1 @c This is part of the Emacs manual.
2 @c Copyright (C) 1985, 86, 87, 93, 94, 95, 97, 00, 2001
3 @c Free Software Foundation, Inc.
4 @c See file emacs.texi for copying conditions.
5 @node Minibuffer, M-x, Basic, Top
6 @chapter The Minibuffer
9 The @dfn{minibuffer} is the facility used by Emacs commands to read
10 arguments more complicated than a single number. Minibuffer arguments
11 can be file names, buffer names, Lisp function names, Emacs command
12 names, Lisp expressions, and many other things, depending on the command
13 reading the argument. You can use the usual Emacs editing commands in
14 the minibuffer to edit the argument text.
17 When the minibuffer is in use, it appears in the echo area, and the
18 terminal's cursor moves there. The beginning of the minibuffer line
19 displays a @dfn{prompt} in a special color, to say what kind of input
20 you should supply and how it will be used. Often this prompt is
21 derived from the name of the command that the argument is for. The
22 prompt normally ends with a colon.
24 @cindex default argument
25 Sometimes a @dfn{default argument} appears in parentheses after the
26 colon; it too is part of the prompt. The default will be used as the
27 argument value if you enter an empty argument (that is, just type
28 @key{RET}). For example, commands that read buffer names always show a
29 default, which is the name of the buffer that will be used if you type
32 The simplest way to enter a minibuffer argument is to type the text
33 you want, terminated by @key{RET} which exits the minibuffer. You can
34 cancel the command that wants the argument, and get out of the
35 minibuffer, by typing @kbd{C-g}.
37 Since the minibuffer uses the screen space of the echo area, it can
38 conflict with other ways Emacs customarily uses the echo area. Here is how
39 Emacs handles such conflicts:
43 If a command gets an error while you are in the minibuffer, this does
44 not cancel the minibuffer. However, the echo area is needed for the
45 error message and therefore the minibuffer itself is hidden for a
46 while. It comes back after a few seconds, or as soon as you type
50 If in the minibuffer you use a command whose purpose is to display a
51 message in the echo area, such as @kbd{C-x =}, the message hides the
52 minibuffer for a while. The minibuffer contents come back after a few
53 seconds, or as soon as you type anything.
56 Echoing of keystrokes does not take place while the minibuffer is in
61 * File: Minibuffer File. Entering file names with the minibuffer.
62 * Edit: Minibuffer Edit. How to edit in the minibuffer.
63 * Completion:: An abbreviation facility for minibuffer input.
64 * Minibuffer History:: Reusing recent minibuffer arguments.
65 * Repetition:: Re-executing commands that used the minibuffer.
69 @section Minibuffers for File Names
71 Sometimes the minibuffer starts out with text in it. For example, when
72 you are supposed to give a file name, the minibuffer starts out containing
73 the @dfn{default directory}, which ends with a slash. This is to inform
74 you which directory the file will be found in if you do not specify a
77 @c Separate paragraph to clean up ugly page break--rms
79 For example, the minibuffer might start out with these contents:
82 Find File: /u2/emacs/src/
86 where @samp{Find File:@: } is the prompt. Typing @kbd{buffer.c} as
87 input specifies the file @file{/u2/emacs/src/buffer.c}. To find files
88 in nearby directories, use @kbd{..}; thus, if you type
89 @kbd{../lisp/simple.el}, you will get the file named
90 @file{/u2/emacs/lisp/simple.el}. Alternatively, you can kill with
91 @kbd{M-@key{DEL}} the directory names you don't want (@pxref{Words}).
93 If you don't want any of the default, you can kill it with @kbd{C-a
94 C-k}. But you don't need to kill the default; you can simply ignore it.
95 Insert an absolute file name, one starting with a slash or a tilde,
96 after the default directory. For example, to specify the file
97 @file{/etc/termcap}, just insert that name, giving these minibuffer
101 Find File: /u2/emacs/src//etc/termcap
105 @cindex // in file name
106 @cindex double slash in file name
107 @cindex slashes repeated in file name
108 GNU Emacs gives a special meaning to a double slash (which is not
109 normally a useful thing to write): it means, ``ignore everything before
110 the second slash in the pair.'' Thus, @samp{/u2/emacs/src/} is ignored
111 in the example above, and you get the file @file{/etc/termcap}.
