fhist - file history
program is used to keep track of the successive versions of a
file. Using this program, you can remember all of your changes to a file, and
get back any one of the old versions. The uses of this ability are:
- You can make a series of tentative edits to the file, and
if necessary back up to the last "good" edit.
- You can delete old subroutines and code from your file
which are obsolete, but still be able to get them back in the future in
case a need for them arises.
- You can compare two versions of the file to see how you
fixed some old problem, so that you can check up on the correctness of the
fix at a later date.
- You get a record of your remarks for each version, so that
you can quickly know what bugs were fixed, and what features were
- The date the file was last edited can be automatically
stored in the file.
program manipulates modules. A module is simply any text file
that you are interested in keeping versions of. For example, a source file
is a module, and so is a documentation file howto.doc
The module name includes the suffix of the file (as in the above examples).
However, pathnames are not part of a module name, so that
cannot be a legal module name. A module name is
limited to 12 characters since the fhist
program needs two extra
characters for its own purpose.
It is possible to have information about the state of the file inserted into the
file. See the -Modify
for more infromation.
The following options are understood:
- -Path pathname
Modules are stored in a directory, called the
module storage directory. The default directory is FHIST
, and therefore
is located relative to your current directory. This is convenient when you are
in a directory containing many modules, and you want a local storage directory
to contain just those modules. If you use the -p
option, then you can
locate the storage directory anywhere you choose. This is useful if you choose
to have a common storage directory for all of your files, independent of where
they actually are used.
The files inside of the storage directory should not be changed by you. Doing so
will probably corrupt your edit history, causing errors when you extract old
revisions. For your information, though, each module is stored as two files in
the directory. The one with the .s
suffix is a copy of the newest
version of the module, with one extra line at the beginning. The one with the
suffix is the edit history of the module, and contains the
information needed to extract previous revisions of the module. Thus if the
edit history is ever corrupted, you will at least have the most recent version
of the module.
- This option may be used to request that the path directory
be created automatically if it does not yet exist. This works for both the
directory specified by the -Path option, and for the default. Intermediate
directories will also be created if necessary.
- This option may be used to specify that the file is binary,
that it may contain NUL characters. It is essential that you have
consistent presence or absence of the -BINary option for each file
when combined with the -CReate, -Update,
-Conditional_Update and -Extract options. Failure to do so
will produce inconsistent results. Note: this is different behaviour to
the fcomp(1) option of the same name. Note: the -BINary
option does not imply the -No‐Keywords option.
To use the fhist
program for the first
time, you need to create your storage directory. Therefore, cd
directory where you want it to be, which is probably the directory containing
the modules you want to save the revisions of. Then create the directory
(or some other name if you don't want to use the default name).
To start using a module under fhist
, you must first use the
option. This creates the initial edit for that module in the
storage directory, with the contents of the specified module as the initial
edit. Thus, if you have a source file prog.c
, then the command:
creates the initial edit of the module. As part of this process, you are asked
to provide remarks about the file. These remarks can be seen later using the
option (described below). After the remarks have been typed, the
contents of the file are then saved. You can then delete the file
if desired, and fhist
would be able to recreate it later.
Or you can leave it there as the working copy of the module.
option may be combined with the -Update
options to create the file if required.
To save another revision of the module, you
use the -Update
option. This updates the files in the storage directory
to include the latest changes. Remarks are again asked for so that you can
document why you made this edit. Thus, to continue the example, after editing
, the command:
will save the changes as a new edit. This command compares the newest version of
the module to the previous version, saves the differences in the .e
file, and copies the new source to the .s
file. At this point, you can
once again delete the prog.c
file if desired, and later get back either
of the two versions of the program.
program handles quota or disk full problems during a create or
update operation without damage occurring to the edit history files. If an
edit cannot be completed because of such problems, the edit is backed out
completely, and you will get an error message about the disk problem. There is
no need for any error recovery in this case, other than retrying the update
when more disk space is available. The fhist
program also disables
signals during the critical file operations, so you do not have to worry about
damaging the edit history files because of attempts to quit out of the
option may be combined with the -Update
options to create the file if required.
