Mercurial > hg > xemacs-beta

comparison lisp/prim/sort.el @ 0:376386a54a3c r19-14

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Import from CVS: tag r19-14
author cvs
date Mon, 13 Aug 2007 08:45:50 +0200
parents
children b82b59fe008d
comparison
equal deleted inserted replaced
-1:000000000000 0:376386a54a3c
1 ;;; sort.el --- commands to sort text in an XEmacs buffer.
2
3 ;; Copyright (C) 1986, 1987, 1994, 1995 Free Software Foundation, Inc.
4
5 ;; Author: Howie Kaye
6 ;; Maintainer: FSF
7 ;; Keywords: unix
8
9 ;; This file is part of XEmacs.
10
11 ;; XEmacs is free software; you can redistribute it and/or modify it
12 ;; under the terms of the GNU General Public License as published by
13 ;; the Free Software Foundation; either version 2, or (at your option)
14 ;; any later version.
15
16 ;; XEmacs is distributed in the hope that it will be useful, but
17 ;; WITHOUT ANY WARRANTY; without even the implied warranty of
18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19 ;; General Public License for more details.
20
21 ;; You should have received a copy of the GNU General Public License
22 ;; along with XEmacs; see the file COPYING. If not, write to the Free
23 ;; Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
24
25 ;;; Synched up with: FSF 19.30.
26
27 ;;; Commentary:
28
29 ;;; This package provides the sorting facilities documented in the XEmacs
30 ;;; Reference Manual.
31
32 ;;; Code:
33
34 (defvar sort-fold-case nil
35 "*Non-nil if the buffer sort functions should ignore case.")
36
37 ;;;###autoload
38 (defun sort-subr (reverse nextrecfun endrecfun &optional startkeyfun endkeyfun)
39 "General text sorting routine to divide buffer into records and sort them.
40 Arguments are REVERSE NEXTRECFUN ENDRECFUN &optional STARTKEYFUN ENDKEYFUN.
41
42 We divide the accessible portion of the buffer into disjoint pieces
43 called sort records. A portion of each sort record (perhaps all of
44 it) is designated as the sort key. The records are rearranged in the
45 buffer in order by their sort keys. The records may or may not be
46 contiguous.
47
48 Usually the records are rearranged in order of ascending sort key.
49 If REVERSE is non-nil, they are rearranged in order of descending sort key.
50
51 The next four arguments are functions to be called to move point
52 across a sort record. They will be called many times from within sort-subr.
53
54 NEXTRECFUN is called with point at the end of the previous record.
55 It moves point to the start of the next record.
56 It should move point to the end of the buffer if there are no more records.
57 The first record is assumed to start at the position of point when sort-subr
58 is called.
59
60 ENDRECFUN is called with point within the record.
61 It should move point to the end of the record.
62
63 STARTKEYFUN moves from the start of the record to the start of the key.
64 It may return either a non-nil value to be used as the key, or
65 else the key is the substring between the values of point after
66 STARTKEYFUN and ENDKEYFUN are called. If STARTKEYFUN is nil, the key
67 starts at the beginning of the record.
68
69 ENDKEYFUN moves from the start of the sort key to the end of the sort key.
70 ENDKEYFUN may be nil if STARTKEYFUN returns a value or if it would be the
71 same as ENDRECFUN."
72 ;; Heuristically try to avoid messages if sorting a small amt of text.
73 (let ((messages (> (- (point-max) (point-min)) 50000)))
74 (save-excursion
75 (if messages (message "Finding sort keys..."))
76 (let* ((sort-lists (sort-build-lists nextrecfun endrecfun
77 startkeyfun endkeyfun))
78 (old (reverse sort-lists))
79 (case-fold-search sort-fold-case))
80 (if (null sort-lists)
81 ()
82 (or reverse (setq sort-lists (nreverse sort-lists)))
83 (if messages (message "Sorting records..."))
84 (setq sort-lists
85 (if (fboundp 'sortcar)
86 (sortcar sort-lists
87 (cond ((numberp (car (car sort-lists)))
88 ;; This handles both ints and floats.
