Using history and command line editing

September 28, 2001 —

One
of the great features of the C shell and the Korn shell is history.
The ability to recall and reuse commands that you have entered earlier
can save many keystrokes of repetitive typing. History and command line
editing are closely related. You may use command line editing to
edit the command that you are currently working on after you have typed
it in, or you may recall a previous command and then edit it.

In this article, I will show you some of the ins and outs of
history and command line editing in the Korn shell. For history to
work, you need to be in the Korn shell, so begin by typing:

ksh

to launch a Korn shell.

The second thing you need to make history work is an easy way to
access it. History is more than just the ability to recall
previous commands; it lets you recall commands and edit them before they are
executed. The Korn shell allows you to use different editing styles
depending on what you're trying to do. The primary
editors are vi, emacs, and gmacs. I recommend vi, as most people are
familiar with it. History and command line editing behave very
differently depending on whether you are using vi, emacs, or gmacs. I
will cover some of these differences.

To set vi up as your history editor type:


set -o vi

If you are going to continue to use history, you will want to include
this in your profile, or the emacs/gmacs equivalents. These set -o
commands only work in the Korn shell:


set -o emacs
or
set -o gmacs

Assuming that you have set -o vi as your history editor,
you may now access history. Press the escape key once (esc) and
release it and press k. The last command that you typed will
appear on the screen:


$ set -o vi

Press enter and that command will be executed. Try a few more
commands so that you can accumulate some history. End with ls -l on the current directory. Now press esc
and k. The ls -l command appears once more. Press k again and
the command just prior to ls -l will appear. Each time you press k
you'll go back one more command in your history. The minus key ( - )
works in exactly the same way. I use the minus key because it's easier for me
to remember. Eventually, if you press k or minus enough times, you'll
end up back at the set -o vi command.

Now press j or the plus key < + > and you will move forward in the command history.
(When you are rapidly running back through history looking for the command you need, and you realize you've passed it, it's good to know how to move forward again.)

If you have set -o emacs or gmacs, the keys are
control-p for previous command and control-n for next command. In vi
you hit esc once and then k or n will move you backward or
forward through the commands. For emacs and gmacs, use control-p or
control-n each time.

To view more of your history, type "history", and a list of your most
recent commands will be displayed along with numbers. If you see the number of a
command that you need, recall it by hitting the esc key, typing the number of the command as displayed by history, and then typing an upper case G. This will place the command with that number on the command line.

Once you have located the command that you want, you can press enter to
execute the command, but you can also edit the command before executing it.

Moving around in vi

There are several vi commands for moving around the command line and
editing the line. I will only cover a few here. For further information
consult the manual on ksh. If you are familiar with vi you will
recognize the commands. The vi editor operates in two main modes. In
command mode single keys move the cursor, delete characters and perform
other actions. In input mode single keys are actually typed into the
command line.

Command Editing -- Cursor Movement in command mode for vi


l (el)    Move the cursor forward one character
h         Move the cursor back one character
w         Move the cursor forward one word
b         Move the cursor back one word
fc        Find character c in the line
0 (zero)  Cursor to the start of the line
$         Cursor to the end of the line

l, h, w, f and b can be preceded by a number, thus 2b moves the cursor
back two words, and 3fx finds the third occurrence of x in the line. In
emacs and gmacs mode, cursor movement is different.

Command Editing -- Cursor Movement in emacs/gmacs


CONTROL-F        Move the cursor forward one character
CONTROL-B        Move the cursor back one character
ESCAPE then f    Move the cursor forward one word
ESCAPE then b    Move the cursor back one word
CONTROL-A        Cursor to the start of the line
CONTROL-E        Cursor to the end of the line

The editing of a command line is not limited to history. You can also
edit a command line that you are currently typing in. For example type
the following, but do not press enter:


ls -l

You should be sitting with the cursor at the end of the command
line. If you have set -o vi then hit the esc key to
exit input mode and enter command mode. Now you can type h a few
times and the cursor will move backward through the command line. If
you have set -o emacs, there is no need to press
esc, just start by typing control-b a few times, and the cursor
will move backwards.

Now we have the cursor where we want it, but how do you actually
edit the command? There are several editing options, and again I am
going to describe just some of the more useful ones.

Command Editing -- Editing keys in command mode for vi


x         Delete the character over the cursor. Can be preceded by a count
X         Delete the character behind the cursor. Can be preceded by a count
~ (tilde) Change case of the character over the cursor. Can be preceded by a
count
u         Undo the last change to the line
U         Undo all changes to the line
a         Enter input mode and begin inserting after the current character
A         Enter input mode and begin inserting at the end of the line
i         Enter input mode and begin inserting before the current character
I         Enter input mode and begin inserting at the beginning of the line

You enter input mode by pressing a, A, i or I. To exit to command mode, press the esc key.

You may also press enter at any time, and the command as it appears on
the command line will be executed.

In emacs and gmacs mode, you are always in input mode and normal characters
that are typed appear in the command line. Editing commands are given
as control + key or as esc + key ("key" is a variable).

Command Editing -- Editing keys for emacs/gmacs


CONTROL-D         Delete the character over the cursor
ESCAPE then d     Delete the current word
CONTROL-C~        Capitalize current character
ESCAPE then l     Convert current character to lower case

In emacs and gmacs mode, you may also press enter at any time, and the
command, as it appears on the command line, will be executed.

Remember that command line editing applies to the current command that
you are typing, or any other command that you have recovered from your
history.

Fun with file name completion

The last piece of command line editing that I want to cover is file name
completion. This is usually used in current command lines but of course it can also
be used when editing a command retrieved from history.

Assume a directory and file structure as follows and you are currently
in the abc directory.


(dir)  	               topdir
         		 |
            -------------------------------
            |            |                |
(dir)      abc          def              ghi
(files)  data.txt     john.ltr         news.doc
         old.txt      john.txt       weather.doc

Type the following but do not press enter:


ls -l d

Now hit esc and then the backslash ( \ ). This causes the command
line editor to attempt to locate a file (or directory) that can be used
to complete the name that has been started. The search finds data.txt
and that is typed in on the command line:


ls -l data.txt

The cursor is left after the txt of data.txt ready
for you to type anything else that is needed to complete your
command line.

This also works for directories. Assume, for example, that you are in the
abc directory and need to edit news.doc in the
ghi directory. You have the option to go to that directory and
begin editing, or simply issue a vi command to edit directly from where you are:


cd ../ghi
vi news.doc


vi ../ghi/news.doc

If you try the following sequence you will see the ( \ ) at work
completing the names for you. First type enough of the command to
identify the directory:


vi ../g

Now type esc followed by ( \ ) and the directory name is completed
for you:


vi ../ghi/

Now type enough to uniquely identify the file by adding an n:


vi ../ghi/n

Press esc followed by ( \ ) and the file name is completed for you:


vi ../ghi/news.doc

This feature is especially useful for long directory and file names and will save you
much typing.

If two or more files match what you have typed so far, the file
completion option will complete as much of the file as it can and then
wait for you to provide more unique input. To edit the file john.txt
from one of the adjacent directories type:


vi ../def/j

Press esc and backslash and it will fill in as much as it can and
then wait for you to indicate whether you want john.txt or
john.ltr:


vi ../def/john.

The sequence that does this for emacs and gmacs is esc followed by a
second esc.

There is much more to command line editing and history, but I have
found that just these few tips greatly speed up my work, and I hope
they do the same for you.

If you like the idea of history, ask your system administrator to
change your shell to a Korn shell, and then be sure to add set -o vi, emacs or gmacs into your
.profile file.

Unix Insider