|
209
|
1 ;;; undo-stack.el --- An "undoable stack" object.
|
|
|
2
|
|
|
3 ;; Copyright (C) 1997 Free Software Foundation, Inc.
|
|
|
4 ;; Copyright (C) 1996 Ben Wing.
|
|
|
5
|
|
|
6 ;; Maintainer: XEmacs Development Team
|
|
|
7 ;; Keywords: extensions, dumped
|
|
|
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
|
|
|
23 ;; Free Software Foundation, 59 Temple Place - Suite 330,
|
|
|
24 ;; Boston, MA 02111-1307, USA.
|
|
|
25
|
|
|
26 ;;; Synched up with: Not in FSF.
|
|
|
27
|
|
|
28 ;;; Commentary:
|
|
|
29
|
|
|
30 ;; This file is dumped with XEmacs.
|
|
|
31
|
|
|
32 ;; An "undoable stack" is an object that can be used to implement
|
|
|
33 ;; a history of positions, with undo and redo. Conceptually, it
|
|
|
34 ;; is the kind of data structure used to keep track of (e.g.)
|
|
|
35 ;; visited Web pages, so that the "Back" and "Forward" operations
|
|
|
36 ;; in the browser work. Basically, I can successively visit a
|
|
|
37 ;; number of Web pages through links, and then hit "Back" a
|
|
|
38 ;; few times to go to previous positions, and then "Forward" a
|
|
|
39 ;; few times to reverse this process. This is similar to an
|
|
|
40 ;; "undo" and "redo" mechanism.
|
|
|
41
|
|
|
42 ;; Note that Emacs does not standardly contain structures like
|
|
|
43 ;; this. Instead, it implements history using either a ring
|
|
|
44 ;; (the kill ring, the mark ring), or something like the undo
|
|
|
45 ;; stack, where successive "undo" operations get recorded as
|
|
|
46 ;; normal modifications, so that if you do a bunch of successive
|
|
|
47 ;; undo's, then something else, then start undoing, you will
|
|
|
48 ;; be redoing all your undo's back to the point before you did
|
|
|
49 ;; the undo's, and then further undo's will act like the previous
|
|
|
50 ;; round of undo's. I think that both of these paradigms are
|
|
|
51 ;; inferior to the "undoable-stack" paradigm because they're
|
|
|
52 ;; confusing and difficult to keep track of.
|
|
|
53
|
|
|
54 ;; Conceptually, imagine a position history like this:
|
|
|
55
|
|
|
56 ;; 1 -> 2 -> 3 -> 4 -> 5 -> 6
|
|
|
57 ;; ^^
|
|
|
58
|
|
|
59 ;; where the arrow indicates where you currently are. "Going back"
|
|
|
60 ;; and "going forward" just amount to moving the arrow. However,
|
|
|
61 ;; what happens if the history state is this:
|
|
|
62
|
|
|
63 ;; 1 -> 2 -> 3 -> 4 -> 5 -> 6
|
|
|
64 ;; ^^
|
|
|
65
|
|
|
66 ;; and then I visit new positions (7) and (8)? In the most general
|
|
|
67 ;; implementation, you've just caused a new branch like this:
|
|
|
68
|
|
|
69 ;; 1 -> 2 -> 3 -> 4 -> 5 -> 6
|
|
|
70 ;; |
|
|
|
71 ;; |
|
|
|
72 ;; 7 -> 8
|
|
|
73 ;; ^^
|
|
|
74
|
|
|
75 ;; But then you can end up with a whole big tree, and you need
|
|
|
76 ;; more sophisticated ways of navigating ("Forward" might involve
|
|
|
77 ;; a choice of paths to follow) and managing its size (if you don't
|
|
|
78 ;; want to keep unlimited history, you have to truncate at some point,
|
|
|
79 ;; and how do you truncate a tree?)
|
|
|
80
|
|
|
81 ;; My solution to this is just to insert the new positions like
|
|
|
82 ;; this:
|
|
|
83
|
|
|
84 ;; 1 -> 2 -> 3 -> 4 -> 7 -> 8 -> 5 -> 6
|
|
|
85 ;; ^^
|
|
|
86
|
|
|
87 ;; (Netscape, I think, would just truncate 5 and 6 completely,
|
|
|
88 ;; but that seems a bit drastic. In the Emacs-standard "ring"
|
|
|
89 ;; structure, this problem is avoided by simply moving 5 and 6
|
|
|
90 ;; to the beginning of the ring. However, it doesn't seem
|
|
|
91 ;; logical to me to have "going back past 1" get you to 6.)
|
|
|
92
|
|
|
93 ;; Now what if we have a "maximum" size of (say) 7 elements?
|
|
|
94 ;; When we add 8, we could truncate either 1 or 6. Since 5 and
|
|
|
95 ;; 6 are "undone" positions, we should presumably truncate
|
|
|
96 ;; them before 1. So, adding 8 truncates 6, adding 9 truncates
|
|
|
97 ;; 5, and adding 10 truncates 1 because there is nothing more
|
|
|
98 ;; that is forward of the insertion point.