89 '<)
90 ((consp (car (car sort-lists)))
91 (function
92 (lambda (a b)
93 (> 0 (compare-buffer-substrings
94 nil (car a) (cdr a)
95 nil (car b) (cdr b))))))
96 (t
97 'string<)))
98 (sort sort-lists
99 (cond ((numberp (car (car sort-lists)))
100 'car-less-than-car)
101 ((consp (car (car sort-lists)))
102 (function (lambda (a b)
103 (> 0 (compare-buffer-substrings
104 nil (car (car a)) (cdr (car a))
105 nil (car (car b)) (cdr (car b)))))))
106 (t
107 (function
108 (lambda (a b)
109 (string< (car a) (car b)))))))))
110 (if reverse (setq sort-lists (nreverse sort-lists)))
111 (if messages (message "Reordering buffer..."))
112 (sort-reorder-buffer sort-lists old)))
113 (if messages (message "Reordering buffer... Done"))))
114 nil)
115
116 ;; Parse buffer into records using the arguments as Lisp expressions;
117 ;; return a list of records. Each record looks like (KEY STARTPOS . ENDPOS)
118 ;; where KEY is the sort key (a number or string),
119 ;; and STARTPOS and ENDPOS are the bounds of this record in the buffer.
120
121 ;; The records appear in the list lastmost first!
122
123 (defun sort-build-lists (nextrecfun endrecfun startkeyfun endkeyfun)
124 (let ((sort-lists ())
125 (start-rec nil)
126 done key)
127 ;; Loop over sort records.
128 ;(goto-char (point-min)) -- it is the caller's responsibility to
129 ;arrange this if necessary
130 (while (not (eobp))
131 (setq start-rec (point)) ;save record start
132 (setq done nil)
133 ;; Get key value, or move to start of key.
134 (setq key (catch 'key
135 (or (and startkeyfun (funcall startkeyfun))
136 ;; If key was not returned as value,
137 ;; move to end of key and get key from the buffer.
138 (let ((start (point)))
139 (funcall (or endkeyfun
140 (prog1 endrecfun (setq done t))))
141 (cons start (point))))))
142 ;; Move to end of this record (start of next one, or end of buffer).
143 (cond ((prog1 done (setq done nil)))
144 (endrecfun (funcall endrecfun))
145 (nextrecfun (funcall nextrecfun) (setq done t)))
146 (if key (setq sort-lists (cons
147 ;; consing optimization in case in which key
148 ;; is same as record.
149 (if (and (consp key)
150 (equal (car key) start-rec)
151 (equal (cdr key) (point)))
152 (cons key key)
153 (cons key (cons start-rec (point))))
154 sort-lists)))
155 (and (not done) nextrecfun (funcall nextrecfun)))
156 sort-lists))
157
158 (defun sort-reorder-buffer (sort-lists old)
159 (let ((inhibit-quit t)
160 (last (point-min))
161 (min (point-min)) (max (point-max)))
162 ;; Make sure insertions done for reordering
163 ;; do not go after any markers at the end of the sorted region,
164 ;; by inserting a space to separate them.
165 (goto-char (point-max))
166 (insert-before-markers " ")
167 (narrow-to-region min (1- (point-max)))
168 (while sort-lists
169 (goto-char (point-max))
170 (insert-buffer-substring (current-buffer)
171 last
172 (nth 1 (car old)))
173 (goto-char (point-max))
174 (insert-buffer-substring (current-buffer)
175 (nth 1 (car sort-lists))
176 (cdr (cdr (car sort-lists))))
177 (setq last (cdr (cdr (car old)))
178 sort-lists (cdr sort-lists)
179 old (cdr old)))
180 (goto-char (point-max))
181 (insert-buffer-substring (current-buffer)
182 last
183 max)
184 ;; Delete the original copy of the text.
185 (delete-region min max)
186 ;; Get rid of the separator " ".
187 (goto-char (point-max))
188 (narrow-to-region min (1+ (point)))
189 (delete-region (point) (1+ (point)))))
190
191 ;;;###autoload
192 (defun sort-lines (reverse beg end)
193 "Sort lines in region alphabetically; argument means descending order.
194 Called from a program, there are three arguments:
195 REVERSE (non-nil means reverse order), BEG and END (region to sort)."
196 (interactive "P\nr")
197 (save-excursion
198 (save-restriction
199 (narrow-to-region beg end)
200 (goto-char (point-min))
201 (sort-subr reverse 'forward-line 'end-of-line))))
202
203 ;;;###autoload
204 (defun sort-paragraphs (reverse beg end)
205 "Sort paragraphs in region alphabetically; argument means descending order.