113 If you set @code{insert-default-directory} to @code{nil}, the default
114 directory is not inserted in the minibuffer. This way, the minibuffer
115 starts out empty. But the name you type, if relative, is still
116 interpreted with respect to the same default directory.
118 @node Minibuffer Edit
119 @section Editing in the Minibuffer
121 The minibuffer is an Emacs buffer (albeit a peculiar one), and the usual
122 Emacs commands are available for editing the text of an argument you are
125 Since @key{RET} in the minibuffer is defined to exit the minibuffer,
126 you can't use it to insert a newline in the minibuffer. To do that,
127 type @kbd{C-o} or @kbd{C-q C-j}. (On text terminals, newline is
128 really the @acronym{ASCII} character control-J.)
130 The minibuffer has its own window which always has space on the screen
131 but acts as if it were not there when the minibuffer is not in use. When
132 the minibuffer is in use, its window is just like the others; you can
133 switch to another window with @kbd{C-x o}, edit text in other windows and
134 perhaps even visit more files, before returning to the minibuffer to submit
135 the argument. You can kill text in another window, return to the
136 minibuffer window, and then yank the text to use it in the argument.
139 @cindex height of minibuffer
140 @cindex size of minibuffer
141 @cindex growing minibuffer
142 @cindex resizing minibuffer
143 There are some restrictions on the use of the minibuffer window,
144 however. You cannot switch buffers in it---the minibuffer and its
145 window are permanently attached. Also, you cannot split or kill the
146 minibuffer window. But you can make it taller in the normal fashion
149 @vindex resize-mini-windows
150 The minibuffer window expands vertically as necessary to hold the
151 text that you put in the minibuffer. If @code{resize-mini-windows} is
152 @code{t} (the default), the window is always resized to fit the size
153 of the text it displays. If its value is the symbol @code{grow-only},
154 the window grows when the size of displayed text increases, but
155 shrinks (back to the normal size) only when the minibuffer becomes
156 inactive. If its value is @code{nil}, you have to adjust the height
159 @vindex max-mini-window-height
160 The variable @code{max-mini-window-height} controls the maximum
161 height for resizing the minibuffer window: a floating-point number
162 specifies a fraction of the frame's height; an integer specifies the
163 maximum number of lines; @code{nil} means do not resize the minibuffer
164 window automatically. The default value is 0.25.
166 If, while in the minibuffer, you issue a command that displays help
167 text of any sort in another window, you can use the @kbd{C-M-v}
168 command while in the minibuffer to scroll the help text.
169 (@kbd{M-@key{PAGEUP}} and @kbd{M-@key{PAGEDOWN}} also operate on that
170 help text.) This lasts until you exit the minibuffer. This feature
171 is especially useful when you display a buffer listing possible
172 completions. @xref{Other Window}.
174 @vindex enable-recursive-minibuffers
175 Emacs normally disallows most commands that use the minibuffer while
176 the minibuffer is active. This rule is to prevent recursive minibuffers
177 from confusing novice users. If you want to be able to use such
178 commands in the minibuffer, set the variable
179 @code{enable-recursive-minibuffers} to a non-@code{nil} value.
185 For certain kinds of arguments, you can use @dfn{completion} to enter
186 the argument value. Completion means that you type part of the
187 argument, then Emacs visibly fills in the rest, or as much as
188 can be determined from the part you have typed.
190 When completion is available, certain keys---@key{TAB}, @key{RET}, and
191 @key{SPC}---are rebound to complete the text in the minibuffer before point
192 into a longer string that it stands for, by matching it against a set of
193 @dfn{completion alternatives} provided by the command reading the
194 argument. @kbd{?} is defined to display a list of possible completions
195 of what you have inserted.
