1 \input texinfo @c -*-texinfo-*-
2 @comment %**start of header
5 @settitle GNU tar @value{VERSION}
14 @c 1. Pay attention to @FIXME{}s and @UNREVISED{}s
15 @c 2. Before creating final variant:
16 @c 2.1. Run `make check-options' to make sure all options are properly
18 @c 2.2. Run `make master-menu' (see comment before the master menu).
20 @include rendition.texi
25 @c Put everything in one index (arbitrarily chosen to be the concept index).
33 This manual is for @acronym{GNU} @command{tar} (version
34 @value{VERSION}, @value{UPDATED}), which creates and extracts files
37 Copyright @copyright{} 1992, 1994, 1995, 1996, 1997, 1999, 2000, 2001,
38 2003, 2004, 2005, 2006, 2007 Free Software Foundation, Inc.
41 Permission is granted to copy, distribute and/or modify this document
42 under the terms of the GNU Free Documentation License, Version 1.1 or
43 any later version published by the Free Software Foundation; with no
44 Invariant Sections, with the Front-Cover Texts being ``A GNU Manual,''
45 and with the Back-Cover Texts as in (a) below. A copy of the license
46 is included in the section entitled "GNU Free Documentation License".
48 (a) The FSF's Back-Cover Text is: ``You are free to copy and modify
49 this GNU Manual. Buying copies from GNU Press supports the FSF in
50 developing GNU and promoting software freedom.''
54 @dircategory Archiving
56 * Tar: (tar). Making tape (or disk) archives.
59 @dircategory Individual utilities
61 * tar: (tar)tar invocation. Invoking @GNUTAR{}.
64 @shorttitlepage @acronym{GNU} @command{tar}
67 @title @acronym{GNU} tar: an archiver tool
68 @subtitle @value{RENDITION} @value{VERSION}, @value{UPDATED}
69 @author John Gilmore, Jay Fenlason et al.
72 @vskip 0pt plus 1filll
78 @top @acronym{GNU} tar: an archiver tool
83 @cindex archiving files
85 The first part of this master menu lists the major nodes in this Info
86 document. The rest of the menu lists all the lower level nodes.
89 @c The master menu goes here.
91 @c NOTE: To update it from within Emacs, make sure mastermenu.el is
92 @c loaded and run texinfo-master-menu.
93 @c To update it from the command line, run
104 * Date input formats::
111 * Configuring Help Summary::
112 * Fixing Snapshot Files::
115 * Free Software Needs Free Documentation::
116 * Copying This Manual::
117 * Index of Command Line Options::
121 --- The Detailed Node Listing ---
125 * Book Contents:: What this Book Contains
126 * Definitions:: Some Definitions
127 * What tar Does:: What @command{tar} Does
128 * Naming tar Archives:: How @command{tar} Archives are Named
129 * Authors:: @GNUTAR{} Authors
130 * Reports:: Reporting bugs or suggestions
132 Tutorial Introduction to @command{tar}
135 * stylistic conventions::
136 * basic tar options:: Basic @command{tar} Operations and Options
137 * frequent operations::
138 * Two Frequent Options::
139 * create:: How to Create Archives
140 * list:: How to List Archives
141 * extract:: How to Extract Members from an Archive
144 Two Frequently Used Options
150 How to Create Archives
152 * prepare for examples::
153 * Creating the archive::
162 How to Extract Members from an Archive
164 * extracting archives::
167 * extracting untrusted archives::
173 * using tar options::
181 The Three Option Styles
183 * Long Options:: Long Option Style
184 * Short Options:: Short Option Style
185 * Old Options:: Old Option Style
186 * Mixing:: Mixing Option Styles
188 All @command{tar} Options
190 * Operation Summary::
192 * Short Option Summary::
204 Advanced @GNUTAR{} Operations
213 How to Add Files to Existing Archives: @option{--append}
215 * appending files:: Appending Files to an Archive
222 Options Used by @option{--create}
224 * override:: Overriding File Metadata.
225 * Ignore Failed Read::
227 Options Used by @option{--extract}
229 * Reading:: Options to Help Read Archives
230 * Writing:: Changing How @command{tar} Writes Files
231 * Scarce:: Coping with Scarce Resources
233 Options to Help Read Archives
235 * read full records::
238 Changing How @command{tar} Writes Files
240 * Dealing with Old Files::
241 * Overwrite Old Files::
246 * Data Modification Times::
247 * Setting Access Permissions::
248 * Directory Modification Times and Permissions::
249 * Writing to Standard Output::
250 * Writing to an External Program::
253 Coping with Scarce Resources
258 Performing Backups and Restoring Files
260 * Full Dumps:: Using @command{tar} to Perform Full Dumps
261 * Incremental Dumps:: Using @command{tar} to Perform Incremental Dumps
262 * Backup Levels:: Levels of Backups
263 * Backup Parameters:: Setting Parameters for Backups and Restoration
264 * Scripted Backups:: Using the Backup Scripts
265 * Scripted Restoration:: Using the Restore Script
267 Setting Parameters for Backups and Restoration
269 * General-Purpose Variables::
270 * Magnetic Tape Control::
272 * backup-specs example:: An Example Text of @file{Backup-specs}
274 Choosing Files and Names for @command{tar}
276 * file:: Choosing the Archive's Name
277 * Selecting Archive Members::
278 * files:: Reading Names from a File
279 * exclude:: Excluding Some Files
280 * wildcards:: Wildcards Patterns and Matching
281 * quoting styles:: Ways of Quoting Special Characters in Names
282 * transform:: Modifying File and Member Names
283 * after:: Operating Only on New Files
284 * recurse:: Descending into Directories
285 * one:: Crossing File System Boundaries
287 Reading Names from a File
293 * problems with exclude::
295 Wildcards Patterns and Matching
297 * controlling pattern-matching::
299 Crossing File System Boundaries
301 * directory:: Changing Directory
302 * absolute:: Absolute File Names
306 * General date syntax:: Common rules.
307 * Calendar date items:: 19 Dec 1994.
308 * Time of day items:: 9:20pm.
309 * Time zone items:: @sc{est}, @sc{pdt}, @sc{gmt}.
310 * Day of week items:: Monday and others.
311 * Relative items in date strings:: next tuesday, 2 years ago.
312 * Pure numbers in date strings:: 19931219, 1440.
313 * Seconds since the Epoch:: @@1078100502.
314 * Specifying time zone rules:: TZ="America/New_York", TZ="UTC0".
315 * Authors of get_date:: Bellovin, Eggert, Salz, Berets, et al.
317 Controlling the Archive Format
319 * Compression:: Using Less Space through Compression
320 * Attributes:: Handling File Attributes
321 * Portability:: Making @command{tar} Archives More Portable
322 * cpio:: Comparison of @command{tar} and @command{cpio}
324 Using Less Space through Compression
326 * gzip:: Creating and Reading Compressed Archives
327 * sparse:: Archiving Sparse Files
329 Making @command{tar} Archives More Portable
331 * Portable Names:: Portable Names
332 * dereference:: Symbolic Links
333 * old:: Old V7 Archives
334 * ustar:: Ustar Archives
335 * gnu:: GNU and old GNU format archives.
336 * posix:: @acronym{POSIX} archives
337 * Checksumming:: Checksumming Problems
338 * Large or Negative Values:: Large files, negative time stamps, etc.
339 * Other Tars:: How to Extract GNU-Specific Data Using
340 Other @command{tar} Implementations
342 @GNUTAR{} and @acronym{POSIX} @command{tar}
344 * PAX keywords:: Controlling Extended Header Keywords.
346 How to Extract GNU-Specific Data Using Other @command{tar} Implementations
348 * Split Recovery:: Members Split Between Volumes
349 * Sparse Recovery:: Sparse Members
351 Tapes and Other Archive Media
353 * Device:: Device selection and switching
354 * Remote Tape Server::
355 * Common Problems and Solutions::
356 * Blocking:: Blocking
357 * Many:: Many archives on one tape
358 * Using Multiple Tapes:: Using Multiple Tapes
359 * label:: Including a Label in the Archive
365 * Format Variations:: Format Variations
366 * Blocking Factor:: The Blocking Factor of an Archive
368 Many Archives on One Tape
370 * Tape Positioning:: Tape Positions and Tape Marks
371 * mt:: The @command{mt} Utility
375 * Multi-Volume Archives:: Archives Longer than One Tape or Disk
376 * Tape Files:: Tape Files
377 * Tarcat:: Concatenate Volumes into a Single Archive
382 * Standard:: Basic Tar Format
383 * Extensions:: @acronym{GNU} Extensions to the Archive Format
384 * Sparse Formats:: Storing Sparse Files
391 * PAX 0:: PAX Format, Versions 0.0 and 0.1
392 * PAX 1:: PAX Format, Version 1.0
396 * Generate Mode:: File Generation Mode.
397 * Status Mode:: File Status Mode.
398 * Exec Mode:: Synchronous Execution mode.
402 * GNU Free Documentation License:: License for copying this manual
408 @chapter Introduction
411 and manipulates @dfn{archives} which are actually collections of
412 many other files; the program provides users with an organized and
413 systematic method for controlling a large amount of data.
414 The name ``tar'' originally came from the phrase ``Tape ARchive'', but
415 archives need not (and these days, typically do not) reside on tapes.
418 * Book Contents:: What this Book Contains
419 * Definitions:: Some Definitions
420 * What tar Does:: What @command{tar} Does
421 * Naming tar Archives:: How @command{tar} Archives are Named
422 * Authors:: @GNUTAR{} Authors
423 * Reports:: Reporting bugs or suggestions
427 @section What this Book Contains
429 The first part of this chapter introduces you to various terms that will
430 recur throughout the book. It also tells you who has worked on @GNUTAR{}
431 and its documentation, and where you should send bug reports
434 The second chapter is a tutorial (@pxref{Tutorial}) which provides a
435 gentle introduction for people who are new to using @command{tar}. It is
436 meant to be self contained, not requiring any reading from subsequent
437 chapters to make sense. It moves from topic to topic in a logical,
438 progressive order, building on information already explained.
440 Although the tutorial is paced and structured to allow beginners to
441 learn how to use @command{tar}, it is not intended solely for beginners.
442 The tutorial explains how to use the three most frequently used
443 operations (@samp{create}, @samp{list}, and @samp{extract}) as well as
444 two frequently used options (@samp{file} and @samp{verbose}). The other
445 chapters do not refer to the tutorial frequently; however, if a section
446 discusses something which is a complex variant of a basic concept, there
447 may be a cross reference to that basic concept. (The entire book,
448 including the tutorial, assumes that the reader understands some basic
449 concepts of using a Unix-type operating system; @pxref{Tutorial}.)
451 The third chapter presents the remaining five operations, and
452 information about using @command{tar} options and option syntax.
454 @FIXME{this sounds more like a @acronym{GNU} Project Manuals Concept [tm] more
455 than the reality. should think about whether this makes sense to say
456 here, or not.} The other chapters are meant to be used as a
457 reference. Each chapter presents everything that needs to be said
458 about a specific topic.
460 One of the chapters (@pxref{Date input formats}) exists in its
461 entirety in other @acronym{GNU} manuals, and is mostly self-contained.
462 In addition, one section of this manual (@pxref{Standard}) contains a
463 big quote which is taken directly from @command{tar} sources.
465 In general, we give both long and short (abbreviated) option names
466 at least once in each section where the relevant option is covered, so
467 that novice readers will become familiar with both styles. (A few
468 options have no short versions, and the relevant sections will
472 @section Some Definitions
476 The @command{tar} program is used to create and manipulate @command{tar}
477 archives. An @dfn{archive} is a single file which contains the contents
478 of many files, while still identifying the names of the files, their
479 owner(s), and so forth. (In addition, archives record access
480 permissions, user and group, size in bytes, and data modification time.
481 Some archives also record the file names in each archived directory, as
482 well as other file and directory information.) You can use @command{tar}
483 to @dfn{create} a new archive in a specified directory.
486 @cindex archive member
489 The files inside an archive are called @dfn{members}. Within this
490 manual, we use the term @dfn{file} to refer only to files accessible in
491 the normal ways (by @command{ls}, @command{cat}, and so forth), and the term
492 @dfn{member} to refer only to the members of an archive. Similarly, a
493 @dfn{file name} is the name of a file, as it resides in the file system,
494 and a @dfn{member name} is the name of an archive member within the
499 The term @dfn{extraction} refers to the process of copying an archive
500 member (or multiple members) into a file in the file system. Extracting
501 all the members of an archive is often called @dfn{extracting the
502 archive}. The term @dfn{unpack} can also be used to refer to the
503 extraction of many or all the members of an archive. Extracting an
504 archive does not destroy the archive's structure, just as creating an
505 archive does not destroy the copies of the files that exist outside of
506 the archive. You may also @dfn{list} the members in a given archive
507 (this is often thought of as ``printing'' them to the standard output,
508 or the command line), or @dfn{append} members to a pre-existing archive.
509 All of these operations can be performed using @command{tar}.
512 @section What @command{tar} Does
515 The @command{tar} program provides the ability to create @command{tar}
516 archives, as well as various other kinds of manipulation. For example,
517 you can use @command{tar} on previously created archives to extract files,
518 to store additional files, or to update or list files which were already
521 Initially, @command{tar} archives were used to store files conveniently on
522 magnetic tape. The name @command{tar} comes from this use; it stands for
523 @code{t}ape @code{ar}chiver. Despite the utility's name, @command{tar} can
524 direct its output to available devices, files, or other programs (using
525 pipes). @command{tar} may even access remote devices or files (as archives).
527 You can use @command{tar} archives in many ways. We want to stress a few
528 of them: storage, backup, and transportation.
530 @FIXME{the following table entries need a bit of work.}
533 Often, @command{tar} archives are used to store related files for
534 convenient file transfer over a network. For example, the
535 @acronym{GNU} Project distributes its software bundled into
536 @command{tar} archives, so that all the files relating to a particular
537 program (or set of related programs) can be transferred as a single
540 A magnetic tape can store several files in sequence. However, the tape
541 has no names for these files; it only knows their relative position on
542 the tape. One way to store several files on one tape and retain their
543 names is by creating a @command{tar} archive. Even when the basic transfer
544 mechanism can keep track of names, as FTP can, the nuisance of handling
545 multiple files, directories, and multiple links makes @command{tar}
548 Archive files are also used for long-term storage. You can think of
549 this as transportation from the present into the future. (It is a
550 science-fiction idiom that you can move through time as well as in
551 space; the idea here is that @command{tar} can be used to move archives in
552 all dimensions, even time!)
555 Because the archive created by @command{tar} is capable of preserving
556 file information and directory structure, @command{tar} is commonly
557 used for performing full and incremental backups of disks. A backup
558 puts a collection of files (possibly pertaining to many users and
559 projects) together on a disk or a tape. This guards against
560 accidental destruction of the information in those files.
561 @GNUTAR{} has special features that allow it to be
562 used to make incremental and full dumps of all the files in a
566 You can create an archive on one system, transfer it to another system,
567 and extract the contents there. This allows you to transport a group of
568 files from one system to another.
571 @node Naming tar Archives
572 @section How @command{tar} Archives are Named
574 Conventionally, @command{tar} archives are given names ending with
575 @samp{.tar}. This is not necessary for @command{tar} to operate properly,
576 but this manual follows that convention in order to accustom readers to
577 it and to make examples more clear.
582 Often, people refer to @command{tar} archives as ``@command{tar} files,'' and
583 archive members as ``files'' or ``entries''. For people familiar with
584 the operation of @command{tar}, this causes no difficulty. However, in
585 this manual, we consistently refer to ``archives'' and ``archive
586 members'' to make learning to use @command{tar} easier for novice users.
589 @section @GNUTAR{} Authors
591 @GNUTAR{} was originally written by John Gilmore,
592 and modified by many people. The @acronym{GNU} enhancements were
593 written by Jay Fenlason, then Joy Kendall, and the whole package has
594 been further maintained by Thomas Bushnell, n/BSG, Fran@,{c}ois
595 Pinard, Paul Eggert, and finally Sergey Poznyakoff with the help of
596 numerous and kind users.
598 We wish to stress that @command{tar} is a collective work, and owes much to
599 all those people who reported problems, offered solutions and other
600 insights, or shared their thoughts and suggestions. An impressive, yet
601 partial list of those contributors can be found in the @file{THANKS}
602 file from the @GNUTAR{} distribution.
604 @FIXME{i want all of these names mentioned, Absolutely. BUT, i'm not
605 sure i want to spell out the history in this detail, at least not for
606 the printed book. i'm just not sure it needs to be said this way.
607 i'll think about it.}
609 @FIXME{History is more important, and surely more interesting, than
610 actual names. Quoting names without history would be meaningless. FP}
612 Jay Fenlason put together a draft of a @GNUTAR{}
613 manual, borrowing notes from the original man page from John Gilmore.
614 This was withdrawn in version 1.11. Thomas Bushnell, n/BSG and Amy
615 Gorin worked on a tutorial and manual for @GNUTAR{}.
616 Fran@,{c}ois Pinard put version 1.11.8 of the manual together by
617 taking information from all these sources and merging them. Melissa
618 Weisshaus finally edited and redesigned the book to create version
619 1.12. The book for versions from 1.14 up to @value{VERSION} were edited
620 by the current maintainer, Sergey Poznyakoff.
622 For version 1.12, Daniel Hagerty contributed a great deal of technical
623 consulting. In particular, he is the primary author of @ref{Backups}.
625 In July, 2003 @GNUTAR{} was put on CVS at savannah.gnu.org
626 (see @url{http://savannah.gnu.org/projects/tar}), and
627 active development and maintenance work has started
628 again. Currently @GNUTAR{} is being maintained by Paul Eggert, Sergey
629 Poznyakoff and Jeff Bailey.
631 Support for @acronym{POSIX} archives was added by Sergey Poznyakoff.
634 @section Reporting bugs or suggestions
637 @cindex reporting bugs
638 If you find problems or have suggestions about this program or manual,
639 please report them to @file{bug-tar@@gnu.org}.
641 When reporting a bug, please be sure to include as much detail as
642 possible, in order to reproduce it. @FIXME{Be more specific, I'd
643 like to make this node as detailed as 'Bug reporting' node in Emacs
647 @chapter Tutorial Introduction to @command{tar}
649 This chapter guides you through some basic examples of three @command{tar}
650 operations: @option{--create}, @option{--list}, and @option{--extract}. If
651 you already know how to use some other version of @command{tar}, then you
652 may not need to read this chapter. This chapter omits most complicated
653 details about how @command{tar} works.
657 * stylistic conventions::
658 * basic tar options:: Basic @command{tar} Operations and Options
659 * frequent operations::
660 * Two Frequent Options::
661 * create:: How to Create Archives
662 * list:: How to List Archives
663 * extract:: How to Extract Members from an Archive
668 @section Assumptions this Tutorial Makes
670 This chapter is paced to allow beginners to learn about @command{tar}
671 slowly. At the same time, we will try to cover all the basic aspects of
672 these three operations. In order to accomplish both of these tasks, we
673 have made certain assumptions about your knowledge before reading this
674 manual, and the hardware you will be using:
678 Before you start to work through this tutorial, you should understand
679 what the terms ``archive'' and ``archive member'' mean
680 (@pxref{Definitions}). In addition, you should understand something
681 about how Unix-type operating systems work, and you should know how to
682 use some basic utilities. For example, you should know how to create,
683 list, copy, rename, edit, and delete files and directories; how to
684 change between directories; and how to figure out where you are in the
685 file system. You should have some basic understanding of directory
686 structure and how files are named according to which directory they are
687 in. You should understand concepts such as standard output and standard
688 input, what various definitions of the term ``argument'' mean, and the
689 differences between relative and absolute file names. @FIXME{and what
693 This manual assumes that you are working from your own home directory
694 (unless we state otherwise). In this tutorial, you will create a
695 directory to practice @command{tar} commands in. When we show file names,
696 we will assume that those names are relative to your home directory.
697 For example, my home directory is @file{/home/fsf/melissa}. All of
698 my examples are in a subdirectory of the directory named by that file
699 name; the subdirectory is called @file{practice}.
702 In general, we show examples of archives which exist on (or can be
703 written to, or worked with from) a directory on a hard disk. In most
704 cases, you could write those archives to, or work with them on any other
705 device, such as a tape drive. However, some of the later examples in
706 the tutorial and next chapter will not work on tape drives.
707 Additionally, working with tapes is much more complicated than working
708 with hard disks. For these reasons, the tutorial does not cover working
709 with tape drives. @xref{Media}, for complete information on using
710 @command{tar} archives with tape drives.
712 @FIXME{this is a cop out. need to add some simple tape drive info.}
715 @node stylistic conventions
716 @section Stylistic Conventions
718 In the examples, @samp{$} represents a typical shell prompt. It
719 precedes lines you should type; to make this more clear, those lines are
720 shown in @kbd{this font}, as opposed to lines which represent the
721 computer's response; those lines are shown in @code{this font}, or
722 sometimes @samp{like this}.
724 @c When we have lines which are too long to be
725 @c displayed in any other way, we will show them like this:
727 @node basic tar options
728 @section Basic @command{tar} Operations and Options
730 @command{tar} can take a wide variety of arguments which specify and define
731 the actions it will have on the particular set of files or the archive.
732 The main types of arguments to @command{tar} fall into one of two classes:
733 operations, and options.
735 Some arguments fall into a class called @dfn{operations}; exactly one of
736 these is both allowed and required for any instance of using @command{tar};
737 you may @emph{not} specify more than one. People sometimes speak of
738 @dfn{operating modes}. You are in a particular operating mode when you
739 have specified the operation which specifies it; there are eight
740 operations in total, and thus there are eight operating modes.
742 The other arguments fall into the class known as @dfn{options}. You are
743 not required to specify any options, and you are allowed to specify more
744 than one at a time (depending on the way you are using @command{tar} at
745 that time). Some options are used so frequently, and are so useful for
746 helping you type commands more carefully that they are effectively
747 ``required''. We will discuss them in this chapter.
749 You can write most of the @command{tar} operations and options in any
750 of three forms: long (mnemonic) form, short form, and old style. Some
751 of the operations and options have no short or ``old'' forms; however,
752 the operations and options which we will cover in this tutorial have
753 corresponding abbreviations. @FIXME{make sure this is still the case,
754 at the end}We will indicate those abbreviations appropriately to get
755 you used to seeing them. (Note that the ``old style'' option forms
756 exist in @GNUTAR{} for compatibility with Unix
757 @command{tar}. In this book we present a full discussion of this way
758 of writing options and operations (@pxref{Old Options}), and we discuss
759 the other two styles of writing options (@xref{Long Options}, and
760 @pxref{Short Options}).
762 In the examples and in the text of this tutorial, we usually use the
763 long forms of operations and options; but the ``short'' forms produce
764 the same result and can make typing long @command{tar} commands easier.
765 For example, instead of typing
768 @kbd{tar --create --verbose --file=afiles.tar apple angst aspic}
774 @kbd{tar -c -v -f afiles.tar apple angst aspic}
780 @kbd{tar -cvf afiles.tar apple angst aspic}
784 For more information on option syntax, see @ref{Advanced tar}. In
785 discussions in the text, when we name an option by its long form, we
786 also give the corresponding short option in parentheses.
788 The term, ``option'', can be confusing at times, since ``operations''
789 are often lumped in with the actual, @emph{optional} ``options'' in certain
790 general class statements. For example, we just talked about ``short and
791 long forms of options and operations''. However, experienced @command{tar}
792 users often refer to these by shorthand terms such as, ``short and long
793 options''. This term assumes that the ``operations'' are included, also.
794 Context will help you determine which definition of ``options'' to use.
796 Similarly, the term ``command'' can be confusing, as it is often used in
797 two different ways. People sometimes refer to @command{tar} ``commands''.
798 A @command{tar} @dfn{command} is the entire command line of user input
799 which tells @command{tar} what to do --- including the operation, options,
800 and any arguments (file names, pipes, other commands, etc.). However,
801 you will also sometimes hear the term ``the @command{tar} command''. When
802 the word ``command'' is used specifically like this, a person is usually
803 referring to the @command{tar} @emph{operation}, not the whole line.
804 Again, use context to figure out which of the meanings the speaker
807 @node frequent operations
808 @section The Three Most Frequently Used Operations
810 Here are the three most frequently used operations (both short and long
811 forms), as well as a brief description of their meanings. The rest of
812 this chapter will cover how to use these operations in detail. We will
813 present the rest of the operations in the next chapter.
818 Create a new @command{tar} archive.
821 List the contents of an archive.
824 Extract one or more members from an archive.
827 @node Two Frequent Options
828 @section Two Frequently Used Options
830 To understand how to run @command{tar} in the three operating modes listed
831 previously, you also need to understand how to use two of the options to
832 @command{tar}: @option{--file} (which takes an archive file as an argument)
833 and @option{--verbose}. (You are usually not @emph{required} to specify
834 either of these options when you run @command{tar}, but they can be very
835 useful in making things more clear and helping you avoid errors.)
844 @unnumberedsubsec The @option{--file} Option
847 @xopindex{file, tutorial}
848 @item --file=@var{archive-name}
849 @itemx -f @var{archive-name}
850 Specify the name of an archive file.
853 You can specify an argument for the @option{--file=@var{archive-name}} (@option{-f @var{archive-name}}) option whenever you
854 use @command{tar}; this option determines the name of the archive file
855 that @command{tar} will work on.
858 If you don't specify this argument, then @command{tar} will examine
859 the environment variable @env{TAPE}. If it is set, its value will be
860 used as the archive name. Otherwise, @command{tar} will use the
861 default archive, determined at the compile time. Usually it is
862 standard output or some physical tape drive attached to your machine
863 (you can verify what the default is by running @kbd{tar
864 --show-defaults}, @pxref{defaults}). If there is no tape drive
865 attached, or the default is not meaningful, then @command{tar} will
866 print an error message. The error message might look roughly like one
870 tar: can't open /dev/rmt8 : No such device or address
871 tar: can't open /dev/rsmt0 : I/O error
875 To avoid confusion, we recommend that you always specify an archive file
876 name by using @option{--file=@var{archive-name}} (@option{-f @var{archive-name}}) when writing your @command{tar} commands.
877 For more information on using the @option{--file=@var{archive-name}} (@option{-f @var{archive-name}}) option, see
880 @node verbose tutorial
881 @unnumberedsubsec The @option{--verbose} Option
884 @xopindex{verbose, introduced}
887 Show the files being worked on as @command{tar} is running.
890 @option{--verbose} (@option{-v}) shows details about the results of running
891 @command{tar}. This can be especially useful when the results might not be
892 obvious. For example, if you want to see the progress of @command{tar} as
893 it writes files into the archive, you can use the @option{--verbose}
894 option. In the beginning, you may find it useful to use
895 @option{--verbose} at all times; when you are more accustomed to
896 @command{tar}, you will likely want to use it at certain times but not at
897 others. We will use @option{--verbose} at times to help make something
898 clear, and we will give many examples both using and not using
899 @option{--verbose} to show the differences.
901 Each instance of @option{--verbose} on the command line increases the
902 verbosity level by one, so if you need more details on the output,
905 When reading archives (@option{--list}, @option{--extract},
906 @option{--diff}), @command{tar} by default prints only the names of
907 the members being extracted. Using @option{--verbose} will show a full,
908 @command{ls} style member listing.
910 In contrast, when writing archives (@option{--create}, @option{--append},
911 @option{--update}), @command{tar} does not print file names by
912 default. So, a single @option{--verbose} option shows the file names
913 being added to the archive, while two @option{--verbose} options
914 enable the full listing.
916 For example, to create an archive in verbose mode:
919 $ @kbd{tar -cvf afiles.tar apple angst aspic}
926 Creating the same archive with the verbosity level 2 could give:
929 $ @kbd{tar -cvvf afiles.tar apple angst aspic}
930 -rw-r--r-- gray/staff 62373 2006-06-09 12:06 apple
931 -rw-r--r-- gray/staff 11481 2006-06-09 12:06 angst
932 -rw-r--r-- gray/staff 23152 2006-06-09 12:06 aspic
936 This works equally well using short or long forms of options. Using
937 long forms, you would simply write out the mnemonic form of the option
941 $ @kbd{tar --create --verbose --verbose @dots{}}
945 Note that you must double the hyphens properly each time.
947 Later in the tutorial, we will give examples using @w{@option{--verbose
950 @anchor{verbose member listing}
951 The full output consists of six fields:
954 @item File type and permissions in symbolic form.
955 These are displayed in the same format as the first column of
956 @command{ls -l} output (@pxref{What information is listed,
957 format=verbose, Verbose listing, fileutils, GNU file utilities}).
959 @item Owner name and group separated by a slash character.
960 If these data are not available (for example, when listing a @samp{v7} format
961 archive), numeric @acronym{ID} values are printed instead.
963 @item Size of the file, in bytes.
965 @item File modification date in ISO 8601 format.
967 @item File modification time.
970 If the name contains any special characters (white space, newlines,
971 etc.) these are displayed in an unambiguous form using so called
972 @dfn{quoting style}. For the detailed discussion of available styles
973 and on how to use them, see @ref{quoting styles}.
975 Depending on the file type, the name can be followed by some
976 additional information, described in the following table:
979 @item -> @var{link-name}
980 The file or archive member is a @dfn{symbolic link} and
981 @var{link-name} is the name of file it links to.
983 @item link to @var{link-name}
984 The file or archive member is a @dfn{hard link} and @var{link-name} is
985 the name of file it links to.
988 The archive member is an old GNU format long link. You will normally
992 The archive member is an old GNU format long name. You will normally
995 @item --Volume Header--
996 The archive member is a GNU @dfn{volume header} (@pxref{Tape Files}).
998 @item --Continued at byte @var{n}--
999 Encountered only at the beginning of a multi-volume archive
1000 (@pxref{Using Multiple Tapes}). This archive member is a continuation
1001 from the previous volume. The number @var{n} gives the offset where
1002 the original file was split.
1004 @item unknown file type @var{c}
1005 An archive member of unknown type. @var{c} is the type character from
1006 the archive header. If you encounter such a message, it means that
1007 either your archive contains proprietary member types @GNUTAR{} is not
1008 able to handle, or the archive is corrupted.
1013 For example, here is an archive listing containing most of the special
1014 suffixes explained above:
1018 V--------- 0/0 1536 2006-06-09 13:07 MyVolume--Volume Header--
1019 -rw-r--r-- gray/staff 456783 2006-06-09 12:06 aspic--Continued at
1021 -rw-r--r-- gray/staff 62373 2006-06-09 12:06 apple
1022 lrwxrwxrwx gray/staff 0 2006-06-09 13:01 angst -> apple
1023 -rw-r--r-- gray/staff 35793 2006-06-09 12:06 blues
1024 hrw-r--r-- gray/staff 0 2006-06-09 12:06 music link to blues
1032 @unnumberedsubsec Getting Help: Using the @option{--help} Option
1038 The @option{--help} option to @command{tar} prints out a very brief list of
1039 all operations and option available for the current version of
1040 @command{tar} available on your system.
1044 @section How to Create Archives
1047 @cindex Creation of the archive
1048 @cindex Archive, creation of
1049 One of the basic operations of @command{tar} is @option{--create} (@option{-c}), which
1050 you use to create a @command{tar} archive. We will explain
1051 @option{--create} first because, in order to learn about the other
1052 operations, you will find it useful to have an archive available to
1055 To make this easier, in this section you will first create a directory
1056 containing three files. Then, we will show you how to create an
1057 @emph{archive} (inside the new directory). Both the directory, and
1058 the archive are specifically for you to practice on. The rest of this
1059 chapter and the next chapter will show many examples using this
1060 directory and the files you will create: some of those files may be
1061 other directories and other archives.
1063 The three files you will archive in this example are called
1064 @file{blues}, @file{folk}, and @file{jazz}. The archive is called
1065 @file{collection.tar}.
1067 This section will proceed slowly, detailing how to use @option{--create}
1068 in @code{verbose} mode, and showing examples using both short and long
1069 forms. In the rest of the tutorial, and in the examples in the next
1070 chapter, we will proceed at a slightly quicker pace. This section
1071 moves more slowly to allow beginning users to understand how
1072 @command{tar} works.
1075 * prepare for examples::
1076 * Creating the archive::
1082 @node prepare for examples
1083 @subsection Preparing a Practice Directory for Examples
1085 To follow along with this and future examples, create a new directory
1086 called @file{practice} containing files called @file{blues}, @file{folk}
1087 and @file{jazz}. The files can contain any information you like:
1088 ideally, they should contain information which relates to their names,
1089 and be of different lengths. Our examples assume that @file{practice}
1090 is a subdirectory of your home directory.
1092 Now @command{cd} to the directory named @file{practice}; @file{practice}
1093 is now your @dfn{working directory}. (@emph{Please note}: Although
1094 the full file name of this directory is
1095 @file{/@var{homedir}/practice}, in our examples we will refer to
1096 this directory as @file{practice}; the @var{homedir} is presumed.
1098 In general, you should check that the files to be archived exist where
1099 you think they do (in the working directory) by running @command{ls}.
1100 Because you just created the directory and the files and have changed to
1101 that directory, you probably don't need to do that this time.
1103 It is very important to make sure there isn't already a file in the
1104 working directory with the archive name you intend to use (in this case,
1105 @samp{collection.tar}), or that you don't care about its contents.
1106 Whenever you use @samp{create}, @command{tar} will erase the current
1107 contents of the file named by @option{--file=@var{archive-name}} (@option{-f @var{archive-name}}) if it exists. @command{tar}
1108 will not tell you if you are about to overwrite an archive unless you
1109 specify an option which does this (@pxref{backup}, for the
1110 information on how to do so). To add files to an existing archive,
1111 you need to use a different option, such as @option{--append} (@option{-r}); see
1112 @ref{append} for information on how to do this.
1114 @node Creating the archive
1115 @subsection Creating the Archive
1117 @xopindex{create, introduced}
1118 To place the files @file{blues}, @file{folk}, and @file{jazz} into an
1119 archive named @file{collection.tar}, use the following command:
1122 $ @kbd{tar --create --file=collection.tar blues folk jazz}
1125 The order of the arguments is not very important, @emph{when using long
1126 option forms}. You could also say:
1129 $ @kbd{tar blues --create folk --file=collection.tar jazz}
1133 However, you can see that this order is harder to understand; this is
1134 why we will list the arguments in the order that makes the commands
1135 easiest to understand (and we encourage you to do the same when you use
1136 @command{tar}, to avoid errors).
1138 Note that the sequence
1139 @option{--file=@-collection.tar} is considered to be @emph{one} argument.
1140 If you substituted any other string of characters for
1141 @kbd{collection.tar}, then that string would become the name of the
1142 archive file you create.
1144 The order of the options becomes more important when you begin to use
1145 short forms. With short forms, if you type commands in the wrong order
1146 (even if you type them correctly in all other ways), you may end up with
1147 results you don't expect. For this reason, it is a good idea to get
1148 into the habit of typing options in the order that makes inherent sense.
1149 @xref{short create}, for more information on this.
1151 In this example, you type the command as shown above: @option{--create}
1152 is the operation which creates the new archive
1153 (@file{collection.tar}), and @option{--file} is the option which lets
1154 you give it the name you chose. The files, @file{blues}, @file{folk},
1155 and @file{jazz}, are now members of the archive, @file{collection.tar}
1156 (they are @dfn{file name arguments} to the @option{--create} operation.
1157 @xref{Choosing}, for the detailed discussion on these.) Now that they are
1158 in the archive, they are called @emph{archive members}, not files.
1159 (@pxref{Definitions,members}).
1161 When you create an archive, you @emph{must} specify which files you
1162 want placed in the archive. If you do not specify any archive
1163 members, @GNUTAR{} will complain.
1165 If you now list the contents of the working directory (@command{ls}), you will
1166 find the archive file listed as well as the files you saw previously:
1169 blues folk jazz collection.tar
1173 Creating the archive @samp{collection.tar} did not destroy the copies of
1174 the files in the directory.
1176 Keep in mind that if you don't indicate an operation, @command{tar} will not
1177 run and will prompt you for one. If you don't name any files, @command{tar}
1178 will complain. You must have write access to the working directory,
1179 or else you will not be able to create an archive in that directory.
1181 @emph{Caution}: Do not attempt to use @option{--create} (@option{-c}) to add files to
1182 an existing archive; it will delete the archive and write a new one.
1183 Use @option{--append} (@option{-r}) instead. @xref{append}.
1185 @node create verbose
1186 @subsection Running @option{--create} with @option{--verbose}
1188 @xopindex{create, using with @option{--verbose}}
1189 @xopindex{verbose, using with @option{--create}}
1190 If you include the @option{--verbose} (@option{-v}) option on the command line,
1191 @command{tar} will list the files it is acting on as it is working. In
1192 verbose mode, the @code{create} example above would appear as:
1195 $ @kbd{tar --create --verbose --file=collection.tar blues folk jazz}
1201 This example is just like the example we showed which did not use
1202 @option{--verbose}, except that @command{tar} generated the remaining lines
1204 (note the different font styles).
1210 In the rest of the examples in this chapter, we will frequently use
1211 @code{verbose} mode so we can show actions or @command{tar} responses that
1212 you would otherwise not see, and which are important for you to
1216 @subsection Short Forms with @samp{create}
1218 As we said before, the @option{--create} (@option{-c}) operation is one of the most
1219 basic uses of @command{tar}, and you will use it countless times.
1220 Eventually, you will probably want to use abbreviated (or ``short'')
1221 forms of options. A full discussion of the three different forms that
1222 options can take appears in @ref{Styles}; for now, here is what the
1223 previous example (including the @option{--verbose} (@option{-v}) option) looks like
1224 using short option forms:
1227 $ @kbd{tar -cvf collection.tar blues folk jazz}
1234 As you can see, the system responds the same no matter whether you use
1235 long or short option forms.
1237 @FIXME{i don't like how this is worded:} One difference between using
1238 short and long option forms is that, although the exact placement of
1239 arguments following options is no more specific when using short forms,
1240 it is easier to become confused and make a mistake when using short
1241 forms. For example, suppose you attempted the above example in the
1245 $ @kbd{tar -cfv collection.tar blues folk jazz}
1249 In this case, @command{tar} will make an archive file called @file{v},
1250 containing the files @file{blues}, @file{folk}, and @file{jazz}, because
1251 the @samp{v} is the closest ``file name'' to the @option{-f} option, and
1252 is thus taken to be the chosen archive file name. @command{tar} will try
1253 to add a file called @file{collection.tar} to the @file{v} archive file;
1254 if the file @file{collection.tar} did not already exist, @command{tar} will
1255 report an error indicating that this file does not exist. If the file
1256 @file{collection.tar} does already exist (e.g., from a previous command
1257 you may have run), then @command{tar} will add this file to the archive.
1258 Because the @option{-v} option did not get registered, @command{tar} will not
1259 run under @samp{verbose} mode, and will not report its progress.
1261 The end result is that you may be quite confused about what happened,
1262 and possibly overwrite a file. To illustrate this further, we will show
1263 you how an example we showed previously would look using short forms.
1268 $ @kbd{tar blues --create folk --file=collection.tar jazz}
1272 is confusing as it is. When shown using short forms, however, it
1273 becomes much more so:
1276 $ @kbd{tar blues -c folk -f collection.tar jazz}
1280 It would be very easy to put the wrong string of characters
1281 immediately following the @option{-f}, but doing that could sacrifice
1284 For this reason, we recommend that you pay very careful attention to
1285 the order of options and placement of file and archive names,
1286 especially when using short option forms. Not having the option name
1287 written out mnemonically can affect how well you remember which option
1288 does what, and therefore where different names have to be placed.
1291 @subsection Archiving Directories
1293 @cindex Archiving Directories
1294 @cindex Directories, Archiving
1295 You can archive a directory by specifying its directory name as a
1296 file name argument to @command{tar}. The files in the directory will be
1297 archived relative to the working directory, and the directory will be
1298 re-created along with its contents when the archive is extracted.
1300 To archive a directory, first move to its superior directory. If you
1301 have followed the previous instructions in this tutorial, you should
1310 This will put you into the directory which contains @file{practice},
1311 i.e., your home directory. Once in the superior directory, you can
1312 specify the subdirectory, @file{practice}, as a file name argument. To
1313 store @file{practice} in the new archive file @file{music.tar}, type:
1316 $ @kbd{tar --create --verbose --file=music.tar practice}
1320 @command{tar} should output:
1327 practice/collection.tar
1330 Note that the archive thus created is not in the subdirectory
1331 @file{practice}, but rather in the current working directory---the
1332 directory from which @command{tar} was invoked. Before trying to archive a
1333 directory from its superior directory, you should make sure you have
1334 write access to the superior directory itself, not only the directory
1335 you are trying archive with @command{tar}. For example, you will probably
1336 not be able to store your home directory in an archive by invoking
1337 @command{tar} from the root directory; @xref{absolute}. (Note
1338 also that @file{collection.tar}, the original archive file, has itself
1339 been archived. @command{tar} will accept any file as a file to be
1340 archived, regardless of its content. When @file{music.tar} is
1341 extracted, the archive file @file{collection.tar} will be re-written
1342 into the file system).
1344 If you give @command{tar} a command such as
1347 $ @kbd{tar --create --file=foo.tar .}
1351 @command{tar} will report @samp{tar: ./foo.tar is the archive; not
1352 dumped}. This happens because @command{tar} creates the archive
1353 @file{foo.tar} in the current directory before putting any files into
1354 it. Then, when @command{tar} attempts to add all the files in the
1355 directory @file{.} to the archive, it notices that the file
1356 @file{./foo.tar} is the same as the archive @file{foo.tar}, and skips
1357 it. (It makes no sense to put an archive into itself.) @GNUTAR{}
1358 will continue in this case, and create the archive
1359 normally, except for the exclusion of that one file. (@emph{Please
1360 note:} Other implementations of @command{tar} may not be so clever;
1361 they will enter an infinite loop when this happens, so you should not
1362 depend on this behavior unless you are certain you are running
1363 @GNUTAR{}. In general, it is wise to always place the archive outside
1364 of the directory being dumped.
1367 @section How to List Archives
1370 Frequently, you will find yourself wanting to determine exactly what a
1371 particular archive contains. You can use the @option{--list}
1372 (@option{-t}) operation to get the member names as they currently
1373 appear in the archive, as well as various attributes of the files at
1374 the time they were archived. For example, you can examine the archive
1375 @file{collection.tar} that you created in the last section with the
1379 $ @kbd{tar --list --file=collection.tar}
1383 The output of @command{tar} would then be:
1392 The archive @file{bfiles.tar} would list as follows:
1401 Be sure to use a @option{--file=@var{archive-name}} (@option{-f
1402 @var{archive-name}}) option just as with @option{--create}
1403 (@option{-c}) to specify the name of the archive.
1405 @xopindex{list, using with @option{--verbose}}
1406 @xopindex{verbose, using with @option{--list}}
1407 If you use the @option{--verbose} (@option{-v}) option with
1408 @option{--list}, then @command{tar} will print out a listing
1409 reminiscent of @w{@samp{ls -l}}, showing owner, file size, and so
1410 forth. This output is described in detail in @ref{verbose member listing}.
1412 If you had used @option{--verbose} (@option{-v}) mode, the example
1413 above would look like:
1416 $ @kbd{tar --list --verbose --file=collection.tar folk}
1417 -rw-r--r-- myself user 62 1990-05-23 10:55 folk
1420 @cindex listing member and file names
1421 @anchor{listing member and file names}
1422 It is important to notice that the output of @kbd{tar --list
1423 --verbose} does not necessarily match that produced by @kbd{tar
1424 --create --verbose} while creating the archive. It is because
1425 @GNUTAR{}, unless told explicitly not to do so, removes some directory
1426 prefixes from file names before storing them in the archive
1427 (@xref{absolute}, for more information). In other
1428 words, in verbose mode @GNUTAR{} shows @dfn{file names} when creating
1429 an archive and @dfn{member names} when listing it. Consider this
1434 $ @kbd{tar cfv archive /etc/mail}
1435 tar: Removing leading `/' from member names
1437 /etc/mail/sendmail.cf
1439 $ @kbd{tar tf archive}
1441 etc/mail/sendmail.cf
1446 @opindex show-stored-names
1447 This default behavior can sometimes be inconvenient. You can force
1448 @GNUTAR{} show member names when creating archive by supplying
1449 @option{--show-stored-names} option.
1452 @item --show-stored-names
1453 Print member (as opposed to @emph{file}) names when creating the archive.
1456 @cindex File name arguments, using @option{--list} with
1457 @xopindex{list, using with file name arguments}
1458 You can specify one or more individual member names as arguments when
1459 using @samp{list}. In this case, @command{tar} will only list the
1460 names of members you identify. For example, @w{@kbd{tar --list
1461 --file=afiles.tar apple}} would only print @file{apple}.
1463 Because @command{tar} preserves file names, these must be specified as
1464 they appear in the archive (i.e., relative to the directory from which
1465 the archive was created). Therefore, it is essential when specifying
1466 member names to @command{tar} that you give the exact member names.
1467 For example, @w{@kbd{tar --list --file=bfiles.tar birds}} would produce an
1468 error message something like @samp{tar: birds: Not found in archive},
1469 because there is no member named @file{birds}, only one named
1470 @file{./birds}. While the names @file{birds} and @file{./birds} name
1471 the same file, @emph{member} names by default are compared verbatim.
1473 However, @w{@kbd{tar --list --file=bfiles.tar baboon}} would respond
1474 with @file{baboon}, because this exact member name is in the archive file
1475 @file{bfiles.tar}. If you are not sure of the exact file name,
1476 use @dfn{globbing patterns}, for example:
1479 $ @kbd{tar --list --file=bfiles.tar --wildcards '*b*'}
1483 will list all members whose name contains @samp{b}. @xref{wildcards},
1484 for a detailed discussion of globbing patterns and related
1485 @command{tar} command line options.
1492 @unnumberedsubsec Listing the Contents of a Stored Directory
1494 To get information about the contents of an archived directory,
1495 use the directory name as a file name argument in conjunction with
1496 @option{--list} (@option{-t}). To find out file attributes, include the
1497 @option{--verbose} (@option{-v}) option.