206 Called from a program, there are three arguments:
207 REVERSE (non-nil means reverse order), BEG and END (region to sort)."
208 (interactive "P\nr")
209 (save-excursion
210 (save-restriction
211 (narrow-to-region beg end)
212 (goto-char (point-min))
213 (sort-subr reverse
214 (function (lambda ()
215 (while (and (not (eobp)) (looking-at paragraph-separate))
216 (forward-line 1))))
217 'forward-paragraph))))
218
219 ;;;###autoload
220 (defun sort-pages (reverse beg end)
221 "Sort pages in region alphabetically; argument means descending order.
222 Called from a program, there are three arguments:
223 REVERSE (non-nil means reverse order), BEG and END (region to sort)."
224 (interactive "P\nr")
225 (save-excursion
226 (save-restriction
227 (narrow-to-region beg end)
228 (goto-char (point-min))
229 (sort-subr reverse
230 (function (lambda () (skip-chars-forward "\n")))
231 'forward-page))))
232
233 (defvar sort-fields-syntax-table nil)
234 (if sort-fields-syntax-table nil
235 (let ((table (make-syntax-table))
236 (i 0))
237 (while (< i 256)
238 (modify-syntax-entry i "w" table)
239 (setq i (1+ i)))
240 (modify-syntax-entry ?\ " " table)
241 (modify-syntax-entry ?\t " " table)
242 (modify-syntax-entry ?\n " " table)
243 (modify-syntax-entry ?\. "_" table) ; for floating pt. numbers. -wsr
244 (setq sort-fields-syntax-table table)))
245
246 ;;;###autoload
247 (defun sort-numeric-fields (field beg end)
248 "Sort lines in region numerically by the ARGth field of each line.
249 Fields are separated by whitespace and numbered from 1 up.
250 Specified field must contain a number in each line of the region.
251 With a negative arg, sorts by the ARGth field counted from the right.
252 Called from a program, there are three arguments:
253 FIELD, BEG and END. BEG and END specify region to sort.
254 If you want to sort floating-point numbers, try `sort-float-fields'."
255 (interactive "p\nr")
256 (sort-fields-1 field beg end
257 (function (lambda ()
258 (sort-skip-fields field)
259 (string-to-number
260 (buffer-substring
261 (point)
262 (save-excursion
263 ;; This is just wrong! Even without floats...
264 ;; (skip-chars-forward "[0-9]")
265 (forward-sexp 1)
266 (point))))))
267 nil))
268
269 ;;;###autoload
270 (defun sort-float-fields (field beg end)
271 "Sort lines in region numerically by the ARGth field of each line.
272 Fields are separated by whitespace and numbered from 1 up. Specified field
273 must contain a floating point number in each line of the region. With a
274 negative arg, sorts by the ARGth field counted from the right. Called from a
275 program, there are three arguments: FIELD, BEG and END. BEG and END specify
276 region to sort."
277 (interactive "p\nr")
278 (sort-fields-1 field beg end
279 (function (lambda ()
280 (sort-skip-fields field)
281 (string-to-number
282 (buffer-substring
283 (point)
284 (save-excursion
285 (re-search-forward
286 "[+-]?[0-9]*\.?[0-9]*\\([eE][+-]?[0-9]+\\)?")
287 (point))))))
288 nil))
289
290 ;;;###autoload
291 (defun sort-fields (field beg end)
292 "Sort lines in region lexicographically by the ARGth field of each line.
293 Fields are separated by whitespace and numbered from 1 up.
294 With a negative arg, sorts by the ARGth field counted from the right.
295 Called from a program, there are three arguments:
296 FIELD, BEG and END. BEG and END specify region to sort."
297 (interactive "p\nr")
298 (sort-fields-1 field beg end
299 (function (lambda ()
300 (sort-skip-fields field)
301 nil))
302 (function (lambda () (skip-chars-forward "^ \t\n")))))
303
304 (defun sort-fields-1 (field beg end startkeyfun endkeyfun)
305 (let ((tbl (syntax-table)))
306 (if (zerop field) (setq field 1))
307 (unwind-protect
308 (save-excursion
309 (save-restriction
310 (narrow-to-region beg end)
311 (goto-char (point-min))
312 (set-syntax-table sort-fields-syntax-table)
313 (sort-subr nil
314 'forward-line 'end-of-line
315 startkeyfun endkeyfun)))
316 (set-syntax-table tbl))))
317
318 ;; Position at the beginning of field N on the current line,
319 ;; assuming point is initially at the beginning of the line.
