1 |
721 |
jeremybenn |
/*
|
2 |
|
|
* Copyright (c) 1993-1994 by Xerox Corporation. All rights reserved.
|
3 |
|
|
*
|
4 |
|
|
* THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
|
5 |
|
|
* OR IMPLIED. ANY USE IS AT YOUR OWN RISK.
|
6 |
|
|
*
|
7 |
|
|
* Permission is hereby granted to use or copy this program
|
8 |
|
|
* for any purpose, provided the above notices are retained on all copies.
|
9 |
|
|
* Permission to modify the code and to distribute modified code is granted,
|
10 |
|
|
* provided the above notices are retained, and a notice that the code was
|
11 |
|
|
* modified is included with the above copyright notice.
|
12 |
|
|
*
|
13 |
|
|
* Author: Hans-J. Boehm (boehm@parc.xerox.com)
|
14 |
|
|
*/
|
15 |
|
|
/* Boehm, October 3, 1994 5:19 pm PDT */
|
16 |
|
|
# include "gc.h"
|
17 |
|
|
# include "cord.h"
|
18 |
|
|
# include <stdlib.h>
|
19 |
|
|
# include <stdio.h>
|
20 |
|
|
# include <string.h>
|
21 |
|
|
|
22 |
|
|
/* An implementation of the cord primitives. These are the only */
|
23 |
|
|
/* Functions that understand the representation. We perform only */
|
24 |
|
|
/* minimal checks on arguments to these functions. Out of bounds */
|
25 |
|
|
/* arguments to the iteration functions may result in client functions */
|
26 |
|
|
/* invoked on garbage data. In most cases, client functions should be */
|
27 |
|
|
/* programmed defensively enough that this does not result in memory */
|
28 |
|
|
/* smashes. */
|
29 |
|
|
|
30 |
|
|
typedef void (* oom_fn)(void);
|
31 |
|
|
|
32 |
|
|
oom_fn CORD_oom_fn = (oom_fn) 0;
|
33 |
|
|
|
34 |
|
|
# define OUT_OF_MEMORY { if (CORD_oom_fn != (oom_fn) 0) (*CORD_oom_fn)(); \
|
35 |
|
|
ABORT("Out of memory\n"); }
|
36 |
|
|
# define ABORT(msg) { fprintf(stderr, "%s\n", msg); abort(); }
|
37 |
|
|
|
38 |
|
|
typedef unsigned long word;
|
39 |
|
|
|
40 |
|
|
typedef union {
|
41 |
|
|
struct Concatenation {
|
42 |
|
|
char null;
|
43 |
|
|
char header;
|
44 |
|
|
char depth; /* concatenation nesting depth. */
|
45 |
|
|
unsigned char left_len;
|
46 |
|
|
/* Length of left child if it is sufficiently */
|
47 |
|
|
/* short; 0 otherwise. */
|
48 |
|
|
# define MAX_LEFT_LEN 255
|
49 |
|
|
word len;
|
50 |
|
|
CORD left; /* length(left) > 0 */
|
51 |
|
|
CORD right; /* length(right) > 0 */
|
52 |
|
|
} concatenation;
|
53 |
|
|
struct Function {
|
54 |
|
|
char null;
|
55 |
|
|
char header;
|
56 |
|
|
char depth; /* always 0 */
|
57 |
|
|
char left_len; /* always 0 */
|
58 |
|
|
word len;
|
59 |
|
|
CORD_fn fn;
|
60 |
|
|
void * client_data;
|
61 |
|
|
} function;
|
62 |
|
|
struct Generic {
|
63 |
|
|
char null;
|
64 |
|
|
char header;
|
65 |
|
|
char depth;
|
66 |
|
|
char left_len;
|
67 |
|
|
word len;
|
68 |
|
|
} generic;
|
69 |
|
|
char string[1];
|
70 |
|
|
} CordRep;
|
71 |
|
|
|
72 |
|
|
# define CONCAT_HDR 1
|
73 |
|
|
|
74 |
|
|
# define FN_HDR 4
|
75 |
|
|
# define SUBSTR_HDR 6
|
76 |
|
|
/* Substring nodes are a special case of function nodes. */
|
77 |
|
|
/* The client_data field is known to point to a substr_args */
|
78 |
|
|
/* structure, and the function is either CORD_apply_access_fn */
|
79 |
|
|
/* or CORD_index_access_fn. */
|
80 |
|
|
|
81 |
|
|
/* The following may be applied only to function and concatenation nodes: */
|
82 |
|
|
#define IS_CONCATENATION(s) (((CordRep *)s)->generic.header == CONCAT_HDR)
|
83 |
|
|
|
84 |
|
|
#define IS_FUNCTION(s) ((((CordRep *)s)->generic.header & FN_HDR) != 0)
|
85 |
|
|
|
86 |
|
|
#define IS_SUBSTR(s) (((CordRep *)s)->generic.header == SUBSTR_HDR)
|
87 |
|
|
|
88 |
|
|
#define LEN(s) (((CordRep *)s) -> generic.len)
|
89 |
|
|
#define DEPTH(s) (((CordRep *)s) -> generic.depth)
|
90 |
|
|
#define GEN_LEN(s) (CORD_IS_STRING(s) ? strlen(s) : LEN(s))
|
91 |
|
|
|
92 |
|
|
#define LEFT_LEN(c) ((c) -> left_len != 0? \
|
93 |
|
|
(c) -> left_len \
|
94 |
|
|
: (CORD_IS_STRING((c) -> left) ? \
|
95 |
|
|
(c) -> len - GEN_LEN((c) -> right) \
|
96 |
|
|
: LEN((c) -> left)))
|
97 |
|
|
|
98 |
|
|
#define SHORT_LIMIT (sizeof(CordRep) - 1)
|
99 |
|
|
/* Cords shorter than this are C strings */
|
100 |
|
|
|
101 |
|
|
|
102 |
|
|
/* Dump the internal representation of x to stdout, with initial */
|
103 |
|
|
/* indentation level n. */
|
104 |
|
|
