_parser.h"
+#include "rc.h"
+#include "zutils.h"
+
+
+namespace {
+
+#include "recursive.cc"
+#include "zcatgrep.cc"
+
+struct Cat_options
+  {
+  int number_lines;		// 0 = no, 1 = nonblank, 2 = all
+  bool show_ends;
+  bool show_nonprinting;
+  bool show_tabs;
+  bool squeeze_blank;
+
+  Cat_options()
+    : number_lines( 0 ), show_ends( false ), show_nonprinting( false ),
+      show_tabs( false ), squeeze_blank( false ) {}
+  };
+
+
+class Line_number		// unlimited size line counter
+  {
+  std::string str;
+  unsigned first_digit_pos;
+
+public:
+  Line_number() : str( "     0\t" ), first_digit_pos( 5 ) {}
+
+  void next()
+    {
+    for( unsigned i = str.size() - 1; i > first_digit_pos; )
+      {
+      if( str[--i] < '9' ) { ++str[i]; return; }
+      str[i] = '0';
+      }
+    if( first_digit_pos > 0 ) str[--first_digit_pos] = '1';
+    else str.insert( str.begin() + first_digit_pos, '1' );
+    }
+
+  int sprint( uint8_t * const buf )
+    {
+    std::memcpy( buf, str.c_str(), str.size() );
+    return str.size();
+    }
+  };
+
+Line_number line_number;
+
+
+void show_help()
+  {
+  std::printf( "zcat copies each file argument to standard output in sequence. If any\n"
+               "file given is compressed, its decompressed content is copied. If a file\n"
+               "given does not exist, and its name does not end with one of the known\n"
+               "extensions, zcat tries the compressed file names corresponding to the\n"
+               "formats supported. If a file fails to decompress, zcat continues copying the\n"
+               "rest of the files.\n"
+               "\nIf a file is specified as '-', data are read from standard input,\n"
+               "decompressed if needed, and sent to standard output. Data read from\n"
+               "standard input must be of the same type; all uncompressed or all in the\n"
+               "same compressed format.\n"
+               "\nIf no files are specified, recursive searches examine the current\n"
+               "working directory, and nonrecursive searches read standard input.\n"
+               "\nThe formats supported are bzip2, gzip, lzip, and xz.\n"
+               "\nUsage: zcat [options] [files]\n"
+               "\nExit status is 0 if no errors occurred, 1 otherwise.\n"
+               "\nOptions:\n"
+               "  -h, --help                   display this help and exit\n"
+               "  -V, --version                output version information and exit\n"
+               "  -A, --show-all               equivalent to '-vET'\n"
+               "  -b, --number-nonblank        number nonblank output lines\n"
+               "  -e                           equivalent to '-vE'\n"
+               "  -E, --show-ends              display '$' at end of each line\n"
+               "  -M, --format=<list>          process only the formats in <list>\n"
+               "  -n, --number                 number all output lines\n"
+               "  -N, --no-rcfile              don't read runtime configuration file\n"
+               "  -O, --force-format=<fmt>     force the format given (bz2, gz, lz, xz)\n"
+               "  -q, --quiet                  suppress all messages\n"
+               "  -r, --recursive              operate recursively on directories\n"
+               "  -R, --dereference-recursive  recursively follow symbolic links\n"
+               "  -s, --squeeze-blank          never more than one single blank line\n"
+               "  -t                           equivalent to '-vT'\n"
+               "  -T, --show-tabs              display TAB characters as '^I'\n"
+               "  -v, --show-nonprinting       use '^' and 'M-' notation, except for LF and TAB\n"
+               "      --verbose                verbose mode (show error messages)\n"
+               "      --bz2=<command>          set compressor and options for bzip2 format\n"
+               "      --gz=<command>           set compressor and options for gzip format\n"
+               "      --lz=<command>           set compressor and options for lzip format\n"
+               "      --xz=<command>           set compressor and options for xz format\n" );
+  show_help_addr();
+  }
+
+
+bool do_cat( const int infd, const int buffer_size,
+             uint8_t * const inbuf, uint8_t * const outbuf,
+             const std::string & input_filename,
+             const Cat_options & cat_options )
+  {
+  static int at_bol = 1;	// at begin of line. 0 = false, 1 = true,
+				// 2 = at begin of second blank line.
+  int inpos = 0;		// positions in buffers
+  int outpos = 0;
+  int rd = -1;			// bytes read by the last readblock
+  unsigned char c;
+
+  while( true )
+    {
+    do {
+      if( outpos >= buffer_size )
+        {
+        if( writeblock( STDOUT_FILENO, outbuf, outpos ) != outpos )
+          { show_error( "Write error", errno ); return false; }
+        outpos = 0;
+        }
+      if( inpos > rd )			// inbuf is empty
+        {
+        rd = readblock( infd, inbuf, buffer_size );
+        if( rd != buffer_size && errno )
+          {
+          show_file_error( input_filename.c_str(), "Read error", errno );
+          return false;
+          }
+        if( rd == 0 )
+          {
+          if( writeblock( STDOUT_FILENO, outbuf, outpos ) != outpos )
+            { show_error( "Write error", errno ); return false; }
+          outpos = 0;
+          return true;
+          }
+        inpos = 0;
+        inbuf[rd] = '\n';		// sentinel newline
+        }
+      else				// a real newline was found
+        {
+        if( at_bol > 1 )
+          {
+          if( cat_options.squeeze_blank ) { c = inbuf[inpos++]; continue; }
+          }
+        else ++at_bol;
+        if( at_bol > 1 && cat_options.number_lines == 2 )
+          {
+          line_number.next();
+          outpos += line_number.sprint( &outbuf[outpos] );
+          }
+        if( cat_options.show_ends ) outbuf[outpos++] = '$';
+        outbuf[outpos++] = '\n';		// output the newline
+        }
+      c = inbuf[inpos++];
+      }
+    while( c == '\n' );
+
+    if( at_bol > 0 && cat_options.number_lines )
+      {
+      line_number.next();
+      outpos += line_number.sprint( &outbuf[outpos] );
+      }
+    at_bol = 0;
+
+    // the loops below continue until a newline (real or sentinel) is found
+
+    if( cat_options.show_nonprinting )
+      while( true )
+        {
+        if( c < 32 || c >= 127 )
+          {
+          if( c == '\n' ) break;
+          if( c != '\t' || cat_options.show_tabs )
+            {
+            if( c >= 128 )
+              { c -= 128; outbuf[outpos++] = 'M'; outbuf[outpos++] = '-'; }
+            if( c < 32 ) { c += 64; outbuf[outpos++] = '^'; }
+            else if( c == 127 ) { c = '?'; outbuf[outpos++] = '^'; }
+            }
+          }
+        outbuf[outpos++] = c;
+        c = inbuf[inpos++];
+        }
+    else				// not quoting
+      while( c != '\n' )
+        {
+        if( c == '\t' && cat_options.show_tabs )
+          { c += 64; outbuf[outpos++] = '^'; }
+        outbuf[outpos++] = c;
+        c = inbuf[inpos++];
+        }
+    }
+  }
+
+
+bool cat( int infd, const int format_index, const std::string & input_filename,
+          const Cat_options & cat_options )
+  {
+  enum { buffer_size = 4096, outbuf_size = (5 * buffer_size) + 256 + 1 };
+    // input buffer with space for sentinel newline at the end
+  uint8_t * const inbuf = new uint8_t[buffer_size+1];
+    // output buffer with space for character quoting, 255-digit line number,
+    // worst case flushing respect to inbuf, and a canary byte.
+  uint8_t * const outbuf = new uint8_t[outbuf_size];
+  outbuf[outbuf_size-1] = 0;
+  Children children;
+  bool error = false;
+
+  if( !set_data_feeder( input_filename, &infd, children, format_index ) ||
+      !do_cat( infd, buffer_size, inbuf, outbuf, input_filename, cat_options ) )
+    error = true;
+  if( !good_status( children, !error ) ) error = true;
+  if( !error && close( infd ) != 0 ) { show_close_error(); error = true; }
+  if( outbuf[outbuf_size-1] != 0 ) internal_error( "buffer overflow." );
+  delete[] outbuf; delete[] inbuf;
+  return !error;
+  }
+
+} // end namespace
+
+
+int main( const int argc, const char * const argv[] )
+  {
+  enum { verbose_opt = 256, bz2_opt, gz_opt, lz_opt, xz_opt };
+  int format_index = -1;
+  int recursive = 0;			// 1 = '-r', 2 = '-R'
+  std::list< std::string > filenames;
+  Cat_options cat_options;
+  program_name = "zcat";
+  invocation_name = ( argc > 0 ) ? argv[0] : program_name;
+
+  const Arg_parser::Option options[] =
+    {
+    { 'A', "show-all",              Arg_parser::no  },	// cat
+    { 'b', "number-nonblank",       Arg_parser::no  },	// cat
+    { 'c', "stdout",                Arg_parser::no  },	// gzip
+    { 'd', "decompress",            Arg_parser::no  },	// gzip
+    { 'e',  0,                      Arg_parser::no  },	// cat
+    { 'E', "show-ends",             Arg_parser::no  },	// cat
+    { 'f', "force",                 Arg_parser::no  },	// gzip
+    { 'h', "help",                  Arg_parser::no  },
+    { 'l', "list",                  Arg_parser::no  },	// gzip
+    { 'L', "license",               Arg_parser::no  },	// gzip
+    { 'M', "format",                Arg_parser::yes },
+    { 'n', "number",                Arg_parser::no  },	// cat
+    { 'N', "no-rcfile",             Arg_parser::no  },
+    { 'O', "force-format",          Arg_parser::yes },
+    { 'q', "quiet",                 Arg_parser::no  },
+    { 'r', "recursive",             Arg_parser::no  },
+    { 'R', "dereference-recursive", Arg_parser::no  },
+    { 's', "squeeze-blank",         Arg_parser::no  },	// cat
+    { 't',  0,                      Arg_parser::no  },	// cat
+    { 'T', "show-tabs",             Arg_parser::no  },	// cat
+    { 'v', "show-nonprinting",      Arg_parser::no  },	// cat
+    { 'V', "version",               Arg_parser::no  },
+    { verbose_opt, "verbose",       Arg_parser::no  },
+    { bz2_opt,     "bz2",           Arg_parser::yes },
+    { gz_opt,      "gz",            Arg_parser::yes },
+    { lz_opt,      "lz",            Arg_parser::yes },
+    { xz_opt,      "xz",            Arg_parser::yes },
+    {  0 ,  0,                      Arg_parser::no  } };
+
+  const Arg_parser parser( argc, argv, options );
+  if( parser.error().size() )				// bad option
+    { show_error( parser.error().c_str(), 0, true ); return 1; }
+
+  maybe_process_config_file( parser );
+
+  int argind = 0;
+  for( ; argind < parser.arguments(); ++argind )
+    {
+    const int code = parser.code( argind );
+    if( !code ) break;					// no more options
+    const std::string & arg = parser.argument( argind );
+    switch( code )
+      {
+      case 'A': cat_options.show_ends = true;
+                cat_options.show_nonprinting = true;
+                cat_options.show_tabs = true; break;
+      case 'b': cat_options.number_lines = 1; break;
+      case 'c': break;
+      case 'd': break;
+      case 'e': cat_options.show_nonprinting = true;	// fall through
+      case 'E': cat_options.show_ends = true; break;
+      case 'f': break;
+      case 'h': show_help(); return 0;
+      case 'l': break;
+      case 'L': break;
+      case 'M': parse_format_list( arg ); break;
+      case 'n': if( cat_options.number_lines == 0 )
+                  { cat_options.number_lines = 2; } break;
+      case 'N': break;
+      case 'O': format_index = parse_format_type( arg ); break;
+      case 'q': verbosity = -1; break;
+      case 'r': recursive = 1; break;
+      case 'R': recursive = 2; break;
+      case 's': cat_options.squeeze_blank = true; break;
+      case 't': cat_options.show_nonprinting = true;	// fall through
+      case 'T': cat_options.show_tabs = true; break;
+      case 'v': cat_options.show_nonprinting = true; break;
+      case 'V': show_version(); return 0;
+      case verbose_opt: if( verbosity < 4 ) ++verbosity; break;
+      case bz2_opt: parse_compressor( arg, fmt_bz2, 1 ); break;
+      case gz_opt: parse_compressor( arg, fmt_gz, 1 ); break;
+      case lz_opt: parse_compressor( arg, fmt_lz, 1 ); break;
+      case xz_opt: parse_compressor( arg, fmt_xz, 1 ); break;
+      default : internal_error( "uncaught option." );
+      }
+    } // end process options
+
+#if defined(__MSVCRT__) || defined(__OS2__)
+  setmode( STDIN_FILENO, O_BINARY );
+  setmode( STDOUT_FILENO, O_BINARY );
+#endif
+
+  for( ; argind < parser.arguments(); ++argind )
+    filenames.push_back( parser.argument( argind ) );
+
+  if( filenames.empty() ) filenames.push_back( recursive ? "." : "-" );
+
+  std::string input_filename;
+  bool error = false;
+  bool stdin_used = false;
+  while( next_filename( filenames, input_filename, error, recursive ) )
+    {
+    int infd;
+    if( input_filename == "." )
+      {
+      if( stdin_used ) continue; else stdin_used = true;
+      infd = STDIN_FILENO; input_filename = "-";
+      }
+    else
+      {
+      infd = open_instream( input_filename, format_index < 0 );
+      if( infd < 0 ) { error = true; continue; }
+      }
+
+    if( !cat( infd, format_index, input_filename, cat_options ) ) error = true;
+
+    if( close( infd ) != 0 )
+      { show_file_error( input_filename.c_str(), "Error closing input file",
+                         errno ); error = true; }
+    }
+
+  if( std::fclose( stdout ) != 0 )
+    {
+    show_error( "Error closing stdout", errno );
+    error = true;
+    }
+  return error;
+  }
diff --git a/zcatgrep.cc b/zcatgrep.cc
new file mode 100644
index 0000000..31d54e6
--- /dev/null
+++ b/zcatgrep.cc
@@ -0,0 +1,59 @@
+/* Common code for zcat and zgrep
+   Copyright (C) 2010-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#ifndef O_BINARY
+#define O_BINARY 0
+#endif
+
+
+int simple_extension_index( const std::string & name )
+  {
+  for( int i = 0; i < num_formats; ++i )
+    {
+    const std::string ext( simple_extensions[i] );
+    if( name.size() > ext.size() &&
+        name.compare( name.size() - ext.size(), ext.size(), ext ) == 0 )
+      return i;
+    }
+  return -1;
+  }
+
+
+int open_instream( std::string & input_filename, const bool search,
+                   const bool no_messages = false )
+  {
+  int infd = open( input_filename.c_str(), O_RDONLY | O_BINARY );
+  if( infd < 0 )
+    {
+    const int saved_errno = errno;
+    if( search && simple_extension_index( input_filename ) < 0 )
+      {
+      for( int i = 0; i < num_formats; ++i )
+        if( enabled_format( format_order[i] ) )