320 (defun sort-skip-fields (n)
321 (if (> n 0)
322 ;; Skip across N - 1 fields.
323 (let ((i (1- n)))
324 (while (> i 0)
325 (skip-chars-forward " \t")
326 (skip-chars-forward "^ \t\n")
327 (setq i (1- i)))
328 (skip-chars-forward " \t")
329 (if (eolp)
330 (error "Line has too few fields: %s"
331 (buffer-substring
332 (save-excursion (beginning-of-line) (point))
333 (save-excursion (end-of-line) (point))))))
334 (end-of-line)
335 ;; Skip back across - N - 1 fields.
336 (let ((i (1- (- n))))
337 (while (> i 0)
338 (skip-chars-backward " \t")
339 (skip-chars-backward "^ \t\n")
340 (setq i (1- i)))
341 (skip-chars-backward " \t"))
342 (if (bolp)
343 (error "Line has too few fields: %s"
344 (buffer-substring
345 (save-excursion (beginning-of-line) (point))
346 (save-excursion (end-of-line) (point)))))
347 ;; Position at the front of the field
348 ;; even if moving backwards.
349 (skip-chars-backward "^ \t\n")))
350
351
352 (defvar sort-regexp-fields-regexp)
353 (defvar sort-regexp-record-end)
354
355 ;; Move to the beginning of the next match for record-regexp,
356 ;; and set sort-regexp-record-end to the end of that match.
357 ;; If the next match is empty and does not advance point,
358 ;; skip one character and try again.
359 (defun sort-regexp-fields-next-record ()
360 (let ((oldpos (point)))
361 (and (re-search-forward sort-regexp-fields-regexp nil 'move)
362 (setq sort-regexp-record-end (match-end 0))
363 (if (= sort-regexp-record-end oldpos)
364 (progn
365 (forward-char 1)
366 (re-search-forward sort-regexp-fields-regexp nil 'move)
367 (setq sort-regexp-record-end (match-end 0)))
368 t)
369 (goto-char (match-beginning 0)))))
370
371 ;;;###autoload
372 (defun sort-regexp-fields (reverse record-regexp key-regexp beg end)
373 "Sort the region lexicographically as specified by RECORD-REGEXP and KEY.
374 RECORD-REGEXP specifies the textual units which should be sorted.
375 For example, to sort lines RECORD-REGEXP would be \"^.*$\"
376 KEY specifies the part of each record (ie each match for RECORD-REGEXP)
377 is to be used for sorting.
378 If it is \"\\\\digit\" then the digit'th \"\\\\(...\\\\)\" match field from
379 RECORD-REGEXP is used.
380 If it is \"\\\\&\" then the whole record is used.
381 Otherwise, it is a regular-expression for which to search within the record.
382 If a match for KEY is not found within a record then that record is ignored.
383
384 With a negative prefix arg sorts in reverse order.
385
386 For example: to sort lines in the region by the first word on each line
387 starting with the letter \"f\",
388 RECORD-REGEXP would be \"^.*$\" and KEY would be \"\\\\=\\<f\\\\w*\\\\>\""
389 ;; using negative prefix arg to mean "reverse" is now inconsistent with
390 ;; other sort-.*fields functions but then again this was before, since it
391 ;; didn't use the magnitude of the arg to specify anything.