197 For example, when @kbd{M-x} uses the minibuffer to read the name of a
198 command, it provides a list of all available Emacs command names to
199 complete against. The completion keys match the minibuffer text
200 against all the command names, find any additional name characters
201 implied by the ones already present in the minibuffer, and add those
202 characters to the ones you have given. This is what makes it possible
203 to type @kbd{M-x ins @key{SPC} b @key{RET}} instead of @kbd{M-x
204 insert-buffer @key{RET}} (for example).
206 Case is normally significant in completion, because it is significant
207 in most of the names that you can complete (buffer names, file names and
208 command names). Thus, @samp{fo} does not complete to @samp{Foo}.
209 Completion does ignore case distinctions for certain arguments in which
210 case does not matter.
213 * Example: Completion Example. Examples of using completion.
214 * Commands: Completion Commands. A list of completion commands.
215 * Strict Completion:: Different types of completion.
216 * Options: Completion Options. Options for completion.
219 @node Completion Example
220 @subsection Completion Example
222 @kindex TAB @r{(completion)}
223 @findex minibuffer-complete
224 A concrete example may help here. If you type @kbd{M-x au @key{TAB}},
225 the @key{TAB} looks for alternatives (in this case, command names) that
226 start with @samp{au}. There are several, including
227 @code{auto-fill-mode} and @code{auto-save-mode}---but they are all the
228 same as far as @code{auto-}, so the @samp{au} in the minibuffer changes
229 to @samp{auto-}.@refill
231 If you type @key{TAB} again immediately, there are multiple
232 possibilities for the very next character---it could be any of
233 @samp{cfilrs}---so no more characters are added; instead, @key{TAB}
234 displays a list of all possible completions in another window.
236 If you go on to type @kbd{f @key{TAB}}, this @key{TAB} sees
237 @samp{auto-f}. The only command name starting this way is
238 @code{auto-fill-mode}, so completion fills in the rest of that. You now
239 have @samp{auto-fill-mode} in the minibuffer after typing just @kbd{au
240 @key{TAB} f @key{TAB}}. Note that @key{TAB} has this effect because in
241 the minibuffer it is bound to the command @code{minibuffer-complete}
242 when completion is available.
244 @node Completion Commands
245 @subsection Completion Commands
247 Here is a list of the completion commands defined in the minibuffer
248 when completion is available.
252 Complete the text before point in the minibuffer as much as possible
253 (@code{minibuffer-complete}).
255 Complete the minibuffer text before point, but don't go beyond one word
256 (@code{minibuffer-complete-word}).
258 Submit the text in the minibuffer as the argument, possibly completing
259 first as described below (@code{minibuffer-complete-and-exit}).
261 Display a list of all possible completions of the text in the minibuffer
262 (@code{minibuffer-list-completions}).
266 @findex minibuffer-complete-word
267 @key{SPC} completes much like @key{TAB}, but never goes beyond the
268 next hyphen or space. If you have @samp{auto-f} in the minibuffer and
269 type @key{SPC}, it finds that the completion is @samp{auto-fill-mode},
270 but it stops completing after @samp{fill-}. This gives
271 @samp{auto-fill-}. Another @key{SPC} at this point completes all the
272 way to @samp{auto-fill-mode}. The command that implements this
273 behavior is called @code{minibuffer-complete-word}.
275 Here are some commands you can use to choose a completion from a
276 window that displays a list of completions:
279 @findex mouse-choose-completion
282 Clicking mouse button 1 or 2 on a completion in the list of possible
283 completions chooses that completion (@code{mouse-choose-completion}).
284 You normally use this command while point is in the minibuffer, but you
285 must click in the list of completions, not in the minibuffer itself.
287 @findex switch-to-completions
290 Typing @key{PRIOR} or @key{PAGE-UP}, or @kbd{M-v}, while in the
291 minibuffer, selects the window showing the completion list buffer
292 (@code{switch-to-completions}). This paves the way for using the
293 commands below. (Selecting that window in the usual ways has the same
294 effect, but this way is more convenient.)
296 @findex choose-completion
298 Typing @key{RET} @emph{in the completion list buffer} chooses the
299 completion that point is in or next to (@code{choose-completion}). To
300 use this command, you must first switch windows to the window that shows
301 the list of completions.