+          {
+          const std::string name( input_filename +
+                                  simple_extensions[format_order[i]] );
+          infd = open( name.c_str(), O_RDONLY | O_BINARY );
+          if( infd >= 0 ) { input_filename = name; break; }
+          }
+      }
+    if( infd < 0 && !no_messages )
+      show_file_error( input_filename.c_str(), "Can't open input file",
+                       saved_errno );
+    }
+  return infd;
+  }
diff --git a/zcmp.cc b/zcmp.cc
new file mode 100644
index 0000000..9daa1b9
--- /dev/null
+++ b/zcmp.cc
@@ -0,0 +1,471 @@
+/* Zcmp - decompress and compare two files byte by byte
+   Copyright (C) 2010-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#define _FILE_OFFSET_BITS 64
+
+#include <algorithm>
+#include <cctype>
+#include <cerrno>
+#include <climits>
+#include <csignal>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <string>
+#include <vector>
+#include <fcntl.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <sys/stat.h>
+#if defined(__MSVCRT__) || defined(__OS2__)
+#include <io.h>
+#endif
+
+#include "arg_parser.h"
+#include "rc.h"
+#include "zutils.h"
+
+#ifndef LLONG_MAX
+#define LLONG_MAX  0x7FFFFFFFFFFFFFFFLL
+#endif
+
+
+namespace {
+
+#include "zcmpdiff.cc"
+
+void show_help()
+  {
+  std::printf( "zcmp compares two files and, if they differ, writes to standard output the\n"
+               "first byte and line number where they differ. Bytes and lines are numbered\n"
+               "starting with 1. A hyphen '-' used as a file argument means standard input.\n"
+               "If any file given is compressed, its decompressed content is used. Compressed\n"
+               "files are decompressed on the fly; no temporary files are created.\n"
+               "\nThe formats supported are bzip2, gzip, lzip, and xz.\n"
+               "\nUsage: zcmp [options] file1 [file2]\n"
+               "\nzcmp compares file1 to file2. The standard input is used only if file1 or\n"
+               "file2 refers to standard input. If file2 is omitted zcmp tries the\n"
+               "following:\n"
+               "\n  - If file1 is compressed, compares its decompressed contents with\n"
+               "  the corresponding uncompressed file (the name of file1 with the\n"
+               "  extension removed).\n"
+               "\n  - If file1 is uncompressed, compares it with the decompressed\n"
+               "  contents of file1.[lz|bz2|gz|xz] (the first one that is found).\n"
+               "\nExit status is 0 if inputs are identical, 1 if different, 2 if trouble.\n"
+               "\nOptions:\n"
+               "  -h, --help                        display this help and exit\n"
+               "  -V, --version                     output version information and exit\n"
+               "  -b, --print-bytes                 print differing bytes\n"
+               "  -i, --ignore-initial=<n>[:<n2>]   ignore differences in the first <n> bytes\n"
+               "  -l, --list                        list position, value of all differing bytes\n"
+               "  -M, --format=<list>               process only the formats in <list>\n"
+               "  -n, --bytes=<n>                   compare at most <n> bytes\n"
+               "  -N, --no-rcfile                   don't read runtime configuration file\n"
+               "  -O, --force-format=[<f1>][,<f2>]  force the formats given (bz2, gz, lz, xz)\n"
+               "  -q, --quiet                       suppress all messages\n"
+               "  -s, --silent                      (same as --quiet)\n"
+               "  -v, --verbose                     verbose mode (same as --list)\n"
+               "      --bz2=<command>               set compressor and options for bzip2 format\n"
+               "      --gz=<command>                set compressor and options for gzip format\n"
+               "      --lz=<command>                set compressor and options for lzip format\n"
+               "      --xz=<command>                set compressor and options for xz format\n"
+               "\nNumbers may be followed by a multiplier: k = kB = 10^3 = 1000,\n"
+               "Ki = KiB = 2^10 = 1024, M = 10^6, Mi = 2^20, G = 10^9, Gi = 2^30, etc...\n" );
+  show_help_addr();
+  }
+
+
+long long getnum( const char * const ptr, const char ** const tailp = 0,
+                  const long long llimit = 0,
+                  const long long ulimit = LLONG_MAX )
+  {
+  char * tail;
+  errno = 0;
+  long long result = strtoll( ptr, &tail, 0 );
+  if( tail == ptr )
+    {
+    show_error( "Bad or missing numerical argument.", 0, true );
+    std::exit( 2 );
+    }
+  if( result < 0 ) errno = ERANGE;
+
+  if( !errno && tail[0] && std::isalpha( tail[0] ) )
+    {
+    const unsigned char ch = *tail++;
+    int factor;
+    bool bsuf;					// 'B' suffix is present
+    if( tail[0] == 'i' ) { ++tail; factor = 1024; } else factor = 1000;
+    if( tail[0] == 'B' ) { ++tail; bsuf = true; } else bsuf = false;
+    int exponent = -1;				// -1 = bad multiplier
+    switch( ch )
+      {
+      case 'Y': exponent = 8; break;
+      case 'Z': exponent = 7; break;
+      case 'E': exponent = 6; break;
+      case 'P': exponent = 5; break;
+      case 'T': exponent = 4; break;
+      case 'G': exponent = 3; break;
+      case 'M': exponent = 2; break;
+      case 'K': if( factor == 1024 ) exponent = 1; break;
+      case 'k': if( factor == 1000 ) exponent = 1; break;
+      case 'B': if( factor == 1000 && !bsuf ) exponent = 0; break;
+      }
+    if( exponent < 0 )
+      {
+      show_error( "Bad multiplier in numerical argument.", 0, true );
+      std::exit( 2 );
+      }
+    for( int i = 0; i < exponent; ++i )
+      {
+      if( ulimit / factor >= result ) result *= factor;
+      else { errno = ERANGE; break; }
+      }
+    }
+  if( !errno && ( result < llimit || result > ulimit ) ) errno = ERANGE;
+  if( errno )
+    {
+    show_error( "Numerical argument out of limits." );
+    std::exit( 2 );
+    }
+  if( tailp ) *tailp = tail;
+  return result;
+  }
+
+
+void parse_ignore_initial( const char * const arg, long long ignore_initial[2] )
+  {
+  const char * tail;
+  ignore_initial[0] = getnum( arg, &tail );
+  if( *tail == ':' || *tail == ',' )
+    ignore_initial[1] = getnum( ++tail );
+  else if( *tail == 0 ) ignore_initial[1] = ignore_initial[0];
+  else
+    {
+    show_error( "Bad separator in argument of '--ignore-initial'", 0, true );
+    std::exit( 2 );
+    }
+  }
+
+
+bool skip_ignore_initial( const long long ignore_initial, const int infd )
+  {
+  if( ignore_initial > 0 )
+    {
+    enum { buffer_size = 4096 };
+    long long rest = ignore_initial;
+    uint8_t buffer[buffer_size];
+    while( rest > 0 )
+      {
+      const int size = std::min( rest, (long long)buffer_size );
+      const int rd = readblock( infd, buffer, size );
+      if( rd != size && errno ) return false;
+      if( rd < size ) break;
+      rest -= rd;
+      }
+    }
+  return true;
+  }
+
+
+// Put into buf the unsigned char c, making unprintable bytes
+// visible by quoting like cat -t does.
+void sprintc( char * const buf, unsigned char c )
+  {
+  int i = 0;
+
+  if( c < 32 || c >= 127 )
+    {
+    if( c >= 128 ) { c -= 128; buf[i++] = 'M'; buf[i++] = '-'; }
+    if( c < 32 ) { c += 64; buf[i++] = '^'; }
+    else if( c == 127 ) { c = '?'; buf[i++] = '^'; }
+    }
+  buf[i++] = c;
+  buf[i++] = 0;
+  }
+
+
+int block_compare( const uint8_t * const buffer0,
+                   const uint8_t * const buffer1,
+                   unsigned long long * const line_numberp )
+  {
+  const uint8_t * p0 = buffer0;
+  const uint8_t * p1 = buffer1;
+
+  if( verbosity == 0 )
+    {
+    int nl_count = 0;
+    while( *p0 == *p1 )
+      { if( *p0 == '\n' ) { ++nl_count; } ++p0; ++p1; }
+    *line_numberp += nl_count;
+    }
+  else while( *p0 == *p1 ) { ++p0; ++p1; }
+  return p0 - buffer0;
+  }
+
+
+int cmp( const long long max_size, const int infd[2],
+         const std::string filenames[2], const bool print_bytes )
+  {
+  const int buffer_size = 4096;
+  unsigned long long byte_number = 1;
+  unsigned long long line_number = 1;
+  // remaining number of bytes to compare
+  long long rest = ( max_size >= 0 ) ? max_size : buffer_size;
+  // buffers with space for sentinels at the end
+  uint8_t * const buffer0 = new uint8_t[2*(buffer_size+1)];
+  uint8_t * const buffer1 = buffer0 + buffer_size + 1;
+  uint8_t * buffer[2];
+  buffer[0] = buffer0; buffer[1] = buffer1;
+  int different = 0;
+
+  while( rest > 0 )
+    {
+    const int size = std::min( (long long)buffer_size, rest );
+    if( max_size >= 0 ) rest -= size;
+    int rd[2];			// number of bytes read from each file
+    for( int i = 0; i < 2; ++i )
+      {
+      rd[i] = readblock( infd[i], buffer[i], size );
+      if( rd[i] != size && errno )
+        {
+        show_file_error( filenames[i].c_str(), "Read error", errno );
+        return 2;
+        }
+      }
+
+    const int min_rd = std::min( rd[0], rd[1] );
+    buffer0[min_rd] = 0;		// sentinels for the block compare
+    buffer1[min_rd] = 1;
+
+    int first_diff = block_compare( buffer0, buffer1, &line_number );
+    byte_number += first_diff;
+
+    if( first_diff < min_rd )
+      {
+      if( verbosity < 0 ) return 1;		// return status only
+      if( verbosity == 0 )			// show first difference
+        {
+        if( !print_bytes )
+          std::printf( "%s %s differ: byte %llu, line %llu\n",
+                       filenames[0].c_str(), filenames[1].c_str(),
+                       byte_number, line_number );
+        else
+          {
+          const unsigned char c0 = buffer0[first_diff];
+          const unsigned char c1 = buffer1[first_diff];
+          char buf0[5], buf1[5];
+          sprintc( buf0, c0 ); sprintc( buf1, c1 );
+          std::printf( "%s %s differ: byte %llu, line %llu is %3o %s %3o %s\n",
+                       filenames[0].c_str(), filenames[1].c_str(),
+                       byte_number, line_number, c0, buf0, c1, buf1 );
+          }
+        std::fflush( stdout );
+        return 1;
+        }
+      else			// verbosity > 0 ; show all differences
+        {
+        different = 1;
+        for( ; first_diff < min_rd; ++byte_number, ++first_diff )
+          {
+          const unsigned char c0 = buffer0[first_diff];
+          const unsigned char c1 = buffer1[first_diff];
+          if( c0 != c1 )
+            {
+            if( !print_bytes )
+              std::printf( "%llu %3o %3o\n", byte_number, c0, c1 );
+            else
+              {
+              char buf0[5], buf1[5];
+              sprintc( buf0, c0 ); sprintc( buf1, c1 );
+              std::printf( "%llu %3o %-4s %3o %s\n",
+                           byte_number, c0, buf0, c1, buf1 );
+              }
+            }
+          }
+        std::fflush( stdout );
+        }
+      }
+
+    if( rd[0] != rd[1] )
+      {
+      if( verbosity >= 0 )
+        std::fprintf( stderr, "%s: EOF on %s\n",
+                      program_name, filenames[rd[1]<rd[0]].c_str() );
+      return 1;
+      }
+    if( min_rd != buffer_size ) break;
+    }
+
+  delete[] buffer0;
+  return different;
+  }
+
+} // end namespace
+
+
+int main( const int argc, const char * const argv[] )
+  {
+  enum { bz2_opt = 256, gz_opt, lz_opt, xz_opt };
+  // number of initial bytes ignored for each file
+  long long ignore_initial[2] = { 0, 0 };
+  long long max_size = -1;			// < 0 means unlimited size
+  int format_types[2] = { -1, -1 };
+  bool print_bytes = false;
+  program_name = "zcmp";
+  invocation_name = ( argc > 0 ) ? argv[0] : program_name;
+
+  const Arg_parser::Option options[] =
+    {
+    { 'b', "print-bytes",    Arg_parser::no  },
+    { 'h', "help",           Arg_parser::no  },
+    { 'i', "ignore-initial", Arg_parser::yes },
+    { 'l', "list",           Arg_parser::no  },
+    { 'M', "format",         Arg_parser::yes },
+    { 'n', "bytes",          Arg_parser::yes },
+    { 'N', "no-rcfile",      Arg_parser::no  },
+    { 'O', "force-format",   Arg_parser::yes },
+    { 'q', "quiet",          Arg_parser::no  },
+    { 's', "silent",         Arg_parser::no  },
+    { 'v', "verbose",        Arg_parser::no  },
+    { 'V', "version",        Arg_parser::no  },
+    { bz2_opt,    "bz2",     Arg_parser::yes },
+    { gz_opt,     "gz",      Arg_parser::yes },
+    { lz_opt,     "lz",      Arg_parser::yes },
+    { xz_opt,     "xz",      Arg_parser::yes },
+    {  0 ,  0,               Arg_parser::no  } };
+
+  const Arg_parser parser( argc, argv, options );
+  if( parser.error().size() )				// bad option
+    { show_error( parser.error().c_str(), 0, true ); return 2; }
+
+  maybe_process_config_file( parser );
+
+  int argind = 0;
+  for( ; argind < parser.arguments(); ++argind )
+    {
+    const int code = parser.code( argind );
+    if( !code ) break;					// no more options
+    const std::string & arg = parser.argument( argind );
+    switch( code )
+      {
+      case 'b': print_bytes = true; break;
+      case 'h': show_help(); return 0;
+      case 'i': parse_ignore_initial( arg.c_str(), ignore_initial ); break;
+      case 'l': verbosity = 1; break;
+      case 'M': parse_format_list( arg ); break;
+      case 'n': max_size = getnum( arg.c_str() ); break;
+      case 'N': break;
+      case 'O': parse_format_types2( arg, format_types ); break;
+      case 'q':
+      case 's': verbosity = -1; break;
+      case 'v': verbosity = 1; break;
+      case 'V': show_version(); return 0;
+      case bz2_opt: parse_compressor( arg, fmt_bz2 ); break;
+      case gz_opt: parse_compressor( arg, fmt_gz ); break;
+      case lz_opt: parse_compressor( arg, fmt_lz ); break;
+      case xz_opt: parse_compressor( arg, fmt_xz ); break;
+      default : internal_error( "uncaught option." );
+      }
+    } // end process options
+
+#if defined(__MSVCRT__) || defined(__OS2__)
+  setmode( STDIN_FILENO, O_BINARY );
+  setmode( STDOUT_FILENO, O_BINARY );
+#endif
+
+  if( argind >= parser.arguments() )
+    { show_error( "No files given.", 0, true ); return 2; }
+  if( argind + 2 < parser.arguments() )
+    { show_error( "Too many files.", 0, true ); return 2; }
+
+  const int files = parser.arguments() - argind;
+  std::string filenames[2];		// file names of the two input files
+  filenames[0] = parser.argument( argind );
+  if( files == 2 ) filenames[1] = parser.argument( argind + 1 );
+
+  int infd[2];				// file descriptors of the two files
+  infd[0] = ( filenames[0] == "-" ) ?