|
|
|
99
|
|
|
100 ;; Interestingly, this method of truncation is almost like
|
|
|
101 ;; how a ring would truncate. A ring would move 5 and 6
|
|
|
102 ;; around to the back, like this:
|
|
|
103
|
|
|
104 ;; 5 -> 6 -> 1 -> 2 -> 3 -> 4 -> 7 -> 8
|
|
|
105 ;; ^^
|
|
|
106
|
|
|
107 ;; However, when 8 is added, the ring truncates 5 instead of
|
|
|
108 ;; 6, which is less than optimal.
|
|
|
109
|
|
|
110 ;; Conceptually, we can implement the "undoable stack" using
|
|
|
111 ;; two stacks of a sort called "truncatable stack", which are
|
|
|
112 ;; just simple stacks, but where you can truncate elements
|
|
|
113 ;; off of the bottom of the stack. Then, the undoable stack
|
|
|
114
|
|
|
115 ;; 1 -> 2 -> 3 -> 4 -> 5 -> 6
|
|
|
116 ;; ^^
|
|
|
117
|
|
|
118 ;; is equivalent to two truncatable stacks:
|
|
|
119
|
|
|
120 ;; 4 <- 3 <- 2 <- 1
|
|
|
121 ;; 5 <- 6
|
|
|
122
|
|
|
123 ;; where I reversed the direction to accord with the probable
|
|
|
124 ;; implementation of a standard list. To do another undo,
|
|
|
125 ;; I pop 4 off of the first stack and move it to the top of
|
|
|
126 ;; the second stack. A redo operation does the opposite.
|
|
|
127 ;; To truncate to the proper size, first chop off 6, then 5,
|
|
|
128 ;; then 1 -- in all cases, truncating off the bottom.
|
|
|
129
|
|
|
130 ;;; Code:
|
|
|
131
|
|
223
|
132 (define-error 'trunc-stack-bottom "Bottom of stack reached")
|
|
209
|
133
|
|
|
134 (defsubst trunc-stack-stack (stack)
|
|
|
135 ;; return the list representing the trunc-stack's elements.
|
|
|
136 ;; the head of the list is the most recent element.
|
|
|
137 (aref stack 1))
|
|
|
138
|
|
|
139 (defsubst trunc-stack-length (stack)
|
|
|
140 ;; return the number of elements in the trunc-stack.
|
|
|
141 (aref stack 2))
|
|
|
142
|
|
|
143 (defsubst set-trunc-stack-stack (stack new)
|
|
|
144 ;; set the list representing the trunc-stack's elements.
|
|
|
145 (aset stack 1 new))
|
|
|
146
|
|
|
147 (defsubst set-trunc-stack-length (stack new)
|
|
|
148 ;; set the length of the trunc-stack.
|
|
|
149 (aset stack 2 new))
|
|
|
150
|
|
|
151 ;; public functions:
|
|
|
152
|
|
|
153 (defun make-trunc-stack ()
|
|
|
154 ;; make an empty trunc-stack.
|
|
|
155 (vector 'trunc-stack nil 0))
|
|
|
156
|
|
|
157 (defun trunc-stack-push (stack el)
|
|
|
158 ;; push a new element onto the head of the trunc-stack.
|
|
|
159 (set-trunc-stack-stack stack (cons el (trunc-stack-stack stack)))
|
|
|
160 (set-trunc-stack-length stack (1+ (trunc-stack-length stack))))
|
|
|
161
|
|
|
162 (defun trunc-stack-top (stack &optional n)
|
|
|
163 ;; return the nth topmost element from the trunc-stack.
|
|
|
164 ;; signal an error if the stack doesn't have that many elements.
|
|
|
165 (or n (setq n 0))
|
|
|
166 (if (>= n (trunc-stack-length stack))
|
|
|
167 (signal-error 'trunc-stack-bottom (list stack))
|
|
|
168 (nth n (trunc-stack-stack stack))))
|
|
|
169
|
|
|
170 (defun trunc-stack-pop (stack)
|
|
|
171 ;; pop and return the topmost element from the stack.
|
|
|
172 (prog1 (trunc-stack-top stack)
|
|
|
173 (set-trunc-stack-stack stack (cdr (trunc-stack-stack stack)))
|
|
|
174 (set-trunc-stack-length stack (1- (trunc-stack-length stack)))))
|
|
|
175
|
|
|
176 (defun trunc-stack-truncate (stack &optional n)
|
|
|
177 ;; truncate N items off the bottom of the stack. If the stack is
|
|
|
178 ;; not that big, it just becomes empty.