void CORD_dump_inner(CORD x, unsigned n)
|
105 |
|
|
{
|
106 |
|
|
register size_t i;
|
107 |
|
|
|
108 |
|
|
for (i = 0; i < (size_t)n; i++) {
|
109 |
|
|
fputs(" ", stdout);
|
110 |
|
|
}
|
111 |
|
|
if (x == 0) {
|
112 |
|
|
fputs("NIL\n", stdout);
|
113 |
|
|
} else if (CORD_IS_STRING(x)) {
|
114 |
|
|
for (i = 0; i <= SHORT_LIMIT; i++) {
|
115 |
|
|
if (x[i] == '\0') break;
|
116 |
|
|
putchar(x[i]);
|
117 |
|
|
}
|
118 |
|
|
if (x[i] != '\0') fputs("...", stdout);
|
119 |
|
|
putchar('\n');
|
120 |
|
|
} else if (IS_CONCATENATION(x)) {
|
121 |
|
|
register struct Concatenation * conc =
|
122 |
|
|
&(((CordRep *)x) -> concatenation);
|
123 |
|
|
printf("Concatenation: %p (len: %d, depth: %d)\n",
|
124 |
|
|
x, (int)(conc -> len), (int)(conc -> depth));
|
125 |
|
|
CORD_dump_inner(conc -> left, n+1);
|
126 |
|
|
CORD_dump_inner(conc -> right, n+1);
|
127 |
|
|
} else /* function */{
|
128 |
|
|
register struct Function * func =
|
129 |
|
|
&(((CordRep *)x) -> function);
|
130 |
|
|
if (IS_SUBSTR(x)) printf("(Substring) ");
|
131 |
|
|
printf("Function: %p (len: %d): ", x, (int)(func -> len));
|
132 |
|
|
for (i = 0; i < 20 && i < func -> len; i++) {
|
133 |
|
|
putchar((*(func -> fn))(i, func -> client_data));
|
134 |
|
|
}
|
135 |
|
|
if (i < func -> len) fputs("...", stdout);
|
136 |
|
|
putchar('\n');
|
137 |
|
|
}
|
138 |
|
|
}
|
139 |
|
|
|
140 |
|
|
/* Dump the internal representation of x to stdout */
|
141 |
|
|
void CORD_dump(CORD x)
|
142 |
|
|
{
|
143 |
|
|
CORD_dump_inner(x, 0);
|
144 |
|
|
fflush(stdout);
|
145 |
|
|
}
|
146 |
|
|
|
147 |
|
|
CORD CORD_cat_char_star(CORD x, const char * y, size_t leny)
|
148 |
|
|
{
|
149 |
|
|
register size_t result_len;
|
150 |
|
|
register size_t lenx;
|
151 |
|
|
register int depth;
|
152 |
|
|
|
153 |
|
|
if (x == CORD_EMPTY) return(y);
|
154 |
|
|
if (leny == 0) return(x);
|
155 |
|
|
if (CORD_IS_STRING(x)) {
|
156 |
|
|
lenx = strlen(x);
|
157 |
|
|
result_len = lenx + leny;
|
158 |
|
|
if (result_len <= SHORT_LIMIT) {
|
159 |
|
|
register char * result = GC_MALLOC_ATOMIC(result_len+1);
|
160 |
|
|
|
161 |
|
|
if (result == 0) OUT_OF_MEMORY;
|
162 |
|
|
memcpy(result, x, lenx);
|
163 |
|
|
memcpy(result + lenx, y, leny);
|
164 |
|
|
result[result_len] = '\0';
|
165 |
|
|
return((CORD) result);
|
166 |
|
|
} else {
|
167 |
|
|
depth = 1;
|
168 |
|
|
}
|
169 |
|
|
} else {
|
170 |
|
|
register CORD right;
|
171 |
|
|
register CORD left;
|
172 |
|
|
register char * new_right;
|
173 |
|
|
register size_t right_len;
|
174 |
|
|
|
175 |
|
|
lenx = LEN(x);
|
176 |
|
|
|
177 |
|
|
if (leny <= SHORT_LIMIT/2
|
178 |
|
|
&& IS_CONCATENATION(x)
|
179 |
|
|
&& CORD_IS_STRING(right = ((CordRep *)x) -> concatenation.right)) {
|
180 |
|
|
/* Merge y into right part of x. */
|
181 |
|
|
if (!CORD_IS_STRING(left = ((CordRep *)x) -> concatenation.left)) {
|
182 |
|
|
right_len = lenx - LEN(left);
|
183 |
|
|
} else if (((CordRep *)x) -> concatenation.left_len != 0) {
|
184 |
|
|
right_len = lenx - ((CordRep *)x) -> concatenation.left_len;
|
185 |
|
|
} else {
|
186 |
|
|
right_len = strlen(right);
|
187 |
|
|
}
|
188 |
|
|
result_len = right_len + leny; /* length of new_right */
|
189 |
|
|
if (result_len <= SHORT_LIMIT) {
|
190 |
|
|
new_right = GC_MALLOC_ATOMIC(result_len + 1);
|
191 |
|
|
memcpy(new_right, right, right_len);
|
192 |
|
|
memcpy(new_right + right_len, y, leny);
|
193 |
|
|
new_right[result_len] = '\0';
|
194 |
|
|
y = new_right;
|
195 |
|
|
leny = result_len;
|
196 |
|
|
x = left;
|
197 |
|
|
lenx -= right_len;
|
198 |
|
|
/* Now fall through to concatenate the two pieces: */
|
199 |
|
|
}
|
200 |
|
|
if (CORD_IS_STRING(x)) {
|
201 |
|
|
depth = 1;
|
202 |
|
|
} else {
|
203 |
|
|
depth = DEPTH(x) + 1;
|
204 |
|
|
}
|
205 |
|
|
} else {
|
206 |
|
|
depth = DEPTH(x) + 1;
|
207 |
|
|
}
|
208 |
|
|
result_len = lenx + leny;
|
209 |
|
|
}
|
210 |
|
|
{
|
211 |
|
|
/* The general case; lenx, result_len is known: */
|
212 |
|
|
register struct Concatenation * result;
|
213 |
|
|
|
214 |
|
|
result = GC_NEW(struct Concatenation);
|
215 |
|
|
if (result == 0) OUT_OF_MEMORY;
|
216 |
|
|
result->header = CONCAT_HDR;
|
217 |
|
|
result->depth = depth;
|
218 |
|
|
if (lenx <= MAX_LEFT_LEN) result->left_len = lenx;
|
219 |
|
|
