2 @c This is part of the GNU Emacs Lisp Reference Manual.
3 @c Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1998 Free Software Foundation, Inc.
4 @c See the file elisp.texi for copying conditions.
5 @setfilename ../info/commands
6 @node Command Loop, Keymaps, Minibuffers, Top
8 @cindex editor command loop
11 When you run Emacs, it enters the @dfn{editor command loop} almost
12 immediately. This loop reads key sequences, executes their definitions,
13 and displays the results. In this chapter, we describe how these things
14 are done, and the subroutines that allow Lisp programs to do them.
17 * Command Overview:: How the command loop reads commands.
18 * Defining Commands:: Specifying how a function should read arguments.
19 * Interactive Call:: Calling a command, so that it will read arguments.
20 * Command Loop Info:: Variables set by the command loop for you to examine.
21 * Input Events:: What input looks like when you read it.
22 * Reading Input:: How to read input events from the keyboard or mouse.
23 * Special Events:: Events processed immediately and individually.
24 * Waiting:: Waiting for user input or elapsed time.
25 * Quitting:: How @kbd{C-g} works. How to catch or defer quitting.
26 * Prefix Command Arguments:: How the commands to set prefix args work.
27 * Recursive Editing:: Entering a recursive edit,
28 and why you usually shouldn't.
29 * Disabling Commands:: How the command loop handles disabled commands.
30 * Command History:: How the command history is set up, and how accessed.
31 * Keyboard Macros:: How keyboard macros are implemented.
34 @node Command Overview
35 @section Command Loop Overview
37 The first thing the command loop must do is read a key sequence, which
38 is a sequence of events that translates into a command. It does this by
39 calling the function @code{read-key-sequence}. Your Lisp code can also
40 call this function (@pxref{Key Sequence Input}). Lisp programs can also
41 do input at a lower level with @code{read-event} (@pxref{Reading One
42 Event}) or discard pending input with @code{discard-input}
43 (@pxref{Event Input Misc}).
45 The key sequence is translated into a command through the currently
46 active keymaps. @xref{Key Lookup}, for information on how this is done.
47 The result should be a keyboard macro or an interactively callable
48 function. If the key is @kbd{M-x}, then it reads the name of another
49 command, which it then calls. This is done by the command
50 @code{execute-extended-command} (@pxref{Interactive Call}).
52 To execute a command requires first reading the arguments for it.
53 This is done by calling @code{command-execute} (@pxref{Interactive
54 Call}). For commands written in Lisp, the @code{interactive}
55 specification says how to read the arguments. This may use the prefix
56 argument (@pxref{Prefix Command Arguments}) or may read with prompting
57 in the minibuffer (@pxref{Minibuffers}). For example, the command
58 @code{find-file} has an @code{interactive} specification which says to
59 read a file name using the minibuffer. The command's function body does
60 not use the minibuffer; if you call this command from Lisp code as a
61 function, you must supply the file name string as an ordinary Lisp
64 If the command is a string or vector (i.e., a keyboard macro) then
65 @code{execute-kbd-macro} is used to execute it. You can call this
66 function yourself (@pxref{Keyboard Macros}).
68 To terminate the execution of a running command, type @kbd{C-g}. This
69 character causes @dfn{quitting} (@pxref{Quitting}).
71 @defvar pre-command-hook
72 The editor command loop runs this normal hook before each command. At
73 that time, @code{this-command} contains the command that is about to
74 run, and @code{last-command} describes the previous command.
78 @defvar post-command-hook
79 The editor command loop runs this normal hook after each command
80 (including commands terminated prematurely by quitting or by errors),
81 and also when the command loop is first entered. At that time,
82 @code{this-command} describes the command that just ran, and
83 @code{last-command} describes the command before that. @xref{Hooks}.
86 Quitting is suppressed while running @code{pre-command-hook} and
87 @code{post-command-hook}. If an error happens while executing one of
88 these hooks, it terminates execution of the hook, and clears the hook
89 variable to @code{nil} so as to prevent an infinite loop of errors.
91 @node Defining Commands
92 @section Defining Commands
93 @cindex defining commands
94 @cindex commands, defining
95 @cindex functions, making them interactive
96 @cindex interactive function
98 A Lisp function becomes a command when its body contains, at top
99 level, a form that calls the special form @code{interactive}. This
100 form does nothing when actually executed, but its presence serves as a
101 flag to indicate that interactive calling is permitted. Its argument
102 controls the reading of arguments for an interactive call.
105 * Using Interactive:: General rules for @code{interactive}.
106 * Interactive Codes:: The standard letter-codes for reading arguments
108 * Interactive Examples:: Examples of how to read interactive arguments.
111 @node Using Interactive
112 @subsection Using @code{interactive}
114 This section describes how to write the @code{interactive} form that
115 makes a Lisp function an interactively-callable command.
117 @defspec interactive arg-descriptor
118 @cindex argument descriptors
119 This special form declares that the function in which it appears is a
120 command, and that it may therefore be called interactively (via
121 @kbd{M-x} or by entering a key sequence bound to it). The argument
122 @var{arg-descriptor} declares how to compute the arguments to the
123 command when the command is called interactively.
125 A command may be called from Lisp programs like any other function, but
126 then the caller supplies the arguments and @var{arg-descriptor} has no
129 The @code{interactive} form has its effect because the command loop
130 (actually, its subroutine @code{call-interactively}) scans through the
131 function definition looking for it, before calling the function. Once
132 the function is called, all its body forms including the
133 @code{interactive} form are executed, but at this time
134 @code{interactive} simply returns @code{nil} without even evaluating its
138 There are three possibilities for the argument @var{arg-descriptor}:
142 It may be omitted or @code{nil}; then the command is called with no
143 arguments. This leads quickly to an error if the command requires one
147 It may be a Lisp expression that is not a string; then it should be a
148 form that is evaluated to get a list of arguments to pass to the
150 @cindex argument evaluation form
152 If this expression reads keyboard input (this includes using the
153 minibuffer), keep in mind that the integer value of point or the mark
154 before reading input may be incorrect after reading input. This is
155 because the current buffer may be receiving subprocess output;
156 if subprocess output arrives while the command is waiting for input,
157 it could relocate point and the mark.
159 Here's an example of what @emph{not} to do:
163 (list (region-beginning) (region-end)
164 (read-string "Foo: " nil 'my-history)))
168 Here's how to avoid the problem, by examining point and the mark only
169 after reading the keyboard input:
173 (let ((string (read-string "Foo: " nil 'my-history)))
174 (list (region-beginning) (region-end) string)))
178 @cindex argument prompt
179 It may be a string; then its contents should consist of a code character
180 followed by a prompt (which some code characters use and some ignore).
181 The prompt ends either with the end of the string or with a newline.
182 Here is a simple example:
185 (interactive "bFrobnicate buffer: ")
189 The code letter @samp{b} says to read the name of an existing buffer,
190 with completion. The buffer name is the sole argument passed to the
191 command. The rest of the string is a prompt.
193 If there is a newline character in the string, it terminates the prompt.
194 If the string does not end there, then the rest of the string should
195 contain another code character and prompt, specifying another argument.
196 You can specify any number of arguments in this way.
199 The prompt string can use @samp{%} to include previous argument values
200 (starting with the first argument) in the prompt. This is done using
201 @code{format} (@pxref{Formatting Strings}). For example, here is how
202 you could read the name of an existing buffer followed by a new name to
207 (interactive "bBuffer to rename: \nsRename buffer %s to: ")
211 @cindex @samp{*} in interactive
212 @cindex read-only buffers in interactive
213 If the first character in the string is @samp{*}, then an error is
214 signaled if the buffer is read-only.
216 @cindex @samp{@@} in interactive
218 If the first character in the string is @samp{@@}, and if the key
219 sequence used to invoke the command includes any mouse events, then
220 the window associated with the first of those events is selected
221 before the command is run.
223 You can use @samp{*} and @samp{@@} together; the order does not matter.
224 Actual reading of arguments is controlled by the rest of the prompt
225 string (starting with the first character that is not @samp{*} or
229 @node Interactive Codes
230 @comment node-name, next, previous, up
231 @subsection Code Characters for @code{interactive}
232 @cindex interactive code description
233 @cindex description for interactive codes
234 @cindex codes, interactive, description of
235 @cindex characters for interactive codes
237 The code character descriptions below contain a number of key words,
238 defined here as follows:
242 @cindex interactive completion
243 Provide completion. @key{TAB}, @key{SPC}, and @key{RET} perform name
244 completion because the argument is read using @code{completing-read}
245 (@pxref{Completion}). @kbd{?} displays a list of possible completions.
248 Require the name of an existing object. An invalid name is not
249 accepted; the commands to exit the minibuffer do not exit if the current
253 @cindex default argument string
254 A default value of some sort is used if the user enters no text in the
255 minibuffer. The default depends on the code character.
258 This code letter computes an argument without reading any input.
259 Therefore, it does not use a prompt string, and any prompt string you
262 Even though the code letter doesn't use a prompt string, you must follow
263 it with a newline if it is not the last code character in the string.
266 A prompt immediately follows the code character. The prompt ends either
267 with the end of the string or with a newline.
270 This code character is meaningful only at the beginning of the
271 interactive string, and it does not look for a prompt or a newline.
272 It is a single, isolated character.
275 @cindex reading interactive arguments
276 Here are the code character descriptions for use with @code{interactive}:
280 Signal an error if the current buffer is read-only. Special.
283 Select the window mentioned in the first mouse event in the key
284 sequence that invoked this command. Special.
287 A function name (i.e., a symbol satisfying @code{fboundp}). Existing,
291 The name of an existing buffer. By default, uses the name of the
292 current buffer (@pxref{Buffers}). Existing, Completion, Default,
296 A buffer name. The buffer need not exist. By default, uses the name of
297 a recently used buffer other than the current buffer. Completion,
301 A character. The cursor does not move into the echo area. Prompt.
304 A command name (i.e., a symbol satisfying @code{commandp}). Existing,
308 @cindex position argument
309 The position of point, as an integer (@pxref{Point}). No I/O.
312 A directory name. The default is the current default directory of the
313 current buffer, @code{default-directory} (@pxref{System Environment}).
314 Existing, Completion, Default, Prompt.
317 The first or next mouse event in the key sequence that invoked the command.
318 More precisely, @samp{e} gets events that are lists, so you can look at
319 the data in the lists. @xref{Input Events}. No I/O.
321 You can use @samp{e} more than once in a single command's interactive
322 specification. If the key sequence that invoked the command has
323 @var{n} events that are lists, the @var{n}th @samp{e} provides the
324 @var{n}th such event. Events that are not lists, such as function keys
325 and @sc{ASCII} characters, do not count where @samp{e} is concerned.
