hash.c revision 1.34
1 /* $NetBSD: hash.c,v 1.34 2020/09/27 21:35:16 rillig 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 /* 73 * This module contains routines to manipulate a hash table. 74 * See hash.h for a definition of the structure of the hash 75 * table. Hash tables grow automatically as the amount of 76 * information increases. 77 */ 78 79 #include "make.h" 80 81 /* "@(#)hash.c 8.1 (Berkeley) 6/6/93" */ 82 MAKE_RCSID("$NetBSD: hash.c,v 1.34 2020/09/27 21:35:16 rillig Exp $"); 83 84 /* 85 * Forward references to local procedures that are used before they're 86 * defined: 87 */ 88 89 static void RebuildTable(Hash_Table *); 90 91 /* 92 * The following defines the ratio of # entries to # buckets 93 * at which we rebuild the table to make it larger. 94 */ 95 96 #define rebuildLimit 3 97 98 /* The hash function(s) */ 99 100 #ifndef HASH 101 /* The default: this one matches Gosling's emacs */ 102 #define HASH(h, key, p) do { \ 103 for (h = 0, p = key; *p;) \ 104 h = (h << 5) - h + (unsigned char)*p++; \ 105 } while (0) 106 107 #endif 108 109 /* Sets up the hash table. 110 * 111 * Input: 112 * t Structure to to hold the table. 113 */ 114 void 115 Hash_InitTable(Hash_Table *t) 116 { 117 size_t n = 16, i; 118 struct Hash_Entry **hp; 119 120 t->numEntries = 0; 121 t->maxchain = 0; 122 t->bucketsSize = n; 123 t->bucketsMask = n - 1; 124 t->buckets = hp = bmake_malloc(sizeof(*hp) * n); 125 for (i = 0; i < n; i++) 126 hp[i] = NULL; 127 } 128 129 /* Removes everything from the hash table and frees up the memory space it 130 * occupied (except for the space in the Hash_Table structure). */ 131 void 132 Hash_DeleteTable(Hash_Table *t) 133 { 134 struct Hash_Entry **hp, *h, *nexth = NULL; 135 int i; 136 137 for (hp = t->buckets, i = t->bucketsSize; --i >= 0;) { 138 for (h = *hp++; h != NULL; h = nexth) { 139 nexth = h->next; 140 free(h); 141 } 142 } 143 free(t->buckets); 144 145 /* 146 * Set up the hash table to cause memory faults on any future access 147 * attempts until re-initialization. 148 */ 149 t->buckets = NULL; 150 } 151 152 /* Searches the hash table for an entry corresponding to the key. 153 * 154 * Input: 155 * t Hash table to search. 156 * key A hash key. 157 * 158 * Results: 159 * Returns a pointer to the entry for key, or NULL if the table contains 160 * no entry for the key. 161 */ 162 Hash_Entry * 163 Hash_FindEntry(Hash_Table *t, const char *key) 164 { 165 Hash_Entry *e; 166 unsigned h; 167 const char *p; 168 int chainlen; 169 170 if (t == NULL || t->buckets == NULL) { 171 return NULL; 172 } 173 HASH(h, key, p); 174 p = key; 175 chainlen = 0; 176 #ifdef DEBUG_HASH_LOOKUP 177 if (DEBUG(HASH)) 178 fprintf(debug_file, "%s: %p h=%x key=%s\n", __func__, 179 t, h, key); 180 #endif 181 for (e = t->buckets[h & t->bucketsMask]; e != NULL; e = e->next) { 182 chainlen++; 183 if (e->namehash == h && strcmp(e->name, p) == 0) 184 break; 185 } 186 if (chainlen > t->maxchain) 187 t->maxchain = chainlen; 188 return e; 189 } 190 191 void * 192 Hash_FindValue(Hash_Table *t, const char *key) 193 { 194 Hash_Entry *he = Hash_FindEntry(t, key); 195 return he != NULL ? he->value : NULL; 196 } 197 198 /* Searches the hash table for an entry corresponding to the key. 199 * If no entry is found, then one is created. 200 * 201 * Input: 202 * t Hash table to search. 203 * key A hash key. 204 * newPtr Filled with TRUE if new entry created, 205 * FALSE otherwise. 206 */ 207 Hash_Entry * 208 Hash_CreateEntry(Hash_Table *t, const char *key, Boolean *newPtr) 209 { 210 Hash_Entry *e; 211 unsigned h; 212 const char *p; 213 int keylen; 214 int chainlen; 215 struct Hash_Entry **hp; 216 217 /* 218 * Hash the key. As a side effect, save the length (strlen) of the 219 * key in case we need to create the entry. 