1499 For example, to find out about files in the directory @file{practice}, in
1500 the archive file @file{music.tar}, type:
1503 $ @kbd{tar --list --verbose --file=music.tar practice}
1506 @command{tar} responds:
1509 drwxrwxrwx myself user 0 1990-05-31 21:49 practice/
1510 -rw-r--r-- myself user 42 1990-05-21 13:29 practice/blues
1511 -rw-r--r-- myself user 62 1990-05-23 10:55 practice/folk
1512 -rw-r--r-- myself user 40 1990-05-21 13:30 practice/jazz
1513 -rw-r--r-- myself user 10240 1990-05-31 21:49 practice/collection.tar
1516 When you use a directory name as a file name argument, @command{tar} acts on
1517 all the files (including sub-directories) in that directory.
1520 @section How to Extract Members from an Archive
1523 @cindex Retrieving files from an archive
1524 @cindex Resurrecting files from an archive
1527 Creating an archive is only half the job---there is no point in storing
1528 files in an archive if you can't retrieve them. The act of retrieving
1529 members from an archive so they can be used and manipulated as
1530 unarchived files again is called @dfn{extraction}. To extract files
1531 from an archive, use the @option{--extract} (@option{--get} or
1532 @option{-x}) operation. As with @option{--create}, specify the name
1533 of the archive with @option{--file} (@option{-f}) option. Extracting
1534 an archive does not modify the archive in any way; you can extract it
1535 multiple times if you want or need to.
1537 Using @option{--extract}, you can extract an entire archive, or specific
1538 files. The files can be directories containing other files, or not. As
1539 with @option{--create} (@option{-c}) and @option{--list} (@option{-t}), you may use the short or the
1540 long form of the operation without affecting the performance.
1543 * extracting archives::
1544 * extracting files::
1546 * extracting untrusted archives::
1547 * failing commands::
1550 @node extracting archives
1551 @subsection Extracting an Entire Archive
1553 To extract an entire archive, specify the archive file name only, with
1554 no individual file names as arguments. For example,
1557 $ @kbd{tar -xvf collection.tar}
1564 -rw-r--r-- me user 28 1996-10-18 16:31 jazz
1565 -rw-r--r-- me user 21 1996-09-23 16:44 blues
1566 -rw-r--r-- me user 20 1996-09-23 16:44 folk
1569 @node extracting files
1570 @subsection Extracting Specific Files
1572 To extract specific archive members, give their exact member names as
1573 arguments, as printed by @option{--list} (@option{-t}). If you had
1574 mistakenly deleted one of the files you had placed in the archive
1575 @file{collection.tar} earlier (say, @file{blues}), you can extract it
1576 from the archive without changing the archive's structure. Its
1577 contents will be identical to the original file @file{blues} that you
1580 First, make sure you are in the @file{practice} directory, and list the
1581 files in the directory. Now, delete the file, @samp{blues}, and list
1582 the files in the directory again.
1584 You can now extract the member @file{blues} from the archive file
1585 @file{collection.tar} like this:
1588 $ @kbd{tar --extract --file=collection.tar blues}
1592 If you list the files in the directory again, you will see that the file
1593 @file{blues} has been restored, with its original permissions, data
1594 modification times, and owner.@footnote{This is only accidentally
1595 true, but not in general. Whereas modification times are always
1596 restored, in most cases, one has to be root for restoring the owner,
1597 and use a special option for restoring permissions. Here, it just
1598 happens that the restoring user is also the owner of the archived
1599 members, and that the current @code{umask} is compatible with original
1600 permissions.} (These parameters will be identical to those which
1601 the file had when you originally placed it in the archive; any changes
1602 you may have made before deleting the file from the file system,
1603 however, will @emph{not} have been made to the archive member.) The
1604 archive file, @samp{collection.tar}, is the same as it was before you
1605 extracted @samp{blues}. You can confirm this by running @command{tar} with
1606 @option{--list} (@option{-t}).
1608 Remember that as with other operations, specifying the exact member
1609 name is important. @w{@kbd{tar --extract --file=bfiles.tar birds}}
1610 will fail, because there is no member named @file{birds}. To extract
1611 the member named @file{./birds}, you must specify @w{@kbd{tar
1612 --extract --file=bfiles.tar ./birds}}. If you don't remember the
1613 exact member names, use @option{--list} (@option{-t}) option
1614 (@pxref{list}). You can also extract those members that match a
1615 specific @dfn{globbing pattern}. For example, to extract from
1616 @file{bfiles.tar} all files that begin with @samp{b}, no matter their
1617 directory prefix, you could type:
1620 $ @kbd{tar -x -f bfiles.tar --wildcards --no-anchored 'b*'}
1624 Here, @option{--wildcards} instructs @command{tar} to treat
1625 command line arguments as globbing patterns and @option{--no-anchored}
1626 informs it that the patterns apply to member names after any @samp{/}
1627 delimiter. The use of globbing patterns is discussed in detail in
1630 You can extract a file to standard output by combining the above options
1631 with the @option{--to-stdout} (@option{-O}) option (@pxref{Writing to Standard
1634 If you give the @option{--verbose} option, then @option{--extract}
1635 will print the names of the archive members as it extracts them.
1638 @subsection Extracting Files that are Directories
1640 Extracting directories which are members of an archive is similar to
1641 extracting other files. The main difference to be aware of is that if
1642 the extracted directory has the same name as any directory already in
1643 the working directory, then files in the extracted directory will be
1644 placed into the directory of the same name. Likewise, if there are
1645 files in the pre-existing directory with the same names as the members
1646 which you extract, the files from the extracted archive will replace
1647 the files already in the working directory (and possible
1648 subdirectories). This will happen regardless of whether or not the
1649 files in the working directory were more recent than those extracted
1650 (there exist, however, special options that alter this behavior
1653 However, if a file was stored with a directory name as part of its file
1654 name, and that directory does not exist under the working directory when
1655 the file is extracted, @command{tar} will create the directory.
1657 We can demonstrate how to use @option{--extract} to extract a directory
1658 file with an example. Change to the @file{practice} directory if you
1659 weren't there, and remove the files @file{folk} and @file{jazz}. Then,
1660 go back to the parent directory and extract the archive
1661 @file{music.tar}. You may either extract the entire archive, or you may
1662 extract only the files you just deleted. To extract the entire archive,
1663 don't give any file names as arguments after the archive name
1664 @file{music.tar}. To extract only the files you deleted, use the
1668 $ @kbd{tar -xvf music.tar practice/folk practice/jazz}
1674 If you were to specify two @option{--verbose} (@option{-v}) options, @command{tar}
1675 would have displayed more detail about the extracted files, as shown
1676 in the example below:
1679 $ @kbd{tar -xvvf music.tar practice/folk practice/jazz}
1680 -rw-r--r-- me user 28 1996-10-18 16:31 practice/jazz
1681 -rw-r--r-- me user 20 1996-09-23 16:44 practice/folk
1685 Because you created the directory with @file{practice} as part of the
1686 file names of each of the files by archiving the @file{practice}
1687 directory as @file{practice}, you must give @file{practice} as part
1688 of the file names when you extract those files from the archive.
1690 @node extracting untrusted archives
1691 @subsection Extracting Archives from Untrusted Sources
1693 Extracting files from archives can overwrite files that already exist.
1694 If you receive an archive from an untrusted source, you should make a
1695 new directory and extract into that directory, so that you don't have
1696 to worry about the extraction overwriting one of your existing files.
1697 For example, if @file{untrusted.tar} came from somewhere else on the
1698 Internet, and you don't necessarily trust its contents, you can
1699 extract it as follows:
1702 $ @kbd{mkdir newdir}
1704 $ @kbd{tar -xvf ../untrusted.tar}
1707 It is also a good practice to examine contents of the archive
1708 before extracting it, using @option{--list} (@option{-t}) option, possibly combined
1709 with @option{--verbose} (@option{-v}).
1711 @node failing commands
1712 @subsection Commands That Will Fail
1714 Here are some sample commands you might try which will not work, and why
1717 If you try to use this command,
1720 $ @kbd{tar -xvf music.tar folk jazz}
1724 you will get the following response:
1727 tar: folk: Not found in archive
1728 tar: jazz: Not found in archive
1733 This is because these files were not originally @emph{in} the parent
1734 directory @file{..}, where the archive is located; they were in the
1735 @file{practice} directory, and their file names reflect this:
1738 $ @kbd{tar -tvf music.tar}
1744 @FIXME{make sure the above works when going through the examples in
1748 Likewise, if you try to use this command,
1751 $ @kbd{tar -tvf music.tar folk jazz}
1755 you would get a similar response. Members with those names are not in the
1756 archive. You must use the correct member names, or wildcards, in order
1757 to extract the files from the archive.
1759 If you have forgotten the correct names of the files in the archive,
1760 use @w{@kbd{tar --list --verbose}} to list them correctly.
1762 @FIXME{more examples, here? hag thinks it's a good idea.}
1765 @section Going Further Ahead in this Manual
1767 @FIXME{need to write up a node here about the things that are going to
1768 be in the rest of the manual.}
1770 @node tar invocation
1771 @chapter Invoking @GNUTAR{}
1774 This chapter is about how one invokes the @GNUTAR{}
1775 command, from the command synopsis (@pxref{Synopsis}). There are
1776 numerous options, and many styles for writing them. One mandatory
1777 option specifies the operation @command{tar} should perform
1778 (@pxref{Operation Summary}), other options are meant to detail how
1779 this operation should be performed (@pxref{Option Summary}).
1780 Non-option arguments are not always interpreted the same way,
1781 depending on what the operation is.
1783 You will find in this chapter everything about option styles and rules for
1784 writing them (@pxref{Styles}). On the other hand, operations and options
1785 are fully described elsewhere, in other chapters. Here, you will find
1786 only synthetic descriptions for operations and options, together with
1787 pointers to other parts of the @command{tar} manual.
1789 Some options are so special they are fully described right in this
1790 chapter. They have the effect of inhibiting the normal operation of
1791 @command{tar} or else, they globally alter the amount of feedback the user
1792 receives about what is going on. These are the @option{--help} and
1793 @option{--version} (@pxref{help}), @option{--verbose} (@pxref{verbose})
1794 and @option{--interactive} options (@pxref{interactive}).
1798 * using tar options::
1808 @section General Synopsis of @command{tar}
1810 The @GNUTAR{} program is invoked as either one of:
1813 @kbd{tar @var{option}@dots{} [@var{name}]@dots{}}
1814 @kbd{tar @var{letter}@dots{} [@var{argument}]@dots{} [@var{option}]@dots{} [@var{name}]@dots{}}
1817 The second form is for when old options are being used.
1819 You can use @command{tar} to store files in an archive, to extract them from
1820 an archive, and to do other types of archive manipulation. The primary
1821 argument to @command{tar}, which is called the @dfn{operation}, specifies
1822 which action to take. The other arguments to @command{tar} are either
1823 @dfn{options}, which change the way @command{tar} performs an operation,
1824 or file names or archive members, which specify the files or members
1825 @command{tar} is to act on.
1827 You can actually type in arguments in any order, even if in this manual
1828 the options always precede the other arguments, to make examples easier
1829 to understand. Further, the option stating the main operation mode
1830 (the @command{tar} main command) is usually given first.
1832 Each @var{name} in the synopsis above is interpreted as an archive member
1833 name when the main command is one of @option{--compare}
1834 (@option{--diff}, @option{-d}), @option{--delete}, @option{--extract}
1835 (@option{--get}, @option{-x}), @option{--list} (@option{-t}) or
1836 @option{--update} (@option{-u}). When naming archive members, you
1837 must give the exact name of the member in the archive, as it is
1838 printed by @option{--list}. For @option{--append} (@option{-r}) and
1839 @option{--create} (@option{-c}), these @var{name} arguments specify
1840 the names of either files or directory hierarchies to place in the archive.
1841 These files or hierarchies should already exist in the file system,
1842 prior to the execution of the @command{tar} command.
1844 @command{tar} interprets relative file names as being relative to the
1845 working directory. @command{tar} will make all file names relative
1846 (by removing leading slashes when archiving or restoring files),
1847 unless you specify otherwise (using the @option{--absolute-names}
1848 option). @xref{absolute}, for more information about
1849 @option{--absolute-names}.
1851 If you give the name of a directory as either a file name or a member
1852 name, then @command{tar} acts recursively on all the files and directories
1853 beneath that directory. For example, the name @file{/} identifies all
1854 the files in the file system to @command{tar}.
1856 The distinction between file names and archive member names is especially
1857 important when shell globbing is used, and sometimes a source of confusion
1858 for newcomers. @xref{wildcards}, for more information about globbing.
1859 The problem is that shells may only glob using existing files in the
1860 file system. Only @command{tar} itself may glob on archive members, so when
1861 needed, you must ensure that wildcard characters reach @command{tar} without
1862 being interpreted by the shell first. Using a backslash before @samp{*}
1863 or @samp{?}, or putting the whole argument between quotes, is usually
1864 sufficient for this.
1866 Even if @var{name}s are often specified on the command line, they
1867 can also be read from a text file in the file system, using the
1868 @option{--files-from=@var{file-of-names}} (@option{-T @var{file-of-names}}) option.
1870 If you don't use any file name arguments, @option{--append} (@option{-r}),
1871 @option{--delete} and @option{--concatenate} (@option{--catenate},
1872 @option{-A}) will do nothing, while @option{--create} (@option{-c})
1873 will usually yield a diagnostic and inhibit @command{tar} execution.
1874 The other operations of @command{tar} (@option{--list},
1875 @option{--extract}, @option{--compare}, and @option{--update})
1876 will act on the entire contents of the archive.
1879 @cindex return status
1880 Besides successful exits, @GNUTAR{} may fail for
1881 many reasons. Some reasons correspond to bad usage, that is, when the
1882 @command{tar} command is improperly written. Errors may be
1883 encountered later, while encountering an error processing the archive
1884 or the files. Some errors are recoverable, in which case the failure
1885 is delayed until @command{tar} has completed all its work. Some
1886 errors are such that it would not meaningful, or at least risky, to
1887 continue processing: @command{tar} then aborts processing immediately.
1888 All abnormal exits, whether immediate or delayed, should always be
1889 clearly diagnosed on @code{stderr}, after a line stating the nature of
1892 Possible exit codes of @GNUTAR{} are summarized in the following
1897 @samp{Successful termination}.
1900 @samp{Some files differ}. If tar was invoked with @option{--compare}
1901 (@option{--diff}, @option{-d}) command line option, this means that
1902 some files in the archive differ from their disk counterparts
1903 (@pxref{compare}). If tar was given @option{--create},
1904 @option{--append} or @option{--update} option, this exit code means
1905 that some files were changed while being archived and so the resulting
1906 archive does not contain the exact copy of the file set.
1909 @samp{Fatal error}. This means that some fatal, unrecoverable error
1913 If @command{tar} has invoked a subprocess and that subprocess exited with a
1914 nonzero exit code, @command{tar} exits with that code as well.
1915 This can happen, for example, if @command{tar} was given some
1916 compression option (@pxref{gzip}) and the external compressor program
1917 failed. Another example is @command{rmt} failure during backup to the
1918 remote device (@pxref{Remote Tape Server}).
1920 @node using tar options
1921 @section Using @command{tar} Options
1923 @GNUTAR{} has a total of eight operating modes which
1924 allow you to perform a variety of tasks. You are required to choose
1925 one operating mode each time you employ the @command{tar} program by
1926 specifying one, and only one operation as an argument to the
1927 @command{tar} command (two lists of four operations each may be found
1928 at @ref{frequent operations} and @ref{Operations}). Depending on
1929 circumstances, you may also wish to customize how the chosen operating
1930 mode behaves. For example, you may wish to change the way the output
1931 looks, or the format of the files that you wish to archive may require
1932 you to do something special in order to make the archive look right.
1934 You can customize and control @command{tar}'s performance by running
1935 @command{tar} with one or more options (such as @option{--verbose}
1936 (@option{-v}), which we used in the tutorial). As we said in the
1937 tutorial, @dfn{options} are arguments to @command{tar} which are (as
1938 their name suggests) optional. Depending on the operating mode, you
1939 may specify one or more options. Different options will have different
1940 effects, but in general they all change details of the operation, such
1941 as archive format, archive name, or level of user interaction. Some
1942 options make sense with all operating modes, while others are
1943 meaningful only with particular modes. You will likely use some
1944 options frequently, while you will only use others infrequently, or
1945 not at all. (A full list of options is available in @pxref{All Options}.)
1947 @vrindex TAR_OPTIONS, environment variable
1948 @anchor{TAR_OPTIONS}
1949 The @env{TAR_OPTIONS} environment variable specifies default options to
1950 be placed in front of any explicit options. For example, if
1951 @code{TAR_OPTIONS} is @samp{-v --unlink-first}, @command{tar} behaves as
1952 if the two options @option{-v} and @option{--unlink-first} had been
1953 specified before any explicit options. Option specifications are
1954 separated by whitespace. A backslash escapes the next character, so it
1955 can be used to specify an option containing whitespace or a backslash.
1957 Note that @command{tar} options are case sensitive. For example, the
1958 options @option{-T} and @option{-t} are different; the first requires an
1959 argument for stating the name of a file providing a list of @var{name}s,
1960 while the second does not require an argument and is another way to
1961 write @option{--list} (@option{-t}).
1963 In addition to the eight operations, there are many options to
1964 @command{tar}, and three different styles for writing both: long (mnemonic)
1965 form, short form, and old style. These styles are discussed below.
1966 Both the options and the operations can be written in any of these three
1969 @FIXME{menu at end of this node. need to think of an actual outline
1970 for this chapter; probably do that after stuff from chapter 4 is
1974 @section The Three Option Styles
1976 There are three styles for writing operations and options to the command
1977 line invoking @command{tar}. The different styles were developed at
1978 different times during the history of @command{tar}. These styles will be
1979 presented below, from the most recent to the oldest.
1981 Some options must take an argument. (For example, @option{--file}
1982 (@option{-f})) takes the name of an archive file as an argument. If
1983 you do not supply an archive file name, @command{tar} will use a
1984 default, but this can be confusing; thus, we recommend that you always
1985 supply a specific archive file name.) Where you @emph{place} the
1986 arguments generally depends on which style of options you choose. We
1987 will detail specific information relevant to each option style in the
1988 sections on the different option styles, below. The differences are
1989 subtle, yet can often be very important; incorrect option placement
1990 can cause you to overwrite a number of important files. We urge you
1991 to note these differences, and only use the option style(s) which
1992 makes the most sense to you until you feel comfortable with the others.
1994 Some options @emph{may} take an argument. Such options may have at
1995 most long and short forms, they do not have old style equivalent. The
1996 rules for specifying an argument for such options are stricter than
1997 those for specifying mandatory arguments. Please, pay special
2001 * Long Options:: Long Option Style
2002 * Short Options:: Short Option Style
2003 * Old Options:: Old Option Style
2004 * Mixing:: Mixing Option Styles
2008 @subsection Long Option Style
2010 Each option has at least one @dfn{long} (or @dfn{mnemonic}) name starting with two
2011 dashes in a row, e.g., @option{--list}. The long names are more clear than
2012 their corresponding short or old names. It sometimes happens that a
2013 single long option has many different names which are
2014 synonymous, such as @option{--compare} and @option{--diff}. In addition,
2015 long option names can be given unique abbreviations. For example,
2016 @option{--cre} can be used in place of @option{--create} because there is no
2017 other long option which begins with @samp{cre}. (One way to find
2018 this out is by trying it and seeing what happens; if a particular
2019 abbreviation could represent more than one option, @command{tar} will tell
2020 you that that abbreviation is ambiguous and you'll know that that
2021 abbreviation won't work. You may also choose to run @samp{tar --help}
2022 to see a list of options. Be aware that if you run @command{tar} with a
2023 unique abbreviation for the long name of an option you didn't want to
2024 use, you are stuck; @command{tar} will perform the command as ordered.)
2026 Long options are meant to be obvious and easy to remember, and their
2027 meanings are generally easier to discern than those of their
2028 corresponding short options (see below). For example:
2031 $ @kbd{tar --create --verbose --blocking-factor=20 --file=/dev/rmt0}
2035 gives a fairly good set of hints about what the command does, even
2036 for those not fully acquainted with @command{tar}.
2038 Long options which require arguments take those arguments
2039 immediately following the option name. There are two ways of
2040 specifying a mandatory argument. It can be separated from the
2041 option name either by an equal sign, or by any amount of
2042 white space characters. For example, the @option{--file} option (which
2043 tells the name of the @command{tar} archive) is given a file such as
2044 @file{archive.tar} as argument by using any of the following notations:
2045 @option{--file=archive.tar} or @option{--file archive.tar}.
2047 In contrast, optional arguments must always be introduced using
2048 an equal sign. For example, the @option{--backup} option takes
2049 an optional argument specifying backup type. It must be used
2050 as @option{--backup=@var{backup-type}}.
2053 @subsection Short Option Style
2055 Most options also have a @dfn{short option} name. Short options start with
2056 a single dash, and are followed by a single character, e.g., @option{-t}
2057 (which is equivalent to @option{--list}). The forms are absolutely
2058 identical in function; they are interchangeable.
2060 The short option names are faster to type than long option names.
2062 Short options which require arguments take their arguments immediately
2063 following the option, usually separated by white space. It is also
2064 possible to stick the argument right after the short option name, using
2065 no intervening space. For example, you might write @w{@option{-f
2066 archive.tar}} or @option{-farchive.tar} instead of using
2067 @option{--file=archive.tar}. Both @option{--file=@var{archive-name}} and
2068 @w{@option{-f @var{archive-name}}} denote the option which indicates a
2069 specific archive, here named @file{archive.tar}.
2071 Short options which take optional arguments take their arguments
2072 immediately following the option letter, @emph{without any intervening
2073 white space characters}.
2075 Short options' letters may be clumped together, but you are not
2076 required to do this (as compared to old options; see below). When
2077 short options are clumped as a set, use one (single) dash for them
2078 all, e.g., @w{@samp{@command{tar} -cvf}}. Only the last option in
2079 such a set is allowed to have an argument@footnote{Clustering many
2080 options, the last of which has an argument, is a rather opaque way to
2081 write options. Some wonder if @acronym{GNU} @code{getopt} should not
2082 even be made helpful enough for considering such usages as invalid.}.
2084 When the options are separated, the argument for each option which requires
2085 an argument directly follows that option, as is usual for Unix programs.
2089 $ @kbd{tar -c -v -b 20 -f /dev/rmt0}
2092 If you reorder short options' locations, be sure to move any arguments
2093 that belong to them. If you do not move the arguments properly, you may
2094 end up overwriting files.
2097 @subsection Old Option Style
2100 Like short options, @dfn{old options} are single letters. However, old options
2101 must be written together as a single clumped set, without spaces separating
2102 them or dashes preceding them@footnote{Beware that if you precede options
2103 with a dash, you are announcing the short option style instead of the
2104 old option style; short options are decoded differently.}. This set
2105 of letters must be the first to appear on the command line, after the
2106 @command{tar} program name and some white space; old options cannot appear
2107 anywhere else. The letter of an old option is exactly the same letter as
2108 the corresponding short option. For example, the old option @samp{t} is
2109 the same as the short option @option{-t}, and consequently, the same as the
2110 long option @option{--list}. So for example, the command @w{@samp{tar
2111 cv}} specifies the option @option{-v} in addition to the operation @option{-c}.
2113 When options that need arguments are given together with the command,
2114 all the associated arguments follow, in the same order as the options.
2115 Thus, the example given previously could also be written in the old
2119 $ @kbd{tar cvbf 20 /dev/rmt0}
2123 Here, @samp{20} is the argument of @option{-b} and @samp{/dev/rmt0} is
2124 the argument of @option{-f}.
2126 On the other hand, this old style syntax makes it difficult to match
2127 option letters with their corresponding arguments, and is often
2128 confusing. In the command @w{@samp{tar cvbf 20 /dev/rmt0}}, for example,
2129 @samp{20} is the argument for @option{-b}, @samp{/dev/rmt0} is the
2130 argument for @option{-f}, and @option{-v} does not have a corresponding
2131 argument. Even using short options like in @w{@samp{tar -c -v -b 20 -f
2132 /dev/rmt0}} is clearer, putting all arguments next to the option they
2135 If you want to reorder the letters in the old option argument, be
2136 sure to reorder any corresponding argument appropriately.
2138 This old way of writing @command{tar} options can surprise even experienced
2139 users. For example, the two commands:
2142 @kbd{tar cfz archive.tar.gz file}
2143 @kbd{tar -cfz archive.tar.gz file}
2147 are quite different. The first example uses @file{archive.tar.gz} as
2148 the value for option @samp{f} and recognizes the option @samp{z}. The
2149 second example, however, uses @file{z} as the value for option
2150 @samp{f} --- probably not what was intended.
2152 Old options are kept for compatibility with old versions of @command{tar}.
2154 This second example could be corrected in many ways, among which the
2155 following are equivalent:
2158 @kbd{tar -czf archive.tar.gz file}
2159 @kbd{tar -cf archive.tar.gz -z file}
2160 @kbd{tar cf archive.tar.gz -z file}
2163 @cindex option syntax, traditional
2164 As far as we know, all @command{tar} programs, @acronym{GNU} and
2165 non-@acronym{GNU}, support old options. @GNUTAR{}
2166 supports them not only for historical reasons, but also because many
2167 people are used to them. For compatibility with Unix @command{tar},
2168 the first argument is always treated as containing command and option
2169 letters even if it doesn't start with @samp{-}. Thus, @samp{tar c} is
2170 equivalent to @w{@samp{tar -c}:} both of them specify the
2171 @option{--create} (@option{-c}) command to create an archive.
2174 @subsection Mixing Option Styles
2176 All three styles may be intermixed in a single @command{tar} command,
2177 so long as the rules for each style are fully
2178 respected@footnote{Before @GNUTAR{} version 1.11.6,
2179 a bug prevented intermixing old style options with long options in
2180 some cases.}. Old style options and either of the modern styles of
2181 options may be mixed within a single @command{tar} command. However,
2182 old style options must be introduced as the first arguments only,
2183 following the rule for old options (old options must appear directly
2184 after the @command{tar} command and some white space). Modern options
2185 may be given only after all arguments to the old options have been
2186 collected. If this rule is not respected, a modern option might be
2187 falsely interpreted as the value of the argument to one of the old
2190 For example, all the following commands are wholly equivalent, and
2191 illustrate the many combinations and orderings of option styles.
2194 @kbd{tar --create --file=archive.tar}
2195 @kbd{tar --create -f archive.tar}
2196 @kbd{tar --create -farchive.tar}
2197 @kbd{tar --file=archive.tar --create}
2198 @kbd{tar --file=archive.tar -c}
2199 @kbd{tar -c --file=archive.tar}
2200 @kbd{tar -c -f archive.tar}
2201 @kbd{tar -c -farchive.tar}
2202 @kbd{tar -cf archive.tar}
2203 @kbd{tar -cfarchive.tar}
2204 @kbd{tar -f archive.tar --create}
2205 @kbd{tar -f archive.tar -c}
2206 @kbd{tar -farchive.tar --create}
2207 @kbd{tar -farchive.tar -c}
2208 @kbd{tar c --file=archive.tar}
2209 @kbd{tar c -f archive.tar}
2210 @kbd{tar c -farchive.tar}
2211 @kbd{tar cf archive.tar}
2212 @kbd{tar f archive.tar --create}
2213 @kbd{tar f archive.tar -c}
2214 @kbd{tar fc archive.tar}
2217 On the other hand, the following commands are @emph{not} equivalent to
2221 @kbd{tar -f -c archive.tar}
2222 @kbd{tar -fc archive.tar}
2223 @kbd{tar -fcarchive.tar}
2224 @kbd{tar -farchive.tarc}
2225 @kbd{tar cfarchive.tar}
2229 These last examples mean something completely different from what the
2230 user intended (judging based on the example in the previous set which
2231 uses long options, whose intent is therefore very clear). The first
2232 four specify that the @command{tar} archive would be a file named
2233 @option{-c}, @samp{c}, @samp{carchive.tar} or @samp{archive.tarc},
2234 respectively. The first two examples also specify a single non-option,
2235 @var{name} argument having the value @samp{archive.tar}. The last
2236 example contains only old style option letters (repeating option
2237 @samp{c} twice), not all of which are meaningful (eg., @samp{.},
2238 @samp{h}, or @samp{i}), with no argument value. @FIXME{not sure i liked
2239 the first sentence of this paragraph..}
2242 @section All @command{tar} Options
2244 The coming manual sections contain an alphabetical listing of all
2245 @command{tar} operations and options, with brief descriptions and cross
2246 references to more in-depth explanations in the body of the manual.
2247 They also contain an alphabetically arranged table of the short option
2248 forms with their corresponding long option. You can use this table as
2249 a reference for deciphering @command{tar} commands in scripts.
2252 * Operation Summary::
2254 * Short Option Summary::
2257 @node Operation Summary
2258 @subsection Operations
2266 Appends files to the end of the archive. @xref{append}.
2268 @opsummary{catenate}
2272 Same as @option{--concatenate}. @xref{concatenate}.
2278 Compares archive members with their counterparts in the file
2279 system, and reports differences in file size, mode, owner,
2280 modification date and contents. @xref{compare}.
2282 @opsummary{concatenate}
2286 Appends other @command{tar} archives to the end of the archive.
2293 Creates a new @command{tar} archive. @xref{create}.
2298 Deletes members from the archive. Don't try this on a archive on a
2299 tape! @xref{delete}.
2305 Same @option{--compare}. @xref{compare}.
2311 Extracts members from the archive into the file system. @xref{extract}.
2317 Same as @option{--extract}. @xref{extract}.
2323 Lists the members in an archive. @xref{list}.
2329 Adds files to the end of the archive, but only if they are newer than
2330 their counterparts already in the archive, or if they do not already
2331 exist in the archive. @xref{update}.
2335 @node Option Summary
2336 @subsection @command{tar} Options
2340 @opsummary{absolute-names}
2341 @item --absolute-names
2344 Normally when creating an archive, @command{tar} strips an initial
2345 @samp{/} from member names. This option disables that behavior.
2348 @opsummary{after-date}
2351 (See @option{--newer}, @pxref{after})
2353 @opsummary{anchored}
2355 A pattern must match an initial subsequence of the name's components.
2356 @xref{controlling pattern-matching}.
2358 @opsummary{atime-preserve}
2359 @item --atime-preserve
2360 @itemx --atime-preserve=replace
2361 @itemx --atime-preserve=system
2363 Attempt to preserve the access time of files when reading them. This
2364 option currently is effective only on files that you own, unless you
2365 have superuser privileges.
2367 @option{--atime-preserve=replace} remembers the access time of a file
2368 before reading it, and then restores the access time afterwards. This
2369 may cause problems if other programs are reading the file at the same
2370 time, as the times of their accesses will be lost. On most platforms
2371 restoring the access time also requires @command{tar} to restore the
2372 data modification time too, so this option may also cause problems if
2373 other programs are writing the file at the same time. (Tar attempts
2374 to detect this situation, but cannot do so reliably due to race
2375 conditions.) Worse, on most platforms restoring the access time also
2376 updates the status change time, which means that this option is
2377 incompatible with incremental backups.
2379 @option{--atime-preserve=system} avoids changing time stamps on files,
2380 without interfering with time stamp updates
2381 caused by other programs, so it works better with incremental backups.
2382 However, it requires a special @code{O_NOATIME} option from the
2383 underlying operating and file system implementation, and it also requires
2384 that searching directories does not update their access times. As of
2385 this writing (November 2005) this works only with Linux, and only with
2386 Linux kernels 2.6.8 and later. Worse, there is currently no reliable
2387 way to know whether this feature actually works. Sometimes
2388 @command{tar} knows that it does not work, and if you use
2389 @option{--atime-preserve=system} then @command{tar} complains and
2390 exits right away. But other times @command{tar} might think that the
2391 option works when it actually does not.
2393 Currently @option{--atime-preserve} with no operand defaults to
2394 @option{--atime-preserve=replace}, but this may change in the future
2395 as support for @option{--atime-preserve=system} improves.
2397 If your operating system does not support
2398 @option{--atime-preserve=@-system}, you might be able to preserve access
2399 times reliably by by using the @command{mount} command. For example,
2400 you can mount the file system read-only, or access the file system via
2401 a read-only loopback mount, or use the @samp{noatime} mount option
2402 available on some systems. However, mounting typically requires
2403 superuser privileges and can be a pain to manage.
2406 @item --backup=@var{backup-type}
2408 Rather than deleting files from the file system, @command{tar} will
2409 back them up using simple or numbered backups, depending upon
2410 @var{backup-type}. @xref{backup}.
2412 @opsummary{block-number}
2413 @item --block-number
2416 With this option present, @command{tar} prints error messages for read errors
2417 with the block number in the archive file. @xref{block-number}.
2419 @opsummary{blocking-factor}
2420 @item --blocking-factor=@var{blocking}
2421 @itemx -b @var{blocking}
2423 Sets the blocking factor @command{tar} uses to @var{blocking} x 512 bytes per
2424 record. @xref{Blocking Factor}.
2430 This option tells @command{tar} to read or write archives through
2431 @code{bzip2}. @xref{gzip}.
2433 @opsummary{checkpoint}
2434 @item --checkpoint[=@var{number}]
2436 This option directs @command{tar} to print periodic checkpoint
2437 messages as it reads through the archive. It is intended for when you
2438 want a visual indication that @command{tar} is still running, but
2439 don't want to see @option{--verbose} output. For a detailed
2440 description, see @ref{Progress information}.
2442 @opsummary{check-links}
2445 If this option was given, @command{tar} will check the number of links
2446 dumped for each processed file. If this number does not match the
2447 total number of hard links for the file, a warning message will be
2448 output @footnote{Earlier versions of @GNUTAR{} understood @option{-l} as a
2449 synonym for @option{--one-file-system}. The current semantics, which
2450 complies to UNIX98, was introduced with version
2451 1.15.91. @xref{Changes}, for more information.}.
2453 @opsummary{compress}
2454 @opsummary{uncompress}
2459 @command{tar} will use the @command{compress} program when reading or
2460 writing the archive. This allows you to directly act on archives
2461 while saving space. @xref{gzip}.
2463 @opsummary{confirmation}
2464 @item --confirmation
2466 (See @option{--interactive}.) @xref{interactive}.
2468 @opsummary{delay-directory-restore}
2469 @item --delay-directory-restore
2471 Delay setting modification times and permissions of extracted
2472 directories until the end of extraction. @xref{Directory Modification Times and Permissions}.
2474 @opsummary{dereference}
2478 When creating a @command{tar} archive, @command{tar} will archive the
2479 file that a symbolic link points to, rather than archiving the
2480 symlink. @xref{dereference}.
2482 @opsummary{directory}
2483 @item --directory=@var{dir}
2486 When this option is specified, @command{tar} will change its current directory
2487 to @var{dir} before performing any operations. When this option is used
2488 during archive creation, it is order sensitive. @xref{directory}.
2491 @item --exclude=@var{pattern}
2493 When performing operations, @command{tar} will skip files that match
2494 @var{pattern}. @xref{exclude}.
2496 @opsummary{exclude-from}
2497 @item --exclude-from=@var{file}
2498 @itemx -X @var{file}
2500 Similar to @option{--exclude}, except @command{tar} will use the list of
2501 patterns in the file @var{file}. @xref{exclude}.
2503 @opsummary{exclude-caches}
2504 @item --exclude-caches
2506 Exclude from dump any directory containing a valid cache directory
2507 tag file, but still dump the directory node and the tag file itself.
2511 @opsummary{exclude-caches-under}
2512 @item --exclude-caches-under
2514 Exclude from dump any directory containing a valid cache directory
2515 tag file, but still dump the directory node itself.
2519 @opsummary{exclude-caches-all}
2520 @item --exclude-caches-all
2522 Exclude from dump any directory containing a valid cache directory
2523 tag file. @xref{exclude}.
2525 @opsummary{exclude-tag}
2526 @item --exclude-tag=@var{file}
2528 Exclude from dump any directory containing file named @var{file}, but
2529 dump the directory node and @var{file} itself. @xref{exclude}.
2531 @opsummary{exclude-tag-under}
2532 @item --exclude-tag-under=@var{file}
2534 Exclude from dump the contents of any directory containing file
2535 named @var{file}, but dump the directory node itself. @xref{exclude}.
2537 @opsummary{exclude-tag-all}
2538 @item --exclude-tag-all=@var{file}
2540 Exclude from dump any directory containing file named @var{file}.
2543 @opsummary{exclude-vcs}
2546 Exclude from dump directories and files, that are internal for some
2547 widely used version control systems.
2552 @item --file=@var{archive}
2553 @itemx -f @var{archive}
2555 @command{tar} will use the file @var{archive} as the @command{tar} archive it
2556 performs operations on, rather than @command{tar}'s compilation dependent
2557 default. @xref{file tutorial}.
2559 @opsummary{files-from}
2560 @item --files-from=@var{file}
2561 @itemx -T @var{file}
2563 @command{tar} will use the contents of @var{file} as a list of archive members
2564 or files to operate on, in addition to those specified on the
2565 command-line. @xref{files}.
2567 @opsummary{force-local}
2570 Forces @command{tar} to interpret the file name given to @option{--file}
2571 as a local file, even if it looks like a remote tape drive name.
2572 @xref{local and remote archives}.
2575 @item --format=@var{format}
2576 @itemx -H @var{format}
2578 Selects output archive format. @var{Format} may be one of the
2583 Creates an archive that is compatible with Unix V7 @command{tar}.
2586 Creates an archive that is compatible with GNU @command{tar} version
2590 Creates archive in GNU tar 1.13 format. Basically it is the same as
2591 @samp{oldgnu} with the only difference in the way it handles long
2595 Creates a @acronym{POSIX.1-1988} compatible archive.
2598 Creates a @acronym{POSIX.1-2001 archive}.
2602 @xref{Formats}, for a detailed discussion of these formats.
2605 @item --group=@var{group}
2607 Files added to the @command{tar} archive will have a group @acronym{ID} of @var{group},
2608 rather than the group from the source file. @var{group} is first decoded
2609 as a group symbolic name, but if this interpretation fails, it has to be
2610 a decimal numeric group @acronym{ID}. @xref{override}.
2612 Also see the comments for the @option{--owner=@var{user}} option.
2622 This option tells @command{tar} to read or write archives through
2623 @command{gzip}, allowing @command{tar} to directly operate on several
2624 kinds of compressed archives transparently. @xref{gzip}.
2630 @command{tar} will print out a short message summarizing the operations and
2631 options to @command{tar} and exit. @xref{help}.
2633 @opsummary{ignore-case}
2635 Ignore case when matching member or file names with
2636 patterns. @xref{controlling pattern-matching}.
2638 @opsummary{ignore-command-error}
2639 @item --ignore-command-error
2640 Ignore exit codes of subprocesses. @xref{Writing to an External Program}.
2642 @opsummary{ignore-failed-read}
2643 @item --ignore-failed-read
2645 Do not exit unsuccessfully merely because an unreadable file was encountered.
2648 @opsummary{ignore-zeros}
2649 @item --ignore-zeros
2652 With this option, @command{tar} will ignore zeroed blocks in the
2653 archive, which normally signals EOF. @xref{Reading}.
2655 @opsummary{incremental}
2659 Informs @command{tar} that it is working with an old
2660 @acronym{GNU}-format incremental backup archive. It is intended
2661 primarily for backwards compatibility only. @xref{Incremental Dumps},
2662 for a detailed discussion of incremental archives.
2664 @opsummary{index-file}
2665 @item --index-file=@var{file}
2667 Send verbose output to @var{file} instead of to standard output.
2669 @opsummary{info-script}
2670 @opsummary{new-volume-script}
2671 @item --info-script=@var{script-file}
2672 @itemx --new-volume-script=@var{script-file}
2673 @itemx -F @var{script-file}
2675 When @command{tar} is performing multi-tape backups, @var{script-file} is run
2676 at the end of each tape. If @var{script-file} exits with nonzero status,
2677 @command{tar} fails immediately. @xref{info-script}, for a detailed
2678 discussion of @var{script-file}.
2680 @opsummary{interactive}
2682 @itemx --confirmation
2685 Specifies that @command{tar} should ask the user for confirmation before
2686 performing potentially destructive options, such as overwriting files.
2689 @opsummary{keep-newer-files}
2690 @item --keep-newer-files
2692 Do not replace existing files that are newer than their archive copies
2693 when extracting files from an archive.
2695 @opsummary{keep-old-files}
2696 @item --keep-old-files
2699 Do not overwrite existing files when extracting files from an archive.
2700 @xref{Keep Old Files}.
2703 @item --label=@var{name}
2704 @itemx -V @var{name}
2706 When creating an archive, instructs @command{tar} to write @var{name}
2707 as a name record in the archive. When extracting or listing archives,
2708 @command{tar} will only operate on archives that have a label matching
2709 the pattern specified in @var{name}. @xref{Tape Files}.
2711 @opsummary{listed-incremental}
2712 @item --listed-incremental=@var{snapshot-file}
2713 @itemx -g @var{snapshot-file}
2715 During a @option{--create} operation, specifies that the archive that
2716 @command{tar} creates is a new @acronym{GNU}-format incremental
2717 backup, using @var{snapshot-file} to determine which files to backup.
2718 With other operations, informs @command{tar} that the archive is in
2719 incremental format. @xref{Incremental Dumps}.
2722 @item --mode=@var{permissions}
2724 When adding files to an archive, @command{tar} will use
2725 @var{permissions} for the archive members, rather than the permissions
2726 from the files. @var{permissions} can be specified either as an octal
2727 number or as symbolic permissions, like with
2728 @command{chmod}. @xref{override}.
2731 @item --mtime=@var{date}
2733 When adding files to an archive, @command{tar} will use @var{date} as
2734 the modification time of members when creating archives, instead of
2735 their actual modification times. The value of @var{date} can be
2736 either a textual date representation (@pxref{Date input formats}) or a
2737 name of the existing file, starting with @samp{/} or @samp{.}. In the
2738 latter case, the modification time of that file is used. @xref{override}.
2740 @opsummary{multi-volume}
2741 @item --multi-volume
2744 Informs @command{tar} that it should create or otherwise operate on a
2745 multi-volume @command{tar} archive. @xref{Using Multiple Tapes}.
2747 @opsummary{new-volume-script}
2748 @item --new-volume-script
2753 @item --newer=@var{date}
2754 @itemx --after-date=@var{date}
2757 When creating an archive, @command{tar} will only add files that have changed
2758 since @var{date}. If @var{date} begins with @samp{/} or @samp{.}, it
2759 is taken to be the name of a file whose data modification time specifies
2760 the date. @xref{after}.
2762 @opsummary{newer-mtime}
2763 @item --newer-mtime=@var{date}
2765 Like @option{--newer}, but add only files whose
2766 contents have changed (as opposed to just @option{--newer}, which will
2767 also back up files for which any status information has
2768 changed). @xref{after}.
2770 @opsummary{no-anchored}
2772 An exclude pattern can match any subsequence of the name's components.
2773 @xref{controlling pattern-matching}.
2775 @opsummary{no-delay-directory-restore}
2776 @item --no-delay-directory-restore
2778 Modification times and permissions of extracted
2779 directories are set when all files from this directory have been
2780 extracted. This is the default.
2781 @xref{Directory Modification Times and Permissions}.
2783 @opsummary{no-ignore-case}
2784 @item --no-ignore-case
2785 Use case-sensitive matching.
2786 @xref{controlling pattern-matching}.
2788 @opsummary{no-ignore-command-error}
2789 @item --no-ignore-command-error
2790 Print warnings about subprocesses that terminated with a nonzero exit
2791 code. @xref{Writing to an External Program}.
2793 @opsummary{no-overwrite-dir}
2794 @item --no-overwrite-dir
2796 Preserve metadata of existing directories when extracting files
2797 from an archive. @xref{Overwrite Old Files}.
2799 @opsummary{no-quote-chars}
2800 @item --no-quote-chars=@var{string}
2801 Remove characters listed in @var{string} from the list of quoted
2802 characters set by the previous @option{--quote-chars} option
2803 (@pxref{quoting styles}).
2805 @opsummary{no-recursion}
2806 @item --no-recursion
2808 With this option, @command{tar} will not recurse into directories.
2811 @opsummary{no-same-owner}
2812 @item --no-same-owner
2815 When extracting an archive, do not attempt to preserve the owner
2816 specified in the @command{tar} archive. This the default behavior
2819 @opsummary{no-same-permissions}
2820 @item --no-same-permissions
2822 When extracting an archive, subtract the user's umask from files from
2823 the permissions specified in the archive. This is the default behavior
2826 @opsummary{no-unquote}
2828 Treat all input file or member names literally, do not interpret
2829 escape sequences. @xref{input name quoting}.
2831 @opsummary{no-wildcards}
2832 @item --no-wildcards
2833 Do not use wildcards.
2834 @xref{controlling pattern-matching}.
2836 @opsummary{no-wildcards-match-slash}
2837 @item --no-wildcards-match-slash
2838 Wildcards do not match @samp{/}.
2839 @xref{controlling pattern-matching}.
2844 When @command{tar} is using the @option{--files-from} option, this option
2845 instructs @command{tar} to expect file names terminated with @acronym{NUL}, so
2846 @command{tar} can correctly work with file names that contain newlines.
2849 @opsummary{numeric-owner}
2850 @item --numeric-owner
2852 This option will notify @command{tar} that it should use numeric user
2853 and group IDs when creating a @command{tar} file, rather than names.
2857 The function of this option depends on the action @command{tar} is
2858 performing. When extracting files, @option{-o} is a synonym for
2859 @option{--no-same-owner}, i.e., it prevents @command{tar} from
2860 restoring ownership of files being extracted.
2862 When creating an archive, it is a synonym for
2863 @option{--old-archive}. This behavior is for compatibility
2864 with previous versions of @GNUTAR{}, and will be
2865 removed in future releases.
2867 @xref{Changes}, for more information.
2869 @opsummary{occurrence}
2870 @item --occurrence[=@var{number}]
2872 This option can be used in conjunction with one of the subcommands
2873 @option{--delete}, @option{--diff}, @option{--extract} or
2874 @option{--list} when a list of files is given either on the command
2875 line or via @option{-T} option.
2877 This option instructs @command{tar} to process only the @var{number}th
2878 occurrence of each named file. @var{Number} defaults to 1, so
2881 tar -x -f archive.tar --occurrence filename
2885 will extract the first occurrence of the member @file{filename} from @file{archive.tar}
2886 and will terminate without scanning to the end of the archive.