328 A file name of an existing file (@pxref{File Names}). The default
329 directory is @code{default-directory}. Existing, Completion, Default,
333 A file name. The file need not exist. Completion, Default, Prompt.
336 A key sequence (@pxref{Keymap Terminology}). This keeps reading events
337 until a command (or undefined command) is found in the current key
338 maps. The key sequence argument is represented as a string or vector.
339 The cursor does not move into the echo area. Prompt.
341 This kind of input is used by commands such as @code{describe-key} and
342 @code{global-set-key}.
345 A key sequence, whose definition you intend to change. This works like
346 @samp{k}, except that it suppresses, for the last input event in the key
347 sequence, the conversions that are normally used (when necessary) to
348 convert an undefined key into a defined one.
351 @cindex marker argument
352 The position of the mark, as an integer. No I/O.
355 Arbitrary text, read in the minibuffer using the current buffer's input
356 method, and returned as a string (@pxref{Input Methods,,, emacs, The GNU
357 Emacs Manual}). Prompt.
360 A number read with the minibuffer. If the input is not a number, the
361 user is asked to try again. The prefix argument, if any, is not used.
365 @cindex raw prefix argument usage
366 The numeric prefix argument; but if there is no prefix argument, read a
367 number as with @kbd{n}. Requires a number. @xref{Prefix Command
371 @cindex numeric prefix argument usage
372 The numeric prefix argument. (Note that this @samp{p} is lower case.)
376 The raw prefix argument. (Note that this @samp{P} is upper case.) No
380 @cindex region argument
381 Point and the mark, as two numeric arguments, smallest first. This is
382 the only code letter that specifies two successive arguments rather than
386 Arbitrary text, read in the minibuffer and returned as a string
387 (@pxref{Text from Minibuffer}). Terminate the input with either
388 @kbd{C-j} or @key{RET}. (@kbd{C-q} may be used to include either of
389 these characters in the input.) Prompt.
392 An interned symbol whose name is read in the minibuffer. Any whitespace
393 character terminates the input. (Use @kbd{C-q} to include whitespace in
394 the string.) Other characters that normally terminate a symbol (e.g.,
395 parentheses and brackets) do not do so here. Prompt.
398 A variable declared to be a user option (i.e., satisfying the predicate
399 @code{user-variable-p}). @xref{High-Level Completion}. Existing,
403 A Lisp object, specified with its read syntax, terminated with a
404 @kbd{C-j} or @key{RET}. The object is not evaluated. @xref{Object from
408 @cindex evaluated expression argument
409 A Lisp form is read as with @kbd{x}, but then evaluated so that its
410 value becomes the argument for the command. Prompt.
413 @node Interactive Examples
414 @comment node-name, next, previous, up
415 @subsection Examples of Using @code{interactive}
416 @cindex examples of using @code{interactive}
417 @cindex @code{interactive}, examples of using
419 Here are some examples of @code{interactive}:
423 (defun foo1 () ; @r{@code{foo1} takes no arguments,}
424 (interactive) ; @r{just moves forward two words.}
430 (defun foo2 (n) ; @r{@code{foo2} takes one argument,}
431 (interactive "p") ; @r{which is the numeric prefix.}
432 (forward-word (* 2 n)))
437 (defun foo3 (n) ; @r{@code{foo3} takes one argument,}
438 (interactive "nCount:") ; @r{which is read with the Minibuffer.}
439 (forward-word (* 2 n)))
444 (defun three-b (b1 b2 b3)
445 "Select three existing buffers.
446 Put them into three windows, selecting the last one."
448 (interactive "bBuffer1:\nbBuffer2:\nbBuffer3:")
449 (delete-other-windows)
450 (split-window (selected-window) 8)
451 (switch-to-buffer b1)
453 (split-window (selected-window) 8)
454 (switch-to-buffer b2)
456 (switch-to-buffer b3))
459 (three-b "*scratch*" "declarations.texi" "*mail*")
464 @node Interactive Call
465 @section Interactive Call
466 @cindex interactive call
468 After the command loop has translated a key sequence into a command it
469 invokes that command using the function @code{command-execute}. If the
470 command is a function, @code{command-execute} calls
471 @code{call-interactively}, which reads the arguments and calls the
472 command. You can also call these functions yourself.
474 @defun commandp object
475 Returns @code{t} if @var{object} is suitable for calling interactively;
476 that is, if @var{object} is a command. Otherwise, returns @code{nil}.
478 The interactively callable objects include strings and vectors (treated
479 as keyboard macros), lambda expressions that contain a top-level call to
480 @code{interactive}, byte-code function objects made from such lambda
481 expressions, autoload objects that are declared as interactive
482 (non-@code{nil} fourth argument to @code{autoload}), and some of the
485 A symbol satisfies @code{commandp} if its function definition satisfies
488 Keys and keymaps are not commands. Rather, they are used to look up
489 commands (@pxref{Keymaps}).
491 See @code{documentation} in @ref{Accessing Documentation}, for a
492 realistic example of using @code{commandp}.
495 @defun call-interactively command &optional record-flag keys
496 This function calls the interactively callable function @var{command},
497 reading arguments according to its interactive calling specifications.
498 An error is signaled if @var{command} is not a function or if it cannot
499 be called interactively (i.e., is not a command). Note that keyboard
500 macros (strings and vectors) are not accepted, even though they are
501 considered commands, because they are not functions.
503 @cindex record command history
504 If @var{record-flag} is non-@code{nil}, then this command and its
505 arguments are unconditionally added to the list @code{command-history}.
506 Otherwise, the command is added only if it uses the minibuffer to read
507 an argument. @xref{Command History}.
509 The argument @var{keys}, if given, specifies the sequence of events to
510 supply if the command inquires which events were used to invoke it.
513 @defun command-execute command &optional record-flag keys
514 @cindex keyboard macro execution
515 This function executes @var{command}. The argument @var{command} must
516 satisfy the @code{commandp} predicate; i.e., it must be an interactively
517 callable function or a keyboard macro.
519 A string or vector as @var{command} is executed with
520 @code{execute-kbd-macro}. A function is passed to
521 @code{call-interactively}, along with the optional @var{record-flag}.
523 A symbol is handled by using its function definition in its place. A
524 symbol with an @code{autoload} definition counts as a command if it was
525 declared to stand for an interactively callable function. Such a
526 definition is handled by loading the specified library and then
527 rechecking the definition of the symbol.
529 The argument @var{keys}, if given, specifies the sequence of events to
530 supply if the command inquires which events were used to invoke it.
533 @deffn Command execute-extended-command prefix-argument
534 @cindex read command name
535 This function reads a command name from the minibuffer using
536 @code{completing-read} (@pxref{Completion}). Then it uses
537 @code{command-execute} to call the specified command. Whatever that
538 command returns becomes the value of @code{execute-extended-command}.
540 @cindex execute with prefix argument
541 If the command asks for a prefix argument, it receives the value
542 @var{prefix-argument}. If @code{execute-extended-command} is called
543 interactively, the current raw prefix argument is used for
544 @var{prefix-argument}, and thus passed on to whatever command is run.
546 @c !!! Should this be @kindex?
548 @code{execute-extended-command} is the normal definition of @kbd{M-x},
549 so it uses the string @w{@samp{M-x }} as a prompt. (It would be better
550 to take the prompt from the events used to invoke
551 @code{execute-extended-command}, but that is painful to implement.) A
552 description of the value of the prefix argument, if any, also becomes
557 (execute-extended-command 1)
558 ---------- Buffer: Minibuffer ----------
559 1 M-x forward-word RET
560 ---------- Buffer: Minibuffer ----------
567 This function returns @code{t} if the containing function (the one whose
568 code includes the call to @code{interactive-p}) was called
569 interactively, with the function @code{call-interactively}. (It makes
570 no difference whether @code{call-interactively} was called from Lisp or
571 directly from the editor command loop.) If the containing function was
572 called by Lisp evaluation (or with @code{apply} or @code{funcall}), then
573 it was not called interactively.
576 The most common use of @code{interactive-p} is for deciding whether to
577 print an informative message. As a special exception,
578 @code{interactive-p} returns @code{nil} whenever a keyboard macro is
579 being run. This is to suppress the informative messages and speed
580 execution of the macro.
588 (when (interactive-p)
596 (setq foobar (list (foo) (interactive-p))))
601 ;; @r{Type @kbd{M-x foo}.}
606 ;; @r{Type @kbd{M-x bar}.}
607 ;; @r{This does not print anything.}
616 The other way to do this sort of job is to make the command take an
617 argument @code{print-message} which should be non-@code{nil} in an
618 interactive call, and use the @code{interactive} spec to make sure it is
619 non-@code{nil}. Here's how:
622 (defun foo (&optional print-message)
628 The numeric prefix argument, provided by @samp{p}, is never @code{nil}.
630 @node Command Loop Info
631 @comment node-name, next, previous, up
632 @section Information from the Command Loop
634 The editor command loop sets several Lisp variables to keep status
635 records for itself and for commands that are run.
638 This variable records the name of the previous command executed by the
639 command loop (the one before the current command). Normally the value
640 is a symbol with a function definition, but this is not guaranteed.
642 The value is copied from @code{this-command} when a command returns to
643 the command loop, except when the command has specified a prefix
644 argument for the following command.
646 This variable is always local to the current terminal and cannot be
647 buffer-local. @xref{Multiple Displays}.
651 @cindex current command
652 This variable records the name of the command now being executed by
653 the editor command loop. Like @code{last-command}, it is normally a symbol
654 with a function definition.
656 The command loop sets this variable just before running a command, and
657 copies its value into @code{last-command} when the command finishes
658 (unless the command specified a prefix argument for the following
661 @cindex kill command repetition
662 Some commands set this variable during their execution, as a flag for
663 whatever command runs next. In particular, the functions for killing text
664 set @code{this-command} to @code{kill-region} so that any kill commands
665 immediately following will know to append the killed text to the
669 If you do not want a particular command to be recognized as the previous
670 command in the case where it got an error, you must code that command to
671 prevent this. One way is to set @code{this-command} to @code{t} at the
672 beginning of the command, and set @code{this-command} back to its proper
673 value at the end, like this:
676 (defun foo (args@dots{})
677 (interactive @dots{})
678 (let ((old-this-command this-command))
679 (setq this-command t)
680 @r{@dots{}do the work@dots{}}
681 (setq this-command old-this-command)))
685 We do not bind @code{this-command} with @code{let} because that would
686 restore the old value in case of error---a feature of @code{let} which
687 in this case does precisely what we want to avoid.
