UNIX and Linux

This section introduces the two most popular UNIX editors: vi and emacs. For each editor, it covers:

• Basic text input and navigation.
• Moving and copying text.
• Searching for and replacing text.
• Other useful commands.
• A quick-reference chart.

vi (pronounced "vee-eye", short for visual, or perhaps vile) is a display-oriented text editor based on an underlying line editor called ex. Although beginners usually find vi somewhat awkward to use, it is useful to learn because it is universally available (being supplied with all UNIX systems). It also uses standard alphanumeric keys for commands, so it can be used on almost any terminal or workstation without having to worry about unusual keyboard mappings. System administrators like users to use vi because it uses very few system resources.

To start vi, enter:

$ vi filename

where filename is the name of the file you want to edit. If the file doesn't exist, vi will create it for you.

The main feature that makes vi unique as an editor is its mode-based operation. vi has two modes: command mode and input mode. In command mode, characters you type perform actions (e.g. moving the cursor, cutting or copying text, etc.) In input mode, characters you type are inserted or overwrite existing text.

When you begin vi, it is in command mode. To put vi into input mode, press i (insert). You can then type text which is inserted at the current cursor location; you can correct mistakes with the backspace key as you type.To get back into command mode, press ESC (the escape key). Another way of inserting text, especially useful when you are at the end of a line is to press a (append).

In command mode, you are able to move the cursor around your document. h, j, k and l move the cursor left, down, up and right respectively (if you are lucky the arrow keys may also work). Other useful keys are ^ and $ which move you to the beginning and end of a line respectively. w skips to the beginning of the next word and b skips back to the beginning of the previous word. To go right to the top of the document, press 1 and then G. To go the bottom of the document, press G. To skip forward a page, press ^F, and to go back a page, press ^B. To go to a particular line number, type the line number and press G, e.g. 55G takes you to line 55.

To delete text, move the cursor over the first character of the group you want to delete and make sure you are in command mode. Press x to delete the current character, dw to delete the next word, d4w to delete the next 4 words, dd to delete the next line, 4dd to delete the next 4 lines, d$ to delete to the end of the line or even dG to delete to the end of the document. If you accidentally delete too much, pressing u will undo the last change.

Occasionally you will want to join two lines together. Press J to do this (trying to press backspace on the beginning of the second line does not have the intuitive effect!)

vi uses buffers to store text that is deleted. There are nine numbered buffers (1-9) as well as the undo buffer. Usually buffer 1 contains the most recent deletion, buffer 2 the next recent, etc.

To cut and paste in vi, delete the text (using e.g. 5dd to delete 5 lines). Then move to the line where you want the text to appear and press p. If you delete something else before you paste, you can still retrieve the delete text by pasting the contents of the delete buffers. You can do this by typing 1p,2p, etc.

To copy and paste, "yank" the text (using e.g. 5yy to copy 5 lines). Then move to the line where you want the text to appear and press p.

In command mode, you can search for text by specifying regular expressions. To search forward, type / and then a regular expression and press enter. To search backwards, begin with a ? instead of a /. To find the next text that matches your regular expression press n.

To search and replace all occurences of pattern1 with pattern2, type :%s/pattern1/pattern2/g. To be asked to confirm each replacement, add a c to this substitution command. Instead of the % you can also give a range of lines (e.g. 1,17) over which you want the substitution to apply.

Programmers might like the :set number command which displays line numbers (:set nonumber turns them off).

To save a file, type :w. To save and quit, type :wq or press ZZ. To force a quit without saving type :q!

To start editing another file, type :e filename .

To execute shell commands from within vi, and then return to vi afterwards, type :!shellcommand. You can use the letter % as a substitute for the name of the file that you are editing (so :!echo % prints the name of the current file).

. repeats the last command.

emacs is a popular editor for UNIX, based operating systems. Unlike vi, it is not a standard UNIX system utility, but is available from the Free Software Foundation.

