1 This is ../info/xemacs.info, produced by makeinfo version 4.0 from
4 INFO-DIR-SECTION XEmacs Editor
6 * XEmacs: (xemacs). XEmacs Editor.
9 This file documents the XEmacs editor.
11 Copyright (C) 1985, 1986, 1988 Richard M. Stallman. Copyright (C)
12 1991, 1992, 1993, 1994 Lucid, Inc. Copyright (C) 1993, 1994 Sun
13 Microsystems, Inc. Copyright (C) 1995 Amdahl Corporation.
15 Permission is granted to make and distribute verbatim copies of this
16 manual provided the copyright notice and this permission notice are
17 preserved on all copies.
19 Permission is granted to copy and distribute modified versions of
20 this manual under the conditions for verbatim copying, provided also
21 that the sections entitled "The GNU Manifesto", "Distribution" and "GNU
22 General Public License" are included exactly as in the original, and
23 provided that the entire resulting derived work is distributed under the
24 terms of a permission notice identical to this one.
26 Permission is granted to copy and distribute translations of this
27 manual into another language, under the above conditions for modified
28 versions, except that the sections entitled "The GNU Manifesto",
29 "Distribution" and "GNU General Public License" may be included in a
30 translation approved by the author instead of in the original English.
33 File: xemacs.info, Node: Scrolling, Next: Horizontal Scrolling, Prev: Display, Up: Display
38 If a buffer contains text that is too large to fit entirely within
39 the window that is displaying the buffer, XEmacs shows a contiguous
40 section of the text. The section shown always contains point.
42 "Scrolling" means moving text up or down in the window so that
43 different parts of the text are visible. Scrolling forward means that
44 text moves up, and new text appears at the bottom. Scrolling backward
45 moves text down and new text appears at the top.
47 Scrolling happens automatically if you move point past the bottom or
48 top of the window. You can also explicitly request scrolling with the
49 commands in this section.
52 Clear frame and redisplay, scrolling the selected window to center
53 point vertically within it (`recenter').
58 Scroll forward (a windowful or a specified number of lines)
64 Scroll backward (`scroll-down').
67 Scroll so point is on line ARG (`recenter').
69 The most basic scrolling command is `C-l' (`recenter') with no
70 argument. It clears the entire frame and redisplays all windows. In
71 addition, it scrolls the selected window so that point is halfway down
72 from the top of the window.
74 The scrolling commands `C-v' and `M-v' let you move all the text in
75 the window up or down a few lines. `C-v' (`scroll-up') with an
76 argument shows you that many more lines at the bottom of the window,
77 moving the text and point up together as `C-l' might. `C-v' with a
78 negative argument shows you more lines at the top of the window.
79 `Meta-v' (`scroll-down') is like `C-v', but moves in the opposite
82 To read the buffer a windowful at a time, use `C-v' with no
83 argument. `C-v' takes the last two lines at the bottom of the window
84 and puts them at the top, followed by nearly a whole windowful of lines
85 not previously visible. Point moves to the new top of the window if it
86 was in the text scrolled off the top. `M-v' with no argument moves
87 backward with similar overlap. The number of lines of overlap across a
88 `C-v' or `M-v' is controlled by the variable
89 `next-screen-context-lines'; by default, it is two.
91 Another way to scroll is using `C-l' with a numeric argument. `C-l'
92 does not clear the frame when given an argument; it only scrolls the
93 selected window. With a positive argument N, `C-l' repositions text to
94 put point N lines down from the top. An argument of zero puts point on
95 the very top line. Point does not move with respect to the text;
96 rather, the text and point move rigidly on the frame. `C-l' with a
97 negative argument puts point that many lines from the bottom of the
98 window. For example, `C-u - 1 C-l' puts point on the bottom line, and
99 `C-u - 5 C-l' puts it five lines from the bottom. Just `C-u' as
100 argument, as in `C-u C-l', scrolls point to the center of the frame.
102 Scrolling happens automatically if point has moved out of the visible
103 portion of the text when it is time to display. Usually scrolling is
104 done to put point vertically centered within the window. However, if
105 the variable `scroll-step' has a non-zero value, an attempt is made to
106 scroll the buffer by that many lines; if that is enough to bring point
107 back into visibility, that is what happens.
109 Scrolling happens automatically if point has moved out of the visible
110 portion of the text when it is time to display. Usually scrolling is
111 done to put point vertically centered within the window. However, if
112 the variable `scroll-step' has a non-zero value, an attempt is made to
113 scroll the buffer by that many lines; if that is enough to bring point
114 back into visibility, that is what happens.
116 If you set `scroll-step' to a small value because you want to use
117 arrow keys to scroll the screen without recentering, the redisplay
118 preemption will likely make XEmacs keep recentering the screen when
119 scrolling fast, regardless of `scroll-step'. To prevent this, set
120 `scroll-conservatively' to a small value, which will have the result of
121 overriding the redisplay preemption.
124 File: xemacs.info, Node: Horizontal Scrolling, Prev: Scrolling, Up: Display
130 Scroll text in current window to the left (`scroll-left').
133 Scroll to the right (`scroll-right').
135 The text in a window can also be scrolled horizontally. This means
136 that each line of text is shifted sideways in the window, and one or
137 more characters at the beginning of each line are not displayed at all.
138 When a window has been scrolled horizontally in this way, text lines
139 are truncated rather than continued (*note Continuation Lines::), with
140 a `$' appearing in the first column when there is text truncated to the
141 left, and in the last column when there is text truncated to the right.
143 The command `C-x <' (`scroll-left') scrolls the selected window to
144 the left by N columns with argument N. With no argument, it scrolls by
145 almost the full width of the window (two columns less, to be precise).