2888 @opsummary{old-archive}
2890 Synonym for @option{--format=v7}.
2892 @opsummary{one-file-system}
2893 @item --one-file-system
2894 Used when creating an archive. Prevents @command{tar} from recursing into
2895 directories that are on different file systems from the current
2896 directory @footnote{Earlier versions of @GNUTAR{} understood @option{-l} as a
2897 synonym for @option{--one-file-system}. This has changed in version
2898 1.15.91. @xref{Changes}, for more information.}.
2900 @opsummary{overwrite}
2903 Overwrite existing files and directory metadata when extracting files
2904 from an archive. @xref{Overwrite Old Files}.
2906 @opsummary{overwrite-dir}
2907 @item --overwrite-dir
2909 Overwrite the metadata of existing directories when extracting files
2910 from an archive. @xref{Overwrite Old Files}.
2913 @item --owner=@var{user}
2915 Specifies that @command{tar} should use @var{user} as the owner of members
2916 when creating archives, instead of the user associated with the source
2917 file. @var{user} is first decoded as a user symbolic name, but if
2918 this interpretation fails, it has to be a decimal numeric user @acronym{ID}.
2921 This option does not affect extraction from archives.
2923 @opsummary{pax-option}
2924 @item --pax-option=@var{keyword-list}
2925 This option is meaningful only with @acronym{POSIX.1-2001} archives
2926 (@pxref{posix}). It modifies the way @command{tar} handles the
2927 extended header keywords. @var{Keyword-list} is a comma-separated
2928 list of keyword options. @xref{PAX keywords}, for a detailed
2931 @opsummary{portability}
2933 @itemx --old-archive
2934 Synonym for @option{--format=v7}.
2938 Same as @option{--format=posix}.
2940 @opsummary{preserve}
2943 Synonymous with specifying both @option{--preserve-permissions} and
2944 @option{--same-order}. @xref{Setting Access Permissions}.
2946 @opsummary{preserve-order}
2947 @item --preserve-order
2949 (See @option{--same-order}; @pxref{Reading}.)
2951 @opsummary{preserve-permissions}
2952 @opsummary{same-permissions}
2953 @item --preserve-permissions
2954 @itemx --same-permissions
2957 When @command{tar} is extracting an archive, it normally subtracts the
2958 users' umask from the permissions specified in the archive and uses
2959 that number as the permissions to create the destination file.
2960 Specifying this option instructs @command{tar} that it should use the
2961 permissions directly from the archive. @xref{Setting Access Permissions}.
2963 @opsummary{quote-chars}
2964 @item --quote-chars=@var{string}
2965 Always quote characters from @var{string}, even if the selected
2966 quoting style would not quote them (@pxref{quoting styles}).
2968 @opsummary{quoting-style}
2969 @item --quoting-style=@var{style}
2970 Set quoting style to use when printing member and file names
2971 (@pxref{quoting styles}). Valid @var{style} values are:
2972 @code{literal}, @code{shell}, @code{shell-always}, @code{c},
2973 @code{escape}, @code{locale}, and @code{clocale}. Default quoting
2974 style is @code{escape}, unless overridden while configuring the
2977 @opsummary{read-full-records}
2978 @item --read-full-records
2981 Specifies that @command{tar} should reblock its input, for reading
2982 from pipes on systems with buggy implementations. @xref{Reading}.
2984 @opsummary{record-size}
2985 @item --record-size=@var{size}
2987 Instructs @command{tar} to use @var{size} bytes per record when accessing the
2988 archive. @xref{Blocking Factor}.
2990 @opsummary{recursion}
2993 With this option, @command{tar} recurses into directories (default).
2996 @opsummary{recursive-unlink}
2997 @item --recursive-unlink
3000 directory hierarchies before extracting directories of the same name
3001 from the archive. @xref{Recursive Unlink}.
3003 @opsummary{remove-files}
3004 @item --remove-files
3006 Directs @command{tar} to remove the source file from the file system after
3007 appending it to an archive. @xref{remove files}.
3009 @opsummary{restrict}
3012 Disable use of some potentially harmful @command{tar} options.
3013 Currently this option disables shell invocation from multi-volume menu
3014 (@pxref{Using Multiple Tapes}).
3016 @opsummary{rmt-command}
3017 @item --rmt-command=@var{cmd}
3019 Notifies @command{tar} that it should use @var{cmd} instead of
3020 the default @file{/usr/libexec/rmt} (@pxref{Remote Tape Server}).
3022 @opsummary{rsh-command}
3023 @item --rsh-command=@var{cmd}
3025 Notifies @command{tar} that is should use @var{cmd} to communicate with remote
3026 devices. @xref{Device}.
3028 @opsummary{same-order}
3030 @itemx --preserve-order
3033 This option is an optimization for @command{tar} when running on machines with
3034 small amounts of memory. It informs @command{tar} that the list of file
3035 arguments has already been sorted to match the order of files in the
3036 archive. @xref{Reading}.
3038 @opsummary{same-owner}
3041 When extracting an archive, @command{tar} will attempt to preserve the owner
3042 specified in the @command{tar} archive with this option present.
3043 This is the default behavior for the superuser; this option has an
3044 effect only for ordinary users. @xref{Attributes}.
3046 @opsummary{same-permissions}
3047 @item --same-permissions
3049 (See @option{--preserve-permissions}; @pxref{Setting Access Permissions}.)
3055 Assume that the archive media supports seeks to arbitrary
3056 locations. Usually @command{tar} determines automatically whether
3057 the archive can be seeked or not. This option is intended for use
3058 in cases when such recognition fails.
3060 @opsummary{show-defaults}
3061 @item --show-defaults
3063 Displays the default options used by @command{tar} and exits
3064 successfully. This option is intended for use in shell scripts.
3065 Here is an example of what you can see using this option:
3068 $ tar --show-defaults
3069 --format=gnu -f- -b20 --quoting-style=escape \
3070 --rmt-command=/usr/libexec/rmt --rsh-command=/usr/bin/rsh
3073 @opsummary{show-omitted-dirs}
3074 @item --show-omitted-dirs
3076 Instructs @command{tar} to mention the directories it is skipping when
3077 operating on a @command{tar} archive. @xref{show-omitted-dirs}.
3079 @opsummary{show-transformed-names}
3080 @opsummary{show-stored-names}
3081 @item --show-transformed-names
3082 @itemx --show-stored-names
3084 Display file or member names after applying any transformations
3085 (@pxref{transform}). In particular, when used in conjunction with one of
3086 the archive creation operations it instructs @command{tar} to list the
3087 member names stored in the archive, as opposed to the actual file
3088 names. @xref{listing member and file names}.
3094 Invokes a @acronym{GNU} extension when adding files to an archive that handles
3095 sparse files efficiently. @xref{sparse}.
3097 @opsummary{sparse-version}
3098 @item --sparse-version=@var{version}
3100 Specifies the @dfn{format version} to use when archiving sparse
3101 files. Implies @option{--sparse}. @xref{sparse}. For the description
3102 of the supported sparse formats, @xref{Sparse Formats}.
3104 @opsummary{starting-file}
3105 @item --starting-file=@var{name}
3106 @itemx -K @var{name}
3108 This option affects extraction only; @command{tar} will skip extracting
3109 files in the archive until it finds one that matches @var{name}.
3112 @opsummary{strip-components}
3113 @item --strip-components=@var{number}
3114 Strip given @var{number} of leading components from file names before
3115 extraction. For example, if archive @file{archive.tar} contained
3116 @file{/some/file/name}, then running
3119 tar --extract --file archive.tar --strip-components=2
3123 would extract this file to file @file{name}.
3125 @opsummary{suffix}, summary
3126 @item --suffix=@var{suffix}
3128 Alters the suffix @command{tar} uses when backing up files from the default
3129 @samp{~}. @xref{backup}.
3131 @opsummary{tape-length}
3132 @item --tape-length=@var{num}
3135 Specifies the length of tapes that @command{tar} is writing as being
3136 @w{@var{num} x 1024} bytes long. @xref{Using Multiple Tapes}.
3138 @opsummary{test-label}
3141 Reads the volume label. If an argument is specified, test whether it
3142 matches the volume label. @xref{--test-label option}.
3144 @opsummary{to-command}
3145 @item --to-command=@var{command}
3147 During extraction @command{tar} will pipe extracted files to the
3148 standard input of @var{command}. @xref{Writing to an External Program}.
3150 @opsummary{to-stdout}
3154 During extraction, @command{tar} will extract files to stdout rather
3155 than to the file system. @xref{Writing to Standard Output}.
3158 @item --totals[=@var{signo}]
3160 Displays the total number of bytes transferred when processing an
3161 archive. If an argument is given, these data are displayed on
3162 request, when signal @var{signo} is delivered to @command{tar}.
3169 Sets the data modification time of extracted files to the extraction time,
3170 rather than the data modification time stored in the archive.
3171 @xref{Data Modification Times}.
3173 @opsummary{transform}
3174 @item --transform=@var{sed-expr}
3176 Transform file or member names using @command{sed} replacement expression
3177 @var{sed-expr}. For example,
3180 $ @kbd{tar cf archive.tar --transform 's,^\./,usr/,' .}
3184 will add to @file{archive} files from the current working directory,
3185 replacing initial @samp{./} prefix with @samp{usr/}. For the detailed
3186 discussion, @xref{transform}.
3188 To see transformed member names in verbose listings, use
3189 @option{--show-transformed-names} option
3190 (@pxref{show-transformed-names}).
3192 @opsummary{uncompress}
3195 (See @option{--compress}. @pxref{gzip})
3200 (See @option{--gzip}. @pxref{gzip})
3202 @opsummary{unlink-first}
3203 @item --unlink-first
3206 Directs @command{tar} to remove the corresponding file from the file
3207 system before extracting it from the archive. @xref{Unlink First}.
3211 Enable unquoting input file or member names (default). @xref{input
3214 @opsummary{use-compress-program}
3215 @item --use-compress-program=@var{prog}
3217 Instructs @command{tar} to access the archive through @var{prog}, which is
3218 presumed to be a compression program of some sort. @xref{gzip}.
3223 Display file modification dates in @acronym{UTC}. This option implies
3230 Specifies that @command{tar} should be more verbose about the
3231 operations it is performing. This option can be specified multiple
3232 times for some operations to increase the amount of information displayed.
3239 Verifies that the archive was correctly written when creating an
3240 archive. @xref{verify}.
3245 Print information about the program's name, version, origin and legal
3246 status, all on standard output, and then exit successfully.
3249 @opsummary{volno-file}
3250 @item --volno-file=@var{file}
3252 Used in conjunction with @option{--multi-volume}. @command{tar} will
3253 keep track of which volume of a multi-volume archive it is working in
3254 @var{file}. @xref{volno-file}.
3256 @opsummary{wildcards}
3258 Use wildcards when matching member names with patterns.
3259 @xref{controlling pattern-matching}.
3261 @opsummary{wildcards-match-slash}
3262 @item --wildcards-match-slash
3263 Wildcards match @samp{/}.
3264 @xref{controlling pattern-matching}.
3267 @node Short Option Summary
3268 @subsection Short Options Cross Reference
3270 Here is an alphabetized list of all of the short option forms, matching
3271 them with the equivalent long option.
3273 @multitable @columnfractions 0.20 0.80
3274 @headitem Short Option @tab Reference
3276 @item -A @tab @ref{--concatenate}.
3278 @item -B @tab @ref{--read-full-records}.
3280 @item -C @tab @ref{--directory}.
3282 @item -F @tab @ref{--info-script}.
3284 @item -G @tab @ref{--incremental}.
3286 @item -K @tab @ref{--starting-file}.
3288 @item -L @tab @ref{--tape-length}.
3290 @item -M @tab @ref{--multi-volume}.
3292 @item -N @tab @ref{--newer}.
3294 @item -O @tab @ref{--to-stdout}.
3296 @item -P @tab @ref{--absolute-names}.
3298 @item -R @tab @ref{--block-number}.
3300 @item -S @tab @ref{--sparse}.
3302 @item -T @tab @ref{--files-from}.
3304 @item -U @tab @ref{--unlink-first}.
3306 @item -V @tab @ref{--label}.
3308 @item -W @tab @ref{--verify}.
3310 @item -X @tab @ref{--exclude-from}.
3312 @item -Z @tab @ref{--compress}.
3314 @item -b @tab @ref{--blocking-factor}.
3316 @item -c @tab @ref{--create}.
3318 @item -d @tab @ref{--compare}.
3320 @item -f @tab @ref{--file}.
3322 @item -g @tab @ref{--listed-incremental}.
3324 @item -h @tab @ref{--dereference}.
3326 @item -i @tab @ref{--ignore-zeros}.
3328 @item -j @tab @ref{--bzip2}.
3330 @item -k @tab @ref{--keep-old-files}.
3332 @item -l @tab @ref{--check-links}.
3334 @item -m @tab @ref{--touch}.
3336 @item -o @tab When creating, @ref{--no-same-owner}, when extracting ---
3337 @ref{--portability}.
3339 The later usage is deprecated. It is retained for compatibility with
3340 the earlier versions of @GNUTAR{}. In future releases
3341 @option{-o} will be equivalent to @option{--no-same-owner} only.
3343 @item -p @tab @ref{--preserve-permissions}.
3345 @item -r @tab @ref{--append}.
3347 @item -s @tab @ref{--same-order}.
3349 @item -t @tab @ref{--list}.
3351 @item -u @tab @ref{--update}.
3353 @item -v @tab @ref{--verbose}.
3355 @item -w @tab @ref{--interactive}.
3357 @item -x @tab @ref{--extract}.
3359 @item -z @tab @ref{--gzip}.
3364 @section @GNUTAR{} documentation
3366 @cindex Getting program version number
3368 @cindex Version of the @command{tar} program
3369 Being careful, the first thing is really checking that you are using
3370 @GNUTAR{}, indeed. The @option{--version} option
3371 causes @command{tar} to print information about its name, version,
3372 origin and legal status, all on standard output, and then exit
3373 successfully. For example, @w{@samp{tar --version}} might print:
3376 tar (GNU tar) @value{VERSION}
3377 Copyright (C) 2006 Free Software Foundation, Inc.
3378 This is free software. You may redistribute copies of it under the terms
3379 of the GNU General Public License <http://www.gnu.org/licenses/gpl.html>.
3380 There is NO WARRANTY, to the extent permitted by law.
3382 Written by John Gilmore and Jay Fenlason.
3386 The first occurrence of @samp{tar} in the result above is the program
3387 name in the package (for example, @command{rmt} is another program),
3388 while the second occurrence of @samp{tar} is the name of the package
3389 itself, containing possibly many programs. The package is currently
3390 named @samp{tar}, after the name of the main program it
3391 contains@footnote{There are plans to merge the @command{cpio} and
3392 @command{tar} packages into a single one which would be called
3393 @code{paxutils}. So, who knows if, one of this days, the
3394 @option{--version} would not output @w{@samp{tar (@acronym{GNU}
3397 @cindex Obtaining help
3398 @cindex Listing all @command{tar} options
3399 @xopindex{help, introduction}
3400 Another thing you might want to do is checking the spelling or meaning
3401 of some particular @command{tar} option, without resorting to this
3402 manual, for once you have carefully read it. @GNUTAR{}
3403 has a short help feature, triggerable through the
3404 @option{--help} option. By using this option, @command{tar} will
3405 print a usage message listing all available options on standard
3406 output, then exit successfully, without doing anything else and
3407 ignoring all other options. Even if this is only a brief summary, it
3408 may be several screens long. So, if you are not using some kind of
3409 scrollable window, you might prefer to use something like:
3412 $ @kbd{tar --help | less}
3416 presuming, here, that you like using @command{less} for a pager. Other
3417 popular pagers are @command{more} and @command{pg}. If you know about some
3418 @var{keyword} which interests you and do not want to read all the
3419 @option{--help} output, another common idiom is doing:
3422 tar --help | grep @var{keyword}
3426 for getting only the pertinent lines. Notice, however, that some
3427 @command{tar} options have long description lines and the above
3428 command will list only the first of them.
3430 The exact look of the option summary displayed by @kbd{tar --help} is
3431 configurable. @xref{Configuring Help Summary}, for a detailed description.
3434 If you only wish to check the spelling of an option, running @kbd{tar
3435 --usage} may be a better choice. This will display a terse list of
3436 @command{tar} option without accompanying explanations.
3438 The short help output is quite succinct, and you might have to get
3439 back to the full documentation for precise points. If you are reading
3440 this paragraph, you already have the @command{tar} manual in some
3441 form. This manual is available in a variety of forms from
3442 @url{http://www.gnu.org/software/tar/manual}. It may be printed out of the @GNUTAR{}
3443 distribution, provided you have @TeX{} already installed somewhere,
3444 and a laser printer around. Just configure the distribution, execute
3445 the command @w{@samp{make dvi}}, then print @file{doc/tar.dvi} the
3446 usual way (contact your local guru to know how). If @GNUTAR{}
3447 has been conveniently installed at your place, this
3448 manual is also available in interactive, hypertextual form as an Info
3449 file. Just call @w{@samp{info tar}} or, if you do not have the
3450 @command{info} program handy, use the Info reader provided within
3451 @acronym{GNU} Emacs, calling @samp{tar} from the main Info menu.
3453 There is currently no @code{man} page for @GNUTAR{}.
3454 If you observe such a @code{man} page on the system you are running,
3455 either it does not belong to @GNUTAR{}, or it has not
3456 been produced by @acronym{GNU}. Some package maintainers convert
3457 @kbd{tar --help} output to a man page, using @command{help2man}. In
3458 any case, please bear in mind that the authoritative source of
3459 information about @GNUTAR{} is this Texinfo documentation.
3462 @section Obtaining @GNUTAR{} default values
3464 @opindex show-defaults
3465 @GNUTAR{} has some predefined defaults that are used when you do not
3466 explicitly specify another values. To obtain a list of such
3467 defaults, use @option{--show-defaults} option. This will output the
3468 values in the form of @command{tar} command line options:
3472 @kbd{tar --show-defaults}
3473 --format=gnu -f- -b20 --quoting-style=escape
3474 --rmt-command=/etc/rmt --rsh-command=/usr/bin/rsh
3479 Notice, that this option outputs only one line. The example output above
3480 has been split to fit page boundaries.
3483 The above output shows that this version of @GNUTAR{} defaults to
3484 using @samp{gnu} archive format (@pxref{Formats}), it uses standard
3485 output as the archive, if no @option{--file} option has been given
3486 (@pxref{file tutorial}), the default blocking factor is 20
3487 (@pxref{Blocking Factor}). It also shows the default locations where
3488 @command{tar} will look for @command{rmt} and @command{rsh} binaries.
3491 @section Checking @command{tar} progress
3493 Typically, @command{tar} performs most operations without reporting any
3494 information to the user except error messages. When using @command{tar}
3495 with many options, particularly ones with complicated or
3496 difficult-to-predict behavior, it is possible to make serious mistakes.
3497 @command{tar} provides several options that make observing @command{tar}
3498 easier. These options cause @command{tar} to print information as it
3499 progresses in its job, and you might want to use them just for being
3500 more careful about what is going on, or merely for entertaining
3501 yourself. If you have encountered a problem when operating on an
3502 archive, however, you may need more information than just an error
3503 message in order to solve the problem. The following options can be
3504 helpful diagnostic tools.
3506 @cindex Verbose operation
3508 Normally, the @option{--list} (@option{-t}) command to list an archive
3509 prints just the file names (one per line) and the other commands are
3510 silent. When used with most operations, the @option{--verbose}
3511 (@option{-v}) option causes @command{tar} to print the name of each
3512 file or archive member as it is processed. This and the other options
3513 which make @command{tar} print status information can be useful in
3514 monitoring @command{tar}.
3516 With @option{--create} or @option{--extract}, @option{--verbose} used
3517 once just prints the names of the files or members as they are processed.
3518 Using it twice causes @command{tar} to print a longer listing
3519 (@xref{verbose member listing}, for the description) for each member.
3520 Since @option{--list} already prints the names of the members,
3521 @option{--verbose} used once with @option{--list} causes @command{tar}
3522 to print an @samp{ls -l} type listing of the files in the archive.
3523 The following examples both extract members with long list output:
3526 $ @kbd{tar --extract --file=archive.tar --verbose --verbose}
3527 $ @kbd{tar xvvf archive.tar}
3530 Verbose output appears on the standard output except when an archive is
3531 being written to the standard output, as with @samp{tar --create
3532 --file=- --verbose} (@samp{tar cfv -}, or even @samp{tar cv}---if the
3533 installer let standard output be the default archive). In that case
3534 @command{tar} writes verbose output to the standard error stream.
3536 If @option{--index-file=@var{file}} is specified, @command{tar} sends
3537 verbose output to @var{file} rather than to standard output or standard
3541 @cindex Obtaining total status information
3543 The @option{--totals} option causes @command{tar} to print on the
3544 standard error the total amount of bytes transferred when processing
3545 an archive. When creating or appending to an archive, this option
3546 prints the number of bytes written to the archive and the average
3547 speed at which they have been written, e.g.:
3551 $ @kbd{tar -c -f archive.tar --totals /home}
3552 Total bytes written: 7924664320 (7.4GiB, 85MiB/s)
3556 When reading an archive, this option displays the number of bytes
3561 $ @kbd{tar -x -f archive.tar --totals}
3562 Total bytes read: 7924664320 (7.4GiB, 95MiB/s)
3566 Finally, when deleting from an archive, the @option{--totals} option
3567 displays both numbers plus number of bytes removed from the archive:
3571 $ @kbd{tar --delete -f foo.tar --totals --wildcards '*~'}
3572 Total bytes read: 9543680 (9.2MiB, 201MiB/s)
3573 Total bytes written: 3829760 (3.7MiB, 81MiB/s)
3574 Total bytes deleted: 1474048
3578 You can also obtain this information on request. When
3579 @option{--totals} is used with an argument, this argument is
3580 interpreted as a symbolic name of a signal, upon delivery of which the
3581 statistics is to be printed:
3584 @item --totals=@var{signo}
3585 Print statistics upon delivery of signal @var{signo}. Valid arguments
3586 are: @code{SIGHUP}, @code{SIGQUIT}, @code{SIGINT}, @code{SIGUSR1} and
3587 @code{SIGUSR2}. Shortened names without @samp{SIG} prefix are also
3591 Both forms of @option{--totals} option can be used simultaneously.
3592 Thus, @kbd{tar -x --totals --totals=USR1} instructs @command{tar} to
3593 extract all members from its default archive and print statistics
3594 after finishing the extraction, as well as when receiving signal
3597 @anchor{Progress information}
3598 @cindex Progress information
3600 The @option{--checkpoint} option prints an occasional message
3601 as @command{tar} reads or writes the archive. It is designed for
3602 those who don't need the more detailed (and voluminous) output of
3603 @option{--block-number} (@option{-R}), but do want visual confirmation
3604 that @command{tar} is actually making forward progress. By default it
3605 prints a message each 10 records read or written. This can be changed
3606 by giving it a numeric argument after an equal sign:
3609 $ @kbd{tar -c --checkpoint=1000} /var
3610 tar: Write checkpoint 1000
3611 tar: Write checkpoint 2000
3612 tar: Write checkpoint 3000
3615 This example shows the default checkpoint message used by
3616 @command{tar}. If you place a dot immediately after the equal
3617 sign, it will print a @samp{.} at each checkpoint. For example:
3620 $ @kbd{tar -c --checkpoint=.1000} /var
3624 @opindex show-omitted-dirs
3625 @anchor{show-omitted-dirs}
3626 The @option{--show-omitted-dirs} option, when reading an archive---with
3627 @option{--list} or @option{--extract}, for example---causes a message
3628 to be printed for each directory in the archive which is skipped.
3629 This happens regardless of the reason for skipping: the directory might
3630 not have been named on the command line (implicitly or explicitly),
3631 it might be excluded by the use of the
3632 @option{--exclude=@var{pattern}} option, or some other reason.
3634 @opindex block-number
3635 @cindex Block number where error occurred
3636 @anchor{block-number}
3637 If @option{--block-number} (@option{-R}) is used, @command{tar} prints, along with
3638 every message it would normally produce, the block number within the
3639 archive where the message was triggered. Also, supplementary messages
3640 are triggered when reading blocks full of NULs, or when hitting end of
3641 file on the archive. As of now, if the archive if properly terminated
3642 with a NUL block, the reading of the file may stop before end of file
3643 is met, so the position of end of file will not usually show when
3644 @option{--block-number} (@option{-R}) is used. Note that @GNUTAR{}
3645 drains the archive before exiting when reading the
3646 archive from a pipe.
3648 @cindex Error message, block number of
3649 This option is especially useful when reading damaged archives, since
3650 it helps pinpoint the damaged sections. It can also be used with
3651 @option{--list} (@option{-t}) when listing a file-system backup tape, allowing you to
3652 choose among several backup tapes when retrieving a file later, in
3653 favor of the tape where the file appears earliest (closest to the
3654 front of the tape). @xref{backup}.
3657 @section Asking for Confirmation During Operations
3658 @cindex Interactive operation
3660 Typically, @command{tar} carries out a command without stopping for
3661 further instructions. In some situations however, you may want to
3662 exclude some files and archive members from the operation (for instance
3663 if disk or storage space is tight). You can do this by excluding
3664 certain files automatically (@pxref{Choosing}), or by performing
3665 an operation interactively, using the @option{--interactive} (@option{-w}) option.
3666 @command{tar} also accepts @option{--confirmation} for this option.
3668 @opindex interactive
3669 When the @option{--interactive} (@option{-w}) option is specified, before
3670 reading, writing, or deleting files, @command{tar} first prints a message
3671 for each such file, telling what operation it intends to take, then asks
3672 for confirmation on the terminal. The actions which require
3673 confirmation include adding a file to the archive, extracting a file
3674 from the archive, deleting a file from the archive, and deleting a file
3675 from disk. To confirm the action, you must type a line of input
3676 beginning with @samp{y}. If your input line begins with anything other
3677 than @samp{y}, @command{tar} skips that file.
3679 If @command{tar} is reading the archive from the standard input,
3680 @command{tar} opens the file @file{/dev/tty} to support the interactive
3683 Verbose output is normally sent to standard output, separate from
3684 other error messages. However, if the archive is produced directly
3685 on standard output, then verbose output is mixed with errors on
3686 @code{stderr}. Producing the archive on standard output may be used
3687 as a way to avoid using disk space, when the archive is soon to be
3688 consumed by another process reading it, say. Some people felt the need
3689 of producing an archive on stdout, still willing to segregate between
3690 verbose output and error output. A possible approach would be using a
3691 named pipe to receive the archive, and having the consumer process to
3692 read from that named pipe. This has the advantage of letting standard
3693 output free to receive verbose output, all separate from errors.
3696 @chapter @GNUTAR{} Operations
3709 @section Basic @GNUTAR{} Operations
3711 The basic @command{tar} operations, @option{--create} (@option{-c}),
3712 @option{--list} (@option{-t}) and @option{--extract} (@option{--get},
3713 @option{-x}), are currently presented and described in the tutorial
3714 chapter of this manual. This section provides some complementary notes
3715 for these operations.
3718 @xopindex{create, complementary notes}
3722 Creating an empty archive would have some kind of elegance. One can
3723 initialize an empty archive and later use @option{--append}
3724 (@option{-r}) for adding all members. Some applications would not
3725 welcome making an exception in the way of adding the first archive
3726 member. On the other hand, many people reported that it is
3727 dangerously too easy for @command{tar} to destroy a magnetic tape with
3728 an empty archive@footnote{This is well described in @cite{Unix-haters
3729 Handbook}, by Simson Garfinkel, Daniel Weise & Steven Strassmann, IDG
3730 Books, ISBN 1-56884-203-1.}. The two most common errors are:
3734 Mistakingly using @code{create} instead of @code{extract}, when the
3735 intent was to extract the full contents of an archive. This error
3736 is likely: keys @kbd{c} and @kbd{x} are right next to each other on
3737 the QWERTY keyboard. Instead of being unpacked, the archive then
3738 gets wholly destroyed. When users speak about @dfn{exploding} an
3739 archive, they usually mean something else :-).
3742 Forgetting the argument to @code{file}, when the intent was to create
3743 an archive with a single file in it. This error is likely because a
3744 tired user can easily add the @kbd{f} key to the cluster of option
3745 letters, by the mere force of habit, without realizing the full
3746 consequence of doing so. The usual consequence is that the single
3747 file, which was meant to be saved, is rather destroyed.
3750 So, recognizing the likelihood and the catastrophic nature of these
3751 errors, @GNUTAR{} now takes some distance from elegance, and
3752 cowardly refuses to create an archive when @option{--create} option is
3753 given, there are no arguments besides options, and
3754 @option{--files-from} (@option{-T}) option is @emph{not} used. To get
3755 around the cautiousness of @GNUTAR{} and nevertheless create an
3756 archive with nothing in it, one may still use, as the value for the
3757 @option{--files-from} option, a file with no names in it, as shown in
3758 the following commands:
3761 @kbd{tar --create --file=empty-archive.tar --files-from=/dev/null}
3762 @kbd{tar cfT empty-archive.tar /dev/null}
3765 @xopindex{extract, complementary notes}
3770 A socket is stored, within a @GNUTAR{} archive, as a pipe.
3772 @item @option{--list} (@option{-t})
3774 @GNUTAR{} now shows dates as @samp{1996-08-30},
3775 while it used to show them as @samp{Aug 30 1996}. Preferably,
3776 people should get used to ISO 8601 dates. Local American dates should
3777 be made available again with full date localization support, once
3778 ready. In the meantime, programs not being localizable for dates
3779 should prefer international dates, that's really the way to go.
3781 Look up @url{http://www.cl.cam.ac.uk/@/~mgk25/@/iso-time.html} if you
3782 are curious, it contains a detailed explanation of the ISO 8601 standard.
3787 @section Advanced @GNUTAR{} Operations
3789 Now that you have learned the basics of using @GNUTAR{}, you may want
3790 to learn about further ways in which @command{tar} can help you.
3792 This chapter presents five, more advanced operations which you probably
3793 won't use on a daily basis, but which serve more specialized functions.
3794 We also explain the different styles of options and why you might want
3795 to use one or another, or a combination of them in your @command{tar}
3796 commands. Additionally, this chapter includes options which allow you to
3797 define the output from @command{tar} more carefully, and provide help and
3798 error correction in special circumstances.
3800 @FIXME{check this after the chapter is actually revised to make sure
3801 it still introduces the info in the chapter correctly : ).}
3813 @subsection The Five Advanced @command{tar} Operations
3816 In the last chapter, you learned about the first three operations to
3817 @command{tar}. This chapter presents the remaining five operations to
3818 @command{tar}: @option{--append}, @option{--update}, @option{--concatenate},
3819 @option{--delete}, and @option{--compare}.
3821 You are not likely to use these operations as frequently as those
3822 covered in the last chapter; however, since they perform specialized
3823 functions, they are quite useful when you do need to use them. We
3824 will give examples using the same directory and files that you created
3825 in the last chapter. As you may recall, the directory is called
3826 @file{practice}, the files are @samp{jazz}, @samp{blues}, @samp{folk},
3827 @samp{rock}, and the two archive files you created are
3828 @samp{collection.tar} and @samp{music.tar}.
3830 We will also use the archive files @samp{afiles.tar} and
3831 @samp{bfiles.tar}. The archive @samp{afiles.tar} contains the members @samp{apple},
3832 @samp{angst}, and @samp{aspic}; @samp{bfiles.tar} contains the members
3833 @samp{./birds}, @samp{baboon}, and @samp{./box}.
3835 Unless we state otherwise, all practicing you do and examples you follow
3836 in this chapter will take place in the @file{practice} directory that
3837 you created in the previous chapter; see @ref{prepare for examples}.
3838 (Below in this section, we will remind you of the state of the examples
3839 where the last chapter left them.)
3841 The five operations that we will cover in this chapter are:
3846 Add new entries to an archive that already exists.
3849 Add more recent copies of archive members to the end of an archive, if
3854 Add one or more pre-existing archives to the end of another archive.
3856 Delete items from an archive (does not work on tapes).
3860 Compare archive members to their counterparts in the file system.
3864 @subsection How to Add Files to Existing Archives: @option{--append}
3868 If you want to add files to an existing archive, you don't need to
3869 create a new archive; you can use @option{--append} (@option{-r}).
3870 The archive must already exist in order to use @option{--append}. (A
3871 related operation is the @option{--update} operation; you can use this
3872 to add newer versions of archive members to an existing archive. To learn how to
3873 do this with @option{--update}, @pxref{update}.)
3875 If you use @option{--append} to add a file that has the same name as an
3876 archive member to an archive containing that archive member, then the
3877 old member is not deleted. What does happen, however, is somewhat
3878 complex. @command{tar} @emph{allows} you to have infinite number of files
3879 with the same name. Some operations treat these same-named members no
3880 differently than any other set of archive members: for example, if you
3881 view an archive with @option{--list} (@option{-t}), you will see all
3882 of those members listed, with their data modification times, owners, etc.
3884 Other operations don't deal with these members as perfectly as you might
3885 prefer; if you were to use @option{--extract} to extract the archive,
3886 only the most recently added copy of a member with the same name as four
3887 other members would end up in the working directory. This is because
3888 @option{--extract} extracts an archive in the order the members appeared
3889 in the archive; the most recently archived members will be extracted
3890 last. Additionally, an extracted member will @emph{replace} a file of
3891 the same name which existed in the directory already, and @command{tar}
3892 will not prompt you about this@footnote{Unless you give it
3893 @option{--keep-old-files} option, or the disk copy is newer than the
3894 the one in the archive and you invoke @command{tar} with
3895 @option{--keep-newer-files} option}. Thus, only the most recently archived
3896 member will end up being extracted, as it will replace the one
3897 extracted before it, and so on.
3899 There exists a special option that allows you to get around this
3900 behavior and extract (or list) only a particular copy of the file.
3901 This is @option{--occurrence} option. If you run @command{tar} with
3902 this option, it will extract only the first copy of the file. You
3903 may also give this option an argument specifying the number of
3904 copy to be extracted. Thus, for example if the archive
3905 @file{archive.tar} contained three copies of file @file{myfile}, then
3909 tar --extract --file archive.tar --occurrence=2 myfile
3913 would extract only the second copy. @xref{Option
3914 Summary,---occurrence}, for the description of @option{--occurrence}
3917 @FIXME{ hag -- you might want to incorporate some of the above into the
3918 MMwtSN node; not sure. i didn't know how to make it simpler...
3920 There are a few ways to get around this. (maybe xref Multiple Members
3921 with the Same Name.}
3923 @cindex Members, replacing with other members
3924 @cindex Replacing members with other members
3925 If you want to replace an archive member, use @option{--delete} to
3926 delete the member you want to remove from the archive, , and then use
3927 @option{--append} to add the member you want to be in the archive. Note
3928 that you can not change the order of the archive; the most recently
3929 added member will still appear last. In this sense, you cannot truly
3930 ``replace'' one member with another. (Replacing one member with another
3931 will not work on certain types of media, such as tapes; see @ref{delete}
3932 and @ref{Media}, for more information.)
3935 * appending files:: Appending Files to an Archive
3939 @node appending files
3940 @subsubsection Appending Files to an Archive
3942 @cindex Adding files to an Archive
3943 @cindex Appending files to an Archive
3944 @cindex Archives, Appending files to
3946 The simplest way to add a file to an already existing archive is the
3947 @option{--append} (@option{-r}) operation, which writes specified
3948 files into the archive whether or not they are already among the
3951 When you use @option{--append}, you @emph{must} specify file name
3952 arguments, as there is no default. If you specify a file that already
3953 exists in the archive, another copy of the file will be added to the
3954 end of the archive. As with other operations, the member names of the
3955 newly added files will be exactly the same as their names given on the
3956 command line. The @option{--verbose} (@option{-v}) option will print
3957 out the names of the files as they are written into the archive.
3959 @option{--append} cannot be performed on some tape drives, unfortunately,
3960 due to deficiencies in the formats those tape drives use. The archive
3961 must be a valid @command{tar} archive, or else the results of using this
3962 operation will be unpredictable. @xref{Media}.
3964 To demonstrate using @option{--append} to add a file to an archive,
3965 create a file called @file{rock} in the @file{practice} directory.
3966 Make sure you are in the @file{practice} directory. Then, run the
3967 following @command{tar} command to add @file{rock} to
3968 @file{collection.tar}:
3971 $ @kbd{tar --append --file=collection.tar rock}
3975 If you now use the @option{--list} (@option{-t}) operation, you will see that
3976 @file{rock} has been added to the archive:
3979 $ @kbd{tar --list --file=collection.tar}
3980 -rw-r--r-- me user 28 1996-10-18 16:31 jazz
3981 -rw-r--r-- me user 21 1996-09-23 16:44 blues
3982 -rw-r--r-- me user 20 1996-09-23 16:44 folk
3983 -rw-r--r-- me user 20 1996-09-23 16:44 rock
3987 @subsubsection Multiple Members with the Same Name
3989 You can use @option{--append} (@option{-r}) to add copies of files
3990 which have been updated since the archive was created. (However, we
3991 do not recommend doing this since there is another @command{tar}
3992 option called @option{--update}; @xref{update}, for more information.
3993 We describe this use of @option{--append} here for the sake of
3994 completeness.) When you extract the archive, the older version will
3995 be effectively lost. This works because files are extracted from an
3996 archive in the order in which they were archived. Thus, when the
3997 archive is extracted, a file archived later in time will replace a
3998 file of the same name which was archived earlier, even though the
3999 older version of the file will remain in the archive unless you delete
4000 all versions of the file.
4002 Supposing you change the file @file{blues} and then append the changed
4003 version to @file{collection.tar}. As you saw above, the original
4004 @file{blues} is in the archive @file{collection.tar}. If you change the
4005 file and append the new version of the file to the archive, there will
4006 be two copies in the archive. When you extract the archive, the older
4007 version of the file will be extracted first, and then replaced by the
4008 newer version when it is extracted.
4010 You can append the new, changed copy of the file @file{blues} to the
4011 archive in this way:
4014 $ @kbd{tar --append --verbose --file=collection.tar blues}
4019 Because you specified the @option{--verbose} option, @command{tar} has
4020 printed the name of the file being appended as it was acted on. Now
4021 list the contents of the archive:
4024 $ @kbd{tar --list --verbose --file=collection.tar}
4025 -rw-r--r-- me user 28 1996-10-18 16:31 jazz
4026 -rw-r--r-- me user 21 1996-09-23 16:44 blues
4027 -rw-r--r-- me user 20 1996-09-23 16:44 folk
4028 -rw-r--r-- me user 20 1996-09-23 16:44 rock
4029 -rw-r--r-- me user 58 1996-10-24 18:30 blues
4033 The newest version of @file{blues} is now at the end of the archive
4034 (note the different creation dates and file sizes). If you extract
4035 the archive, the older version of the file @file{blues} will be
4036 replaced by the newer version. You can confirm this by extracting
4037 the archive and running @samp{ls} on the directory.
4039 If you wish to extract the first occurrence of the file @file{blues}
4040 from the archive, use @option{--occurrence} option, as shown in
4041 the following example:
4044 $ @kbd{tar --extract -vv --occurrence --file=collection.tar blues}
4045 -rw-r--r-- me user 21 1996-09-23 16:44 blues
4048 @xref{Writing}, for more information on @option{--extract} and
4049 @xref{Option Summary, --occurrence}, for the description of
4050 @option{--occurrence} option.
4053 @subsection Updating an Archive
4055 @cindex Updating an archive
4058 In the previous section, you learned how to use @option{--append} to
4059 add a file to an existing archive. A related operation is
4060 @option{--update} (@option{-u}). The @option{--update} operation
4061 updates a @command{tar} archive by comparing the date of the specified
4062 archive members against the date of the file with the same name. If
4063 the file has been modified more recently than the archive member, then
4064 the newer version of the file is added to the archive (as with
4067 Unfortunately, you cannot use @option{--update} with magnetic tape drives.
4068 The operation will fail.
4070 @FIXME{other examples of media on which --update will fail? need to ask
4071 charles and/or mib/thomas/dave shevett..}
4073 Both @option{--update} and @option{--append} work by adding to the end
4074 of the archive. When you extract a file from the archive, only the
4075 version stored last will wind up in the file system, unless you use
4076 the @option{--backup} option. @xref{multiple}, for a detailed discussion.
4083 @subsubsection How to Update an Archive Using @option{--update}
4085 You must use file name arguments with the @option{--update}
4086 (@option{-u}) operation. If you don't specify any files,
4087 @command{tar} won't act on any files and won't tell you that it didn't
4088 do anything (which may end up confusing you).
4090 @c note: the above parenthetical added because in fact, this
4091 @c behavior just confused the author. :-)
4093 To see the @option{--update} option at work, create a new file,
4094 @file{classical}, in your practice directory, and some extra text to the
4095 file @file{blues}, using any text editor. Then invoke @command{tar} with
4096 the @samp{update} operation and the @option{--verbose} (@option{-v})
4097 option specified, using the names of all the files in the practice
4098 directory as file name arguments:
4101 $ @kbd{tar --update -v -f collection.tar blues folk rock classical}
4108 Because we have specified verbose mode, @command{tar} prints out the names
4109 of the files it is working on, which in this case are the names of the
4110 files that needed to be updated. If you run @samp{tar --list} and look
4111 at the archive, you will see @file{blues} and @file{classical} at its
4112 end. There will be a total of two versions of the member @samp{blues};
4113 the one at the end will be newer and larger, since you added text before
4116 (The reason @command{tar} does not overwrite the older file when updating
4117 it is because writing to the middle of a section of tape is a difficult
4118 process. Tapes are not designed to go backward. @xref{Media}, for more
4119 information about tapes.
4121 @option{--update} (@option{-u}) is not suitable for performing backups for two
4122 reasons: it does not change directory content entries, and it
4123 lengthens the archive every time it is used. The @GNUTAR{}
4124 options intended specifically for backups are more
4125 efficient. If you need to run backups, please consult @ref{Backups}.
4128 @subsection Combining Archives with @option{--concatenate}
4130 @cindex Adding archives to an archive
4131 @cindex Concatenating Archives
4132 @opindex concatenate
4134 @c @cindex @option{-A} described
4135 Sometimes it may be convenient to add a second archive onto the end of
4136 an archive rather than adding individual files to the archive. To add
4137 one or more archives to the end of another archive, you should use the
4138 @option{--concatenate} (@option{--catenate}, @option{-A}) operation.
4140 To use @option{--concatenate}, give the first archive with
4141 @option{--file} option and name the rest of archives to be
4142 concatenated on the command line. The members, and their member
4143 names, will be copied verbatim from those archives to the first one.
4144 @footnote{This can cause multiple members to have the same name, for
4145 information on how this affects reading the archive, @ref{multiple}.}
4146 The new, concatenated archive will be called by the same name as the
4147 one given with the @option{--file} option. As usual, if you omit
4148 @option{--file}, @command{tar} will use the value of the environment
4149 variable @env{TAPE}, or, if this has not been set, the default archive name.
4151 @FIXME{There is no way to specify a new name...}
4153 To demonstrate how @option{--concatenate} works, create two small archives
4154 called @file{bluesrock.tar} and @file{folkjazz.tar}, using the relevant
4155 files from @file{practice}:
4158 $ @kbd{tar -cvf bluesrock.tar blues rock}
4161 $ @kbd{tar -cvf folkjazz.tar folk jazz}
4167 If you like, You can run @samp{tar --list} to make sure the archives
4168 contain what they are supposed to:
4171 $ @kbd{tar -tvf bluesrock.tar}
4172 -rw-r--r-- melissa user 105 1997-01-21 19:42 blues
4173 -rw-r--r-- melissa user 33 1997-01-20 15:34 rock
4174 $ @kbd{tar -tvf jazzfolk.tar}
4175 -rw-r--r-- melissa user 20 1996-09-23 16:44 folk
4176 -rw-r--r-- melissa user 65 1997-01-30 14:15 jazz
4179 We can concatenate these two archives with @command{tar}:
4183 $ @kbd{tar --concatenate --file=bluesrock.tar jazzfolk.tar}
4186 If you now list the contents of the @file{bluesrock.tar}, you will see
4187 that now it also contains the archive members of @file{jazzfolk.tar}:
4190 $ @kbd{tar --list --file=bluesrock.tar}
4197 When you use @option{--concatenate}, the source and target archives must
4198 already exist and must have been created using compatible format
4199 parameters. Notice, that @command{tar} does not check whether the
4200 archives it concatenates have compatible formats, it does not
4201 even check if the files are really tar archives.
4203 Like @option{--append} (@option{-r}), this operation cannot be performed on some
4204 tape drives, due to deficiencies in the formats those tape drives use.
4206 @cindex @code{concatenate} vs @command{cat}
4207 @cindex @command{cat} vs @code{concatenate}
4208 It may seem more intuitive to you to want or try to use @command{cat} to
4209 concatenate two archives instead of using the @option{--concatenate}
4210 operation; after all, @command{cat} is the utility for combining files.
4212 However, @command{tar} archives incorporate an end-of-file marker which
4213 must be removed if the concatenated archives are to be read properly as
4214 one archive. @option{--concatenate} removes the end-of-archive marker
4215 from the target archive before each new archive is appended. If you use
4216 @command{cat} to combine the archives, the result will not be a valid
4217 @command{tar} format archive. If you need to retrieve files from an
4218 archive that was added to using the @command{cat} utility, use the
4219 @option{--ignore-zeros} (@option{-i}) option. @xref{Ignore Zeros}, for further
4220 information on dealing with archives improperly combined using the
4221 @command{cat} shell utility.
4224 @subsection Removing Archive Members Using @option{--delete}
4226 @cindex Deleting files from an archive
4227 @cindex Removing files from an archive
4230 You can remove members from an archive by using the @option{--delete}
4231 option. Specify the name of the archive with @option{--file}
4232 (@option{-f}) and then specify the names of the members to be deleted;
4233 if you list no member names, nothing will be deleted. The
4234 @option{--verbose} option will cause @command{tar} to print the names
4235 of the members as they are deleted. As with @option{--extract}, you
4236 must give the exact member names when using @samp{tar --delete}.
4237 @option{--delete} will remove all versions of the named file from the
4238 archive. The @option{--delete} operation can run very slowly.