- -Input filename
In either the -CReate
options, the file containing the new version of the module defaults to the
same name as the module. In the example, the module prog.c
and updated from the data in the file prog.c
. When you wish the data to
come from some other file, you can use the -Input
specifies the input file to use for the data. For example, if you wanted to
, but from a filename called newprog.c
, then the
fhist prog.c -u -i newprog.c
would save a new revision of module prog.c
, but with the data that was in
the file newprog.c
. In this case, the file prog.c
does not have
to exist, and isn't referenced even if it did exist. Again, once the update is
complete, you could delete the newprog.c
file if desired and then later
you can retrieve its contents.
Remarks can be read from a file instead of
from the terminal. The -Remarks
option can be used to specify a file
name containing the remarks. If there is no file name following the
option, then no remarks at all are used. The command:
would create a new revision of prog.c
without asking for or saving any
remarks about the edit.
- -Remark_String text
- It is also possible to specify the remarks directly on the
command line. You may only use this option once.
- -Extract [ edit ]
To retrieve a previous revision of a module,
you specify the name of the module and use the -Extract
specify the edit number you want retrieved. Edit numbers are assigned
sequentially starting with 1. Thus the initial version of the module has edit
number 1, the first revision has edit number 2, and so on until the latest
revision. If the -Extract
option is not used, or if no edit number is
supplied for it, then the latest edit number is extracted. Therefore, this is
the default action if no options at all are specified.
Edit numbers can also be zero, negative, or be a name with an optional offset.
The number zero represents the latest edit number, and negative numbers
indicate edit numbers backwards from the latest edit number. Edit names
represent edit numbers whose name had been set by using the -Name
option. For example, if edit number 10 was associated with the name
, then the edit name foo
represents 10, foo‐4
represents edit number6, and foo+2
represents edit number 12. The
special reserved names oldest
refer to the oldest and
newest versions of the module in the edit history.
As an example of retrievals, assume that you have saved ten versions of the
. The following commands will then extract the versions of
the file with the specified edit numbers:
- fhist prog.c
- version 10 (the latest)
- fhist prog.c -e 9
- version 9 (the version just prior)
- fhist prog.c -e oldest
- version 1 (the oldest version)
- fhist prog.c -e -2
- version 8 (latest version - 2)
The output filename is again defaulted to the module name. So when the module
is extracted, the specified version of the module is written to
In order to prevent accidental overwriting of a file, the fhist
will by default ask you if overwriting is permitted if that would occur. A
common mistake is to edit prog.c
, and then try to update the module,
but forget to specify the -u
option. Then the fhist
would try to extract the newest version of the module, and thus overwrite the
file with the new changes. Asking the question allows you to notice your
mistake, and prevent the overwriting.
- -Output filename
You can change the output filename using the
option. Thus, the command:
fhist prog.c -o newprog.c
will extract the latest version of the module prog.c
, and put it into the
. Once again, the file "prog.c" is ignored,
whether or not it existed.
This option will force overwriting of the file, thus never asking you if
overwriting is permitted. This is often useful in shell scripts, or when
you are sure that you want to overwrite any existing file.
This option is the no‐overwrite option, and will cause any existing
files to not be overwritten, again without asking you. This is
useful if you already have some of the modules in your directory, and you
want to extract the rest of the modules without overwriting the ones you
already have. Specifying both -Fore_Write and -No_Write is
- -Terminal [ edit ]
This option is used to output an extracted module to the standard output,
instead of writing it to a file. This is useful in order to view the
beginning of a version of the file. This can be interrupted if you do not
want to see the whole file.