result->len = result_len;
|
220 |
|
|
result->left = x;
|
221 |
|
|
result->right = y;
|
222 |
|
|
if (depth >= MAX_DEPTH) {
|
223 |
|
|
return(CORD_balance((CORD)result));
|
224 |
|
|
} else {
|
225 |
|
|
return((CORD) result);
|
226 |
|
|
}
|
227 |
|
|
}
|
228 |
|
|
}
|
229 |
|
|
|
230 |
|
|
|
231 |
|
|
CORD CORD_cat(CORD x, CORD y)
|
232 |
|
|
{
|
233 |
|
|
register size_t result_len;
|
234 |
|
|
register int depth;
|
235 |
|
|
register size_t lenx;
|
236 |
|
|
|
237 |
|
|
if (x == CORD_EMPTY) return(y);
|
238 |
|
|
if (y == CORD_EMPTY) return(x);
|
239 |
|
|
if (CORD_IS_STRING(y)) {
|
240 |
|
|
return(CORD_cat_char_star(x, y, strlen(y)));
|
241 |
|
|
} else if (CORD_IS_STRING(x)) {
|
242 |
|
|
lenx = strlen(x);
|
243 |
|
|
depth = DEPTH(y) + 1;
|
244 |
|
|
} else {
|
245 |
|
|
register int depthy = DEPTH(y);
|
246 |
|
|
|
247 |
|
|
lenx = LEN(x);
|
248 |
|
|
depth = DEPTH(x) + 1;
|
249 |
|
|
if (depthy >= depth) depth = depthy + 1;
|
250 |
|
|
}
|
251 |
|
|
result_len = lenx + LEN(y);
|
252 |
|
|
{
|
253 |
|
|
register struct Concatenation * result;
|
254 |
|
|
|
255 |
|
|
result = GC_NEW(struct Concatenation);
|
256 |
|
|
if (result == 0) OUT_OF_MEMORY;
|
257 |
|
|
result->header = CONCAT_HDR;
|
258 |
|
|
result->depth = depth;
|
259 |
|
|
if (lenx <= MAX_LEFT_LEN) result->left_len = lenx;
|
260 |
|
|
result->len = result_len;
|
261 |
|
|
result->left = x;
|
262 |
|
|
result->right = y;
|
263 |
|
|
if (depth >= MAX_DEPTH) {
|
264 |
|
|
return(CORD_balance((CORD)result));
|
265 |
|
|
} else {
|
266 |
|
|
return((CORD) result);
|
267 |
|
|
}
|
268 |
|
|
}
|
269 |
|
|
}
|
270 |
|
|
|
271 |
|
|
|
272 |
|
|
|
273 |
|
|
CORD CORD_from_fn(CORD_fn fn, void * client_data, size_t len)
|
274 |
|
|
{
|
275 |
|
|
if (len <= 0) return(0);
|
276 |
|
|
if (len <= SHORT_LIMIT) {
|
277 |
|
|
register char * result;
|
278 |
|
|
register size_t i;
|
279 |
|
|
char buf[SHORT_LIMIT+1];
|
280 |
|
|
register char c;
|
281 |
|
|
|
282 |
|
|
for (i = 0; i < len; i++) {
|
283 |
|
|
c = (*fn)(i, client_data);
|
284 |
|
|
if (c == '\0') goto gen_case;
|
285 |
|
|
buf[i] = c;
|
286 |
|
|
}
|
287 |
|
|
buf[i] = '\0';
|
288 |
|
|
result = GC_MALLOC_ATOMIC(len+1);
|
289 |
|
|
if (result == 0) OUT_OF_MEMORY;
|
290 |
|
|
strcpy(result, buf);
|
291 |
|
|
result[len] = '\0';
|
292 |
|
|
return((CORD) result);
|
293 |
|
|
}
|
294 |
|
|
gen_case:
|
295 |
|
|
{
|
296 |
|
|
register struct Function * result;
|
297 |
|
|
|
298 |
|
|
result = GC_NEW(struct Function);
|
299 |
|
|
if (result == 0) OUT_OF_MEMORY;
|
300 |
|
|
result->header = FN_HDR;
|
301 |
|
|
/* depth is already 0 */
|
302 |
|
|
result->len = len;
|
303 |
|
|
result->fn = fn;
|
304 |
|
|
result->client_data = client_data;
|
305 |
|
|
return((CORD) result);
|
306 |
|
|
}
|
307 |
|
|
}
|
308 |
|
|
|
309 |
|
|
size_t CORD_len(CORD x)
|
310 |
|
|
{
|
311 |
|
|
if (x == 0) {
|
312 |
|
|
return(0);
|
313 |
|
|
} else {
|
314 |
|
|
return(GEN_LEN(x));
|
315 |
|
|
}
|
316 |
|
|
}
|
317 |
|
|
|
318 |
|
|
struct substr_args {
|
319 |
|
|
CordRep * sa_cord;
|
320 |
|
|
size_t sa_index;
|
321 |
|
|
};
|
322 |
|
|
|
323 |
|
|
char CORD_index_access_fn(size_t i, void * client_data)
|
324 |
|
|
{
|
325 |
|
|
register struct substr_args *descr = (struct substr_args *)client_data;
|
326 |
|
|
|
327 |
|
|
return(((char *)(descr->sa_cord))[i + descr->sa_index]);
|
328 |
|
|
}
|
329 |
|
|
|
330 |
|
|
char CORD_apply_access_fn(size_t i, void * client_data)
|
331 |
|
|
{
|
332 |
|
|
register struct substr_args *descr = (struct substr_args *)client_data;
|
333 |
|
|
register struct Function * fn_cord = &(descr->sa_cord->function);
|
334 |
|
|
|
335 |
|
|
return((*(fn_cord->fn))(i + descr->sa_index, fn_cord->client_data));
|
336 |
|
|
}
|
337 |
|
|
|
338 |
|
|
/* A version of CORD_substr that simply returns a function node, thus */
|
339 |
|
|
/* postponing its work. The fourth argument is a function that may */
|
340 |
|
|
/* be used for efficient access to the ith character. */
|
341 |
|
|
/* Assumes i >= 0 and i + n < length(x). */
|
342 |
|
|
CORD CORD_substr_closure(CORD x, size_t i, size_t n, CORD_fn f)
|
343 |
|
|
{
|
344 |
|
|
register struct substr_args * sa = GC_NEW(struct substr_args);
|
345 |
|
|
CORD result;
|
346 |
|
|
|
347 |
|
|