4240 Unlike other operations, @option{--delete} has no short form.
4242 @cindex Tapes, using @option{--delete} and
4243 @cindex Deleting from tape archives
4244 This operation will rewrite the archive. You can only use
4245 @option{--delete} on an archive if the archive device allows you to
4246 write to any point on the media, such as a disk; because of this, it
4247 does not work on magnetic tapes. Do not try to delete an archive member
4248 from a magnetic tape; the action will not succeed, and you will be
4249 likely to scramble the archive and damage your tape. There is no safe
4250 way (except by completely re-writing the archive) to delete files from
4251 most kinds of magnetic tape. @xref{Media}.
4253 To delete all versions of the file @file{blues} from the archive
4254 @file{collection.tar} in the @file{practice} directory, make sure you
4255 are in that directory, and then,
4258 $ @kbd{tar --list --file=collection.tar}
4263 $ @kbd{tar --delete --file=collection.tar blues}
4264 $ @kbd{tar --list --file=collection.tar}
4271 @FIXME{Check if the above listing is actually produced after running
4272 all the examples on collection.tar.}
4274 The @option{--delete} option has been reported to work properly when
4275 @command{tar} acts as a filter from @code{stdin} to @code{stdout}.
4278 @subsection Comparing Archive Members with the File System
4279 @cindex Verifying the currency of an archive
4283 The @option{--compare} (@option{-d}), or @option{--diff} operation compares
4284 specified archive members against files with the same names, and then
4285 reports differences in file size, mode, owner, modification date and
4286 contents. You should @emph{only} specify archive member names, not file
4287 names. If you do not name any members, then @command{tar} will compare the
4288 entire archive. If a file is represented in the archive but does not
4289 exist in the file system, @command{tar} reports a difference.
4291 You have to specify the record size of the archive when modifying an
4292 archive with a non-default record size.
4294 @command{tar} ignores files in the file system that do not have
4295 corresponding members in the archive.
4297 The following example compares the archive members @file{rock},
4298 @file{blues} and @file{funk} in the archive @file{bluesrock.tar} with
4299 files of the same name in the file system. (Note that there is no file,
4300 @file{funk}; @command{tar} will report an error message.)
4303 $ @kbd{tar --compare --file=bluesrock.tar rock blues funk}
4306 tar: funk not found in archive
4309 The spirit behind the @option{--compare} (@option{--diff},
4310 @option{-d}) option is to check whether the archive represents the
4311 current state of files on disk, more than validating the integrity of
4312 the archive media. For this later goal, @xref{verify}.
4314 @node create options
4315 @section Options Used by @option{--create}
4317 @xopindex{create, additional options}
4318 The previous chapter described the basics of how to use
4319 @option{--create} (@option{-c}) to create an archive from a set of files.
4320 @xref{create}. This section described advanced options to be used with
4324 * override:: Overriding File Metadata.
4325 * Ignore Failed Read::
4329 @subsection Overriding File Metadata
4331 As described above, a @command{tar} archive keeps, for each member it contains,
4332 its @dfn{metadata}, such as modification time, mode and ownership of
4333 the file. @GNUTAR{} allows to replace these data with other values
4334 when adding files to the archive. The options described in this
4335 section affect creation of archives of any type. For POSIX archives,
4336 see also @ref{PAX keywords}, for additional ways of controlling
4337 metadata, stored in the archive.
4341 @item --mode=@var{permissions}
4343 When adding files to an archive, @command{tar} will use
4344 @var{permissions} for the archive members, rather than the permissions
4345 from the files. @var{permissions} can be specified either as an octal
4346 number or as symbolic permissions, like with
4347 @command{chmod} (@xref{File permissions, Permissions, File
4348 permissions, fileutils, @acronym{GNU} file utilities}. This reference
4349 also has useful information for those not being overly familiar with
4350 the UNIX permission system). Using latter syntax allows for
4351 more flexibility. For example, the value @samp{a+rw} adds read and write
4352 permissions for everybody, while retaining executable bits on directories
4353 or on any other file already marked as executable:
4356 $ @kbd{tar -c -f archive.tar --mode='a+rw' .}
4359 @item --mtime=@var{date}
4362 When adding files to an archive, @command{tar} will use @var{date} as
4363 the modification time of members when creating archives, instead of
4364 their actual modification times. The argument @var{date} can be
4365 either a textual date representation in almost arbitrary format
4366 (@pxref{Date input formats}) or a name of the existing file, starting
4367 with @samp{/} or @samp{.}. In the latter case, the modification time
4368 of that file will be used.
4370 The following example will set the modification date to 00:00:00 UTC,
4374 $ @kbd{tar -c -f archive.tar --mtime='1970-01-01' .}
4378 When used with @option{--verbose} (@pxref{verbose tutorial}) @GNUTAR{}
4379 will try to convert the specified date back to its textual
4380 representation and compare it with the one given with
4381 @option{--mtime} options. If the two dates differ, @command{tar} will
4382 print a warning saying what date it will use. This is to help user
4383 ensure he is using the right date.
4388 $ @kbd{tar -c -f archive.tar -v --mtime=yesterday .}
4389 tar: Option --mtime: Treating date `yesterday' as 2006-06-20
4394 @item --owner=@var{user}
4397 Specifies that @command{tar} should use @var{user} as the owner of members
4398 when creating archives, instead of the user associated with the source
4399 file. The argument @var{user} can be either an existing user symbolic
4400 name, or a decimal numeric user @acronym{ID}.
4402 There is no value indicating a missing number, and @samp{0} usually means
4403 @code{root}. Some people like to force @samp{0} as the value to offer in
4404 their distributions for the owner of files, because the @code{root} user is
4405 anonymous anyway, so that might as well be the owner of anonymous
4406 archives. For example:
4410 $ @kbd{tar -c -f archive.tar --owner=0 .}
4412 $ @kbd{tar -c -f archive.tar --owner=root .}
4416 @item --group=@var{group}
4419 Files added to the @command{tar} archive will have a group @acronym{ID} of @var{group},
4420 rather than the group from the source file. The argument @var{group}
4421 can be either an existing group symbolic name, or a decimal numeric group @acronym{ID}.
4424 @node Ignore Failed Read
4425 @subsection Ignore Fail Read
4428 @item --ignore-failed-read
4429 @opindex ignore-failed-read
4430 Do not exit with nonzero on unreadable files or directories.
4433 @node extract options
4434 @section Options Used by @option{--extract}
4437 @xopindex{extract, additional options}
4438 The previous chapter showed how to use @option{--extract} to extract
4439 an archive into the file system. Various options cause @command{tar} to
4440 extract more information than just file contents, such as the owner,
4441 the permissions, the modification date, and so forth. This section
4442 presents options to be used with @option{--extract} when certain special
4443 considerations arise. You may review the information presented in
4444 @ref{extract} for more basic information about the
4445 @option{--extract} operation.
4448 * Reading:: Options to Help Read Archives
4449 * Writing:: Changing How @command{tar} Writes Files
4450 * Scarce:: Coping with Scarce Resources
4454 @subsection Options to Help Read Archives
4455 @cindex Options when reading archives
4458 @cindex Reading incomplete records
4459 @cindex Records, incomplete
4460 @opindex read-full-records
4461 Normally, @command{tar} will request data in full record increments from
4462 an archive storage device. If the device cannot return a full record,
4463 @command{tar} will report an error. However, some devices do not always
4464 return full records, or do not require the last record of an archive to
4465 be padded out to the next record boundary. To keep reading until you
4466 obtain a full record, or to accept an incomplete record if it contains
4467 an end-of-archive marker, specify the @option{--read-full-records} (@option{-B}) option
4468 in conjunction with the @option{--extract} or @option{--list} operations.
4471 The @option{--read-full-records} (@option{-B}) option is turned on by default when
4472 @command{tar} reads an archive from standard input, or from a remote
4473 machine. This is because on @acronym{BSD} Unix systems, attempting to read a
4474 pipe returns however much happens to be in the pipe, even if it is
4475 less than was requested. If this option were not enabled, @command{tar}
4476 would fail as soon as it read an incomplete record from the pipe.
4478 If you're not sure of the blocking factor of an archive, you can
4479 read the archive by specifying @option{--read-full-records} (@option{-B}) and
4480 @option{--blocking-factor=@var{512-size}} (@option{-b
4481 @var{512-size}}), using a blocking factor larger than what the archive
4482 uses. This lets you avoid having to determine the blocking factor
4483 of an archive. @xref{Blocking Factor}.
4486 * read full records::
4490 @node read full records
4491 @unnumberedsubsubsec Reading Full Records
4493 @FIXME{need sentence or so of intro here}
4496 @opindex read-full-records
4497 @item --read-full-records
4499 Use in conjunction with @option{--extract} (@option{--get},
4500 @option{-x}) to read an archive which contains incomplete records, or
4501 one which has a blocking factor less than the one specified.
4505 @unnumberedsubsubsec Ignoring Blocks of Zeros
4507 @cindex End-of-archive blocks, ignoring
4508 @cindex Ignoring end-of-archive blocks
4509 @opindex ignore-zeros
4510 Normally, @command{tar} stops reading when it encounters a block of zeros
4511 between file entries (which usually indicates the end of the archive).
4512 @option{--ignore-zeros} (@option{-i}) allows @command{tar} to
4513 completely read an archive which contains a block of zeros before the
4514 end (i.e., a damaged archive, or one that was created by concatenating
4515 several archives together).
4517 The @option{--ignore-zeros} (@option{-i}) option is turned off by default because many
4518 versions of @command{tar} write garbage after the end-of-archive entry,
4519 since that part of the media is never supposed to be read. @GNUTAR{}
4520 does not write after the end of an archive, but seeks to
4521 maintain compatibility among archiving utilities.
4524 @item --ignore-zeros
4526 To ignore blocks of zeros (i.e., end-of-archive entries) which may be
4527 encountered while reading an archive. Use in conjunction with
4528 @option{--extract} or @option{--list}.
4532 @subsection Changing How @command{tar} Writes Files
4535 @FIXME{Introductory paragraph}
4538 * Dealing with Old Files::
4539 * Overwrite Old Files::
4541 * Keep Newer Files::
4543 * Recursive Unlink::
4544 * Data Modification Times::
4545 * Setting Access Permissions::
4546 * Directory Modification Times and Permissions::
4547 * Writing to Standard Output::
4548 * Writing to an External Program::
4552 @node Dealing with Old Files
4553 @unnumberedsubsubsec Options Controlling the Overwriting of Existing Files
4555 @xopindex{overwrite-dir, introduced}
4556 When extracting files, if @command{tar} discovers that the extracted
4557 file already exists, it normally replaces the file by removing it before
4558 extracting it, to prevent confusion in the presence of hard or symbolic
4559 links. (If the existing file is a symbolic link, it is removed, not
4560 followed.) However, if a directory cannot be removed because it is
4561 nonempty, @command{tar} normally overwrites its metadata (ownership,
4562 permission, etc.). The @option{--overwrite-dir} option enables this
4563 default behavior. To be more cautious and preserve the metadata of
4564 such a directory, use the @option{--no-overwrite-dir} option.
4566 @cindex Overwriting old files, prevention
4567 @xopindex{keep-old-files, introduced}
4568 To be even more cautious and prevent existing files from being replaced, use
4569 the @option{--keep-old-files} (@option{-k}) option. It causes @command{tar} to refuse
4570 to replace or update a file that already exists, i.e., a file with the
4571 same name as an archive member prevents extraction of that archive
4572 member. Instead, it reports an error.
4574 @xopindex{overwrite, introduced}
4575 To be more aggressive about altering existing files, use the
4576 @option{--overwrite} option. It causes @command{tar} to overwrite
4577 existing files and to follow existing symbolic links when extracting.
4579 @cindex Protecting old files
4580 Some people argue that @GNUTAR{} should not hesitate
4581 to overwrite files with other files when extracting. When extracting
4582 a @command{tar} archive, they expect to see a faithful copy of the
4583 state of the file system when the archive was created. It is debatable
4584 that this would always be a proper behavior. For example, suppose one
4585 has an archive in which @file{usr/local} is a link to
4586 @file{usr/local2}. Since then, maybe the site removed the link and
4587 renamed the whole hierarchy from @file{/usr/local2} to
4588 @file{/usr/local}. Such things happen all the time. I guess it would
4589 not be welcome at all that @GNUTAR{} removes the
4590 whole hierarchy just to make room for the link to be reinstated
4591 (unless it @emph{also} simultaneously restores the full
4592 @file{/usr/local2}, of course!) @GNUTAR{} is indeed
4593 able to remove a whole hierarchy to reestablish a symbolic link, for
4594 example, but @emph{only if} @option{--recursive-unlink} is specified
4595 to allow this behavior. In any case, single files are silently
4598 @xopindex{unlink-first, introduced}
4599 Finally, the @option{--unlink-first} (@option{-U}) option can improve performance in
4600 some cases by causing @command{tar} to remove files unconditionally
4601 before extracting them.
4603 @node Overwrite Old Files
4604 @unnumberedsubsubsec Overwrite Old Files
4609 Overwrite existing files and directory metadata when extracting files
4612 This causes @command{tar} to write extracted files into the file system without
4613 regard to the files already on the system; i.e., files with the same
4614 names as archive members are overwritten when the archive is extracted.
4615 It also causes @command{tar} to extract the ownership, permissions,
4616 and time stamps onto any preexisting files or directories.
4617 If the name of a corresponding file name is a symbolic link, the file
4618 pointed to by the symbolic link will be overwritten instead of the
4619 symbolic link itself (if this is possible). Moreover, special devices,
4620 empty directories and even symbolic links are automatically removed if
4621 they are in the way of extraction.
4623 Be careful when using the @option{--overwrite} option, particularly when
4624 combined with the @option{--absolute-names} (@option{-P}) option, as this combination
4625 can change the contents, ownership or permissions of any file on your
4626 system. Also, many systems do not take kindly to overwriting files that
4627 are currently being executed.
4629 @opindex overwrite-dir
4630 @item --overwrite-dir
4631 Overwrite the metadata of directories when extracting files from an
4632 archive, but remove other files before extracting.
4635 @node Keep Old Files
4636 @unnumberedsubsubsec Keep Old Files
4639 @opindex keep-old-files
4640 @item --keep-old-files
4642 Do not replace existing files from archive. The
4643 @option{--keep-old-files} (@option{-k}) option prevents @command{tar}
4644 from replacing existing files with files with the same name from the
4645 archive. The @option{--keep-old-files} option is meaningless with
4646 @option{--list} (@option{-t}). Prevents @command{tar} from replacing
4647 files in the file system during extraction.
4650 @node Keep Newer Files
4651 @unnumberedsubsubsec Keep Newer Files
4654 @opindex keep-newer-files
4655 @item --keep-newer-files
4656 Do not replace existing files that are newer than their archive
4657 copies. This option is meaningless with @option{--list} (@option{-t}).
4661 @unnumberedsubsubsec Unlink First
4664 @opindex unlink-first
4665 @item --unlink-first
4667 Remove files before extracting over them.
4668 This can make @command{tar} run a bit faster if you know in advance
4669 that the extracted files all need to be removed. Normally this option
4670 slows @command{tar} down slightly, so it is disabled by default.
4673 @node Recursive Unlink
4674 @unnumberedsubsubsec Recursive Unlink
4677 @opindex recursive-unlink
4678 @item --recursive-unlink
4679 When this option is specified, try removing files and directory hierarchies
4680 before extracting over them. @emph{This is a dangerous option!}
4683 If you specify the @option{--recursive-unlink} option,
4684 @command{tar} removes @emph{anything} that keeps you from extracting a file
4685 as far as current permissions will allow it. This could include removal
4686 of the contents of a full directory hierarchy.
4688 @node Data Modification Times
4689 @unnumberedsubsubsec Setting Data Modification Times
4691 @cindex Data modification times of extracted files
4692 @cindex Modification times of extracted files
4693 Normally, @command{tar} sets the data modification times of extracted
4694 files to the corresponding times recorded for the files in the archive, but
4695 limits the permissions of extracted files by the current @code{umask}
4698 To set the data modification times of extracted files to the time when
4699 the files were extracted, use the @option{--touch} (@option{-m}) option in
4700 conjunction with @option{--extract} (@option{--get}, @option{-x}).
4706 Sets the data modification time of extracted archive members to the time
4707 they were extracted, not the time recorded for them in the archive.
4708 Use in conjunction with @option{--extract} (@option{--get}, @option{-x}).
4711 @node Setting Access Permissions
4712 @unnumberedsubsubsec Setting Access Permissions
4714 @cindex Permissions of extracted files
4715 @cindex Modes of extracted files
4716 To set the modes (access permissions) of extracted files to those
4717 recorded for those files in the archive, use @option{--same-permissions}
4718 in conjunction with the @option{--extract} (@option{--get},
4719 @option{-x}) operation.
4722 @opindex preserve-permissions
4723 @opindex same-permissions
4724 @item --preserve-permissions
4725 @itemx --same-permissions
4726 @c @itemx --ignore-umask
4728 Set modes of extracted archive members to those recorded in the
4729 archive, instead of current umask settings. Use in conjunction with
4730 @option{--extract} (@option{--get}, @option{-x}).
4733 @node Directory Modification Times and Permissions
4734 @unnumberedsubsubsec Directory Modification Times and Permissions
4736 After successfully extracting a file member, @GNUTAR{} normally
4737 restores its permissions and modification times, as described in the
4738 previous sections. This cannot be done for directories, because
4739 after extracting a directory @command{tar} will almost certainly
4740 extract files into that directory and this will cause the directory
4741 modification time to be updated. Moreover, restoring that directory
4742 permissions may not permit file creation within it. Thus, restoring
4743 directory permissions and modification times must be delayed at least
4744 until all files have been extracted into that directory. @GNUTAR{}
4745 restores directories using the following approach.
4747 The extracted directories are created with the mode specified in the
4748 archive, as modified by the umask of the user, which gives sufficient
4749 permissions to allow file creation. The meta-information about the
4750 directory is recorded in the temporary list of directories. When
4751 preparing to extract next archive member, @GNUTAR{} checks if the
4752 directory prefix of this file contains the remembered directory. If
4753 it does not, the program assumes that all files have been extracted
4754 into that directory, restores its modification time and permissions
4755 and removes its entry from the internal list. This approach allows
4756 to correctly restore directory meta-information in the majority of
4757 cases, while keeping memory requirements sufficiently small. It is
4758 based on the fact, that most @command{tar} archives use the predefined
4759 order of members: first the directory, then all the files and
4760 subdirectories in that directory.
4762 However, this is not always true. The most important exception are
4763 incremental archives (@pxref{Incremental Dumps}). The member order in
4764 an incremental archive is reversed: first all directory members are
4765 stored, followed by other (non-directory) members. So, when extracting
4766 from incremental archives, @GNUTAR{} alters the above procedure. It
4767 remembers all restored directories, and restores their meta-data
4768 only after the entire archive has been processed. Notice, that you do
4769 not need to specify any special options for that, as @GNUTAR{}
4770 automatically detects archives in incremental format.
4772 There may be cases, when such processing is required for normal archives
4773 too. Consider the following example:
4777 $ @kbd{tar --no-recursion -cvf archive \
4778 foo foo/file1 bar bar/file foo/file2}
4787 During the normal operation, after encountering @file{bar}
4788 @GNUTAR{} will assume that all files from the directory @file{foo}
4789 were already extracted and will therefore restore its timestamp and
4790 permission bits. However, after extracting @file{foo/file2} the
4791 directory timestamp will be offset again.
4793 To correctly restore directory meta-information in such cases, use
4794 @option{delay-directory-restore} command line option:
4797 @opindex delay-directory-restore
4798 @item --delay-directory-restore
4799 Delays restoring of the modification times and permissions of extracted
4800 directories until the end of extraction. This way, correct
4801 meta-information is restored even if the archive has unusual member
4804 @opindex no-delay-directory-restore
4805 @item --no-delay-directory-restore
4806 Cancel the effect of the previous @option{--delay-directory-restore}.
4807 Use this option if you have used @option{--delay-directory-restore} in
4808 @env{TAR_OPTIONS} variable (@pxref{TAR_OPTIONS}) and wish to
4809 temporarily disable it.
4812 @node Writing to Standard Output
4813 @unnumberedsubsubsec Writing to Standard Output
4815 @cindex Writing extracted files to standard output
4816 @cindex Standard output, writing extracted files to
4817 To write the extracted files to the standard output, instead of
4818 creating the files on the file system, use @option{--to-stdout} (@option{-O}) in
4819 conjunction with @option{--extract} (@option{--get}, @option{-x}). This option is useful if you are
4820 extracting files to send them through a pipe, and do not need to
4821 preserve them in the file system. If you extract multiple members,
4822 they appear on standard output concatenated, in the order they are
4823 found in the archive.
4829 Writes files to the standard output. Use only in conjunction with
4830 @option{--extract} (@option{--get}, @option{-x}). When this option is
4831 used, instead of creating the files specified, @command{tar} writes
4832 the contents of the files extracted to its standard output. This may
4833 be useful if you are only extracting the files in order to send them
4834 through a pipe. This option is meaningless with @option{--list}
4838 This can be useful, for example, if you have a tar archive containing
4839 a big file and don't want to store the file on disk before processing
4840 it. You can use a command like this:
4843 tar -xOzf foo.tgz bigfile | process
4846 or even like this if you want to process the concatenation of the files:
4849 tar -xOzf foo.tgz bigfile1 bigfile2 | process
4852 However, @option{--to-command} may be more convenient for use with
4853 multiple files. See the next section.
4855 @node Writing to an External Program
4856 @unnumberedsubsubsec Writing to an External Program
4858 You can instruct @command{tar} to send the contents of each extracted
4859 file to the standard input of an external program:
4863 @item --to-command=@var{command}
4864 Extract files and pipe their contents to the standard input of
4865 @var{command}. When this option is used, instead of creating the
4866 files specified, @command{tar} invokes @var{command} and pipes the
4867 contents of the files to its standard output. @var{Command} may
4868 contain command line arguments. The program is executed via
4869 @code{sh -c}. Notice, that @var{command} is executed once for each regular file
4870 extracted. Non-regular files (directories, etc.) are ignored when this
4874 The command can obtain the information about the file it processes
4875 from the following environment variables:
4878 @vrindex TAR_FILETYPE, to-command environment
4880 Type of the file. It is a single letter with the following meaning:
4882 @multitable @columnfractions 0.10 0.90
4883 @item f @tab Regular file
4884 @item d @tab Directory
4885 @item l @tab Symbolic link
4886 @item h @tab Hard link
4887 @item b @tab Block device
4888 @item c @tab Character device
4891 Currently only regular files are supported.
4893 @vrindex TAR_MODE, to-command environment
4895 File mode, an octal number.
4897 @vrindex TAR_FILENAME, to-command environment
4899 The name of the file.
4901 @vrindex TAR_REALNAME, to-command environment
4903 Name of the file as stored in the archive.
4905 @vrindex TAR_UNAME, to-command environment
4907 Name of the file owner.
4909 @vrindex TAR_GNAME, to-command environment
4911 Name of the file owner group.
4913 @vrindex TAR_ATIME, to-command environment
4915 Time of last access. It is a decimal number, representing seconds
4916 since the epoch. If the archive provides times with nanosecond
4917 precision, the nanoseconds are appended to the timestamp after a
4920 @vrindex TAR_MTIME, to-command environment
4922 Time of last modification.
4924 @vrindex TAR_CTIME, to-command environment
4926 Time of last status change.
4928 @vrindex TAR_SIZE, to-command environment
4932 @vrindex TAR_UID, to-command environment
4934 UID of the file owner.
4936 @vrindex TAR_GID, to-command environment
4938 GID of the file owner.
4941 In addition to these variables, @env{TAR_VERSION} contains the
4942 @GNUTAR{} version number.
4944 If @var{command} exits with a non-0 status, @command{tar} will print
4945 an error message similar to the following:
4948 tar: 2345: Child returned status 1
4951 Here, @samp{2345} is the PID of the finished process.
4953 If this behavior is not wanted, use @option{--ignore-command-error}:
4956 @opindex ignore-command-error
4957 @item --ignore-command-error
4958 Ignore exit codes of subprocesses. Notice that if the program
4959 exits on signal or otherwise terminates abnormally, the error message
4960 will be printed even if this option is used.
4962 @opindex no-ignore-command-error
4963 @item --no-ignore-command-error
4964 Cancel the effect of any previous @option{--ignore-command-error}
4965 option. This option is useful if you have set
4966 @option{--ignore-command-error} in @env{TAR_OPTIONS}
4967 (@pxref{TAR_OPTIONS}) and wish to temporarily cancel it.
4971 @unnumberedsubsubsec Removing Files
4973 @FIXME{The section is too terse. Something more to add? An example,
4977 @opindex remove-files
4978 @item --remove-files
4979 Remove files after adding them to the archive.
4983 @subsection Coping with Scarce Resources
4986 @cindex Small memory
4987 @cindex Running out of space
4995 @unnumberedsubsubsec Starting File
4998 @opindex starting-file
4999 @item --starting-file=@var{name}
5000 @itemx -K @var{name}
5001 Starts an operation in the middle of an archive. Use in conjunction
5002 with @option{--extract} (@option{--get}, @option{-x}) or @option{--list} (@option{-t}).
5005 @cindex Middle of the archive, starting in the
5006 If a previous attempt to extract files failed due to lack of disk
5007 space, you can use @option{--starting-file=@var{name}} (@option{-K
5008 @var{name}}) to start extracting only after member @var{name} of the
5009 archive. This assumes, of course, that there is now free space, or
5010 that you are now extracting into a different file system. (You could
5011 also choose to suspend @command{tar}, remove unnecessary files from
5012 the file system, and then restart the same @command{tar} operation.
5013 In this case, @option{--starting-file} is not necessary.
5014 @xref{Incremental Dumps}, @xref{interactive}, and @ref{exclude}.)
5017 @unnumberedsubsubsec Same Order
5020 @cindex Large lists of file names on small machines
5022 @opindex preserve-order
5024 @itemx --preserve-order
5026 To process large lists of file names on machines with small amounts of
5027 memory. Use in conjunction with @option{--compare} (@option{--diff},
5028 @option{-d}), @option{--list} (@option{-t}) or @option{--extract}
5029 (@option{--get}, @option{-x}).
5032 The @option{--same-order} (@option{--preserve-order}, @option{-s}) option tells @command{tar} that the list of file
5033 names to be listed or extracted is sorted in the same order as the
5034 files in the archive. This allows a large list of names to be used,
5035 even on a small machine that would not otherwise be able to hold all
5036 the names in memory at the same time. Such a sorted list can easily be
5037 created by running @samp{tar -t} on the archive and editing its output.
5039 This option is probably never needed on modern computer systems.
5042 @section Backup options
5044 @cindex backup options
5046 @GNUTAR{} offers options for making backups of files
5047 before writing new versions. These options control the details of
5048 these backups. They may apply to the archive itself before it is
5049 created or rewritten, as well as individual extracted members. Other
5050 @acronym{GNU} programs (@command{cp}, @command{install}, @command{ln},
5051 and @command{mv}, for example) offer similar options.
5053 Backup options may prove unexpectedly useful when extracting archives
5054 containing many members having identical name, or when extracting archives
5055 on systems having file name limitations, making different members appear
5056 has having similar names through the side-effect of name truncation.
5057 (This is true only if we have a good scheme for truncated backup names,
5058 which I'm not sure at all: I suspect work is needed in this area.)
5059 When any existing file is backed up before being overwritten by extraction,
5060 then clashing files are automatically be renamed to be unique, and the
5061 true name is kept for only the last file of a series of clashing files.
5062 By using verbose mode, users may track exactly what happens.
5064 At the detail level, some decisions are still experimental, and may
5065 change in the future, we are waiting comments from our users. So, please
5066 do not learn to depend blindly on the details of the backup features.
5067 For example, currently, directories themselves are never renamed through
5068 using these options, so, extracting a file over a directory still has
5069 good chances to fail. Also, backup options apply to created archives,
5070 not only to extracted members. For created archives, backups will not
5071 be attempted when the archive is a block or character device, or when it
5072 refers to a remote file.
5074 For the sake of simplicity and efficiency, backups are made by renaming old
5075 files prior to creation or extraction, and not by copying. The original
5076 name is restored if the file creation fails. If a failure occurs after a
5077 partial extraction of a file, both the backup and the partially extracted
5081 @item --backup[=@var{method}]
5083 @vindex VERSION_CONTROL
5085 Back up files that are about to be overwritten or removed.
5086 Without this option, the original versions are destroyed.
5088 Use @var{method} to determine the type of backups made.
5089 If @var{method} is not specified, use the value of the @env{VERSION_CONTROL}
5090 environment variable. And if @env{VERSION_CONTROL} is not set,
5091 use the @samp{existing} method.
5093 @vindex version-control @r{Emacs variable}
5094 This option corresponds to the Emacs variable @samp{version-control};
5095 the same values for @var{method} are accepted as in Emacs. This option
5096 also allows more descriptive names. The valid @var{method}s are:
5101 @cindex numbered @r{backup method}
5102 Always make numbered backups.
5106 @cindex existing @r{backup method}
5107 Make numbered backups of files that already have them, simple backups
5112 @cindex simple @r{backup method}
5113 Always make simple backups.
5117 @item --suffix=@var{suffix}
5119 @cindex backup suffix
5120 @vindex SIMPLE_BACKUP_SUFFIX
5121 Append @var{suffix} to each backup file made with @option{--backup}. If this
5122 option is not specified, the value of the @env{SIMPLE_BACKUP_SUFFIX}
5123 environment variable is used. And if @env{SIMPLE_BACKUP_SUFFIX} is not
5124 set, the default is @samp{~}, just as in Emacs.
5129 @section Notable @command{tar} Usages
5132 @FIXME{Using Unix file linking capability to recreate directory
5133 structures---linking files into one subdirectory and then
5134 @command{tar}ring that directory.}
5136 @FIXME{Nice hairy example using absolute-names, newer, etc.}
5139 You can easily use archive files to transport a group of files from
5140 one system to another: put all relevant files into an archive on one
5141 computer system, transfer the archive to another system, and extract
5142 the contents there. The basic transfer medium might be magnetic tape,
5143 Internet FTP, or even electronic mail (though you must encode the
5144 archive with @command{uuencode} in order to transport it properly by
5145 mail). Both machines do not have to use the same operating system, as
5146 long as they both support the @command{tar} program.
5148 For example, here is how you might copy a directory's contents from
5149 one disk to another, while preserving the dates, modes, owners and
5150 link-structure of all the files therein. In this case, the transfer
5151 medium is a @dfn{pipe}, which is one a Unix redirection mechanism:
5154 $ @kbd{(cd sourcedir; tar -cf - .) | (cd targetdir; tar -xf -)}
5158 You can avoid subshells by using @option{-C} option:
5161 $ @kbd{tar -C sourcedir -cf - . | tar -C targetdir -xf -}
5165 The command also works using short option forms:
5168 $ @kbd{(cd sourcedir; tar --create --file=- . ) \
5169 | (cd targetdir; tar --extract --file=-)}
5171 $ @kbd{tar --directory sourcedir --create --file=- . ) \
5172 | tar --directory targetdir --extract --file=-}
5176 This is one of the easiest methods to transfer a @command{tar} archive.
5179 @section Looking Ahead: The Rest of this Manual
5181 You have now seen how to use all eight of the operations available to
5182 @command{tar}, and a number of the possible options. The next chapter
5183 explains how to choose and change file and archive names, how to use
5184 files to store names of other files which you can then call as
5185 arguments to @command{tar} (this can help you save time if you expect to
5186 archive the same list of files a number of times), and so forth.
5187 @FIXME{in case it's not obvious, i'm making this up in some sense
5188 based on my limited memory of what the next chapter *really* does. i
5189 just wanted to flesh out this final section a little bit so i'd
5190 remember to stick it in here. :-)}
5192 If there are too many files to conveniently list on the command line,
5193 you can list the names in a file, and @command{tar} will read that file.
5196 There are various ways of causing @command{tar} to skip over some files,
5197 and not archive them. @xref{Choosing}.
5200 @chapter Performing Backups and Restoring Files
5203 @GNUTAR{} is distributed along with the scripts
5204 which the Free Software Foundation uses for performing backups. There
5205 is no corresponding scripts available yet for doing restoration of
5206 files. Even if there is a good chance those scripts may be satisfying
5207 to you, they are not the only scripts or methods available for doing
5208 backups and restore. You may well create your own, or use more
5209 sophisticated packages dedicated to that purpose.
5211 Some users are enthusiastic about @code{Amanda} (The Advanced Maryland
5212 Automatic Network Disk Archiver), a backup system developed by James
5213 da Silva @file{jds@@cs.umd.edu} and available on many Unix systems.
5214 This is free software, and it is available at these places:
5217 http://www.cs.umd.edu/projects/amanda/amanda.html
5218 ftp://ftp.cs.umd.edu/pub/amanda
5223 Here is a possible plan for a future documentation about the backuping
5224 scripts which are provided within the @GNUTAR{}
5230 @item what are dumps
5231 @item different levels of dumps
5233 @item full dump = dump everything
5234 @item level 1, level 2 dumps etc
5235 A level @var{n} dump dumps everything changed since the last level
5238 @item how to use scripts for dumps (ie, the concept)
5240 @item scripts to run after editing backup specs (details)
5242 @item Backup Specs, what is it.
5244 @item how to customize
5245 @item actual text of script [/sp/dump/backup-specs]
5249 @item rsh doesn't work
5250 @item rtape isn't installed
5253 @item the @option{--incremental} option of tar
5256 @item write protection
5257 @item types of media, different sizes and types, useful for different things
5258 @item files and tape marks
5259 one tape mark between files, two at end.
5260 @item positioning the tape
5261 MT writes two at end of write,
5262 backspaces over one when writing again.
5268 This chapter documents both the provided shell scripts and @command{tar}
5269 options which are more specific to usage as a backup tool.
5271 To @dfn{back up} a file system means to create archives that contain
5272 all the files in that file system. Those archives can then be used to
5273 restore any or all of those files (for instance if a disk crashes or a
5274 file is accidentally deleted). File system @dfn{backups} are also
5278 * Full Dumps:: Using @command{tar} to Perform Full Dumps
5279 * Incremental Dumps:: Using @command{tar} to Perform Incremental Dumps
5280 * Backup Levels:: Levels of Backups
5281 * Backup Parameters:: Setting Parameters for Backups and Restoration
5282 * Scripted Backups:: Using the Backup Scripts
5283 * Scripted Restoration:: Using the Restore Script
5287 @section Using @command{tar} to Perform Full Dumps
5293 @cindex corrupted archives
5294 Full dumps should only be made when no other people or programs
5295 are modifying files in the file system. If files are modified while
5296 @command{tar} is making the backup, they may not be stored properly in
5297 the archive, in which case you won't be able to restore them if you
5298 have to. (Files not being modified are written with no trouble, and do
5299 not corrupt the entire archive.)
5301 You will want to use the @option{--label=@var{archive-label}}
5302 (@option{-V @var{archive-label}}) option to give the archive a
5303 volume label, so you can tell what this archive is even if the label
5304 falls off the tape, or anything like that.
5306 Unless the file system you are dumping is guaranteed to fit on
5307 one volume, you will need to use the @option{--multi-volume} (@option{-M}) option.
5308 Make sure you have enough tapes on hand to complete the backup.
5310 If you want to dump each file system separately you will need to use
5311 the @option{--one-file-system} option to prevent
5312 @command{tar} from crossing file system boundaries when storing
5315 The @option{--incremental} (@option{-G}) (@pxref{Incremental Dumps})
5316 option is not needed, since this is a complete copy of everything in
5317 the file system, and a full restore from this backup would only be
5318 done onto a completely
5321 Unless you are in a hurry, and trust the @command{tar} program (and your
5322 tapes), it is a good idea to use the @option{--verify} (@option{-W})
5323 option, to make sure your files really made it onto the dump properly.
5324 This will also detect cases where the file was modified while (or just
5325 after) it was being archived. Not all media (notably cartridge tapes)
5326 are capable of being verified, unfortunately.
5328 @node Incremental Dumps
5329 @section Using @command{tar} to Perform Incremental Dumps
5331 @dfn{Incremental backup} is a special form of @GNUTAR{} archive that
5332 stores additional metadata so that exact state of the file system
5333 can be restored when extracting the archive.
5335 @GNUTAR{} currently offers two options for handling incremental
5336 backups: @option{--listed-incremental=@var{snapshot-file}} (@option{-g
5337 @var{snapshot-file}}) and @option{--incremental} (@option{-G}).
5339 @opindex listed-incremental
5340 The option @option{--listed-incremental} instructs tar to operate on
5341 an incremental archive with additional metadata stored in a standalone
5342 file, called a @dfn{snapshot file}. The purpose of this file is to help
5343 determine which files have been changed, added or deleted since the
5344 last backup, so that the next incremental backup will contain only
5345 modified files. The name of the snapshot file is given as an argument
5349 @item --listed-incremental=@var{file}
5350 @itemx -g @var{file}
5351 Handle incremental backups with snapshot data in @var{file}.
5354 To create an incremental backup, you would use
5355 @option{--listed-incremental} together with @option{--create}
5356 (@pxref{create}). For example:
5359 $ @kbd{tar --create \
5360 --file=archive.1.tar \
5361 --listed-incremental=/var/log/usr.snar \
5365 This will create in @file{archive.1.tar} an incremental backup of
5366 the @file{/usr} file system, storing additional metadata in the file
5367 @file{/var/log/usr.snar}. If this file does not exist, it will be
5368 created. The created archive will then be a @dfn{level 0 backup};
5369 please see the next section for more on backup levels.
5371 Otherwise, if the file @file{/var/log/usr.snar} exists, it
5372 determines which files are modified. In this case only these files will be
5373 stored in the archive. Suppose, for example, that after running the
5374 above command, you delete file @file{/usr/doc/old} and create
5375 directory @file{/usr/local/db} with the following contents:
5378 $ @kbd{ls /usr/local/db}
5383 Some time later you create another incremental backup. You will
5387 $ @kbd{tar --create \
5388 --file=archive.2.tar \
5389 --listed-incremental=/var/log/usr.snar \
5391 tar: usr/local/db: Directory is new
5398 The created archive @file{archive.2.tar} will contain only these
5399 three members. This archive is called a @dfn{level 1 backup}. Notice
5400 that @file{/var/log/usr.snar} will be updated with the new data, so if
5401 you plan to create more @samp{level 1} backups, it is necessary to
5402 create a working copy of the snapshot file before running
5403 @command{tar}. The above example will then be modified as follows:
5406 $ @kbd{cp /var/log/usr.snar /var/log/usr.snar-1}
5407 $ @kbd{tar --create \
5408 --file=archive.2.tar \
5409 --listed-incremental=/var/log/usr.snar-1 \
5413 Incremental dumps depend crucially on time stamps, so the results are
5414 unreliable if you modify a file's time stamps during dumping (e.g.,
5415 with the @option{--atime-preserve=replace} option), or if you set the clock
5418 @cindex Device numbers, using in incremental backups
5419 Metadata stored in snapshot files include device numbers, which,
5420 obviously is supposed to be a non-volatile value. However, it turns
5421 out that NFS devices have undependable values when an automounter
5422 gets in the picture. This can lead to a great deal of spurious
5423 redumping in incremental dumps, so it is somewhat useless to compare
5424 two NFS devices numbers over time. The solution implemented currently
5425 is to considers all NFS devices as being equal when it comes to
5426 comparing directories; this is fairly gross, but there does not seem
5427 to be a better way to go.
5429 If you are using the @i{Linux} kernel, the device numbers can also
5430 change when upgrading to some newer versions of the kernel. This can
5431 cause the next backup to be full backup on the affected filesystems.
5432 @xref{Fixing Snapshot Files}, for the information on how to handle this case.
5434 Note that incremental archives use @command{tar} extensions and may
5435 not be readable by non-@acronym{GNU} versions of the @command{tar} program.
5437 @xopindex{listed-incremental, using with @option{--extract}}
5438 @xopindex{extract, using with @option{--listed-incremental}}
5439 To extract from the incremental dumps, use
5440 @option{--listed-incremental} together with @option{--extract}
5441 option (@pxref{extracting files}). In this case, @command{tar} does
5442 not need to access snapshot file, since all the data necessary for
5443 extraction are stored in the archive itself. So, when extracting, you
5444 can give whatever argument to @option{--listed-incremental}, the usual
5445 practice is to use @option{--listed-incremental=/dev/null}.
5446 Alternatively, you can use @option{--incremental}, which needs no
5447 arguments. In general, @option{--incremental} (@option{-G}) can be
5448 used as a shortcut for @option{--listed-incremental} when listing or
5449 extracting incremental backups (for more information, regarding this
5450 option, @pxref{incremental-op}).
5452 When extracting from the incremental backup @GNUTAR{} attempts to
5453 restore the exact state the file system had when the archive was
5454 created. In particular, it will @emph{delete} those files in the file
5455 system that did not exist in their directories when the archive was
5456 created. If you have created several levels of incremental files,
5457 then in order to restore the exact contents the file system had when
5458 the last level was created, you will need to restore from all backups
5459 in turn. Continuing our example, to restore the state of @file{/usr}
5460 file system, one would do@footnote{Notice, that since both archives
5461 were created without @option{-P} option (@pxref{absolute}), these
5462 commands should be run from the root file system.}:
5465 $ @kbd{tar --extract \
5466 --listed-incremental=/dev/null \
5467 --file archive.1.tar}
5468 $ @kbd{tar --extract \
5469 --listed-incremental=/dev/null \
5470 --file archive.2.tar}
5473 To list the contents of an incremental archive, use @option{--list}
5474 (@pxref{list}), as usual. To obtain more information about the
5475 archive, use @option{--listed-incremental} or @option{--incremental}
5476 combined with two @option{--verbose} options@footnote{Two
5477 @option{--verbose} options were selected to avoid breaking usual
5478 verbose listing output (@option{--list --verbose}) when using in
5481 @xopindex{incremental, using with @option{--list}}
5482 @xopindex{listed-incremental, using with @option{--list}}
5483 @xopindex{list, using with @option{--incremental}}
5484 @xopindex{list, using with @option{--listed-incremental}}
5485 Versions of @GNUTAR{} up to 1.15.1 used to dump verbatim binary
5486 contents of the DUMPDIR header (with terminating nulls) when
5487 @option{--incremental} or @option{--listed-incremental} option was
5488 given, no matter what the verbosity level. This behavior, and,
5489 especially, the binary output it produced were considered inconvenient
5490 and were changed in version 1.16}:
5493 @kbd{tar --list --incremental --verbose --verbose archive.tar}
5496 This command will print, for each directory in the archive, the list
5497 of files in that directory at the time the archive was created. This
5498 information is put out in a format which is both human-readable and
5499 unambiguous for a program: each file name is printed as
5506 where @var{x} is a letter describing the status of the file: @samp{Y}
5507 if the file is present in the archive, @samp{N} if the file is not
5508 included in the archive, or a @samp{D} if the file is a directory (and
5509 is included in the archive). @xref{Dumpdir}, for the detailed
5510 description of dumpdirs and status codes. Each such
5511 line is terminated by a newline character. The last line is followed
5512 by an additional newline to indicate the end of the data.
5514 @anchor{incremental-op}The option @option{--incremental} (@option{-G})
5515 gives the same behavior as @option{--listed-incremental} when used
5516 with @option{--list} and @option{--extract} options. When used with
5517 @option{--create} option, it creates an incremental archive without
5518 creating snapshot file. Thus, it is impossible to create several
5519 levels of incremental backups with @option{--incremental} option.
5522 @section Levels of Backups
5524 An archive containing all the files in the file system is called a
5525 @dfn{full backup} or @dfn{full dump}. You could insure your data by
5526 creating a full dump every day. This strategy, however, would waste a
5527 substantial amount of archive media and user time, as unchanged files
5528 are daily re-archived.
5530 It is more efficient to do a full dump only occasionally. To back up
5531 files between full dumps, you can use @dfn{incremental dumps}. A @dfn{level
5532 one} dump archives all the files that have changed since the last full
5535 A typical dump strategy would be to perform a full dump once a week,
5536 and a level one dump once a day. This means some versions of files
5537 will in fact be archived more than once, but this dump strategy makes
5538 it possible to restore a file system to within one day of accuracy by
5539 only extracting two archives---the last weekly (full) dump and the
5540 last daily (level one) dump. The only information lost would be in
5541 files changed or created since the last daily backup. (Doing dumps
5542 more than once a day is usually not worth the trouble).
5544 @GNUTAR{} comes with scripts you can use to do full
5545 and level-one (actually, even level-two and so on) dumps. Using
5546 scripts (shell programs) to perform backups and restoration is a
5547 convenient and reliable alternative to typing out file name lists
5548 and @command{tar} commands by hand.
5550 Before you use these scripts, you need to edit the file
5551 @file{backup-specs}, which specifies parameters used by the backup
5552 scripts and by the restore script. This file is usually located
5553 in @file{/etc/backup} directory. @xref{Backup Parameters}, for its
5554 detailed description. Once the backup parameters are set, you can
5555 perform backups or restoration by running the appropriate script.
5557 The name of the backup script is @code{backup}. The name of the
5558 restore script is @code{restore}. The following sections describe
5559 their use in detail.
5561 @emph{Please Note:} The backup and restoration scripts are
5562 designed to be used together. While it is possible to restore files by
5563 hand from an archive which was created using a backup script, and to create
5564 an archive by hand which could then be extracted using the restore script,
5565 it is easier to use the scripts. @xref{Incremental Dumps}, before
5566 making such an attempt.
5568 @node Backup Parameters
5569 @section Setting Parameters for Backups and Restoration
5571 The file @file{backup-specs} specifies backup parameters for the
5572 backup and restoration scripts provided with @command{tar}. You must
5573 edit @file{backup-specs} to fit your system configuration and schedule
5574 before using these scripts.
5576 Syntactically, @file{backup-specs} is a shell script, containing
5577 mainly variable assignments. However, any valid shell construct
5578 is allowed in this file. Particularly, you may wish to define
5579 functions within that script (e.g., see @code{RESTORE_BEGIN} below).
5580 For more information about shell script syntax, please refer to
5581 @url{http://www.opengroup.org/onlinepubs/009695399/utilities/xcu_chap02.html#ta
5582 g_02, the definition of the Shell Command Language}. See also
5583 @ref{Top,,Bash Features,bashref,Bash Reference Manual}.