- -Modify number
When extracting a file, the fhist
program looks for and updates special character sequences in the first few
lines of the file. These special sequences are used for documentation
purposes, such as describing the edit number the file is from. For speed of
extraction and updating, these sequences are usually limited to the first 25
lines of the file, since the fhist
program then does not have to
examine the entire file. The -Modify
option can be used to change the
number of lines to be modified from the default value of 25. Specifying zero
totally disables the special character sequences, whereas specifying a very
large number will cause the sequences to be checked for each line of the file
(and thus slow the fhist
Each special sequence is of the form [# keyword value, keyword value, ...,
keyword value #] ,
where each keyword
describes an item, and each
is the value for the preceding keyword. The keywords can be in
upper or lower case, or both. The single space following the [#
following each comma, and preceding the #]
must be present. If the
sequence is wrong, an unknown keyword is used, the line is longer than 200
characters, or more than four keywords are used, then the whole line will not
be changed. The current keywords which can be used are the following:
- The edit number
- The date that the edit was created
- The user name of the user who created the edit
- The module name
In order to use this special character sequence, you simply insert it into your
module inside of a comment (within the first few lines). When this is done,
the value parts of the sequence can be null. For example, if you want to put a
special sequence into a program called delete.c
, then you could edit
the first few lines as follows:
* Delete - program to delete files
* [# Edit, Date #]
When an extract is done, the proper edit number and date are automatically
inserted as the new values. Thus, if you extract edit 23 of the module
which had been created on 8 August 89, then the resulting file
* Delete - program to delete files
* [# Edit 23, Date 8‐Aug‐89 #]
When updating a module, it is never necessary to edit these sequences, as any
old values will be removed and replaced with the new ones. Also, when using
options (described below), lines with these
sequences compare as if the values were null, and thus will not cause spurious
During an update, the special character sequences are read and any edit value
found is compared against the current edit number of the module. If they
differ, then the update fails. This provides an interlock check for the case
of two users extracting the same version of a file, editing it, and then both
updating it without knowledge of each other. In this case, the second user
would fail, and then he can merge his edits with the previous user's edit and
then retry the update. This checking is disabled if there is no special
character sequence containing the edit keyword, the edit number value is null,
or if the -Forced_Update
option is used to indicate that the check is
- This option may be used to disable the use of the keyword
special character sequences described above. Text containing keyword
sequences is treated as plain text. Note: the -No_Keywords option
does not imply the -BINary option.
- -Name string
This option is used to associate a name for the newest version of a module.
It can be given along with the -CReate, -Update, or
-Difference_Update options, to specify a name for the new version
of the module. It can also be given by itself in order to specify a name
for the newest version of a module. Each edit number can have many names
associated with it, so this will not remove any previously defined name
for the edit. This option is useful to correlate many modules together.
For example, when a new version of a program is ready to be released, you
could give each module of the program the same name release1. Then
in the future, you can recreate the sources making up that release by
extracting the edits with the name release1 for every module. Edit
names cannot begin with a digit, and cannot contain plus or minus signs.
These rules prevent ambiguous parsing of edit numbers for the
-Extract, -Terminal, -ALL, and -List
- -List [ edit1 [ edit2 ]]
This option prints a list of edits for the module, giving the user name,
date, user remarks, and names specified for the edits. If no edit number
is supplied, then all edits are printed in reverse order. If a single edit
number is supplied, then only that edit number is printed. If two edit
numbers are supplied, then all edits in the specified range are printed.
The output from this option defaults to the terminal. You can use the
-Output option to save the results to a file.
- -Difference [ edit1 [ edit2 ]]
This option is used to display the differences
between two versions of a module, or a file and a version of a module. There
are three modes for this action, depending on how many edit numbers are
supplied. These modes are illustrated by the following examples:
- fhist foo.c -d
- Compare latest version against file "foo.c"
- fhist foo.c -d 3
- Compare version 3 against file "foo.c"
- fhist foo.c -d 3 4
- Compare version 3 against version 4
This option accepts the -Input
option to specify the file to be compared.
When using the -Difference
option, the output defaults to the terminal.
Therefore, you must use -Output
if you wish the differences saved to a
file. Using -Quick
will only output a quick
summary of the changes, instead of the detailed changes. This summary only
supplies the number of lines inserted, deleted, and unchanged between the
files. Using -What
will display all of both
files, showing in detail what the differences are using change bars.
option may need to write one or two temporary files in
order to extract old versions of a module to be compared. These files have
names like T$n_nnn .
