Scattered all around your computer are thousands of text files. To a new user, this may seem inconsequential, but almost everything in Slackware Linux uses a plain-text file for configuration. This allows users to make changes to the system quickly, easily, and intuitively. In chapter 5, we looked at a few commands such as cat and less that can be used to read these files, but what if we want to make changes to them? For that, we need a text editor, and vi is up to the task. In short, vi is one of the oldest and most powerful text editors still used today. It's beloved by system administrators, programmers, hobbiests, and others the world over. In fact, nearly this entire book was written using vi; only the next chapter on emacs was written with that editor. A little further explanation is needed to learn exactly what vi is today though, as Slackware Linux technically doesn't include vi. Rather, Slackware includes two vi "clones", elvis(1) and vim(1). These clones add many additional features to vi such as syntax highlighting, binary editing modes, and network support. We won't go too deeply into all these details. By default, if you execute vi on Slackware Linux, you'll be using elvis, so all examples in this chapter will assume that is what you are using. If you've used another Linux distribution before, you may be more familiar with vim. If so, you might wish to change the symlink for /usr/bin/vi to point to /usr/bin/vim, or add an alias to your shell's startup scripts. vim is generally considered to be more feature-rich than elvis, but elvis is a much smaller program and contains more features than most users will ever need. vi is very powerful, but also somewhat cumbersome and challening for a new user to learn. However, mastering vi is an important skill for any self-respecting system administrator to learn, as vi is included on nearly every Linux distribution, every BSD system, and every UNIX system in existance. It's even included in Mac OS X. Once you've learned vi, you'll not have to learn another text editor to work on any of these systems. In fact, vi clones have even been ported to Microsoft Windows systems, so you can use it there too.

New users are often frustrated when using vi for the first time. When invoked without any arguments, vi will display a screen something like this. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Command At this point, the user will being typing and expect the keys he presses to appear in the document. Instead, something really strange happens. The reason for this is simple. vi has different operation "modes". There is a command mode and an insert mode. Command mode is the default; in this mode, each keystroke performs a particular action such as moving the cursor around, deleting text, yanking (copying) text, searching, etc.

Ok, so you've decided that you want to learn how to use vi. The first thing to do is learn how to open and save files. Opening files is actually pretty easy. Simply type the filename as an argument on the command-line and vi will happily load it for you. For example, vi chapter_11.xml will open the file chapter_11.xml and load its content onto the screen, simple enough. But what if we've finished with one document and wish to save it? We can do that in command mode using the :w command. When in command mode, pressing the : key temporarily positions the cursor on the very bottom line of the window and allows you to enter special commands. (This is technically known as ex-mode after the venerable ex application which we will not document here.) The command to save your current work is :w. Once this is done, vi will write your changes to the buffer back into the file. If you wish to open another document, simply use the :e other_document command and vi will happily open it for you. If you've made changes to the buffer but haven't saved it yet, :e will fail and print a warning message on the bottom line. You can bypass this with the :e! command. Most ex-mode commands in vi can be "forced" by adding ! to them. This tells vi that you want to abandon any changes you've made to the buffer and open the other document immediately. But what if I don't like my changes and want to quit or start over? That's easily done as well. Executing the :e! command without any arguments will re-open the current document from the beginning. Quitting vi is as simple as running the :q command if you haven't made any changes to the buffer, or :q! if you'd like to quit and abandon those changes.

Moving around in vi is perhaps the hardest thing for a new user to learn. vi does not traditionally use the directional arrow keys for cursor movement, although in Slackware Linux that is an option. Rather, movement is simply another command issued in command-mode. The reason for this is rather simple. vi actually predates the inclusion of directional arrow keys on keyboards. Thus, movement of the cursor had to be accomplished by using the few keys available, so the right-hand "home row" keys of h, j, k, and l were chosen. These keys will move the cursor about whenever vi is in command mode. Here's a short table to help you remember how they work. Command Result h Move the cursor one character left. j Move the cursor one line down k Move the cursor one line up l Move the cursor one character right Moving around is a little more powerful than that though. Like many command keys, these movement keys accept numerical arguments. For example, 10j will move the cursor down 10 lines. You can also move to the end or beginning of the current line with $ and ^, respectively.

Now that we're able to open and save documents, as well as move around in them, it's time to learn how to edit them. The primary means of editing is to enter insert mode using either the i or a command keys. These either insert text at the cursor's current location, or append it after the cursor's current location. Once into insert mode, you can type any text normally and it will be placed into your document. You can return to command mode in order to save your changes by pressing the ESC key.

Since vi can be difficult to learn, I've prepared a short cheat sheat that should help you with the basics until you begin to feel comfortable. Command Result h Move the cursor one character left. j Move the cursor one line down k Move the cursor one line up l Move the cursor one character right 10j Move the cursor ten lines down G Move to the end of the file ^ Move to the beginning of the line $ Move to the end of the line dd Remove a line (and store it in the copy buffer) 5dd Remove 5 lines (and store them in the copy buffer) dw Remove a single word (and store it in the copy buffer) 5dw Remove five words (and store them in the copy buffer) yw Yank (copy) a single word (and store it in the copy buffer) 5yw Yank five words (and store them in the copy buffer) p Paste the contents of the copy buffer at the cursor's location P Paste the contents of the copy buffer above the cursor's location r Replace a single character R Replace multiple characters x Delete a character X Delete the previous character u Undo the last action :s'old'new'g Replace all occurances of 'old' with 'new' /asdf Locate next occurance of asdf :q Quit (without saving) :w Save the current document :w file Save the current document as 'file' :x Save and quit