689 @defun this-command-keys
690 This function returns a string or vector containing the key sequence
691 that invoked the present command, plus any previous commands that
692 generated the prefix argument for this command. The value is a string
693 if all those events were characters. @xref{Input Events}.
698 ;; @r{Now use @kbd{C-u C-x C-e} to evaluate that.}
704 @defun this-command-keys-vector
705 Like @code{this-command-keys}, except that it always returns
706 the events in a vector, so you do never need to deal with the complexities
707 of storing input events in a string (@pxref{Strings of Events}).
710 @defvar last-nonmenu-event
711 This variable holds the last input event read as part of a key sequence,
712 not counting events resulting from mouse menus.
714 One use of this variable is for telling @code{x-popup-menu} where to pop
715 up a menu. It is also used internally by @code{y-or-n-p}
716 (@pxref{Yes-or-No Queries}).
719 @defvar last-command-event
720 @defvarx last-command-char
721 This variable is set to the last input event that was read by the
722 command loop as part of a command. The principal use of this variable
723 is in @code{self-insert-command}, which uses it to decide which
729 ;; @r{Now use @kbd{C-u C-x C-e} to evaluate that.}
735 The value is 5 because that is the @sc{ASCII} code for @kbd{C-e}.
737 The alias @code{last-command-char} exists for compatibility with
742 @defvar last-event-frame
743 This variable records which frame the last input event was directed to.
744 Usually this is the frame that was selected when the event was
745 generated, but if that frame has redirected input focus to another
746 frame, the value is the frame to which the event was redirected.
751 @section Input Events
755 The Emacs command loop reads a sequence of @dfn{input events} that
756 represent keyboard or mouse activity. The events for keyboard activity
757 are characters or symbols; mouse events are always lists. This section
758 describes the representation and meaning of input events in detail.
761 This function returns non-@code{nil} if @var{object} is an input event.
766 * Keyboard Events:: Ordinary characters--keys with symbols on them.
767 * Function Keys:: Function keys--keys with names, not symbols.
768 * Mouse Events:: Overview of mouse events.
769 * Click Events:: Pushing and releasing a mouse button.
770 * Drag Events:: Moving the mouse before releasing the button.
771 * Button-Down Events:: A button was pushed and not yet released.
772 * Repeat Events:: Double and triple click (or drag, or down).
773 * Motion Events:: Just moving the mouse, not pushing a button.
774 * Focus Events:: Moving the mouse between frames.
775 * Misc Events:: Other events window systems can generate.
776 * Event Examples:: Examples of the lists for mouse events.
777 * Classifying Events:: Finding the modifier keys in an event symbol.
779 * Accessing Events:: Functions to extract info from events.
780 * Strings of Events:: Special considerations for putting
781 keyboard character events in a string.
784 @node Keyboard Events
785 @subsection Keyboard Events
787 There are two kinds of input you can get from the keyboard: ordinary
788 keys, and function keys. Ordinary keys correspond to characters; the
789 events they generate are represented in Lisp as characters. The event
790 type of a character event is the character itself (an integer); see
791 @ref{Classifying Events}.
793 @cindex modifier bits (of input character)
794 @cindex basic code (of input character)
795 An input character event consists of a @dfn{basic code} between 0 and
796 524287, plus any or all of these @dfn{modifier bits}:
807 bit in the character code indicates a character
808 typed with the meta key held down.
818 bit in the character code indicates a non-@sc{ASCII}
821 @sc{ASCII} control characters such as @kbd{C-a} have special basic
822 codes of their own, so Emacs needs no special bit to indicate them.
823 Thus, the code for @kbd{C-a} is just 1.
825 But if you type a control combination not in @sc{ASCII}, such as
826 @kbd{%} with the control key, the numeric value you get is the code
834 (assuming the terminal supports non-@sc{ASCII}
845 bit in the character code indicates an @sc{ASCII} control
846 character typed with the shift key held down.
848 For letters, the basic code itself indicates upper versus lower case;
849 for digits and punctuation, the shift key selects an entirely different
850 character with a different basic code. In order to keep within the
851 @sc{ASCII} character set whenever possible, Emacs avoids using the
858 bit for those characters.
860 However, @sc{ASCII} provides no way to distinguish @kbd{C-A} from
861 @kbd{C-a}, so Emacs uses the
868 bit in @kbd{C-A} and not in
879 bit in the character code indicates a character
880 typed with the hyper key held down.
890 bit in the character code indicates a character
891 typed with the super key held down.
901 bit in the character code indicates a character typed with
902 the alt key held down. (On some terminals, the key labeled @key{ALT}
903 is actually the meta key.)
906 It is best to avoid mentioning specific bit numbers in your program.
907 To test the modifier bits of a character, use the function
908 @code{event-modifiers} (@pxref{Classifying Events}). When making key
909 bindings, you can use the read syntax for characters with modifier bits
910 (@samp{\C-}, @samp{\M-}, and so on). For making key bindings with
911 @code{define-key}, you can use lists such as @code{(control hyper ?x)} to
912 specify the characters (@pxref{Changing Key Bindings}). The function
913 @code{event-convert-list} converts such a list into an event type
914 (@pxref{Classifying Events}).
917 @subsection Function Keys
919 @cindex function keys
920 Most keyboards also have @dfn{function keys}---keys that have names or
921 symbols that are not characters. Function keys are represented in Emacs
922 Lisp as symbols; the symbol's name is the function key's label, in lower
923 case. For example, pressing a key labeled @key{F1} places the symbol
924 @code{f1} in the input stream.
926 The event type of a function key event is the event symbol itself.
927 @xref{Classifying Events}.
929 Here are a few special cases in the symbol-naming convention for
933 @item @code{backspace}, @code{tab}, @code{newline}, @code{return}, @code{delete}
934 These keys correspond to common @sc{ASCII} control characters that have
935 special keys on most keyboards.
937 In @sc{ASCII}, @kbd{C-i} and @key{TAB} are the same character. If the
938 terminal can distinguish between them, Emacs conveys the distinction to
939 Lisp programs by representing the former as the integer 9, and the
940 latter as the symbol @code{tab}.
942 Most of the time, it's not useful to distinguish the two. So normally
943 @code{function-key-map} (@pxref{Translating Input}) is set up to map
944 @code{tab} into 9. Thus, a key binding for character code 9 (the
945 character @kbd{C-i}) also applies to @code{tab}. Likewise for the other
946 symbols in this group. The function @code{read-char} likewise converts
947 these events into characters.
949 In @sc{ASCII}, @key{BS} is really @kbd{C-h}. But @code{backspace}
950 converts into the character code 127 (@key{DEL}), not into code 8
951 (@key{BS}). This is what most users prefer.
953 @item @code{left}, @code{up}, @code{right}, @code{down}
955 @item @code{kp-add}, @code{kp-decimal}, @code{kp-divide}, @dots{}
956 Keypad keys (to the right of the regular keyboard).
957 @item @code{kp-0}, @code{kp-1}, @dots{}
958 Keypad keys with digits.
959 @item @code{kp-f1}, @code{kp-f2}, @code{kp-f3}, @code{kp-f4}
961 @item @code{kp-home}, @code{kp-left}, @code{kp-up}, @code{kp-right}, @code{kp-down}
962 Keypad arrow keys. Emacs normally translates these into the
963 corresponding non-keypad keys @code{home}, @code{left}, @dots{}
964 @item @code{kp-prior}, @code{kp-next}, @code{kp-end}, @code{kp-begin}, @code{kp-insert}, @code{kp-delete}
965 Additional keypad duplicates of keys ordinarily found elsewhere. Emacs
966 normally translates these into the like-named non-keypad keys.
969 You can use the modifier keys @key{ALT}, @key{CTRL}, @key{HYPER},
970 @key{META}, @key{SHIFT}, and @key{SUPER} with function keys. The way to
971 represent them is with prefixes in the symbol name:
977 The control modifier.
988 Thus, the symbol for the key @key{F3} with @key{META} held down is
989 @code{M-f3}. When you use more than one prefix, we recommend you
990 write them in alphabetical order; but the order does not matter in
991 arguments to the key-binding lookup and modification functions.
994 @subsection Mouse Events
996 Emacs supports four kinds of mouse events: click events, drag events,
997 button-down events, and motion events. All mouse events are represented
998 as lists. The @sc{car} of the list is the event type; this says which
999 mouse button was involved, and which modifier keys were used with it.
1000 The event type can also distinguish double or triple button presses
1001 (@pxref{Repeat Events}). The rest of the list elements give position
1002 and time information.
1004 For key lookup, only the event type matters: two events of the same type
1005 necessarily run the same command. The command can access the full
1006 values of these events using the @samp{e} interactive code.
1007 @xref{Interactive Codes}.
1009 A key sequence that starts with a mouse event is read using the keymaps
1010 of the buffer in the window that the mouse was in, not the current
1011 buffer. This does not imply that clicking in a window selects that
1012 window or its buffer---that is entirely under the control of the command
1013 binding of the key sequence.
1016 @subsection Click Events
1018 @cindex mouse click event
1020 When the user presses a mouse button and releases it at the same
1021 location, that generates a @dfn{click} event. Mouse click events have
1026 (@var{window} @var{buffer-pos} (@var{x} . @var{y}) @var{timestamp})
1030 Here is what the elements normally mean:
1033 @item @var{event-type}
1034 This is a symbol that indicates which mouse button was used. It is
1035 one of the symbols @code{mouse-1}, @code{mouse-2}, @dots{}, where the
1036 buttons are numbered left to right.
1038 You can also use prefixes @samp{A-}, @samp{C-}, @samp{H-}, @samp{M-},
1039 @samp{S-} and @samp{s-} for modifiers alt, control, hyper, meta, shift
1040 and super, just as you would with function keys.
1042 This symbol also serves as the event type of the event. Key bindings
1043 describe events by their types; thus, if there is a key binding for
1044 @code{mouse-1}, that binding would apply to all events whose
1045 @var{event-type} is @code{mouse-1}.
1048 This is the window in which the click occurred.
1050 @item @var{x}, @var{y}
1051 These are the pixel-denominated coordinates of the click, relative to
1052 the top left corner of @var{window}, which is @code{(0 . 0)}.
1054 @item @var{buffer-pos}
1055 This is the buffer position of the character clicked on.
1057 @item @var{timestamp}
1058 This is the time at which the event occurred, in milliseconds. (Since
1059 this value wraps around the entire range of Emacs Lisp integers in about
1060 five hours, it is useful only for relating the times of nearby events.)
1062 @item @var{click-count}
1063 This is the number of rapid repeated presses so far of the same mouse
1064 button. @xref{Repeat Events}.
1067 The meanings of @var{buffer-pos}, @var{x} and @var{y} are somewhat
1068 different when the event location is in a special part of the screen,
1069 such as the mode line or a scroll bar.