if (sa == 0) OUT_OF_MEMORY;
|
348 |
|
|
sa->sa_cord = (CordRep *)x;
|
349 |
|
|
sa->sa_index = i;
|
350 |
|
|
result = CORD_from_fn(f, (void *)sa, n);
|
351 |
|
|
((CordRep *)result) -> function.header = SUBSTR_HDR;
|
352 |
|
|
return (result);
|
353 |
|
|
}
|
354 |
|
|
|
355 |
|
|
# define SUBSTR_LIMIT (10 * SHORT_LIMIT)
|
356 |
|
|
/* Substrings of function nodes and flat strings shorter than */
|
357 |
|
|
/* this are flat strings. Othewise we use a functional */
|
358 |
|
|
/* representation, which is significantly slower to access. */
|
359 |
|
|
|
360 |
|
|
/* A version of CORD_substr that assumes i >= 0, n > 0, and i + n < length(x).*/
|
361 |
|
|
CORD CORD_substr_checked(CORD x, size_t i, size_t n)
|
362 |
|
|
{
|
363 |
|
|
if (CORD_IS_STRING(x)) {
|
364 |
|
|
if (n > SUBSTR_LIMIT) {
|
365 |
|
|
return(CORD_substr_closure(x, i, n, CORD_index_access_fn));
|
366 |
|
|
} else {
|
367 |
|
|
register char * result = GC_MALLOC_ATOMIC(n+1);
|
368 |
|
|
|
369 |
|
|
if (result == 0) OUT_OF_MEMORY;
|
370 |
|
|
strncpy(result, x+i, n);
|
371 |
|
|
result[n] = '\0';
|
372 |
|
|
return(result);
|
373 |
|
|
}
|
374 |
|
|
} else if (IS_CONCATENATION(x)) {
|
375 |
|
|
register struct Concatenation * conc
|
376 |
|
|
= &(((CordRep *)x) -> concatenation);
|
377 |
|
|
register size_t left_len;
|
378 |
|
|
register size_t right_len;
|
379 |
|
|
|
380 |
|
|
left_len = LEFT_LEN(conc);
|
381 |
|
|
right_len = conc -> len - left_len;
|
382 |
|
|
if (i >= left_len) {
|
383 |
|
|
if (n == right_len) return(conc -> right);
|
384 |
|
|
return(CORD_substr_checked(conc -> right, i - left_len, n));
|
385 |
|
|
} else if (i+n <= left_len) {
|
386 |
|
|
if (n == left_len) return(conc -> left);
|
387 |
|
|
return(CORD_substr_checked(conc -> left, i, n));
|
388 |
|
|
} else {
|
389 |
|
|
/* Need at least one character from each side. */
|
390 |
|
|
register CORD left_part;
|
391 |
|
|
register CORD right_part;
|
392 |
|
|
register size_t left_part_len = left_len - i;
|
393 |
|
|
|
394 |
|
|
if (i == 0) {
|
395 |
|
|
left_part = conc -> left;
|
396 |
|
|
} else {
|
397 |
|
|
left_part = CORD_substr_checked(conc -> left, i, left_part_len);
|
398 |
|
|
}
|
399 |
|
|
if (i + n == right_len + left_len) {
|
400 |
|
|
right_part = conc -> right;
|
401 |
|
|
} else {
|
402 |
|
|
right_part = CORD_substr_checked(conc -> right, 0,
|
403 |
|
|
n - left_part_len);
|
404 |
|
|
}
|
405 |
|
|
return(CORD_cat(left_part, right_part));
|
406 |
|
|
}
|
407 |
|
|
} else /* function */ {
|
408 |
|
|
if (n > SUBSTR_LIMIT) {
|
409 |
|
|
if (IS_SUBSTR(x)) {
|
410 |
|
|
/* Avoid nesting substring nodes. */
|
411 |
|
|
register struct Function * f = &(((CordRep *)x) -> function);
|
412 |
|
|
register struct substr_args *descr =
|
413 |
|
|
(struct substr_args *)(f -> client_data);
|
414 |
|
|
|
415 |
|
|
return(CORD_substr_closure((CORD)descr->sa_cord,
|
416 |
|
|
i + descr->sa_index,
|
417 |
|
|
n, f -> fn));
|
418 |
|
|
} else {
|
419 |
|
|
return(CORD_substr_closure(x, i, n, CORD_apply_access_fn));
|
420 |
|
|
}
|
421 |
|
|
} else {
|
422 |
|
|
char * result;
|
423 |
|
|
register struct Function * f = &(((CordRep *)x) -> function);
|
424 |
|
|
char buf[SUBSTR_LIMIT+1];
|
425 |
|
|
register char * p = buf;
|
426 |
|
|
register char c;
|
427 |
|
|
register int j;
|
428 |
|
|
register int lim = i + n;
|
429 |
|
|
|
430 |
|
|
for (j = i; j < lim; j++) {
|
431 |
|
|
c = (*(f -> fn))(j, f -> client_data);
|
432 |
|
|
if (c == '\0') {
|
433 |
|
|
return(CORD_substr_closure(x, i, n, CORD_apply_access_fn));
|
434 |
|
|
}
|
435 |
|
|
*p++ = c;
|
436 |
|
|
}
|
437 |
|
|
*p = '\0';
|
438 |
|
|
result = GC_MALLOC_ATOMIC(n+1);
|
439 |
|
|
if (result == 0) OUT_OF_MEMORY;
|
440 |
|
|
strcpy(result, buf);
|
441 |
|
|
return(result);
|
442 |
|
|
}
|
443 |
|
|
}
|
444 |
|
|
}
|
445 |
|
|
|
446 |
|
|
CORD CORD_substr(CORD x, size_t i, size_t n)
|
447 |
|
|
{
|
448 |
|
|
register size_t len = CORD_len(x);
|
449 |
|
|
|
450 |
|
|
if (i >= len || n <= 0) return(0);
|
451 |
|
|
/* n < 0 is impossible in a correct C implementation, but */
|
452 |
|
|
/* quite possible under SunOS 4.X. */
|
453 |
|
|
if (i + n > len) n = len - i;
|
454 |
|
|
# ifndef __STDC__
|
455 |
|
|
if (i < 0) ABORT("CORD_substr: second arg. negative");
|
456 |
|
|
