hash.c revision 1.3 1 /* $NetBSD: hash.c,v 1.3 2003/08/07 10:04:37 agc Exp $ */
2
3 /*
4 * Copyright (c) 1988, 1989, 1990 The Regents of the University of California.
5 * All rights reserved.
6 *
7 * This code is derived from software contributed to Berkeley by
8 * Adam de Boor.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 * 3. Neither the name of the University nor the names of its contributors
19 * may be used to endorse or promote products derived from this software
20 * without specific prior written permission.
21 *
22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 * SUCH DAMAGE.
33 */
34
35 /*
36 * Copyright (c) 1988, 1989 by Adam de Boor
37 * Copyright (c) 1989 by Berkeley Softworks
38 * All rights reserved.
39 *
40 * This code is derived from software contributed to Berkeley by
41 * Adam de Boor.
42 *
43 * Redistribution and use in source and binary forms, with or without
44 * modification, are permitted provided that the following conditions
45 * are met:
46 * 1. Redistributions of source code must retain the above copyright
47 * notice, this list of conditions and the following disclaimer.
48 * 2. Redistributions in binary form must reproduce the above copyright
49 * notice, this list of conditions and the following disclaimer in the
50 * documentation and/or other materials provided with the distribution.
51 * 3. All advertising materials mentioning features or use of this software
52 * must display the following acknowledgement:
53 * This product includes software developed by the University of
54 * California, Berkeley and its contributors.
55 * 4. Neither the name of the University nor the names of its contributors
56 * may be used to endorse or promote products derived from this software
57 * without specific prior written permission.
58 *
59 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
60 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
61 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
62 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
63 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
64 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
65 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
66 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
67 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
68 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
69 * SUCH DAMAGE.
70 */
71
72 #ifdef MAKE_BOOTSTRAP
73 static char rcsid[] = "$NetBSD: hash.c,v 1.3 2003/08/07 10:04:37 agc Exp $";
74 #else
75 #include <sys/cdefs.h>
76 #ifndef lint
77 #if 0
78 static char sccsid[] = "@(#)hash.c 8.1 (Berkeley) 6/6/93";
79 #else
80 __RCSID("$NetBSD: hash.c,v 1.3 2003/08/07 10:04:37 agc Exp $");
81 #endif
82 #endif /* not lint */
83 #endif
84
85 #include <sys/types.h>
86
87 #include <stdlib.h>
88 #include <string.h>
89 #include <unistd.h>
90
91 /* hash.c --
92 *
93 * This module contains routines to manipulate a hash table.
94 * See hash.h for a definition of the structure of the hash
95 * table. Hash tables grow automatically as the amount of
96 * information increases.
97 */
98 #include "hash.h"
99 #include "ealloc.h"
100
101 /*
102 * Forward references to local procedures that are used before they're
103 * defined:
104 */
105
106 static void RebuildTable(Hash_Table *);
107
108 /*
109 * The following defines the ratio of # entries to # buckets
110 * at which we rebuild the table to make it larger.
111 */
112
113 #define rebuildLimit 8
114
115 /*
116 *---------------------------------------------------------
117 *
118 * Hash_InitTable --
119 *
120 * This routine just sets up the hash table.
121 *
122 * Input:
123 * t Structure to use to hold table.
124 * numBuckets How many buckets to create for starters. This number
125 * is rounded up to a power of two. If <= 0, a reasonable
126 * default is chosen. The table will grow in size later
127 * as needed.
128 *
129 * Results:
130 * None.
131 *
132 * Side Effects:
133 * Memory is allocated for the initial bucket area.
134 *
135 *---------------------------------------------------------
136 */
137
138 void
139 Hash_InitTable(Hash_Table *t, int numBuckets)
140 {
141 int i;
142 struct Hash_Entry **hp;
143
144 /*
145 * Round up the size to a power of two.
146 */
147 if (numBuckets <= 0)
148 i = 16;
149 else {
150 for (i = 2; i < numBuckets; i <<= 1)
151 continue;
152 }
153 t->numEntries = 0;
154 t->size = i;
155 t->mask = i - 1;
156 t->bucketPtr = hp = (struct Hash_Entry **)emalloc(sizeof(*hp) * i);
157 while (--i >= 0)
158 *hp++ = NULL;
159 }
160
161 /*
162 *---------------------------------------------------------
163 *
164 * Hash_DeleteTable --
165 *
166 * This routine removes everything from a hash table
167 * and frees up the memory space it occupied (except for
168 * the space in the Hash_Table structure).