146 `C-x >' (`scroll-right') scrolls similarly to the right. The window
147 cannot be scrolled any farther to the right once it is displaying
148 normally (with each line starting at the window's left margin);
149 attempting to do so has no effect.
152 File: xemacs.info, Node: Selective Display, Next: Display Vars, Prev: Display, Up: Display
157 XEmacs can hide lines indented more than a certain number of columns
158 (you specify how many columns). This allows you to get an overview of
161 To hide lines, type `C-x $' (`set-selective-display') with a numeric
162 argument N. (*Note Arguments::, for information on giving the
163 argument.) Lines with at least N columns of indentation disappear from
164 the screen. The only indication of their presence are three dots
165 (`...'), which appear at the end of each visible line that is followed
166 by one or more invisible ones.
168 The invisible lines are still present in the buffer, and most editing
169 commands see them as usual, so it is very easy to put point in the
170 middle of invisible text. When this happens, the cursor appears at the
171 end of the previous line, after the three dots. If point is at the end
172 of the visible line, before the newline that ends it, the cursor
173 appears before the three dots.
175 The commands `C-n' and `C-p' move across the invisible lines as if
178 To make everything visible again, type `C-x $' with no argument.
181 File: xemacs.info, Node: Display Vars, Prev: Selective Display, Up: Display
183 Variables Controlling Display
184 =============================
186 This section contains information for customization only. Beginning
187 users should skip it.
189 When you reenter XEmacs after suspending, XEmacs normally clears the
190 screen and redraws the entire display. On some terminals with more than
191 one page of memory, it is possible to arrange the termcap entry so that
192 the `ti' and `te' strings (output to the terminal when XEmacs is
193 entered and exited, respectively) switch between pages of memory so as
194 to use one page for XEmacs and another page for other output. In that
195 case, you might want to set the variable `no-redraw-on-reenter' to
196 non-`nil' so that XEmacs will assume, when resumed, that the screen
197 page it is using still contains what XEmacs last wrote there.
199 The variable `echo-keystrokes' controls the echoing of
200 multi-character keys; its value is the number of seconds of pause
201 required to cause echoing to start, or zero, meaning don't echo at all.
204 If the variable `ctl-arrow' is `nil', control characters in the
205 buffer are displayed with octal escape sequences, all except newline and
206 tab. If its value is `t', then control characters will be printed with
207 an up-arrow, for example `^A'.
209 If its value is not `t' and not `nil', then characters whose code is
210 greater than 160 (that is, the space character (32) with its high bit
211 set) will be assumed to be printable, and will be displayed without
212 alteration. This is the default when running under X Windows, since
213 XEmacs assumes an ISO/8859-1 character set (also known as "Latin1").
214 The `ctl-arrow' variable may also be set to an integer, in which case
215 all characters whose codes are greater than or equal to that value will
216 be assumed to be printable.
218 Altering the value of `ctl-arrow' makes it local to the current
219 buffer; until that time, the default value is in effect. *Note
222 Normally, a tab character in the buffer is displayed as whitespace
223 which extends to the next display tab stop position, and display tab
224 stops come at intervals equal to eight spaces. The number of spaces
225 per tab is controlled by the variable `tab-width', which is made local
226 by changing it, just like `ctl-arrow'. Note that how the tab character
227 in the buffer is displayed has nothing to do with the definition of
230 If you set the variable `selective-display-ellipses' to `nil', the
231 three dots at the end of a line that precedes invisible lines do not
232 appear. There is no visible indication of the invisible lines. This
233 variable becomes local automatically when set.
236 File: xemacs.info, Node: Search, Next: Fixit, Prev: Display, Up: Top
238 Searching and Replacement
239 *************************
241 Like other editors, Emacs has commands for searching for occurrences
242 of a string. The principal search command is unusual in that it is
243 "incremental": it begins to search before you have finished typing the
244 search string. There are also non-incremental search commands more like
245 those of other editors.
247 Besides the usual `replace-string' command that finds all
248 occurrences of one string and replaces them with another, Emacs has a
249 fancy replacement command called `query-replace' which asks
250 interactively which occurrences to replace.
254 * Incremental Search:: Search happens as you type the string.
255 * Non-Incremental Search:: Specify entire string and then search.
256 * Word Search:: Search for sequence of words.
257 * Regexp Search:: Search for match for a regexp.
258 * Regexps:: Syntax of regular expressions.
259 * Search Case:: To ignore case while searching, or not.
260 * Replace:: Search, and replace some or all matches.
261 * Other Repeating Search:: Operating on all matches for some regexp.
264 File: xemacs.info, Node: Incremental Search, Next: Non-Incremental Search, Prev: Search, Up: Search
269 An incremental search begins searching as soon as you type the first
270 character of the search string. As you type in the search string, Emacs
271 shows you where the string (as you have typed it so far) is found.
272 When you have typed enough characters to identify the place you want,
273 you can stop. Depending on what you do next, you may or may not need to
274 terminate the search explicitly with a <RET>.
277 Incremental search forward (`isearch-forward').
280 Incremental search backward (`isearch-backward').
282 `C-s' starts an incremental search. `C-s' reads characters from the
283 keyboard and positions the cursor at the first occurrence of the
284 characters that you have typed. If you type `C-s' and then `F', the
285 cursor moves right after the first `F'. Type an `O', and see the
286 cursor move to after the first `FO'. After another `O', the cursor is
287 after the first `FOO' after the place where you started the search.