+    STDIN_FILENO : open_instream( filenames[0] );
+  if( infd[0] < 0 ) return 2;
+
+  if( files == 2 )
+    {
+    if( check_identical( filenames[0].c_str(), filenames[1].c_str() ) )
+      {
+      if( ignore_initial[0] == ignore_initial[1] ) return 0;
+      else { show_error( "Can't compare parts of same file." ); return 2; }
+      }
+    infd[1] = ( filenames[1] == "-" ) ?
+      STDIN_FILENO : open_instream( filenames[1] );
+    if( infd[1] < 0 ) return 2;
+    }
+  else
+    {
+    if( filenames[0] == "-" )
+      { show_error( "Missing operand after '-'.", 0, true ); return 2; }
+    if( format_types[0] >= 0 || format_types[1] >= 0 )
+      { show_error( "Two files must be given when format is specified.", 0, true );
+        return 2; }
+    filenames[1] = filenames[0];
+    infd[1] = open_other_instream( filenames[1] );
+    if( infd[1] < 0 )
+      {
+      if( verbosity >= 0 )
+        std::fprintf( stderr, "%s: Can't find file to compare with '%s'.\n",
+                      program_name, filenames[0].c_str() );
+      show_error( 0, 0, true ); return 2;
+      }
+    }
+
+  int old_infd[2];		// copy of file descriptors of the two files
+  old_infd[0] = infd[0]; old_infd[1] = infd[1];
+  Children children[2];
+  if( !set_data_feeder( filenames[0], &infd[0], children[0], format_types[0] ) ||
+      !set_data_feeder( filenames[1], &infd[1], children[1], format_types[1] ) )
+    return 2;
+
+  for( int i = 0; i < 2; ++i )
+    if( !skip_ignore_initial( ignore_initial[i], infd[i] ) )
+      {
+      show_file_error( filenames[i].c_str(),
+                       "Read error skipping initial bytes", errno );
+      return 2;
+      }
+
+  int retval = cmp( max_size, infd, filenames, print_bytes );
+
+  for( int i = 0; i < 2; ++i )
+    if( !good_status( children[i], retval == 0 && max_size < 0 ) ) retval = 2;
+
+  for( int i = 0; i < 2; ++i )
+    {
+    if( close( infd[i] ) != 0 )
+      { show_close_error(); retval = 2; }
+    if( filenames[i] != "-" && close( old_infd[i] ) != 0 )
+      {
+      show_file_error( filenames[i].c_str(), "Error closing input file", errno );
+      retval = 2;
+      }
+    }
+  if( std::fclose( stdout ) != 0 )
+    {
+    show_error( "Error closing stdout", errno );
+    retval = 2;
+    }
+
+  return retval;
+  }
diff --git a/zcmpdiff.cc b/zcmpdiff.cc
new file mode 100644
index 0000000..fceb8cf
--- /dev/null
+++ b/zcmpdiff.cc
@@ -0,0 +1,70 @@
+/* Common code for zcmp and zdiff
+   Copyright (C) 2010-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#ifndef O_BINARY
+#define O_BINARY 0
+#endif
+
+
+int open_instream( const std::string & input_filename )
+  {
+  const int infd = open( input_filename.c_str(), O_RDONLY | O_BINARY );
+  if( infd < 0 )
+    show_file_error( input_filename.c_str(), "Can't open input file", errno );
+  return infd;
+  }
+
+
+int open_other_instream( std::string & name )
+  {
+  const int eindex = extension_index( name );
+  if( eindex >= 0 && enabled_format( -1 ) )
+    {					// open uncompressed version
+    name.resize( name.size() - std::strlen( extension_from( eindex ) ) );
+    name += extension_to( eindex );
+    return open( name.c_str(), O_RDONLY | O_BINARY );
+    }
+  if( eindex < 0 )			// search compressed version
+    for( int i = 0; i < num_formats; ++i )
+      if( enabled_format( format_order[i] ) )
+        {
+        const std::string s( name + simple_extensions[format_order[i]] );
+        const int infd = open( s.c_str(), O_RDONLY | O_BINARY );
+        if( infd >= 0 ) { name = s; return infd; }
+        }
+  return -1;
+  }
+
+
+void parse_format_types2( const std::string & arg, int format_types[2] )
+  {
+  const unsigned i = std::min( arg.find( ',' ), arg.size() );
+  if( i > 0 ) format_types[0] = parse_format_type( arg.substr( 0, i ) );
+  else format_types[0] = -1;
+  if( i + 1 < arg.size() ) format_types[1] =
+    parse_format_type( arg.substr( i + 1 ) );
+  else format_types[1] = -1;
+  }
+
+
+bool check_identical( const char * const name1, const char * const name2 )
+  {
+  if( std::strcmp( name1, name2 ) == 0 ) return true;
+  struct stat stat1, stat2;
+  if( stat( name1, &stat1 ) || stat( name2, &stat2 ) ) return false;
+  return ( stat1.st_ino == stat2.st_ino && stat1.st_dev == stat2.st_dev );
+  }
diff --git a/zdiff.cc b/zdiff.cc
new file mode 100644
index 0000000..a173971
--- /dev/null
+++ b/zdiff.cc
@@ -0,0 +1,440 @@
+/* Zdiff - decompress and compare two files line by line
+   Copyright (C) 2010-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#define _FILE_OFFSET_BITS 64
+
+#include <algorithm>
+#include <cctype>
+#include <cerrno>
+#include <climits>
+#include <csignal>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <string>
+#include <vector>
+#include <fcntl.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <sys/stat.h>
+#if defined(__MSVCRT__) || defined(__OS2__)
+#include <io.h>
+#endif
+
+#include "arg_parser.h"
+#include "rc.h"
+#include "zutils.h"
+
+// 'verbosity' is always 0 in zdiff; no --verbose or --quiet available.
+
+namespace {
+
+std::string fifonames[2];	// names of the two fifos passed to diff
+
+#include "zcmpdiff.cc"
+
+void show_help()
+  {
+  std::printf( "zdiff compares two files and, if they differ, writes to standard output the\n"
+               "differences line by line. A hyphen '-' used as a file argument means standard\n"
+               "input. If any file given is compressed, its decompressed content is used.\n"
+               "zdiff is a front end to the program diff and has the limitation that messages\n"
+               "from diff refer to temporary file names instead of those specified.\n"
+               "\nThe formats supported are bzip2, gzip, lzip, and xz.\n"
+               "\nUsage: zdiff [options] file1 [file2]\n"
+               "\nzdiff compares file1 to file2. The standard input is used only if file1 or\n"
+               "file2 refers to standard input. If file2 is omitted zdiff tries the\n"
+               "following:\n"
+               "\n  - If file1 is compressed, compares its decompressed contents with\n"
+               "  the corresponding uncompressed file (the name of file1 with the\n"
+               "  extension removed).\n"
+               "\n  - If file1 is uncompressed, compares it with the decompressed\n"
+               "  contents of file1.[lz|bz2|gz|xz] (the first one that is found).\n"
+               "\nExit status is 0 if inputs are identical, 1 if different, 2 if trouble.\n"
+               "Some options only work if the diff program used supports them.\n"
+               "\nOptions:\n"
+               "  -h, --help                        display this help and exit\n"
+               "  -V, --version                     output version information and exit\n"
+               "  -a, --text                        treat all files as text\n"
+               "  -b, --ignore-space-change         ignore changes in the amount of white space\n"
+               "  -B, --ignore-blank-lines          ignore changes whose lines are all blank\n"
+               "  -c                                use the context output format\n"
+               "  -C, --context=<n>                 same as -c but use <n> lines of context\n"
+               "  -d, --minimal                     try hard to find a smaller set of changes\n"
+               "  -E, --ignore-tab-expansion        ignore changes due to tab expansion\n"
+               "  -i, --ignore-case                 ignore case differences in file contents\n"
+               "  -M, --format=<list>               process only the formats in <list>\n"
+               "  -N, --no-rcfile                   don't read runtime configuration file\n"
+               "  -O, --force-format=[<f1>][,<f2>]  force the formats given (bz2, gz, lz, xz)\n"
+               "  -p, --show-c-function             show which C function each change is in\n"
+               "  -q, --brief                       output only whether files differ\n"
+               "  -s, --report-identical-files      report when two files are identical\n"
+               "  -t, --expand-tabs                 expand tabs to spaces in output\n"
+               "  -T, --initial-tab                 make tabs line up by prepending a tab\n"
+               "  -u                                use the unified output format\n"
+               "  -U, --unified=<n>                 same as -u but use <n> lines of context\n"
+               "  -w, --ignore-all-space            ignore all white space\n"
+               "  -W, --width=<n>                   output at most <n> print columns\n"
+               "  -y, --side-by-side                output in two columns\n"
+               "      --bz2=<command>               set compressor and options for bzip2 format\n"
+               "      --gz=<command>                set compressor and options for gzip format\n"
+               "      --lz=<command>                set compressor and options for lzip format\n"
+               "      --xz=<command>                set compressor and options for xz format\n" );
+  show_help_addr();
+  }
+
+
+const char * my_basename( const char * filename )
+  {
+  const char * c = filename;
+  while( *c ) { if( *c == '/' ) { filename = c + 1; } ++c; }
+  return filename;
+  }
+
+
+extern "C" void remove_fifos()
+  {
+  if( fifonames[0].size() )
+    { std::remove( fifonames[0].c_str() ); fifonames[0].clear(); }
+  if( fifonames[1].size() )
+    { std::remove( fifonames[1].c_str() ); fifonames[1].clear(); }
+  }
+
+
+/* Set fifonames[i] to "${TMPDIR}/<coded_pid>[_-]<basename(filenames[i])>"
+   and create FIFOs. The pid is coded in little endian order.