220 */ 221 HASH(h, key, p); 222 keylen = p - key; 223 p = key; 224 chainlen = 0; 225 #ifdef DEBUG_HASH_LOOKUP 226 if (DEBUG(HASH)) 227 fprintf(debug_file, "%s: %p h=%x key=%s\n", __func__, 228 t, h, key); 229 #endif 230 for (e = t->buckets[h & t->bucketsMask]; e != NULL; e = e->next) { 231 chainlen++; 232 if (e->namehash == h && strcmp(e->name, p) == 0) { 233 if (newPtr != NULL) 234 *newPtr = FALSE; 235 break; 236 } 237 } 238 if (chainlen > t->maxchain) 239 t->maxchain = chainlen; 240 if (e) 241 return e; 242 243 /* 244 * The desired entry isn't there. Before allocating a new entry, 245 * expand the table if necessary (and this changes the resulting 246 * bucket chain). 247 */ 248 if (t->numEntries >= rebuildLimit * t->bucketsSize) 249 RebuildTable(t); 250 e = bmake_malloc(sizeof(*e) + keylen); 251 hp = &t->buckets[h & t->bucketsMask]; 252 e->next = *hp; 253 *hp = e; 254 Hash_SetValue(e, NULL); 255 e->namehash = h; 256 (void)strcpy(e->name, p); 257 t->numEntries++; 258 259 if (newPtr != NULL) 260 *newPtr = TRUE; 261 return e; 262 } 263 264 /* Delete the given hash table entry and free memory associated with it. */ 265 void 266 Hash_DeleteEntry(Hash_Table *t, Hash_Entry *e) 267 { 268 Hash_Entry **hp, *p; 269 270 if (e == NULL) 271 return; 272 for (hp = &t->buckets[e->namehash & t->bucketsMask]; 273 (p = *hp) != NULL; hp = &p->next) { 274 if (p == e) { 275 *hp = p->next; 276 free(p); 277 t->numEntries--; 278 return; 279 } 280 } 281 (void)write(2, "bad call to Hash_DeleteEntry\n", 29); 282 abort(); 283 } 284 285 /* Sets things up for enumerating all entries in the hash table. 286 * 287 * Input: 288 * t Table to be searched. 289 * searchPtr Area in which to keep state about search. 290 * 291 * Results: 292 * The return value is the address of the first entry in 293 * the hash table, or NULL if the table is empty. 294 */ 295 Hash_Entry * 296 Hash_EnumFirst(Hash_Table *t, Hash_Search *searchPtr) 297 { 298 searchPtr->table = t; 299 searchPtr->nextBucket = 0; 300 searchPtr->entry = NULL; 301 return Hash_EnumNext(searchPtr); 302 } 303 304 /* Returns the next entry in the hash table, or NULL if the end of the table 305 * is reached. 306 * 307 * Input: 308 * searchPtr Area used to keep state about search. 309 */ 310 Hash_Entry * 311 Hash_EnumNext(Hash_Search *searchPtr) 312 { 313 Hash_Entry *e; 314 Hash_Table *t = searchPtr->table; 315 316 /* 317 * The entry field points to the most recently returned 318 * entry, or is NULL if we are starting up. If not NULL, we have 319 * to start at the next one in the chain. 320 */ 321 e = searchPtr->entry; 322 if (e != NULL) 323 e = e->next; 324 /* 325 * If the chain ran out, or if we are starting up, we need to 326 * find the next nonempty chain. 327 */ 328 while (e == NULL) { 329 if (searchPtr->nextBucket >= t->bucketsSize) 330 return NULL; 331 e = t->buckets[searchPtr->nextBucket++]; 332 } 333 searchPtr->entry = e; 334 return e; 335 } 336 337 /* Makes a new hash table that is larger than the old one. The entire hash 338 * table is moved, so any bucket numbers from the old table become invalid. */ 339 static void 340 RebuildTable(Hash_Table *t) 341 { 342 Hash_Entry *e, *next = NULL, **hp, **xp; 343 int i, mask; 344 Hash_Entry **oldhp; 345 int oldsize; 346 347 oldhp = t->buckets; 348 oldsize = i = t->bucketsSize; 349 i <<= 1; 350 t->bucketsSize = i; 351 t->bucketsMask = mask = i - 1; 352 t->buckets = hp = bmake_malloc(sizeof(*hp) * i); 353 while (--i >= 0) 354 *hp++ = NULL; 355 for (hp = oldhp, i = oldsize; --i >= 0;) { 356 for (e = *hp++; e != NULL; e = next) { 357 next = e->next; 358 xp = &t->buckets[e->namehash & mask]; 359 e->next = *xp; 360 *xp = e; 361 } 362 } 363 free(oldhp); 364 if (DEBUG(HASH)) 365 fprintf(debug_file, "%s: %p size=%d entries=%d maxchain=%d\n", 366 __func__, t, t->bucketsSize, t->numEntries, t->maxchain); 367 t->maxchain = 0; 368 } 369 370 void 371 Hash_ForEach(Hash_Table *t, void (*action)(void *, void *), void *data) 372 { 373 Hash_Search search; 374 Hash_Entry *e; 375 376 for (e = Hash_EnumFirst(t, &search); 377 e != NULL; 378 e = Hash_EnumNext(&search)) 379 action(Hash_GetValue(e), data); 380 } 381 382 void 383 Hash_DebugStats(Hash_Table *t, const char *name) 384 { 385 if (DEBUG(HASH)) 386 fprintf(debug_file, "Hash_Table %s: size=%d numEntries=%d maxchain=%d\n", 387 name, t->bucketsSize, t->numEntries, t->maxchain); 388 } 389
Indexes created Mon Sep 29 15:09:53 GMT 2025