288 Meanwhile, the search string `FOO' has been echoed in the echo area.
290 The echo area display ends with three dots when actual searching is
291 going on. When search is waiting for more input, the three dots are
292 removed. (On slow terminals, the three dots are not displayed.)
294 If you make a mistake in typing the search string, you can erase
295 characters with <DEL>. Each <DEL> cancels the last character of the
296 search string. This does not happen until Emacs is ready to read
297 another input character; first it must either find, or fail to find,
298 the character you want to erase. If you do not want to wait for this
299 to happen, use `C-g' as described below.
301 When you are satisfied with the place you have reached, you can type
302 <RET> (or <C-m>), which stops searching, leaving the cursor where the
303 search brought it. Any command not specially meaningful in searches
304 also stops the search and is then executed. Thus, typing `C-a' exits
305 the search and then moves to the beginning of the line. <RET> is
306 necessary only if the next command you want to type is a printing
307 character, <DEL>, <ESC>, or another control character that is special
308 within searches (`C-q', `C-w', `C-r', `C-s', or `C-y').
310 Sometimes you search for `FOO' and find it, but were actually
311 looking for a different occurrence of it. To move to the next
312 occurrence of the search string, type another `C-s'. Do this as often
313 as necessary. If you overshoot, you can cancel some `C-s' characters
316 After you exit a search, you can search for the same string again by
317 typing just `C-s C-s': the first `C-s' is the key that invokes
318 incremental search, and the second `C-s' means "search again".
320 If the specified string is not found at all, the echo area displays
321 the text `Failing I-Search'. The cursor is after the place where Emacs
322 found as much of your string as it could. Thus, if you search for
323 `FOOT', and there is no `FOOT', the cursor may be after the `FOO' in
324 `FOOL'. At this point there are several things you can do. If you
325 mistyped the search string, correct it. If you like the place you have
326 found, you can type <RET> or some other Emacs command to "accept what
327 the search offered". Or you can type `C-g', which removes from the
328 search string the characters that could not be found (the `T' in
329 `FOOT'), leaving those that were found (the `FOO' in `FOOT'). A second
330 `C-g' at that point cancels the search entirely, returning point to
331 where it was when the search started.
333 If a search is failing and you ask to repeat it by typing another
334 `C-s', it starts again from the beginning of the buffer. Repeating a
335 failing backward search with `C-r' starts again from the end. This is
336 called "wrapping around". `Wrapped' appears in the search prompt once
339 The `C-g' "quit" character does special things during searches; just
340 what it does depends on the status of the search. If the search has
341 found what you specified and is waiting for input, `C-g' cancels the
342 entire search. The cursor moves back to where you started the search.
343 If `C-g' is typed when there are characters in the search string that
344 have not been found--because Emacs is still searching for them, or
345 because it has failed to find them--then the search string characters
346 which have not been found are discarded from the search string. The
347 search is now successful and waiting for more input, so a second `C-g'
348 cancels the entire search.
350 To search for a control character such as `C-s' or <DEL> or <ESC>,
351 you must quote it by typing `C-q' first. This function of `C-q' is
352 analogous to its meaning as an Emacs command: it causes the following
353 character to be treated the way a graphic character would normally be
354 treated in the same context.
356 To search backwards, you can use `C-r' instead of `C-s' to start the
357 search; `C-r' is the key that runs the command (`isearch-backward') to
358 search backward. You can also use `C-r' to change from searching
359 forward to searching backwards. Do this if a search fails because the
360 place you started was too far down in the file. Repeated `C-r' keeps
361 looking for more occurrences backwards. `C-s' starts going forward
362 again. You can cancel `C-r' in a search with <DEL>.
364 The characters `C-y' and `C-w' can be used in incremental search to
365 grab text from the buffer into the search string. This makes it
366 convenient to search for another occurrence of text at point. `C-w'
367 copies the word after point as part of the search string, advancing
368 point over that word. Another `C-s' to repeat the search will then
369 search for a string including that word. `C-y' is similar to `C-w' but
370 copies the rest of the current line into the search string.
372 The characters `M-p' and `M-n' can be used in an incremental search
373 to recall things which you have searched for in the past. A list of
374 the last 16 things you have searched for is retained, and `M-p' and
375 `M-n' let you cycle through that ring.
377 The character `M-<TAB>' does completion on the elements in the
378 search history ring. For example, if you know that you have recently
379 searched for the string `POTATOE', you could type `C-s P O M-<TAB>'.
380 If you had searched for other strings beginning with `PO' then you
381 would be shown a list of them, and would need to type more to select
384 You can change any of the special characters in incremental search
385 via the normal keybinding mechanism: simply add a binding to the
386 `isearch-mode-map'. For example, to make the character `C-b' mean
387 "search backwards" while in isearch-mode, do this:
389 (define-key isearch-mode-map "\C-b" 'isearch-repeat-backward)
391 These are the default bindings of isearch-mode:
394 Delete a character from the incremental search string
395 (`isearch-delete-char').
398 Exit incremental search (`isearch-exit').
401 Quote special characters for incremental search
402 (`isearch-quote-char').
405 Repeat incremental search forward (`isearch-repeat-forward').
408 Repeat incremental search backward (`isearch-repeat-backward').
411 Pull rest of line from buffer into search string
412 (`isearch-yank-line').
415 Pull next word from buffer into search string
416 (`isearch-yank-word').
419 Cancels input back to what has been found successfully, or aborts
420 the isearch (`isearch-abort').
423 Recall the previous element in the isearch history ring
424 (`isearch-ring-retreat').
427 Recall the next element in the isearch history ring
428 (`isearch-ring-advance').
431 Do completion on the elements in the isearch history ring
432 (`isearch-complete').