+*/
+bool set_fifonames( const std::string filenames[2] )
+  {
+  enum { num_codes = 36 };
+  const char * const codes = "0123456789abcdefghijklmnopqrstuvwxyz";
+  const char * p = std::getenv( "TMPDIR" );
+
+  if( p ) { fifonames[0] = p; fifonames[0] += '/'; }
+  else fifonames[0] = "/tmp/";
+  int n = getpid();
+  do fifonames[0] += codes[n % num_codes]; while( n /= num_codes );
+  const unsigned pos = fifonames[0].size();
+  fifonames[0] += '_';
+  fifonames[1] = fifonames[0];
+  fifonames[0] += my_basename( filenames[0].c_str() );
+  fifonames[1] += my_basename( filenames[1].c_str() );
+  if( fifonames[1] == fifonames[0] ) fifonames[1][pos] = '-';
+
+  for( int i = 0; i < 2; ++i )
+    if( mkfifo( fifonames[i].c_str(), S_IRUSR | S_IWUSR ) != 0 )
+      {
+      if( errno == EEXIST )
+        {
+        std::remove( fifonames[i].c_str() );
+        if( mkfifo( fifonames[i].c_str(), S_IRUSR | S_IWUSR ) == 0 )
+          continue;
+        }
+      show_file_error( fifonames[i].c_str(), "Can't create FIFO", errno );
+      return false;
+      }
+  return true;
+  }
+
+
+bool set_data_feeder( const std::string & filename,
+                      const std::string & fifoname, const int infd,
+                      Children & children, int format_index )
+  {
+  uint8_t magic_data[magic_buf_size];
+  int magic_size = 0;
+  if( format_index < 0 )
+    format_index = test_format( infd, magic_data, &magic_size );
+  children.compressor_name = get_compressor_name( format_index );
+
+  if( children.compressor_name )	// compressed
+    {
+    int fda[2];				// pipe from feeder to compressor
+    if( pipe( fda ) < 0 )
+      { show_error( "Can't create pipe", errno ); return false; }
+    const pid_t pid = fork();
+    if( pid == 0 )			// child 1 (compressor feeder)
+      {
+      if( close( fda[0] ) != 0 ||
+          !feed_data( filename, infd, fda[1], magic_data, magic_size ) )
+        _exit( 2 );
+      if( close( fda[1] ) != 0 )
+        { show_close_error(); _exit( 2 ); }
+      _exit( 0 );
+      }
+    if( pid < 0 )			// parent
+      { show_fork_error( "data feeder" ); return false; }
+
+    const pid_t pid2 = fork();
+    if( pid2 == 0 )			// child 2 (compressor)
+      {
+      const int outfd = open( fifoname.c_str(), O_WRONLY | O_BINARY );
+      if( outfd < 0 )
+        {
+        if( verbosity >= 0 )
+          std::fprintf( stderr, "%s: Can't open FIFO '%s' for writing: %s\n",
+                        program_name, fifoname.c_str(), std::strerror( errno ) );
+        _exit( 2 );
+        }
+      if( dup2( fda[0], STDIN_FILENO ) >= 0 &&
+          dup2( outfd, STDOUT_FILENO ) >= 0 &&
+          close( fda[0] ) == 0 && close( fda[1] ) == 0 &&
+          close( outfd ) == 0 )
+        {
+        const std::vector< std::string > & compressor_args =
+          get_compressor_args( format_index );
+        const int size = compressor_args.size();
+        const char ** const argv = new const char *[size+3];
+        argv[0] = children.compressor_name;
+        for( int i = 0; i < size; ++i )
+          argv[i+1] = compressor_args[i].c_str();
+        argv[size+1] = ( verbosity >= 0 ) ? "-d" : "-dq";
+        argv[size+2] = 0;
+        execvp( argv[0], (char **)argv );
+        }
+      show_exec_error( children.compressor_name );
+      _exit( 2 );
+      }
+    if( pid2 < 0 )			// parent
+      { show_fork_error( children.compressor_name ); return false; }
+
+    close( fda[0] ); close( fda[1] );
+    children.pid[0] = pid;
+    children.pid[1] = pid2;
+    }
+  else					// uncompressed
+    {
+    const pid_t pid = fork();
+    if( pid == 0 )			// child (feeder)
+      {
+      const int outfd = open( fifoname.c_str(), O_WRONLY | O_BINARY );
+      if( outfd < 0 )
+        {
+        if( verbosity >= 0 )
+          std::fprintf( stderr, "%s: Can't open FIFO '%s' for writing: %s\n",
+                        program_name, fifoname.c_str(), std::strerror( errno ) );
+        _exit( 2 );
+        }
+      if( !feed_data( filename, infd, outfd, magic_data, magic_size ) )
+        _exit( 2 );
+      if( close( outfd ) != 0 )
+        { show_close_error(); _exit( 2 ); }
+      _exit( 0 );
+      }
+    if( pid < 0 )			// parent
+      { show_fork_error( "data feeder" ); return false; }
+    children.pid[0] = pid;
+    children.pid[1] = 0;
+    }
+  return true;
+  }
+
+
+extern "C" void signal_handler( int sig )
+  {
+  remove_fifos();
+  std::signal( sig, SIG_DFL );
+  std::raise( sig );
+  }
+
+
+void set_signals()
+  {
+  std::signal( SIGHUP, signal_handler );
+  std::signal( SIGINT, signal_handler );
+  std::signal( SIGTERM, signal_handler );
+  }
+
+} // end namespace
+
+
+int main( const int argc, const char * const argv[] )
+  {
+  enum { bz2_opt = 256, gz_opt, lz_opt, xz_opt };
+  std::vector< const char * > diff_args;	// args to diff, maybe empty
+  int format_types[2] = { -1, -1 };
+  program_name = "zdiff";
+  invocation_name = ( argc > 0 ) ? argv[0] : program_name;
+
+  const Arg_parser::Option options[] =
+    {
+    { 'a', "text",                   Arg_parser::no  },
+    { 'b', "ignore-space-change",    Arg_parser::no  },
+    { 'B', "ignore-blank-lines",     Arg_parser::no  },
+    { 'c',  0,                       Arg_parser::no  },
+    { 'C', "context",                Arg_parser::yes },
+    { 'd', "minimal",                Arg_parser::no  },
+    { 'E', "ignore-tab-expansion",   Arg_parser::no  },
+    { 'h', "help",                   Arg_parser::no  },
+    { 'i', "ignore-case",            Arg_parser::no  },
+    { 'M', "format",                 Arg_parser::yes },
+    { 'N', "no-rcfile",              Arg_parser::no  },
+    { 'O', "force-format",           Arg_parser::yes },
+    { 'p', "show-c-function",        Arg_parser::no  },
+    { 'q', "brief",                  Arg_parser::no  },
+    { 's', "report-identical-files", Arg_parser::no  },
+    { 't', "expand-tabs",            Arg_parser::no  },
+    { 'T', "initial-tab",            Arg_parser::no  },
+    { 'u',  0,                       Arg_parser::no  },
+    { 'U', "unified",                Arg_parser::yes },
+    { 'V', "version",                Arg_parser::no  },
+    { 'w', "ignore-all-space",       Arg_parser::no  },
+    { 'W', "width",                  Arg_parser::yes },
+    { 'y', "side-by-side",           Arg_parser::no  },
+    { bz2_opt,    "bz2",             Arg_parser::yes },
+    { gz_opt,     "gz",              Arg_parser::yes },
+    { lz_opt,     "lz",              Arg_parser::yes },
+    { xz_opt,     "xz",              Arg_parser::yes },
+    {  0 ,  0,                       Arg_parser::no  } };
+
+  const Arg_parser parser( argc, argv, options );
+  if( parser.error().size() )				// bad option
+    { show_error( parser.error().c_str(), 0, true ); return 2; }
+
+  maybe_process_config_file( parser );
+
+  int argind = 0;
+  for( ; argind < parser.arguments(); ++argind )
+    {
+    const int code = parser.code( argind );
+    if( !code ) break;					// no more options
+    const std::string & arg = parser.argument( argind );
+    switch( code )
+      {
+      case 'a': diff_args.push_back( "-a" ); break;
+      case 'b': diff_args.push_back( "-b" ); break;
+      case 'B': diff_args.push_back( "-B" ); break;
+      case 'c': diff_args.push_back( "-c" ); break;
+      case 'C': diff_args.push_back( "-C" );
+                diff_args.push_back( arg.c_str() ); break;
+      case 'd': diff_args.push_back( "-d" ); break;
+      case 'E': diff_args.push_back( "-E" ); break;
+      case 'h': show_help(); return 0;
+      case 'i': diff_args.push_back( "-i" ); break;
+      case 'M': parse_format_list( arg ); break;
+      case 'N': break;
+      case 'O': parse_format_types2( arg, format_types ); break;
+      case 'p': diff_args.push_back( "-p" ); break;
+      case 'q': diff_args.push_back( "-q" ); break;
+      case 's': diff_args.push_back( "-s" ); break;
+      case 't': diff_args.push_back( "-t" ); break;
+      case 'T': diff_args.push_back( "-T" ); break;
+      case 'u': diff_args.push_back( "-u" ); break;
+      case 'U': diff_args.push_back( "-U" );
+                diff_args.push_back( arg.c_str() ); break;
+      case 'V': show_version(); return 0;
+      case 'w': diff_args.push_back( "-w" ); break;
+      case 'W': diff_args.push_back( "-W" );
+                diff_args.push_back( arg.c_str() ); break;
+      case 'y': diff_args.push_back( "-y" ); break;
+      case bz2_opt: parse_compressor( arg, fmt_bz2 ); break;
+      case gz_opt: parse_compressor( arg, fmt_gz ); break;
+      case lz_opt: parse_compressor( arg, fmt_lz ); break;
+      case xz_opt: parse_compressor( arg, fmt_xz ); break;
+      default : internal_error( "uncaught option." );
+      }
+    } // end process options
+
+#if defined(__MSVCRT__) || defined(__OS2__)
+  setmode( STDIN_FILENO, O_BINARY );
+  setmode( STDOUT_FILENO, O_BINARY );
+#endif
+
+  if( argind >= parser.arguments() )
+    { show_error( "No files given.", 0, true ); return 2; }
+  if( argind + 2 < parser.arguments() )
+    { show_error( "Too many files.", 0, true ); return 2; }
+
+  const int files = parser.arguments() - argind;
+  std::string filenames[2];		// file names of the two input files
+  filenames[0] = parser.argument( argind );
+  if( files == 2 ) filenames[1] = parser.argument( argind + 1 );
+
+  int infd[2];				// file descriptors of the two files
+  infd[0] = ( filenames[0] == "-" ) ?
+    STDIN_FILENO : open_instream( filenames[0] );
+  if( infd[0] < 0 ) return 2;
+
+  if( files == 2 )
+    {
+    if( check_identical( filenames[0].c_str(), filenames[1].c_str() ) )
+      return 0;
+    infd[1] = ( filenames[1] == "-" ) ?
+      STDIN_FILENO : open_instream( filenames[1] );
+    if( infd[1] < 0 ) return 2;
+    }
+  else
+    {
+    if( filenames[0] == "-" )
+      { show_error( "Missing operand after '-'.", 0, true ); return 2; }
+    if( format_types[0] >= 0 || format_types[1] >= 0 )
+      { show_error( "Two files must be given when format is specified.", 0, true );
+        return 2; }
+    filenames[1] = filenames[0];
+    infd[1] = open_other_instream( filenames[1] );
+    if( infd[1] < 0 )
+      {
+      if( verbosity >= 0 )
+        std::fprintf( stderr, "%s: Can't find file to compare with '%s'.\n",
+                      program_name, filenames[0].c_str() );
+      show_error( 0, 0, true ); return 2;
+      }
+    }
+
+  std::atexit( remove_fifos );
+  set_signals();
+  if( !set_fifonames( filenames ) ) return 2;
+
+  Children children[2];
+  if( !set_data_feeder( filenames[0], fifonames[0], infd[0], children[0],
+                        format_types[0] ) ||
+      !set_data_feeder( filenames[1], fifonames[1], infd[1], children[1],
+                        format_types[1] ) )
+    return 2;
+
+  const pid_t diff_pid = fork();
+  if( diff_pid == 0 )			// child (diff)
+    {
+    const char ** const argv = new const char *[diff_args.size()+5];
+    argv[0] = DIFF;
+    for( unsigned i = 0; i < diff_args.size(); ++i )
+      argv[i+1] = diff_args[i];
+    argv[diff_args.size()+1] = "--";
+    argv[diff_args.size()+2] = fifonames[0].c_str();
+    argv[diff_args.size()+3] = fifonames[1].c_str();
+    argv[diff_args.size()+4] = 0;
+    execvp( argv[0], (char **)argv );
+    show_exec_error( DIFF );
+    _exit( 2 );
+    }
+  if( diff_pid < 0 )			// parent
+    { show_fork_error( DIFF ); return 2; }
+
+  int retval = wait_for_child( diff_pid, DIFF );
+
+  for( int i = 0; i < 2; ++i )
+    if( !good_status( children[i], retval == 0 ) ) retval = 2;
+
+  for( int i = 0; i < 2; ++i )
+    if( filenames[i] != "-" && close( infd[i] ) != 0 )
+      {
+      show_file_error( filenames[i].c_str(), "Error closing input file", errno );
+      retval = 2;
+      }
+
+  return retval;
+  }
diff --git a/zegrep.in b/zegrep.in
new file mode 100644
index 0000000..0cac12e
--- /dev/null
+++ b/zegrep.in
@@ -0,0 +1,3 @@
+#! /bin/sh
+bindir=`echo "$0" | sed -e 's,[^/]*$,,'`
+exec "${bindir}"zgrep -E "$@"
diff --git a/zfgrep.in b/zfgrep.in
new file mode 100644
index 0000000..c1a96d8
--- /dev/null
+++ b/zfgrep.in
@@ -0,0 +1,3 @@
+#! /bin/sh
+bindir=`echo "$0" | sed -e 's,[^/]*$,,'`
+exec "${bindir}"zgrep -F "$@"
diff --git a/zgrep.cc b/zgrep.cc
new file mode 100644
index 0000000..1454e77
--- /dev/null
+++ b/zgrep.cc
@@ -0,0 +1,401 @@
+/* Zgrep - search compressed files for a regular expression
+   Copyright (C) 2010-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#define _FILE_OFFSET_BITS 64
+
+#include <cerrno>
+#include <climits>
+#include <csignal>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <list>
+#include <string>
+#include <vector>
+#include <dirent.h>
+#include <fcntl.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <sys/stat.h>
+#if defined(__MSVCRT__) || defined(__OS2__)
+#include <io.h>
+#endif
+
+#include "arg_parser.h"
+#include "rc.h"
+#include "zutils.h"
+
+
+namespace {
+
+#include "recursive.cc"
+#include "zcatgrep.cc"
+
+void show_help()
+  {
+  std::printf( "zgrep is a front end to the program grep that allows transparent search\n"
+               "on any combination of compressed and uncompressed files. If any file\n"
+               "given is compressed, its decompressed content is used. If a file given\n"
+               "does not exist, and its name does not end with one of the known\n"
+               "extensions, zgrep tries the compressed file names corresponding to the\n"
+               "formats supported. If a file fails to decompress, zgrep continues\n"
+               "searching the rest of the files.\n"
+               "\nIf a file is specified as '-', data are read from standard input,\n"
+               "decompressed if needed, and fed to grep. Data read from standard input\n"
+               "must be of the same type; all uncompressed or all in the same\n"
+               "compressed format.\n"
+               "\nIf no files are specified, recursive searches examine the current\n"
+               "working directory, and nonrecursive searches read standard input.\n"
+               "\nThe formats supported are bzip2, gzip, lzip, and xz.\n"
+               "\nUsage: zgrep [options] <pattern> [files]\n"
+               "\nExit status is 0 if match, 1 if no match, 2 if trouble.\n"
+               "Some options only work if the grep program used supports them.\n"
+               "\nOptions:\n"
+               "      --help                   display this help and exit\n"
+               "  -V, --version                output version information and exit\n"
+               "  -a, --text                   treat all files as text\n"
+               "  -A, --after-context=<n>      print <n> lines of trailing context\n"
+               "  -b, --byte-offset            print the byte offset of each line\n"
+               "  -B, --before-context=<n>     print <n> lines of leading context\n"
+               "  -c, --count                  only print a count of matching lines per file\n"
+               "  -C, --context=<n>            print <n> lines of output context\n"
+               "      --color[=<when>]         show matched strings in color\n"
+               "  -e, --regexp=<pattern>       use <pattern> as the pattern to match\n"
+               "  -E, --extended-regexp        <pattern> is an extended regular expression\n"
+               "  -f, --file=<file>            obtain patterns from <file>\n"
+               "  -F, --fixed-strings          <pattern> is a set of newline-separated strings\n"