They are deleted again just before differences are
output, so that stopping the output before it is complete will not leave these
files around. The temporary files are usually written to the current
directory. If this is not reasonable because of permission or quota problems,
then you can specify the directory for writing the temporary files into. This
is done by defining the TMPDIR
environment variable to be the path of
This option combines the effects of the
options. It displays the differences
between a file and the latest version of a module. If there are any
differences, it then proceeds to perform an update of the module with that
file, asking for remarks as usual. This option is very useful when used with
wildcarded module names. Then you can update just those modules which were
changed by an edit session, and see the changes for each module before typing
the appropriate remark for each module.
You may specify both of the -Difference
you may use this option. The results are identical.
This option conditionally updates a module.
That is, it will only do an update if there are any differences between a file
and the latest version of a module. This is convenient when related changes
are made to many modules in a directory, and one command using wildcards can
update just those modules that were changed.
option may be combined with the -Update
options to create the file if required.
This option is used to remove files which match the newest versions of
modules. If a file exists which matches the newest version of a module,
then the file is deleted, otherwise it is kept. This option is used to
clean up a work directory after building a new version of a product. This
option is especially useful when used with the -ALL option. It will
also accept the -Input option to specify a directory containing the
files to be cleaned.
This option is used to find out if a file does
not match the latest version of a module. If so, a message is given. If the
file does match, no output occurs. This option is thus useful to determine
which files have been modified and in need of updating. The -ALL
is defaulted for this option, since it is usually used for all modules. For
will report on all files which are different than the latest modules. If
is specified, then the output will consist of the module names
with no other output. This is useful for the backquote operator in shell
scripts for referencing the modules which are out of date. The -CHeck
option will also accept the -Input
- -PRune edit
This option is used to permanently remove
early edits from an edit history. This is useful if you wish to cut down on
the amount of disk space taken by an edit history file, or when you want to
start another release of a file, and want a copy of the edit history file for
that new release. The option takes an edit number to preserve, and all edits
in the edit history file before that edit are deleted, and can no longer be
referenced. For example, to keep only the current edit plus the previous 10
edits of the module file
, you could use the command:
Since the -PRune
option is unrecoverable (unless backup files are
available), the fhist
program asks the user to verify that the prune is
really wanted. The -Forced_Update
option can be used to bypass this
This option can be used with any of the action
options. It means perform the operation for all modules in the module storage
directory. Alternatively, you can specify multiple module names on the command
line, and the actions will be performed with those modules. You cannot specify
and module names.
When using multiple modules or the -ALL
option, the -Input
options have a slightly different meaning. In these cases, the
arguments are a directory name which contains
filenames with the same name as the module names. If the argument is not a
directory, then an error is given. This feature is useful for example, to
extract all the modules and place them into some remote directory, as in:
You should be careful when specifying numeric edit numbers for multiple modules.
Most probably, a particular edit number is not appropriate for multiple
modules, since changes corresponding to a particular edit number are not
usually related. Using named edits avoids these problems. As an example, if
you wanted to extract every module which had an edit that was named
, then you could use the command:
Some other useful examples of commands which use multiple modules are:
fhist *.c -create
fhist -check -all
fhist -cu -all
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:
- No output at all (except for errors).
- Single‐line output describing action (default).
- Detailed status as action proceeds.
Give some help on how to use the fhist program.
Show what version of fhist 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
Options and other command line arguments may be mixed arbitrarily on the command
The GNU long option names are understood. Since all option names for
are long, this means ignoring the extra leading '-'. The "
" convention is also understood.
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 the command
is actually equivilant to the command
fhist -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:
In general, fhist can handle all text files you throw at it, even international
text with unusual encodings. However, fhist is unable
to cope elegantly
with files which contain the NUL character.
(1) program simply prints a warning, and continues, you need to
know that it converts NUL characters into an 0x80 value before performing the
(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.
(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.
program will exit with a status of 1 on any error. The
program will only exit with a status of 0 if there are no errors.
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
A File Comparison Program, Webb Miller
and Eugene W. Myers, Software Practice and Experience, Volume 15, No. 11,
fhist 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
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/>.
|David I. Bell