434 Any other character which is normally inserted into a buffer when
435 typed is automatically added to the search string in isearch-mode.
437 Slow Terminal Incremental Search
438 --------------------------------
440 Incremental search on a slow terminal uses a modified style of
441 display that is designed to take less time. Instead of redisplaying
442 the buffer at each place the search gets to, it creates a new
443 single-line window and uses that to display the line the search has
444 found. The single-line window appears as soon as point gets outside of
445 the text that is already on the screen.
447 When the search is terminated, the single-line window is removed.
448 Only at this time the window in which the search was done is
449 redisplayed to show its new value of point.
451 The three dots at the end of the search string, normally used to
452 indicate that searching is going on, are not displayed in slow style
455 The slow terminal style of display is used when the terminal baud
456 rate is less than or equal to the value of the variable
457 `search-slow-speed', initially 1200.
459 The number of lines to use in slow terminal search display is
460 controlled by the variable `search-slow-window-lines'. Its normal
464 File: xemacs.info, Node: Non-Incremental Search, Next: Word Search, Prev: Incremental Search, Up: Search
466 Non-Incremental Search
467 ======================
469 Emacs also has conventional non-incremental search commands, which
470 require you type the entire search string before searching begins.
472 `C-s <RET> STRING <RET>'
475 `C-r <RET> STRING <RET>'
476 Search backward for STRING.
478 To do a non-incremental search, first type `C-s <RET>' (or `C-s
479 C-m'). This enters the minibuffer to read the search string.
480 Terminate the string with <RET> to start the search. If the string is
481 not found, the search command gets an error.
483 By default, `C-s' invokes incremental search, but if you give it an
484 empty argument, which would otherwise be useless, it invokes
485 non-incremental search. Therefore, `C-s <RET>' invokes non-incremental
486 search. `C-r <RET>' also works this way.
488 Forward and backward non-incremental searches are implemented by the
489 commands `search-forward' and `search-backward'. You can bind these
490 commands to keys. The reason that incremental search is programmed to
491 invoke them as well is that `C-s <RET>' is the traditional sequence of
492 characters used in Emacs to invoke non-incremental search.
494 Non-incremental searches performed using `C-s <RET>' do not call
495 `search-forward' right away. They first check if the next character is
496 `C-w', which requests a word search. *Note Word Search::.
499 File: xemacs.info, Node: Word Search, Next: Regexp Search, Prev: Non-Incremental Search, Up: Search
504 Word search looks for a sequence of words without regard to how the
505 words are separated. More precisely, you type a string of many words,
506 using single spaces to separate them, and the string is found even if
507 there are multiple spaces, newlines or other punctuation between the
510 Word search is useful in editing documents formatted by text
511 formatters. If you edit while looking at the printed, formatted
512 version, you can't tell where the line breaks are in the source file.
513 Word search, allows you to search without having to know the line
516 `C-s <RET> C-w WORDS <RET>'
517 Search for WORDS, ignoring differences in punctuation.
519 `C-r <RET> C-w WORDS <RET>'
520 Search backward for WORDS, ignoring differences in punctuation.
522 Word search is a special case of non-incremental search. It is
523 invoked with `C-s <RET> C-w' followed by the search string, which must
524 always be terminated with another <RET>. Being non-incremental, this
525 search does not start until the argument is terminated. It works by
526 constructing a regular expression and searching for that. *Note Regexp
529 You can do a backward word search with `C-r <RET> C-w'.
531 Forward and backward word searches are implemented by the commands
532 `word-search-forward' and `word-search-backward'. You can bind these
533 commands to keys. The reason that incremental search is programmed to
534 invoke them as well is that `C-s <RET> C-w' is the traditional Emacs
535 sequence of keys for word search.
538 File: xemacs.info, Node: Regexp Search, Next: Regexps, Prev: Word Search, Up: Search
540 Regular Expression Search
541 =========================
543 A "regular expression" ("regexp", for short) is a pattern that
544 denotes a (possibly infinite) set of strings. Searching for matches
545 for a regexp is a powerful operation that editors on Unix systems have
546 traditionally offered.
548 To gain a thorough understanding of regular expressions and how to
549 use them to best advantage, we recommend that you study `Mastering
550 Regular Expressions, by Jeffrey E.F. Friedl, O'Reilly and Associates,
551 1997'. (It's known as the "Hip Owls" book, because of the picture on its
552 cover.) You might also read the manuals to *Note (gawk)Top::, *Note
553 (ed)Top::, `sed', `grep', *Note (perl)Top::, *Note (regex)Top::, *Note
554 (rx)Top::, `pcre', and *Note (flex)Top::, which also make good use of
557 The XEmacs regular expression syntax most closely resembles that of
558 `ed', or `grep', the GNU versions of which all utilize the GNU `regex'
559 library. XEmacs' version of `regex' has recently been extended with
560 some Perl-like capabilities, described in the next section.
562 In XEmacs, you can search for the next match for a regexp either
563 incrementally or not.
565 Incremental search for a regexp is done by typing `M-C-s'
566 (`isearch-forward-regexp'). This command reads a search string
567 incrementally just like `C-s', but it treats the search string as a
568 regexp rather than looking for an exact match against the text in the
569 buffer. Each time you add text to the search string, you make the
570 regexp longer, and the new regexp is searched for. A reverse regexp
571 search command `isearch-backward-regexp' also exists, bound to `M-C-r'.
573 All of the control characters that do special things within an
574 ordinary incremental search have the same functionality in incremental
575 regexp search. Typing `C-s' or `C-r' immediately after starting a
576 search retrieves the last incremental search regexp used: incremental
577 regexp and non-regexp searches have independent defaults.