+               "  -h, --no-filename            suppress the prefixing filename on output\n"
+               "  -H, --with-filename          print the filename for each match\n"
+               "  -i, --ignore-case            ignore case distinctions\n"
+               "  -I                           ignore binary files\n"
+               "  -l, --files-with-matches     only print names of files containing matches\n"
+               "  -L, --files-without-match    only print names of files containing no matches\n"
+               "  -m, --max-count=<n>          stop after <n> matches\n"
+               "  -M, --format=<list>          process only the formats in <list>\n"
+               "  -n, --line-number            print the line number of each line\n"
+               "  -N, --no-rcfile              don't read runtime configuration file\n"
+               "  -o, --only-matching          show only the part of a line matching <pattern>\n"
+               "  -O, --force-format=<fmt>     force the format given (bz2, gz, lz, xz)\n"
+               "  -q, --quiet                  suppress all messages\n"
+               "  -r, --recursive              operate recursively on directories\n"
+               "  -R, --dereference-recursive  recursively follow symbolic links\n"
+               "  -s, --no-messages            suppress error messages\n"
+               "  -v, --invert-match           select non-matching lines\n"
+               "      --verbose                verbose mode (show error messages)\n"
+               "  -w, --word-regexp            match only whole words\n"
+               "  -x, --line-regexp            match only whole lines\n"
+               "      --bz2=<command>          set compressor and options for bzip2 format\n"
+               "      --gz=<command>           set compressor and options for gzip format\n"
+               "      --lz=<command>           set compressor and options for lzip format\n"
+               "      --xz=<command>           set compressor and options for xz format\n"
+               "\nNumbers may be followed by a multiplier: k = kB = 10^3 = 1000,\n"
+               "Ki = KiB = 2^10 = 1024, M = 10^6, Mi = 2^20, G = 10^9, Gi = 2^30, etc...\n" );
+  show_help_addr();
+  }
+
+
+int zgrep_stdin( int infd, const int format_index,
+                 const std::vector< const char * > & grep_args )
+  {
+  Children children;
+  if( !set_data_feeder( "", &infd, children, format_index ) ) return 2;
+  const pid_t grep_pid = fork();
+  if( grep_pid == 0 )			// child (grep)
+    {
+    if( dup2( infd, STDIN_FILENO ) >= 0 && close( infd ) == 0 )
+      {
+      const char ** const argv = new const char *[grep_args.size()+2];
+      argv[0] = GREP;
+      for( unsigned i = 0; i < grep_args.size(); ++i )
+        argv[i+1] = grep_args[i];
+      argv[grep_args.size()+1] = 0;
+      execvp( argv[0], (char **)argv );
+      }
+    show_exec_error( GREP );
+    _exit( 2 );
+    }
+  if( grep_pid < 0 )			// parent
+    { show_fork_error( GREP ); return 2; }
+
+  int retval = wait_for_child( grep_pid, GREP );
+
+  if( !good_status( children, retval == 1 ) ) retval = 2;
+
+  if( close( infd ) != 0 )
+    { show_close_error(); return 2; }
+  return retval;
+  }
+
+
+int zgrep_file( int infd, const int format_index,
+                const std::string & input_filename,
+                const std::vector< const char * > & grep_args,
+                const int list_mode, const bool show_name )
+  {
+  Children children;
+  if( !set_data_feeder( input_filename, &infd, children, format_index ) )
+    return 2;
+  int fda[2];				// pipe from grep
+  if( pipe( fda ) < 0 )
+    { show_error( "Can't create pipe", errno ); return 2; }
+  const pid_t grep_pid = fork();
+  if( grep_pid == 0 )			// child (grep)
+    {
+    if( dup2( infd, STDIN_FILENO ) >= 0 &&
+        dup2( fda[1], STDOUT_FILENO ) >= 0 &&
+        close( infd ) == 0 && close( fda[0] ) == 0 && close( fda[1] ) == 0 )
+      {
+      const char ** const argv = new const char *[grep_args.size()+2];
+      argv[0] = GREP;
+      for( unsigned i = 0; i < grep_args.size(); ++i )
+        argv[i+1] = grep_args[i];
+      argv[grep_args.size()+1] = 0;
+      execvp( argv[0], (char **)argv );
+      }
+    show_exec_error( GREP );
+    _exit( 2 );
+    }
+  if( grep_pid < 0 )			// parent
+    { show_fork_error( GREP ); return 2; }
+
+  close( fda[1] );
+  enum { buffer_size = 256 };
+  uint8_t buffer[buffer_size];
+  bool line_begin = true;
+  while( true )
+    {
+    const int size = readblock( fda[0], buffer, buffer_size );
+    if( size != buffer_size && errno )
+      { show_error( "Read error", errno ); return 2; }
+    if( size > 0 && !list_mode )
+      {
+      if( show_name )
+        for( int i = 0; i < size; ++i )
+          {
+          if( line_begin )
+            { line_begin = false; std::printf( "%s:", input_filename.c_str() ); }
+          if( buffer[i] == '\n' ) line_begin = true;
+          putchar( buffer[i] );
+          }
+      else if( std::fwrite( buffer, 1, size, stdout ) != (unsigned)size )
+        { std::fflush( stdout ); show_error( "Write error", errno ); return 2; }
+      std::fflush( stdout );
+      }
+    if( size < buffer_size ) break;		// end of grep's output
+    }
+
+  int retval = wait_for_child( grep_pid, GREP );
+
+  if( !good_status( children, retval == 1 ) ) retval = 2;
+
+  if( list_mode && (retval == 0) == (list_mode == 1) )
+    { std::printf( "%s\n", input_filename.c_str() ); std::fflush( stdout ); }
+  if( close( infd ) != 0 )
+    { show_close_error(); return 2; }
+  if( close( fda[0] ) != 0 )
+    { show_close_error( GREP ); return 2; }
+  return retval;
+  }
+
+} // end namespace
+
+
+int main( const int argc, const char * const argv[] )
+  {
+  enum { help_opt = 256, verbose_opt, color_opt,
+         bz2_opt, gz_opt, lz_opt, xz_opt };
+  int format_index = -1;
+  int list_mode = 0;		// 1 = list matches, -1 = list non-matches
+  int recursive = 0;		// 1 = '-r', 2 = '-R'
+  int show_name = -1;		// tri-state bool
+  bool no_messages = false;
+  std::list< std::string > filenames;
+  std::vector< const char * > grep_args;	// args to grep, maybe empty
+  std::string color_option;		// needed because of optional arg
+  program_name = "zgrep";
+  invocation_name = ( argc > 0 ) ? argv[0] : program_name;
+
+  const Arg_parser::Option options[] =
+    {
+    { 'a', "text",                  Arg_parser::no  },	// grep GNU
+    { 'A', "after-context",         Arg_parser::yes },	// grep GNU
+    { 'b', "byte-offset",           Arg_parser::no  },	// grep GNU
+    { 'B', "before-context",        Arg_parser::yes },	// grep GNU
+    { 'c', "count",                 Arg_parser::no  },	// grep
+    { 'C', "context",               Arg_parser::yes },	// grep GNU
+    { 'e', "regexp",                Arg_parser::yes },	// grep
+    { 'E', "extended-regexp",       Arg_parser::no  },	// grep
+    { 'f', "file ",                 Arg_parser::yes },	// grep
+    { 'F', "fixed-strings",         Arg_parser::no  },	// grep
+    { 'h', "no-filename",           Arg_parser::no  },	// grep GNU
+    { 'H', "with-filename",         Arg_parser::no  },	// grep GNU
+    { 'i', "ignore-case",           Arg_parser::no  },	// grep
+    { 'I',  0,                      Arg_parser::no  },	// grep GNU
+    { 'l', "files-with-matches",    Arg_parser::no  },	// grep
+    { 'L', "files-without-match",   Arg_parser::no  },	// grep GNU
+    { 'm', "max-count",             Arg_parser::yes },	// grep GNU
+    { 'M', "format",                Arg_parser::yes },
+    { 'n', "line-number",           Arg_parser::no  },	// grep
+    { 'N', "no-rcfile",             Arg_parser::no  },
+    { 'o', "only-matching",         Arg_parser::no  },	// grep
+    { 'O', "force-format",          Arg_parser::yes },
+    { 'q', "quiet",                 Arg_parser::no  },
+    { 'r', "recursive",             Arg_parser::no  },
+    { 'R', "dereference-recursive", Arg_parser::no  },
+    { 's', "no-messages",           Arg_parser::no  },	// grep
+    { 'v', "invert-match",          Arg_parser::no  },	// grep
+    { 'V', "version",               Arg_parser::no  },
+    { 'w', "word-regexp",           Arg_parser::no  },	// grep GNU
+    { 'x', "line-regexp",           Arg_parser::no  },	// grep
+    { help_opt,    "help",          Arg_parser::no  },
+    { verbose_opt, "verbose",       Arg_parser::no  },
+    { color_opt,   "color",         Arg_parser::maybe },
+    { bz2_opt,     "bz2",           Arg_parser::yes },
+    { gz_opt,      "gz",            Arg_parser::yes },
+    { lz_opt,      "lz",            Arg_parser::yes },
+    { xz_opt,      "xz",            Arg_parser::yes },
+    {  0 ,  0,                      Arg_parser::no  } };
+
+  const Arg_parser parser( argc, argv, options );
+  if( parser.error().size() )				// bad option
+    { show_error( parser.error().c_str(), 0, true ); return 2; }
+
+  maybe_process_config_file( parser );
+
+  int argind = 0;
+  bool pattern_found = false;
+  for( ; argind < parser.arguments(); ++argind )
+    {
+    const int code = parser.code( argind );
+    if( !code ) break;					// no more options
+    const std::string & arg = parser.argument( argind );
+    switch( code )
+      {
+      case 'a': grep_args.push_back( "-a" ); break;
+      case 'A': grep_args.push_back( "-A" );
+                grep_args.push_back( arg.c_str() ); break;
+      case 'b': grep_args.push_back( "-b" ); break;
+      case 'B': grep_args.push_back( "-B" );
+                grep_args.push_back( arg.c_str() ); break;
+      case 'c': grep_args.push_back( "-c" ); break;
+      case 'C': grep_args.push_back( "-C" );
+                grep_args.push_back( arg.c_str() ); break;
+      case 'e': grep_args.push_back( "-e" );
+                grep_args.push_back( arg.c_str() ); pattern_found = true; break;
+      case 'E': grep_args.push_back( "-E" ); break;
+      case 'f': grep_args.push_back( "-f" );
+                grep_args.push_back( arg.c_str() ); pattern_found = true; break;
+      case 'F': grep_args.push_back( "-F" ); break;
+      case 'h': show_name = false; break;
+      case 'H': show_name = true; break;
+      case 'i': grep_args.push_back( "-i" ); break;
+      case 'I': grep_args.push_back( "-I" ); break;
+      case 'l': grep_args.push_back( "-l" ); list_mode = 1; break;
+      case 'L': grep_args.push_back( "-L" ); list_mode = -1; break;
+      case 'm': grep_args.push_back( "-m" );
+                grep_args.push_back( arg.c_str() ); break;
+      case 'M': parse_format_list( arg ); break;
+      case 'n': grep_args.push_back( "-n" ); break;
+      case 'N': break;
+      case 'o': grep_args.push_back( "-o" ); break;
+      case 'O': format_index = parse_format_type( arg ); break;
+      case 'q': grep_args.push_back( "-q" ); verbosity = -1; break;
+      case 'r': recursive = 1; break;
+      case 'R': recursive = 2; break;
+      case 's': grep_args.push_back( "-s" ); no_messages = true; break;
+      case 'v': grep_args.push_back( "-v" ); break;
+      case 'V': show_version(); return 0;
+      case 'w': grep_args.push_back( "-w" ); break;
+      case 'x': grep_args.push_back( "-x" ); break;
+      case help_opt   : show_help(); return 0;
+      case verbose_opt: if( verbosity < 4 ) ++verbosity;
+                        no_messages = false; break;
+      case color_opt: color_option = "--color";
+        if( !arg.empty() ) { color_option += '='; color_option += arg; }
+        break;
+      case bz2_opt: parse_compressor( arg, fmt_bz2 ); break;
+      case gz_opt: parse_compressor( arg, fmt_gz ); break;
+      case lz_opt: parse_compressor( arg, fmt_lz ); break;
+      case xz_opt: parse_compressor( arg, fmt_xz ); break;
+      default : internal_error( "uncaught option." );
+      }
+    } // end process options
+
+  if( !color_option.empty() )		// push the last value set
+    grep_args.push_back( color_option.c_str() );
+
+#if defined(__MSVCRT__) || defined(__OS2__)
+  setmode( STDIN_FILENO, O_BINARY );
+  setmode( STDOUT_FILENO, O_BINARY );
+#endif
+
+  if( !pattern_found )
+    {
+    if( argind >= parser.arguments() )
+      { show_error( "Pattern not found." ); return 2; }
+    const std::string & arg = parser.argument( argind++ );
+    if( arg.size() && arg[0] == '-' ) grep_args.push_back( "-e" );
+    grep_args.push_back( arg.c_str() );
+    }
+
+  for( ; argind < parser.arguments(); ++argind )
+    filenames.push_back( parser.argument( argind ) );
+
+  if( filenames.empty() ) filenames.push_back( recursive ? "." : "-" );
+
+  if( show_name < 0 ) show_name = ( filenames.size() != 1 || recursive );
+
+  std::string input_filename;
+  int retval = 1;
+  bool error = false;
+  bool stdin_used = false;
+  while( next_filename( filenames, input_filename, error, recursive,
+                        false, no_messages ) )
+    {
+    int infd;
+    if( input_filename == "." )
+      {
+      if( stdin_used ) continue; else stdin_used = true;
+      infd = STDIN_FILENO; input_filename = "-";
+      }
+    else
+      {
+      infd = open_instream( input_filename, format_index < 0, no_messages );
+      if( infd < 0 ) { error = true; continue; }
+      }
+
+    int tmp;
+    if( infd == STDIN_FILENO )
+      tmp = zgrep_stdin( infd, format_index, grep_args );
+    else tmp = zgrep_file( infd, format_index, input_filename, grep_args,
+                           list_mode, show_name );
+    if( tmp == 0 || ( tmp == 2 && retval == 1 ) ) retval = tmp;
+
+    if( close( infd ) != 0 )
+      { show_file_error( input_filename.c_str(), "Error closing input file",
+                         errno ); error = true; }
+    if( retval == 0 && verbosity < 0 ) break;
+    }
+
+  if( std::fclose( stdout ) != 0 )
+    {
+    show_error( "Error closing stdout", errno );
+    error = true;
+    }
+  if( error && ( retval != 0 || verbosity >= 0 ) ) retval = 2;
+  return retval;
+  }
diff --git a/ztest.cc b/ztest.cc
new file mode 100644
index 0000000..812278e
--- /dev/null
+++ b/ztest.cc
@@ -0,0 +1,335 @@
+/* Ztest - verify the integrity of compressed files
+   Copyright (C) 2010-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#define _FILE_OFFSET_BITS 64
+
+#include <cerrno>
+#include <climits>
+#include <csignal>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <list>
+#include <string>
+#include <vector>
+#include <dirent.h>
+#include <fcntl.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <sys/stat.h>
+#if defined(__MSVCRT__) || defined(__OS2__)
+#include <io.h>
+#endif
+
+#include "arg_parser.h"
+#include "rc.h"
+#include "zutils.h"
+
+#ifndef O_BINARY
+#define O_BINARY 0
+#endif
+
+
+namespace {
+
+#include "recursive.cc"
+
+void show_help()
+  {
+  std::printf( "ztest verifies the integrity of the compressed files specified.\n"
+               "Uncompressed files are ignored. If a file is specified as '-', the\n"
+               "integrity of compressed data read from standard input is verified. Data\n"
+               "read from standard input must be all in the same compressed format. If\n"
+               "a file fails to decompress, does not exist, can't be opened, or is a\n"
+               "terminal, ztest continues verifying the rest of the files. A final\n"
+               "diagnostic is shown at verbosity level 1 or higher if any file fails the\n"
+               "test when testing multiple files.\n"
+               "\nIf no files are specified, recursive searches examine the current\n"
+               "working directory, and nonrecursive searches read standard input.\n"