303 @findex next-completion
305 Typing the right-arrow key @key{RIGHT} @emph{in the completion list
306 buffer} moves point to the following completion (@code{next-completion}).
308 @findex previous-completion
310 Typing the left-arrow key @key{LEFT} @emph{in the completion list
311 buffer} moves point toward the beginning of the buffer, to the previous
312 completion (@code{previous-completion}).
315 @node Strict Completion
316 @subsection Strict Completion
318 There are three different ways that @key{RET} can work in completing
319 minibuffers, depending on how the argument will be used.
323 @dfn{Strict} completion is used when it is meaningless to give any
324 argument except one of the known alternatives. For example, when
325 @kbd{C-x k} reads the name of a buffer to kill, it is meaningless to
326 give anything but the name of an existing buffer. In strict
327 completion, @key{RET} refuses to exit if the text in the minibuffer
328 does not complete to an exact match.
331 @dfn{Cautious} completion is similar to strict completion, except that
332 @key{RET} exits only if the text was an exact match already, not
333 needing completion. If the text is not an exact match, @key{RET} does
334 not exit, but it does complete the text. If it completes to an exact
335 match, a second @key{RET} will exit.
337 Cautious completion is used for reading file names for files that must
341 @dfn{Permissive} completion is used when any string whatever is
342 meaningful, and the list of completion alternatives is just a guide.
343 For example, when @kbd{C-x C-f} reads the name of a file to visit, any
344 file name is allowed, in case you want to create a file. In
345 permissive completion, @key{RET} takes the text in the minibuffer
346 exactly as given, without completing it.
349 The completion commands display a list of all possible completions in
350 a window whenever there is more than one possibility for the very next
351 character. Also, typing @kbd{?} explicitly requests such a list. If
352 the list of completions is long, you can scroll it with @kbd{C-M-v}
353 (@pxref{Other Window}).
355 @node Completion Options
356 @subsection Completion Options
358 @vindex completion-ignored-extensions
359 @cindex ignored file names, in completion
360 When completion is done on file names, certain file names are usually
361 ignored. The variable @code{completion-ignored-extensions} contains a
362 list of strings; a file whose name ends in any of those strings is
363 ignored as a possible completion. The standard value of this variable
364 has several elements including @code{".o"}, @code{".elc"}, @code{".dvi"}
365 and @code{"~"}. The effect is that, for example, @samp{foo} can
366 complete to @samp{foo.c} even though @samp{foo.o} exists as well.
367 However, if @emph{all} the possible completions end in ``ignored''
368 strings, then they are not ignored. Ignored extensions do not apply to
369 lists of completions---those always mention all possible completions.
371 If an element of the list in @code{completion-ignored-extensions} ends
372 in a slash @file{/}, it indicates a subdirectory that should be ignored
373 when completing file names. (Elements of
374 @code{completion-ignored-extensions} which do not end in a slash are
375 never considered when a completion candidate is a directory; thus,
376 completion returns directories whose names end in @file{.elc} even
377 though there's an element @code{".elc"} in the list.)
379 @vindex completion-auto-help
380 Normally, a completion command that cannot determine even one
381 additional character automatically displays a list of all possible
382 completions. If the variable @code{completion-auto-help} is set to
383 @code{nil}, this automatic display is disabled, so you must type
384 @kbd{?} to display the list of completions.
386 @cindex Partial Completion mode
387 @vindex partial-completion-mode
388 @findex partial-completion-mode
389 Partial Completion mode implements a more powerful kind of
390 completion that can complete multiple words in parallel. For example,
391 it can complete the command name abbreviation @code{p-b} into
392 @code{print-buffer}, because no other command starts with two words
393 whose initials are @samp{p} and @samp{b}.
395 Partial completion of directories in file names uses @samp{*} to
396 indicate the places for completion; thus, @file{/u*/b*/f*} might
397 complete to @file{/usr/bin/foo}.
