fmerge - merge files
fmerge [ option... ] basefile fileA fileB
The fmerge program is used to compare the changes between two different
descendants of a base file, and creates an output file which contains
both sets of changes. This is useful when two users both take the same
version of a file and make independent edits to it, and then later want
to create a file which contains both sets of edits. In such a use, the
original file that both sets of edits is derived from is called the
base file. The two files containing the edits are called file A and
fmerge basefile fileA fileB -o outputfile
produces the output file which contains the edits contained in file A
and file B, based on the base file . If the -Output option was not
used, or if no outputfile is specified, then the merged lines are typed
to the standard output. The order of specifying file A and file B is
The fmerge program can also be used to remove earlier edits made to a
module. To do this, make the version containing the edits you want to
delete be the basefile. Make the version previous to the edit you want
deleted be file A. Finally, make the most recent version of the file
which contains the other edits (including the one you want deleted) be
file B. Then the result of merging will be the newest version of the
module minus the changes made by the edit you wanted removed. For
example, if three successive versions of some module have the names
edit10, edit11 and edit12, and you want the changes done by edit11 to
be undone, but still want the changes done by edit12, then you use the
fmerge edit11 edit10 edit12 -o outputfile
While merging the two sets of edits, fmerge may discover conflicts. A
conflict occurs when the same line of the base file is changed by both
of the two sets of edits. The change can be due to new lines being
inserted, lines being deleted, or both. When conflicts occur, the
output file contains conflict identification lines, which are lines
containing the string ////. These lines indicate the region where
the two sets of edits are incompatible. You must then edit the output
file and remove these lines, and in addition correct the conflicts
manually in order to produce the correct result.
The following options are understood:
-Conflicts [ conflictfile ]
Since conflicts due to deletions are invisible in the output
file, and inserts do not specify which of the two edits
inserted the lines, there is an alternative output format from
the fmerge program. This output format describes what happens
to each line of the base file, so that conflicts are easier to
detect and fix. The command:
fmerge basefile fileA fileB -c conflictfile
produces the file describing the results of the merge in
detail. If the -Conflicts option is specified without any
conflictfile name, then the conflicts are send to the standard
If there are conflicts, and the -Conflicts options is not
specified, the fmerge program will exit with a status of 1.
The conflict file contains lines which contain three characters
and then some text. The first three characters describe what
is happening to the base file at that point. These characters
are the following:
IA This line was inserted by file A.
DA This line was deleted by file A.
IB This line was inserted by file B.
DB This line was deleted by file B.
This line is unchanged.
X This is a conflict identification line.
U There are unspecified unchanged lines here.
Each set of conflicts is flagged by three identification lines.
The first line indicates the beginning of the conflict, and
specifies the line numbers for the base file and two divergent
files. The second conflict identification line separates lines
changed by file A from lines changed by file B. The third
conflict identification indicates the end of the conflict.
You can edit this conflict file to remove the conflicts. This
involves deleting the conflict identification lines, and
changing the conflicting lines as necessary to fix the
conflict. While doing this, remember to leave three blank
characters at the front of any new lines you insert while
correcting the conflicts. When you are done, there should be
no lines which begin with an ’X’ in the file. All other lines
can remain. Then you can use the command:
fmerge conflictfile -o outputfile
to create the new output file which has the desired data. Once
again, if no -Output option or outputfile is used, the output
is send to the standard output.
Besides physical conflicts, there can be logical conflicts.
These are changes made to different lines in the base file such
that the program is no longer correct. Such conflicts cannot
be detected by a program, and so these must be checked
manually. In order to make this process easier, the -Unchanged
option can be used to reduce the size of the conflict file to
only include regions near changed lines. This file can then be
examined in order to detect possible logical conflicts. As an
example, the command:
fmerge basefile fileA fileB -c -u 3
will send to the standard output all changes made by either
sets of edits, with only three unchanged lines surrounding each
When using the -Unchanged option, the conflict file will
contain lines starting with ’U’. These represent unchanged
lines, and the number following the letter is the number of
unchanged lines. The resulting conflict file cannot be read to
produce an output file because of the missing lines. If this
is attempted, an error will be generated.