+               "\nThe formats supported are bzip2, gzip, lzip, and xz.\n"
+               "\nNote that error detection in the xz format is broken. First, some xz\n"
+               "files lack integrity information. Second, not all xz decompressors can\n"
+               "verify the integrity of all xz files. Third, section 2.1.1.2 'Stream\n"
+               "Flags' of the xz format specification allows xz decompressors to produce\n"
+               "garbage output without issuing any warning. Therefore, xz files can't\n"
+               "always be verified as reliably as files in the other formats can.\n"
+               "\nUsage: ztest [options] [files]\n"
+               "\nExit status is 0 if all compressed files verify OK, 1 if environmental\n"
+               "problems (file not found, invalid flags, I/O errors, etc), 2 if any\n"
+               "compressed file is corrupt or invalid.\n"
+               "\nOptions:\n"
+               "  -h, --help                   display this help and exit\n"
+               "  -V, --version                output version information and exit\n"
+               "  -M, --format=<list>          process only the formats in <list>\n"
+               "  -N, --no-rcfile              don't read runtime configuration file\n"
+               "  -O, --force-format=<fmt>     force the format given (bz2, gz, lz, xz)\n"
+               "  -q, --quiet                  suppress all messages\n"
+               "  -r, --recursive              operate recursively on directories\n"
+               "  -R, --dereference-recursive  recursively follow symbolic links\n"
+               "  -v, --verbose                be verbose (a 2nd -v gives more)\n"
+               "      --bz2=<command>          set compressor and options for bzip2 format\n"
+               "      --gz=<command>           set compressor and options for gzip format\n"
+               "      --lz=<command>           set compressor and options for lzip format\n"
+               "      --xz=<command>           set compressor and options for xz format\n" );
+  show_help_addr();
+  }
+
+
+int open_instream( const std::string & input_filename )
+  {
+  const int infd = open( input_filename.c_str(), O_RDONLY | O_BINARY );
+  if( infd < 0 )
+    show_file_error( input_filename.c_str(), "Can't open input file", errno );
+  return infd;
+  }
+
+
+int ztest_stdin( const int infd, int format_index,
+                 const std::vector< const char * > & ztest_args )
+  {
+  uint8_t magic_data[magic_buf_size];
+  int magic_size = 0;
+  if( format_index < 0 )
+    format_index = test_format( infd, magic_data, &magic_size );
+  const char * const compressor_name = get_compressor_name( format_index );
+  if( !compressor_name )
+    { show_error( "Unknown data format read from stdin." ); return 2; }
+  int fda[2];				// pipe from feeder
+  if( pipe( fda ) < 0 )
+    { show_error( "Can't create pipe", errno ); return 1; }
+
+  const pid_t pid = fork();
+  if( pid == 0 )			// child1 (compressor feeder)
+    {
+    if( close( fda[0] ) != 0 ||
+        !feed_data( "", infd, fda[1], magic_data, magic_size ) )
+      _exit( 1 );
+    if( close( fda[1] ) != 0 )
+      { show_close_error(); _exit( 1 ); }
+    _exit( 0 );
+    }
+  if( pid < 0 )				// parent
+    { show_fork_error( "data feeder" ); return 1; }
+
+  const pid_t pid2 = fork();
+  if( pid2 == 0 )			// child2 (compressor)
+    {
+    if( dup2( fda[0], STDIN_FILENO ) >= 0 &&
+        close( fda[0] ) == 0 && close( fda[1] ) == 0 )
+      {
+      const std::vector< std::string > & compressor_args =
+        get_compressor_args( format_index );
+      const int size = compressor_args.size();
+      const int size2 = ztest_args.size();
+      const char ** const argv = new const char *[size+size2+3];
+      argv[0] = compressor_name;
+      for( int i = 0; i < size; ++i )
+        argv[i+1] = compressor_args[i].c_str();
+      for( int i = 0; i < size2; ++i )
+        argv[i+size+1] = ztest_args[i];
+      argv[size+size2+1] = "-t";
+      argv[size+size2+2] = 0;
+      execvp( argv[0], (char **)argv );
+      }
+    show_exec_error( compressor_name );
+    _exit( 1 );
+    }
+  if( pid2 < 0 )			// parent
+    { show_fork_error( compressor_name ); return 1; }
+
+  close( fda[0] ); close( fda[1] );
+  const bool isgzxz = ( format_index == fmt_gz || format_index == fmt_xz );
+  int retval = wait_for_child( pid2, compressor_name, 1, isgzxz );
+  if( retval == 0 && wait_for_child( pid, "data feeder" ) != 0 )
+    retval = 1;
+  return retval;
+  }
+
+
+int ztest_file( const int infd, int format_index,
+                const std::string & input_filename,
+                const std::vector< const char * > & ztest_args )
+  {
+  static int disable_xz = -1;				// tri-state bool
+  uint8_t magic_data[magic_buf_size];
+  int magic_size = 0;
+  if( format_index < 0 )
+    format_index = test_format( infd, magic_data, &magic_size );
+  const char * const compressor_name = get_compressor_name( format_index );
+  if( !compressor_name )
+    return 0;				// ignore this file
+  if( format_index == fmt_xz )
+    {
+    if( disable_xz < 0 )
+      {
+      std::string command( compressor_name ); command += " -V > /dev/null 2>&1";
+      disable_xz = ( std::system( command.c_str() ) != 0 );
+      }
+    if( disable_xz ) return 0;		// ignore this file if no xz installed
+    }
+
+  const pid_t pid = fork();
+
+  if( pid == 0 )			// child (compressor)
+    {
+    const std::vector< std::string > & compressor_args =
+      get_compressor_args( format_index );
+    const int size = compressor_args.size();
+    const int size2 = ztest_args.size();
+    const char ** const argv = new const char *[size+size2+5];
+    argv[0] = compressor_name;
+    for( int i = 0; i < size; ++i )
+      argv[i+1] = compressor_args[i].c_str();
+    for( int i = 0; i < size2; ++i )
+      argv[i+size+1] = ztest_args[i];
+    argv[size+size2+1] = "-t";
+    argv[size+size2+2] = "--";
+    argv[size+size2+3] = input_filename.c_str();
+    argv[size+size2+4] = 0;
+    execvp( argv[0], (char **)argv );
+    show_exec_error( compressor_name );
+    _exit( 1 );
+    }
+  if( pid < 0 )				// parent
+    { show_fork_error( compressor_name ); return 1; }
+
+  const bool isgzxz = ( format_index == fmt_gz || format_index == fmt_xz );
+  return wait_for_child( pid, compressor_name, 1, isgzxz );
+  }
+
+} // end namespace
+
+
+int main( const int argc, const char * const argv[] )
+  {
+  enum { bz2_opt = 256, gz_opt, lz_opt, xz_opt };
+  int format_index = -1;
+  int recursive = 0;				// 1 = '-r', 2 = '-R'
+  std::list< std::string > filenames;
+  std::vector< const char * > ztest_args;	// args to ztest, maybe empty
+  program_name = "ztest";
+  invocation_name = ( argc > 0 ) ? argv[0] : program_name;
+
+  const Arg_parser::Option options[] =
+    {
+    { 'h', "help",                  Arg_parser::no  },
+    { 'M', "format",                Arg_parser::yes },
+    { 'N', "no-rcfile",             Arg_parser::no  },
+    { 'O', "force-format",          Arg_parser::yes },
+    { 'q', "quiet",                 Arg_parser::no  },
+    { 'r', "recursive",             Arg_parser::no  },
+    { 'R', "dereference-recursive", Arg_parser::no  },
+    { 'v', "verbose",               Arg_parser::no  },
+    { 'V', "version",               Arg_parser::no  },
+    { bz2_opt,    "bz2",            Arg_parser::yes },
+    { gz_opt,     "gz",             Arg_parser::yes },
+    { lz_opt,     "lz",             Arg_parser::yes },
+    { xz_opt,     "xz",             Arg_parser::yes },
+    {  0 ,  0,                      Arg_parser::no  } };
+
+  const Arg_parser parser( argc, argv, options );
+  if( parser.error().size() )				// bad option
+    { show_error( parser.error().c_str(), 0, true ); return 1; }
+
+  maybe_process_config_file( parser );
+
+  int argind = 0;
+  for( ; argind < parser.arguments(); ++argind )
+    {
+    const int code = parser.code( argind );
+    if( !code ) break;					// no more options
+    const std::string & arg = parser.argument( argind );
+    switch( code )
+      {
+      case 'h': show_help(); return 0;
+      case 'M': parse_format_list( arg ); break;
+      case 'N': break;
+      case 'O': format_index = parse_format_type( arg ); break;
+      case 'q': verbosity = -1; ztest_args.push_back( "-q" ); break;
+      case 'r': recursive = 1; break;
+      case 'R': recursive = 2; break;
+      case 'v': if( verbosity < 4 ) ++verbosity;
+                ztest_args.push_back( "-v" ); break;
+      case 'V': show_version(); return 0;
+      case bz2_opt: parse_compressor( arg, fmt_bz2, 1 ); break;
+      case gz_opt: parse_compressor( arg, fmt_gz, 1 ); break;
+      case lz_opt: parse_compressor( arg, fmt_lz, 1 ); break;
+      case xz_opt: parse_compressor( arg, fmt_xz, 1 ); break;
+      default : internal_error( "uncaught option." );
+      }
+    } // end process options
+
+#if defined(__MSVCRT__) || defined(__OS2__)
+  setmode( STDIN_FILENO, O_BINARY );
+  setmode( STDOUT_FILENO, O_BINARY );
+#endif
+
+  for( ; argind < parser.arguments(); ++argind )
+    filenames.push_back( parser.argument( argind ) );
+
+  if( filenames.empty() ) filenames.push_back( recursive ? "." : "-" );
+
+  std::string input_filename;
+  int files_tested = 0, failed_tests = 0;
+  int retval = 0;
+  bool error = false;
+  bool stdin_used = false;
+  while( next_filename( filenames, input_filename, error, recursive ) )
+    {
+    int infd;
+    if( input_filename == "." )
+      {
+      if( stdin_used ) continue; else stdin_used = true;
+      infd = STDIN_FILENO; input_filename = "-";
+      }
+    else
+      {
+      infd = open_instream( input_filename );
+      if( infd < 0 ) { error = true; continue; }
+      }
+
+    if( isatty( infd ) )			// for example /dev/tty
+      {
+      show_file_error( input_filename == "-" ? "(stdin)" : input_filename.c_str(),
+                       "I won't read compressed data from a terminal." );
+      close( infd ); error = true; continue;
+      }
+
+    int tmp;
+    if( infd == STDIN_FILENO )
+      tmp = ztest_stdin( infd, format_index, ztest_args );
+    else tmp = ztest_file( infd, format_index, input_filename, ztest_args );
+    if( tmp > retval ) retval = tmp;
+    ++files_tested; if( tmp ) ++failed_tests;
+
+    if( close( infd ) != 0 )
+      { show_file_error( input_filename.c_str(), "Error closing input file",
+                         errno ); error = true; }
+    }
+
+  if( std::fclose( stdout ) != 0 )	// in case decompressor writes to stdout
+    {
+    show_error( "Error closing stdout", errno );
+    error = true;
+    }
+  if( error && retval == 0 ) retval = 1;
+  if( failed_tests > 0 && verbosity >= 1 && files_tested > 1 )
+    std::fprintf( stderr, "%s: warning: %d %s failed the test.\n",
+                  program_name, failed_tests,
+                  ( failed_tests == 1 ) ? "file" : "files" );
+  return retval;
+  }
diff --git a/zupdate.cc b/zupdate.cc
new file mode 100644
index 0000000..a605f35
--- /dev/null
+++ b/zupdate.cc
@@ -0,0 +1,412 @@
+/* Zupdate - recompress bzip2, gzip, xz files to lzip format
+   Copyright (C) 2013-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#define _FILE_OFFSET_BITS 64
+
+#include <cerrno>
+#include <climits>
+#include <csignal>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <list>
+#include <string>
+#include <vector>
+#include <dirent.h>
+#include <fcntl.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <utime.h>
+#include <sys/stat.h>
+#include <sys/wait.h>
+#if defined(__MSVCRT__) || defined(__OS2__)
+#include <io.h>
+#endif
+
+#include "arg_parser.h"
+#include "rc.h"
+
+#ifndef O_BINARY
+#define O_BINARY 0
+#endif
+
+
+namespace {
+
+#include "recursive.cc"
+
+void show_help()
+  {
+  std::printf( "zupdate recompresses files from bzip2, gzip, and xz formats to lzip\n"
+               "format. Each original is compared with the new file and then deleted.\n"
+               "Only regular files with standard file name extensions are recompressed,\n"
+               "other files are ignored. Compressed files are decompressed and then\n"
+               "recompressed on the fly; no temporary files are created. The lzip format\n"
+               "is chosen as destination because it is the most appropriate for\n"
+               "long-term data archiving.\n"
+               "\nIf no files are specified, recursive searches examine the current\n"
+               "working directory, and nonrecursive searches do nothing.\n"
+               "\nIf the lzip compressed version of a file already exists, the file is\n"
+               "skipped unless the option '--force' is given. In this case, if the\n"
+               "comparison with the existing lzip version fails, an error is returned\n"
+               "and the original file is not deleted. The operation of zupdate is meant\n"
+               "to be safe and not cause any data loss. Therefore, existing lzip\n"
+               "compressed files are never overwritten nor deleted.\n"
+               "\nThe names of the original files must have one of the following extensions:\n"
+               "'.bz2', '.gz', or '.xz', which are recompressed to '.lz';\n"
+               "'.tbz', '.tbz2', '.tgz', or '.txz', which are recompressed to '.tlz'.\n"
+               "\nUsage: zupdate [options] [files]\n"
+               "\nExit status is 0 if all the compressed files were successfully recompressed\n"
+               "(if needed), compared, and deleted (if requested). Non-zero otherwise.\n"
+               "\nOptions:\n"
+               "  -h, --help                   display this help and exit\n"
+               "  -V, --version                output version information and exit\n"
+               "  -f, --force                  don't skip a file even if the .lz exists\n"
+               "  -k, --keep                   keep (don't delete) input files\n"
+               "  -l, --lzip-verbose           pass one option -v to the lzip compressor\n"
+               "  -M, --format=<list>          process only the formats in <list>\n"
+               "  -N, --no-rcfile              don't read runtime configuration file\n"
+               "  -q, --quiet                  suppress all messages\n"
+               "  -r, --recursive              operate recursively on directories\n"
+               "  -R, --dereference-recursive  recursively follow symbolic links\n"
+               "  -v, --verbose                be verbose (a 2nd -v gives more)\n"
+               "  -0 .. -9                     set compression level [default 9]\n"
+               "      --bz2=<command>          set compressor and options for bzip2 format\n"
+               "      --gz=<command>           set compressor and options for gzip format\n"
+               "      --lz=<command>           set compressor and options for lzip format\n"
+               "      --xz=<command>           set compressor and options for xz format\n" );
+  show_help_addr();
+  }
+
+
+int cant_execute( const std::string & command, const int status )
+  {
+  if( verbosity >= 0 )
+    {
+    if( WIFEXITED( status ) )
+      std::fprintf( stderr, "%s: Error executing '%s'. Exit status = %d\n",
+                    program_name, command.c_str(), WEXITSTATUS( status ) );
+    else
+      std::fprintf( stderr, "%s: Can't execute '%s'\n",
+                    program_name, command.c_str() );
+    }
+  return 1;
+  }
+
+
+// Set permissions, owner, and times.