5585 The shell variables controlling behavior of @code{backup} and
5586 @code{restore} are described in the following subsections.
5589 * General-Purpose Variables::
5590 * Magnetic Tape Control::
5592 * backup-specs example:: An Example Text of @file{Backup-specs}
5595 @node General-Purpose Variables
5596 @subsection General-Purpose Variables
5598 @defvr {Backup variable} ADMINISTRATOR
5599 The user name of the backup administrator. @code{Backup} scripts
5600 sends a backup report to this address.
5603 @defvr {Backup variable} BACKUP_HOUR
5604 The hour at which the backups are done. This can be a number from 0
5605 to 23, or the time specification in form @var{hours}:@var{minutes},
5606 or the string @samp{now}.
5608 This variable is used by @code{backup}. Its value may be overridden
5609 using @option{--time} option (@pxref{Scripted Backups}).
5612 @defvr {Backup variable} TAPE_FILE
5614 The device @command{tar} writes the archive to. If @var{TAPE_FILE}
5615 is a remote archive (@pxref{remote-dev}), backup script will suppose
5616 that your @command{mt} is able to access remote devices. If @var{RSH}
5617 (@pxref{RSH}) is set, @option{--rsh-command} option will be added to
5618 invocations of @command{mt}.
5621 @defvr {Backup variable} BLOCKING
5623 The blocking factor @command{tar} will use when writing the dump archive.
5624 @xref{Blocking Factor}.
5627 @defvr {Backup variable} BACKUP_DIRS
5629 A list of file systems to be dumped (for @code{backup}), or restored
5630 (for @code{restore}). You can include any directory
5631 name in the list --- subdirectories on that file system will be
5632 included, regardless of how they may look to other networked machines.
5633 Subdirectories on other file systems will be ignored.
5635 The host name specifies which host to run @command{tar} on, and should
5636 normally be the host that actually contains the file system. However,
5637 the host machine must have @GNUTAR{} installed, and
5638 must be able to access the directory containing the backup scripts and
5639 their support files using the same file name that is used on the
5640 machine where the scripts are run (i.e., what @command{pwd} will print
5641 when in that directory on that machine). If the host that contains
5642 the file system does not have this capability, you can specify another
5643 host as long as it can access the file system through NFS.
5645 If the list of file systems is very long you may wish to put it
5646 in a separate file. This file is usually named
5647 @file{/etc/backup/dirs}, but this name may be overridden in
5648 @file{backup-specs} using @code{DIRLIST} variable.
5651 @defvr {Backup variable} DIRLIST
5653 The name of the file that contains a list of file systems to backup
5654 or restore. By default it is @file{/etc/backup/dirs}.
5657 @defvr {Backup variable} BACKUP_FILES
5659 A list of individual files to be dumped (for @code{backup}), or restored
5660 (for @code{restore}). These should be accessible from the machine on
5661 which the backup script is run.
5663 If the list of file systems is very long you may wish to store it
5664 in a separate file. This file is usually named
5665 @file{/etc/backup/files}, but this name may be overridden in
5666 @file{backup-specs} using @code{FILELIST} variable.
5669 @defvr {Backup variable} FILELIST
5671 The name of the file that contains a list of individual files to backup
5672 or restore. By default it is @file{/etc/backup/files}.
5675 @defvr {Backup variable} MT
5677 Full file name of @command{mt} binary.
5680 @defvr {Backup variable} RSH
5682 Full file name of @command{rsh} binary or its equivalent. You may wish to
5683 set it to @code{ssh}, to improve security. In this case you will have
5684 to use public key authentication.
5687 @defvr {Backup variable} RSH_COMMAND
5689 Full file name of @command{rsh} binary on remote machines. This will
5690 be passed via @option{--rsh-command} option to the remote invocation
5694 @defvr {Backup variable} VOLNO_FILE
5696 Name of temporary file to hold volume numbers. This needs to be accessible
5697 by all the machines which have file systems to be dumped.
5700 @defvr {Backup variable} XLIST
5702 Name of @dfn{exclude file list}. An @dfn{exclude file list} is a file
5703 located on the remote machine and containing the list of files to
5704 be excluded from the backup. Exclude file lists are searched in
5705 /etc/tar-backup directory. A common use for exclude file lists
5706 is to exclude files containing security-sensitive information
5707 (e.g., @file{/etc/shadow} from backups).
5709 This variable affects only @code{backup}.
5712 @defvr {Backup variable} SLEEP_TIME
5714 Time to sleep between dumps of any two successive file systems
5716 This variable affects only @code{backup}.
5719 @defvr {Backup variable} DUMP_REMIND_SCRIPT
5721 Script to be run when it's time to insert a new tape in for the next
5722 volume. Administrators may want to tailor this script for their site.
5723 If this variable isn't set, @GNUTAR{} will display its built-in
5724 prompt, and will expect confirmation from the console. For the
5725 description of the default prompt, see @ref{change volume prompt}.
5729 @defvr {Backup variable} SLEEP_MESSAGE
5731 Message to display on the terminal while waiting for dump time. Usually
5732 this will just be some literal text.
5735 @defvr {Backup variable} TAR
5737 Full file name of the @GNUTAR{} executable. If this is not set, backup
5738 scripts will search @command{tar} in the current shell path.
5741 @node Magnetic Tape Control
5742 @subsection Magnetic Tape Control
5744 Backup scripts access tape device using special @dfn{hook functions}.
5745 These functions take a single argument -- the name of the tape
5746 device. Their names are kept in the following variables:
5748 @defvr {Backup variable} MT_BEGIN
5749 The name of @dfn{begin} function. This function is called before
5750 accessing the drive. By default it retensions the tape:
5756 mt -f "$1" retension
5761 @defvr {Backup variable} MT_REWIND
5762 The name of @dfn{rewind} function. The default definition is as
5775 @defvr {Backup variable} MT_OFFLINE
5776 The name of the function switching the tape off line. By default
5777 it is defined as follows:
5780 MT_OFFLINE=mt_offline
5788 @defvr {Backup variable} MT_STATUS
5789 The name of the function used to obtain the status of the archive device,
5790 including error count. Default definition:
5802 @subsection User Hooks
5804 @dfn{User hooks} are shell functions executed before and after
5805 each @command{tar} invocation. Thus, there are @dfn{backup
5806 hooks}, which are executed before and after dumping each file
5807 system, and @dfn{restore hooks}, executed before and
5808 after restoring a file system. Each user hook is a shell function
5809 taking four arguments:
5811 @deffn {User Hook Function} hook @var{level} @var{host} @var{fs} @var{fsname}
5816 Current backup or restore level.
5819 Name or IP address of the host machine being dumped or restored.
5822 Full file name of the file system being dumped or restored.
5825 File system name with directory separators replaced with colons. This
5826 is useful, e.g., for creating unique files.
5830 Following variables keep the names of user hook functions
5832 @defvr {Backup variable} DUMP_BEGIN
5833 Dump begin function. It is executed before dumping the file system.
5836 @defvr {Backup variable} DUMP_END
5837 Executed after dumping the file system.
5840 @defvr {Backup variable} RESTORE_BEGIN
5841 Executed before restoring the file system.
5844 @defvr {Backup variable} RESTORE_END
5845 Executed after restoring the file system.
5848 @node backup-specs example
5849 @subsection An Example Text of @file{Backup-specs}
5851 The following is an example of @file{backup-specs}:
5854 # site-specific parameters for file system backup.
5856 ADMINISTRATOR=friedman
5858 TAPE_FILE=/dev/nrsmt0
5860 # Use @code{ssh} instead of the less secure @code{rsh}
5862 RSH_COMMAND=/usr/bin/ssh
5864 # Override MT_STATUS function:
5870 # Disable MT_OFFLINE function
5887 apple-gunkies:/com/mailer/gnu
5888 apple-gunkies:/com/archive/gnu"
5890 BACKUP_FILES="/com/mailer/aliases /com/mailer/league*[a-z]"
5894 @node Scripted Backups
5895 @section Using the Backup Scripts
5897 The syntax for running a backup script is:
5900 backup --level=@var{level} --time=@var{time}
5903 The @option{level} option requests the dump level. Thus, to produce
5904 a full dump, specify @code{--level=0} (this is the default, so
5905 @option{--level} may be omitted if its value is @code{0}).
5906 @footnote{For backward compatibility, the @code{backup} will also
5907 try to deduce the requested dump level from the name of the
5908 script itself. If the name consists of a string @samp{level-}
5909 followed by a single decimal digit, that digit is taken as
5910 the dump level number. Thus, you may create a link from @code{backup}
5911 to @code{level-1} and then run @code{level-1} whenever you need to
5912 create a level one dump.}
5914 The @option{--time} option determines when should the backup be
5915 run. @var{Time} may take three forms:
5918 @item @var{hh}:@var{mm}
5920 The dump must be run at @var{hh} hours @var{mm} minutes.
5924 The dump must be run at @var{hh} hours
5928 The dump must be run immediately.
5931 You should start a script with a tape or disk mounted. Once you
5932 start a script, it prompts you for new tapes or disks as it
5933 needs them. Media volumes don't have to correspond to archive
5934 files --- a multi-volume archive can be started in the middle of a
5935 tape that already contains the end of another multi-volume archive.
5936 The @code{restore} script prompts for media by its archive volume,
5937 so to avoid an error message you should keep track of which tape
5938 (or disk) contains which volume of the archive (@pxref{Scripted
5941 The backup scripts write two files on the file system. The first is a
5942 record file in @file{/etc/tar-backup/}, which is used by the scripts
5943 to store and retrieve information about which files were dumped. This
5944 file is not meant to be read by humans, and should not be deleted by
5945 them. @xref{Snapshot Files}, for a more detailed explanation of this
5948 The second file is a log file containing the names of the file systems
5949 and files dumped, what time the backup was made, and any error
5950 messages that were generated, as well as how much space was left in
5951 the media volume after the last volume of the archive was written.
5952 You should check this log file after every backup. The file name is
5953 @file{log-@var{mm-dd-yyyy}-level-@var{n}}, where @var{mm-dd-yyyy}
5954 represents current date, and @var{n} represents current dump level number.
5956 The script also prints the name of each system being dumped to the
5959 Following is the full list of options accepted by @code{backup}
5963 @item -l @var{level}
5964 @itemx --level=@var{level}
5965 Do backup level @var{level} (default 0).
5969 Force backup even if today's log file already exists.
5971 @item -v[@var{level}]
5972 @itemx --verbose[=@var{level}]
5973 Set verbosity level. The higher the level is, the more debugging
5974 information will be output during execution. Default @var{level}
5975 is 100, which means the highest debugging level.
5977 @item -t @var{start-time}
5978 @itemx --time=@var{start-time}
5979 Wait till @var{time}, then do backup.
5983 Display short help message and exit.
5987 Display information about the program's name, version, origin and legal
5988 status, all on standard output, and then exit successfully.
5992 @node Scripted Restoration
5993 @section Using the Restore Script
5995 To restore files that were archived using a scripted backup, use the
5996 @code{restore} script. Its usage is quite straightforward. In the
5997 simplest form, invoke @code{restore --all}, it will
5998 then restore all the file systems and files specified in
5999 @file{backup-specs} (@pxref{General-Purpose Variables,BACKUP_DIRS}).
6001 You may select the file systems (and/or files) to restore by
6002 giving @code{restore} list of @dfn{patterns} in its command
6003 line. For example, running
6010 will restore all file systems on the machine @samp{albert}. A more
6011 complicated example:
6014 restore 'albert:*' '*:/var'
6018 This command will restore all file systems on the machine @samp{albert}
6019 as well as @file{/var} file system on all machines.
6021 By default @code{restore} will start restoring files from the lowest
6022 available dump level (usually zero) and will continue through
6023 all available dump levels. There may be situations where such a
6024 thorough restore is not necessary. For example, you may wish to
6025 restore only files from the recent level one backup. To do so,
6026 use @option{--level} option, as shown in the example below:
6032 The full list of options accepted by @code{restore} follows:
6037 Restore all file systems and files specified in @file{backup-specs}
6039 @item -l @var{level}
6040 @itemx --level=@var{level}
6041 Start restoring from the given backup level, instead of the default 0.
6043 @item -v[@var{level}]
6044 @itemx --verbose[=@var{level}]
6045 Set verbosity level. The higher the level is, the more debugging
6046 information will be output during execution. Default @var{level}
6047 is 100, which means the highest debugging level.
6051 Display short help message and exit.
6055 Display information about the program's name, version, origin and legal
6056 status, all on standard output, and then exit successfully.
6059 You should start the restore script with the media containing the
6060 first volume of the archive mounted. The script will prompt for other
6061 volumes as they are needed. If the archive is on tape, you don't need
6062 to rewind the tape to to its beginning---if the tape head is
6063 positioned past the beginning of the archive, the script will rewind
6064 the tape as needed. @xref{Tape Positioning}, for a discussion of tape
6068 @strong{Warning:} The script will delete files from the active file
6069 system if they were not in the file system when the archive was made.
6072 @xref{Incremental Dumps}, for an explanation of how the script makes
6076 @chapter Choosing Files and Names for @command{tar}
6079 Certain options to @command{tar} enable you to specify a name for your
6080 archive. Other options let you decide which files to include or exclude
6081 from the archive, based on when or whether files were modified, whether
6082 the file names do or don't match specified patterns, or whether files
6083 are in specified directories.
6085 This chapter discusses these options in detail.
6088 * file:: Choosing the Archive's Name
6089 * Selecting Archive Members::
6090 * files:: Reading Names from a File
6091 * exclude:: Excluding Some Files
6092 * wildcards:: Wildcards Patterns and Matching
6093 * quoting styles:: Ways of Quoting Special Characters in Names
6094 * transform:: Modifying File and Member Names
6095 * after:: Operating Only on New Files
6096 * recurse:: Descending into Directories
6097 * one:: Crossing File System Boundaries
6101 @section Choosing and Naming Archive Files
6104 @cindex Naming an archive
6105 @cindex Archive Name
6106 @cindex Choosing an archive file
6107 @cindex Where is the archive?
6108 By default, @command{tar} uses an archive file name that was compiled when
6109 it was built on the system; usually this name refers to some physical
6110 tape drive on the machine. However, the person who installed @command{tar}
6111 on the system may not have set the default to a meaningful value as far as
6112 most users are concerned. As a result, you will usually want to tell
6113 @command{tar} where to find (or create) the archive. The
6114 @option{--file=@var{archive-name}} (@option{-f @var{archive-name}})
6115 option allows you to either specify or name a file to use as the archive
6116 instead of the default archive file location.
6119 @xopindex{file, short description}
6120 @item --file=@var{archive-name}
6121 @itemx -f @var{archive-name}
6122 Name the archive to create or operate on. Use in conjunction with
6126 For example, in this @command{tar} command,
6129 $ @kbd{tar -cvf collection.tar blues folk jazz}
6133 @file{collection.tar} is the name of the archive. It must directly
6134 follow the @option{-f} option, since whatever directly follows @option{-f}
6135 @emph{will} end up naming the archive. If you neglect to specify an
6136 archive name, you may end up overwriting a file in the working directory
6137 with the archive you create since @command{tar} will use this file's name
6138 for the archive name.
6140 An archive can be saved as a file in the file system, sent through a
6141 pipe or over a network, or written to an I/O device such as a tape,
6142 floppy disk, or CD write drive.
6144 @cindex Writing new archives
6145 @cindex Archive creation
6146 If you do not name the archive, @command{tar} uses the value of the
6147 environment variable @env{TAPE} as the file name for the archive. If
6148 that is not available, @command{tar} uses a default, compiled-in archive
6149 name, usually that for tape unit zero (i.e., @file{/dev/tu00}).
6151 @cindex Standard input and output
6152 @cindex tar to standard input and output
6153 If you use @file{-} as an @var{archive-name}, @command{tar} reads the
6154 archive from standard input (when listing or extracting files), or
6155 writes it to standard output (when creating an archive). If you use
6156 @file{-} as an @var{archive-name} when modifying an archive,
6157 @command{tar} reads the original archive from its standard input and
6158 writes the entire new archive to its standard output.
6160 The following example is a convenient way of copying directory
6161 hierarchy from @file{sourcedir} to @file{targetdir}.
6164 $ @kbd{(cd sourcedir; tar -cf - .) | (cd targetdir; tar -xpf -)}
6167 The @option{-C} option allows to avoid using subshells:
6170 $ @kbd{tar -C sourcedir -cf - . | tar -C targetdir -xpf -}
6173 In both examples above, the leftmost @command{tar} invocation archives
6174 the contents of @file{sourcedir} to the standard output, while the
6175 rightmost one reads this archive from its standard input and
6176 extracts it. The @option{-p} option tells it to restore permissions
6177 of the extracted files.
6179 @cindex Remote devices
6180 @cindex tar to a remote device
6182 To specify an archive file on a device attached to a remote machine,
6186 @kbd{--file=@var{hostname}:/@var{dev}/@var{file-name}}
6190 @command{tar} will complete the remote connection, if possible, and
6191 prompt you for a username and password. If you use
6192 @option{--file=@@@var{hostname}:/@var{dev}/@var{file-name}}, @command{tar}
6193 will complete the remote connection, if possible, using your username
6194 as the username on the remote machine.
6196 @cindex Local and remote archives
6197 @anchor{local and remote archives}
6198 If the archive file name includes a colon (@samp{:}), then it is assumed
6199 to be a file on another machine. If the archive file is
6200 @samp{@var{user}@@@var{host}:@var{file}}, then @var{file} is used on the
6201 host @var{host}. The remote host is accessed using the @command{rsh}
6202 program, with a username of @var{user}. If the username is omitted
6203 (along with the @samp{@@} sign), then your user name will be used.
6204 (This is the normal @command{rsh} behavior.) It is necessary for the
6205 remote machine, in addition to permitting your @command{rsh} access, to
6206 have the @file{rmt} program installed (This command is included in
6207 the @GNUTAR{} distribution and by default is installed under
6208 @file{@var{prefix}/libexec/rmt}, were @var{prefix} means your
6209 installation prefix). If you need to use a file whose name includes a
6210 colon, then the remote tape drive behavior
6211 can be inhibited by using the @option{--force-local} option.
6213 When the archive is being created to @file{/dev/null}, @GNUTAR{}
6214 tries to minimize input and output operations. The Amanda backup
6215 system, when used with @GNUTAR{}, has an initial sizing pass which
6218 @node Selecting Archive Members
6219 @section Selecting Archive Members
6220 @cindex Specifying files to act on
6221 @cindex Specifying archive members
6223 @dfn{File Name arguments} specify which files in the file system
6224 @command{tar} operates on, when creating or adding to an archive, or which
6225 archive members @command{tar} operates on, when reading or deleting from
6226 an archive. @xref{Operations}.
6228 To specify file names, you can include them as the last arguments on
6229 the command line, as follows:
6231 @kbd{tar} @var{operation} [@var{option1} @var{option2} @dots{}] [@var{file name-1} @var{file name-2} @dots{}]
6234 If a file name begins with dash (@samp{-}), precede it with
6235 @option{--add-file} option to prevent it from being treated as an
6238 @anchor{input name quoting}
6239 By default @GNUTAR{} attempts to @dfn{unquote} each file or member
6240 name, replacing @dfn{escape sequences} according to the following
6243 @multitable @columnfractions 0.20 0.60
6244 @headitem Escape @tab Replaced with
6245 @item \a @tab Audible bell (@acronym{ASCII} 7)
6246 @item \b @tab Backspace (@acronym{ASCII} 8)
6247 @item \f @tab Form feed (@acronym{ASCII} 12)
6248 @item \n @tab New line (@acronym{ASCII} 10)
6249 @item \r @tab Carriage return (@acronym{ASCII} 13)
6250 @item \t @tab Horizontal tabulation (@acronym{ASCII} 9)
6251 @item \v @tab Vertical tabulation (@acronym{ASCII} 11)
6252 @item \? @tab @acronym{ASCII} 127
6253 @item \@var{n} @tab @acronym{ASCII} @var{n} (@var{n} should be an octal number
6257 A backslash followed by any other symbol is retained.
6259 This default behavior is controlled by the following command line
6265 Enable unquoting input file or member names (default).
6269 Disable unquoting input file or member names.
6272 If you specify a directory name as a file name argument, all the files
6273 in that directory are operated on by @command{tar}.
6275 If you do not specify files, @command{tar} behavior differs depending
6276 on the operation mode as described below:
6278 When @command{tar} is invoked with @option{--create} (@option{-c}),
6279 @command{tar} will stop immediately, reporting the following:
6283 $ @kbd{tar cf a.tar}
6284 tar: Cowardly refusing to create an empty archive
6285 Try `tar --help' or `tar --usage' for more information.
6289 If you specify either @option{--list} (@option{-t}) or
6290 @option{--extract} (@option{--get}, @option{-x}), @command{tar}
6291 operates on all the archive members in the archive.
6293 If run with @option{--diff} option, tar will compare the archive with
6294 the contents of the current working directory.
6296 If you specify any other operation, @command{tar} does nothing.
6298 By default, @command{tar} takes file names from the command line. However,
6299 there are other ways to specify file or member names, or to modify the
6300 manner in which @command{tar} selects the files or members upon which to
6301 operate. In general, these methods work both for specifying the names
6302 of files and archive members.
6305 @section Reading Names from a File
6307 @cindex Reading file names from a file
6308 @cindex Lists of file names
6309 @cindex File Name arguments, alternatives
6310 Instead of giving the names of files or archive members on the command
6311 line, you can put the names into a file, and then use the
6312 @option{--files-from=@var{file-of-names}} (@option{-T
6313 @var{file-of-names}}) option to @command{tar}. Give the name of the
6314 file which contains the list of files to include as the argument to
6315 @option{--files-from}. In the list, the file names should be separated by
6316 newlines. You will frequently use this option when you have generated
6317 the list of files to archive with the @command{find} utility.
6321 @item --files-from=@var{file-name}
6322 @itemx -T @var{file-name}
6323 Get names to extract or create from file @var{file-name}.
6326 If you give a single dash as a file name for @option{--files-from}, (i.e.,
6327 you specify either @code{--files-from=-} or @code{-T -}), then the file
6328 names are read from standard input.
6330 Unless you are running @command{tar} with @option{--create}, you can not use
6331 both @code{--files-from=-} and @code{--file=-} (@code{-f -}) in the same
6334 Any number of @option{-T} options can be given in the command line.
6336 The following example shows how to use @command{find} to generate a list of
6337 files smaller than 400K in length and put that list into a file
6338 called @file{small-files}. You can then use the @option{-T} option to
6339 @command{tar} to specify the files from that file, @file{small-files}, to
6340 create the archive @file{little.tgz}. (The @option{-z} option to
6341 @command{tar} compresses the archive with @command{gzip}; @pxref{gzip} for
6345 $ @kbd{find . -size -400 -print > small-files}
6346 $ @kbd{tar -c -v -z -T small-files -f little.tgz}
6350 In the file list given by @option{-T} option, any file name beginning
6351 with @samp{-} character is considered a @command{tar} option and is
6352 processed accordingly.@footnote{Versions of @GNUTAR{} up to 1.15.1
6353 recognized only @option{-C} option in file lists, and only if the
6354 option and its argument occupied two consecutive lines.} For example,
6355 the common use of this feature is to change to another directory by
6356 specifying @option{-C} option:
6366 $ @kbd{tar -c -f foo.tar --files-from list}
6371 In this example, @command{tar} will first switch to @file{/etc}
6372 directory and add files @file{passwd} and @file{hosts} to the
6373 archive. Then it will change to @file{/lib} directory and will archive
6374 the file @file{libc.a}. Thus, the resulting archive @file{foo.tar} will
6379 $ @kbd{tar tf foo.tar}
6387 @xopindex{directory, using in @option{--files-from} argument}
6388 Notice that the option parsing algorithm used with @option{-T} is
6389 stricter than the one used by shell. Namely, when specifying option
6390 arguments, you should observe the following rules:
6394 When using short (single-letter) option form, its argument must
6395 immediately follow the option letter, without any intervening
6396 whitespace. For example: @code{-Cdir}.
6399 When using long option form, the option argument must be separated
6400 from the option by a single equal sign. No whitespace is allowed on
6401 any side of the equal sign. For example: @code{--directory=dir}.
6404 For both short and long option forms, the option argument can be given
6405 on the next line after the option name, e.g.:
6426 If you happen to have a file whose name starts with @samp{-},
6427 precede it with @option{--add-file} option to prevent it from
6428 being recognized as an option. For example: @code{--add-file=--my-file}.
6435 @subsection @code{NUL} Terminated File Names
6437 @cindex File names, terminated by @code{NUL}
6438 @cindex @code{NUL} terminated file names
6439 The @option{--null} option causes
6440 @option{--files-from=@var{file-of-names}} (@option{-T @var{file-of-names}})
6441 to read file names terminated by a @code{NUL} instead of a newline, so
6442 files whose names contain newlines can be archived using
6443 @option{--files-from}.
6448 Only consider @code{NUL} terminated file names, instead of files that
6449 terminate in a newline.
6452 The @option{--null} option is just like the one in @acronym{GNU}
6453 @command{xargs} and @command{cpio}, and is useful with the
6454 @option{-print0} predicate of @acronym{GNU} @command{find}. In
6455 @command{tar}, @option{--null} also disables special handling for
6456 file names that begin with dash.
6458 This example shows how to use @command{find} to generate a list of files
6459 larger than 800K in length and put that list into a file called
6460 @file{long-files}. The @option{-print0} option to @command{find} is just
6461 like @option{-print}, except that it separates files with a @code{NUL}
6462 rather than with a newline. You can then run @command{tar} with both the
6463 @option{--null} and @option{-T} options to specify that @command{tar} get the
6464 files from that file, @file{long-files}, to create the archive
6465 @file{big.tgz}. The @option{--null} option to @command{tar} will cause
6466 @command{tar} to recognize the @code{NUL} separator between files.
6469 $ @kbd{find . -size +800 -print0 > long-files}
6470 $ @kbd{tar -c -v --null --files-from=long-files --file=big.tar}
6473 @FIXME{say anything else here to conclude the section?}
6476 @section Excluding Some Files
6479 @cindex File names, excluding files by
6480 @cindex Excluding files by name and pattern
6481 @cindex Excluding files by file system
6482 To avoid operating on files whose names match a particular pattern,
6483 use the @option{--exclude} or @option{--exclude-from} options.
6487 @item --exclude=@var{pattern}
6488 Causes @command{tar} to ignore files that match the @var{pattern}.
6492 The @option{--exclude=@var{pattern}} option prevents any file or
6493 member whose name matches the shell wildcard (@var{pattern}) from
6495 For example, to create an archive with all the contents of the directory
6496 @file{src} except for files whose names end in @file{.o}, use the
6497 command @samp{tar -cf src.tar --exclude='*.o' src}.
6499 You may give multiple @option{--exclude} options.
6502 @opindex exclude-from
6503 @item --exclude-from=@var{file}
6504 @itemx -X @var{file}
6505 Causes @command{tar} to ignore files that match the patterns listed in
6509 @findex exclude-from
6510 Use the @option{--exclude-from} option to read a
6511 list of patterns, one per line, from @var{file}; @command{tar} will
6512 ignore files matching those patterns. Thus if @command{tar} is
6513 called as @w{@samp{tar -c -X foo .}} and the file @file{foo} contains a
6514 single line @file{*.o}, no files whose names end in @file{.o} will be
6515 added to the archive.
6517 Notice, that lines from @var{file} are read verbatim. One of the
6518 frequent errors is leaving some extra whitespace after a file name,
6519 which is difficult to catch using text editors.
6521 However, empty lines are OK.
6523 @cindex version control system, excluding files
6524 @cindex VCS, excluding files
6525 @cindex SCCS, excluding files
6526 @cindex RCS, excluding files
6527 @cindex CVS, excluding files
6528 @cindex SVN, excluding files
6529 @cindex git, excluding files
6531 @opindex exclude-vcs
6533 Exclude files and directories used by some version control systems.
6536 As of version @value{VERSION}, the following files are excluded:
6539 @item @file{CVS/}, and everything under it
6540 @item @file{RCS/}, and everything under it
6541 @item @file{SCCS/}, and everything under it
6542 @item @file{.git/}, and everything under it
6543 @item @file{.gitignore}
6544 @item @file{.cvsignore}
6545 @item @file{.svn/}, and everything under it
6546 @item @file{.arch-ids/}, and everything under it
6547 @item @file{@{arch@}/}, and everything under it
6548 @item @file{=RELEASE-ID}
6549 @item @file{=meta-update}
6550 @item @file{=update}
6553 @findex exclude-caches
6554 When creating an archive, the @option{--exclude-caches} option family
6555 causes @command{tar} to exclude all directories that contain a @dfn{cache
6556 directory tag}. A cache directory tag is a short file with the
6557 well-known name @file{CACHEDIR.TAG} and having a standard header
6558 specified in @url{http://www.brynosaurus.com/cachedir/spec.html}.
6559 Various applications write cache directory tags into directories they
6560 use to hold regenerable, non-precious data, so that such data can be
6561 more easily excluded from backups.
6563 There are three @samp{exclude-caches} options, each providing a different
6564 exclusion semantics:
6567 @opindex exclude-caches
6568 @item --exclude-caches
6569 Do not archive the contents of the directory, but archive the
6570 directory itself and the @file{CACHEDIR.TAG} file.
6572 @opindex exclude-caches-under
6573 @item --exclude-caches-under
6574 Do not archive the contents of the directory, nor the
6575 @file{CACHEDIR.TAG} file, archive only the directory itself.
6577 @opindex exclude-caches-all
6578 @item --exclude-caches-all
6579 Omit directories containing @file{CACHEDIR.TAG} file entirely.
6583 Another option family, @option{--exclude-tag}, provides a generalization of
6584 this concept. It takes a single argument, a file name to look for.
6585 Any directory that contains this file will be excluded from the dump.
6586 Similarly to @samp{exclude-caches}, there are three options in this
6590 @opindex exclude-tag
6591 @item --exclude-tag=@var{file}
6592 Do not dump the contents of the directory, but dump the
6593 directory itself and the @var{file}.
6595 @opindex exclude-tag-under
6596 @item --exclude-tag-under=@var{file}
6597 Do not dump the contents of the directory, nor the
6598 @var{file}, archive only the directory itself.
6600 @opindex exclude-tag-all
6601 @item --exclude-tag-all=@var{file}
6602 Omit directories containing @var{file} file entirely.
6605 Multiple @option{--exclude-tag*} options can be given.
6607 For example, given this directory:
6622 The @option{--exclude-tag} will produce the following:
6625 $ @kbd{tar -cf archive.tar --exclude-tag=tagfile -v dir}
6630 tar: dir/folk/: contains a cache directory tag tagfile;
6635 Both the @file{dir/folk} directory and its tagfile are preserved in
6636 the archive, however the rest of files in this directory are not.
6638 Now, using the @option{--exclude-tag-under} option will exclude
6639 @file{tagfile} from the dump, while still preserving the directory
6640 itself, as shown in this example:
6643 $ @kbd{tar -cf archive.tar --exclude-tag-under=tagfile -v dir}
6648 ./tar: dir/folk/: contains a cache directory tag tagfile;
6652 Finally, using @option{--exclude-tag-all} omits the @file{dir/folk}
6656 $ @kbd{tar -cf archive.tar --exclude-tag-all=tagfile -v dir}
6660 ./tar: dir/folk/: contains a cache directory tag tagfile;
6661 directory not dumped
6665 * problems with exclude::
6668 @node problems with exclude
6669 @unnumberedsubsec Problems with Using the @code{exclude} Options
6671 @xopindex{exclude, potential problems with}
6672 Some users find @samp{exclude} options confusing. Here are some common
6677 The main operating mode of @command{tar} does not act on a file name
6678 explicitly listed on the command line, if one of its file name
6679 components is excluded. In the example above, if
6680 you create an archive and exclude files that end with @samp{*.o}, but
6681 explicitly name the file @samp{dir.o/foo} after all the options have been
6682 listed, @samp{dir.o/foo} will be excluded from the archive.
6685 You can sometimes confuse the meanings of @option{--exclude} and
6686 @option{--exclude-from}. Be careful: use @option{--exclude} when files
6687 to be excluded are given as a pattern on the command line. Use
6688 @option{--exclude-from} to introduce the name of a file which contains
6689 a list of patterns, one per line; each of these patterns can exclude
6690 zero, one, or many files.
6693 When you use @option{--exclude=@var{pattern}}, be sure to quote the
6694 @var{pattern} parameter, so @GNUTAR{} sees wildcard characters
6695 like @samp{*}. If you do not do this, the shell might expand the
6696 @samp{*} itself using files at hand, so @command{tar} might receive a
6697 list of files instead of one pattern, or none at all, making the
6698 command somewhat illegal. This might not correspond to what you want.
6703 $ @kbd{tar -c -f @var{archive.tar} --exclude '*.o' @var{directory}}
6711 $ @kbd{tar -c -f @var{archive.tar} --exclude *.o @var{directory}}
6715 You must use use shell syntax, or globbing, rather than @code{regexp}
6716 syntax, when using exclude options in @command{tar}. If you try to use
6717 @code{regexp} syntax to describe files to be excluded, your command
6721 @FIXME{The change in semantics must have occurred before 1.11,
6722 so I doubt if it is worth mentioning at all. Anyway, should at
6723 least specify in which version the semantics changed.}
6724 In earlier versions of @command{tar}, what is now the
6725 @option{--exclude-from} option was called @option{--exclude} instead.
6726 Now, @option{--exclude} applies to patterns listed on the command
6727 line and @option{--exclude-from} applies to patterns listed in a
6733 @section Wildcards Patterns and Matching
6735 @dfn{Globbing} is the operation by which @dfn{wildcard} characters,
6736 @samp{*} or @samp{?} for example, are replaced and expanded into all
6737 existing files matching the given pattern. @GNUTAR{} can use wildcard
6738 patterns for matching (or globbing) archive members when extracting
6739 from or listing an archive. Wildcard patterns are also used for
6740 verifying volume labels of @command{tar} archives. This section has the
6741 purpose of explaining wildcard syntax for @command{tar}.
6743 @FIXME{the next few paragraphs need work.}
6745 A @var{pattern} should be written according to shell syntax, using wildcard
6746 characters to effect globbing. Most characters in the pattern stand
6747 for themselves in the matched string, and case is significant: @samp{a}
6748 will match only @samp{a}, and not @samp{A}. The character @samp{?} in the
6749 pattern matches any single character in the matched string. The character
6750 @samp{*} in the pattern matches zero, one, or more single characters in
6751 the matched string. The character @samp{\} says to take the following
6752 character of the pattern @emph{literally}; it is useful when one needs to
6753 match the @samp{?}, @samp{*}, @samp{[} or @samp{\} characters, themselves.
6755 The character @samp{[}, up to the matching @samp{]}, introduces a character
6756 class. A @dfn{character class} is a list of acceptable characters
6757 for the next single character of the matched string. For example,
6758 @samp{[abcde]} would match any of the first five letters of the alphabet.
6759 Note that within a character class, all of the ``special characters''
6760 listed above other than @samp{\} lose their special meaning; for example,
6761 @samp{[-\\[*?]]} would match any of the characters, @samp{-}, @samp{\},
6762 @samp{[}, @samp{*}, @samp{?}, or @samp{]}. (Due to parsing constraints,
6763 the characters @samp{-} and @samp{]} must either come @emph{first} or
6764 @emph{last} in a character class.)
6766 @cindex Excluding characters from a character class
6767 @cindex Character class, excluding characters from
6768 If the first character of the class after the opening @samp{[}
6769 is @samp{!} or @samp{^}, then the meaning of the class is reversed.
6770 Rather than listing character to match, it lists those characters which
6771 are @emph{forbidden} as the next single character of the matched string.
6773 Other characters of the class stand for themselves. The special
6774 construction @samp{[@var{a}-@var{e}]}, using an hyphen between two
6775 letters, is meant to represent all characters between @var{a} and
6778 @FIXME{need to add a sentence or so here to make this clear for those
6779 who don't have dan around.}
6781 Periods (@samp{.}) or forward slashes (@samp{/}) are not considered
6782 special for wildcard matches. However, if a pattern completely matches
6783 a directory prefix of a matched string, then it matches the full matched
6784 string: thus, excluding a directory also excludes all the files beneath it.
6787 * controlling pattern-matching::
6790 @node controlling pattern-matching
6791 @unnumberedsubsec Controlling Pattern-Matching
6793 For the purposes of this section, we call @dfn{exclusion members} all
6794 member names obtained while processing @option{--exclude} and
6795 @option{--exclude-from} options, and @dfn{inclusion members} those
6796 member names that were given in the command line or read from the file
6797 specified with @option{--files-from} option.
6799 These two pairs of member lists are used in the following operations:
6800 @option{--diff}, @option{--extract}, @option{--list},
6803 There are no inclusion members in create mode (@option{--create} and
6804 @option{--append}), since in this mode the names obtained from the
6805 command line refer to @emph{files}, not archive members.
6807 By default, inclusion members are compared with archive members
6808 literally @footnote{Notice that earlier @GNUTAR{} versions used
6809 globbing for inclusion members, which contradicted to UNIX98
6810 specification and was not documented. @xref{Changes}, for more
6811 information on this and other changes.} and exclusion members are
6812 treated as globbing patterns. For example:
6816 $ @kbd{tar tf foo.tar}
6821 # @i{Member names are used verbatim:}
6822 $ @kbd{tar -xf foo.tar -v '[remarks]'}
6824 # @i{Exclude member names are globbed:}
6825 $ @kbd{tar -xf foo.tar -v --exclude '*.c'}
6831 This behavior can be altered by using the following options:
6836 Treat all member names as wildcards.
6838 @opindex no-wildcards
6839 @item --no-wildcards
6840 Treat all member names as literal strings.
6843 Thus, to extract files whose names end in @samp{.c}, you can use:
6846 $ @kbd{tar -xf foo.tar -v --wildcards '*.c'}
6852 Notice quoting of the pattern to prevent the shell from interpreting
6855 The effect of @option{--wildcards} option is canceled by
6856 @option{--no-wildcards}. This can be used to pass part of
6857 the command line arguments verbatim and other part as globbing
6858 patterns. For example, the following invocation:
6861 $ @kbd{tar -xf foo.tar --wildcards '*.txt' --no-wildcards '[remarks]'}
6865 instructs @command{tar} to extract from @file{foo.tar} all files whose
6866 names end in @samp{.txt} and the file named @file{[remarks]}.
6868 Normally, a pattern matches a name if an initial subsequence of the
6869 name's components matches the pattern, where @samp{*}, @samp{?}, and
6870 @samp{[...]} are the usual shell wildcards, @samp{\} escapes wildcards,
6871 and wildcards can match @samp{/}.
6873 Other than optionally stripping leading @samp{/} from names
6874 (@pxref{absolute}), patterns and names are used as-is. For
6875 example, trailing @samp{/} is not trimmed from a user-specified name
6876 before deciding whether to exclude it.
6878 However, this matching procedure can be altered by the options listed
6879 below. These options accumulate. For example:
6882 --ignore-case --exclude='makefile' --no-ignore-case ---exclude='readme'
6886 ignores case when excluding @samp{makefile}, but not when excluding
6891 @opindex no-anchored
6893 @itemx --no-anchored
6894 If anchored, a pattern must match an initial subsequence
6895 of the name's components. Otherwise, the pattern can match any
6896 subsequence. Default is @option{--no-anchored} for exclusion members
6897 and @option{--anchored} inclusion members.
6899 @opindex ignore-case
6900 @opindex no-ignore-case
6902 @itemx --no-ignore-case
6903 When ignoring case, upper-case patterns match lower-case names and vice versa.
6904 When not ignoring case (the default), matching is case-sensitive.
6906 @opindex wildcards-match-slash
6907 @opindex no-wildcards-match-slash
6908 @item --wildcards-match-slash
6909 @itemx --no-wildcards-match-slash
6910 When wildcards match slash (the default for exclusion members), a
6911 wildcard like @samp{*} in the pattern can match a @samp{/} in the
6912 name. Otherwise, @samp{/} is matched only by @samp{/}.
6916 The @option{--recursion} and @option{--no-recursion} options
6917 (@pxref{recurse}) also affect how member patterns are interpreted. If
6918 recursion is in effect, a pattern matches a name if it matches any of
6919 the name's parent directories.
6921 The following table summarizes pattern-matching default values:
6923 @multitable @columnfractions .3 .7
6924 @headitem Members @tab Default settings
6925 @item Inclusion @tab @option{--no-wildcards --anchored --no-wildcards-match-slash}
6926 @item Exclusion @tab @option{--wildcards --no-anchored --wildcards-match-slash}
6929 @node quoting styles
6930 @section Quoting Member Names
6932 When displaying member names, @command{tar} takes care to avoid
6933 ambiguities caused by certain characters. This is called @dfn{name
6934 quoting}. The characters in question are:
6937 @item Non-printable control characters:
6939 @multitable @columnfractions 0.20 0.10 0.60
6940 @headitem Character @tab @acronym{ASCII} @tab Character name
6941 @item \a @tab 7 @tab Audible bell
6942 @item \b @tab 8 @tab Backspace
6943 @item \f @tab 12 @tab Form feed
6944 @item \n @tab 10 @tab New line
6945 @item \r @tab 13 @tab Carriage return
6946 @item \t @tab 9 @tab Horizontal tabulation
6947 @item \v @tab 11 @tab Vertical tabulation
6950 @item Space (@acronym{ASCII} 32)
6952 @item Single and double quotes (@samp{'} and @samp{"})
6954 @item Backslash (@samp{\})
6957 The exact way @command{tar} uses to quote these characters depends on
6958 the @dfn{quoting style}. The default quoting style, called
6959 @dfn{escape} (see below), uses backslash notation to represent control
6960 characters, space and backslash. Using this quoting style, control
6961 characters are represented as listed in column @samp{Character} in the
6962 above table, a space is printed as @samp{\ } and a backslash as @samp{\\}.
6964 @GNUTAR{} offers seven distinct quoting styles, which can be selected
6965 using @option{--quoting-style} option:
6968 @item --quoting-style=@var{style}
6969 @opindex quoting-style
6971 Sets quoting style. Valid values for @var{style} argument are:
6972 literal, shell, shell-always, c, escape, locale, clocale.
6975 These styles are described in detail below. To illustrate their
6976 effect, we will use an imaginary tar archive @file{arch.tar}
6977 containing the following members:
6981 # 1. Contains horizontal tabulation character.
6983 # 2. Contains newline character
6986 # 3. Contains a space
6988 # 4. Contains double quotes
6990 # 5. Contains single quotes
6992 # 6. Contains a backslash character:
6997 Here is how usual @command{ls} command would have listed them, if they
6998 had existed in the current working directory:
7016 No quoting, display each character as is:
7020 $ @kbd{tar tf arch.tar --quoting-style=literal}
7033 Display characters the same way Bourne shell does:
7034 control characters, except @samp{\t} and @samp{\n}, are printed using
7035 backslash escapes, @samp{\t} and @samp{\n} are printed as is, and a
7036 single quote is printed as @samp{\'}. If a name contains any quoted
7037 characters, it is enclosed in single quotes. In particular, if a name
7038 contains single quotes, it is printed as several single-quoted strings:
7042 $ @kbd{tar tf arch.tar --quoting-style=shell}
7045 './a'\''single'\''quote'
7055 Same as @samp{shell}, but the names are always enclosed in single
7060 $ @kbd{tar tf arch.tar --quoting-style=shell-always}
7063 './a'\''single'\''quote'
7073 Use the notation of the C programming language. All names are
7074 enclosed in double quotes. Control characters are quoted using
7075 backslash notations, double quotes are represented as @samp{\"},
7076 backslash characters are represented as @samp{\\}. Single quotes and
7077 spaces are not quoted:
7081 $ @kbd{tar tf arch.tar --quoting-style=c}
7085 "./a\"double\"quote"
7093 Control characters are printed using backslash notation, a space is
7094 printed as @samp{\ } and a backslash as @samp{\\}. This is the
7095 default quoting style, unless it was changed when configured the
7100 $ @kbd{tar tf arch.tar --quoting-style=escape}
7112 Control characters, single quote and backslash are printed using
7113 backslash notation. All names are quoted using left and right
7114 quotation marks, appropriate to the current locale. If it does not
7115 define quotation marks, use @samp{`} as left and @samp{'} as right
7116 quotation marks. Any occurrences of the right quotation mark in a
7117 name are escaped with @samp{\}, for example:
7123 $ @kbd{tar tf arch.tar --quoting-style=locale}
7126 `./a\'single\'quote'
7135 Same as @samp{locale}, but @samp{"} is used for both left and right
7136 quotation marks, if not provided by the currently selected locale:
7140 $ @kbd{tar tf arch.tar --quoting-style=clocale}
7144 "./a\"double\"quote"
7152 You can specify which characters should be quoted in addition to those
7153 implied by the current quoting style:
7156 @item --quote-chars=@var{string}
7157 Always quote characters from @var{string}, even if the selected
7158 quoting style would not quote them.
7161 For example, using @samp{escape} quoting (compare with the usual
7162 escape listing above):
7166 $ @kbd{tar tf arch.tar --quoting-style=escape --quote-chars=' "'}
7177 To disable quoting of such additional characters, use the following
7181 @item --no-quote-chars=@var{string}
7182 Remove characters listed in @var{string} from the list of quoted
7183 characters set by the previous @option{--quote-chars} option.
7186 This option is particularly useful if you have added
7187 @option{--quote-chars} to your @env{TAR_OPTIONS} (@pxref{TAR_OPTIONS})
7188 and wish to disable it for the current invocation.
7190 Note, that @option{--no-quote-chars} does @emph{not} disable those
7191 characters that are quoted by default in the selected quoting style.
7194 @section Modifying File and Member Names
7196 @command{Tar} archives contain detailed information about files stored
7197 in them and full file names are part of that information. When
7198 storing file to an archive, its file name is recorded in the archive
7199 along with the actual file contents. When restoring from an archive,
7200 a file is created on disk with exactly the same name as that stored
7201 in the archive. In the majority of cases this is the desired behavior
7202 of a file archiver. However, there are some cases when it is not.