579 Non-incremental search for a regexp is done by the functions
580 `re-search-forward' and `re-search-backward'. You can invoke them with
581 `M-x' or bind them to keys. You can also call `re-search-forward' by
582 way of incremental regexp search with `M-C-s <RET>'; similarly for
583 `re-search-backward' with `M-C-r <RET>'.
586 File: xemacs.info, Node: Regexps, Next: Search Case, Prev: Regexp Search, Up: Search
588 Syntax of Regular Expressions
589 =============================
591 Regular expressions have a syntax in which a few characters are
592 special constructs and the rest are "ordinary". An ordinary character
593 is a simple regular expression that matches that character and nothing
594 else. The special characters are `.', `*', `+', `?', `[', `]', `^',
595 `$', and `\'; no new special characters will be defined in the future.
596 Any other character appearing in a regular expression is ordinary,
597 unless a `\' precedes it.
599 For example, `f' is not a special character, so it is ordinary, and
600 therefore `f' is a regular expression that matches the string `f' and
601 no other string. (It does _not_ match the string `ff'.) Likewise, `o'
602 is a regular expression that matches only `o'.
604 Any two regular expressions A and B can be concatenated. The result
605 is a regular expression that matches a string if A matches some amount
606 of the beginning of that string and B matches the rest of the string.
608 As a simple example, we can concatenate the regular expressions `f'
609 and `o' to get the regular expression `fo', which matches only the
610 string `fo'. Still trivial. To do something more powerful, you need
611 to use one of the special characters. Here is a list of them:
614 is a special character that matches any single character except a
615 newline. Using concatenation, we can make regular expressions
616 like `a.b', which matches any three-character string that begins
617 with `a' and ends with `b'.
620 is not a construct by itself; it is a quantifying suffix operator
621 that means to repeat the preceding regular expression as many
622 times as possible. In `fo*', the `*' applies to the `o', so `fo*'
623 matches one `f' followed by any number of `o's. The case of zero
624 `o's is allowed: `fo*' does match `f'.
626 `*' always applies to the _smallest_ possible preceding
627 expression. Thus, `fo*' has a repeating `o', not a repeating `fo'.
629 The matcher processes a `*' construct by matching, immediately, as
630 many repetitions as can be found; it is "greedy". Then it
631 continues with the rest of the pattern. If that fails,
632 backtracking occurs, discarding some of the matches of the
633 `*'-modified construct in case that makes it possible to match the
634 rest of the pattern. For example, in matching `ca*ar' against the
635 string `caaar', the `a*' first tries to match all three `a's; but
636 the rest of the pattern is `ar' and there is only `r' left to
637 match, so this try fails. The next alternative is for `a*' to
638 match only two `a's. With this choice, the rest of the regexp
639 matches successfully.
641 Nested repetition operators can be extremely slow if they specify
642 backtracking loops. For example, it could take hours for the
643 regular expression `\(x+y*\)*a' to match the sequence
644 `xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxz'. The slowness is because
645 Emacs must try each imaginable way of grouping the 35 `x''s before
646 concluding that none of them can work. To make sure your regular
647 expressions run fast, check nested repetitions carefully.
650 is a quantifying suffix operator similar to `*' except that the
651 preceding expression must match at least once. It is also
652 "greedy". So, for example, `ca+r' matches the strings `car' and
653 `caaaar' but not the string `cr', whereas `ca*r' matches all three
657 is a quantifying suffix operator similar to `*', except that the
658 preceding expression can match either once or not at all. For
659 example, `ca?r' matches `car' or `cr', but does not match anything
663 works just like `*', except that rather than matching the longest
664 match, it matches the shortest match. `*?' is known as a
665 "non-greedy" quantifier, a regexp construct borrowed from Perl.
667 This construct is very useful for when you want to match the text
668 inside a pair of delimiters. For instance, `/\*.*?\*/' will match
669 C comments in a string. This could not easily be achieved without
670 the use of a non-greedy quantifier.
672 This construct has not been available prior to XEmacs 20.4. It is
673 not available in FSF Emacs.
676 is the non-greedy version of `+'.
679 is the non-greedy version of `?'.
682 serves as an interval quantifier, analogous to `*' or `+', but
683 specifies that the expression must match at least N times, but no
684 more than M times. This syntax is supported by most Unix regexp
685 utilities, and has been introduced to XEmacs for the version 20.3.
687 Unfortunately, the non-greedy version of this quantifier does not
688 exist currently, although it does in Perl.
691 `[' begins a "character set", which is terminated by a `]'. In
692 the simplest case, the characters between the two brackets form
693 the set. Thus, `[ad]' matches either one `a' or one `d', and
694 `[ad]*' matches any string composed of just `a's and `d's
695 (including the empty string), from which it follows that `c[ad]*r'
696 matches `cr', `car', `cdr', `caddaar', etc.
698 The usual regular expression special characters are not special
699 inside a character set. A completely different set of special
700 characters exists inside character sets: `]', `-' and `^'.
702 `-' is used for ranges of characters. To write a range, write two
703 characters with a `-' between them. Thus, `[a-z]' matches any
704 lower case letter. Ranges may be intermixed freely with individual
705 characters, as in `[a-z$%.]', which matches any lower case letter
706 or `$', `%', or a period.
708 To include a `]' in a character set, make it the first character.
709 For example, `[]a]' matches `]' or `a'. To include a `-', write
710 `-' as the first character in the set, or put it immediately after
711 a range. (You can replace one individual character C with the
712 range `C-C' to make a place to put the `-'.) There is no way to
713 write a set containing just `-' and `]'.
