hash.c revision 1.40
1 /* $NetBSD: hash.c,v 1.40 2020/10/04 17:50:41 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 1.40 2020/10/04 17:50:41 rillig Exp $"); 83 84 /* 85 * The ratio of # entries to # buckets at which we rebuild the table to 86 * make it larger. 87 */ 88 #define rebuildLimit 3 89 90 /* This hash function matches Gosling's emacs. */ 91 static unsigned int 92 hash(const char *key, size_t *out_keylen) 93 { 94 unsigned h = 0; 95 const char *p = key; 96 while (*p != '\0') 97 h = (h << 5) - h + (unsigned char)*p++; 98 if (out_keylen != NULL) 99 *out_keylen = (size_t)(p - key); 100 return h; 101 } 102 103 /* Sets up the hash table. */ 104 void 105 Hash_InitTable(Hash_Table *t) 106 { 107 size_t n = 16, i; 108 struct Hash_Entry **hp; 109 110 t->numEntries = 0; 111 t->maxchain = 0; 112 t->bucketsSize = n; 113 t->bucketsMask = n - 1; 114 t->buckets = hp = bmake_malloc(sizeof(*hp) * n); 115 for (i = 0; i < n; i++) 116 hp[i] = NULL; 117 } 118 119 /* Removes everything from the hash table and frees up the memory space it 120 * occupied (except for the space in the Hash_Table structure). */ 121 void 122 Hash_DeleteTable(Hash_Table *t) 123 { 124 struct Hash_Entry **hp, *h, *nexth = NULL; 125 int i; 126 127 for (hp = t->buckets, i = t->bucketsSize; --i >= 0;) { 128 for (h = *hp++; h != NULL; h = nexth) { 129 nexth = h->next; 130 free(h); 131 } 132 } 133 free(t->buckets); 134 135 /* 136 * Set up the hash table to cause memory faults on any future access 137 * attempts until re-initialization. 138 */ 139 t->buckets = NULL; 140 } 141 142 /* Searches the hash table for an entry corresponding to the key. 143 * 144 * Input: 145 * t Hash table to search. 146 * key A hash key. 147 * 148 * Results: 149 * Returns a pointer to the entry for key, or NULL if the table contains 150 * no entry for the key. 151 */ 152 Hash_Entry * 153 Hash_FindEntry(Hash_Table *t, const char *key) 154 { 155 Hash_Entry *e; 156 unsigned h; 157 int chainlen; 158 159 h = hash(key, NULL); 160 chainlen = 0; 161 #ifdef DEBUG_HASH_LOOKUP 162 DEBUG4(HASH, "%s: %p h=%x key=%s\n", __func__, t, h, key); 163 #endif 164 for (e = t->buckets[h & t->bucketsMask]; e != NULL; e = e->next) { 165 chainlen++; 166 if (e->namehash == h && strcmp(e->name, key) == 0) 167 break; 168 } 169 if (chainlen > t->maxchain) 170 t->maxchain = chainlen; 171 return e; 172 } 173 174 void * 175 Hash_FindValue(Hash_Table *t, const char *key) 176 { 177 Hash_Entry *he = Hash_FindEntry(t, key); 178 return he != NULL ? he->value : NULL; 179 } 180 181 /* Makes a new hash table that is larger than the old one. The entire hash 182 * table is moved, so any bucket numbers from the old table become invalid. */ 183 static void 184 RebuildTable(Hash_Table *t) 185 { 186 Hash_Entry *e, *next = NULL, **hp, **xp; 187 int i, mask; 188 Hash_Entry **oldhp; 189 int oldsize; 190 191 oldhp = t->buckets; 192 oldsize = i = t->bucketsSize; 193 i <<= 1; 194 t->bucketsSize = i; 195 t->bucketsMask = mask = i - 1; 196 t->buckets = hp = bmake_malloc(sizeof(*hp) * i); 197 while (--i >= 0) 198 *hp++ = NULL; 199 for (hp = oldhp, i = oldsize; --i >= 0;) { 200 for (e = *hp++; e != NULL; e = next) { 201 next = e->next; 202 xp = &t->buckets[e->namehash & mask]; 203 e->next = *xp; 204 *xp = e; 205 } 206 } 207 free(oldhp); 208 DEBUG5(HASH, "%s: %p size=%d entries=%d maxchain=%d\n", 209 __func__, t, t->bucketsSize, t->numEntries, t->maxchain); 210 t->maxchain = 0; 211 } 212 213 /* Searches the hash table for an entry corresponding to the key. 214 * If no entry is found, then one is created. 215 * 216 * Input: 217 * t Hash table to search. 218 * key A hash key. 219 * newPtr Filled with TRUE if new entry created, 220 * FALSE otherwise. 