/* Possible only if both client and C implementation are buggy. */
|
457 |
|
|
/* But empirically this happens frequently. */
|
458 |
|
|
# endif
|
459 |
|
|
return(CORD_substr_checked(x, i, n));
|
460 |
|
|
}
|
461 |
|
|
|
462 |
|
|
/* See cord.h for definition. We assume i is in range. */
|
463 |
|
|
int CORD_iter5(CORD x, size_t i, CORD_iter_fn f1,
|
464 |
|
|
CORD_batched_iter_fn f2, void * client_data)
|
465 |
|
|
{
|
466 |
|
|
if (x == 0) return(0);
|
467 |
|
|
if (CORD_IS_STRING(x)) {
|
468 |
|
|
register const char *p = x+i;
|
469 |
|
|
|
470 |
|
|
if (*p == '\0') ABORT("2nd arg to CORD_iter5 too big");
|
471 |
|
|
if (f2 != CORD_NO_FN) {
|
472 |
|
|
return((*f2)(p, client_data));
|
473 |
|
|
} else {
|
474 |
|
|
while (*p) {
|
475 |
|
|
if ((*f1)(*p, client_data)) return(1);
|
476 |
|
|
p++;
|
477 |
|
|
}
|
478 |
|
|
return(0);
|
479 |
|
|
}
|
480 |
|
|
} else if (IS_CONCATENATION(x)) {
|
481 |
|
|
register struct Concatenation * conc
|
482 |
|
|
= &(((CordRep *)x) -> concatenation);
|
483 |
|
|
|
484 |
|
|
|
485 |
|
|
if (i > 0) {
|
486 |
|
|
register size_t left_len = LEFT_LEN(conc);
|
487 |
|
|
|
488 |
|
|
if (i >= left_len) {
|
489 |
|
|
return(CORD_iter5(conc -> right, i - left_len, f1, f2,
|
490 |
|
|
client_data));
|
491 |
|
|
}
|
492 |
|
|
}
|
493 |
|
|
if (CORD_iter5(conc -> left, i, f1, f2, client_data)) {
|
494 |
|
|
return(1);
|
495 |
|
|
}
|
496 |
|
|
return(CORD_iter5(conc -> right, 0, f1, f2, client_data));
|
497 |
|
|
} else /* function */ {
|
498 |
|
|
register struct Function * f = &(((CordRep *)x) -> function);
|
499 |
|
|
register size_t j;
|
500 |
|
|
register size_t lim = f -> len;
|
501 |
|
|
|
502 |
|
|
for (j = i; j < lim; j++) {
|
503 |
|
|
if ((*f1)((*(f -> fn))(j, f -> client_data), client_data)) {
|
504 |
|
|
return(1);
|
505 |
|
|
}
|
506 |
|
|
}
|
507 |
|
|
return(0);
|
508 |
|
|
}
|
509 |
|
|
}
|
510 |
|
|
|
511 |
|
|
#undef CORD_iter
|
512 |
|
|
int CORD_iter(CORD x, CORD_iter_fn f1, void * client_data)
|
513 |
|
|
{
|
514 |
|
|
return(CORD_iter5(x, 0, f1, CORD_NO_FN, client_data));
|
515 |
|
|
}
|
516 |
|
|
|
517 |
|
|
int CORD_riter4(CORD x, size_t i, CORD_iter_fn f1, void * client_data)
|
518 |
|
|
{
|
519 |
|
|
if (x == 0) return(0);
|
520 |
|
|
if (CORD_IS_STRING(x)) {
|
521 |
|
|
register const char *p = x + i;
|
522 |
|
|
register char c;
|
523 |
|
|
|
524 |
|
|
for(;;) {
|
525 |
|
|
c = *p;
|
526 |
|
|
if (c == '\0') ABORT("2nd arg to CORD_riter4 too big");
|
527 |
|
|
if ((*f1)(c, client_data)) return(1);
|
528 |
|
|
if (p == x) break;
|
529 |
|
|
p--;
|
530 |
|
|
}
|
531 |
|
|
return(0);
|
532 |
|
|
} else if (IS_CONCATENATION(x)) {
|
533 |
|
|
register struct Concatenation * conc
|
534 |
|
|
= &(((CordRep *)x) -> concatenation);
|
535 |
|
|
register CORD left_part = conc -> left;
|
536 |
|
|
register size_t left_len;
|
537 |
|
|
|
538 |
|
|
left_len = LEFT_LEN(conc);
|
539 |
|
|
if (i >= left_len) {
|
540 |
|
|
if (CORD_riter4(conc -> right, i - left_len, f1, client_data)) {
|
541 |
|
|
return(1);
|
542 |
|
|
}
|
543 |
|
|
return(CORD_riter4(left_part, left_len - 1, f1, client_data));
|
544 |
|
|
} else {
|
545 |
|
|
return(CORD_riter4(left_part, i, f1, client_data));
|
546 |
|
|
}
|
547 |
|
|
} else /* function */ {
|
548 |
|
|
register struct Function * f = &(((CordRep *)x) -> function);
|
549 |
|
|
register size_t j;
|
550 |
|
|
|
551 |
|
|
for (j = i; ; j--) {
|
552 |
|
|
if ((*f1)((*(f -> fn))(j, f -> client_data), client_data)) {
|
553 |
|
|
return(1);
|
554 |
|
|
}
|
555 |
|
|
if (j == 0) return(0);
|
556 |
|
|
}
|
557 |
|
|
}
|
558 |
|
|
}
|
559 |
|
|
|
560 |
|
|
int CORD_riter(CORD x, CORD_iter_fn f1, void * client_data)
|
561 |
|
|
{
|
562 |
|
|
return(CORD_riter4(x, CORD_len(x) - 1, f1, client_data));
|
563 |
|
|
}
|
564 |
|
|
|
565 |
|
|
/*
|
566 |
|
|
* The following functions are concerned with balancing cords.
|
567 |
|
|
* Strategy:
|
568 |
|
|
* Scan the cord from left to right, keeping the cord scanned so far
|
569 |
|
|
* as a forest of balanced trees of exponentialy decreasing length.
|
570 |
|
|
* When a new subtree needs to be added to the forest, we concatenate all
|
571 |
|
|
* shorter ones to the new tree in the appropriate order, and then insert
|
572 |
|
|
* the result into the forest.