399 To enable this mode, use the command @kbd{M-x
400 partial-completion-mode}, or customize the variable
401 @code{partial-completion-mode}. This binds the partial completion
402 commands to @key{TAB}, @key{SPC}, @key{RET}, and @kbd{?}. The usual
403 completion commands are available on @kbd{M-@key{TAB}} (or
404 @kbd{C-M-i}), @kbd{M-@key{SPC}}, @kbd{M-@key{RET}} and @kbd{M-?}.
406 @vindex PC-include-file-path
407 @vindex PC-disable-includes
408 Another feature of Partial Completion mode is to extend
409 @code{find-file} so that the @samp{<@var{include}>} stands for the
410 file named @var{include} in some directory in the path
411 @code{PC-include-file-path}. If you set @code{PC-disable-includes} to
412 non-@code{nil}, this feature is disabled.
414 @cindex Icomplete mode
415 @findex icomplete-mode
416 Icomplete mode presents a constantly-updated display that tells you
417 what completions are available for the text you've entered so far. The
418 command to enable or disable this minor mode is @kbd{M-x
421 @node Minibuffer History
422 @section Minibuffer History
423 @cindex minibuffer history
424 @cindex history of minibuffer input
426 Every argument that you enter with the minibuffer is saved on a
427 @dfn{minibuffer history list} so that you can use it again later in
428 another argument. Special commands load the text of an earlier argument
429 in the minibuffer. They discard the old minibuffer contents, so you can
430 think of them as moving through the history of previous arguments.
435 Move to the next earlier argument string saved in the minibuffer history
436 (@code{previous-history-element}).
439 Move to the next later argument string saved in the minibuffer history
440 (@code{next-history-element}).
441 @item M-r @var{regexp} @key{RET}
442 Move to an earlier saved argument in the minibuffer history that has a
443 match for @var{regexp} (@code{previous-matching-history-element}).
444 @item M-s @var{regexp} @key{RET}
445 Move to a later saved argument in the minibuffer history that has a
446 match for @var{regexp} (@code{next-matching-history-element}).
449 @kindex M-p @r{(minibuffer history)}
450 @kindex M-n @r{(minibuffer history)}
451 @findex next-history-element
452 @findex previous-history-element
453 The simplest way to reuse the saved arguments in the history list is
454 to move through the history list one element at a time. While in the
455 minibuffer, use @kbd{M-p} or up-arrow
456 (@code{previous-history-element}) to ``move to'' the next earlier
457 minibuffer input, and use @kbd{M-n} or down-arrow
458 (@code{next-history-element}) to ``move to'' the next later input.
459 These commands don't move the cursor, they bring different saved
460 strings into the minibuffer. But you can think of them as ``moving''
461 through the history list.
463 The previous input that you fetch from the history entirely replaces
464 the contents of the minibuffer. To use it as the argument, exit the
465 minibuffer as usual with @key{RET}. You can also edit the text before
466 you reuse it; this does not change the history element that you
467 ``moved'' to, but your new argument does go at the end of the history
468 list in its own right.
470 For many minibuffer arguments there is a ``default'' value. In some
471 cases, the minibuffer history commands know the default value. Then you
472 can insert the default value into the minibuffer as text by using
473 @kbd{M-n} to move ``into the future'' in the history. Eventually we
474 hope to make this feature available whenever the minibuffer has a
477 @findex previous-matching-history-element
478 @findex next-matching-history-element
479 @kindex M-r @r{(minibuffer history)}
480 @kindex M-s @r{(minibuffer history)}
481 There are also commands to search forward or backward through the
482 history; they search for history elements that match a regular
483 expression that you specify with the minibuffer. @kbd{M-r}
484 (@code{previous-matching-history-element}) searches older elements in
485 the history, while @kbd{M-s} (@code{next-matching-history-element})
486 searches newer elements. By special dispensation, these commands can
487 use the minibuffer to read their arguments even though you are already
488 in the minibuffer when you issue them. As with incremental searching,
489 an upper-case letter in the regular expression makes the search
490 case-sensitive (@pxref{Search Case}).