221 */ 222 Hash_Entry * 223 Hash_CreateEntry(Hash_Table *t, const char *key, Boolean *newPtr) 224 { 225 Hash_Entry *e; 226 unsigned h; 227 size_t keylen; 228 int chainlen; 229 struct Hash_Entry **hp; 230 231 /* 232 * Hash the key. As a side effect, save the length (strlen) of the 233 * key in case we need to create the entry. 234 */ 235 h = hash(key, &keylen); 236 chainlen = 0; 237 #ifdef DEBUG_HASH_LOOKUP 238 DEBUG4(HASH, "%s: %p h=%x key=%s\n", __func__, t, h, key); 239 #endif 240 for (e = t->buckets[h & t->bucketsMask]; e != NULL; e = e->next) { 241 chainlen++; 242 if (e->namehash == h && strcmp(e->name, key) == 0) { 243 if (newPtr != NULL) 244 *newPtr = FALSE; 245 break; 246 } 247 } 248 if (chainlen > t->maxchain) 249 t->maxchain = chainlen; 250 if (e) 251 return e; 252 253 /* 254 * The desired entry isn't there. Before allocating a new entry, 255 * expand the table if necessary (and this changes the resulting 256 * bucket chain). 257 */ 258 if (t->numEntries >= rebuildLimit * t->bucketsSize) 259 RebuildTable(t); 260 261 e = bmake_malloc(sizeof(*e) + keylen); 262 hp = &t->buckets[h & t->bucketsMask]; 263 e->next = *hp; 264 *hp = e; 265 Hash_SetValue(e, NULL); 266 e->namehash = h; 267 memcpy(e->name, key, keylen + 1); 268 t->numEntries++; 269 270 if (newPtr != NULL) 271 *newPtr = TRUE; 272 return e; 273 } 274 275 /* Delete the given hash table entry and free memory associated with it. */ 276 void 277 Hash_DeleteEntry(Hash_Table *t, Hash_Entry *e) 278 { 279 Hash_Entry **hp, *p; 280 281 for (hp = &t->buckets[e->namehash & t->bucketsMask]; 282 (p = *hp) != NULL; hp = &p->next) { 283 if (p == e) { 284 *hp = p->next; 285 free(p); 286 t->numEntries--; 287 return; 288 } 289 } 290 abort(); 291 } 292 293 /* Sets things up for enumerating all entries in the hash table. 294 * 295 * Input: 296 * t Table to be searched. 297 * searchPtr Area in which to keep state about search. 298 * 299 * Results: 300 * The return value is the address of the first entry in 301 * the hash table, or NULL if the table is empty. 302 */ 303 Hash_Entry * 304 Hash_EnumFirst(Hash_Table *t, Hash_Search *searchPtr) 305 { 306 searchPtr->table = t; 307 searchPtr->nextBucket = 0; 308 searchPtr->entry = NULL; 309 return Hash_EnumNext(searchPtr); 310 } 311 312 /* Returns the next entry in the hash table, or NULL if the end of the table 313 * is reached. 314 * 315 * Input: 316 * searchPtr Area used to keep state about search. 317 */ 318 Hash_Entry * 319 Hash_EnumNext(Hash_Search *searchPtr) 320 { 321 Hash_Entry *e; 322 Hash_Table *t = searchPtr->table; 323 324 /* 325 * The entry field points to the most recently returned 326 * entry, or is NULL if we are starting up. If not NULL, we have 327 * to start at the next one in the chain. 328 */ 329 e = searchPtr->entry; 330 if (e != NULL) 331 e = e->next; 332 /* 333 * If the chain ran out, or if we are starting up, we need to 334 * find the next nonempty chain. 335 */ 336 while (e == NULL) { 337 if (searchPtr->nextBucket >= t->bucketsSize) 338 return NULL; 339 e = t->buckets[searchPtr->nextBucket++]; 340 } 341 searchPtr->entry = e; 342 return e; 343 } 344 345 void 346 Hash_ForEach(Hash_Table *t, void (*action)(void *, void *), void *data) 347 { 348 Hash_Search search; 349 Hash_Entry *e; 350 351 for (e = Hash_EnumFirst(t, &search); 352 e != NULL; 353 e = Hash_EnumNext(&search)) 354 action(Hash_GetValue(e), data); 355 } 356 357 void 358 Hash_DebugStats(Hash_Table *t, const char *name) 359 { 360 DEBUG4(HASH, "Hash_Table %s: size=%d numEntries=%d maxchain=%d\n", 361 name, t->bucketsSize, t->numEntries, t->maxchain); 362 } 363
Indexes created Fri Sep 26 20:09:58 GMT 2025