715 To include `^' in a set, put it anywhere but at the beginning of
719 `[^' begins a "complement character set", which matches any
720 character except the ones specified. Thus, `[^a-z0-9A-Z]' matches
721 all characters _except_ letters and digits.
723 `^' is not special in a character set unless it is the first
724 character. The character following the `^' is treated as if it
725 were first (thus, `-' and `]' are not special there).
727 Note that a complement character set can match a newline, unless
728 newline is mentioned as one of the characters not to match.
731 is a special character that matches the empty string, but only at
732 the beginning of a line in the text being matched. Otherwise it
733 fails to match anything. Thus, `^foo' matches a `foo' that occurs
734 at the beginning of a line.
736 When matching a string instead of a buffer, `^' matches at the
737 beginning of the string or after a newline character `\n'.
740 is similar to `^' but matches only at the end of a line. Thus,
741 `x+$' matches a string of one `x' or more at the end of a line.
743 When matching a string instead of a buffer, `$' matches at the end
744 of the string or before a newline character `\n'.
747 has two functions: it quotes the special characters (including
748 `\'), and it introduces additional special constructs.
750 Because `\' quotes special characters, `\$' is a regular
751 expression that matches only `$', and `\[' is a regular expression
752 that matches only `[', and so on.
754 *Please note:* For historical compatibility, special characters are
755 treated as ordinary ones if they are in contexts where their special
756 meanings make no sense. For example, `*foo' treats `*' as ordinary
757 since there is no preceding expression on which the `*' can act. It is
758 poor practice to depend on this behavior; quote the special character
759 anyway, regardless of where it appears.
761 For the most part, `\' followed by any character matches only that
762 character. However, there are several exceptions: characters that,
763 when preceded by `\', are special constructs. Such characters are
764 always ordinary when encountered on their own. Here is a table of `\'
768 specifies an alternative. Two regular expressions A and B with
769 `\|' in between form an expression that matches anything that
770 either A or B matches.
772 Thus, `foo\|bar' matches either `foo' or `bar' but no other string.
774 `\|' applies to the largest possible surrounding expressions.
775 Only a surrounding `\( ... \)' grouping can limit the grouping
778 Full backtracking capability exists to handle multiple uses of
782 is a grouping construct that serves three purposes:
784 1. To enclose a set of `\|' alternatives for other operations.
785 Thus, `\(foo\|bar\)x' matches either `foox' or `barx'.
787 2. To enclose an expression for a suffix operator such as `*' to
788 act on. Thus, `ba\(na\)*' matches `bananana', etc., with any
789 (zero or more) number of `na' strings.
791 3. To record a matched substring for future reference.
793 This last application is not a consequence of the idea of a
794 parenthetical grouping; it is a separate feature that happens to be
795 assigned as a second meaning to the same `\( ... \)' construct
796 because there is no conflict in practice between the two meanings.
797 Here is an explanation of this feature:
800 matches the same text that matched the DIGITth occurrence of a `\(
803 In other words, after the end of a `\( ... \)' construct. the
804 matcher remembers the beginning and end of the text matched by that
805 construct. Then, later on in the regular expression, you can use
806 `\' followed by DIGIT to match that same text, whatever it may
809 The strings matching the first nine `\( ... \)' constructs
810 appearing in a regular expression are assigned numbers 1 through 9
811 in the order that the open parentheses appear in the regular
812 expression. So you can use `\1' through `\9' to refer to the text
813 matched by the corresponding `\( ... \)' constructs.
815 For example, `\(.*\)\1' matches any newline-free string that is
816 composed of two identical halves. The `\(.*\)' matches the first
817 half, which may be anything, but the `\1' that follows must match
821 is called a "shy" grouping operator, and it is used just like `\(
822 ... \)', except that it does not cause the matched substring to be
823 recorded for future reference.
825 This is useful when you need a lot of grouping `\( ... \)'
826 constructs, but only want to remember one or two - or if you have
827 more than nine groupings and need to use backreferences to refer to
828 the groupings at the end.
830 Using `\(?: ... \)' rather than `\( ... \)' when you don't need
831 the captured substrings ought to speed up your programs some,
832 since it shortens the code path followed by the regular expression
833 engine, as well as the amount of memory allocation and string
834 copying it must do. The actual performance gain to be observed
835 has not been measured or quantified as of this writing.
837 The shy grouping operator has been borrowed from Perl, and has not
838 been available prior to XEmacs 20.3, nor is it available in FSF
842 matches any word-constituent character. The editor syntax table
843 determines which characters these are. *Note Syntax::.
846 matches any character that is not a word constituent.
849 matches any character whose syntax is CODE. Here CODE is a
850 character that represents a syntax code: thus, `w' for word
851 constituent, `-' for whitespace, `(' for open parenthesis, etc.
852 *Note Syntax::, for a list of syntax codes and the characters that
856 matches any character whose syntax is not CODE.
858 The following regular expression constructs match the empty
859 string--that is, they don't use up any characters--but whether they
860 match depends on the context.
863 matches the empty string, but only at the beginning of the buffer
864 or string being matched against.
867 matches the empty string, but only at the end of the buffer or
868 string being matched against.
871 matches the empty string, but only at point. (This construct is
872 not defined when matching against a string.)
875 matches the empty string, but only at the beginning or end of a
876 word. Thus, `\bfoo\b' matches any occurrence of `foo' as a
877 separate word. `\bballs?\b' matches `ball' or `balls' as a
881 matches the empty string, but _not_ at the beginning or end of a
885 matches the empty string, but only at the beginning of a word.
888 matches the empty string, but only at the end of a word.