An emacs zealot will tell you how emacs provides advanced facilities that go beyond simple insertion and deletion of text: you can view two are more files at the same time, compile and debug programs in almost any programming language, typeset documents, run shell commands, read manual pages, email and news and even browse the web from inside emacs.

emacs is also very flexible - you can redefine keystrokes and commands easily, and (for the more ambitious) you can even write Lisp programs to add new commands and display modes to emacs, since emacs has its own Lisp interpreter. In fact most of the editing commands of Emacs are written in Lisp already; the few exceptions are written in C for efficiency. However, users do not need to know how to program Lisp to use emacs (while it is true that only a programmer can write a substantial extension to emacs, it is easy for anyone to use it afterwards).

Critics of emacs point out that it uses a relatively large amount of system resources compared to vi and that it has quite a complicated command structure (joking that emacs stands for Escape-Meta-Alt-Control-Shift).

In practice most users tend to use both editors - vi to quickly edit short scripts and programs and emacs for more complex jobs that require reference to more than one file simultaneously.

On UNIX systems, emacs can run in graphical mode under the X Windows system, in which case some helpful menus and mouse button command mappings are provided. However, most of its facilities are also available on a text terminal.

To start emacs, enter:

$ emacs filename

where filename is the name of the file you want to edit. If the file doesn't exist, emacs will create it for you.

Text input and navigation in emacs is mostly a matter of using the arrow keys to position the cursor and typing some text. Issuing more complex emacs commands usually involves pressing the Ctrl key (sometimes also labelled Control or Ctl) or the Meta key (sometimes also labelled Alt). emacs commands are usually described in this way:

• C-<chr> means hold the Ctrl key while typing the character <chr>. Thus, C-f would be: hold the Ctrl key and type f.
• M-<chr> means hold the Meta or Alt key down while typing <chr>. If there is no Meta or Alt key, instead press and release the ESC key and then type <chr>.

One initially annoying feature of emacs is that its help facility has been installed on C-h (Ctrl h). Unfortunately this is also the code for the backspace key on most systems. You can, however, easily make the key work as expected by creating a .emacs file (a file always read on start up) in your home directory containing the following line:

(global-set-key "\C-h" 'backward-delete-char-untabify)

Here is a .emacs file that contains this line as well as several other useful facilities (see Section 6.3.6).

To access the help system you can still type M-x help or (for a comprehensive tutorial) M-x help-with-tutorial

Useful navigation commands are C-a (beginning of line), C-e (end of line), C-v (forward page), M-v (backwards page), M-< (beginning of document) and M-> (end of document). C-d will delete the character under the cursor, while C-k will delete to the end of the line. Text deleted with C-k is placed in a buffer which can be "yanked" back later with C-y.

The easiest way to cut and paste text is to go to the start of the text and press C-k until you have deleted the text you want. Then move to the spot where you want to paste the text and press C-y to restore the text. If you make a mistake, C-u will undo the change (emacs supports several levels of undo).

Another way to delete a chunk of text is to go the start of the text and press C-SPC (SPC is the spacebar; this sets a mark). Then go to the end of the text you wish to delete and press C-w. Restore the text in the right spot with C-y.

To copy and paste, delete the target text as above, and then use C-y twice (once to restore the original text, and once to create the copy).

To search forwards and backwards incrementally, use C-s and C-r respectively. Pressing C-s or C-r again will repeat the operation. When you have found the text you want, press enter, or press C-g to cancel the operation and return your cursor to the position where the search started.

To replace a string, type M-x replace-string (you may want to modify your .emacs file so that this command is on C-x r). M-% performs a query search and replace.

Pressing C-g will cancel any emacs command in progress.

To save a file, press C-x C-s. To start editing a new file, press C-x C-f.

To bring up two windows (or "buffers" in emacs-speak), press C-x 2 (C-x 1 gets you back to 1). C-x o switches between buffers on the same screen. C-x b lets you switch between all active buffers, whether you can see them or not. C-x C-k deletes a buffer that you are finished with.

M-x shell brings up a UNIX shell inside a buffer (you may like to put this on C-x C-u). M-x goto-line skips to a particular line (you may like to put this on C-x g). M-x compile will attempt to compile a program (using the make utility or some other command that you can specify). If you are doing a lot of programming you will probably want to put this on a key like C-x c. M-q will reformat a paragraph. C-x C-c quits emacs.

There are many other editors for UNIX systems. Two popular alternatives to vi and emacs are nano and pico.