493 We may change the precise way these commands read their arguments.
494 Perhaps they will search for a match for the string given so far in the
495 minibuffer; perhaps they will search for a literal match rather than a
496 regular expression match; perhaps they will only accept matches at the
497 beginning of a history element; perhaps they will read the string to
498 search for incrementally like @kbd{C-s}. To find out what interface is
499 actually available, type @kbd{C-h f previous-matching-history-element}.
502 All uses of the minibuffer record your input on a history list, but
503 there are separate history lists for different kinds of arguments. For
504 example, there is a list for file names, used by all the commands that
505 read file names. (As a special feature, this history list records
506 the absolute file name, no more and no less, even if that is not how
507 you entered the file name.)
509 There are several other very specific history lists, including one for
510 command names read by @kbd{M-x}, one for buffer names, one for arguments
511 of commands like @code{query-replace}, and one for compilation commands
512 read by @code{compile}. Finally, there is one ``miscellaneous'' history
513 list that most minibuffer arguments use.
515 @vindex history-length
516 The variable @code{history-length} specifies the maximum length of a
517 minibuffer history list; once a list gets that long, the oldest element
518 is deleted each time an element is added. If the value of
519 @code{history-length} is @code{t}, though, there is no maximum length
520 and elements are never deleted.
522 @vindex history-delete-duplicates
523 The variable @code{history-delete-duplicates} specifies whether to
524 delete duplicates in history. If the value of @code{history-delete-duplicates}
525 is @code{t}, that means when adding a new history element, all
526 previous identical elements are deleted.
529 @section Repeating Minibuffer Commands
530 @cindex command history
531 @cindex history of commands
533 Every command that uses the minibuffer at least once is recorded on a
534 special history list, together with the values of its arguments, so that
535 you can repeat the entire command. In particular, every use of
536 @kbd{M-x} is recorded there, since @kbd{M-x} uses the minibuffer to read
539 @findex list-command-history
541 @item C-x @key{ESC} @key{ESC}
542 Re-execute a recent minibuffer command (@code{repeat-complex-command}).
543 @item M-x list-command-history
544 Display the entire command history, showing all the commands
545 @kbd{C-x @key{ESC} @key{ESC}} can repeat, most recent first.
549 @findex repeat-complex-command
550 @kbd{C-x @key{ESC} @key{ESC}} is used to re-execute a recent
551 minibuffer-using command. With no argument, it repeats the last such
552 command. A numeric argument specifies which command to repeat; one
553 means the last one, and larger numbers specify earlier ones.
555 @kbd{C-x @key{ESC} @key{ESC}} works by turning the previous command
556 into a Lisp expression and then entering a minibuffer initialized with
557 the text for that expression. If you type just @key{RET}, the command
558 is repeated as before. You can also change the command by editing the
559 Lisp expression. Whatever expression you finally submit is what will be
560 executed. The repeated command is added to the front of the command
561 history unless it is identical to the most recently executed command
564 Even if you don't understand Lisp syntax, it will probably be obvious
565 which command is displayed for repetition. If you do not change the
566 text, it will repeat exactly as before.
568 Once inside the minibuffer for @kbd{C-x @key{ESC} @key{ESC}}, you can
569 use the minibuffer history commands (@kbd{M-p}, @kbd{M-n}, @kbd{M-r},
570 @kbd{M-s}; @pxref{Minibuffer History}) to move through the history list
571 of saved entire commands. After finding the desired previous command,
572 you can edit its expression as usual and then resubmit it by typing
575 @vindex isearch-resume-in-command-history
576 Incremental search does not, strictly speaking, use the minibuffer,
577 but it does something similar. Although it behaves like a complex command,
578 it normally does not appear in the history list for @kbd{C-x
579 @key{ESC} @key{ESC}}. You can make it appear in the history by
580 setting @code{isearch-resume-in-command-history} to a non-@code{nil}
581 value. @xref{Incremental Search}.
583 @vindex command-history
584 The list of previous minibuffer-using commands is stored as a Lisp
585 list in the variable @code{command-history}. Each element is a Lisp
586 expression which describes one command and its arguments. Lisp programs
587 can re-execute a command by calling @code{eval} with the
588 @code{command-history} element.
591 arch-tag: ba913cfd-b70e-400f-b663-22b2c309227f