7204 First of all, it is often unsafe to extract archive members with
7205 absolute file names or those that begin with a @file{../}. @GNUTAR{}
7206 takes special precautions when extracting such names and provides a
7207 special option for handling them, which is described in
7210 Secondly, you may wish to extract file names without some leading
7211 directory components, or with otherwise modified names. In other
7212 cases it is desirable to store files under differing names in the
7215 @GNUTAR{} provides two options for these needs.
7218 @opindex strip-components
7219 @item --strip-components=@var{number}
7220 Strip given @var{number} of leading components from file names before
7224 For example, suppose you have archived whole @file{/usr} hierarchy to
7225 a tar archive named @file{usr.tar}. Among other files, this archive
7226 contains @file{usr/include/stdlib.h}, which you wish to extract to
7227 the current working directory. To do so, you type:
7230 $ @kbd{tar -xf usr.tar --strip=2 usr/include/stdlib.h}
7233 The option @option{--strip=2} instructs @command{tar} to strip the
7234 two leading components (@file{usr/} and @file{include/}) off the file
7237 If you add to the above invocation @option{--verbose} (@option{-v})
7238 option, you will note that the verbose listing still contains the
7239 full file name, with the two removed components still in place. This
7240 can be inconvenient, so @command{tar} provides a special option for
7241 altering this behavior:
7243 @anchor{show-transformed-names}
7245 @opindex show-transformed-names
7246 @item --show-transformed-names
7247 Display file or member names with all requested transformations
7256 $ @kbd{tar -xf usr.tar -v --strip=2 usr/include/stdlib.h}
7257 usr/include/stdlib.h
7258 $ @kbd{tar -xf usr.tar -v --strip=2 --show-transformed usr/include/stdlib.h}
7263 Notice that in both cases the file is @file{stdlib.h} extracted to the
7264 current working directory, @option{--show-transformed-names} affects
7265 only the way its name is displayed.
7267 This option is especially useful for verifying whether the invocation
7268 will have the desired effect. Thus, before running
7271 $ @kbd{tar -x --strip=@var{n}}
7275 it is often advisable to run
7278 $ @kbd{tar -t -v --show-transformed --strip=@var{n}}
7282 to make sure the command will produce the intended results.
7284 In case you need to apply more complex modifications to the file name,
7285 @GNUTAR{} provides a general-purpose transformation option:
7289 @item --transform=@var{expression}
7290 Modify file names using supplied @var{expression}.
7294 The @var{expression} is a @command{sed}-like replace expression of the
7298 s/@var{regexp}/@var{replace}/[@var{flags}]
7302 where @var{regexp} is a @dfn{regular expression}, @var{replace} is a
7303 replacement for each file name part that matches @var{regexp}. Both
7304 @var{regexp} and @var{replace} are described in detail in
7305 @ref{The "s" Command, The "s" Command, The `s' Command, sed, GNU sed}.
7307 Supported @var{flags} are:
7311 Apply the replacement to @emph{all} matches to the @var{regexp}, not
7315 Use case-insensitive matching
7318 @var{regexp} is an @dfn{extended regular expression} (@pxref{Extended
7319 regexps, Extended regular expressions, Extended regular expressions,
7323 Only replace the @var{number}th match of the @var{regexp}.
7325 Note: the @var{posix} standard does not specify what should happen
7326 when you mix the @samp{g} and @var{number} modifiers. @GNUTAR{}
7327 follows the GNU @command{sed} implementation in this regard, so
7328 the interaction is defined to be: ignore matches before the
7329 @var{number}th, and then match and replace all matches from the
7334 Any delimiter can be used in lieue of @samp{/}, the only requirement being
7335 that it be used consistently throughout the expression. For example,
7336 the following two expressions are equivalent:
7345 Changing delimiters is often useful when the @var{regex} contains
7346 slashes. For example, it is more convenient to write @code{s,/,-,} than
7349 Here are several examples of @option{--transform} usage:
7352 @item Extract @file{usr/} hierarchy into @file{usr/local/}:
7355 $ @kbd{tar --transform='s,usr/,usr/local/,' -x -f arch.tar}
7358 @item Strip two leading directory components (equivalent to
7359 @option{--strip-components=2}):
7362 $ @kbd{tar --transform='s,/*[^/]*/[^/]*/,,' -x -f arch.tar}
7365 @item Prepend @file{/prefix/} to each file name:
7368 $ @kbd{tar --transform 's,^,/prefix/,' -x -f arch.tar}
7371 @item Convert each file name to lower case:
7374 $ @kbd{tar --transform 's/.*/\L&/' -x -f arch.tar}
7379 Unlike @option{--strip-components}, @option{--transform} can be used
7380 in any @GNUTAR{} operation mode. For example, the following command
7381 adds files to the archive while replacing the leading @file{usr/}
7382 component with @file{var/}:
7385 $ @kbd{tar -cf arch.tar --transform='s,^usr/,var/,' /}
7388 To test @option{--transform} effect we suggest using
7389 @option{--show-transformed-names} option:
7392 $ @kbd{tar -cf arch.tar --transform='s,^usr/,var/,' \
7393 --verbose --show-transformed-names /}
7396 If both @option{--strip-components} and @option{--transform} are used
7397 together, then @option{--transform} is applied first, and the required
7398 number of components is then stripped from its result.
7401 @section Operating Only on New Files
7404 @cindex Excluding file by age
7405 @cindex Data Modification time, excluding files by
7406 @cindex Modification time, excluding files by
7407 @cindex Age, excluding files by
7408 The @option{--after-date=@var{date}} (@option{--newer=@var{date}},
7409 @option{-N @var{date}}) option causes @command{tar} to only work on
7410 files whose data modification or status change times are newer than
7411 the @var{date} given. If @var{date} starts with @samp{/} or @samp{.},
7412 it is taken to be a file name; the data modification time of that file
7413 is used as the date. If you use this option when creating or appending
7414 to an archive, the archive will only include new files. If you use
7415 @option{--after-date} when extracting an archive, @command{tar} will
7416 only extract files newer than the @var{date} you specify.
7418 If you only want @command{tar} to make the date comparison based on
7419 modification of the file's data (rather than status
7420 changes), then use the @option{--newer-mtime=@var{date}} option.
7422 You may use these options with any operation. Note that these options
7423 differ from the @option{--update} (@option{-u}) operation in that they
7424 allow you to specify a particular date against which @command{tar} can
7425 compare when deciding whether or not to archive the files.
7430 @item --after-date=@var{date}
7431 @itemx --newer=@var{date}
7432 @itemx -N @var{date}
7433 Only store files newer than @var{date}.
7435 Acts on files only if their data modification or status change times are
7436 later than @var{date}. Use in conjunction with any operation.
7438 If @var{date} starts with @samp{/} or @samp{.}, it is taken to be a file
7439 name; the data modification time of that file is used as the date.
7441 @opindex newer-mtime
7442 @item --newer-mtime=@var{date}
7443 Acts like @option{--after-date}, but only looks at data modification times.
7446 These options limit @command{tar} to operate only on files which have
7447 been modified after the date specified. A file's status is considered to have
7448 changed if its contents have been modified, or if its owner,
7449 permissions, and so forth, have been changed. (For more information on
7450 how to specify a date, see @ref{Date input formats}; remember that the
7451 entire date argument must be quoted if it contains any spaces.)
7453 Gurus would say that @option{--after-date} tests both the data
7454 modification time (@code{mtime}, the time the contents of the file
7455 were last modified) and the status change time (@code{ctime}, the time
7456 the file's status was last changed: owner, permissions, etc.@:)
7457 fields, while @option{--newer-mtime} tests only the @code{mtime}
7460 To be precise, @option{--after-date} checks @emph{both} @code{mtime} and
7461 @code{ctime} and processes the file if either one is more recent than
7462 @var{date}, while @option{--newer-mtime} only checks @code{mtime} and
7463 disregards @code{ctime}. Neither does it use @code{atime} (the last time the
7464 contents of the file were looked at).
7466 Date specifiers can have embedded spaces. Because of this, you may need
7467 to quote date arguments to keep the shell from parsing them as separate
7468 arguments. For example, the following command will add to the archive
7469 all the files modified less than two days ago:
7472 $ @kbd{tar -cf foo.tar --newer-mtime '2 days ago'}
7475 When any of these options is used with the option @option{--verbose}
7476 (@pxref{verbose tutorial}) @GNUTAR{} will try to convert the specified
7477 date back to its textual representation and compare that with the
7478 one given with the option. If the two dates differ, @command{tar} will
7479 print a warning saying what date it will use. This is to help user
7480 ensure he is using the right date. For example:
7484 $ @kbd{tar -c -f archive.tar --after-date='10 days ago' .}
7485 tar: Option --after-date: Treating date `10 days ago' as 2006-06-11
7491 @strong{Please Note:} @option{--after-date} and @option{--newer-mtime}
7492 should not be used for incremental backups. @xref{Incremental Dumps},
7493 for proper way of creating incremental backups.
7497 @section Descending into Directories
7499 @cindex Avoiding recursion in directories
7500 @cindex Descending directories, avoiding
7501 @cindex Directories, avoiding recursion
7502 @cindex Recursion in directories, avoiding
7504 @FIXME{arrggh! this is still somewhat confusing to me. :-< }
7506 Usually, @command{tar} will recursively explore all directories (either
7507 those given on the command line or through the @option{--files-from}
7508 option) for the various files they contain. However, you may not always
7509 want @command{tar} to act this way.
7511 @opindex no-recursion
7512 The @option{--no-recursion} option inhibits @command{tar}'s recursive descent
7513 into specified directories. If you specify @option{--no-recursion}, you can
7514 use the @command{find} utility for hunting through levels of directories to
7515 construct a list of file names which you could then pass to @command{tar}.
7516 @command{find} allows you to be more selective when choosing which files to
7517 archive; see @ref{files}, for more information on using @command{find} with
7518 @command{tar}, or look.
7521 @item --no-recursion
7522 Prevents @command{tar} from recursively descending directories.
7526 Requires @command{tar} to recursively descend directories.
7527 This is the default.
7530 When you use @option{--no-recursion}, @GNUTAR{} grabs
7531 directory entries themselves, but does not descend on them
7532 recursively. Many people use @command{find} for locating files they
7533 want to back up, and since @command{tar} @emph{usually} recursively
7534 descends on directories, they have to use the @samp{@w{-not -type d}}
7535 test in their @command{find} invocation (@pxref{Type, Type, Type test,
7536 find, Finding Files}), as they usually do not want all the files in a
7537 directory. They then use the @option{--files-from} option to archive
7538 the files located via @command{find}.
7540 The problem when restoring files archived in this manner is that the
7541 directories themselves are not in the archive; so the
7542 @option{--same-permissions} (@option{--preserve-permissions},
7543 @option{-p}) option does not affect them---while users might really
7544 like it to. Specifying @option{--no-recursion} is a way to tell
7545 @command{tar} to grab only the directory entries given to it, adding
7546 no new files on its own. To summarize, if you use @command{find} to
7547 create a list of files to be stored in an archive, use it as follows:
7551 $ @kbd{find @var{dir} @var{tests} | \
7552 tar -cf @var{archive} -T - --no-recursion}
7556 The @option{--no-recursion} option also applies when extracting: it
7557 causes @command{tar} to extract only the matched directory entries, not
7558 the files under those directories.
7560 The @option{--no-recursion} option also affects how globbing patterns
7561 are interpreted (@pxref{controlling pattern-matching}).
7563 The @option{--no-recursion} and @option{--recursion} options apply to
7564 later options and operands, and can be overridden by later occurrences
7565 of @option{--no-recursion} and @option{--recursion}. For example:
7568 $ @kbd{tar -cf jams.tar --no-recursion grape --recursion grape/concord}
7572 creates an archive with one entry for @file{grape}, and the recursive
7573 contents of @file{grape/concord}, but no entries under @file{grape}
7574 other than @file{grape/concord}.
7577 @section Crossing File System Boundaries
7578 @cindex File system boundaries, not crossing
7581 @command{tar} will normally automatically cross file system boundaries in
7582 order to archive files which are part of a directory tree. You can
7583 change this behavior by running @command{tar} and specifying
7584 @option{--one-file-system}. This option only affects files that are
7585 archived because they are in a directory that is being archived;
7586 @command{tar} will still archive files explicitly named on the command line
7587 or through @option{--files-from}, regardless of where they reside.
7590 @opindex one-file-system
7591 @item --one-file-system
7592 Prevents @command{tar} from crossing file system boundaries when
7593 archiving. Use in conjunction with any write operation.
7596 The @option{--one-file-system} option causes @command{tar} to modify its
7597 normal behavior in archiving the contents of directories. If a file in
7598 a directory is not on the same file system as the directory itself, then
7599 @command{tar} will not archive that file. If the file is a directory
7600 itself, @command{tar} will not archive anything beneath it; in other words,
7601 @command{tar} will not cross mount points.
7603 This option is useful for making full or incremental archival backups of
7604 a file system. If this option is used in conjunction with
7605 @option{--verbose} (@option{-v}), files that are excluded are
7606 mentioned by name on the standard error.
7609 * directory:: Changing Directory
7610 * absolute:: Absolute File Names
7614 @subsection Changing the Working Directory
7616 @FIXME{need to read over this node now for continuity; i've switched
7617 things around some.}
7619 @cindex Changing directory mid-stream
7620 @cindex Directory, changing mid-stream
7621 @cindex Working directory, specifying
7622 To change the working directory in the middle of a list of file names,
7623 either on the command line or in a file specified using
7624 @option{--files-from} (@option{-T}), use @option{--directory} (@option{-C}).
7625 This will change the working directory to the specified directory
7626 after that point in the list.
7630 @item --directory=@var{directory}
7631 @itemx -C @var{directory}
7632 Changes the working directory in the middle of a command line.
7638 $ @kbd{tar -c -f jams.tar grape prune -C food cherry}
7642 will place the files @file{grape} and @file{prune} from the current
7643 directory into the archive @file{jams.tar}, followed by the file
7644 @file{cherry} from the directory @file{food}. This option is especially
7645 useful when you have several widely separated files that you want to
7646 store in the same archive.
7648 Note that the file @file{cherry} is recorded in the archive under the
7649 precise name @file{cherry}, @emph{not} @file{food/cherry}. Thus, the
7650 archive will contain three files that all appear to have come from the
7651 same directory; if the archive is extracted with plain @samp{tar
7652 --extract}, all three files will be written in the current directory.
7654 Contrast this with the command,
7657 $ @kbd{tar -c -f jams.tar grape prune -C food red/cherry}
7661 which records the third file in the archive under the name
7662 @file{red/cherry} so that, if the archive is extracted using
7663 @samp{tar --extract}, the third file will be written in a subdirectory
7664 named @file{orange-colored}.
7666 You can use the @option{--directory} option to make the archive
7667 independent of the original name of the directory holding the files.
7668 The following command places the files @file{/etc/passwd},
7669 @file{/etc/hosts}, and @file{/lib/libc.a} into the archive
7673 $ @kbd{tar -c -f foo.tar -C /etc passwd hosts -C /lib libc.a}
7677 However, the names of the archive members will be exactly what they were
7678 on the command line: @file{passwd}, @file{hosts}, and @file{libc.a}.
7679 They will not appear to be related by file name to the original
7680 directories where those files were located.
7682 Note that @option{--directory} options are interpreted consecutively. If
7683 @option{--directory} specifies a relative file name, it is interpreted
7684 relative to the then current directory, which might not be the same as
7685 the original current working directory of @command{tar}, due to a previous
7686 @option{--directory} option.
7688 When using @option{--files-from} (@pxref{files}), you can put various
7689 @command{tar} options (including @option{-C}) in the file list. Notice,
7690 however, that in this case the option and its argument may not be
7691 separated by whitespace. If you use short option, its argument must
7692 either follow the option letter immediately, without any intervening
7693 whitespace, or occupy the next line. Otherwise, if you use long
7694 option, separate its argument by an equal sign.
7696 For instance, the file list for the above example will be:
7709 To use it, you would invoke @command{tar} as follows:
7712 $ @kbd{tar -c -f foo.tar --files-from list}
7715 The interpretation of @option{--directory} is disabled by
7716 @option{--null} option.
7719 @subsection Absolute File Names
7723 @opindex absolute-names
7724 @item --absolute-names
7726 Do not strip leading slashes from file names, and permit file names
7727 containing a @file{..} file name component.
7730 By default, @GNUTAR{} drops a leading @samp{/} on
7731 input or output, and complains about file names containing a @file{..}
7732 component. This option turns off this behavior.
7734 When @command{tar} extracts archive members from an archive, it strips any
7735 leading slashes (@samp{/}) from the member name. This causes absolute
7736 member names in the archive to be treated as relative file names. This
7737 allows you to have such members extracted wherever you want, instead of
7738 being restricted to extracting the member in the exact directory named
7739 in the archive. For example, if the archive member has the name
7740 @file{/etc/passwd}, @command{tar} will extract it as if the name were
7741 really @file{etc/passwd}.
7743 File names containing @file{..} can cause problems when extracting, so
7744 @command{tar} normally warns you about such files when creating an
7745 archive, and rejects attempts to extracts such files.
7747 Other @command{tar} programs do not do this. As a result, if you
7748 create an archive whose member names start with a slash, they will be
7749 difficult for other people with a non-@GNUTAR{}
7750 program to use. Therefore, @GNUTAR{} also strips
7751 leading slashes from member names when putting members into the
7752 archive. For example, if you ask @command{tar} to add the file
7753 @file{/bin/ls} to an archive, it will do so, but the member name will
7754 be @file{bin/ls}.@footnote{A side effect of this is that when
7755 @option{--create} is used with @option{--verbose} the resulting output
7756 is not, generally speaking, the same as the one you'd get running
7757 @kbd{tar --list} command. This may be important if you use some
7758 scripts for comparing both outputs. @xref{listing member and file names},
7759 for the information on how to handle this case.}
7761 If you use the @option{--absolute-names} (@option{-P}) option,
7762 @command{tar} will do none of these transformations.
7764 To archive or extract files relative to the root directory, specify
7765 the @option{--absolute-names} (@option{-P}) option.
7767 Normally, @command{tar} acts on files relative to the working
7768 directory---ignoring superior directory names when archiving, and
7769 ignoring leading slashes when extracting.
7771 When you specify @option{--absolute-names} (@option{-P}),
7772 @command{tar} stores file names including all superior directory
7773 names, and preserves leading slashes. If you only invoked
7774 @command{tar} from the root directory you would never need the
7775 @option{--absolute-names} option, but using this option
7776 may be more convenient than switching to root.
7778 @FIXME{Should be an example in the tutorial/wizardry section using this
7779 to transfer files between systems.}
7781 @FIXME{Is write access an issue?}
7784 @item --absolute-names
7785 Preserves full file names (including superior directory names) when
7786 archiving files. Preserves leading slash when extracting files.
7790 @FIXME{this is still horrible; need to talk with dan on monday.}
7792 @command{tar} prints out a message about removing the @samp{/} from
7793 file names. This message appears once per @GNUTAR{}
7794 invocation. It represents something which ought to be told; ignoring
7795 what it means can cause very serious surprises, later.
7797 Some people, nevertheless, do not want to see this message. Wanting to
7798 play really dangerously, one may of course redirect @command{tar} standard
7799 error to the sink. For example, under @command{sh}:
7802 $ @kbd{tar -c -f archive.tar /home 2> /dev/null}
7806 Another solution, both nicer and simpler, would be to change to
7807 the @file{/} directory first, and then avoid absolute notation.
7811 $ @kbd{(cd / && tar -c -f archive.tar home)}
7813 $ @kbd{tar -c -f archive.tar -C / home}
7816 @include getdate.texi
7819 @chapter Controlling the Archive Format
7821 @cindex Tar archive formats
7822 Due to historical reasons, there are several formats of tar archives.
7823 All of them are based on the same principles, but have some subtle
7824 differences that often make them incompatible with each other.
7826 GNU tar is able to create and handle archives in a variety of formats.
7827 The most frequently used formats are (in alphabetical order):
7831 Format used by @GNUTAR{} versions up to 1.13.25. This format derived
7832 from an early @acronym{POSIX} standard, adding some improvements such as
7833 sparse file handling and incremental archives. Unfortunately these
7834 features were implemented in a way incompatible with other archive
7837 Archives in @samp{gnu} format are able to hold file names of unlimited
7841 Format used by @GNUTAR{} of versions prior to 1.12.
7844 Archive format, compatible with the V7 implementation of tar. This
7845 format imposes a number of limitations. The most important of them
7849 @item The maximum length of a file name is limited to 99 characters.
7850 @item The maximum length of a symbolic link is limited to 99 characters.
7851 @item It is impossible to store special files (block and character
7852 devices, fifos etc.)
7853 @item Maximum value of user or group @acronym{ID} is limited to 2097151 (7777777
7855 @item V7 archives do not contain symbolic ownership information (user
7856 and group name of the file owner).
7859 This format has traditionally been used by Automake when producing
7860 Makefiles. This practice will change in the future, in the meantime,
7861 however this means that projects containing file names more than 99
7862 characters long will not be able to use @GNUTAR{} @value{VERSION} and
7863 Automake prior to 1.9.
7866 Archive format defined by @acronym{POSIX.1-1988} specification. It stores
7867 symbolic ownership information. It is also able to store
7868 special files. However, it imposes several restrictions as well:
7871 @item The maximum length of a file name is limited to 256 characters,
7872 provided that the file name can be split at a directory separator in
7873 two parts, first of them being at most 155 bytes long. So, in most
7874 cases the maximum file name length will be shorter than 256
7876 @item The maximum length of a symbolic link name is limited to
7878 @item Maximum size of a file the archive is able to accommodate
7880 @item Maximum value of UID/GID is 2097151.
7881 @item Maximum number of bits in device major and minor numbers is 21.
7885 Format used by J@"org Schilling @command{star}
7886 implementation. @GNUTAR{} is able to read @samp{star} archives but
7887 currently does not produce them.
7890 Archive format defined by @acronym{POSIX.1-2001} specification. This is the
7891 most flexible and feature-rich format. It does not impose any
7892 restrictions on file sizes or file name lengths. This format is quite
7893 recent, so not all tar implementations are able to handle it properly.
7894 However, this format is designed in such a way that any tar
7895 implementation able to read @samp{ustar} archives will be able to read
7896 most @samp{posix} archives as well, with the only exception that any
7897 additional information (such as long file names etc.) will in such
7898 case be extracted as plain text files along with the files it refers to.
7900 This archive format will be the default format for future versions
7905 The following table summarizes the limitations of each of these
7908 @multitable @columnfractions .10 .20 .20 .20 .20
7909 @headitem Format @tab UID @tab File Size @tab File Name @tab Devn
7910 @item gnu @tab 1.8e19 @tab Unlimited @tab Unlimited @tab 63
7911 @item oldgnu @tab 1.8e19 @tab Unlimited @tab Unlimited @tab 63
7912 @item v7 @tab 2097151 @tab 8GB @tab 99 @tab n/a
7913 @item ustar @tab 2097151 @tab 8GB @tab 256 @tab 21
7914 @item posix @tab Unlimited @tab Unlimited @tab Unlimited @tab Unlimited
7917 The default format for @GNUTAR{} is defined at compilation
7918 time. You may check it by running @command{tar --help}, and examining
7919 the last lines of its output. Usually, @GNUTAR{} is configured
7920 to create archives in @samp{gnu} format, however, future version will
7921 switch to @samp{posix}.
7924 * Compression:: Using Less Space through Compression
7925 * Attributes:: Handling File Attributes
7926 * Portability:: Making @command{tar} Archives More Portable
7927 * cpio:: Comparison of @command{tar} and @command{cpio}
7931 @section Using Less Space through Compression
7934 * gzip:: Creating and Reading Compressed Archives
7935 * sparse:: Archiving Sparse Files
7939 @subsection Creating and Reading Compressed Archives
7940 @cindex Compressed archives
7941 @cindex Storing archives in compressed format
7943 @GNUTAR{} is able to create and read compressed archives. It supports
7944 @command{gzip} and @command{bzip2} compression programs. For backward
7945 compatibility, it also supports @command{compress} command, although
7946 we strongly recommend against using it, since there is a patent
7947 covering the algorithm it uses and you could be sued for patent
7948 infringement merely by running @command{compress}! Besides, it is less
7949 effective than @command{gzip} and @command{bzip2}.
7951 Creating a compressed archive is simple: you just specify a
7952 @dfn{compression option} along with the usual archive creation
7953 commands. The compression option is @option{-z} (@option{--gzip}) to
7954 create a @command{gzip} compressed archive, @option{-j}
7955 (@option{--bzip2}) to create a @command{bzip2} compressed archive, and
7956 @option{-Z} (@option{--compress}) to use @command{compress} program.
7960 $ @kbd{tar cfz archive.tar.gz .}
7963 Reading compressed archive is even simpler: you don't need to specify
7964 any additional options as @GNUTAR{} recognizes its format
7965 automatically. Thus, the following commands will list and extract the
7966 archive created in previous example:
7969 # List the compressed archive
7970 $ @kbd{tar tf archive.tar.gz}
7971 # Extract the compressed archive
7972 $ @kbd{tar xf archive.tar.gz}
7975 The only case when you have to specify a decompression option while
7976 reading the archive is when reading from a pipe or from a tape drive
7977 that does not support random access. However, in this case @GNUTAR{}
7978 will indicate which option you should use. For example:
7981 $ @kbd{cat archive.tar.gz | tar tf -}
7982 tar: Archive is compressed. Use -z option
7983 tar: Error is not recoverable: exiting now
7986 If you see such diagnostics, just add the suggested option to the
7987 invocation of @GNUTAR{}:
7990 $ @kbd{cat archive.tar.gz | tar tfz -}
7993 Notice also, that there are several restrictions on operations on
7994 compressed archives. First of all, compressed archives cannot be
7995 modified, i.e., you cannot update (@option{--update} (@option{-u})) them or delete
7996 (@option{--delete}) members from them. Likewise, you cannot append
7997 another @command{tar} archive to a compressed archive using
7998 @option{--append} (@option{-r})). Secondly, multi-volume archives cannot be
8001 The following table summarizes compression options used by @GNUTAR{}.
8009 Filter the archive through @command{gzip}.
8011 You can use @option{--gzip} and @option{--gunzip} on physical devices
8012 (tape drives, etc.) and remote files as well as on normal files; data
8013 to or from such devices or remote files is reblocked by another copy
8014 of the @command{tar} program to enforce the specified (or default) record
8015 size. The default compression parameters are used; if you need to
8016 override them, set @env{GZIP} environment variable, e.g.:
8019 $ @kbd{GZIP=--best tar cfz archive.tar.gz subdir}
8023 Another way would be to avoid the @option{--gzip} (@option{--gunzip}, @option{--ungzip}, @option{-z}) option and run
8024 @command{gzip} explicitly:
8027 $ @kbd{tar cf - subdir | gzip --best -c - > archive.tar.gz}
8030 @cindex corrupted archives
8031 About corrupted compressed archives: @command{gzip}'ed files have no
8032 redundancy, for maximum compression. The adaptive nature of the
8033 compression scheme means that the compression tables are implicitly
8034 spread all over the archive. If you lose a few blocks, the dynamic
8035 construction of the compression tables becomes unsynchronized, and there
8036 is little chance that you could recover later in the archive.
8038 There are pending suggestions for having a per-volume or per-file
8039 compression in @GNUTAR{}. This would allow for viewing the
8040 contents without decompression, and for resynchronizing decompression at
8041 every volume or file, in case of corrupted archives. Doing so, we might
8042 lose some compressibility. But this would have make recovering easier.
8043 So, there are pros and cons. We'll see!
8048 Filter the archive through @code{bzip2}. Otherwise like @option{--gzip}.
8055 Filter the archive through @command{compress}. Otherwise like @option{--gzip}.
8057 The @acronym{GNU} Project recommends you not use
8058 @command{compress}, because there is a patent covering the algorithm it
8059 uses. You could be sued for patent infringement merely by running
8062 @opindex use-compress-program
8063 @item --use-compress-program=@var{prog}
8064 Use external compression program @var{prog}. Use this option if you
8065 have a compression program that @GNUTAR{} does not support. There
8066 are two requirements to which @var{prog} should comply:
8068 First, when called without options, it should read data from standard
8069 input, compress it and output it on standard output.
8071 Secondly, if called with @option{-d} argument, it should do exactly
8072 the opposite, i.e., read the compressed data from the standard input
8073 and produce uncompressed data on the standard output.
8076 @cindex gpg, using with tar
8077 @cindex gnupg, using with tar
8078 @cindex Using encrypted archives
8079 The @option{--use-compress-program} option, in particular, lets you
8080 implement your own filters, not necessarily dealing with
8081 compression/decompression. For example, suppose you wish to implement
8082 PGP encryption on top of compression, using @command{gpg} (@pxref{Top,
8083 gpg, gpg ---- encryption and signing tool, gpg, GNU Privacy Guard
8084 Manual}). The following script does that:
8090 -d) gpg --decrypt - | gzip -d -c;;
8091 '') gzip -c | gpg -s ;;
8092 *) echo "Unknown option $1">&2; exit 1;;
8097 Suppose you name it @file{gpgz} and save it somewhere in your
8098 @env{PATH}. Then the following command will create a compressed
8099 archive signed with your private key:
8102 $ @kbd{tar -cf foo.tar.gpgz --use-compress=gpgz .}
8106 Likewise, the following command will list its contents:
8109 $ @kbd{tar -tf foo.tar.gpgz --use-compress=gpgz .}
8113 The above is based on the following discussion:
8115 I have one question, or maybe it's a suggestion if there isn't a way
8116 to do it now. I would like to use @option{--gzip}, but I'd also like
8117 the output to be fed through a program like @acronym{GNU}
8118 @command{ecc} (actually, right now that's @samp{exactly} what I'd like
8119 to use :-)), basically adding ECC protection on top of compression.
8120 It seems as if this should be quite easy to do, but I can't work out
8121 exactly how to go about it. Of course, I can pipe the standard output
8122 of @command{tar} through @command{ecc}, but then I lose (though I
8123 haven't started using it yet, I confess) the ability to have
8124 @command{tar} use @command{rmt} for it's I/O (I think).
8126 I think the most straightforward thing would be to let me specify a
8127 general set of filters outboard of compression (preferably ordered,
8128 so the order can be automatically reversed on input operations, and
8129 with the options they require specifiable), but beggars shouldn't be
8130 choosers and anything you decide on would be fine with me.
8132 By the way, I like @command{ecc} but if (as the comments say) it can't
8133 deal with loss of block sync, I'm tempted to throw some time at adding
8134 that capability. Supposing I were to actually do such a thing and
8135 get it (apparently) working, do you accept contributed changes to
8136 utilities like that? (Leigh Clayton @file{loc@@soliton.com}, May 1995).
8138 Isn't that exactly the role of the
8139 @option{--use-compress-prog=@var{program}} option?
8140 I never tried it myself, but I suspect you may want to write a
8141 @var{prog} script or program able to filter stdin to stdout to
8142 way you want. It should recognize the @option{-d} option, for when
8143 extraction is needed rather than creation.
8145 It has been reported that if one writes compressed data (through the
8146 @option{--gzip} or @option{--compress} options) to a DLT and tries to use
8147 the DLT compression mode, the data will actually get bigger and one will
8148 end up with less space on the tape.
8152 @subsection Archiving Sparse Files
8153 @cindex Sparse Files
8155 Files in the file system occasionally have @dfn{holes}. A @dfn{hole}
8156 in a file is a section of the file's contents which was never written.
8157 The contents of a hole reads as all zeros. On many operating systems,
8158 actual disk storage is not allocated for holes, but they are counted
8159 in the length of the file. If you archive such a file, @command{tar}
8160 could create an archive longer than the original. To have @command{tar}
8161 attempt to recognize the holes in a file, use @option{--sparse}
8162 (@option{-S}). When you use this option, then, for any file using
8163 less disk space than would be expected from its length, @command{tar}
8164 searches the file for consecutive stretches of zeros. It then records
8165 in the archive for the file where the consecutive stretches of zeros
8166 are, and only archives the ``real contents'' of the file. On
8167 extraction (using @option{--sparse} is not needed on extraction) any
8168 such files have holes created wherever the continuous stretches of zeros
8169 were found. Thus, if you use @option{--sparse}, @command{tar} archives
8170 won't take more space than the original.
8176 This option instructs @command{tar} to test each file for sparseness
8177 before attempting to archive it. If the file is found to be sparse it
8178 is treated specially, thus allowing to decrease the amount of space
8179 used by its image in the archive.
8181 This option is meaningful only when creating or updating archives. It
8182 has no effect on extraction.
8185 Consider using @option{--sparse} when performing file system backups,
8186 to avoid archiving the expanded forms of files stored sparsely in the
8189 Even if your system has no sparse files currently, some may be
8190 created in the future. If you use @option{--sparse} while making file
8191 system backups as a matter of course, you can be assured the archive
8192 will never take more space on the media than the files take on disk
8193 (otherwise, archiving a disk filled with sparse files might take
8194 hundreds of tapes). @xref{Incremental Dumps}.
8196 However, be aware that @option{--sparse} option presents a serious
8197 drawback. Namely, in order to determine if the file is sparse
8198 @command{tar} has to read it before trying to archive it, so in total
8199 the file is read @strong{twice}. So, always bear in mind that the
8200 time needed to process all files with this option is roughly twice
8201 the time needed to archive them without it.
8202 @FIXME{A technical note:
8204 Programs like @command{dump} do not have to read the entire file; by
8205 examining the file system directly, they can determine in advance
8206 exactly where the holes are and thus avoid reading through them. The
8207 only data it need read are the actual allocated data blocks.
8208 @GNUTAR{} uses a more portable and straightforward
8209 archiving approach, it would be fairly difficult that it does
8210 otherwise. Elizabeth Zwicky writes to @file{comp.unix.internals}, on
8214 What I did say is that you cannot tell the difference between a hole and an
8215 equivalent number of nulls without reading raw blocks. @code{st_blocks} at
8216 best tells you how many holes there are; it doesn't tell you @emph{where}.
8217 Just as programs may, conceivably, care what @code{st_blocks} is (care
8218 to name one that does?), they may also care where the holes are (I have
8219 no examples of this one either, but it's equally imaginable).
8221 I conclude from this that good archivers are not portable. One can
8222 arguably conclude that if you want a portable program, you can in good
8223 conscience restore files with as many holes as possible, since you can't
8228 @cindex sparse formats, defined
8229 When using @samp{POSIX} archive format, @GNUTAR{} is able to store
8230 sparse files using in three distinct ways, called @dfn{sparse
8231 formats}. A sparse format is identified by its @dfn{number},
8232 consisting, as usual of two decimal numbers, delimited by a dot. By
8233 default, format @samp{1.0} is used. If, for some reason, you wish to
8234 use an earlier format, you can select it using
8235 @option{--sparse-version} option.
8238 @opindex sparse-version
8239 @item --sparse-version=@var{version}
8241 Select the format to store sparse files in. Valid @var{version} values
8242 are: @samp{0.0}, @samp{0.1} and @samp{1.0}. @xref{Sparse Formats},
8243 for a detailed description of each format.
8246 Using @option{--sparse-format} option implies @option{--sparse}.
8249 @section Handling File Attributes
8252 When @command{tar} reads files, it updates their access times. To
8253 avoid this, use the @option{--atime-preserve[=METHOD]} option, which can either
8254 reset the access time retroactively or avoid changing it in the first
8257 Handling of file attributes
8260 @opindex atime-preserve
8261 @item --atime-preserve
8262 @itemx --atime-preserve=replace
8263 @itemx --atime-preserve=system
8264 Preserve the access times of files that are read. This works only for
8265 files that you own, unless you have superuser privileges.
8267 @option{--atime-preserve=replace} works on most systems, but it also
8268 restores the data modification time and updates the status change
8269 time. Hence it doesn't interact with incremental dumps nicely
8270 (@pxref{Incremental Dumps}), and it can set access or data modification times
8271 incorrectly if other programs access the file while @command{tar} is
8274 @option{--atime-preserve=system} avoids changing the access time in
8275 the first place, if the operating system supports this.
8276 Unfortunately, this may or may not work on any given operating system
8277 or file system. If @command{tar} knows for sure it won't work, it
8278 complains right away.
8280 Currently @option{--atime-preserve} with no operand defaults to
8281 @option{--atime-preserve=replace}, but this is intended to change to
8282 @option{--atime-preserve=system} when the latter is better-supported.
8287 Do not extract data modification time.
8289 When this option is used, @command{tar} leaves the data modification times
8290 of the files it extracts as the times when the files were extracted,
8291 instead of setting it to the times recorded in the archive.
8293 This option is meaningless with @option{--list} (@option{-t}).
8297 Create extracted files with the same ownership they have in the
8300 This is the default behavior for the superuser,
8301 so this option is meaningful only for non-root users, when @command{tar}
8302 is executed on those systems able to give files away. This is
8303 considered as a security flaw by many people, at least because it
8304 makes quite difficult to correctly account users for the disk space
8305 they occupy. Also, the @code{suid} or @code{sgid} attributes of
8306 files are easily and silently lost when files are given away.
8308 When writing an archive, @command{tar} writes the user @acronym{ID} and user name
8309 separately. If it can't find a user name (because the user @acronym{ID} is not
8310 in @file{/etc/passwd}), then it does not write one. When restoring,
8311 it tries to look the name (if one was written) up in
8312 @file{/etc/passwd}. If it fails, then it uses the user @acronym{ID} stored in
8313 the archive instead.
8315 @opindex no-same-owner
8316 @item --no-same-owner
8318 Do not attempt to restore ownership when extracting. This is the
8319 default behavior for ordinary users, so this option has an effect
8320 only for the superuser.
8322 @opindex numeric-owner
8323 @item --numeric-owner
8324 The @option{--numeric-owner} option allows (ANSI) archives to be written
8325 without user/group name information or such information to be ignored
8326 when extracting. It effectively disables the generation and/or use
8327 of user/group name information. This option forces extraction using
8328 the numeric ids from the archive, ignoring the names.
8330 This is useful in certain circumstances, when restoring a backup from
8331 an emergency floppy with different passwd/group files for example.
8332 It is otherwise impossible to extract files with the right ownerships
8333 if the password file in use during the extraction does not match the
8334 one belonging to the file system(s) being extracted. This occurs,
8335 for example, if you are restoring your files after a major crash and
8336 had booted from an emergency floppy with no password file or put your
8337 disk into another machine to do the restore.
8339 The numeric ids are @emph{always} saved into @command{tar} archives.
8340 The identifying names are added at create time when provided by the
8341 system, unless @option{--old-archive} (@option{-o}) is used. Numeric ids could be
8342 used when moving archives between a collection of machines using
8343 a centralized management for attribution of numeric ids to users
8344 and groups. This is often made through using the NIS capabilities.
8346 When making a @command{tar} file for distribution to other sites, it
8347 is sometimes cleaner to use a single owner for all files in the
8348 distribution, and nicer to specify the write permission bits of the
8349 files as stored in the archive independently of their actual value on
8350 the file system. The way to prepare a clean distribution is usually
8351 to have some Makefile rule creating a directory, copying all needed
8352 files in that directory, then setting ownership and permissions as
8353 wanted (there are a lot of possible schemes), and only then making a
8354 @command{tar} archive out of this directory, before cleaning
8355 everything out. Of course, we could add a lot of options to
8356 @GNUTAR{} for fine tuning permissions and ownership.
8357 This is not the good way, I think. @GNUTAR{} is
8358 already crowded with options and moreover, the approach just explained
8359 gives you a great deal of control already.
8361 @xopindex{same-permissions, short description}
8362 @xopindex{preserve-permissions, short description}
8364 @itemx --same-permissions
8365 @itemx --preserve-permissions
8366 Extract all protection information.
8368 This option causes @command{tar} to set the modes (access permissions) of
8369 extracted files exactly as recorded in the archive. If this option
8370 is not used, the current @code{umask} setting limits the permissions
8371 on extracted files. This option is by default enabled when
8372 @command{tar} is executed by a superuser.
8375 This option is meaningless with @option{--list} (@option{-t}).
8379 Same as both @option{--same-permissions} and @option{--same-order}.
8381 The @option{--preserve} option has no equivalent short option name.
8382 It is equivalent to @option{--same-permissions} plus @option{--same-order}.
8384 @FIXME{I do not see the purpose of such an option. (Neither I. FP.)
8385 Neither do I. --Sergey}
8390 @section Making @command{tar} Archives More Portable
8392 Creating a @command{tar} archive on a particular system that is meant to be
8393 useful later on many other machines and with other versions of @command{tar}
8394 is more challenging than you might think. @command{tar} archive formats
8395 have been evolving since the first versions of Unix. Many such formats
8396 are around, and are not always compatible with each other. This section
8397 discusses a few problems, and gives some advice about making @command{tar}
8398 archives more portable.
8400 One golden rule is simplicity. For example, limit your @command{tar}
8401 archives to contain only regular files and directories, avoiding
8402 other kind of special files. Do not attempt to save sparse files or
8403 contiguous files as such. Let's discuss a few more problems, in turn.
8405 @FIXME{Discuss GNU extensions (incremental backups, multi-volume
8406 archives and archive labels) in GNU and PAX formats.}
8409 * Portable Names:: Portable Names
8410 * dereference:: Symbolic Links
8411 * old:: Old V7 Archives
8412 * ustar:: Ustar Archives
8413 * gnu:: GNU and old GNU format archives.
8414 * posix:: @acronym{POSIX} archives
8415 * Checksumming:: Checksumming Problems
8416 * Large or Negative Values:: Large files, negative time stamps, etc.
8417 * Other Tars:: How to Extract GNU-Specific Data Using
8418 Other @command{tar} Implementations
8421 @node Portable Names
8422 @subsection Portable Names
8424 Use portable file and member names. A name is portable if it contains
8425 only @acronym{ASCII} letters and digits, @samp{/}, @samp{.}, @samp{_}, and
8426 @samp{-}; it cannot be empty, start with @samp{-} or @samp{//}, or
8427 contain @samp{/-}. Avoid deep directory nesting. For portability to
8428 old Unix hosts, limit your file name components to 14 characters or
8431 If you intend to have your @command{tar} archives to be read under
8432 MSDOS, you should not rely on case distinction for file names, and you
8433 might use the @acronym{GNU} @command{doschk} program for helping you
8434 further diagnosing illegal MSDOS names, which are even more limited
8438 @subsection Symbolic Links
8439 @cindex File names, using symbolic links
8440 @cindex Symbolic link as file name
8442 @opindex dereference
8443 Normally, when @command{tar} archives a symbolic link, it writes a
8444 block to the archive naming the target of the link. In that way, the
8445 @command{tar} archive is a faithful record of the file system contents.
8446 @option{--dereference} (@option{-h}) is used with @option{--create} (@option{-c}), and causes
8447 @command{tar} to archive the files symbolic links point to, instead of
8448 the links themselves. When this option is used, when @command{tar}
8449 encounters a symbolic link, it will archive the linked-to file,
8450 instead of simply recording the presence of a symbolic link.
8452 The name under which the file is stored in the file system is not
8453 recorded in the archive. To record both the symbolic link name and
8454 the file name in the system, archive the file under both names. If
8455 all links were recorded automatically by @command{tar}, an extracted file
8456 might be linked to a file name that no longer exists in the file
8459 If a linked-to file is encountered again by @command{tar} while creating
8460 the same archive, an entire second copy of it will be stored. (This
8461 @emph{might} be considered a bug.)
8463 So, for portable archives, do not archive symbolic links as such,
8464 and use @option{--dereference} (@option{-h}): many systems do not support
8465 symbolic links, and moreover, your distribution might be unusable if
8466 it contains unresolved symbolic links.
8469 @subsection Old V7 Archives
8470 @cindex Format, old style
8471 @cindex Old style format
8472 @cindex Old style archives
8473 @cindex v7 archive format
8475 Certain old versions of @command{tar} cannot handle additional
8476 information recorded by newer @command{tar} programs. To create an
8477 archive in V7 format (not ANSI), which can be read by these old
8478 versions, specify the @option{--format=v7} option in
8479 conjunction with the @option{--create} (@option{-c}) (@command{tar} also
8480 accepts @option{--portability} or @option{--old-archive} for this
8481 option). When you specify it,
8482 @command{tar} leaves out information about directories, pipes, fifos,
8483 contiguous files, and device files, and specifies file ownership by
8484 group and user IDs instead of group and user names.
8486 When updating an archive, do not use @option{--format=v7}
8487 unless the archive was created using this option.
8489 In most cases, a @emph{new} format archive can be read by an @emph{old}
8490 @command{tar} program without serious trouble, so this option should
8491 seldom be needed. On the other hand, most modern @command{tar}s are
8492 able to read old format archives, so it might be safer for you to
8493 always use @option{--format=v7} for your distributions. Notice,
8494 however, that @samp{ustar} format is a better alternative, as it is
8495 free from many of @samp{v7}'s drawbacks.
8498 @subsection Ustar Archive Format
8500 @cindex ustar archive format
8501 Archive format defined by @acronym{POSIX}.1-1988 specification is called
8502 @code{ustar}. Although it is more flexible than the V7 format, it
8503 still has many restrictions (@xref{Formats,ustar}, for the detailed
8504 description of @code{ustar} format). Along with V7 format,
8505 @code{ustar} format is a good choice for archives intended to be read
8506 with other implementations of @command{tar}.
8508 To create archive in @code{ustar} format, use @option{--format=ustar}
8509 option in conjunction with the @option{--create} (@option{-c}).
8512 @subsection @acronym{GNU} and old @GNUTAR{} format
8514 @cindex GNU archive format
8515 @cindex Old GNU archive format
8516 @GNUTAR{} was based on an early draft of the
8517 @acronym{POSIX} 1003.1 @code{ustar} standard. @acronym{GNU} extensions to
8518 @command{tar}, such as the support for file names longer than 100
8519 characters, use portions of the @command{tar} header record which were
8520 specified in that @acronym{POSIX} draft as unused. Subsequent changes in
8521 @acronym{POSIX} have allocated the same parts of the header record for
8522 other purposes. As a result, @GNUTAR{} format is
8523 incompatible with the current @acronym{POSIX} specification, and with
8524 @command{tar} programs that follow it.
8526 In the majority of cases, @command{tar} will be configured to create
8527 this format by default. This will change in future releases, since
8528 we plan to make @samp{POSIX} format the default.
8530 To force creation a @GNUTAR{} archive, use option
8531 @option{--format=gnu}.