890 Here is a complicated regexp used by Emacs to recognize the end of a
891 sentence together with any whitespace that follows. It is given in Lisp
892 syntax to enable you to distinguish the spaces from the tab characters.
893 In Lisp syntax, the string constant begins and ends with a
894 double-quote. `\"' stands for a double-quote as part of the regexp,
895 `\\' for a backslash as part of the regexp, `\t' for a tab and `\n' for
898 "[.?!][]\"')]*\\($\\|\t\\| \\)[ \t\n]*"
900 This regexp contains four parts: a character set matching period, `?'
901 or `!'; a character set matching close-brackets, quotes or parentheses,
902 repeated any number of times; an alternative in backslash-parentheses
903 that matches end-of-line, a tab or two spaces; and a character set
904 matching whitespace characters, repeated any number of times.
907 File: xemacs.info, Node: Search Case, Next: Replace, Prev: Regexps, Up: Search
912 All searches in Emacs normally ignore the case of the text they are
913 searching through; if you specify searching for `FOO', `Foo' and `foo'
914 are also considered a match. Regexps, and in particular character
915 sets, are included: `[aB]' matches `a' or `A' or `b' or `B'.
917 If you want a case-sensitive search, set the variable
918 `case-fold-search' to `nil'. Then all letters must match exactly,
919 including case. `case-fold-search' is a per-buffer variable; altering
920 it affects only the current buffer, but there is a default value which
921 you can change as well. *Note Locals::. You can also use Case
922 Sensitive Search from the Options menu on your screen.
925 File: xemacs.info, Node: Replace, Next: Other Repeating Search, Prev: Search Case, Up: Search
930 Global search-and-replace operations are not needed as often in
931 Emacs as they are in other editors, but they are available. In
932 addition to the simple `replace-string' command which is like that
933 found in most editors, there is a `query-replace' command which asks
934 you, for each occurrence of a pattern, whether to replace it.
936 The replace commands all replace one string (or regexp) with one
937 replacement string. It is possible to perform several replacements in
938 parallel using the command `expand-region-abbrevs'. *Note Expanding
943 * Unconditional Replace:: Replacing all matches for a string.
944 * Regexp Replace:: Replacing all matches for a regexp.
945 * Replacement and Case:: How replacements preserve case of letters.
946 * Query Replace:: How to use querying.
949 File: xemacs.info, Node: Unconditional Replace, Next: Regexp Replace, Prev: Replace, Up: Replace
951 Unconditional Replacement
952 -------------------------
954 `M-x replace-string <RET> STRING <RET> NEWSTRING <RET>'
955 Replace every occurrence of STRING with NEWSTRING.
957 `M-x replace-regexp <RET> REGEXP <RET> NEWSTRING <RET>'
958 Replace every match for REGEXP with NEWSTRING.
960 To replace every instance of `foo' after point with `bar', use the
961 command `M-x replace-string' with the two arguments `foo' and `bar'.
962 Replacement occurs only after point: if you want to cover the whole
963 buffer you must go to the beginning first. By default, all occurrences
964 up to the end of the buffer are replaced. To limit replacement to part
965 of the buffer, narrow to that part of the buffer before doing the
966 replacement (*note Narrowing::).
968 When `replace-string' exits, point is left at the last occurrence
969 replaced. The value of point when the `replace-string' command was
970 issued is remembered on the mark ring; `C-u C-<SPC>' moves back there.
972 A numeric argument restricts replacement to matches that are
973 surrounded by word boundaries.
976 File: xemacs.info, Node: Regexp Replace, Next: Replacement and Case, Prev: Unconditional Replace, Up: Replace
981 `replace-string' replaces exact matches for a single string. The
982 similar command `replace-regexp' replaces any match for a specified
985 In `replace-regexp', the NEWSTRING need not be constant. It can
986 refer to all or part of what is matched by the REGEXP. `\&' in
987 NEWSTRING stands for the entire text being replaced. `\D' in
988 NEWSTRING, where D is a digit, stands for whatever matched the D'th
989 parenthesized grouping in REGEXP. For example,
991 M-x replace-regexp <RET> c[ad]+r <RET> \&-safe <RET>
993 would replace (for example) `cadr' with `cadr-safe' and `cddr' with
996 M-x replace-regexp <RET> \(c[ad]+r\)-safe <RET> \1 <RET>
998 would perform exactly the opposite replacements. To include a `\' in
999 the text to replace with, you must give `\\'.
1002 File: xemacs.info, Node: Replacement and Case, Next: Query Replace, Prev: Regexp Replace, Up: Replace
1004 Replace Commands and Case
1005 -------------------------
1007 If the arguments to a replace command are in lower case, the command
1008 preserves case when it makes a replacement. Thus, the following
1011 M-x replace-string <RET> foo <RET> bar <RET>
1013 replaces a lower-case `foo' with a lower case `bar', `FOO' with `BAR',
1014 and `Foo' with `Bar'. If upper-case letters are used in the second
1015 argument, they remain upper-case every time that argument is inserted.
1016 If upper-case letters are used in the first argument, the second
1017 argument is always substituted exactly as given, with no case
1018 conversion. Likewise, if the variable `case-replace' is set to `nil',
1019 replacement is done without case conversion. If `case-fold-search' is
1020 set to `nil', case is significant in matching occurrences of `foo' to
1021 replace; also, case conversion of the replacement string is not done.