+void set_permissions( const char * const rname, const struct stat & in_stats )
+  {
+  bool warning = false;
+  const mode_t mode = in_stats.st_mode;
+  // chown will in many cases return with EPERM, which can be safely ignored.
+  if( chown( rname, in_stats.st_uid, in_stats.st_gid ) == 0 )
+    { if( chmod( rname, mode ) != 0 ) warning = true; }
+  else
+    if( errno != EPERM ||
+        chmod( rname, mode & ~( S_ISUID | S_ISGID | S_ISVTX ) ) != 0 )
+      warning = true;
+  struct utimbuf t;
+  t.actime = in_stats.st_atime;
+  t.modtime = in_stats.st_mtime;
+  if( utime( rname, &t ) != 0 ) warning = true;
+  if( warning && verbosity >= 2 )
+    show_error( "Can't change output file attributes." );
+  }
+
+
+    // Returns 0 for success, -1 for file skipped, 1 for error.
+int zupdate_file( const std::string & name, const char * const lzip_name,
+                  const std::vector< std::string > & lzip_args2,
+                  const bool force, const bool keep_input_files,
+                  const bool no_rcfile )
+  {
+  static int disable_xz = -1;				// tri-state bool
+  int format_index = -1;
+  std::string rname;					// recompressed name
+
+  const int eindex = extension_index( name );		// search extension
+  if( eindex >= 0 )
+    {
+    format_index = extension_format( eindex );
+    if( format_index == fmt_lz )
+      {
+      if( verbosity >= 2 )
+        std::fprintf( stderr, "%s: Input file '%s' already has '%s' suffix.\n",
+                      program_name, name.c_str(), extension_from( eindex ) );
+      return 0;					// ignore this file
+      }
+    rname.assign( name, 0, name.size() - std::strlen( extension_from( eindex ) ) );
+    rname += ( std::strcmp( extension_to( eindex ), ".tar" ) == 0 ) ?
+             ".tlz" : ".lz";			// keep combined extension
+    }
+  const char * const compressor_name = get_compressor_name( format_index );
+  if( !compressor_name )
+    {
+    if( verbosity >= 2 )
+      std::fprintf( stderr, "%s: Unknown extension in file name '%s' -- ignored.\n",
+                    program_name, name.c_str() );
+    return 0;						// ignore this file
+    }
+
+  struct stat in_stats;
+  if( stat( name.c_str(), &in_stats ) != 0 )		// check input file
+    {
+    if( verbosity >= 0 )
+      std::fprintf( stderr, "%s: Can't stat input file '%s': %s\n",
+                    program_name, name.c_str(), std::strerror( errno ) );
+    return 1;
+    }
+  if( !S_ISREG( in_stats.st_mode ) )
+    {
+    if( verbosity >= 0 )
+      std::fprintf( stderr, "%s: Input file '%s' is not a regular file.\n",
+                    program_name, name.c_str() );
+    return 1;
+    }
+
+  struct stat st;				// not used
+  const std::string rname2( rname + ".lz" );	// produced by lzip < 1.20
+  const bool lz_exists = ( stat( rname.c_str(), &st ) == 0 );
+  // don't modify an existing 'rname.lz'
+  const bool lz_lz_exists = ( stat( rname2.c_str(), &st ) == 0 );
+  if( lz_exists && !force )
+    {
+    if( verbosity >= 0 )
+      std::fprintf( stderr, "%s: Output file '%s' already exists, skipping.\n",
+                    program_name, rname.c_str() );
+    return -1;
+    }
+
+  if( format_index == fmt_xz )
+    {
+    if( disable_xz < 0 )
+      {
+      std::string command( compressor_name ); command += " -V > /dev/null 2>&1";
+      disable_xz = ( std::system( command.c_str() ) != 0 );
+      }
+    if( disable_xz ) return 0;		// ignore this file if no xz installed
+    }
+
+  if( !lz_exists )					// recompress
+    {
+    if( verbosity >= 1 )
+      std::fprintf( stderr, "Recompressing file '%s'\n", name.c_str() );
+    int fda[2];			// pipe between decompressor and compressor
+    if( pipe( fda ) < 0 )
+      { show_error( "Can't create pipe", errno ); return 1; }
+
+    const pid_t pid = fork();
+    if( pid == 0 )				// child1 (decompressor)
+      {
+      if( dup2( fda[1], STDOUT_FILENO ) >= 0 &&
+          close( fda[0] ) == 0 && close( fda[1] ) == 0 )
+        {
+        const std::vector< std::string > & compressor_args =
+          get_compressor_args( format_index );
+        const int size = compressor_args.size();
+        const char ** const argv = new const char *[size+5];
+        argv[0] = compressor_name;
+        for( int i = 0; i < size; ++i ) argv[i+1] = compressor_args[i].c_str();
+        argv[size+1] = "-cd";
+        argv[size+2] = "--";
+        argv[size+3] = name.c_str();
+        argv[size+4] = 0;
+        execvp( argv[0], (char **)argv );
+        }
+      show_exec_error( compressor_name );
+      _exit( 1 );
+      }
+    if( pid < 0 )				// parent
+      { show_fork_error( compressor_name ); return 1; }
+
+    const pid_t pid2 = fork();
+    if( pid2 == 0 )				// child2 (lzip compressor)
+      {
+      if( dup2( fda[0], STDIN_FILENO ) >= 0 &&
+          close( fda[0] ) == 0 && close( fda[1] ) == 0 )
+        {
+        const std::vector< std::string > & lzip_args =
+          get_compressor_args( fmt_lz );
+        const int size = lzip_args.size();
+        const int size2 = lzip_args2.size();
+        const char ** const argv = new const char *[size+size2+5];
+        argv[0] = lzip_name;
+        argv[1] = "-9";
+        for( int i = 0; i < size; ++i ) argv[i+2] = lzip_args[i].c_str();
+        for( int i = 0; i < size2; ++i ) argv[i+size+2] = lzip_args2[i].c_str();
+        argv[size+size2+2] = "-o";
+        argv[size+size2+3] = rname.c_str();
+        argv[size+size2+4] = 0;
+        execvp( argv[0], (char **)argv );
+        }
+      show_exec_error( lzip_name );
+      _exit( 1 );
+      }
+    if( pid2 < 0 )				// parent
+      { show_fork_error( lzip_name ); return 1; }
+
+    close( fda[0] ); close( fda[1] );
+    int retval = wait_for_child( pid, compressor_name );
+    int retval2 = wait_for_child( pid2, lzip_name );
+    if( retval || retval2 )
+      { if( !lz_lz_exists ) std::remove( rname2.c_str() );	// lzip < 1.20
+        std::remove( rname.c_str() ); return 1; }
+    if( stat( rname.c_str(), &st ) != 0 &&
+        ( lz_lz_exists || stat( rname2.c_str(), &st ) != 0 ||
+          std::rename( rname2.c_str(), rname.c_str() ) != 0 ) )
+      { show_file_error( rname.c_str(), "Error renaming output file", errno );
+        return 1; }						// lzip < 1.11
+    set_permissions( rname.c_str(), in_stats );
+    }
+
+  {
+  if( lz_exists && verbosity >= 1 )
+    std::fprintf( stderr, "Comparing file '%s'\n", name.c_str() );
+  std::string zcmp_command( invocation_name );
+  unsigned i =  zcmp_command.size();
+  while( i > 0 && zcmp_command[i-1] != '/' ) --i;
+  zcmp_command.resize( i ); zcmp_command.insert( zcmp_command.begin(), '\'' );
+  zcmp_command += "zcmp' ";				// '[dir/]zcmp'
+  if( no_rcfile ) zcmp_command += "-N ";
+  if( verbosity < 0 ) zcmp_command += "-q ";
+  zcmp_command += '\''; zcmp_command += name;
+  zcmp_command += "' '"; zcmp_command += rname; zcmp_command += '\'';
+  int status = std::system( zcmp_command.c_str() );
+  if( status != 0 )
+    { if( !lz_exists ) std::remove( rname.c_str() );
+      return cant_execute( zcmp_command, status ); }
+  }
+
+  if( !keep_input_files && std::remove( name.c_str() ) != 0 && errno != ENOENT )
+    {
+    if( verbosity >= 0 )
+      std::fprintf( stderr, "%s: Can't delete input file '%s': %s\n",
+                    program_name, name.c_str(), std::strerror( errno ) );
+    return 1;
+    }
+  return 0;
+  }
+
+} // end namespace
+
+
+int main( const int argc, const char * const argv[] )
+  {
+  enum { bz2_opt = 256, gz_opt, lz_opt, xz_opt };
+  int recursive = 0;				// 1 = '-r', 2 = '-R'
+  std::list< std::string > filenames;
+  std::vector< std::string > lzip_args2;	// args to lzip, maybe empty
+  bool force = false;
+  bool keep_input_files = false;
+  bool no_rcfile = false;
+  program_name = "zupdate";
+  invocation_name = ( argc > 0 ) ? argv[0] : program_name;
+
+  const Arg_parser::Option options[] =
+    {
+    { '0',  0,                      Arg_parser::no  },
+    { '1',  0,                      Arg_parser::no  },
+    { '2',  0,                      Arg_parser::no  },
+    { '3',  0,                      Arg_parser::no  },
+    { '4',  0,                      Arg_parser::no  },
+    { '5',  0,                      Arg_parser::no  },
+    { '6',  0,                      Arg_parser::no  },
+    { '7',  0,                      Arg_parser::no  },
+    { '8',  0,                      Arg_parser::no  },
+    { '9',  0,                      Arg_parser::no  },
+    { 'f', "force",                 Arg_parser::no  },
+    { 'h', "help",                  Arg_parser::no  },
+    { 'k', "keep",                  Arg_parser::no  },
+    { 'l', "lzip-verbose",          Arg_parser::no  },
+    { 'M', "format",                Arg_parser::yes },
+    { 'N', "no-rcfile",             Arg_parser::no  },
+    { 'q', "quiet",                 Arg_parser::no  },
+    { 'r', "recursive",             Arg_parser::no  },
+    { 'R', "dereference-recursive", Arg_parser::no  },
+    { 'v', "verbose",               Arg_parser::no  },
+    { 'V', "version",               Arg_parser::no  },
+    { bz2_opt,    "bz2",            Arg_parser::yes },
+    { gz_opt,     "gz",             Arg_parser::yes },
+    { lz_opt,     "lz",             Arg_parser::yes },
+    { xz_opt,     "xz",             Arg_parser::yes },
+    {  0 ,  0,                      Arg_parser::no  } };
+
+  const Arg_parser parser( argc, argv, options );
+  if( parser.error().size() )				// bad option
+    { show_error( parser.error().c_str(), 0, true ); return 1; }
+
+  maybe_process_config_file( parser );
+
+  int argind = 0;
+  for( ; argind < parser.arguments(); ++argind )
+    {
+    const int code = parser.code( argind );
+    if( !code ) break;					// no more options
+    const std::string & arg = parser.argument( argind );
+    switch( code )
+      {
+      case '0': case '1': case '2': case '3': case '4':
+      case '5': case '6': case '7': case '8': case '9':
+                lzip_args2.push_back( "-" ); lzip_args2.back() += code; break;
+      case 'f': force = true; break;
+      case 'h': show_help(); return 0;
+      case 'k': keep_input_files = true; break;
+      case 'l': lzip_args2.push_back( "-v" ); break;
+      case 'M': parse_format_list( arg ); break;
+      case 'N': no_rcfile = true; break;
+      case 'q': verbosity = -1; lzip_args2.push_back( "-q" ); break;
+      case 'r': recursive = 1; break;
+      case 'R': recursive = 2; break;
+      case 'v': if( verbosity < 4 ) ++verbosity; break;
+      case 'V': show_version(); return 0;
+      case bz2_opt: parse_compressor( arg, fmt_bz2, 1 ); break;
+      case gz_opt: parse_compressor( arg, fmt_gz, 1 ); break;
+      case lz_opt: parse_compressor( arg, fmt_lz, 1 ); break;
+      case xz_opt: parse_compressor( arg, fmt_xz, 1 ); break;
+      default : internal_error( "uncaught option." );
+      }
+    } // end process options
+
+#if defined(__MSVCRT__) || defined(__OS2__)
+  setmode( STDIN_FILENO, O_BINARY );
+  setmode( STDOUT_FILENO, O_BINARY );
+#endif
+
+  const char * const lzip_name = get_compressor_name( fmt_lz );
+  if( !lzip_name )
+    { show_error( "Missing name of compressor for lzip format." ); return 1; }
+
+  for( ; argind < parser.arguments(); ++argind )
+    filenames.push_back( parser.argument( argind ) );
+
+  if( filenames.empty() && recursive ) filenames.push_back( "." );
+
+  std::string input_filename;
+  int retval = 0;
+  bool error = false;
+  while( next_filename( filenames, input_filename, error, recursive, true ) )
+    {
+    int tmp = zupdate_file( input_filename, lzip_name, lzip_args2, force,
+                            keep_input_files, no_rcfile );
+    if( tmp < 0 ) error = true;
+    if( tmp > retval ) retval = tmp;
+    if( tmp > 0 ) break;
+    }
+  if( error && retval == 0 ) retval = 1;
+  return retval;
+  }
diff --git a/zutils.cc b/zutils.cc
new file mode 100644
index 0000000..54090ff
--- /dev/null
+++ b/zutils.cc
@@ -0,0 +1,283 @@
+/* Zutils - Utilities dealing with compressed files
+   Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#define _FILE_OFFSET_BITS 64
+
+#include <cerrno>
+#include <csignal>
+#include <cstdio>
+#include <cstdlib>
+#include <cstring>
+#include <string>
+#include <vector>
+#include <stdint.h>
+#include <unistd.h>
+#include <sys/wait.h>
+
+#include "rc.h"
+#include "zutils.h"
+
+
+namespace {
+
+inline bool isvalid_ds( const uint8_t ds )	// lzip valid dictionary_size
+  {
+  enum { min_dictionary_size = 1 << 12,
+         max_dictionary_size = 1 << 29 };
+  unsigned dictionary_size = ( 1 << ( ds & 0x1F ) );
+  if( dictionary_size > min_dictionary_size )
+    dictionary_size -= ( dictionary_size / 16 ) * ( ( ds >> 5 ) & 7 );
+  return ( dictionary_size >= min_dictionary_size &&
+           dictionary_size <= max_dictionary_size );
+  }
+
+
+/* Returns -1 if child not terminated, 2 in case of error, or exit status of
+   child process 'pid'.