8534 @subsection @GNUTAR{} and @acronym{POSIX} @command{tar}
8536 @cindex POSIX archive format
8537 @cindex PAX archive format
8538 Starting from version 1.14 @GNUTAR{} features full support for
8539 @acronym{POSIX.1-2001} archives.
8541 A @acronym{POSIX} conformant archive will be created if @command{tar}
8542 was given @option{--format=posix} (@option{--format=pax}) option. No
8543 special option is required to read and extract from a @acronym{POSIX}
8547 * PAX keywords:: Controlling Extended Header Keywords.
8551 @subsubsection Controlling Extended Header Keywords
8555 @item --pax-option=@var{keyword-list}
8556 Handle keywords in @acronym{PAX} extended headers. This option is
8557 equivalent to @option{-o} option of the @command{pax} utility.
8560 @var{Keyword-list} is a comma-separated
8561 list of keyword options, each keyword option taking one of
8562 the following forms:
8565 @item delete=@var{pattern}
8566 When used with one of archive-creation commands,
8567 this option instructs @command{tar} to omit from extended header records
8568 that it produces any keywords matching the string @var{pattern}.
8570 When used in extract or list mode, this option instructs tar
8571 to ignore any keywords matching the given @var{pattern} in the extended
8572 header records. In both cases, matching is performed using the pattern
8573 matching notation described in @acronym{POSIX 1003.2}, 3.13
8574 (@pxref{wildcards}). For example:
8577 --pax-option delete=security.*
8580 would suppress security-related information.
8582 @item exthdr.name=@var{string}
8584 This keyword allows user control over the name that is written into the
8585 ustar header blocks for the extended headers. The name is obtained
8586 from @var{string} after making the following substitutions:
8588 @multitable @columnfractions .25 .55
8589 @headitem Meta-character @tab Replaced By
8590 @item %d @tab The directory name of the file, equivalent to the
8591 result of the @command{dirname} utility on the translated file name.
8592 @item %f @tab The name of the file with the directory information
8593 stripped, equivalent to the result of the @command{basename} utility
8594 on the translated file name.
8595 @item %p @tab The process @acronym{ID} of the @command{tar} process.
8596 @item %% @tab A @samp{%} character.
8599 Any other @samp{%} characters in @var{string} produce undefined
8602 If no option @samp{exthdr.name=string} is specified, @command{tar}
8603 will use the following default value:
8609 @item globexthdr.name=@var{string}
8610 This keyword allows user control over the name that is written into
8611 the ustar header blocks for global extended header records. The name
8612 is obtained from the contents of @var{string}, after making
8613 the following substitutions:
8615 @multitable @columnfractions .25 .55
8616 @headitem Meta-character @tab Replaced By
8617 @item %n @tab An integer that represents the
8618 sequence number of the global extended header record in the archive,
8620 @item %p @tab The process @acronym{ID} of the @command{tar} process.
8621 @item %% @tab A @samp{%} character.
8624 Any other @samp{%} characters in @var{string} produce undefined results.
8626 If no option @samp{globexthdr.name=string} is specified, @command{tar}
8627 will use the following default value:
8630 $TMPDIR/GlobalHead.%p.%n
8634 where @samp{$TMPDIR} represents the value of the @var{TMPDIR}
8635 environment variable. If @var{TMPDIR} is not set, @command{tar}
8638 @item @var{keyword}=@var{value}
8639 When used with one of archive-creation commands, these keyword/value pairs
8640 will be included at the beginning of the archive in a global extended
8641 header record. When used with one of archive-reading commands,
8642 @command{tar} will behave as if it has encountered these keyword/value
8643 pairs at the beginning of the archive in a global extended header
8646 @item @var{keyword}:=@var{value}
8647 When used with one of archive-creation commands, these keyword/value pairs
8648 will be included as records at the beginning of an extended header for
8649 each file. This is effectively equivalent to @var{keyword}=@var{value}
8650 form except that it creates no global extended header records.
8652 When used with one of archive-reading commands, @command{tar} will
8653 behave as if these keyword/value pairs were included as records at the
8654 end of each extended header; thus, they will override any global or
8655 file-specific extended header record keywords of the same names.
8656 For example, in the command:
8659 tar --format=posix --create \
8660 --file archive --pax-option gname:=user .
8663 the group name will be forced to a new value for all files
8664 stored in the archive.
8668 @subsection Checksumming Problems
8670 SunOS and HP-UX @command{tar} fail to accept archives created using
8671 @GNUTAR{} and containing non-@acronym{ASCII} file names, that
8672 is, file names having characters with the eight bit set, because they
8673 use signed checksums, while @GNUTAR{} uses unsigned
8674 checksums while creating archives, as per @acronym{POSIX} standards. On
8675 reading, @GNUTAR{} computes both checksums and
8676 accept any. It is somewhat worrying that a lot of people may go
8677 around doing backup of their files using faulty (or at least
8678 non-standard) software, not learning about it until it's time to
8679 restore their missing files with an incompatible file extractor, or
8682 @GNUTAR{} compute checksums both ways, and accept
8683 any on read, so @acronym{GNU} tar can read Sun tapes even with their
8684 wrong checksums. @GNUTAR{} produces the standard
8685 checksum, however, raising incompatibilities with Sun. That is to
8686 say, @GNUTAR{} has not been modified to
8687 @emph{produce} incorrect archives to be read by buggy @command{tar}'s.
8688 I've been told that more recent Sun @command{tar} now read standard
8689 archives, so maybe Sun did a similar patch, after all?
8691 The story seems to be that when Sun first imported @command{tar}
8692 sources on their system, they recompiled it without realizing that
8693 the checksums were computed differently, because of a change in
8694 the default signing of @code{char}'s in their compiler. So they
8695 started computing checksums wrongly. When they later realized their
8696 mistake, they merely decided to stay compatible with it, and with
8697 themselves afterwards. Presumably, but I do not really know, HP-UX
8698 has chosen that their @command{tar} archives to be compatible with Sun's.
8699 The current standards do not favor Sun @command{tar} format. In any
8700 case, it now falls on the shoulders of SunOS and HP-UX users to get
8701 a @command{tar} able to read the good archives they receive.
8703 @node Large or Negative Values
8704 @subsection Large or Negative Values
8705 @cindex large values
8706 @cindex future time stamps
8707 @cindex negative time stamps
8710 The above sections suggest to use @samp{oldest possible} archive
8711 format if in doubt. However, sometimes it is not possible. If you
8712 attempt to archive a file whose metadata cannot be represented using
8713 required format, @GNUTAR{} will print error message and ignore such a
8714 file. You will than have to switch to a format that is able to
8715 handle such values. The format summary table (@pxref{Formats}) will
8718 In particular, when trying to archive files larger than 8GB or with
8719 timestamps not in the range 1970-01-01 00:00:00 through 2242-03-16
8720 12:56:31 @sc{utc}, you will have to chose between @acronym{GNU} and
8721 @acronym{POSIX} archive formats. When considering which format to
8722 choose, bear in mind that the @acronym{GNU} format uses
8723 two's-complement base-256 notation to store values that do not fit
8724 into standard @acronym{ustar} range. Such archives can generally be
8725 read only by a @GNUTAR{} implementation. Moreover, they sometimes
8726 cannot be correctly restored on another hosts even by @GNUTAR{}. For
8727 example, using two's complement representation for negative time
8728 stamps that assumes a signed 32-bit @code{time_t} generates archives
8729 that are not portable to hosts with differing @code{time_t}
8732 On the other hand, @acronym{POSIX} archives, generally speaking, can
8733 be extracted by any tar implementation that understands older
8734 @acronym{ustar} format. The only exception are files larger than 8GB.
8736 @FIXME{Describe how @acronym{POSIX} archives are extracted by non
8740 @subsection How to Extract GNU-Specific Data Using Other @command{tar} Implementations
8742 In previous sections you became acquainted with various quirks
8743 necessary to make your archives portable. Sometimes you may need to
8744 extract archives containing GNU-specific members using some
8745 third-party @command{tar} implementation or an older version of
8746 @GNUTAR{}. Of course your best bet is to have @GNUTAR{} installed,
8747 but if it is for some reason impossible, this section will explain
8748 how to cope without it.
8750 When we speak about @dfn{GNU-specific} members we mean two classes of
8751 them: members split between the volumes of a multi-volume archive and
8752 sparse members. You will be able to always recover such members if
8753 the archive is in PAX format. In addition split members can be
8754 recovered from archives in old GNU format. The following subsections
8755 describe the required procedures in detail.
8758 * Split Recovery:: Members Split Between Volumes
8759 * Sparse Recovery:: Sparse Members
8762 @node Split Recovery
8763 @subsubsection Extracting Members Split Between Volumes
8765 @cindex Mutli-volume archives, extracting using non-GNU tars
8766 If a member is split between several volumes of an old GNU format archive
8767 most third party @command{tar} implementation will fail to extract
8768 it. To extract it, use @command{tarcat} program (@pxref{Tarcat}).
8769 This program is available from
8770 @uref{http://www.gnu.org/@/software/@/tar/@/utils/@/tarcat.html, @GNUTAR{}
8771 home page}. It concatenates several archive volumes into a single
8772 valid archive. For example, if you have three volumes named from
8773 @file{vol-1.tar} to @file{vol-3.tar}, you can do the following to
8774 extract them using a third-party @command{tar}:
8777 $ @kbd{tarcat vol-1.tar vol-2.tar vol-3.tar | tar xf -}
8780 @cindex Mutli-volume archives in PAX format, extracting using non-GNU tars
8781 You could use this approach for most (although not all) PAX
8782 format archives as well. However, extracting split members from a PAX
8783 archive is a much easier task, because PAX volumes are constructed in
8784 such a way that each part of a split member is extracted to a
8785 different file by @command{tar} implementations that are not aware of
8786 GNU extensions. More specifically, the very first part retains its
8787 original name, and all subsequent parts are named using the pattern:
8790 %d/GNUFileParts.%p/%f.%n
8794 where symbols preceeded by @samp{%} are @dfn{macro characters} that
8795 have the following meaning:
8797 @multitable @columnfractions .25 .55
8798 @headitem Meta-character @tab Replaced By
8799 @item %d @tab The directory name of the file, equivalent to the
8800 result of the @command{dirname} utility on its full name.
8801 @item %f @tab The file name of the file, equivalent to the result
8802 of the @command{basename} utility on its full name.
8803 @item %p @tab The process @acronym{ID} of the @command{tar} process that
8804 created the archive.
8805 @item %n @tab Ordinal number of this particular part.
8808 For example, if the file @file{var/longfile} was split during archive
8809 creation between three volumes, and the creator @command{tar} process
8810 had process @acronym{ID} @samp{27962}, then the member names will be:
8814 var/GNUFileParts.27962/longfile.1
8815 var/GNUFileParts.27962/longfile.2
8818 When you extract your archive using a third-party @command{tar}, these
8819 files will be created on your disk, and the only thing you will need
8820 to do to restore your file in its original form is concatenate them in
8821 the proper order, for example:
8826 $ @kbd{cat GNUFileParts.27962/longfile.1 \
8827 GNUFileParts.27962/longfile.2 >> longfile}
8828 $ rm -f GNUFileParts.27962
8832 Notice, that if the @command{tar} implementation you use supports PAX
8833 format archives, it will probably emit warnings about unknown keywords
8834 during extraction. They will look like this:
8839 Unknown extended header keyword 'GNU.volume.filename' ignored.
8840 Unknown extended header keyword 'GNU.volume.size' ignored.
8841 Unknown extended header keyword 'GNU.volume.offset' ignored.
8846 You can safely ignore these warnings.
8848 If your @command{tar} implementation is not PAX-aware, you will get
8849 more warnings and more files generated on your disk, e.g.:
8853 $ @kbd{tar xf vol-1.tar}
8854 var/PaxHeaders.27962/longfile: Unknown file type 'x', extracted as
8856 Unexpected EOF in archive
8857 $ @kbd{tar xf vol-2.tar}
8858 tmp/GlobalHead.27962.1: Unknown file type 'g', extracted as normal file
8859 GNUFileParts.27962/PaxHeaders.27962/sparsefile.1: Unknown file type
8860 'x', extracted as normal file
8864 Ignore these warnings. The @file{PaxHeaders.*} directories created
8865 will contain files with @dfn{extended header keywords} describing the
8866 extracted files. You can delete them, unless they describe sparse
8867 members. Read further to learn more about them.
8869 @node Sparse Recovery
8870 @subsubsection Extracting Sparse Members
8872 @cindex sparse files, extracting with non-GNU tars
8873 Any @command{tar} implementation will be able to extract sparse members from a
8874 PAX archive. However, the extracted files will be @dfn{condensed},
8875 i.e., any zero blocks will be removed from them. When we restore such
8876 a condensed file to its original form, by adding zero blocks (or
8877 @dfn{holes}) back to their original locations, we call this process
8878 @dfn{expanding} a compressed sparse file.
8881 To expand a file, you will need a simple auxiliary program called
8882 @command{xsparse}. It is available in source form from
8883 @uref{http://www.gnu.org/@/software/@/tar/@/utils/@/xsparse.html, @GNUTAR{}
8886 @cindex sparse files v.1.0, extracting with non-GNU tars
8887 Let's begin with archive members in @dfn{sparse format
8888 version 1.0}@footnote{@xref{PAX 1}.}, which are the easiest to expand.
8889 The condensed file will contain both file map and file data, so no
8890 additional data will be needed to restore it. If the original file
8891 name was @file{@var{dir}/@var{name}}, then the condensed file will be
8892 named @file{@var{dir}/@/GNUSparseFile.@var{n}/@/@var{name}}, where
8893 @var{n} is a decimal number@footnote{technically speaking, @var{n} is a
8894 @dfn{process @acronym{ID}} of the @command{tar} process which created the
8895 archive (@pxref{PAX keywords}).}.
8897 To expand a version 1.0 file, run @command{xsparse} as follows:
8900 $ @kbd{xsparse @file{cond-file}}
8904 where @file{cond-file} is the name of the condensed file. The utility
8905 will deduce the name for the resulting expanded file using the
8906 following algorithm:
8909 @item If @file{cond-file} does not contain any directories,
8910 @file{../cond-file} will be used;
8912 @item If @file{cond-file} has the form
8913 @file{@var{dir}/@var{t}/@var{name}}, where both @var{t} and @var{name}
8914 are simple names, with no @samp{/} characters in them, the output file
8915 name will be @file{@var{dir}/@var{name}}.
8917 @item Otherwise, if @file{cond-file} has the form
8918 @file{@var{dir}/@var{name}}, the output file name will be
8922 In the unlikely case when this algorithm does not suit your needs,
8923 you can explicitly specify output file name as a second argument to
8927 $ @kbd{xsparse @file{cond-file} @file{out-file}}
8930 It is often a good idea to run @command{xsparse} in @dfn{dry run} mode
8931 first. In this mode, the command does not actually expand the file,
8932 but verbosely lists all actions it would be taking to do so. The dry
8933 run mode is enabled by @option{-n} command line argument:
8937 $ @kbd{xsparse -n /home/gray/GNUSparseFile.6058/sparsefile}
8938 Reading v.1.0 sparse map
8939 Expanding file `/home/gray/GNUSparseFile.6058/sparsefile' to
8940 `/home/gray/sparsefile'
8945 To actually expand the file, you would run:
8948 $ @kbd{xsparse /home/gray/GNUSparseFile.6058/sparsefile}
8952 The program behaves the same way all UNIX utilities do: it will keep
8953 quiet unless it has simething important to tell you (e.g. an error
8954 condition or something). If you wish it to produce verbose output,
8955 similar to that from the dry run mode, use @option{-v} option:
8959 $ @kbd{xsparse -v /home/gray/GNUSparseFile.6058/sparsefile}
8960 Reading v.1.0 sparse map
8961 Expanding file `/home/gray/GNUSparseFile.6058/sparsefile' to
8962 `/home/gray/sparsefile'
8967 Additionally, if your @command{tar} implementation has extracted the
8968 @dfn{extended headers} for this file, you can instruct @command{xstar}
8969 to use them in order to verify the integrity of the expanded file.
8970 The option @option{-x} sets the name of the extended header file to
8971 use. Continuing our example:
8975 $ @kbd{xsparse -v -x /home/gray/PaxHeaders.6058/sparsefile \
8976 /home/gray/GNUSparseFile.6058/sparsefile}
8977 Reading extended header file
8978 Found variable GNU.sparse.major = 1
8979 Found variable GNU.sparse.minor = 0
8980 Found variable GNU.sparse.name = sparsefile
8981 Found variable GNU.sparse.realsize = 217481216
8982 Reading v.1.0 sparse map
8983 Expanding file `/home/gray/GNUSparseFile.6058/sparsefile' to
8984 `/home/gray/sparsefile'
8989 @anchor{extracting sparse v.0.x}
8990 @cindex sparse files v.0.1, extracting with non-GNU tars
8991 @cindex sparse files v.0.0, extracting with non-GNU tars
8992 An @dfn{extended header} is a special @command{tar} archive header
8993 that precedes an archive member and contains a set of
8994 @dfn{variables}, describing the member properties that cannot be
8995 stored in the standard @code{ustar} header. While optional for
8996 expanding sparse version 1.0 members, the use of extended headers is
8997 mandatory when expanding sparse members in older sparse formats: v.0.0
8998 and v.0.1 (The sparse formats are described in detail in @ref{Sparse
8999 Formats}.) So, for these formats, the question is: how to obtain
9000 extended headers from the archive?
9002 If you use a @command{tar} implementation that does not support PAX
9003 format, extended headers for each member will be extracted as a
9004 separate file. If we represent the member name as
9005 @file{@var{dir}/@var{name}}, then the extended header file will be
9006 named @file{@var{dir}/@/PaxHeaders.@var{n}/@/@var{name}}, where
9007 @var{n} is an integer number.
9009 Things become more difficult if your @command{tar} implementation
9010 does support PAX headers, because in this case you will have to
9011 manually extract the headers. We recommend the following algorithm:
9015 Consult the documentation of your @command{tar} implementation for an
9016 option that prints @dfn{block numbers} along with the archive
9017 listing (analogous to @GNUTAR{}'s @option{-R} option). For example,
9018 @command{star} has @option{-block-number}.
9021 Obtain verbose listing using the @samp{block number} option, and
9022 find block numbers of the sparse member in question and the member
9023 immediately following it. For example, running @command{star} on our
9028 $ @kbd{star -t -v -block-number -f arc.tar}
9030 star: Unknown extended header keyword 'GNU.sparse.size' ignored.
9031 star: Unknown extended header keyword 'GNU.sparse.numblocks' ignored.
9032 star: Unknown extended header keyword 'GNU.sparse.name' ignored.
9033 star: Unknown extended header keyword 'GNU.sparse.map' ignored.
9034 block 56: 425984 -rw-r--r-- gray/users Jun 25 14:46 2006 GNUSparseFile.28124/sparsefile
9035 block 897: 65391 -rw-r--r-- gray/users Jun 24 20:06 2006 README
9041 (as usual, ignore the warnings about unknown keywords.)
9044 Let @var{size} be the size of the sparse member, @var{Bs} be its block number
9045 and @var{Bn} be the block number of the next member.
9049 @var{N} = @var{Bs} - @var{Bn} - @var{size}/512 - 2
9053 This number gives the size of the extended header part in tar @dfn{blocks}.
9054 In our example, this formula gives: @code{897 - 56 - 425984 / 512 - 2
9058 Use @command{dd} to extract the headers:
9061 @kbd{dd if=@var{archive} of=@var{hname} bs=512 skip=@var{Bs} count=@var{N}}
9065 where @var{archive} is the archive name, @var{hname} is a name of the
9066 file to store the extended header in, @var{Bs} and @var{N} are
9067 computed in previous steps.
9069 In our example, this command will be
9072 $ @kbd{dd if=arc.tar of=xhdr bs=512 skip=56 count=7}
9076 Finally, you can expand the condensed file, using the obtained header:
9080 $ @kbd{xsparse -v -x xhdr GNUSparseFile.6058/sparsefile}
9081 Reading extended header file
9082 Found variable GNU.sparse.size = 217481216
9083 Found variable GNU.sparse.numblocks = 208
9084 Found variable GNU.sparse.name = sparsefile
9085 Found variable GNU.sparse.map = 0,2048,1050624,2048,@dots{}
9086 Expanding file `GNUSparseFile.28124/sparsefile' to `sparsefile'
9092 @section Comparison of @command{tar} and @command{cpio}
9095 @FIXME{Reorganize the following material}
9097 The @command{cpio} archive formats, like @command{tar}, do have maximum
9098 file name lengths. The binary and old @acronym{ASCII} formats have a maximum file
9099 length of 256, and the new @acronym{ASCII} and @acronym{CRC ASCII} formats have a max
9100 file length of 1024. @acronym{GNU} @command{cpio} can read and write archives
9101 with arbitrary file name lengths, but other @command{cpio} implementations
9102 may crash unexplainedly trying to read them.
9104 @command{tar} handles symbolic links in the form in which it comes in @acronym{BSD};
9105 @command{cpio} doesn't handle symbolic links in the form in which it comes
9106 in System V prior to SVR4, and some vendors may have added symlinks
9107 to their system without enhancing @command{cpio} to know about them.
9108 Others may have enhanced it in a way other than the way I did it
9109 at Sun, and which was adopted by AT&T (and which is, I think, also
9110 present in the @command{cpio} that Berkeley picked up from AT&T and put
9111 into a later @acronym{BSD} release---I think I gave them my changes).
9113 (SVR4 does some funny stuff with @command{tar}; basically, its @command{cpio}
9114 can handle @command{tar} format input, and write it on output, and it
9115 probably handles symbolic links. They may not have bothered doing
9116 anything to enhance @command{tar} as a result.)
9118 @command{cpio} handles special files; traditional @command{tar} doesn't.
9120 @command{tar} comes with V7, System III, System V, and @acronym{BSD} source;
9121 @command{cpio} comes only with System III, System V, and later @acronym{BSD}
9122 (4.3-tahoe and later).
9124 @command{tar}'s way of handling multiple hard links to a file can handle
9125 file systems that support 32-bit inumbers (e.g., the @acronym{BSD} file system);
9126 @command{cpio}s way requires you to play some games (in its ``binary''
9127 format, i-numbers are only 16 bits, and in its ``portable @acronym{ASCII}'' format,
9128 they're 18 bits---it would have to play games with the "file system @acronym{ID}"
9129 field of the header to make sure that the file system @acronym{ID}/i-number pairs
9130 of different files were always different), and I don't know which
9131 @command{cpio}s, if any, play those games. Those that don't might get
9132 confused and think two files are the same file when they're not, and
9133 make hard links between them.
9135 @command{tar}s way of handling multiple hard links to a file places only
9136 one copy of the link on the tape, but the name attached to that copy
9137 is the @emph{only} one you can use to retrieve the file; @command{cpio}s
9138 way puts one copy for every link, but you can retrieve it using any
9142 What type of check sum (if any) is used, and how is this calculated.
9145 See the attached manual pages for @command{tar} and @command{cpio} format.
9146 @command{tar} uses a checksum which is the sum of all the bytes in the
9147 @command{tar} header for a file; @command{cpio} uses no checksum.
9150 If anyone knows why @command{cpio} was made when @command{tar} was present
9154 It wasn't. @command{cpio} first showed up in PWB/UNIX 1.0; no
9155 generally-available version of UNIX had @command{tar} at the time. I don't
9156 know whether any version that was generally available @emph{within AT&T}
9157 had @command{tar}, or, if so, whether the people within AT&T who did
9158 @command{cpio} knew about it.
9160 On restore, if there is a corruption on a tape @command{tar} will stop at
9161 that point, while @command{cpio} will skip over it and try to restore the
9164 The main difference is just in the command syntax and header format.
9166 @command{tar} is a little more tape-oriented in that everything is blocked
9167 to start on a record boundary.
9170 Is there any differences between the ability to recover crashed
9171 archives between the two of them. (Is there any chance of recovering
9172 crashed archives at all.)
9175 Theoretically it should be easier under @command{tar} since the blocking
9176 lets you find a header with some variation of @samp{dd skip=@var{nn}}.
9177 However, modern @command{cpio}'s and variations have an option to just
9178 search for the next file header after an error with a reasonable chance
9179 of resyncing. However, lots of tape driver software won't allow you to
9180 continue past a media error which should be the only reason for getting
9181 out of sync unless a file changed sizes while you were writing the
9185 If anyone knows why @command{cpio} was made when @command{tar} was present
9186 at the unix scene, please tell me about this too.
9189 Probably because it is more media efficient (by not blocking everything
9190 and using only the space needed for the headers where @command{tar}
9191 always uses 512 bytes per file header) and it knows how to archive
9194 You might want to look at the freely available alternatives. The
9195 major ones are @command{afio}, @GNUTAR{}, and
9196 @command{pax}, each of which have their own extensions with some
9197 backwards compatibility.
9199 Sparse files were @command{tar}red as sparse files (which you can
9200 easily test, because the resulting archive gets smaller, and
9201 @acronym{GNU} @command{cpio} can no longer read it).
9204 @chapter Tapes and Other Archive Media
9207 A few special cases about tape handling warrant more detailed
9208 description. These special cases are discussed below.
9210 Many complexities surround the use of @command{tar} on tape drives. Since
9211 the creation and manipulation of archives located on magnetic tape was
9212 the original purpose of @command{tar}, it contains many features making
9213 such manipulation easier.
9215 Archives are usually written on dismountable media---tape cartridges,
9216 mag tapes, or floppy disks.
9218 The amount of data a tape or disk holds depends not only on its size,
9219 but also on how it is formatted. A 2400 foot long reel of mag tape
9220 holds 40 megabytes of data when formatted at 1600 bits per inch. The
9221 physically smaller EXABYTE tape cartridge holds 2.3 gigabytes.
9223 Magnetic media are re-usable---once the archive on a tape is no longer
9224 needed, the archive can be erased and the tape or disk used over.
9225 Media quality does deteriorate with use, however. Most tapes or disks
9226 should be discarded when they begin to produce data errors. EXABYTE
9227 tape cartridges should be discarded when they generate an @dfn{error
9228 count} (number of non-usable bits) of more than 10k.
9230 Magnetic media are written and erased using magnetic fields, and
9231 should be protected from such fields to avoid damage to stored data.
9232 Sticking a floppy disk to a filing cabinet using a magnet is probably
9236 * Device:: Device selection and switching
9237 * Remote Tape Server::
9238 * Common Problems and Solutions::
9239 * Blocking:: Blocking
9240 * Many:: Many archives on one tape
9241 * Using Multiple Tapes:: Using Multiple Tapes
9242 * label:: Including a Label in the Archive
9244 * Write Protection::
9248 @section Device Selection and Switching
9252 @item -f [@var{hostname}:]@var{file}
9253 @itemx --file=[@var{hostname}:]@var{file}
9254 Use archive file or device @var{file} on @var{hostname}.
9257 This option is used to specify the file name of the archive @command{tar}
9260 If the file name is @samp{-}, @command{tar} reads the archive from standard
9261 input (when listing or extracting), or writes it to standard output
9262 (when creating). If the @samp{-} file name is given when updating an
9263 archive, @command{tar} will read the original archive from its standard
9264 input, and will write the entire new archive to its standard output.
9266 If the file name contains a @samp{:}, it is interpreted as
9267 @samp{hostname:file name}. If the @var{hostname} contains an @dfn{at}
9268 sign (@samp{@@}), it is treated as @samp{user@@hostname:file name}. In
9269 either case, @command{tar} will invoke the command @command{rsh} (or
9270 @command{remsh}) to start up an @command{/usr/libexec/rmt} on the remote
9271 machine. If you give an alternate login name, it will be given to the
9273 Naturally, the remote machine must have an executable
9274 @command{/usr/libexec/rmt}. This program is free software from the
9275 University of California, and a copy of the source code can be found
9276 with the sources for @command{tar}; it's compiled and installed by default.
9277 The exact path to this utility is determined when configuring the package.
9278 It is @file{@var{prefix}/libexec/rmt}, where @var{prefix} stands for
9279 your installation prefix. This location may also be overridden at
9280 runtime by using @option{rmt-command=@var{command}} option (@xref{Option Summary,
9281 ---rmt-command}, for detailed description of this option. @xref{Remote
9282 Tape Server}, for the description of @command{rmt} command).
9284 If this option is not given, but the environment variable @env{TAPE}
9285 is set, its value is used; otherwise, old versions of @command{tar}
9286 used a default archive name (which was picked when @command{tar} was
9287 compiled). The default is normally set up to be the @dfn{first} tape
9288 drive or other transportable I/O medium on the system.
9290 Starting with version 1.11.5, @GNUTAR{} uses
9291 standard input and standard output as the default device, and I will
9292 not try anymore supporting automatic device detection at installation
9293 time. This was failing really in too many cases, it was hopeless.
9294 This is now completely left to the installer to override standard
9295 input and standard output for default device, if this seems
9296 preferable. Further, I think @emph{most} actual usages of
9297 @command{tar} are done with pipes or disks, not really tapes,
9298 cartridges or diskettes.
9300 Some users think that using standard input and output is running
9301 after trouble. This could lead to a nasty surprise on your screen if
9302 you forget to specify an output file name---especially if you are going
9303 through a network or terminal server capable of buffering large amounts
9304 of output. We had so many bug reports in that area of configuring
9305 default tapes automatically, and so many contradicting requests, that
9306 we finally consider the problem to be portably intractable. We could
9307 of course use something like @samp{/dev/tape} as a default, but this
9308 is @emph{also} running after various kind of trouble, going from hung
9309 processes to accidental destruction of real tapes. After having seen
9310 all this mess, using standard input and output as a default really
9311 sounds like the only clean choice left, and a very useful one too.
9313 @GNUTAR{} reads and writes archive in records, I
9314 suspect this is the main reason why block devices are preferred over
9315 character devices. Most probably, block devices are more efficient
9316 too. The installer could also check for @samp{DEFTAPE} in
9317 @file{<sys/mtio.h>}.
9320 @xopindex{force-local, short description}
9322 Archive file is local even if it contains a colon.
9324 @opindex rsh-command
9325 @item --rsh-command=@var{command}
9326 Use remote @var{command} instead of @command{rsh}. This option exists
9327 so that people who use something other than the standard @command{rsh}
9328 (e.g., a Kerberized @command{rsh}) can access a remote device.
9330 When this command is not used, the shell command found when
9331 the @command{tar} program was installed is used instead. This is
9332 the first found of @file{/usr/ucb/rsh}, @file{/usr/bin/remsh},
9333 @file{/usr/bin/rsh}, @file{/usr/bsd/rsh} or @file{/usr/bin/nsh}.
9334 The installer may have overridden this by defining the environment
9335 variable @env{RSH} @emph{at installation time}.
9338 Specify drive and density.
9340 @xopindex{multi-volume, short description}
9342 @itemx --multi-volume
9343 Create/list/extract multi-volume archive.
9345 This option causes @command{tar} to write a @dfn{multi-volume} archive---one
9346 that may be larger than will fit on the medium used to hold it.
9347 @xref{Multi-Volume Archives}.
9349 @xopindex{tape-length, short description}
9351 @itemx --tape-length=@var{num}
9352 Change tape after writing @var{num} x 1024 bytes.
9354 This option might be useful when your tape drivers do not properly
9355 detect end of physical tapes. By being slightly conservative on the
9356 maximum tape length, you might avoid the problem entirely.
9358 @xopindex{info-script, short description}
9359 @xopindex{new-volume-script, short description}
9361 @itemx --info-script=@var{file}
9362 @itemx --new-volume-script=@var{file}
9363 Execute @file{file} at end of each tape. This implies
9364 @option{--multi-volume} (@option{-M}). @xref{info-script}, for a detailed
9365 description of this option.
9368 @node Remote Tape Server
9369 @section The Remote Tape Server
9371 @cindex remote tape drive
9373 In order to access the tape drive on a remote machine, @command{tar}
9374 uses the remote tape server written at the University of California at
9375 Berkeley. The remote tape server must be installed as
9376 @file{@var{prefix}/libexec/rmt} on any machine whose tape drive you
9377 want to use. @command{tar} calls @command{rmt} by running an
9378 @command{rsh} or @command{remsh} to the remote machine, optionally
9379 using a different login name if one is supplied.
9381 A copy of the source for the remote tape server is provided. It is
9382 Copyright @copyright{} 1983 by the Regents of the University of
9383 California, but can be freely distributed. It is compiled and
9384 installed by default.
9386 @cindex absolute file names
9387 Unless you use the @option{--absolute-names} (@option{-P}) option,
9388 @GNUTAR{} will not allow you to create an archive that contains
9389 absolute file names (a file name beginning with @samp{/}.) If you try,
9390 @command{tar} will automatically remove the leading @samp{/} from the
9391 file names it stores in the archive. It will also type a warning
9392 message telling you what it is doing.
9394 When reading an archive that was created with a different
9395 @command{tar} program, @GNUTAR{} automatically
9396 extracts entries in the archive which have absolute file names as if
9397 the file names were not absolute. This is an important feature. A
9398 visitor here once gave a @command{tar} tape to an operator to restore;
9399 the operator used Sun @command{tar} instead of @GNUTAR{},
9400 and the result was that it replaced large portions of
9401 our @file{/bin} and friends with versions from the tape; needless to
9402 say, we were unhappy about having to recover the file system from
9405 For example, if the archive contained a file @file{/usr/bin/computoy},
9406 @GNUTAR{} would extract the file to @file{usr/bin/computoy},
9407 relative to the current directory. If you want to extract the files in
9408 an archive to the same absolute names that they had when the archive
9409 was created, you should do a @samp{cd /} before extracting the files
9410 from the archive, or you should either use the @option{--absolute-names}
9411 option, or use the command @samp{tar -C / @dots{}}.
9413 @cindex Ultrix 3.1 and write failure
9414 Some versions of Unix (Ultrix 3.1 is known to have this problem),
9415 can claim that a short write near the end of a tape succeeded,
9416 when it actually failed. This will result in the -M option not
9417 working correctly. The best workaround at the moment is to use a
9418 significantly larger blocking factor than the default 20.
9420 In order to update an archive, @command{tar} must be able to backspace the
9421 archive in order to reread or rewrite a record that was just read (or
9422 written). This is currently possible only on two kinds of files: normal
9423 disk files (or any other file that can be backspaced with @samp{lseek}),
9424 and industry-standard 9-track magnetic tape (or any other kind of tape
9425 that can be backspaced with the @code{MTIOCTOP} @code{ioctl}.
9427 This means that the @option{--append}, @option{--concatenate}, and
9428 @option{--delete} commands will not work on any other kind of file.
9429 Some media simply cannot be backspaced, which means these commands and
9430 options will never be able to work on them. These non-backspacing
9431 media include pipes and cartridge tape drives.
9433 Some other media can be backspaced, and @command{tar} will work on them
9434 once @command{tar} is modified to do so.
9436 Archives created with the @option{--multi-volume}, @option{--label}, and
9437 @option{--incremental} (@option{-G}) options may not be readable by other version
9438 of @command{tar}. In particular, restoring a file that was split over
9439 a volume boundary will require some careful work with @command{dd}, if
9440 it can be done at all. Other versions of @command{tar} may also create
9441 an empty file whose name is that of the volume header. Some versions
9442 of @command{tar} may create normal files instead of directories archived
9443 with the @option{--incremental} (@option{-G}) option.
9445 @node Common Problems and Solutions
9446 @section Some Common Problems and their Solutions
9453 no such file or directory
9456 errors from @command{tar}:
9457 directory checksum error
9460 errors from media/system:
9471 @dfn{Block} and @dfn{record} terminology is rather confused, and it
9472 is also confusing to the expert reader. On the other hand, readers
9473 who are new to the field have a fresh mind, and they may safely skip
9474 the next two paragraphs, as the remainder of this manual uses those
9475 two terms in a quite consistent way.
9477 John Gilmore, the writer of the public domain @command{tar} from which
9478 @GNUTAR{} was originally derived, wrote (June 1995):
9481 The nomenclature of tape drives comes from IBM, where I believe
9482 they were invented for the IBM 650 or so. On IBM mainframes, what
9483 is recorded on tape are tape blocks. The logical organization of
9484 data is into records. There are various ways of putting records into
9485 blocks, including @code{F} (fixed sized records), @code{V} (variable
9486 sized records), @code{FB} (fixed blocked: fixed size records, @var{n}
9487 to a block), @code{VB} (variable size records, @var{n} to a block),
9488 @code{VSB} (variable spanned blocked: variable sized records that can
9489 occupy more than one block), etc. The @code{JCL} @samp{DD RECFORM=}
9490 parameter specified this to the operating system.
9492 The Unix man page on @command{tar} was totally confused about this.
9493 When I wrote @code{PD TAR}, I used the historically correct terminology
9494 (@command{tar} writes data records, which are grouped into blocks).
9495 It appears that the bogus terminology made it into @acronym{POSIX} (no surprise
9496 here), and now Fran@,{c}ois has migrated that terminology back
9497 into the source code too.
9500 The term @dfn{physical block} means the basic transfer chunk from or
9501 to a device, after which reading or writing may stop without anything
9502 being lost. In this manual, the term @dfn{block} usually refers to
9503 a disk physical block, @emph{assuming} that each disk block is 512
9504 bytes in length. It is true that some disk devices have different
9505 physical blocks, but @command{tar} ignore these differences in its own
9506 format, which is meant to be portable, so a @command{tar} block is always
9507 512 bytes in length, and @dfn{block} always mean a @command{tar} block.
9508 The term @dfn{logical block} often represents the basic chunk of
9509 allocation of many disk blocks as a single entity, which the operating
9510 system treats somewhat atomically; this concept is only barely used
9513 The term @dfn{physical record} is another way to speak of a physical
9514 block, those two terms are somewhat interchangeable. In this manual,
9515 the term @dfn{record} usually refers to a tape physical block,
9516 @emph{assuming} that the @command{tar} archive is kept on magnetic tape.
9517 It is true that archives may be put on disk or used with pipes,
9518 but nevertheless, @command{tar} tries to read and write the archive one
9519 @dfn{record} at a time, whatever the medium in use. One record is made
9520 up of an integral number of blocks, and this operation of putting many
9521 disk blocks into a single tape block is called @dfn{reblocking}, or
9522 more simply, @dfn{blocking}. The term @dfn{logical record} refers to
9523 the logical organization of many characters into something meaningful
9524 to the application. The term @dfn{unit record} describes a small set
9525 of characters which are transmitted whole to or by the application,
9526 and often refers to a line of text. Those two last terms are unrelated
9527 to what we call a @dfn{record} in @GNUTAR{}.
9529 When writing to tapes, @command{tar} writes the contents of the archive
9530 in chunks known as @dfn{records}. To change the default blocking
9531 factor, use the @option{--blocking-factor=@var{512-size}} (@option{-b
9532 @var{512-size}}) option. Each record will then be composed of
9533 @var{512-size} blocks. (Each @command{tar} block is 512 bytes.
9534 @xref{Standard}.) Each file written to the archive uses at least one
9535 full record. As a result, using a larger record size can result in
9536 more wasted space for small files. On the other hand, a larger record
9537 size can often be read and written much more efficiently.
9539 Further complicating the problem is that some tape drives ignore the
9540 blocking entirely. For these, a larger record size can still improve
9541 performance (because the software layers above the tape drive still
9542 honor the blocking), but not as dramatically as on tape drives that
9545 When reading an archive, @command{tar} can usually figure out the
9546 record size on itself. When this is the case, and a non-standard
9547 record size was used when the archive was created, @command{tar} will
9548 print a message about a non-standard blocking factor, and then operate
9549 normally. On some tape devices, however, @command{tar} cannot figure
9550 out the record size itself. On most of those, you can specify a
9551 blocking factor (with @option{--blocking-factor}) larger than the
9552 actual blocking factor, and then use the @option{--read-full-records}
9553 (@option{-B}) option. (If you specify a blocking factor with
9554 @option{--blocking-factor} and don't use the
9555 @option{--read-full-records} option, then @command{tar} will not
9556 attempt to figure out the recording size itself.) On some devices,
9557 you must always specify the record size exactly with
9558 @option{--blocking-factor} when reading, because @command{tar} cannot
9559 figure it out. In any case, use @option{--list} (@option{-t}) before
9560 doing any extractions to see whether @command{tar} is reading the archive
9563 @command{tar} blocks are all fixed size (512 bytes), and its scheme for
9564 putting them into records is to put a whole number of them (one or
9565 more) into each record. @command{tar} records are all the same size;
9566 at the end of the file there's a block containing all zeros, which
9567 is how you tell that the remainder of the last record(s) are garbage.
9569 In a standard @command{tar} file (no options), the block size is 512
9570 and the record size is 10240, for a blocking factor of 20. What the
9571 @option{--blocking-factor} option does is sets the blocking factor,
9572 changing the record size while leaving the block size at 512 bytes.
9573 20 was fine for ancient 800 or 1600 bpi reel-to-reel tape drives;
9574 most tape drives these days prefer much bigger records in order to
9575 stream and not waste tape. When writing tapes for myself, some tend
9576 to use a factor of the order of 2048, say, giving a record size of
9577 around one megabyte.
9579 If you use a blocking factor larger than 20, older @command{tar}
9580 programs might not be able to read the archive, so we recommend this
9581 as a limit to use in practice. @GNUTAR{}, however,
9582 will support arbitrarily large record sizes, limited only by the
9583 amount of virtual memory or the physical characteristics of the tape
9587 * Format Variations:: Format Variations
9588 * Blocking Factor:: The Blocking Factor of an Archive
9591 @node Format Variations
9592 @subsection Format Variations
9593 @cindex Format Parameters
9594 @cindex Format Options
9595 @cindex Options, archive format specifying
9596 @cindex Options, format specifying
9599 Format parameters specify how an archive is written on the archive
9600 media. The best choice of format parameters will vary depending on
9601 the type and number of files being archived, and on the media used to
9604 To specify format parameters when accessing or creating an archive,
9605 you can use the options described in the following sections.
9606 If you do not specify any format parameters, @command{tar} uses
9607 default parameters. You cannot modify a compressed archive.
9608 If you create an archive with the @option{--blocking-factor} option
9609 specified (@pxref{Blocking Factor}), you must specify that
9610 blocking-factor when operating on the archive. @xref{Formats}, for other
9611 examples of format parameter considerations.
9613 @node Blocking Factor
9614 @subsection The Blocking Factor of an Archive
9615 @cindex Blocking Factor
9617 @cindex Number of blocks per record
9618 @cindex Number of bytes per record
9619 @cindex Bytes per record
9620 @cindex Blocks per record
9623 @opindex blocking-factor
9624 The data in an archive is grouped into blocks, which are 512 bytes.
9625 Blocks are read and written in whole number multiples called
9626 @dfn{records}. The number of blocks in a record (i.e., the size of a
9627 record in units of 512 bytes) is called the @dfn{blocking factor}.
9628 The @option{--blocking-factor=@var{512-size}} (@option{-b
9629 @var{512-size}}) option specifies the blocking factor of an archive.
9630 The default blocking factor is typically 20 (i.e., 10240 bytes), but
9631 can be specified at installation. To find out the blocking factor of
9632 an existing archive, use @samp{tar --list --file=@var{archive-name}}.
9633 This may not work on some devices.
9635 Records are separated by gaps, which waste space on the archive media.
9636 If you are archiving on magnetic tape, using a larger blocking factor
9637 (and therefore larger records) provides faster throughput and allows you
9638 to fit more data on a tape (because there are fewer gaps). If you are
9639 archiving on cartridge, a very large blocking factor (say 126 or more)
9640 greatly increases performance. A smaller blocking factor, on the other
9641 hand, may be useful when archiving small files, to avoid archiving lots
9642 of nulls as @command{tar} fills out the archive to the end of the record.
9643 In general, the ideal record size depends on the size of the
9644 inter-record gaps on the tape you are using, and the average size of the
9645 files you are archiving. @xref{create}, for information on
9648 @FIXME{Need example of using a cartridge with blocking factor=126 or more.}
9650 Archives with blocking factors larger than 20 cannot be read
9651 by very old versions of @command{tar}, or by some newer versions
9652 of @command{tar} running on old machines with small address spaces.
9653 With @GNUTAR{}, the blocking factor of an archive is limited
9654 only by the maximum record size of the device containing the archive,
9655 or by the amount of available virtual memory.
9657 Also, on some systems, not using adequate blocking factors, as sometimes
9658 imposed by the device drivers, may yield unexpected diagnostics. For
9659 example, this has been reported:
9662 Cannot write to /dev/dlt: Invalid argument
9666 In such cases, it sometimes happen that the @command{tar} bundled by
9667 the system is aware of block size idiosyncrasies, while @GNUTAR{}
9668 requires an explicit specification for the block size,
9669 which it cannot guess. This yields some people to consider
9670 @GNUTAR{} is misbehaving, because by comparison,
9671 @cite{the bundle @command{tar} works OK}. Adding @w{@kbd{-b 256}},
9672 for example, might resolve the problem.
9674 If you use a non-default blocking factor when you create an archive, you
9675 must specify the same blocking factor when you modify that archive. Some
9676 archive devices will also require you to specify the blocking factor when
9677 reading that archive, however this is not typically the case. Usually, you
9678 can use @option{--list} (@option{-t}) without specifying a blocking factor---@command{tar}
9679 reports a non-default record size and then lists the archive members as
9680 it would normally. To extract files from an archive with a non-standard
9681 blocking factor (particularly if you're not sure what the blocking factor
9682 is), you can usually use the @option{--read-full-records} (@option{-B}) option while
9683 specifying a blocking factor larger then the blocking factor of the archive
9684 (i.e., @samp{tar --extract --read-full-records --blocking-factor=300}.
9685 @xref{list}, for more information on the @option{--list} (@option{-t})
9686 operation. @xref{Reading}, for a more detailed explanation of that option.
9689 @item --blocking-factor=@var{number}
9690 @itemx -b @var{number}
9691 Specifies the blocking factor of an archive. Can be used with any
9692 operation, but is usually not necessary with @option{--list} (@option{-t}).
9698 @item -b @var{blocks}
9699 @itemx --blocking-factor=@var{blocks}
9700 Set record size to @math{@var{blocks} * 512} bytes.
9702 This option is used to specify a @dfn{blocking factor} for the archive.
9703 When reading or writing the archive, @command{tar}, will do reads and writes
9704 of the archive in records of @math{@var{block}*512} bytes. This is true
9705 even when the archive is compressed. Some devices requires that all
9706 write operations be a multiple of a certain size, and so, @command{tar}
9707 pads the archive out to the next record boundary.
