Learning GNU Emacs, 3rd Edition

Customizing Your Key Bindings

Perhaps the most common things that Emacs users want to customize are the keystrokes that cause commands to run. Keystrokes are associated with commands via key bindings.

Actually, every keystroke runs a command in Emacs. Printable character keys (letters, numerals, punctuation, and spaces) run the self-insert-command, which merely causes the key just pressed to be inserted at the cursor in the current buffer. (You could play a nasty April Fool's joke on a naïve Emacs user by changing the bindings of their printable characters.)

The default set of key bindings is adequate for most purposes, of course, but there are various cases in which you may want to add or change key bindings. Emacs contains literally hundreds of commands, only some of which have key bindings. As you know, you can access those that don't have bindings by typing M-x command-name Enter.

If, however, you intend to use an unbound command often, you may want to bind it to a keystroke sequence for convenience. You may want to set special keys, such as arrow, numeric keypad, or function keys, to perform commands you use often.

The other important concept you need to know now is that of a keymap, which is a collection of key bindings. The most basic default key bindings in Emacs are kept in a keymap called global-map. There is also the concept of a local keymap, which is specific to a single buffer. Local keymaps are used to implement commands in modes (like C mode, text mode, shell mode, ...