|
|
|
179 (or n (setq n 1))
|
|
|
180 (if (> n 0)
|
|
|
181 (let ((len (trunc-stack-length stack)))
|
|
|
182 (if (>= n len)
|
|
|
183 (progn
|
|
|
184 (set-trunc-stack-length stack 0)
|
|
|
185 (set-trunc-stack-stack stack nil))
|
|
|
186 (setcdr (nthcdr (1- (- len n)) (trunc-stack-stack stack)) nil)
|
|
|
187 (set-trunc-stack-length stack (- len n))))))
|
|
|
188
|
|
|
189 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
|
|
|
190
|
|
|
191 ;;; FMH! FMH! FMH! This object-oriented stuff doesn't really work
|
|
|
192 ;;; properly without built-in structures (vectors suck) and without
|
|
251
|
193 ;;; public and private functions and fields.
|
|
209
|
194
|
|
|
195 (defsubst undoable-stack-max (stack)
|
|
|
196 (aref stack 1))
|
|
|
197
|
|
|
198 (defsubst undoable-stack-a (stack)
|
|
|
199 (aref stack 2))
|
|
|
200
|
|
|
201 (defsubst undoable-stack-b (stack)
|
|
|
202 (aref stack 3))
|
|
|
203
|
|
|
204 ;; public functions:
|
|
|
205
|
|
|
206 (defun make-undoable-stack (max)
|
|
|
207 ;; make an empty undoable stack of max size MAX.
|
|
|
208 (vector 'undoable-stack max (make-trunc-stack) (make-trunc-stack)))
|
|
|
209
|
|
|
210 (defsubst set-undoable-stack-max (stack new)
|
|
|
211 ;; change the max size of an undoable stack.
|
|
|
212 (aset stack 1 new))
|
|
|
213
|
|
|
214 (defun undoable-stack-a-top (stack)
|
|
|
215 ;; return the topmost element off the "A" stack of an undoable stack.
|
|
|
216 ;; this is the most recent position pushed on the undoable stack.
|
|
|
217 (trunc-stack-top (undoable-stack-a stack)))
|
|
|
218
|
|
|
219 (defun undoable-stack-a-length (stack)
|
|
|
220 (trunc-stack-length (undoable-stack-a stack)))
|
|
|
221
|
|
|
222 (defun undoable-stack-b-top (stack)
|
|
|
223 ;; return the topmost element off the "B" stack of an undoable stack.
|
|
|
224 ;; this is the position that will become the most recent position,
|
|
|
225 ;; after a redo operation.
|
|
|
226 (trunc-stack-top (undoable-stack-b stack)))
|
|
|
227
|
|
|
228 (defun undoable-stack-b-length (stack)
|
|
|
229 (trunc-stack-length (undoable-stack-b stack)))
|
|
|
230
|
|
|
231 (defun undoable-stack-push (stack el)
|
|
|
232 ;; push an element onto the stack.
|
|
|
233 (let*
|
|
|
234 ((lena (trunc-stack-length (undoable-stack-a stack)))
|
|
|
235 (lenb (trunc-stack-length (undoable-stack-b stack)))
|
|
|
236 (max (undoable-stack-max stack))
|
|
|
237 (len (+ lena lenb)))
|
|
|
238 ;; maybe truncate some elements. We have to deal with the
|
|
|
239 ;; possibility that we have more elements than our max
|
|
|
240 ;; (someone might have reduced the max).
|
|
|
241 (if (>= len max)
|
|
|
242 (let ((must-nuke (1+ (- len max))))
|
|
|
243 ;; chop off must-nuke elements from the B stack.
|
|
|
244 (trunc-stack-truncate (undoable-stack-b stack) must-nuke)
|
|
|
245 ;; but if there weren't that many elements to chop,
|
|
|
246 ;; take the rest off the A stack.
|
|
|
247 (if (< lenb must-nuke)
|
|
|
248 (trunc-stack-truncate (undoable-stack-a stack)
|
|
|
249 (- must-nuke lenb)))))
|
|
|
250 (trunc-stack-push (undoable-stack-a stack) el)))
|
|
|
251
|
|
|
252 (defun undoable-stack-pop (stack)
|
|
|
253 ;; pop an element off the stack.
|
|
|
254 (trunc-stack-pop (undoable-stack-a stack)))
|
|
|
255
|
|
|
256 (defun undoable-stack-undo (stack)
|
|
|
257 ;; transfer an element from the top of A to the top of B.
|
|
|
258 ;; return value is undefined.
|
|
|
259 (trunc-stack-push (undoable-stack-b stack)
|
|
|
260 (trunc-stack-pop (undoable-stack-a stack))))
|
|
|
261
|
|
|
262 (defun undoable-stack-redo (stack)
|
|
|
263 ;; transfer an element from the top of B to the top of A.
|
|
|
264 ;; return value is undefined.
|
|
|
265 (trunc-stack-push (undoable-stack-a stack)
|
|
|
266 (trunc-stack-pop (undoable-stack-b stack))))
|
|
|
267
|
|
|
268
|
|
|
269 ;;; undo-stack.el ends here
|