169 *
170 * Results:
171 * None.
172 *
173 * Side Effects:
174 * Lots of memory is freed up.
175 *
176 *---------------------------------------------------------
177 */
178
179 void
180 Hash_DeleteTable(Hash_Table *t)
181 {
182 struct Hash_Entry **hp, *h, *nexth;
183 int i;
184
185 nexth = NULL;
186 for (hp = t->bucketPtr, i = t->size; --i >= 0;) {
187 for (h = *hp++; h != NULL; h = nexth) {
188 nexth = h->next;
189 free((char *)h);
190 }
191 }
192 free((char *)t->bucketPtr);
193
194 /*
195 * Set up the hash table to cause memory faults on any future access
196 * attempts until re-initialization.
197 */
198 t->bucketPtr = NULL;
199 }
200
201 /*
202 *---------------------------------------------------------
203 *
204 * Hash_FindEntry --
205 *
206 * Searches a hash table for an entry corresponding to key.
207 *
208 * Input:
209 * t Hash table to search.
210 * key A hash key.
211 *
212 * Results:
213 * The return value is a pointer to the entry for key,
214 * if key was present in the table. If key was not
215 * present, NULL is returned.
216 *
217 * Side Effects:
218 * None.
219 *
220 *---------------------------------------------------------
221 */
222
223 Hash_Entry *
224 Hash_FindEntry(Hash_Table *t, char *key)
225 {
226 Hash_Entry *e;
227 unsigned h;
228 char *p;
229
230 for (h = 0, p = key; *p;)
231 h = (h << 5) - h + *p++;
232 p = key;
233 for (e = t->bucketPtr[h & t->mask]; e != NULL; e = e->next)
234 if (e->namehash == h && strcmp(e->name, p) == 0)
235 return (e);
236 return (NULL);
237 }
238
239 /*
240 *---------------------------------------------------------
241 *
242 * Hash_CreateEntry --
243 *
244 * Searches a hash table for an entry corresponding to
245 * key. If no entry is found, then one is created.
246 *
247 * Input:
248 * t Hash table to search.
249 * key A hash key.
250 * newPtr Filled in with 1 if new entry created, 0 otherwise.
251 *
252 * Results:
253 * The return value is a pointer to the entry. If *newPtr
254 * isn't NULL, then *newPtr is filled in with TRUE if a
255 * new entry was created, and FALSE if an entry already existed
256 * with the given key.
257 *
258 * Side Effects:
259 * Memory may be allocated, and the hash buckets may be modified.
260 *---------------------------------------------------------
261 */
262
263 Hash_Entry *
264 Hash_CreateEntry(Hash_Table *t, char *key, int *newPtr)
265 {
266 Hash_Entry *e;
267 unsigned h;
268 char *p;
269 int keylen;
270 struct Hash_Entry **hp;
271
272 /*
273 * Hash the key. As a side effect, save the length (strlen) of the
274 * key in case we need to create the entry.
275 */
276 for (h = 0, p = key; *p;)
277 h = (h << 5) - h + *p++;
278 keylen = p - key;
279 p = key;
280 for (e = t->bucketPtr[h & t->mask]; e != NULL; e = e->next) {
281 if (e->namehash == h && strcmp(e->name, p) == 0) {
282 if (newPtr != NULL)
283 *newPtr = 0;
284 return (e);
285 }
286 }
287
288 /*
289 * The desired entry isn't there. Before allocating a new entry,
290 * expand the table if necessary (and this changes the resulting
291 * bucket chain).
292 */
293 if (t->numEntries >= rebuildLimit * t->size)
294 RebuildTable(t);
295 e = (Hash_Entry *) emalloc(sizeof(*e) + keylen);
296 hp = &t->bucketPtr[h & t->mask];
297 e->next = *hp;
298 *hp = e;
299 e->clientData = NULL;
300 e->namehash = h;
301 (void) strcpy(e->name, p);
302 t->numEntries++;
303
304 if (newPtr != NULL)
305 *newPtr = 1;
306 return (e);
307 }
308
309 /*
310 *---------------------------------------------------------
311 *
312 * Hash_DeleteEntry --
313 *
314 * Delete the given hash table entry and free memory associated with
315 * it.