1071 If the location is in a scroll bar, then @var{buffer-pos} is the symbol
1072 @code{vertical-scroll-bar} or @code{horizontal-scroll-bar}, and the pair
1073 @code{(@var{x} . @var{y})} is replaced with a pair @code{(@var{portion}
1074 . @var{whole})}, where @var{portion} is the distance of the click from
1075 the top or left end of the scroll bar, and @var{whole} is the length of
1076 the entire scroll bar.
1078 If the position is on a mode line or the vertical line separating
1079 @var{window} from its neighbor to the right, then @var{buffer-pos} is
1080 the symbol @code{mode-line} or @code{vertical-line}. For the mode line,
1081 @var{y} does not have meaningful data. For the vertical line, @var{x}
1082 does not have meaningful data.
1084 In one special case, @var{buffer-pos} is a list containing a symbol (one
1085 of the symbols listed above) instead of just the symbol. This happens
1086 after the imaginary prefix keys for the event are inserted into the
1087 input stream. @xref{Key Sequence Input}.
1090 @subsection Drag Events
1092 @cindex mouse drag event
1094 With Emacs, you can have a drag event without even changing your
1095 clothes. A @dfn{drag event} happens every time the user presses a mouse
1096 button and then moves the mouse to a different character position before
1097 releasing the button. Like all mouse events, drag events are
1098 represented in Lisp as lists. The lists record both the starting mouse
1099 position and the final position, like this:
1103 (@var{window1} @var{buffer-pos1} (@var{x1} . @var{y1}) @var{timestamp1})
1104 (@var{window2} @var{buffer-pos2} (@var{x2} . @var{y2}) @var{timestamp2})
1108 For a drag event, the name of the symbol @var{event-type} contains the
1109 prefix @samp{drag-}. For example, dragging the mouse with button 2 held
1110 down generates a @code{drag-mouse-2} event. The second and third
1111 elements of the event give the starting and ending position of the drag.
1112 Aside from that, the data have the same meanings as in a click event
1113 (@pxref{Click Events}). You can access the second element of any mouse
1114 event in the same way, with no need to distinguish drag events from
1117 The @samp{drag-} prefix follows the modifier key prefixes such as
1118 @samp{C-} and @samp{M-}.
1120 If @code{read-key-sequence} receives a drag event that has no key
1121 binding, and the corresponding click event does have a binding, it
1122 changes the drag event into a click event at the drag's starting
1123 position. This means that you don't have to distinguish between click
1124 and drag events unless you want to.
1126 @node Button-Down Events
1127 @subsection Button-Down Events
1128 @cindex button-down event
1130 Click and drag events happen when the user releases a mouse button.
1131 They cannot happen earlier, because there is no way to distinguish a
1132 click from a drag until the button is released.
1134 If you want to take action as soon as a button is pressed, you need to
1135 handle @dfn{button-down} events.@footnote{Button-down is the
1136 conservative antithesis of drag.} These occur as soon as a button is
1137 pressed. They are represented by lists that look exactly like click
1138 events (@pxref{Click Events}), except that the @var{event-type} symbol
1139 name contains the prefix @samp{down-}. The @samp{down-} prefix follows
1140 modifier key prefixes such as @samp{C-} and @samp{M-}.
1142 The function @code{read-key-sequence} ignores any button-down events
1143 that don't have command bindings; therefore, the Emacs command loop
1144 ignores them too. This means that you need not worry about defining
1145 button-down events unless you want them to do something. The usual
1146 reason to define a button-down event is so that you can track mouse
1147 motion (by reading motion events) until the button is released.
1148 @xref{Motion Events}.
1151 @subsection Repeat Events
1152 @cindex repeat events
1153 @cindex double-click events
1154 @cindex triple-click events
1156 If you press the same mouse button more than once in quick succession
1157 without moving the mouse, Emacs generates special @dfn{repeat} mouse
1158 events for the second and subsequent presses.
1160 The most common repeat events are @dfn{double-click} events. Emacs
1161 generates a double-click event when you click a button twice; the event
1162 happens when you release the button (as is normal for all click
1165 The event type of a double-click event contains the prefix
1166 @samp{double-}. Thus, a double click on the second mouse button with
1167 @key{meta} held down comes to the Lisp program as
1168 @code{M-double-mouse-2}. If a double-click event has no binding, the
1169 binding of the corresponding ordinary click event is used to execute
1170 it. Thus, you need not pay attention to the double click feature
1171 unless you really want to.
1173 When the user performs a double click, Emacs generates first an ordinary
1174 click event, and then a double-click event. Therefore, you must design
1175 the command binding of the double click event to assume that the
1176 single-click command has already run. It must produce the desired
1177 results of a double click, starting from the results of a single click.
1179 This is convenient, if the meaning of a double click somehow ``builds
1180 on'' the meaning of a single click---which is recommended user interface
1181 design practice for double clicks.
1183 If you click a button, then press it down again and start moving the
1184 mouse with the button held down, then you get a @dfn{double-drag} event
1185 when you ultimately release the button. Its event type contains
1186 @samp{double-drag} instead of just @samp{drag}. If a double-drag event
1187 has no binding, Emacs looks for an alternate binding as if the event
1188 were an ordinary drag.
1190 Before the double-click or double-drag event, Emacs generates a
1191 @dfn{double-down} event when the user presses the button down for the
1192 second time. Its event type contains @samp{double-down} instead of just
1193 @samp{down}. If a double-down event has no binding, Emacs looks for an
1194 alternate binding as if the event were an ordinary button-down event.
1195 If it finds no binding that way either, the double-down event is
1198 To summarize, when you click a button and then press it again right
1199 away, Emacs generates a down event and a click event for the first
1200 click, a double-down event when you press the button again, and finally
1201 either a double-click or a double-drag event.
1203 If you click a button twice and then press it again, all in quick
1204 succession, Emacs generates a @dfn{triple-down} event, followed by
1205 either a @dfn{triple-click} or a @dfn{triple-drag}. The event types of
1206 these events contain @samp{triple} instead of @samp{double}. If any
1207 triple event has no binding, Emacs uses the binding that it would use
1208 for the corresponding double event.
1210 If you click a button three or more times and then press it again, the
1211 events for the presses beyond the third are all triple events. Emacs
1212 does not have separate event types for quadruple, quintuple, etc.@:
1213 events. However, you can look at the event list to find out precisely
1214 how many times the button was pressed.
1216 @defun event-click-count event
1217 This function returns the number of consecutive button presses that led
1218 up to @var{event}. If @var{event} is a double-down, double-click or
1219 double-drag event, the value is 2. If @var{event} is a triple event,
1220 the value is 3 or greater. If @var{event} is an ordinary mouse event
1221 (not a repeat event), the value is 1.
1224 @defvar double-click-time
1225 To generate repeat events, successive mouse button presses must be at
1226 the same screen position, and the number of milliseconds between
1227 successive button presses must be less than the value of
1228 @code{double-click-time}. Setting @code{double-click-time} to
1229 @code{nil} disables multi-click detection entirely. Setting it to
1230 @code{t} removes the time limit; Emacs then detects multi-clicks by
1235 @subsection Motion Events
1236 @cindex motion event
1237 @cindex mouse motion events
1239 Emacs sometimes generates @dfn{mouse motion} events to describe motion
1240 of the mouse without any button activity. Mouse motion events are
1241 represented by lists that look like this:
1244 (mouse-movement (@var{window} @var{buffer-pos} (@var{x} . @var{y}) @var{timestamp}))
1247 The second element of the list describes the current position of the
1248 mouse, just as in a click event (@pxref{Click Events}).
1250 The special form @code{track-mouse} enables generation of motion events
1251 within its body. Outside of @code{track-mouse} forms, Emacs does not
1252 generate events for mere motion of the mouse, and these events do not
1253 appear. @xref{Mouse Tracking}.
1256 @subsection Focus Events
1259 Window systems provide general ways for the user to control which window
1260 gets keyboard input. This choice of window is called the @dfn{focus}.
1261 When the user does something to switch between Emacs frames, that
1262 generates a @dfn{focus event}. The normal definition of a focus event,
1263 in the global keymap, is to select a new frame within Emacs, as the user
1264 would expect. @xref{Input Focus}.
1266 Focus events are represented in Lisp as lists that look like this:
1269 (switch-frame @var{new-frame})
1273 where @var{new-frame} is the frame switched to.
1275 Most X window managers are set up so that just moving the mouse into a
1276 window is enough to set the focus there. Emacs appears to do this,
1277 because it changes the cursor to solid in the new frame. However, there
1278 is no need for the Lisp program to know about the focus change until
1279 some other kind of input arrives. So Emacs generates a focus event only
1280 when the user actually types a keyboard key or presses a mouse button in
1281 the new frame; just moving the mouse between frames does not generate a
1284 A focus event in the middle of a key sequence would garble the
1285 sequence. So Emacs never generates a focus event in the middle of a key
1286 sequence. If the user changes focus in the middle of a key
1287 sequence---that is, after a prefix key---then Emacs reorders the events
1288 so that the focus event comes either before or after the multi-event key
1289 sequence, and not within it.
1292 @subsection Miscellaneous Window System Events
1294 A few other event types represent occurrences within the window system.
1297 @cindex @code{delete-frame} event
1298 @item (delete-frame (@var{frame}))
1299 This kind of event indicates that the user gave the window manager
1300 a command to delete a particular window, which happens to be an Emacs frame.
1302 The standard definition of the @code{delete-frame} event is to delete @var{frame}.
1304 @cindex @code{iconify-frame} event
1305 @item (iconify-frame (@var{frame}))
1306 This kind of event indicates that the user iconified @var{frame} using
1307 the window manager. Its standard definition is @code{ignore}; since the
1308 frame has already been iconified, Emacs has no work to do. The purpose
1309 of this event type is so that you can keep track of such events if you
1312 @cindex @code{make-frame-visible} event
1313 @item (make-frame-visible (@var{frame}))
1314 This kind of event indicates that the user deiconified @var{frame} using
1315 the window manager. Its standard definition is @code{ignore}; since the
1316 frame has already been made visible, Emacs has no work to do.
1319 If one of these events arrives in the middle of a key sequence---that
1320 is, after a prefix key---then Emacs reorders the events so that this
1321 event comes either before or after the multi-event key sequence, not
1324 @node Event Examples
1325 @subsection Event Examples
1327 If the user presses and releases the left mouse button over the same
1328 location, that generates a sequence of events like this:
1331 (down-mouse-1 (#<window 18 on NEWS> 2613 (0 . 38) -864320))
1332 (mouse-1 (#<window 18 on NEWS> 2613 (0 . 38) -864180))
1335 While holding the control key down, the user might hold down the
1336 second mouse button, and drag the mouse from one line to the next.