|
573 |
|
|
* Crucial invariants:
|
574 |
|
|
* 1. The concatenation of the forest (in decreasing order) with the
|
575 |
|
|
* unscanned part of the rope is equal to the rope being balanced.
|
576 |
|
|
* 2. All trees in the forest are balanced.
|
577 |
|
|
* 3. forest[i] has depth at most i.
|
578 |
|
|
*/
|
579 |
|
|
|
580 |
|
|
typedef struct {
|
581 |
|
|
CORD c;
|
582 |
|
|
size_t len; /* Actual length of c */
|
583 |
|
|
} ForestElement;
|
584 |
|
|
|
585 |
|
|
static size_t min_len [ MAX_DEPTH ];
|
586 |
|
|
|
587 |
|
|
static int min_len_init = 0;
|
588 |
|
|
|
589 |
|
|
int CORD_max_len;
|
590 |
|
|
|
591 |
|
|
typedef ForestElement Forest [ MAX_DEPTH ];
|
592 |
|
|
/* forest[i].len >= fib(i+1) */
|
593 |
|
|
/* The string is the concatenation */
|
594 |
|
|
/* of the forest in order of DECREASING */
|
595 |
|
|
/* indices. */
|
596 |
|
|
|
597 |
|
|
void CORD_init_min_len()
|
598 |
|
|
{
|
599 |
|
|
register int i;
|
600 |
|
|
register size_t last, previous, current;
|
601 |
|
|
|
602 |
|
|
min_len[0] = previous = 1;
|
603 |
|
|
min_len[1] = last = 2;
|
604 |
|
|
for (i = 2; i < MAX_DEPTH; i++) {
|
605 |
|
|
current = last + previous;
|
606 |
|
|
if (current < last) /* overflow */ current = last;
|
607 |
|
|
min_len[i] = current;
|
608 |
|
|
previous = last;
|
609 |
|
|
last = current;
|
610 |
|
|
}
|
611 |
|
|
CORD_max_len = last - 1;
|
612 |
|
|
min_len_init = 1;
|
613 |
|
|
}
|
614 |
|
|
|
615 |
|
|
|
616 |
|
|
void CORD_init_forest(ForestElement * forest, size_t max_len)
|
617 |
|
|
{
|
618 |
|
|
register int i;
|
619 |
|
|
|
620 |
|
|
for (i = 0; i < MAX_DEPTH; i++) {
|
621 |
|
|
forest[i].c = 0;
|
622 |
|
|
if (min_len[i] > max_len) return;
|
623 |
|
|
}
|
624 |
|
|
ABORT("Cord too long");
|
625 |
|
|
}
|
626 |
|
|
|
627 |
|
|
/* Add a leaf to the appropriate level in the forest, cleaning */
|
628 |
|
|
/* out lower levels as necessary. */
|
629 |
|
|
/* Also works if x is a balanced tree of concatenations; however */
|
630 |
|
|
/* in this case an extra concatenation node may be inserted above x; */
|
631 |
|
|
/* This node should not be counted in the statement of the invariants. */
|
632 |
|
|
void CORD_add_forest(ForestElement * forest, CORD x, size_t len)
|
633 |
|
|
{
|
634 |
|
|
register int i = 0;
|
635 |
|
|
register CORD sum = CORD_EMPTY;
|
636 |
|
|
register size_t sum_len = 0;
|
637 |
|
|
|
638 |
|
|
while (len > min_len[i + 1]) {
|
639 |
|
|
if (forest[i].c != 0) {
|
640 |
|
|
sum = CORD_cat(forest[i].c, sum);
|
641 |
|
|
sum_len += forest[i].len;
|
642 |
|
|
forest[i].c = 0;
|
643 |
|
|
}
|
644 |
|
|
i++;
|
645 |
|
|
}
|
646 |
|
|
/* Sum has depth at most 1 greter than what would be required */
|
647 |
|
|
/* for balance. */
|
648 |
|
|
sum = CORD_cat(sum, x);
|
649 |
|
|
sum_len += len;
|
650 |
|
|
/* If x was a leaf, then sum is now balanced. To see this */
|
651 |
|
|
/* consider the two cases in which forest[i-1] either is or is */
|
652 |
|
|
/* not empty. */
|
653 |
|
|
while (sum_len >= min_len[i]) {
|
654 |
|
|
if (forest[i].c != 0) {
|
655 |
|
|
sum = CORD_cat(forest[i].c, sum);
|
656 |
|
|
sum_len += forest[i].len;
|
657 |
|
|
/* This is again balanced, since sum was balanced, and has */
|
658 |
|
|
/* allowable depth that differs from i by at most 1. */
|
659 |
|
|
forest[i].c = 0;
|
660 |
|
|
}
|
661 |
|
|
i++;
|
662 |
|
|
}
|
663 |
|
|
i--;
|
664 |
|
|
forest[i].c = sum;
|
665 |
|
|
forest[i].len = sum_len;
|
666 |
|
|
}
|
667 |
|
|
|
668 |
|
|
CORD CORD_concat_forest(ForestElement * forest, size_t expected_len)
|
669 |
|
|
{
|
670 |
|
|
register int i = 0;
|
671 |
|
|
CORD sum = 0;
|
672 |
|
|
size_t sum_len = 0;
|
673 |
|
|
|
674 |
|
|
while (sum_len != expected_len) {
|
675 |
|
|
if (forest[i].c != 0) {
|
676 |
|
|
sum = CORD_cat(forest[i].c, sum);
|
677 |
|
|
sum_len += forest[i].len;
|
678 |
|
|
}
|
679 |
|
|
i++;
|
680 |
|
|
}
|
681 |
|
|
return(sum);
|
682 |
|
|
}
|
683 |
|
|
|
684 |
|
|
/* Insert the frontier of x into forest. Balanced subtrees are */
|
685 |
|
|
/* treated as leaves. This potentially adds one to the depth */
|
686 |
|
|
/* of the final tree. */
|
687 |
|
|