316 *
317 * Results:
318 * None.
319 *
320 * Side Effects:
321 * Hash chain that entry lives in is modified and memory is freed.
322 *
323 *---------------------------------------------------------
324 */
325
326 void
327 Hash_DeleteEntry(Hash_Table *t, Hash_Entry *e)
328 {
329 Hash_Entry **hp, *p;
330
331 if (e == NULL)
332 return;
333 for (hp = &t->bucketPtr[e->namehash & t->mask];
334 (p = *hp) != NULL; hp = &p->next) {
335 if (p == e) {
336 *hp = p->next;
337 free((char *)p);
338 t->numEntries--;
339 return;
340 }
341 }
342 (void)write(2, "bad call to Hash_DeleteEntry\n", 29);
343 abort();
344 }
345
346 /*
347 *---------------------------------------------------------
348 *
349 * Hash_EnumFirst --
350 * This procedure sets things up for a complete search
351 * of all entries recorded in the hash table.
352 *
353 * Input:
354 * t Table to be searched.
355 * searchPtr Area in which to keep state about search.
356 *
357 * Results:
358 * The return value is the address of the first entry in
359 * the hash table, or NULL if the table is empty.
360 *
361 * Side Effects:
362 * The information in searchPtr is initialized so that successive
363 * calls to Hash_Next will return successive HashEntry's
364 * from the table.
365 *
366 *---------------------------------------------------------
367 */
368
369 Hash_Entry *
370 Hash_EnumFirst(Hash_Table *t, Hash_Search *searchPtr)
371 {
372
373 searchPtr->tablePtr = t;
374 searchPtr->nextIndex = 0;
375 searchPtr->hashEntryPtr = NULL;
376 return Hash_EnumNext(searchPtr);
377 }
378
379 /*
380 *---------------------------------------------------------
381 *
382 * Hash_EnumNext --
383 * This procedure returns successive entries in the hash table.
384 *
385 * Results:
386 * The return value is a pointer to the next HashEntry
387 * in the table, or NULL when the end of the table is
388 * reached.
389 *
390 * Side Effects:
391 * The information in searchPtr is modified to advance to the
392 * next entry.
393 *
394 *---------------------------------------------------------
395 */
396
397 Hash_Entry *
398 Hash_EnumNext(Hash_Search *searchPtr)
399 {
400 Hash_Entry *e;
401 Hash_Table *t = searchPtr->tablePtr;
402
403 /*
404 * The hashEntryPtr field points to the most recently returned
405 * entry, or is nil if we are starting up. If not nil, we have
406 * to start at the next one in the chain.
407 */
408 e = searchPtr->hashEntryPtr;
409 if (e != NULL)
410 e = e->next;
411 /*
412 * If the chain ran out, or if we are starting up, we need to
413 * find the next nonempty chain.
414 */
415 while (e == NULL) {
416 if (searchPtr->nextIndex >= t->size)
417 return (NULL);
418 e = t->bucketPtr[searchPtr->nextIndex++];
419 }
420 searchPtr->hashEntryPtr = e;
421 return (e);
422 }
423
424 /*
425 *---------------------------------------------------------
426 *
427 * RebuildTable --
428 * This local routine makes a new hash table that
429 * is larger than the old one.
430 *
431 * Results:
432 * None.
433 *
434 * Side Effects:
435 * The entire hash table is moved, so any bucket numbers
436 * from the old table are invalid.
437 *
438 *---------------------------------------------------------
439 */
440
441 static void
442 RebuildTable(Hash_Table *t)
443 {
444 Hash_Entry *e, *next, **hp, **xp;
445 int i, mask;
446 Hash_Entry **oldhp;
447 int oldsize;
448
449 next = NULL;
450 oldhp = t->bucketPtr;
451 oldsize = i = t->size;
452 i <<= 1;
453 t->size = i;
454 t->mask = mask = i - 1;
455 t->bucketPtr = hp = (struct Hash_Entry **) emalloc(sizeof(*hp) * i);
456 while (--i >= 0)
457 *hp++ = NULL;
458 for (hp = oldhp, i = oldsize; --i >= 0;) {
459 for (e = *hp++; e != NULL; e = next) {
460 next = e->next;
461 xp = &t->bucketPtr[e->namehash & mask];
462 e->next = *xp;
463 *xp = e;
464 }
465 }
466 free((char *)oldhp);
467 }
468