1337 That produces two events, as shown here:
1340 (C-down-mouse-2 (#<window 18 on NEWS> 3440 (0 . 27) -731219))
1341 (C-drag-mouse-2 (#<window 18 on NEWS> 3440 (0 . 27) -731219)
1342 (#<window 18 on NEWS> 3510 (0 . 28) -729648))
1345 While holding down the meta and shift keys, the user might press the
1346 second mouse button on the window's mode line, and then drag the mouse
1347 into another window. That produces a pair of events like these:
1350 (M-S-down-mouse-2 (#<window 18 on NEWS> mode-line (33 . 31) -457844))
1351 (M-S-drag-mouse-2 (#<window 18 on NEWS> mode-line (33 . 31) -457844)
1352 (#<window 20 on carlton-sanskrit.tex> 161 (33 . 3)
1356 @node Classifying Events
1357 @subsection Classifying Events
1360 Every event has an @dfn{event type}, which classifies the event for
1361 key binding purposes. For a keyboard event, the event type equals the
1362 event value; thus, the event type for a character is the character, and
1363 the event type for a function key symbol is the symbol itself. For
1364 events that are lists, the event type is the symbol in the @sc{car} of
1365 the list. Thus, the event type is always a symbol or a character.
1367 Two events of the same type are equivalent where key bindings are
1368 concerned; thus, they always run the same command. That does not
1369 necessarily mean they do the same things, however, as some commands look
1370 at the whole event to decide what to do. For example, some commands use
1371 the location of a mouse event to decide where in the buffer to act.
1373 Sometimes broader classifications of events are useful. For example,
1374 you might want to ask whether an event involved the @key{META} key,
1375 regardless of which other key or mouse button was used.
1377 The functions @code{event-modifiers} and @code{event-basic-type} are
1378 provided to get such information conveniently.
1380 @defun event-modifiers event
1381 This function returns a list of the modifiers that @var{event} has. The
1382 modifiers are symbols; they include @code{shift}, @code{control},
1383 @code{meta}, @code{alt}, @code{hyper} and @code{super}. In addition,
1384 the modifiers list of a mouse event symbol always contains one of
1385 @code{click}, @code{drag}, and @code{down}.
1387 The argument @var{event} may be an entire event object, or just an event
1390 Here are some examples:
1393 (event-modifiers ?a)
1395 (event-modifiers ?\C-a)
1397 (event-modifiers ?\C-%)
1399 (event-modifiers ?\C-\S-a)
1400 @result{} (control shift)
1401 (event-modifiers 'f5)
1403 (event-modifiers 's-f5)
1405 (event-modifiers 'M-S-f5)
1406 @result{} (meta shift)
1407 (event-modifiers 'mouse-1)
1409 (event-modifiers 'down-mouse-1)
1413 The modifiers list for a click event explicitly contains @code{click},
1414 but the event symbol name itself does not contain @samp{click}.
1417 @defun event-basic-type event
1418 This function returns the key or mouse button that @var{event}
1419 describes, with all modifiers removed. For example:
1422 (event-basic-type ?a)
1424 (event-basic-type ?A)
1426 (event-basic-type ?\C-a)
1428 (event-basic-type ?\C-\S-a)
1430 (event-basic-type 'f5)
1432 (event-basic-type 's-f5)
1434 (event-basic-type 'M-S-f5)
1436 (event-basic-type 'down-mouse-1)
1441 @defun mouse-movement-p object
1442 This function returns non-@code{nil} if @var{object} is a mouse movement
1446 @defun event-convert-list list
1447 This function converts a list of modifier names and a basic event type
1448 to an event type which specifies all of them. For example,
1451 (event-convert-list '(control ?a))
1453 (event-convert-list '(control meta ?a))
1454 @result{} -134217727
1455 (event-convert-list '(control super f1))
1460 @node Accessing Events
1461 @subsection Accessing Events
1463 This section describes convenient functions for accessing the data in
1464 a mouse button or motion event.
1466 These two functions return the starting or ending position of a
1467 mouse-button event, as a list of this form:
1470 (@var{window} @var{buffer-position} (@var{x} . @var{y}) @var{timestamp})
1473 @defun event-start event
1474 This returns the starting position of @var{event}.
1476 If @var{event} is a click or button-down event, this returns the
1477 location of the event. If @var{event} is a drag event, this returns the
1478 drag's starting position.
1481 @defun event-end event
1482 This returns the ending position of @var{event}.
1484 If @var{event} is a drag event, this returns the position where the user
1485 released the mouse button. If @var{event} is a click or button-down
1486 event, the value is actually the starting position, which is the only
1487 position such events have.
1490 These five functions take a position list as described above, and
1491 return various parts of it.
1493 @defun posn-window position
1494 Return the window that @var{position} is in.
1497 @defun posn-point position
1498 Return the buffer position in @var{position}. This is an integer.
1501 @defun posn-x-y position
1502 Return the pixel-based x and y coordinates in @var{position}, as a cons
1503 cell @code{(@var{x} . @var{y})}.
1506 @defun posn-col-row position
1507 Return the row and column (in units of characters) of @var{position}, as
1508 a cons cell @code{(@var{col} . @var{row})}. These are computed from the
1509 @var{x} and @var{y} values actually found in @var{position}.
1512 @defun posn-timestamp position
1513 Return the timestamp in @var{position}.
1516 These functions are useful for decoding scroll bar events.
1518 @defun scroll-bar-event-ratio event
1519 This function returns the fractional vertical position of a scroll bar
1520 event within the scroll bar. The value is a cons cell
1521 @code{(@var{portion} . @var{whole})} containing two integers whose ratio
1522 is the fractional position.
1525 @defun scroll-bar-scale ratio total
1526 This function multiplies (in effect) @var{ratio} by @var{total},
1527 rounding the result to an integer. The argument @var{ratio} is not a
1528 number, but rather a pair @code{(@var{num} . @var{denom})}---typically a
1529 value returned by @code{scroll-bar-event-ratio}.
1531 This function is handy for scaling a position on a scroll bar into a
1532 buffer position. Here's how to do that:
1537 (posn-x-y (event-start event))
1538 (- (point-max) (point-min))))
1541 Recall that scroll bar events have two integers forming ratio in place
1542 of a pair of x and y coordinates.
1545 @node Strings of Events
1546 @subsection Putting Keyboard Events in Strings
1548 In most of the places where strings are used, we conceptualize the
1549 string as containing text characters---the same kind of characters found
1550 in buffers or files. Occasionally Lisp programs use strings that
1551 conceptually contain keyboard characters; for example, they may be key
1552 sequences or keyboard macro definitions. However, storing keyboard
1553 characters in a string is a complex matter, for reasons of historical
1554 compatibility, and it is not always possible.
1556 We recommend that new programs avoid dealing with these complexities
1557 by not storing keyboard events in strings. Here is how to do that:
1561 Use vectors instead of strings for key sequences, when you plan to use
1562 them for anything other than as arguments @code{lookup-key} and
1563 @code{define-key}. For example, you can use
1564 @code{read-key-sequence-vector} instead of @code{read-key-sequence}, and
1565 @code{this-command-keys-vector} instead of @code{this-command-keys}.
1568 Use vectors to write key sequence constants containing meta characters,
1569 even when passing them directly to @code{define-key}.
1572 When you have to look at the contents of a key sequence that might be a
1573 string, use @code{listify-key-sequence} (@pxref{Event Input Misc})
1574 first, to convert it to a list.
1577 The complexities stem from the modifier bits that keyboard input
1578 characters can include. Aside from the Meta modifier, none of these
1579 modifier bits can be included in a string, and the Meta modifier is
1580 allowed only in special cases.
1582 The earliest GNU Emacs versions represented meta characters as codes
1583 in the range of 128 to 255. At that time, the basic character codes
1584 ranged from 0 to 127, so all keyboard character codes did fit in a
1585 string. Many Lisp programs used @samp{\M-} in string constants to stand
1586 for meta characters, especially in arguments to @code{define-key} and
1587 similar functions, and key sequences and sequences of events were always
1588 represented as strings.
1590 When we added support for larger basic character codes beyond 127, and
1591 additional modifier bits, we had to change the representation of meta
1592 characters. Now the flag that represents the Meta modifier in a
1600 and such numbers cannot be included in a string.
1602 To support programs with @samp{\M-} in string constants, there are
1603 special rules for including certain meta characters in a string.
1604 Here are the rules for interpreting keyboard
1608 If the keyboard character value is in the range of 0 to 127, it can go
1609 in the string unchanged.
1612 The meta variants of those characters, with codes in the range of
1626 can also go in the string, but you must change their
1627 numeric values. You must set the
1641 bit, resulting in a value between 128 and 255. Only a unibyte string
1642 can include these codes.
1645 Non-@sc{ASCII} characters above 256 can be included in a multibyte string.
1648 Other keyboard character events cannot fit in a string. This includes
1649 keyboard events in the range of 128 to 255.
1652 Functions such as @code{read-key-sequence} that construct strings of
1653 keyboard input characters follow these rules: they construct vectors
1654 instead of strings, when the events won't fit in a string.
1656 When you use the read syntax @samp{\M-} in a string, it produces a
1657 code in the range of 128 to 255---the same code that you get if you
1658 modify the corresponding keyboard event to put it in the string. Thus,
1659 meta events in strings work consistently regardless of how they get into
1662 However, most programs would do well to avoid these issues by
1663 following the recommendations at the beginning of this section.
1666 @section Reading Input
1668 The editor command loop reads key sequences using the function
1669 @code{read-key-sequence}, which uses @code{read-event}. These and other
1670 functions for event input are also available for use in Lisp programs.
1671 See also @code{momentary-string-display} in @ref{Temporary Displays},
1672 and @code{sit-for} in @ref{Waiting}. @xref{Terminal Input}, for
1673 functions and variables for controlling terminal input modes and
1674 debugging terminal input. @xref{Translating Input}, for features you
1675 can use for translating or modifying input events while reading them.
1677 For higher-level input facilities, see @ref{Minibuffers}.
1680 * Key Sequence Input:: How to read one key sequence.
1681 * Reading One Event:: How to read just one event.
1682 * Quoted Character Input:: Asking the user to specify a character.
1683 * Event Input Misc:: How to reread or throw away input events.
1686 @node Key Sequence Input
1687 @subsection Key Sequence Input
1688 @cindex key sequence input
1690 The command loop reads input a key sequence at a time, by calling
1691 @code{read-key-sequence}. Lisp programs can also call this function;
1692 for example, @code{describe-key} uses it to read the key to describe.