void CORD_balance_insert(CORD x, size_t len, ForestElement * forest)
|
688 |
|
|
{
|
689 |
|
|
register int depth;
|
690 |
|
|
|
691 |
|
|
if (CORD_IS_STRING(x)) {
|
692 |
|
|
CORD_add_forest(forest, x, len);
|
693 |
|
|
} else if (IS_CONCATENATION(x)
|
694 |
|
|
&& ((depth = DEPTH(x)) >= MAX_DEPTH
|
695 |
|
|
|| len < min_len[depth])) {
|
696 |
|
|
register struct Concatenation * conc
|
697 |
|
|
= &(((CordRep *)x) -> concatenation);
|
698 |
|
|
size_t left_len = LEFT_LEN(conc);
|
699 |
|
|
|
700 |
|
|
CORD_balance_insert(conc -> left, left_len, forest);
|
701 |
|
|
CORD_balance_insert(conc -> right, len - left_len, forest);
|
702 |
|
|
} else /* function or balanced */ {
|
703 |
|
|
CORD_add_forest(forest, x, len);
|
704 |
|
|
}
|
705 |
|
|
}
|
706 |
|
|
|
707 |
|
|
|
708 |
|
|
CORD CORD_balance(CORD x)
|
709 |
|
|
{
|
710 |
|
|
Forest forest;
|
711 |
|
|
register size_t len;
|
712 |
|
|
|
713 |
|
|
if (x == 0) return(0);
|
714 |
|
|
if (CORD_IS_STRING(x)) return(x);
|
715 |
|
|
if (!min_len_init) CORD_init_min_len();
|
716 |
|
|
len = LEN(x);
|
717 |
|
|
CORD_init_forest(forest, len);
|
718 |
|
|
CORD_balance_insert(x, len, forest);
|
719 |
|
|
return(CORD_concat_forest(forest, len));
|
720 |
|
|
}
|
721 |
|
|
|
722 |
|
|
|
723 |
|
|
/* Position primitives */
|
724 |
|
|
|
725 |
|
|
/* Private routines to deal with the hard cases only: */
|
726 |
|
|
|
727 |
|
|
/* P contains a prefix of the path to cur_pos. Extend it to a full */
|
728 |
|
|
/* path and set up leaf info. */
|
729 |
|
|
/* Return 0 if past the end of cord, 1 o.w. */
|
730 |
|
|
void CORD__extend_path(register CORD_pos p)
|
731 |
|
|
{
|
732 |
|
|
register struct CORD_pe * current_pe = &(p[0].path[p[0].path_len]);
|
733 |
|
|
register CORD top = current_pe -> pe_cord;
|
734 |
|
|
register size_t pos = p[0].cur_pos;
|
735 |
|
|
register size_t top_pos = current_pe -> pe_start_pos;
|
736 |
|
|
register size_t top_len = GEN_LEN(top);
|
737 |
|
|
|
738 |
|
|
/* Fill in the rest of the path. */
|
739 |
|
|
while(!CORD_IS_STRING(top) && IS_CONCATENATION(top)) {
|
740 |
|
|
register struct Concatenation * conc =
|
741 |
|
|
&(((CordRep *)top) -> concatenation);
|
742 |
|
|
register size_t left_len;
|
743 |
|
|
|
744 |
|
|
left_len = LEFT_LEN(conc);
|
745 |
|
|
current_pe++;
|
746 |
|
|
if (pos >= top_pos + left_len) {
|
747 |
|
|
current_pe -> pe_cord = top = conc -> right;
|
748 |
|
|
current_pe -> pe_start_pos = top_pos = top_pos + left_len;
|
749 |
|
|
top_len -= left_len;
|
750 |
|
|
} else {
|
751 |
|
|
current_pe -> pe_cord = top = conc -> left;
|
752 |
|
|
current_pe -> pe_start_pos = top_pos;
|
753 |
|
|
top_len = left_len;
|
754 |
|
|
}
|
755 |
|
|
p[0].path_len++;
|
756 |
|
|
}
|
757 |
|
|
/* Fill in leaf description for fast access. */
|
758 |
|
|
if (CORD_IS_STRING(top)) {
|
759 |
|
|
p[0].cur_leaf = top;
|
760 |
|
|
p[0].cur_start = top_pos;
|
761 |
|
|
p[0].cur_end = top_pos + top_len;
|
762 |
|
|
} else {
|
763 |
|
|
p[0].cur_end = 0;
|
764 |
|
|
}
|
765 |
|
|
if (pos >= top_pos + top_len) p[0].path_len = CORD_POS_INVALID;
|
766 |
|
|
}
|
767 |
|
|
|
768 |
|
|
char CORD__pos_fetch(register CORD_pos p)
|
769 |
|
|
{
|
770 |
|
|
/* Leaf is a function node */
|
771 |
|
|
struct CORD_pe * pe = &((p)[0].path[(p)[0].path_len]);
|
772 |
|
|
CORD leaf = pe -> pe_cord;
|
773 |
|
|
register struct Function * f = &(((CordRep *)leaf) -> function);
|
774 |
|
|
|
775 |
|
|
if (!IS_FUNCTION(leaf)) ABORT("CORD_pos_fetch: bad leaf");
|
776 |
|
|
return ((*(f -> fn))(p[0].cur_pos - pe -> pe_start_pos, f -> client_data));
|
777 |
|
|
}
|
778 |
|
|
|
779 |
|
|
void CORD__next(register CORD_pos p)
|
780 |
|
|
{
|
781 |
|
|
register size_t cur_pos = p[0].cur_pos + 1;
|
782 |
|
|
register struct CORD_pe * current_pe = &((p)[0].path[(p)[0].path_len]);
|
783 |
|
|
register CORD leaf = current_pe -> pe_cord;
|
784 |
|
|
|
785 |
|
|
/* Leaf is not a string or we're at end of leaf */
|
786 |
|
|
p[0].cur_pos = cur_pos;
|
787 |
|
|
if (!CORD_IS_STRING(leaf)) {
|
788 |
|
|
/* Function leaf */
|
789 |
|
|
register struct Function * f = &(((CordRep *)leaf) -> function);
|
790 |
|
|