+*/
+int child_status( const pid_t pid, const char * const name )
+  {
+  int status;
+  while( true )
+    {
+    const int tmp = waitpid( pid, &status, WNOHANG );
+    if( tmp == -1 && errno != EINTR )
+      {
+      if( verbosity >= 0 )
+        std::fprintf( stderr, "%s: Error checking status of '%s': %s\n",
+                      program_name, name, std::strerror( errno ) );
+      _exit( 2 );
+      }
+    if( tmp == 0 ) return -1;			// child not terminated
+    if( tmp == pid ) break;			// child terminated
+    }
+  if( WIFEXITED( status ) ) return WEXITSTATUS( status );
+  if( WIFSIGNALED( status ) && WTERMSIG( status ) == SIGPIPE ) return 0;
+  return 2;
+  }
+
+} // end namespace
+
+
+/* Returns the number of bytes really read.
+   If (returned value < size) and (errno == 0), means EOF was reached.
+*/
+int readblock( const int fd, uint8_t * const buf, const int size )
+  {
+  int sz = 0;
+  errno = 0;
+  while( sz < size )
+    {
+    const int n = read( fd, buf + sz, size - sz );
+    if( n > 0 ) sz += n;
+    else if( n == 0 ) break;				// EOF
+    else if( errno != EINTR ) break;
+    errno = 0;
+    }
+  return sz;
+  }
+
+
+/* Returns the number of bytes really written.
+   If (returned value < size), it is always an error.
+*/
+int writeblock( const int fd, const uint8_t * const buf, const int size )
+  {
+  int sz = 0;
+  errno = 0;
+  while( sz < size )
+    {
+    const int n = write( fd, buf + sz, size - sz );
+    if( n > 0 ) sz += n;
+    else if( n < 0 && errno != EINTR ) break;
+    errno = 0;
+    }
+  return sz;
+  }
+
+
+// Empty filename means stdin.
+//
+bool feed_data( const std::string & filename, const int infd, const int outfd,
+                const uint8_t * magic_data, const int magic_size )
+  {
+  if( magic_size && writeblock( outfd, magic_data, magic_size ) != magic_size )
+    { show_error( "Write error", errno ); return false; }
+  enum { buffer_size = 4096 };
+  uint8_t buffer[buffer_size];
+  while( true )
+    {
+    const int size = readblock( infd, buffer, buffer_size );
+    if( size != buffer_size && errno )
+      { const char * const name = filename.empty() ? "-" : filename.c_str();
+        show_file_error( name, "Read error", errno ); return false; }
+    if( size > 0 && writeblock( outfd, buffer, size ) != size )
+      { show_error( "Write error", errno ); return false; }
+    if( size < buffer_size ) break;
+    }
+  return true;
+  }
+
+
+bool good_status( const Children & children, const bool finished )
+  {
+  bool error = false;
+  for( int i = 0; i < 2; ++i )
+    {
+    const pid_t pid = children.pid[i];
+    if( pid )
+      {
+      const char * const name =
+        ( i == 0 ) ? "data feeder" : children.compressor_name;
+      // even if compressor finished, trailing data may remain in data feeder
+      if( i == 0 || !finished )
+        {
+        const int tmp = child_status( pid, name );
+        if( tmp < 0 )				// child not terminated
+          { kill( pid, SIGTERM ); wait_for_child( pid, name ); }
+        else if( tmp != 0 ) error = true;	// child status != 0
+        }
+      else
+        if( wait_for_child( pid, name ) != 0 ) error = true;
+      }
+    }
+  return !error;
+  }
+
+
+bool set_data_feeder( const std::string & filename, int * const infdp,
+                      Children & children, int format_index )
+  {
+  uint8_t magic_data[magic_buf_size];
+  int magic_size = 0;
+  if( format_index < 0 )
+    format_index = test_format( *infdp, magic_data, &magic_size );
+  children.compressor_name = get_compressor_name( format_index );
+
+  if( children.compressor_name )	// compressed
+    {
+    int fda[2];				// pipe from feeder
+    int fda2[2];			// pipe from compressor
+    if( pipe( fda ) < 0 || pipe( fda2 ) < 0 )
+      { show_error( "Can't create pipe", errno ); return false; }
+    const int old_infd = *infdp;
+    *infdp = fda2[0];
+    const pid_t pid = fork();
+    if( pid == 0 )			// child 1 (compressor feeder)
+      {
+      if( close( fda[0] ) != 0 ||
+          close( fda2[0] ) != 0 || close( fda2[1] ) != 0 ||
+          !feed_data( filename, old_infd, fda[1], magic_data, magic_size ) )
+        _exit( 2 );
+      if( close( fda[1] ) != 0 )
+        { show_close_error(); _exit( 2 ); }
+      _exit( 0 );
+      }
+    if( pid < 0 )			// parent
+      { show_fork_error( "data feeder" ); return false; }
+
+    const pid_t pid2 = fork();
+    if( pid2 == 0 )			// child 2 (compressor)
+      {
+      if( dup2( fda[0], STDIN_FILENO ) >= 0 &&
+          dup2( fda2[1], STDOUT_FILENO ) >= 0 &&
+          close( fda[0] ) == 0 && close( fda[1] ) == 0 &&
+          close( fda2[0] ) == 0 && close( fda2[1] ) == 0 )
+        {
+        const std::vector< std::string > & compressor_args =
+          get_compressor_args( format_index );
+        const int size = compressor_args.size();
+        const char ** const argv = new const char *[size+3];
+        argv[0] = children.compressor_name;
+        for( int i = 0; i < size; ++i )
+          argv[i+1] = compressor_args[i].c_str();
+        argv[size+1] = ( verbosity >= 0 ) ? "-d" : "-dq";
+        argv[size+2] = 0;
+        execvp( argv[0], (char **)argv );
+        }
+      show_exec_error( children.compressor_name );
+      _exit( 2 );
+      }
+    if( pid2 < 0 )			// parent
+      { show_fork_error( children.compressor_name ); return false; }
+
+    close( fda[0] ); close( fda[1] ); close( fda2[1] );
+    children.pid[0] = pid;
+    children.pid[1] = pid2;
+    }
+  else					// uncompressed
+    {
+    int fda[2];				// pipe from feeder
+    if( pipe( fda ) < 0 )
+      { show_error( "Can't create pipe", errno ); return false; }
+    const int old_infd = *infdp;
+    *infdp = fda[0];
+    const pid_t pid = fork();
+    if( pid == 0 )			// child (feeder)
+      {
+      if( close( fda[0] ) != 0 ||
+          !feed_data( filename, old_infd, fda[1], magic_data, magic_size ) )
+        _exit( 2 );
+      if( close( fda[1] ) != 0 )
+        { show_close_error(); _exit( 2 ); }
+      _exit( 0 );
+      }
+    if( pid < 0 )			// parent
+      { show_fork_error( "data feeder" ); return false; }
+    close( fda[1] );
+    children.pid[0] = pid;
+    children.pid[1] = 0;
+    }
+  return true;
+  }
+
+
+// Returns format index or -1 if uncompressed
+//
+int test_format( const int infd, uint8_t magic_data[],
+                 int * const magic_sizep )
+  {
+  enum { bzip2_magic_size = 3,
+          gzip_magic_size = 2,
+          lzip_magic_size = 5,
+            xz_magic_size = 5 };
+  const uint8_t bzip2_magic[bzip2_magic_size] =
+    { 0x42, 0x5A, 0x68 };				// "BZh"
+  const uint8_t gzip_magic[gzip_magic_size] =
+    { 0x1F, 0x8B };
+  const uint8_t lzip_magic[lzip_magic_size] =
+    { 0x4C, 0x5A, 0x49, 0x50, 0x01 };			// "LZIP\001"
+  const uint8_t xz_magic[xz_magic_size] =
+    { 0xFD, 0x37, 0x7A, 0x58, 0x5A };			// 0xFD, "7zXZ"
+
+  *magic_sizep = readblock( infd, magic_data, magic_buf_size );
+  if( *magic_sizep == magic_buf_size )
+    {
+    if( std::memcmp( magic_data, bzip2_magic, bzip2_magic_size ) == 0 &&
+        magic_data[3] >= '1' && magic_data[3] <= '9' &&
+        std::memcmp( magic_data + 4, "1AY&SY", 6 ) == 0 )
+      return fmt_bz2;
+    if( std::memcmp( magic_data, gzip_magic, gzip_magic_size ) == 0 )
+      return fmt_gz;
+    if( std::memcmp( magic_data, lzip_magic, lzip_magic_size ) == 0 &&
+        isvalid_ds( magic_data[lzip_magic_size] ) )
+      return fmt_lz;
+    if( std::memcmp( magic_data, xz_magic, xz_magic_size ) == 0 )
+      return fmt_xz;
+    }
+  return -1;
+  }
diff --git a/zutils.h b/zutils.h
new file mode 100644
index 0000000..064af51
--- /dev/null
+++ b/zutils.h
@@ -0,0 +1,37 @@
+/* Zutils - Utilities dealing with compressed files
+   Copyright (C) 2009-2021 Antonio Diaz Diaz.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 2 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+int readblock( const int fd, uint8_t * const buf, const int size );
+int writeblock( const int fd, const uint8_t * const buf, const int size );
+bool feed_data( const std::string & filename, const int infd, const int outfd,
+                const uint8_t * magic_data, const int magic_size );
+
+struct Children
+  {
+  const char * compressor_name;
+  pid_t pid[2];				// data feeder, compressor
+  };
+bool good_status( const Children & children, const bool finished );
+bool set_data_feeder( const std::string & filename, int * const infdp,
+                      Children & children, int format_index );
+
+enum { magic_buf_size = 10 };		// >= longest extended magic (bzip2)
+
+// Returns format index or -1 if uncompressed
+//
+int test_format( const int infd, uint8_t magic_data[],
+                 int * const magic_sizep );
diff --git a/zutilsrc b/zutilsrc
new file mode 100644
index 0000000..04a1d69
--- /dev/null
+++ b/zutilsrc
@@ -0,0 +1,16 @@
+#
+#                 Runtime Configuration file for Zutils
+#
+# Zutils looks for this file in:
+#	1 - $HOME/.zutilsrc
+#	2 - ${sysconfdir}/zutilsrc
+
+# This file sets the compressor and options to be used for each format.
+# The command line options override compressors specified in this file.
+# Syntax: <format> = <compressor> [options]
+# Uncomment each line you want to take effect.
+
+#	bz2 = lbzip2 -n2
+#	gz = pigz -p2
+#	lz = plzip -n2
+#	xz = pixz -p2