1694 @defun read-key-sequence prompt
1695 @cindex key sequence
1696 This function reads a key sequence and returns it as a string or
1697 vector. It keeps reading events until it has accumulated a complete key
1698 sequence; that is, enough to specify a non-prefix command using the
1699 currently active keymaps.
1701 If the events are all characters and all can fit in a string, then
1702 @code{read-key-sequence} returns a string (@pxref{Strings of Events}).
1703 Otherwise, it returns a vector, since a vector can hold all kinds of
1704 events---characters, symbols, and lists. The elements of the string or
1705 vector are the events in the key sequence.
1707 The argument @var{prompt} is either a string to be displayed in the echo
1708 area as a prompt, or @code{nil}, meaning not to display a prompt.
1710 In the example below, the prompt @samp{?} is displayed in the echo area,
1711 and the user types @kbd{C-x C-f}.
1714 (read-key-sequence "?")
1717 ---------- Echo Area ----------
1719 ---------- Echo Area ----------
1725 The function @code{read-key-sequence} suppresses quitting: @kbd{C-g}
1726 typed while reading with this function works like any other character,
1727 and does not set @code{quit-flag}. @xref{Quitting}.
1730 @defun read-key-sequence-vector prompt
1731 This is like @code{read-key-sequence} except that it always
1732 returns the key sequence as a vector, never as a string.
1733 @xref{Strings of Events}.
1736 @cindex upper case key sequence
1737 @cindex downcasing in @code{lookup-key}
1738 If an input character is an upper-case letter and has no key binding,
1739 but its lower-case equivalent has one, then @code{read-key-sequence}
1740 converts the character to lower case. Note that @code{lookup-key} does
1741 not perform case conversion in this way.
1743 The function @code{read-key-sequence} also transforms some mouse events.
1744 It converts unbound drag events into click events, and discards unbound
1745 button-down events entirely. It also reshuffles focus events and
1746 miscellaneous window events so that they never appear in a key sequence
1747 with any other events.
1749 When mouse events occur in special parts of a window, such as a mode
1750 line or a scroll bar, the event type shows nothing special---it is the
1751 same symbol that would normally represent that combination of mouse
1752 button and modifier keys. The information about the window part is kept
1753 elsewhere in the event---in the coordinates. But
1754 @code{read-key-sequence} translates this information into imaginary
1755 ``prefix keys'', all of which are symbols: @code{mode-line},
1756 @code{vertical-line}, @code{horizontal-scroll-bar} and
1757 @code{vertical-scroll-bar}. You can define meanings for mouse clicks in
1758 special window parts by defining key sequences using these imaginary
1761 For example, if you call @code{read-key-sequence} and then click the
1762 mouse on the window's mode line, you get two events, like this:
1765 (read-key-sequence "Click on the mode line: ")
1766 @result{} [mode-line
1768 (#<window 6 on NEWS> mode-line
1769 (40 . 63) 5959987))]
1772 @defvar num-input-keys
1774 This variable's value is the number of key sequences processed so far in
1775 this Emacs session. This includes key sequences read from the terminal
1776 and key sequences read from keyboard macros being executed.
1779 @tindex num-nonmacro-input-events
1780 @defvar num-nonmacro-input-events
1781 This variable holds the total number of input events received so far
1782 from the terminal---not counting those generated by keyboard macros.
1785 @node Reading One Event
1786 @subsection Reading One Event
1788 The lowest level functions for command input are those that read a
1792 This function reads and returns the next event of command input, waiting
1793 if necessary until an event is available. Events can come directly from
1794 the user or from a keyboard macro.
1796 The function @code{read-event} does not display any message to indicate
1797 it is waiting for input; use @code{message} first, if you wish to
1798 display one. If you have not displayed a message, @code{read-event}
1799 prompts by echoing: it displays descriptions of the events that led to
1800 or were read by the current command. @xref{The Echo Area}.
1802 If @code{cursor-in-echo-area} is non-@code{nil}, then @code{read-event}
1803 moves the cursor temporarily to the echo area, to the end of any message
1804 displayed there. Otherwise @code{read-event} does not move the cursor.
1806 Here is what happens if you call @code{read-event} and then press the
1807 right-arrow function key:
1818 This function reads and returns a character of command input. It
1819 discards any events that are not characters, until it gets a character.
1821 In the first example, the user types the character @kbd{1} (@sc{ASCII}
1822 code 49). The second example shows a keyboard macro definition that
1823 calls @code{read-char} from the minibuffer using @code{eval-expression}.
1824 @code{read-char} reads the keyboard macro's very next character, which
1825 is @kbd{1}. Then @code{eval-expression} displays its return value in
1835 ;; @r{We assume here you use @kbd{M-:} to evaluate this.}
1836 (symbol-function 'foo)
1837 @result{} "^[:(read-char)^M1"
1840 (execute-kbd-macro 'foo)
1847 @node Quoted Character Input
1848 @subsection Quoted Character Input
1849 @cindex quoted character input
1851 You can use the function @code{read-quoted-char} to ask the user to
1852 specify a character, and allow the user to specify a control or meta
1853 character conveniently, either literally or as an octal character code.
1854 The command @code{quoted-insert} uses this function.
1856 @defun read-quoted-char &optional prompt
1857 @cindex octal character input
1858 @cindex control characters, reading
1859 @cindex nonprinting characters, reading
1860 This function is like @code{read-char}, except that if the first
1861 character read is an octal digit (0-7), it reads any number of octal
1862 digits (but stopping if a non-octal digit is found), and returns the
1863 character represented by that numeric character code.
1865 Quitting is suppressed when the first character is read, so that the
1866 user can enter a @kbd{C-g}. @xref{Quitting}.
1868 If @var{prompt} is supplied, it specifies a string for prompting the
1869 user. The prompt string is always displayed in the echo area, followed
1870 by a single @samp{-}.
1872 In the following example, the user types in the octal number 177 (which
1876 (read-quoted-char "What character")
1879 ---------- Echo Area ----------
1880 What character-@kbd{177}
1881 ---------- Echo Area ----------
1889 @node Event Input Misc
1890 @subsection Miscellaneous Event Input Features
1892 This section describes how to ``peek ahead'' at events without using
1893 them up, how to check for pending input, and how to discard pending
1896 @defvar unread-command-events
1898 @cindex peeking at input
1899 This variable holds a list of events waiting to be read as command
1900 input. The events are used in the order they appear in the list, and
1901 removed one by one as they are used.
1903 The variable is needed because in some cases a function reads an event
1904 and then decides not to use it. Storing the event in this variable
1905 causes it to be processed normally, by the command loop or by the
1906 functions to read command input.
1908 @cindex prefix argument unreading
1909 For example, the function that implements numeric prefix arguments reads
1910 any number of digits. When it finds a non-digit event, it must unread
1911 the event so that it can be read normally by the command loop.
1912 Likewise, incremental search uses this feature to unread events with no
1913 special meaning in a search, because these events should exit the search
1914 and then execute normally.
1916 The reliable and easy way to extract events from a key sequence so as to
1917 put them in @code{unread-command-events} is to use
1918 @code{listify-key-sequence} (@pxref{Strings of Events}).
1920 Normally you add events to the front of this list, so that the events
1921 most recently unread will be reread first.
1924 @defun listify-key-sequence key
1925 This function converts the string or vector @var{key} to a list of
1926 individual events, which you can put in @code{unread-command-events}.
1929 @defvar unread-command-char
1930 This variable holds a character to be read as command input.
1931 A value of -1 means ``empty''.
1933 This variable is mostly obsolete now that you can use
1934 @code{unread-command-events} instead; it exists only to support programs
1935 written for Emacs versions 18 and earlier.
1938 @defun input-pending-p
1939 @cindex waiting for command key input
1940 This function determines whether any command input is currently
1941 available to be read. It returns immediately, with value @code{t} if
1942 there is available input, @code{nil} otherwise. On rare occasions it
1943 may return @code{t} when no input is available.
1946 @defvar last-input-event
1947 @defvarx last-input-char
1948 This variable records the last terminal input event read, whether
1949 as part of a command or explicitly by a Lisp program.
1951 In the example below, the Lisp program reads the character @kbd{1},
1952 @sc{ASCII} code 49. It becomes the value of @code{last-input-event},
1953 while @kbd{C-e} (we assume @kbd{C-x C-e} command is used to evaluate
1954 this expression) remains the value of @code{last-command-event}.
1958 (progn (print (read-char))
1959 (print last-command-event)
1967 The alias @code{last-input-char} exists for compatibility with
1971 @defun discard-input
1973 @cindex discard input
1974 @cindex terminate keyboard macro
1975 This function discards the contents of the terminal input buffer and
1976 cancels any keyboard macro that might be in the process of definition.
1977 It returns @code{nil}.
1979 In the following example, the user may type a number of characters right
1980 after starting the evaluation of the form. After the @code{sleep-for}
1981 finishes sleeping, @code{discard-input} discards any characters typed
1985 (progn (sleep-for 2)
1991 @node Special Events
1992 @section Special Events
1994 @cindex special events
1995 Special events are handled at a very low level---as soon as they are
1996 read. The @code{read-event} function processes these events itself, and
1999 Events that are handled in this way do not echo, they are never grouped
2000 into key sequences, and they never appear in the value of
2001 @code{last-command-event} or @code{(this-command-keys)}. They do not
2002 discard a numeric argument, they cannot be unread with
2003 @code{unread-command-events}, they may not appear in a keyboard macro,
2004 and they are not recorded in a keyboard macro while you are defining
2007 These events do, however, appear in @code{last-input-event} immediately
2008 after they are read, and this is the way for the event's definition to
2009 find the actual event.
2011 The events types @code{iconify-frame}, @code{make-frame-visible} and
2012 @code{delete-frame} are normally handled in this way. The keymap which
2013 defines how to handle special events---and which events are special---is
2014 in the variable @code{special-event-map} (@pxref{Active Keymaps}).
2017 @section Waiting for Elapsed Time or Input
2021 The wait functions are designed to wait for a certain amount of time
2022 to pass or until there is input. For example, you may wish to pause in
2023 the middle of a computation to allow the user time to view the display.
2024 @code{sit-for} pauses and updates the screen, and returns immediately if
2025 input comes in, while @code{sleep-for} pauses without updating the
2028 @defun sit-for seconds &optional millisec nodisp
2029 This function performs redisplay (provided there is no pending input
2030 from the user), then waits @var{seconds} seconds, or until input is
2031 available. The value is @code{t} if @code{sit-for} waited the full
2032 time with no input arriving (see @code{input-pending-p} in @ref{Event
2033 Input Misc}). Otherwise, the value is @code{nil}.
2035 The argument @var{seconds} need not be an integer. If it is a floating
2036 point number, @code{sit-for} waits for a fractional number of seconds.