It is possible to use both -Output and -Conflicts in the same
command. Thus you can produce the output file which you hope
is correct, and also produce the conflict file which you can
use to check for logical conflicts.
-Verbose [ number ]
This option can be specified with any other action, and outputs
status information about the progress of the action. This is
useful for debugging of problems, or just for amusement when
the system is slow or a large file is being processed. It
accepts a numeric argument to indicate the verbosity for
output. The levels are as follows:
0 No output at all (except for errors).
1 Single‐line output describing action (default).
2 Detailed status as action proceeds.
This option restricts the number of physical conflicts that are
allowed before failing. This is used if you are not interested
in the results if there are too many conflicts.
Give some help on how to use the fmerge program.
Ignore all conflicts.
The option may be used to suppress conflicts which make
identical deletes, or identical inserts, or identical changes.
This is often desirable when merging two source code branches.
Show what version of fmerge is running.
All options may be abbreviated; the abbreviation is documented as the
upper case letters, all lower case letters and underscores (_) are
optional. You must use consecutive sequences of optional letters.
All options are case insensitive, you may type them in upper case or
lower case or a combination of both, case is not important.
For example: the arguments "-help, "-HELP" and "-h" are all interpreted
to mean the -Help option. The argument "-hlp" will not be understood,
because consecutive optional characters were not supplied.
Options and other command line arguments may be mixed arbitrarily on
the command line.
The GNU long option names are understood. Since all option names for
fmerge are long, this means ignoring the extra leading ’-’. The
"-option=value" convention is also understood.
FILE NAME EXPANSION
As a convenience, if a pathname begins with a period and a environment
variable exists with that name, then the value of the environment
variable will be used as the actual pathname. For example, if a
environment variable of .FOO has the value this.is.a.long.name, then
fmerge -o .FOO
is actually equivilant to the command
fmerge -o this.is.a.long.name
If you want to prevent the expansion of a pathname which begins with a
period, then you can use an alternate form for the pathname, as in:
fmerge -o ./.FOO
In general, fmerge can handle all text files you throw at it, even
international text with unusual encodings. However, fmerge is unable
to cope elegantly with files which contain the NUL character.
The fcomp(1) program simply prints a warning, and continues, you need
to know that it converts NUL characters into an 0x80 value before
performing the comparison.
The fmerge(1) program also converts the NUL character to an 0x80 value
before merging, after a warning, and any output file will contain this
value, rather than the original NUL character.
The fhist(1) program, however, generates a fatal error if any input
file contains NUL characters. This is intended to protect your source
files for unintentional corruption. Use -BINary for files which
absolutely must contain NUL characters.
The fmerge program will exit with a status of 1 on any error. The
fmerge program will only exit with a status of 0 if there are no
This program is based on the algorithm in
An O(ND) Difference Algorithm and Its Variations, Eugene W.
Myers, TR 85‐6, 10‐April‐1985, Department of Computer Science,
The University of Arizona, Tuscon, Arizona 85721.
A File Comparison Program, Webb Miller and Eugene W. Myers,
Software Practice and Experience, Volume 15, No. 11, November
fmerge version 1.18.D001
Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
2000, 2001, 2002, 2003, 2004, 2005, 2006, 2008, 2009 Peter Miller;
This program is derived from a work
Copyright (C) 1990 David I. Bell.
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 3 of the License, or (at your
option) any later version.
This program is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program. If not, see <http://www.gnu.org/licenses/>.
Peter Miller Web: http://miller.emu.id.au/pmiller/
/\/\* E‐Mail: firstname.lastname@example.org
David I. Bell Web: http://www.canb.auug.org.au/~dbell