1024 File: xemacs.info, Node: Query Replace, Prev: Replacement and Case, Up: Replace
1029 `M-% STRING <RET> NEWSTRING <RET>'
1030 `M-x query-replace <RET> STRING <RET> NEWSTRING <RET>'
1031 Replace some occurrences of STRING with NEWSTRING.
1033 `M-x query-replace-regexp <RET> REGEXP <RET> NEWSTRING <RET>'
1034 Replace some matches for REGEXP with NEWSTRING.
1036 If you want to change only some of the occurrences of `foo' to
1037 `bar', not all of them, you can use `query-replace' instead of `M-%'.
1038 This command finds occurrences of `foo' one by one, displays each
1039 occurrence, and asks you whether to replace it. A numeric argument to
1040 `query-replace' tells it to consider only occurrences that are bounded
1041 by word-delimiter characters.
1043 Aside from querying, `query-replace' works just like
1044 `replace-string', and `query-replace-regexp' works just like
1047 The things you can type when you are shown an occurrence of STRING
1048 or a match for REGEXP are:
1051 to replace the occurrence with NEWSTRING. This preserves case,
1052 just like `replace-string', provided `case-replace' is non-`nil',
1056 to skip to the next occurrence without replacing this one.
1059 to replace this occurrence and display the result. You are then
1060 prompted for another input character. However, since the
1061 replacement has already been made, <DEL> and <SPC> are equivalent.
1062 At this point, you can type `C-r' (see below) to alter the
1063 replaced text. To undo the replacement, you can type `C-x u'.
1064 This exits the `query-replace'. If you want to do further
1065 replacement you must use `C-x ESC' to restart (*note Repetition::).
1068 to exit without doing any more replacements.
1071 to replace this occurrence and then exit.
1074 to replace all remaining occurrences without asking again.
1077 to go back to the location of the previous occurrence (or what
1078 used to be an occurrence), in case you changed it by mistake.
1079 This works by popping the mark ring. Only one `^' in a row is
1080 allowed, because only one previous replacement location is kept
1081 during `query-replace'.
1084 to enter a recursive editing level, in case the occurrence needs
1085 to be edited rather than just replaced with NEWSTRING. When you
1086 are done, exit the recursive editing level with `C-M-c' and the
1087 next occurrence will be displayed. *Note Recursive Edit::.
1090 to delete the occurrence, and then enter a recursive editing level
1091 as in `C-r'. Use the recursive edit to insert text to replace the
1092 deleted occurrence of STRING. When done, exit the recursive
1093 editing level with `C-M-c' and the next occurrence will be
1097 to redisplay the screen and then give another answer.
1100 to display a message summarizing these options, then give another
1103 If you type any other character, Emacs exits the `query-replace', and
1104 executes the character as a command. To restart the `query-replace',
1105 use `C-x <ESC>', which repeats the `query-replace' because it used the
1106 minibuffer to read its arguments. *Note C-x ESC: Repetition.
1109 File: xemacs.info, Node: Other Repeating Search, Prev: Replace, Up: Search
1111 Other Search-and-Loop Commands
1112 ==============================
1114 Here are some other commands that find matches for a regular
1115 expression. They all operate from point to the end of the buffer.
1118 Print each line that follows point and contains a match for the
1119 specified regexp. A numeric argument specifies the number of
1120 context lines to print before and after each matching line; the
1123 The buffer `*Occur*' containing the output serves as a menu for
1124 finding occurrences in their original context. Find an occurrence
1125 as listed in `*Occur*', position point there, and type `C-c C-c';
1126 this switches to the buffer that was searched and moves point to
1127 the original of the same occurrence.
1129 `M-x list-matching-lines'
1130 Synonym for `M-x occur'.
1133 Print the number of matches following point for the specified
1136 `M-x delete-non-matching-lines'
1137 Delete each line that follows point and does not contain a match
1138 for the specified regexp.
1140 `M-x delete-matching-lines'
1141 Delete each line that follows point and contains a match for the
1145 File: xemacs.info, Node: Fixit, Next: Files, Prev: Search, Up: Top
1147 Commands for Fixing Typos
1148 *************************
1150 This chapter describes commands that are especially useful when you
1151 catch a mistake in your text just after you have made it, or when you
1152 change your mind while composing text on line.
1156 * Kill Errors:: Commands to kill a batch of recently entered text.
1157 * Transpose:: Exchanging two characters, words, lines, lists...
1158 * Fixing Case:: Correcting case of last word entered.
1159 * Spelling:: Apply spelling checker to a word, or a whole file.
1162 File: xemacs.info, Node: Kill Errors, Next: Transpose, Prev: Fixit, Up: Fixit
1164 Killing Your Mistakes
1165 =====================
1168 Delete last character (`delete-backward-char').
1171 Kill last word (`backward-kill-word').
1174 Kill to beginning of sentence (`backward-kill-sentence').
1176 The <DEL> character (`delete-backward-char') is the most important
1177 correction command. When used among graphic (self-inserting)
1178 characters, it can be thought of as canceling the last character typed.
1180 When your mistake is longer than a couple of characters, it might be
1181 more convenient to use `M-<DEL>' or `C-x <DEL>'. `M-<DEL>' kills back
1182 to the start of the last word, and `C-x <DEL>' kills back to the start
1183 of the last sentence. `C-x <DEL>' is particularly useful when you are
1184 thinking of what to write as you type it, in case you change your mind
1185 about phrasing. `M-<DEL>' and `C-x <DEL>' save the killed text for
1186 `C-y' and `M-y' to retrieve. *Note Yanking::.
1188 `M-<DEL>' is often useful even when you have typed only a few
1189 characters wrong, if you know you are confused in your typing and aren't
1190 sure exactly what you typed. At such a time, you cannot correct with
1191 <DEL> except by looking at the screen to see what you did. It requires
1192 less thought to kill the whole word and start over.