2037 Some systems support only a whole number of seconds; on these systems,
2038 @var{seconds} is rounded down.
2040 The optional argument @var{millisec} specifies an additional waiting
2041 period measured in milliseconds. This adds to the period specified by
2042 @var{seconds}. If the system doesn't support waiting fractions of a
2043 second, you get an error if you specify nonzero @var{millisec}.
2045 @cindex forcing redisplay
2046 Redisplay is always preempted if input arrives, and does not happen at
2047 all if input is available before it starts. Thus, there is no way to
2048 force screen updating if there is pending input; however, if there is no
2049 input pending, you can force an update with no delay by using
2052 If @var{nodisp} is non-@code{nil}, then @code{sit-for} does not
2053 redisplay, but it still returns as soon as input is available (or when
2054 the timeout elapses).
2056 Iconifying or deiconifying a frame makes @code{sit-for} return, because
2057 that generates an event. @xref{Misc Events}.
2059 The usual purpose of @code{sit-for} is to give the user time to read
2060 text that you display.
2063 @defun sleep-for seconds &optional millisec
2064 This function simply pauses for @var{seconds} seconds without updating
2065 the display. It pays no attention to available input. It returns
2068 The argument @var{seconds} need not be an integer. If it is a floating
2069 point number, @code{sleep-for} waits for a fractional number of seconds.
2070 Some systems support only a whole number of seconds; on these systems,
2071 @var{seconds} is rounded down.
2073 The optional argument @var{millisec} specifies an additional waiting
2074 period measured in milliseconds. This adds to the period specified by
2075 @var{seconds}. If the system doesn't support waiting fractions of a
2076 second, you get an error if you specify nonzero @var{millisec}.
2078 Use @code{sleep-for} when you wish to guarantee a delay.
2081 @xref{Time of Day}, for functions to get the current time.
2088 Typing @kbd{C-g} while a Lisp function is running causes Emacs to
2089 @dfn{quit} whatever it is doing. This means that control returns to the
2090 innermost active command loop.
2092 Typing @kbd{C-g} while the command loop is waiting for keyboard input
2093 does not cause a quit; it acts as an ordinary input character. In the
2094 simplest case, you cannot tell the difference, because @kbd{C-g}
2095 normally runs the command @code{keyboard-quit}, whose effect is to quit.
2096 However, when @kbd{C-g} follows a prefix key, they combine to form an
2097 undefined key. The effect is to cancel the prefix key as well as any
2100 In the minibuffer, @kbd{C-g} has a different definition: it aborts out
2101 of the minibuffer. This means, in effect, that it exits the minibuffer
2102 and then quits. (Simply quitting would return to the command loop
2103 @emph{within} the minibuffer.) The reason why @kbd{C-g} does not quit
2104 directly when the command reader is reading input is so that its meaning
2105 can be redefined in the minibuffer in this way. @kbd{C-g} following a
2106 prefix key is not redefined in the minibuffer, and it has its normal
2107 effect of canceling the prefix key and prefix argument. This too
2108 would not be possible if @kbd{C-g} always quit directly.
2110 When @kbd{C-g} does directly quit, it does so by setting the variable
2111 @code{quit-flag} to @code{t}. Emacs checks this variable at appropriate
2112 times and quits if it is not @code{nil}. Setting @code{quit-flag}
2113 non-@code{nil} in any way thus causes a quit.
2115 At the level of C code, quitting cannot happen just anywhere; only at the
2116 special places that check @code{quit-flag}. The reason for this is
2117 that quitting at other places might leave an inconsistency in Emacs's
2118 internal state. Because quitting is delayed until a safe place, quitting
2119 cannot make Emacs crash.
2121 Certain functions such as @code{read-key-sequence} or
2122 @code{read-quoted-char} prevent quitting entirely even though they wait
2123 for input. Instead of quitting, @kbd{C-g} serves as the requested
2124 input. In the case of @code{read-key-sequence}, this serves to bring
2125 about the special behavior of @kbd{C-g} in the command loop. In the
2126 case of @code{read-quoted-char}, this is so that @kbd{C-q} can be used
2127 to quote a @kbd{C-g}.
2129 You can prevent quitting for a portion of a Lisp function by binding
2130 the variable @code{inhibit-quit} to a non-@code{nil} value. Then,
2131 although @kbd{C-g} still sets @code{quit-flag} to @code{t} as usual, the
2132 usual result of this---a quit---is prevented. Eventually,
2133 @code{inhibit-quit} will become @code{nil} again, such as when its
2134 binding is unwound at the end of a @code{let} form. At that time, if
2135 @code{quit-flag} is still non-@code{nil}, the requested quit happens
2136 immediately. This behavior is ideal when you wish to make sure that
2137 quitting does not happen within a ``critical section'' of the program.
2139 @cindex @code{read-quoted-char} quitting
2140 In some functions (such as @code{read-quoted-char}), @kbd{C-g} is
2141 handled in a special way that does not involve quitting. This is done
2142 by reading the input with @code{inhibit-quit} bound to @code{t}, and
2143 setting @code{quit-flag} to @code{nil} before @code{inhibit-quit}
2144 becomes @code{nil} again. This excerpt from the definition of
2145 @code{read-quoted-char} shows how this is done; it also shows that
2146 normal quitting is permitted after the first character of input.
2149 (defun read-quoted-char (&optional prompt)
2150 "@dots{}@var{documentation}@dots{}"
2151 (let ((message-log-max nil) done (first t) (code 0) char)
2153 (let ((inhibit-quit first)
2155 (and prompt (message "%s-" prompt))
2156 (setq char (read-event))
2157 (if inhibit-quit (setq quit-flag nil)))
2158 @r{@dots{}set the variable @code{code}@dots{}})
2163 If this variable is non-@code{nil}, then Emacs quits immediately, unless
2164 @code{inhibit-quit} is non-@code{nil}. Typing @kbd{C-g} ordinarily sets
2165 @code{quit-flag} non-@code{nil}, regardless of @code{inhibit-quit}.
2168 @defvar inhibit-quit
2169 This variable determines whether Emacs should quit when @code{quit-flag}
2170 is set to a value other than @code{nil}. If @code{inhibit-quit} is
2171 non-@code{nil}, then @code{quit-flag} has no special effect.
2174 @deffn Command keyboard-quit
2175 This function signals the @code{quit} condition with @code{(signal 'quit
2176 nil)}. This is the same thing that quitting does. (See @code{signal}
2180 You can specify a character other than @kbd{C-g} to use for quitting.
2181 See the function @code{set-input-mode} in @ref{Terminal Input}.
2183 @node Prefix Command Arguments
2184 @section Prefix Command Arguments
2185 @cindex prefix argument
2186 @cindex raw prefix argument
2187 @cindex numeric prefix argument
2189 Most Emacs commands can use a @dfn{prefix argument}, a number
2190 specified before the command itself. (Don't confuse prefix arguments
2191 with prefix keys.) The prefix argument is at all times represented by a
2192 value, which may be @code{nil}, meaning there is currently no prefix
2193 argument. Each command may use the prefix argument or ignore it.
2195 There are two representations of the prefix argument: @dfn{raw} and
2196 @dfn{numeric}. The editor command loop uses the raw representation
2197 internally, and so do the Lisp variables that store the information, but
2198 commands can request either representation.
2200 Here are the possible values of a raw prefix argument:
2204 @code{nil}, meaning there is no prefix argument. Its numeric value is
2205 1, but numerous commands make a distinction between @code{nil} and the
2209 An integer, which stands for itself.
2212 A list of one element, which is an integer. This form of prefix
2213 argument results from one or a succession of @kbd{C-u}'s with no
2214 digits. The numeric value is the integer in the list, but some
2215 commands make a distinction between such a list and an integer alone.
2218 The symbol @code{-}. This indicates that @kbd{M--} or @kbd{C-u -} was
2219 typed, without following digits. The equivalent numeric value is
2220 @minus{}1, but some commands make a distinction between the integer
2221 @minus{}1 and the symbol @code{-}.
2224 We illustrate these possibilities by calling the following function with
2229 (defun display-prefix (arg)
2230 "Display the value of the raw prefix arg."
2237 Here are the results of calling @code{display-prefix} with various
2238 raw prefix arguments:
2241 M-x display-prefix @print{} nil
2243 C-u M-x display-prefix @print{} (4)
2245 C-u C-u M-x display-prefix @print{} (16)
2247 C-u 3 M-x display-prefix @print{} 3
2249 M-3 M-x display-prefix @print{} 3 ; @r{(Same as @code{C-u 3}.)}
2251 C-u - M-x display-prefix @print{} -
2253 M-- M-x display-prefix @print{} - ; @r{(Same as @code{C-u -}.)}
2255 C-u - 7 M-x display-prefix @print{} -7
2257 M-- 7 M-x display-prefix @print{} -7 ; @r{(Same as @code{C-u -7}.)}
2260 Emacs uses two variables to store the prefix argument:
2261 @code{prefix-arg} and @code{current-prefix-arg}. Commands such as
2262 @code{universal-argument} that set up prefix arguments for other
2263 commands store them in @code{prefix-arg}. In contrast,
2264 @code{current-prefix-arg} conveys the prefix argument to the current
2265 command, so setting it has no effect on the prefix arguments for future
2268 Normally, commands specify which representation to use for the prefix
2269 argument, either numeric or raw, in the @code{interactive} declaration.
2270 (@xref{Using Interactive}.) Alternatively, functions may look at the
2271 value of the prefix argument directly in the variable
2272 @code{current-prefix-arg}, but this is less clean.
2274 @defun prefix-numeric-value arg
2275 This function returns the numeric meaning of a valid raw prefix argument
2276 value, @var{arg}. The argument may be a symbol, a number, or a list.
2277 If it is @code{nil}, the value 1 is returned; if it is @code{-}, the
2278 value @minus{}1 is returned; if it is a number, that number is returned;
2279 if it is a list, the @sc{car} of that list (which should be a number) is
2283 @defvar current-prefix-arg
2284 This variable holds the raw prefix argument for the @emph{current}
2285 command. Commands may examine it directly, but the usual method for
2286 accessing it is with @code{(interactive "P")}.
2290 The value of this variable is the raw prefix argument for the
2291 @emph{next} editing command. Commands such as @code{universal-argument}
2292 that specify prefix arguments for the following command work by setting
2296 The following commands exist to set up prefix arguments for the
2297 following command. Do not call them for any other reason.