392 (interactive "P\nsRegexp specifying records to sort:
393 sRegexp specifying key within record: \nr")
394 (cond ((or (equal key-regexp "") (equal key-regexp "\\&"))
395 (setq key-regexp 0))
396 ((string-match "\\`\\\\[1-9]\\'" key-regexp)
397 (setq key-regexp (- (aref key-regexp 1) ?0))))
398 (save-excursion
399 (save-restriction
400 (narrow-to-region beg end)
401 (goto-char (point-min))
402 (let (sort-regexp-record-end
403 (sort-regexp-fields-regexp record-regexp))
404 (re-search-forward sort-regexp-fields-regexp)
405 (setq sort-regexp-record-end (point))
406 (goto-char (match-beginning 0))
407 (sort-subr reverse
408 'sort-regexp-fields-next-record
409 (function (lambda ()
410 (goto-char sort-regexp-record-end)))
411 (function (lambda ()
412 (let ((n 0))
413 (cond ((numberp key-regexp)
414 (setq n key-regexp))
415 ((re-search-forward
416 key-regexp sort-regexp-record-end t)
417 (setq n 0))
418 (t (throw 'key nil)))
419 (condition-case ()
420 (if (fboundp 'buffer-substring-lessp)
421 (cons (match-beginning n)
422 (match-end n))
423 (buffer-substring (match-beginning n)
424 (match-end n)))
425 ;; if there was no such register
426 (error (throw 'key nil)))))))))))
427
428
429 (defvar sort-columns-subprocess t)
430
431 ;;;###autoload
432 (defun sort-columns (reverse &optional beg end)
433 "Sort lines in region alphabetically by a certain range of columns.
434 For the purpose of this command, the region includes
435 the entire line that point is in and the entire line the mark is in.
436 The column positions of point and mark bound the range of columns to sort on.
437 A prefix argument means sort into reverse order.
438
439 Note that `sort-columns' rejects text that contains tabs,
440 because tabs could be split across the specified columns
441 and it doesn't know how to handle that. Also, when possible,
442 it uses the `sort' utility program, which doesn't understand tabs.
443 Use \\[untabify] to convert tabs to spaces before sorting."
444 (interactive "P\nr")
445 (save-excursion
446 (let (beg1 end1 col-beg1 col-end1 col-start col-end)
447 (goto-char (min beg end))
448 (setq col-beg1 (current-column))
449 (beginning-of-line)
450 (setq beg1 (point))
451 (goto-char (max beg end))
452 (setq col-end1 (current-column))
453 (forward-line)
454 (setq end1 (point))
455 (setq col-start (min col-beg1 col-end1))
456 (setq col-end (max col-beg1 col-end1))
457 (if (search-backward "\t" beg1 t)
458 (error
459 "sort-columns does not work with tabs. Use M-x untabify."))
460 (if (not (eq system-type 'vax-vms))
461 ;; Use the sort utility if we can; it is 4 times as fast.
462 (call-process-region beg1 end1 "sort" t t nil
463 (if reverse "-rt\n" "-t\n")
464 (concat "+0." (int-to-string col-start))
465 (concat "-0." (int-to-string col-end)))
466 ;; On VMS, use Emacs's own facilities.
467 (save-excursion
468 (save-restriction
469 (narrow-to-region beg1 end1)
470 (goto-char beg1)
471 (sort-subr reverse 'forward-line 'end-of-line
472 (function (lambda () (move-to-column col-start) nil))
473 (function (lambda () (move-to-column col-end) nil)))))))))
474
475 ;;;###autoload
476 (defun reverse-region (beg end)
477 "Reverse the order of lines in a region.
478 From a program takes two point or marker arguments, BEG and END."
479 (interactive "r")
480 (if (> beg end)
481 (let (mid) (setq mid end end beg beg mid)))
482 (save-excursion
483 ;; put beg at the start of a line and end and the end of one --
484 ;; the largest possible region which fits this criteria
485 (goto-char beg)
486 (or (bolp) (forward-line 1))
487 (setq beg (point))
488 (goto-char end)
489 ;; the test for bolp is for those times when end is on an empty line;
490 ;; it is probably not the case that the line should be included in the
491 ;; reversal; it isn't difficult to add it afterward.
492 (or (and (eolp) (not (bolp))) (progn (forward-line -1) (end-of-line)))
493 (setq end (point-marker))
494 ;; the real work. this thing cranks through memory on large regions.
495 (let (ll (do t))
496 (while do
497 (goto-char beg)
498 (setq ll (cons (buffer-substring (point) (progn (end-of-line) (point)))
499 ll))
500 (setq do (/= (point) end))
501 (delete-region beg (if do (1+ (point)) (point))))
502 (while (cdr ll)
503 (insert (car ll) "\n")
504 (setq ll (cdr ll)))
505 (insert (car ll)))))
506
507 (provide 'sort)
508
509 ;;; sort.el ends here