9709 The default blocking factor is set when @command{tar} is compiled, and is
9710 typically 20. Blocking factors larger than 20 cannot be read by very
9711 old versions of @command{tar}, or by some newer versions of @command{tar}
9712 running on old machines with small address spaces.
9714 With a magnetic tape, larger records give faster throughput and fit
9715 more data on a tape (because there are fewer inter-record gaps).
9716 If the archive is in a disk file or a pipe, you may want to specify
9717 a smaller blocking factor, since a large one will result in a large
9718 number of null bytes at the end of the archive.
9720 When writing cartridge or other streaming tapes, a much larger
9721 blocking factor (say 126 or more) will greatly increase performance.
9722 However, you must specify the same blocking factor when reading or
9723 updating the archive.
9725 Apparently, Exabyte drives have a physical block size of 8K bytes.
9726 If we choose our blocksize as a multiple of 8k bytes, then the problem
9727 seems to disappear. Id est, we are using block size of 112 right
9728 now, and we haven't had the problem since we switched@dots{}
9730 With @GNUTAR{} the blocking factor is limited only
9731 by the maximum record size of the device containing the archive, or by
9732 the amount of available virtual memory.
9734 However, deblocking or reblocking is virtually avoided in a special
9735 case which often occurs in practice, but which requires all the
9736 following conditions to be simultaneously true:
9739 the archive is subject to a compression option,
9741 the archive is not handled through standard input or output, nor
9742 redirected nor piped,
9744 the archive is directly handled to a local disk, instead of any special
9747 @option{--blocking-factor} is not explicitly specified on the @command{tar}
9751 If the output goes directly to a local disk, and not through
9752 stdout, then the last write is not extended to a full record size.
9753 Otherwise, reblocking occurs. Here are a few other remarks on this
9759 @command{gzip} will complain about trailing garbage if asked to
9760 uncompress a compressed archive on tape, there is an option to turn
9761 the message off, but it breaks the regularity of simply having to use
9762 @samp{@var{prog} -d} for decompression. It would be nice if gzip was
9763 silently ignoring any number of trailing zeros. I'll ask Jean-loup
9764 Gailly, by sending a copy of this message to him.
9767 @command{compress} does not show this problem, but as Jean-loup pointed
9768 out to Michael, @samp{compress -d} silently adds garbage after
9769 the result of decompression, which tar ignores because it already
9770 recognized its end-of-file indicator. So this bug may be safely
9774 @samp{gzip -d -q} will be silent about the trailing zeros indeed,
9775 but will still return an exit status of 2 which tar reports in turn.
9776 @command{tar} might ignore the exit status returned, but I hate doing
9777 that, as it weakens the protection @command{tar} offers users against
9778 other possible problems at decompression time. If @command{gzip} was
9779 silently skipping trailing zeros @emph{and} also avoiding setting the
9780 exit status in this innocuous case, that would solve this situation.
9783 @command{tar} should become more solid at not stopping to read a pipe at
9784 the first null block encountered. This inelegantly breaks the pipe.
9785 @command{tar} should rather drain the pipe out before exiting itself.
9788 @xopindex{ignore-zeros, short description}
9790 @itemx --ignore-zeros
9791 Ignore blocks of zeros in archive (means EOF).
9793 The @option{--ignore-zeros} (@option{-i}) option causes @command{tar} to ignore blocks
9794 of zeros in the archive. Normally a block of zeros indicates the
9795 end of the archive, but when reading a damaged archive, or one which
9796 was created by concatenating several archives together, this option
9797 allows @command{tar} to read the entire archive. This option is not on
9798 by default because many versions of @command{tar} write garbage after
9801 Note that this option causes @command{tar} to read to the end of the
9802 archive file, which may sometimes avoid problems when multiple files
9803 are stored on a single physical tape.
9805 @xopindex{read-full-records, short description}
9807 @itemx --read-full-records
9808 Reblock as we read (for reading 4.2@acronym{BSD} pipes).
9810 If @option{--read-full-records} is used, @command{tar}
9811 will not panic if an attempt to read a record from the archive does
9812 not return a full record. Instead, @command{tar} will keep reading
9813 until it has obtained a full
9816 This option is turned on by default when @command{tar} is reading
9817 an archive from standard input, or from a remote machine. This is
9818 because on @acronym{BSD} Unix systems, a read of a pipe will return however
9819 much happens to be in the pipe, even if it is less than @command{tar}
9820 requested. If this option was not used, @command{tar} would fail as
9821 soon as it read an incomplete record from the pipe.
9823 This option is also useful with the commands for updating an archive.
9829 @FIXME{Appropriate options should be moved here from elsewhere.}
9831 @cindex blocking factor
9832 @cindex tape blocking
9834 When handling various tapes or cartridges, you have to take care of
9835 selecting a proper blocking, that is, the number of disk blocks you
9836 put together as a single tape block on the tape, without intervening
9837 tape gaps. A @dfn{tape gap} is a small landing area on the tape
9838 with no information on it, used for decelerating the tape to a
9839 full stop, and for later regaining the reading or writing speed.
9840 When the tape driver starts reading a record, the record has to
9841 be read whole without stopping, as a tape gap is needed to stop the
9842 tape motion without loosing information.
9844 @cindex Exabyte blocking
9845 @cindex DAT blocking
9846 Using higher blocking (putting more disk blocks per tape block) will use
9847 the tape more efficiently as there will be less tape gaps. But reading
9848 such tapes may be more difficult for the system, as more memory will be
9849 required to receive at once the whole record. Further, if there is a
9850 reading error on a huge record, this is less likely that the system will
9851 succeed in recovering the information. So, blocking should not be too
9852 low, nor it should be too high. @command{tar} uses by default a blocking of
9853 20 for historical reasons, and it does not really matter when reading or
9854 writing to disk. Current tape technology would easily accommodate higher
9855 blockings. Sun recommends a blocking of 126 for Exabytes and 96 for DATs.
9856 We were told that for some DLT drives, the blocking should be a multiple
9857 of 4Kb, preferably 64Kb (@w{@kbd{-b 128}}) or 256 for decent performance.
9858 Other manufacturers may use different recommendations for the same tapes.
9859 This might also depends of the buffering techniques used inside modern
9860 tape controllers. Some imposes a minimum blocking, or a maximum blocking.
9861 Others request blocking to be some exponent of two.
9863 So, there is no fixed rule for blocking. But blocking at read time
9864 should ideally be the same as blocking used at write time. At one place
9865 I know, with a wide variety of equipment, they found it best to use a
9866 blocking of 32 to guarantee that their tapes are fully interchangeable.
9868 I was also told that, for recycled tapes, prior erasure (by the same
9869 drive unit that will be used to create the archives) sometimes lowers
9870 the error rates observed at rewriting time.
9872 I might also use @option{--number-blocks} instead of
9873 @option{--block-number}, so @option{--block} will then expand to
9874 @option{--blocking-factor} unambiguously.
9877 @section Many Archives on One Tape
9879 @FIXME{Appropriate options should be moved here from elsewhere.}
9881 @findex ntape @r{device}
9882 Most tape devices have two entries in the @file{/dev} directory, or
9883 entries that come in pairs, which differ only in the minor number for
9884 this device. Let's take for example @file{/dev/tape}, which often
9885 points to the only or usual tape device of a given system. There might
9886 be a corresponding @file{/dev/nrtape} or @file{/dev/ntape}. The simpler
9887 name is the @emph{rewinding} version of the device, while the name
9888 having @samp{nr} in it is the @emph{no rewinding} version of the same
9891 A rewinding tape device will bring back the tape to its beginning point
9892 automatically when this device is opened or closed. Since @command{tar}
9893 opens the archive file before using it and closes it afterwards, this
9894 means that a simple:
9897 $ @kbd{tar cf /dev/tape @var{directory}}
9901 will reposition the tape to its beginning both prior and after saving
9902 @var{directory} contents to it, thus erasing prior tape contents and
9903 making it so that any subsequent write operation will destroy what has
9906 @cindex tape positioning
9907 So, a rewinding device is normally meant to hold one and only one file.
9908 If you want to put more than one @command{tar} archive on a given tape, you
9909 will need to avoid using the rewinding version of the tape device. You
9910 will also have to pay special attention to tape positioning. Errors in
9911 positioning may overwrite the valuable data already on your tape. Many
9912 people, burnt by past experiences, will only use rewinding devices and
9913 limit themselves to one file per tape, precisely to avoid the risk of
9914 such errors. Be fully aware that writing at the wrong position on a
9915 tape loses all information past this point and most probably until the
9916 end of the tape, and this destroyed information @emph{cannot} be
9919 To save @var{directory-1} as a first archive at the beginning of a
9920 tape, and leave that tape ready for a second archive, you should use:
9923 $ @kbd{mt -f /dev/nrtape rewind}
9924 $ @kbd{tar cf /dev/nrtape @var{directory-1}}
9928 @dfn{Tape marks} are special magnetic patterns written on the tape
9929 media, which are later recognizable by the reading hardware. These
9930 marks are used after each file, when there are many on a single tape.
9931 An empty file (that is to say, two tape marks in a row) signal the
9932 logical end of the tape, after which no file exist. Usually,
9933 non-rewinding tape device drivers will react to the close request issued
9934 by @command{tar} by first writing two tape marks after your archive, and by
9935 backspacing over one of these. So, if you remove the tape at that time
9936 from the tape drive, it is properly terminated. But if you write
9937 another file at the current position, the second tape mark will be
9938 erased by the new information, leaving only one tape mark between files.
9940 So, you may now save @var{directory-2} as a second archive after the
9941 first on the same tape by issuing the command:
9944 $ @kbd{tar cf /dev/nrtape @var{directory-2}}
9948 and so on for all the archives you want to put on the same tape.
9950 Another usual case is that you do not write all the archives the same
9951 day, and you need to remove and store the tape between two archive
9952 sessions. In general, you must remember how many files are already
9953 saved on your tape. Suppose your tape already has 16 files on it, and
9954 that you are ready to write the 17th. You have to take care of skipping
9955 the first 16 tape marks before saving @var{directory-17}, say, by using
9959 $ @kbd{mt -f /dev/nrtape rewind}
9960 $ @kbd{mt -f /dev/nrtape fsf 16}
9961 $ @kbd{tar cf /dev/nrtape @var{directory-17}}
9964 In all the previous examples, we put aside blocking considerations, but
9965 you should do the proper things for that as well. @xref{Blocking}.
9968 * Tape Positioning:: Tape Positions and Tape Marks
9969 * mt:: The @command{mt} Utility
9972 @node Tape Positioning
9973 @subsection Tape Positions and Tape Marks
9976 Just as archives can store more than one file from the file system,
9977 tapes can store more than one archive file. To keep track of where
9978 archive files (or any other type of file stored on tape) begin and
9979 end, tape archive devices write magnetic @dfn{tape marks} on the
9980 archive media. Tape drives write one tape mark between files,
9981 two at the end of all the file entries.
9983 If you think of data as a series of records "rrrr"'s, and tape marks as
9984 "*"'s, a tape might look like the following:
9987 rrrr*rrrrrr*rrrrr*rr*rrrrr**-------------------------
9990 Tape devices read and write tapes using a read/write @dfn{tape
9991 head}---a physical part of the device which can only access one
9992 point on the tape at a time. When you use @command{tar} to read or
9993 write archive data from a tape device, the device will begin reading
9994 or writing from wherever on the tape the tape head happens to be,
9995 regardless of which archive or what part of the archive the tape
9996 head is on. Before writing an archive, you should make sure that no
9997 data on the tape will be overwritten (unless it is no longer needed).
9998 Before reading an archive, you should make sure the tape head is at
9999 the beginning of the archive you want to read. You can do it manually
10000 via @code{mt} utility (@pxref{mt}). The @code{restore} script does
10001 that automatically (@pxref{Scripted Restoration}).
10003 If you want to add new archive file entries to a tape, you should
10004 advance the tape to the end of the existing file entries, backspace
10005 over the last tape mark, and write the new archive file. If you were
10006 to add two archives to the example above, the tape might look like the
10010 rrrr*rrrrrr*rrrrr*rr*rrrrr*rrr*rrrr**----------------
10014 @subsection The @command{mt} Utility
10017 @FIXME{Is it true that this only works on non-block devices?
10018 should explain the difference, (fixed or variable).}
10019 @xref{Blocking Factor}.
10021 You can use the @command{mt} utility to advance or rewind a tape past a
10022 specified number of archive files on the tape. This will allow you
10023 to move to the beginning of an archive before extracting or reading
10024 it, or to the end of all the archives before writing a new one.
10025 @FIXME{Why isn't there an "advance 'til you find two tape marks
10028 The syntax of the @command{mt} command is:
10031 @kbd{mt [-f @var{tapename}] @var{operation} [@var{number}]}
10034 where @var{tapename} is the name of the tape device, @var{number} is
10035 the number of times an operation is performed (with a default of one),
10036 and @var{operation} is one of the following:
10038 @FIXME{is there any use for record operations?}
10043 Writes @var{number} tape marks at the current position on the tape.
10046 Moves tape position forward @var{number} files.
10049 Moves tape position back @var{number} files.
10052 Rewinds the tape. (Ignores @var{number}).
10056 Rewinds the tape and takes the tape device off-line. (Ignores @var{number}).
10059 Prints status information about the tape unit.
10063 @FIXME{Is there a better way to frob the spacing on the list?}
10065 If you don't specify a @var{tapename}, @command{mt} uses the environment
10066 variable @env{TAPE}; if @env{TAPE} is not set, @command{mt} will use
10067 the default device specified in your @file{sys/mtio.h} file
10068 (@code{DEFTAPE} variable). If this is not defined, the program will
10069 display a descriptive error message and exit with code 1.
10071 @command{mt} returns a 0 exit status when the operation(s) were
10072 successful, 1 if the command was unrecognized, and 2 if an operation
10075 @node Using Multiple Tapes
10076 @section Using Multiple Tapes
10078 Often you might want to write a large archive, one larger than will fit
10079 on the actual tape you are using. In such a case, you can run multiple
10080 @command{tar} commands, but this can be inconvenient, particularly if you
10081 are using options like @option{--exclude=@var{pattern}} or dumping entire file systems.
10082 Therefore, @command{tar} provides a special mode for creating
10083 multi-volume archives.
10085 @dfn{Multi-volume} archive is a single @command{tar} archive, stored
10086 on several media volumes of fixed size. Although in this section we will
10087 often call @samp{volume} a @dfn{tape}, there is absolutely no
10088 requirement for multi-volume archives to be stored on tapes. Instead,
10089 they can use whatever media type the user finds convenient, they can
10090 even be located on files.
10092 When creating a multi-volume archive, @GNUTAR{} continues to fill
10093 current volume until it runs out of space, then it switches to
10094 next volume (usually the operator is queried to replace the tape on
10095 this point), and continues working on the new volume. This operation
10096 continues until all requested files are dumped. If @GNUTAR{} detects
10097 end of media while dumping a file, such a file is archived in split
10098 form. Some very big files can even be split across several volumes.
10100 Each volume is itself a valid @GNUTAR{} archive, so it can be read
10101 without any special options. Consequently any file member residing
10102 entirely on one volume can be extracted or otherwise operated upon
10103 without needing the other volume. Sure enough, to extract a split
10104 member you would need all volumes its parts reside on.
10106 Multi-volume archives suffer from several limitations. In particular,
10107 they cannot be compressed.
10109 @GNUTAR{} is able to create multi-volume archives of two formats
10110 (@pxref{Formats}): @samp{GNU} and @samp{POSIX}.
10113 * Multi-Volume Archives:: Archives Longer than One Tape or Disk
10114 * Tape Files:: Tape Files
10115 * Tarcat:: Concatenate Volumes into a Single Archive
10119 @node Multi-Volume Archives
10120 @subsection Archives Longer than One Tape or Disk
10121 @cindex Multi-volume archives
10123 @opindex multi-volume
10124 To create an archive that is larger than will fit on a single unit of
10125 the media, use the @option{--multi-volume} (@option{-M}) option in conjunction with
10126 the @option{--create} option (@pxref{create}). A @dfn{multi-volume}
10127 archive can be manipulated like any other archive (provided the
10128 @option{--multi-volume} option is specified), but is stored on more
10129 than one tape or disk.
10131 When you specify @option{--multi-volume}, @command{tar} does not report an
10132 error when it comes to the end of an archive volume (when reading), or
10133 the end of the media (when writing). Instead, it prompts you to load
10134 a new storage volume. If the archive is on a magnetic tape, you
10135 should change tapes when you see the prompt; if the archive is on a
10136 floppy disk, you should change disks; etc.
10139 @item --multi-volume
10141 Creates a multi-volume archive, when used in conjunction with
10142 @option{--create} (@option{-c}). To perform any other operation on a multi-volume
10143 archive, specify @option{--multi-volume} in conjunction with that
10148 $ @kbd{tar --create --multi-volume --file=/dev/tape @var{files}}
10152 The method @command{tar} uses to detect end of tape is not perfect, and
10153 fails on some operating systems or on some devices. If @command{tar}
10154 cannot detect the end of the tape itself, you can use
10155 @option{--tape-length} option to inform it about the capacity of the
10158 @anchor{tape-length}
10160 @opindex tape-length
10161 @item --tape-length=@var{size}
10162 @itemx -L @var{size}
10163 Set maximum length of a volume. The @var{size} argument should then
10164 be the usable size of the tape in units of 1024 bytes. This option
10165 selects @option{--multi-volume} automatically. For example:
10168 $ @kbd{tar --create --tape-length=41943040 --file=/dev/tape @var{files}}
10172 @anchor{change volume prompt}
10173 When @GNUTAR{} comes to the end of a storage media, it asks you to
10174 change the volume. The built-in prompt for POSIX locale
10175 is@footnote{If you run @GNUTAR{} under a different locale, the
10176 translation to the locale's language will be used.}:
10179 Prepare volume #@var{n} for `@var{archive}' and hit return:
10183 where @var{n} is the ordinal number of the volume to be created and
10184 @var{archive} is archive file or device name.
10186 When prompting for a new tape, @command{tar} accepts any of the following
10191 Request @command{tar} to explain possible responses
10193 Request @command{tar} to exit immediately.
10194 @item n @var{file-name}
10195 Request @command{tar} to write the next volume on the file @var{file-name}.
10197 Request @command{tar} to run a subshell. This option can be disabled
10198 by giving @option{--restrict} command line option to
10199 @command{tar}@footnote{@xref{--restrict}, for more information about
10202 Request @command{tar} to begin writing the next volume.
10205 (You should only type @samp{y} after you have changed the tape;
10206 otherwise @command{tar} will write over the volume it just finished.)
10208 @cindex Volume number file
10210 @anchor{volno-file}
10211 @opindex volno-file
10212 The volume number used by @command{tar} in its tape-changing prompt
10213 can be changed; if you give the
10214 @option{--volno-file=@var{file-of-number}} option, then
10215 @var{file-of-number} should be an non-existing file to be created, or
10216 else, a file already containing a decimal number. That number will be
10217 used as the volume number of the first volume written. When
10218 @command{tar} is finished, it will rewrite the file with the
10219 now-current volume number. (This does not change the volume number
10220 written on a tape label, as per @ref{label}, it @emph{only} affects
10221 the number used in the prompt.)
10223 @cindex End-of-archive info script
10224 @cindex Info script
10225 @anchor{info-script}
10226 @opindex info-script
10227 @opindex new-volume-script
10228 If you want more elaborate behavior than this, you can write a special
10229 @dfn{new volume script}, that will be responsible for changing the
10230 volume, and instruct @command{tar} to use it instead of its normal
10231 prompting procedure:
10234 @item --info-script=@var{script-name}
10235 @itemx --new-volume-script=@var{script-name}
10236 @itemx -F @var{script-name}
10237 Specify the full name of the volume script to use. The script can be
10238 used to eject cassettes, or to broadcast messages such as
10239 @samp{Someone please come change my tape} when performing unattended
10243 The @var{script-name} is executed without any command line
10244 arguments. It inherits @command{tar}'s shell environment.
10245 Additional data is passed to it via the following
10246 environment variables:
10249 @vrindex TAR_VERSION, info script environment variable
10251 @GNUTAR{} version number.
10253 @vrindex TAR_ARCHIVE, info script environment variable
10255 The name of the archive @command{tar} is processing.
10257 @vrindex TAR_VOLUME, info script environment variable
10259 Ordinal number of the volume @command{tar} is about to start.
10261 @vrindex TAR_SUBCOMMAND, info script environment variable
10262 @item TAR_SUBCOMMAND
10263 Short option describing the operation @command{tar} is executing
10264 @xref{Operations}, for a complete list of subcommand options.
10266 @vrindex TAR_FORMAT, info script environment variable
10268 Format of the archive being processed. @xref{Formats}, for a complete
10269 list of archive format names.
10271 @vrindex TAR_FD, info script environment variable
10273 File descriptor which can be used to communicate the new volume
10274 name to @command{tar}.
10277 The volume script can instruct @command{tar} to use new archive name,
10278 by writing in to file descriptor @env{$TAR_FD} (see below for an example).
10280 If the info script fails, @command{tar} exits; otherwise, it begins
10281 writing the next volume.
10283 If you want @command{tar} to cycle through a series of files or tape
10284 drives, there are three approaches to choose from. First of all, you
10285 can give @command{tar} multiple @option{--file} options. In this case
10286 the specified files will be used, in sequence, as the successive
10287 volumes of the archive. Only when the first one in the sequence needs
10288 to be used again will @command{tar} prompt for a tape change (or run
10289 the info script). For example, suppose someone has two tape drives on
10290 a system named @file{/dev/tape0} and @file{/dev/tape1}. For having
10291 @GNUTAR{} to switch to the second drive when it needs to write the
10292 second tape, and then back to the first tape, etc., just do either of:
10295 $ @kbd{tar --create --multi-volume --file=/dev/tape0 --file=/dev/tape1 @var{files}}
10296 $ @kbd{tar cMff /dev/tape0 /dev/tape1 @var{files}}
10299 The second method is to use the @samp{n} response to the tape-change
10302 Finally, the most flexible approach is to use a volume script, that
10303 writes new archive name to the file descriptor @env{$TAR_FD}. For example, the
10304 following volume script will create a series of archive files, named
10305 @file{@var{archive}-@var{vol}}, where @var{archive} is the name of the
10306 archive being created (as given by @option{--file} option) and
10307 @var{vol} is the ordinal number of the archive being created:
10312 echo Preparing volume $TAR_VOLUME of $TAR_ARCHIVE.
10314 name=`expr $TAR_ARCHIVE : '\(.*\)-.*'`
10315 case $TAR_SUBCOMMAND in
10317 -d|-x|-t) test -r $@{name:-$TAR_ARCHIVE@}-$TAR_VOLUME || exit 1
10322 echo $@{name:-$TAR_ARCHIVE@}-$TAR_VOLUME >&$TAR_FD
10326 The same script can be used while listing, comparing or extracting
10327 from the created archive. For example:
10331 # @r{Create a multi-volume archive:}
10332 $ @kbd{tar -c -L1024 -f archive.tar -F new-volume .}
10333 # @r{Extract from the created archive:}
10334 $ @kbd{tar -x -f archive.tar -F new-volume .}
10339 Notice, that the first command had to use @option{-L} option, since
10340 otherwise @GNUTAR{} will end up writing everything to file
10341 @file{archive.tar}.
10343 You can read each individual volume of a multi-volume archive as if it
10344 were an archive by itself. For example, to list the contents of one
10345 volume, use @option{--list}, without @option{--multi-volume} specified.
10346 To extract an archive member from one volume (assuming it is described
10347 that volume), use @option{--extract}, again without
10348 @option{--multi-volume}.
10350 If an archive member is split across volumes (i.e., its entry begins on
10351 one volume of the media and ends on another), you need to specify
10352 @option{--multi-volume} to extract it successfully. In this case, you
10353 should load the volume where the archive member starts, and use
10354 @samp{tar --extract --multi-volume}---@command{tar} will prompt for later
10355 volumes as it needs them. @xref{extracting archives}, for more
10356 information about extracting archives.
10358 Multi-volume archives can be modified like any other archive. To add
10359 files to a multi-volume archive, you need to only mount the last
10360 volume of the archive media (and new volumes, if needed). For all
10361 other operations, you need to use the entire archive.
10363 If a multi-volume archive was labeled using
10364 @option{--label=@var{archive-label}} (@pxref{label}) when it was
10365 created, @command{tar} will not automatically label volumes which are
10366 added later. To label subsequent volumes, specify
10367 @option{--label=@var{archive-label}} again in conjunction with the
10368 @option{--append}, @option{--update} or @option{--concatenate} operation.
10370 Notice that multi-volume support is a GNU extension and the archives
10371 created in this mode should be read only using @GNUTAR{}. If you
10372 absolutely have to process such archives using a third-party @command{tar}
10373 implementation, read @ref{Split Recovery}.
10376 @subsection Tape Files
10379 To give the archive a name which will be recorded in it, use the
10380 @option{--label=@var{volume-label}} (@option{-V @var{volume-label}})
10381 option. This will write a special block identifying
10382 @var{volume-label} as the name of the archive to the front of the
10383 archive which will be displayed when the archive is listed with
10384 @option{--list}. If you are creating a multi-volume archive with
10385 @option{--multi-volume} (@pxref{Using Multiple Tapes}), then the
10386 volume label will have @samp{Volume @var{nnn}} appended to the name
10387 you give, where @var{nnn} is the number of the volume of the archive.
10388 (If you use the @option{--label=@var{volume-label}}) option when
10389 reading an archive, it checks to make sure the label on the tape
10390 matches the one you give. @xref{label}.
10392 When @command{tar} writes an archive to tape, it creates a single
10393 tape file. If multiple archives are written to the same tape, one
10394 after the other, they each get written as separate tape files. When
10395 extracting, it is necessary to position the tape at the right place
10396 before running @command{tar}. To do this, use the @command{mt} command.
10397 For more information on the @command{mt} command and on the organization
10398 of tapes into a sequence of tape files, see @ref{mt}.
10400 People seem to often do:
10403 @kbd{--label="@var{some-prefix} `date +@var{some-format}`"}
10406 or such, for pushing a common date in all volumes or an archive set.
10409 @subsection Concatenate Volumes into a Single Archive
10412 Sometimes it is necessary to convert existing @GNUTAR{} multi-volume
10413 archive to a single @command{tar} archive. Simply concatenating all
10414 volumes into one will not work, since each volume carries an additional
10415 information at the beginning. @GNUTAR{} is shipped with the shell
10416 script @command{tarcat} designed for this purpose.
10418 The script takes a list of files comprising a multi-volume archive
10419 and creates the resulting archive at the standard output. For example:
10422 @kbd{tarcat vol.1 vol.2 vol.3 | tar tf -}
10425 The script implements a simple heuristics to determine the format of
10426 the first volume file and to decide how to process the rest of the
10427 files. However, it makes no attempt to verify whether the files are
10428 given in order or even if they are valid @command{tar} archives.
10429 It uses @command{dd} and does not filter its standard error, so you
10430 will usually see lots of spurious messages.
10432 @FIXME{The script is not installed. Should we install it?}
10435 @section Including a Label in the Archive
10436 @cindex Labeling an archive
10437 @cindex Labels on the archive media
10438 @cindex Labeling multi-volume archives
10442 To avoid problems caused by misplaced paper labels on the archive
10443 media, you can include a @dfn{label} entry---an archive member which
10444 contains the name of the archive---in the archive itself. Use the
10445 @option{--label=@var{archive-label}} (@option{-V @var{archive-label}})
10446 option in conjunction with the @option{--create} operation to include
10447 a label entry in the archive as it is being created.
10450 @item --label=@var{archive-label}
10451 @itemx -V @var{archive-label}
10452 Includes an @dfn{archive-label} at the beginning of the archive when
10453 the archive is being created, when used in conjunction with the
10454 @option{--create} operation. Checks to make sure the archive label
10455 matches the one specified (when used in conjunction with any other
10459 If you create an archive using both
10460 @option{--label=@var{archive-label}} (@option{-V @var{archive-label}})
10461 and @option{--multi-volume} (@option{-M}), each volume of the archive
10462 will have an archive label of the form @samp{@var{archive-label}
10463 Volume @var{n}}, where @var{n} is 1 for the first volume, 2 for the
10464 next, and so on. @xref{Using Multiple Tapes}, for information on
10465 creating multiple volume archives.
10467 @cindex Volume label, listing
10468 @cindex Listing volume label
10469 The volume label will be displayed by @option{--list} along with
10470 the file contents. If verbose display is requested, it will also be
10471 explicitly marked as in the example below:
10475 $ @kbd{tar --verbose --list --file=iamanarchive}
10476 V--------- 0 0 0 1992-03-07 12:01 iamalabel--Volume Header--
10477 -rw-r--r-- ringo user 40 1990-05-21 13:30 iamafilename
10481 @opindex test-label
10482 @anchor{--test-label option}
10483 However, @option{--list} option will cause listing entire
10484 contents of the archive, which may be undesirable (for example, if the
10485 archive is stored on a tape). You can request checking only the volume
10486 by specifying @option{--test-label} option. This option reads only the
10487 first block of an archive, so it can be used with slow storage
10488 devices. For example:
10492 $ @kbd{tar --test-label --file=iamanarchive}
10497 If @option{--test-label} is used with a single command line
10498 argument, @command{tar} compares the volume label with the
10499 argument. It exits with code 0 if the two strings match, and with code
10500 2 otherwise. In this case no output is displayed. For example:
10504 $ @kbd{tar --test-label --file=iamanarchive 'iamalable'}
10506 $ @kbd{tar --test-label --file=iamanarchive 'iamalable' alabel}
10511 If you request any operation, other than @option{--create}, along
10512 with using @option{--label} option, @command{tar} will first check if
10513 the archive label matches the one specified and will refuse to proceed
10514 if it does not. Use this as a safety precaution to avoid accidentally
10515 overwriting existing archives. For example, if you wish to add files
10516 to @file{archive}, presumably labeled with string @samp{My volume},
10521 $ @kbd{tar -rf archive --label 'My volume' .}
10522 tar: Archive not labeled to match `My volume'
10527 in case its label does not match. This will work even if
10528 @file{archive} is not labeled at all.
10530 Similarly, @command{tar} will refuse to list or extract the
10531 archive if its label doesn't match the @var{archive-label}
10532 specified. In those cases, @var{archive-label} argument is interpreted
10533 as a globbing-style pattern which must match the actual magnetic
10534 volume label. @xref{exclude}, for a precise description of how match
10535 is attempted@footnote{Previous versions of @command{tar} used full
10536 regular expression matching, or before that, only exact string
10537 matching, instead of wildcard matchers. We decided for the sake of
10538 simplicity to use a uniform matching device through
10539 @command{tar}.}. If the switch @option{--multi-volume} (@option{-M}) is being used,
10540 the volume label matcher will also suffix @var{archive-label} by
10541 @w{@samp{ Volume [1-9]*}} if the initial match fails, before giving
10542 up. Since the volume numbering is automatically added in labels at
10543 creation time, it sounded logical to equally help the user taking care
10544 of it when the archive is being read.
10546 The @option{--label} was once called @option{--volume}, but is not
10547 available under that name anymore.
10549 You can also use @option{--label} to get a common information on
10550 all tapes of a series. For having this information different in each
10551 series created through a single script used on a regular basis, just
10552 manage to get some date string as part of the label. For example:
10556 $ @kbd{tar cfMV /dev/tape "Daily backup for `date +%Y-%m-%d`"}
10557 $ @kbd{tar --create --file=/dev/tape --multi-volume \
10558 --volume="Daily backup for `date +%Y-%m-%d`"}
10562 Also note that each label has its own date and time, which corresponds
10563 to when @GNUTAR{} initially attempted to write it,
10564 often soon after the operator launches @command{tar} or types the
10565 carriage return telling that the next tape is ready. Comparing date
10566 labels does give an idea of tape throughput only if the delays for
10567 rewinding tapes and the operator switching them were negligible, which
10568 is usually not the case.
10571 @section Verifying Data as It is Stored
10572 @cindex Verifying a write operation
10573 @cindex Double-checking a write operation
10578 @opindex verify, short description
10579 Attempt to verify the archive after writing.
10582 This option causes @command{tar} to verify the archive after writing it.
10583 Each volume is checked after it is written, and any discrepancies
10584 are recorded on the standard error output.
10586 Verification requires that the archive be on a back-space-able medium.
10587 This means pipes, some cartridge tape drives, and some other devices
10588 cannot be verified.
10590 You can insure the accuracy of an archive by comparing files in the
10591 system with archive members. @command{tar} can compare an archive to the
10592 file system as the archive is being written, to verify a write
10593 operation, or can compare a previously written archive, to insure that
10596 @xopindex{verify, using with @option{--create}}
10597 @xopindex{create, using with @option{--verify}}
10598 To check for discrepancies in an archive immediately after it is
10599 written, use the @option{--verify} (@option{-W}) option in conjunction with
10600 the @option{--create} operation. When this option is
10601 specified, @command{tar} checks archive members against their counterparts
10602 in the file system, and reports discrepancies on the standard error.
10604 To verify an archive, you must be able to read it from before the end
10605 of the last written entry. This option is useful for detecting data
10606 errors on some tapes. Archives written to pipes, some cartridge tape
10607 drives, and some other devices cannot be verified.
10609 One can explicitly compare an already made archive with the file
10610 system by using the @option{--compare} (@option{--diff}, @option{-d})
10611 option, instead of using the more automatic @option{--verify} option.
10614 Note that these two options have a slightly different intent. The
10615 @option{--compare} option checks how identical are the logical contents of some
10616 archive with what is on your disks, while the @option{--verify} option is
10617 really for checking if the physical contents agree and if the recording
10618 media itself is of dependable quality. So, for the @option{--verify}
10619 operation, @command{tar} tries to defeat all in-memory cache pertaining to
10620 the archive, while it lets the speed optimization undisturbed for the
10621 @option{--compare} option. If you nevertheless use @option{--compare} for
10622 media verification, you may have to defeat the in-memory cache yourself,
10623 maybe by opening and reclosing the door latch of your recording unit,
10624 forcing some doubt in your operating system about the fact this is really
10625 the same volume as the one just written or read.
10627 The @option{--verify} option would not be necessary if drivers were indeed
10628 able to detect dependably all write failures. This sometimes require many
10629 magnetic heads, some able to read after the writes occurred. One would
10630 not say that drivers unable to detect all cases are necessarily flawed,
10631 as long as programming is concerned.
10633 The @option{--verify} (@option{-W}) option will not work in
10634 conjunction with the @option{--multi-volume} (@option{-M}) option or
10635 the @option{--append} (@option{-r}), @option{--update} (@option{-u})
10636 and @option{--delete} operations. @xref{Operations}, for more
10637 information on these operations.
10639 Also, since @command{tar} normally strips leading @samp{/} from file
10640 names (@pxref{absolute}), a command like @samp{tar --verify -cf
10641 /tmp/foo.tar /etc} will work as desired only if the working directory is
10642 @file{/}, as @command{tar} uses the archive's relative member names
10643 (e.g., @file{etc/motd}) when verifying the archive.
10645 @node Write Protection
10646 @section Write Protection
10648 Almost all tapes and diskettes, and in a few rare cases, even disks can
10649 be @dfn{write protected}, to protect data on them from being changed.
10650 Once an archive is written, you should write protect the media to prevent
10651 the archive from being accidentally overwritten or deleted. (This will
10652 protect the archive from being changed with a tape or floppy drive---it
10653 will not protect it from magnet fields or other physical hazards).
10655 The write protection device itself is usually an integral part of the
10656 physical media, and can be a two position (write enabled/write
10657 disabled) switch, a notch which can be popped out or covered, a ring
10658 which can be removed from the center of a tape reel, or some other
10659 changeable feature.
10664 This appendix lists some important user-visible changes between
10665 version @GNUTAR{} @value{VERSION} and previous versions. An up-to-date
10666 version of this document is available at
10667 @uref{http://www.gnu.org/@/software/@/tar/manual/changes.html,the
10668 @GNUTAR{} documentation page}.
10671 @item Use of globbing patterns when listing and extracting.
10673 Previous versions of GNU tar assumed shell-style globbing when
10674 extracting from or listing an archive. For example:
10677 $ @kbd{tar xf foo.tar '*.c'}
10680 would extract all files whose names end in @samp{.c}. This behavior
10681 was not documented and was incompatible with traditional tar
10682 implementations. Therefore, starting from version 1.15.91, GNU tar
10683 no longer uses globbing by default. For example, the above invocation
10684 is now interpreted as a request to extract from the archive the file
10687 To facilitate transition to the new behavior for those users who got
10688 used to the previous incorrect one, @command{tar} will print a warning
10689 if it finds out that a requested member was not found in the archive
10690 and its name looks like a globbing pattern. For example:
10693 $ @kbd{tar xf foo.tar '*.c'}
10694 tar: Pattern matching characters used in file names. Please,
10695 tar: use --wildcards to enable pattern matching, or --no-wildcards to
10696 tar: suppress this warning.
10697 tar: *.c: Not found in archive
10698 tar: Error exit delayed from previous errors
10701 To treat member names as globbing patterns, use --wildcards option.
10702 If you want to tar to mimic the behavior of versions prior to 1.15.91,
10703 add this option to your @env{TAR_OPTIONS} variable.
10705 @xref{wildcards}, for the detailed discussion of the use of globbing
10706 patterns by @GNUTAR{}.
10708 @item Use of short option @option{-o}.
10710 Earlier versions of @GNUTAR{} understood @option{-o} command line
10711 option as a synonym for @option{--old-archive}.
10713 @GNUTAR{} starting from version 1.13.90 understands this option as
10714 a synonym for @option{--no-same-owner}. This is compatible with
10715 UNIX98 @command{tar} implementations.
10717 However, to facilitate transition, @option{-o} option retains its
10718 old semantics when it is used with one of archive-creation commands.
10719 Users are encouraged to use @option{--format=oldgnu} instead.
10721 It is especially important, since versions of @acronym{GNU} Automake
10722 up to and including 1.8.4 invoke tar with this option to produce
10723 distribution tarballs. @xref{Formats,v7}, for the detailed discussion
10724 of this issue and its implications.
10726 @FIXME{Change the first argument to tar-formats when the new Automake is
10727 out. The proposition to add @anchor{} to the appropriate place of its
10728 docs was accepted by Automake people --Sergey 2006-05-25}.
10729 @xref{Options, tar-v7, Changing Automake's Behavior,
10730 automake, GNU Automake}, for a description on how to use various
10731 archive formats with @command{automake}.
10733 Future versions of @GNUTAR{} will understand @option{-o} only as a
10734 synonym for @option{--no-same-owner}.
10736 @item Use of short option @option{-l}
10738 Earlier versions of @GNUTAR{} understood @option{-l} option as a
10739 synonym for @option{--one-file-system}. Since such usage contradicted
10740 to UNIX98 specification and harmed compatibility with other
10741 implementation, it was declared deprecated in version 1.14. However,
10742 to facilitate transition to its new semantics, it was supported by
10743 versions 1.15 and 1.15.90. The present use of @option{-l} as a short
10744 variant of @option{--check-links} was introduced in version 1.15.91.
10746 @item Use of options @option{--portability} and @option{--old-archive}
10748 These options are deprecated. Please use @option{--format=v7} instead.
10750 @item Use of option @option{--posix}
10752 This option is deprecated. Please use @option{--format=posix} instead.
10755 @node Configuring Help Summary
10756 @appendix Configuring Help Summary
10758 Running @kbd{tar --help} displays the short @command{tar} option
10759 summary (@pxref{help}). This summary is organized by @dfn{groups} of
10760 semantically close options. The options within each group are printed
10761 in the following order: a short option, eventually followed by a list
10762 of corresponding long option names, followed by a short description of
10763 the option. For example, here is an excerpt from the actual @kbd{tar
10767 Main operation mode:
10769 -A, --catenate, --concatenate append tar files to an archive
10770 -c, --create create a new archive
10771 -d, --diff, --compare find differences between archive and
10773 --delete delete from the archive
10776 @vrindex ARGP_HELP_FMT, environment variable
10777 The exact visual representation of the help output is configurable via
10778 @env{ARGP_HELP_FMT} environment variable. The value of this variable
10779 is a comma-separated list of @dfn{format variable} assignments. There
10780 are two kinds of format variables. An @dfn{offset variable} keeps the
10781 offset of some part of help output text from the leftmost column on
10782 the screen. A @dfn{boolean} variable is a flag that toggles some
10783 output feature on or off. Depending on the type of the corresponding
10784 variable, there are two kinds of assignments:
10787 @item Offset assignment
10789 The assignment to an offset variable has the following syntax:
10792 @var{variable}=@var{value}
10796 where @var{variable} is the variable name, and @var{value} is a
10797 numeric value to be assigned to the variable.
10799 @item Boolean assignment
10801 To assign @code{true} value to a variable, simply put this variable name. To
10802 assign @code{false} value, prefix the variable name with @samp{no-}. For
10807 # Assign @code{true} value:
10809 # Assign @code{false} value:
10815 Following variables are declared:
10817 @deftypevr {Help Output} boolean dup-args
10818 If true, arguments for an option are shown with both short and long
10819 options, even when a given option has both forms, for example:
10822 -f ARCHIVE, --file=ARCHIVE use archive file or device ARCHIVE
10825 If false, then if an option has both short and long forms, the
10826 argument is only shown with the long one, for example:
10829 -f, --file=ARCHIVE use archive file or device ARCHIVE
10833 and a message indicating that the argument is applicable to both
10834 forms is printed below the options. This message can be disabled
10835 using @code{dup-args-note} (see below).
10837 The default is false.
10840 @deftypevr {Help Output} boolean dup-args-note
10841 If this variable is true, which is the default, the following notice
10842 is displayed at the end of the help output:
10845 Mandatory or optional arguments to long options are also mandatory or
10846 optional for any corresponding short options.
10849 Setting @code{no-dup-args-note} inhibits this message. Normally, only one of
10850 variables @code{dup-args} or @code{dup-args-note} should be set.
10853 @deftypevr {Help Output} offset short-opt-col
10854 Column in which short options start. Default is 2.
10858 $ @kbd{tar --help|grep ARCHIVE}
10859 -f, --file=ARCHIVE use archive file or device ARCHIVE
10860 $ @kbd{ARGP_HELP_FMT=short-opt-col=6 tar --help|grep ARCHIVE}
10861 -f, --file=ARCHIVE use archive file or device ARCHIVE
10866 @deftypevr {Help Output} offset long-opt-col
10867 Column in which long options start. Default is 6. For example:
10871 $ @kbd{tar --help|grep ARCHIVE}
10872 -f, --file=ARCHIVE use archive file or device ARCHIVE
10873 $ @kbd{ARGP_HELP_FMT=long-opt-col=16 tar --help|grep ARCHIVE}
10874 -f, --file=ARCHIVE use archive file or device ARCHIVE
10879 @deftypevr {Help Output} offset doc-opt-col
10880 Column in which @dfn{doc options} start. A doc option isn't actually
10881 an option, but rather an arbitrary piece of documentation that is
10882 displayed in much the same manner as the options. For example, in
10883 the description of @option{--format} option:
10887 -H, --format=FORMAT create archive of the given format.
10889 FORMAT is one of the following:
10891 gnu GNU tar 1.13.x format
10892 oldgnu GNU format as per tar <= 1.12
10893 pax POSIX 1003.1-2001 (pax) format
10895 ustar POSIX 1003.1-1988 (ustar) format
10896 v7 old V7 tar format
10901 the format names are doc options. Thus, if you set
10902 @kbd{ARGP_HELP_FMT=doc-opt-col=6} the above part of the help output
10903 will look as follows:
10907 -H, --format=FORMAT create archive of the given format.
10909 FORMAT is one of the following:
10911 gnu GNU tar 1.13.x format
10912 oldgnu GNU format as per tar <= 1.12
10913 pax POSIX 1003.1-2001 (pax) format
10915 ustar POSIX 1003.1-1988 (ustar) format
10916 v7 old V7 tar format
10921 @deftypevr {Help Output} offset opt-doc-col
10922 Column in which option description starts. Default is 29.
10926 $ @kbd{tar --help|grep ARCHIVE}
10927 -f, --file=ARCHIVE use archive file or device ARCHIVE
10928 $ @kbd{ARGP_HELP_FMT=opt-doc-col=19 tar --help|grep ARCHIVE}
10929 -f, --file=ARCHIVE use archive file or device ARCHIVE
10930 $ @kbd{ARGP_HELP_FMT=opt-doc-col=9 tar --help|grep ARCHIVE}
10932 use archive file or device ARCHIVE
10937 Notice, that the description starts on a separate line if
10938 @code{opt-doc-col} value is too small.
10941 @deftypevr {Help Output} offset header-col
10942 Column in which @dfn{group headers} are printed. A group header is a
10943 descriptive text preceding an option group. For example, in the
10947 Main operation mode:
10949 -A, --catenate, --concatenate append tar files to
10951 -c, --create create a new archive
10954 @samp{Main operation mode:} is the group header.
10956 The default value is 1.
10959 @deftypevr {Help Output} offset usage-indent
10960 Indentation of wrapped usage lines. Affects @option{--usage}
10961 output. Default is 12.
10964 @deftypevr {Help Output} offset rmargin
10965 Right margin of the text output. Used for wrapping.
10968 @node Fixing Snapshot Files
10969 @appendix Fixing Snapshot Files
10970 @include tar-snapshot-edit.texi
10972 @node Tar Internals
10973 @appendix Tar Internals
10974 @include intern.texi
10978 @include genfile.texi
10980 @node Free Software Needs Free Documentation
10981 @appendix Free Software Needs Free Documentation
10982 @include freemanuals.texi
10984 @node Copying This Manual
10985 @appendix Copying This Manual
10988 * GNU Free Documentation License:: License for copying this manual
10993 @node Index of Command Line Options
10994 @appendix Index of Command Line Options
10996 This appendix contains an index of all @GNUTAR{} long command line
10997 options. The options are listed without the preceding double-dash.
10998 For a cross-reference of short command line options, @ref{Short Option Summary}.
11011 @c Local variables:
11012 @c texinfo-column-for-description: 32
projects / chaz / tar / blob