register size_t start_pos = current_pe -> pe_start_pos;
|
791 |
|
|
register size_t end_pos = start_pos + f -> len;
|
792 |
|
|
|
793 |
|
|
if (cur_pos < end_pos) {
|
794 |
|
|
/* Fill cache and return. */
|
795 |
|
|
register size_t i;
|
796 |
|
|
register size_t limit = cur_pos + FUNCTION_BUF_SZ;
|
797 |
|
|
register CORD_fn fn = f -> fn;
|
798 |
|
|
register void * client_data = f -> client_data;
|
799 |
|
|
|
800 |
|
|
if (limit > end_pos) {
|
801 |
|
|
limit = end_pos;
|
802 |
|
|
}
|
803 |
|
|
for (i = cur_pos; i < limit; i++) {
|
804 |
|
|
p[0].function_buf[i - cur_pos] =
|
805 |
|
|
(*fn)(i - start_pos, client_data);
|
806 |
|
|
}
|
807 |
|
|
p[0].cur_start = cur_pos;
|
808 |
|
|
p[0].cur_leaf = p[0].function_buf;
|
809 |
|
|
p[0].cur_end = limit;
|
810 |
|
|
return;
|
811 |
|
|
}
|
812 |
|
|
}
|
813 |
|
|
/* End of leaf */
|
814 |
|
|
/* Pop the stack until we find two concatenation nodes with the */
|
815 |
|
|
/* same start position: this implies we were in left part. */
|
816 |
|
|
{
|
817 |
|
|
while (p[0].path_len > 0
|
818 |
|
|
&& current_pe[0].pe_start_pos != current_pe[-1].pe_start_pos) {
|
819 |
|
|
p[0].path_len--;
|
820 |
|
|
current_pe--;
|
821 |
|
|
}
|
822 |
|
|
if (p[0].path_len == 0) {
|
823 |
|
|
p[0].path_len = CORD_POS_INVALID;
|
824 |
|
|
return;
|
825 |
|
|
}
|
826 |
|
|
}
|
827 |
|
|
p[0].path_len--;
|
828 |
|
|
CORD__extend_path(p);
|
829 |
|
|
}
|
830 |
|
|
|
831 |
|
|
void CORD__prev(register CORD_pos p)
|
832 |
|
|
{
|
833 |
|
|
register struct CORD_pe * pe = &(p[0].path[p[0].path_len]);
|
834 |
|
|
|
835 |
|
|
if (p[0].cur_pos == 0) {
|
836 |
|
|
p[0].path_len = CORD_POS_INVALID;
|
837 |
|
|
return;
|
838 |
|
|
}
|
839 |
|
|
p[0].cur_pos--;
|
840 |
|
|
if (p[0].cur_pos >= pe -> pe_start_pos) return;
|
841 |
|
|
|
842 |
|
|
/* Beginning of leaf */
|
843 |
|
|
|
844 |
|
|
/* Pop the stack until we find two concatenation nodes with the */
|
845 |
|
|
/* different start position: this implies we were in right part. */
|
846 |
|
|
{
|
847 |
|
|
register struct CORD_pe * current_pe = &((p)[0].path[(p)[0].path_len]);
|
848 |
|
|
|
849 |
|
|
while (p[0].path_len > 0
|
850 |
|
|
&& current_pe[0].pe_start_pos == current_pe[-1].pe_start_pos) {
|
851 |
|
|
p[0].path_len--;
|
852 |
|
|
current_pe--;
|
853 |
|
|
}
|
854 |
|
|
}
|
855 |
|
|
p[0].path_len--;
|
856 |
|
|
CORD__extend_path(p);
|
857 |
|
|
}
|
858 |
|
|
|
859 |
|
|
#undef CORD_pos_fetch
|
860 |
|
|
#undef CORD_next
|
861 |
|
|
#undef CORD_prev
|
862 |
|
|
#undef CORD_pos_to_index
|
863 |
|
|
#undef CORD_pos_to_cord
|
864 |
|
|
#undef CORD_pos_valid
|
865 |
|
|
|
866 |
|
|
char CORD_pos_fetch(register CORD_pos p)
|
867 |
|
|
{
|
868 |
|
|
if (p[0].cur_start <= p[0].cur_pos && p[0].cur_pos < p[0].cur_end) {
|
869 |
|
|
return(p[0].cur_leaf[p[0].cur_pos - p[0].cur_start]);
|
870 |
|
|
} else {
|
871 |
|
|
return(CORD__pos_fetch(p));
|
872 |
|
|
}
|
873 |
|
|
}
|
874 |
|
|
|
875 |
|
|
void CORD_next(CORD_pos p)
|
876 |
|
|
{
|
877 |
|
|
if (p[0].cur_pos < p[0].cur_end - 1) {
|
878 |
|
|
p[0].cur_pos++;
|
879 |
|
|
} else {
|
880 |
|
|
CORD__next(p);
|
881 |
|
|
}
|
882 |
|
|
}
|
883 |
|
|
|
884 |
|
|
void CORD_prev(CORD_pos p)
|
885 |
|
|
{
|
886 |
|
|
if (p[0].cur_end != 0 && p[0].cur_pos > p[0].cur_start) {
|
887 |
|
|
p[0].cur_pos--;
|
888 |
|
|
} else {
|
889 |
|
|
CORD__prev(p);
|
890 |
|
|
}
|
891 |
|
|
}
|
892 |
|
|
|
893 |
|
|
size_t CORD_pos_to_index(CORD_pos p)
|
894 |
|
|
{
|
895 |
|
|
return(p[0].cur_pos);
|
896 |
|
|
}
|
897 |
|
|
|
898 |
|
|
CORD CORD_pos_to_cord(CORD_pos p)
|
899 |
|
|
{
|
900 |
|
|
return(p[0].path[0].pe_cord);
|
901 |
|
|
}
|
902 |
|
|
|
903 |
|
|
int CORD_pos_valid(CORD_pos p)
|
904 |
|
|
{
|
905 |
|
|
return(p[0].path_len != CORD_POS_INVALID);
|
906 |
|
|
}
|
907 |
|
|
|
908 |
|
|
void CORD_set_pos(CORD_pos p, CORD x, size_t i)
|
909 |
|
|
{
|
910 |
|
|
if (x == CORD_EMPTY) {
|
911 |
|
|
p[0].path_len = CORD_POS_INVALID;
|
912 |
|
|
return;
|
913 |
|
|
}
|
914 |
|
|
p[0].path[0].pe_cord = x;
|
915 |
|
|
p[0].path[0].pe_start_pos = 0;
|
916 |
|
|
p[0].path_len = 0;
|
917 |
|
|
p[0].cur_pos = i;
|
918 |
|
|
CORD__extend_path(p);
|
919 |
|
|
}
|