Files
tar/doc/dumpdir.texi
Paul Eggert 0ab5e64ac0 tar: remove trailing white space from source files
* ChangeLog.1, ChangeLog.CVS, Makefile.am, NEWS, README:
* README-hacking, directory, doc/Makefile.am, doc/dumpdir.texi:
* doc/gendocs_template, doc/intern.texi, doc/mastermenu.el:
* doc/snapshot.texi, doc/sparse.texi, doc/tar-snapshot-edit.texi:
* doc/value.texi, lib/Makefile.am, scripts/backup-specs:
* scripts/dump-remind.in, scripts/tar-snapshot-edit, scripts/tarcat:
* scripts/xsparse.c, src/arith.h, src/buffer.c, src/compare.c:
* src/create.c, src/delete.c, src/exit.c, src/suffix.c, src/tar.c:
* src/tar.h, src/update.c, src/warning.c, src/xheader.c:
* tests/append01.at, tests/append02.at, tests/atlocal.in:
* tests/delete03.at, tests/exclude.at, tests/exclude06.at:
* tests/extrac04.at, tests/extrac05.at, tests/extrac06.at:
* tests/extrac07.at, tests/filerem01.at, tests/filerem02.at:
* tests/incr01.at, tests/incr02.at, tests/incr03.at, tests/incr06.at:
* tests/label02.at, tests/label03.at, tests/label04.at:
* tests/label05.at, tests/link02.at, tests/link03.at:
* tests/listed01.at, tests/listed02.at, tests/long01.at:
* tests/longv7.at, tests/multiv01.at, tests/multiv02.at:
* tests/multiv03.at, tests/multiv05.at, tests/multiv06.at:
* tests/multiv07.at, tests/multiv08.at, tests/options.at:
* tests/options02.at, tests/remfiles03.at, tests/rename01.at:
* tests/rename02.at, tests/rename03.at, tests/rename04.at:
* tests/rename05.at, tests/same-order01.at, tests/same-order02.at:
* tests/shortfile.at, tests/shortupd.at, tests/sparse01.at:
* tests/sparse02.at, tests/sparsemv.at, tests/sparsemvp.at:
* tests/star/README, tests/star/gtarfail2.at:
* tests/star/multi-fail.at:
* tests/star/pax-big-10g.at, tests/star/quicktest.sh:
* tests/star/ustar-big-2g.at, tests/star/ustar-big-8g.at:
* tests/update01.at, tests/update02.at, tests/volsize.at:
* tests/volume.at:
Remove trailing spaces and tabs from lines, and remove
trailing empty lines from files.  This makes it a bit easier
to share code among coreutils and other projects that do this.
2010-08-19 15:50:07 -07:00

133 lines
3.7 KiB
Plaintext

@c This is part of the paxutils manual.
@c Copyright (C) 2006, 2007 Free Software Foundation, Inc.
@c Written by Sergey Poznyakoff
@c This file is distributed under GFDL 1.1 or any later version
@c published by the Free Software Foundation.
Incremental archives keep information about contents of each
dumped directory in special data blocks called @dfn{dumpdirs}.
Dumpdir is a sequence of entries of the following form:
@smallexample
@var{C} @var{filename} \0
@end smallexample
@noindent
where @var{C} is one of the @dfn{control codes} described below,
@var{filename} is the name of the file @var{C} operates upon, and
@samp{\0} represents a nul character (ASCII 0). The white space
characters were added for readability, real dumpdirs do not contain
them.
Each dumpdir ends with a single nul character.
The following table describes control codes and their meanings:
@table @samp
@item Y
@var{filename} is contained in the archive.
@item N
@var{filename} was present in the directory at the time the archive
was made, yet it was not dumped to the archive, because it had not
changed since the last backup.
@item D
@var{filename} is a directory.
@item R
This code requests renaming of the @var{filename} to the name
specified with the @samp{T} command, that immediately follows it.
@item T
Specify target file name for @samp{R} command (see below).
@item X
Specify @dfn{temporary directory} name for a rename operation (see below).
@end table
Codes @samp{Y}, @samp{N} and @samp{D} require @var{filename} argument
to be a relative file name to the directory this dumpdir describes,
whereas codes @samp{R}, @samp{T} and @samp{X} require their argument
to be an absolute file name.
The three codes @samp{R}, @samp{T} and @samp{X} specify a
@dfn{renaming operation}. In the simplest case it is:
@smallexample
R@file{source}\0T@file{dest}\0
@end smallexample
@noindent
which means ``rename file @file{source} to file @file{dest}''.
However, there are cases that require using a @dfn{temporary
directory}. For example, consider the following scenario:
@enumerate 1
@item
Previous run dumped a directory @file{foo} which contained the
following three directories:
@smallexample
a
b
c
@end smallexample
@item
They were renamed @emph{cyclically}, so that:
@example
@file{a} became @file{b}
@file{b} became @file{c}
@file{c} became @file{a}
@end example
@item
New incremental dump was made.
@end enumerate
This case cannot be handled by three successive renames, since
renaming @file{a} to @file{b} will destroy the existing directory.
To correctly process it, @GNUTAR{} needs a temporary directory, so
it creates the following dumpdir (newlines have been added for
readability):
@smallexample
@group
Xfoo\0
Rfoo/a\0T\0
Rfoo/b\0Tfoo/c\0
Rfoo/c\0Tfoo/a\0
R\0Tfoo/a\0
@end group
@end smallexample
The first command, @samp{Xfoo\0}, instructs the extractor to create a
temporary directory in the directory @file{foo}. Second command,
@samp{Rfoo/aT\0}, says ``rename file @file{foo/a} to the temporary
directory that has just been created'' (empty file name after a
command means use temporary directory). Third and fourth commands
work as usual, and, finally, the last command, @samp{R\0Tfoo/a\0}
tells tar to rename the temporary directory to @file{foo/a}.
The exact placement of a dumpdir in the archive depends on the
archive format (@pxref{Formats}):
@itemize
@item PAX archives
In PAX archives, dumpdir is stored in the extended header of the
corresponding directory, in variable @code{GNU.dumpdir}.
@item GNU and old GNU archives
These formats implement special header type @samp{D}, which is similar
to ustar header @samp{5} (directory), except that it precedes a data
block containing the dumpdir.
@end itemize
@c End of dumpdir.texi