2299 @deffn Command universal-argument
2300 This command reads input and specifies a prefix argument for the
2301 following command. Don't call this command yourself unless you know
2305 @deffn Command digit-argument arg
2306 This command adds to the prefix argument for the following command. The
2307 argument @var{arg} is the raw prefix argument as it was before this
2308 command; it is used to compute the updated prefix argument. Don't call
2309 this command yourself unless you know what you are doing.
2312 @deffn Command negative-argument arg
2313 This command adds to the numeric argument for the next command. The
2314 argument @var{arg} is the raw prefix argument as it was before this
2315 command; its value is negated to form the new prefix argument. Don't
2316 call this command yourself unless you know what you are doing.
2319 @node Recursive Editing
2320 @section Recursive Editing
2321 @cindex recursive command loop
2322 @cindex recursive editing level
2323 @cindex command loop, recursive
2325 The Emacs command loop is entered automatically when Emacs starts up.
2326 This top-level invocation of the command loop never exits; it keeps
2327 running as long as Emacs does. Lisp programs can also invoke the
2328 command loop. Since this makes more than one activation of the command
2329 loop, we call it @dfn{recursive editing}. A recursive editing level has
2330 the effect of suspending whatever command invoked it and permitting the
2331 user to do arbitrary editing before resuming that command.
2333 The commands available during recursive editing are the same ones
2334 available in the top-level editing loop and defined in the keymaps.
2335 Only a few special commands exit the recursive editing level; the others
2336 return to the recursive editing level when they finish. (The special
2337 commands for exiting are always available, but they do nothing when
2338 recursive editing is not in progress.)
2340 All command loops, including recursive ones, set up all-purpose error
2341 handlers so that an error in a command run from the command loop will
2344 @cindex minibuffer input
2345 Minibuffer input is a special kind of recursive editing. It has a few
2346 special wrinkles, such as enabling display of the minibuffer and the
2347 minibuffer window, but fewer than you might suppose. Certain keys
2348 behave differently in the minibuffer, but that is only because of the
2349 minibuffer's local map; if you switch windows, you get the usual Emacs
2352 @cindex @code{throw} example
2354 @cindex exit recursive editing
2356 To invoke a recursive editing level, call the function
2357 @code{recursive-edit}. This function contains the command loop; it also
2358 contains a call to @code{catch} with tag @code{exit}, which makes it
2359 possible to exit the recursive editing level by throwing to @code{exit}
2360 (@pxref{Catch and Throw}). If you throw a value other than @code{t},
2361 then @code{recursive-edit} returns normally to the function that called
2362 it. The command @kbd{C-M-c} (@code{exit-recursive-edit}) does this.
2363 Throwing a @code{t} value causes @code{recursive-edit} to quit, so that
2364 control returns to the command loop one level up. This is called
2365 @dfn{aborting}, and is done by @kbd{C-]} (@code{abort-recursive-edit}).
2367 Most applications should not use recursive editing, except as part of
2368 using the minibuffer. Usually it is more convenient for the user if you
2369 change the major mode of the current buffer temporarily to a special
2370 major mode, which should have a command to go back to the previous mode.
2371 (The @kbd{e} command in Rmail uses this technique.) Or, if you wish to
2372 give the user different text to edit ``recursively'', create and select
2373 a new buffer in a special mode. In this mode, define a command to
2374 complete the processing and go back to the previous buffer. (The
2375 @kbd{m} command in Rmail does this.)
2377 Recursive edits are useful in debugging. You can insert a call to
2378 @code{debug} into a function definition as a sort of breakpoint, so that
2379 you can look around when the function gets there. @code{debug} invokes
2380 a recursive edit but also provides the other features of the debugger.
2382 Recursive editing levels are also used when you type @kbd{C-r} in
2383 @code{query-replace} or use @kbd{C-x q} (@code{kbd-macro-query}).
2385 @defun recursive-edit
2386 @cindex suspend evaluation
2387 This function invokes the editor command loop. It is called
2388 automatically by the initialization of Emacs, to let the user begin
2389 editing. When called from a Lisp program, it enters a recursive editing
2392 In the following example, the function @code{simple-rec} first
2393 advances point one word, then enters a recursive edit, printing out a
2394 message in the echo area. The user can then do any editing desired, and
2395 then type @kbd{C-M-c} to exit and continue executing @code{simple-rec}.
2398 (defun simple-rec ()
2400 (message "Recursive edit in progress")
2403 @result{} simple-rec
2409 @deffn Command exit-recursive-edit
2410 This function exits from the innermost recursive edit (including
2411 minibuffer input). Its definition is effectively @code{(throw 'exit
2415 @deffn Command abort-recursive-edit
2416 This function aborts the command that requested the innermost recursive
2417 edit (including minibuffer input), by signaling @code{quit}
2418 after exiting the recursive edit. Its definition is effectively
2419 @code{(throw 'exit t)}. @xref{Quitting}.
2422 @deffn Command top-level
2423 This function exits all recursive editing levels; it does not return a
2424 value, as it jumps completely out of any computation directly back to
2425 the main command loop.
2428 @defun recursion-depth
2429 This function returns the current depth of recursive edits. When no
2430 recursive edit is active, it returns 0.
2433 @node Disabling Commands
2434 @section Disabling Commands
2435 @cindex disabled command
2437 @dfn{Disabling a command} marks the command as requiring user
2438 confirmation before it can be executed. Disabling is used for commands
2439 which might be confusing to beginning users, to prevent them from using
2440 the commands by accident.
2443 The low-level mechanism for disabling a command is to put a
2444 non-@code{nil} @code{disabled} property on the Lisp symbol for the
2445 command. These properties are normally set up by the user's
2446 @file{.emacs} file with Lisp expressions such as this:
2449 (put 'upcase-region 'disabled t)
2453 For a few commands, these properties are present by default and may be
2454 removed by the @file{.emacs} file.
2456 If the value of the @code{disabled} property is a string, the message
2457 saying the command is disabled includes that string. For example:
2460 (put 'delete-region 'disabled
2461 "Text deleted this way cannot be yanked back!\n")
2464 @xref{Disabling,,, emacs, The GNU Emacs Manual}, for the details on
2465 what happens when a disabled command is invoked interactively.
2466 Disabling a command has no effect on calling it as a function from Lisp
2469 @deffn Command enable-command command
2470 Allow @var{command} to be executed without special confirmation from now
2471 on, and (if the user confirms) alter the user's @file{.emacs} file so
2472 that this will apply to future sessions.
2475 @deffn Command disable-command command
2476 Require special confirmation to execute @var{command} from now on, and
2477 (if the user confirms) alter the user's @file{.emacs} file so that this
2478 will apply to future sessions.
2481 @defvar disabled-command-hook
2482 When the user invokes a disabled command interactively, this normal hook
2483 is run instead of the disabled command. The hook functions can use
2484 @code{this-command-keys} to determine what the user typed to run the
2485 command, and thus find the command itself. @xref{Hooks}.
2487 By default, @code{disabled-command-hook} contains a function that asks
2488 the user whether to proceed.
2491 @node Command History
2492 @section Command History
2493 @cindex command history
2494 @cindex complex command
2495 @cindex history of commands
2497 The command loop keeps a history of the complex commands that have
2498 been executed, to make it convenient to repeat these commands. A
2499 @dfn{complex command} is one for which the interactive argument reading
2500 uses the minibuffer. This includes any @kbd{M-x} command, any
2501 @kbd{M-:} command, and any command whose @code{interactive}
2502 specification reads an argument from the minibuffer. Explicit use of
2503 the minibuffer during the execution of the command itself does not cause
2504 the command to be considered complex.
2506 @defvar command-history
2507 This variable's value is a list of recent complex commands, each
2508 represented as a form to evaluate. It continues to accumulate all
2509 complex commands for the duration of the editing session, but all but
2510 the first (most recent) thirty elements are deleted when a garbage
2511 collection takes place (@pxref{Garbage Collection}).
2516 @result{} ((switch-to-buffer "chistory.texi")
2517 (describe-key "^X^[")
2518 (visit-tags-table "~/emacs/src/")
2519 (find-tag "repeat-complex-command"))
2524 This history list is actually a special case of minibuffer history
2525 (@pxref{Minibuffer History}), with one special twist: the elements are
2526 expressions rather than strings.
2528 There are a number of commands devoted to the editing and recall of
2529 previous commands. The commands @code{repeat-complex-command}, and
2530 @code{list-command-history} are described in the user manual
2531 (@pxref{Repetition,,, emacs, The GNU Emacs Manual}). Within the
2532 minibuffer, the usual minibuffer history commands are available.
2534 @node Keyboard Macros
2535 @section Keyboard Macros
2536 @cindex keyboard macros
2538 A @dfn{keyboard macro} is a canned sequence of input events that can
2539 be considered a command and made the definition of a key. The Lisp
2540 representation of a keyboard macro is a string or vector containing the
2541 events. Don't confuse keyboard macros with Lisp macros
2544 @defun execute-kbd-macro kbdmacro &optional count
2545 This function executes @var{kbdmacro} as a sequence of events. If
2546 @var{kbdmacro} is a string or vector, then the events in it are executed
2547 exactly as if they had been input by the user. The sequence is
2548 @emph{not} expected to be a single key sequence; normally a keyboard
2549 macro definition consists of several key sequences concatenated.
2551 If @var{kbdmacro} is a symbol, then its function definition is used in
2552 place of @var{kbdmacro}. If that is another symbol, this process repeats.
2553 Eventually the result should be a string or vector. If the result is
2554 not a symbol, string, or vector, an error is signaled.
2556 The argument @var{count} is a repeat count; @var{kbdmacro} is executed that
2557 many times. If @var{count} is omitted or @code{nil}, @var{kbdmacro} is
2558 executed once. If it is 0, @var{kbdmacro} is executed over and over until it
2559 encounters an error or a failing search.
2561 @xref{Reading One Event}, for an example of using @code{execute-kbd-macro}.
2564 @defvar executing-macro
2565 This variable contains the string or vector that defines the keyboard
2566 macro that is currently executing. It is @code{nil} if no macro is
2567 currently executing. A command can test this variable so as to behave
2568 differently when run from an executing macro. Do not set this variable
2572 @defvar defining-kbd-macro
2573 This variable indicates whether a keyboard macro is being defined. A
2574 command can test this variable so as to behave differently while a macro
2575 is being defined. The commands @code{start-kbd-macro} and
2576 @code{end-kbd-macro} set this variable---do not set it yourself.
2578 The variable is always local to the current terminal and cannot be
2579 buffer-local. @xref{Multiple Displays}.
2582 @defvar last-kbd-macro
2583 This variable is the definition of the most recently defined keyboard
2584 macro. Its value is a string or vector, or @code{nil}.
2586 The variable is always local to the current terminal and cannot be
2587 buffer-local. @xref{Multiple Displays}.