hash.c revision 1.41
1 /* $NetBSD: hash.c,v 1.41 2020/10/04 18:16:09 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 1.41 2020/10/04 18:16:09 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 static Hash_Entry * 104 HashTable_Find(Hash_Table *t, unsigned int h, const char *key) 105 { 106 Hash_Entry *e; 107 int chainlen = 0; 108 109 #ifdef DEBUG_HASH_LOOKUP 110 DEBUG4(HASH, "%s: %p h=%x key=%s\n", __func__, t, h, key); 111 #endif 112 113 for (e = t->buckets[h & t->bucketsMask]; e != NULL; e = e->next) { 114 chainlen++; 115 if (e->namehash == h && strcmp(e->name, key) == 0) 116 break; 117 } 118 119 if (chainlen > t->maxchain) 120 t->maxchain = chainlen; 121 122 return e; 123 } 124 125 /* Sets up the hash table. */ 126 void 127 Hash_InitTable(Hash_Table *t) 128 { 129 size_t n = 16, i; 130 struct Hash_Entry **hp; 131 132 t->numEntries = 0; 133 t->maxchain = 0; 134 t->bucketsSize = n; 135 t->bucketsMask = n - 1; 136 t->buckets = hp = bmake_malloc(sizeof(*hp) * n); 137 for (i = 0; i < n; i++) 138 hp[i] = NULL; 139 } 140 141 /* Removes everything from the hash table and frees up the memory space it 142 * occupied (except for the space in the Hash_Table structure). */ 143 void 144 Hash_DeleteTable(Hash_Table *t) 145 { 146 struct Hash_Entry **hp, *h, *nexth = NULL; 147 int i; 148 149 for (hp = t->buckets, i = t->bucketsSize; --i >= 0;) { 150 for (h = *hp++; h != NULL; h = nexth) { 151 nexth = h->next; 152 free(h); 153 } 154 } 155 free(t->buckets); 156 157 /* 158 * Set up the hash table to cause memory faults on any future access 159 * attempts until re-initialization. 160 */ 161 t->buckets = NULL; 162 } 163 164 /* Searches the hash table for an entry corresponding to the key. 165 * 166 * Input: 167 * t Hash table to search. 168 * key A hash key. 169 * 170 * Results: 171 * Returns a pointer to the entry for key, or NULL if the table contains 172 * no entry for the key. 173 */ 174 Hash_Entry * 175 Hash_FindEntry(Hash_Table *t, const char *key) 176 { 177 unsigned int h = hash(key, NULL); 178 return HashTable_Find(t, h, key); 179 } 180 181 void * 182 Hash_FindValue(Hash_Table *t, const char *key) 183 { 184 Hash_Entry *he = Hash_FindEntry(t, key); 185 return he != NULL ? he->value : NULL; 186 } 187 188 /* Makes a new hash table that is larger than the old one. The entire hash 189 * table is moved, so any bucket numbers from the old table become invalid. */ 190 static void 191 RebuildTable(Hash_Table *t) 192 { 193 Hash_Entry *e, *next = NULL, **hp, **xp; 194 int i, mask; 195 Hash_Entry **oldhp; 196 int oldsize; 197 198 oldhp = t->buckets; 199 oldsize = i = t->bucketsSize; 200 i <<= 1; 201 t->bucketsSize = i; 202 t->bucketsMask = mask = i - 1; 203 t->buckets = hp = bmake_malloc(sizeof(*hp) * i); 204 while (--i >= 0) 205 *hp++ = NULL; 206 for (hp = oldhp, i = oldsize; --i >= 0;) { 207 for (e = *hp++; e != NULL; e = next) { 208 next = e->next; 209 xp = &t->buckets[e->namehash & mask]; 210 e->next = *xp; 211 *xp = e; 212 } 213 } 214 free(oldhp); 215 DEBUG5(HASH, "%s: %p size=%d entries=%d maxchain=%d\n", 216 __func__, t, t->bucketsSize, t->numEntries, t->maxchain); 217 t->maxchain = 0; 218 } 219 220 /* Searches the hash table for an entry corresponding to the key. 221 * If no entry is found, then one is created. 222 * 223 * Input: 224 * t Hash table to search. 225 * key A hash key. 226 * newPtr Filled with TRUE if new entry created, 227 * FALSE otherwise. 228 */ 229 Hash_Entry * 230 Hash_CreateEntry(Hash_Table *t, const char *key, Boolean *newPtr) 231 { 232 Hash_Entry *e; 233 unsigned h; 234 size_t keylen; 235 struct Hash_Entry **hp; 236 237 h = hash(key, &keylen); 238 e = HashTable_Find(t, h, key); 239 if (e) { 240 if (newPtr != NULL) 241 *newPtr = FALSE; 242 return e; 243 } 244 245 /* 246 * The desired entry isn't there. Before allocating a new entry, 247 * expand the table if necessary (and this changes the resulting 248 * bucket chain). 249 */ 250 if (t->numEntries >= rebuildLimit * t->bucketsSize) 251 RebuildTable(t); 252 253 e = bmake_malloc(sizeof(*e) + keylen); 254 hp = &t->buckets[h & t->bucketsMask]; 255 e->next = *hp; 256 *hp = e; 257 Hash_SetValue(e, NULL); 258 e->namehash = h; 259 memcpy(e->name, key, keylen + 1); 260 t->numEntries++; 261 262 if (newPtr != NULL) 263 *newPtr = TRUE; 264 return e; 265 } 266 267 /* Delete the given hash table entry and free memory associated with it. */ 268 void 269 Hash_DeleteEntry(Hash_Table *t, Hash_Entry *e) 270 { 271 Hash_Entry **hp, *p; 272 273 for (hp = &t->buckets[e->namehash & t->bucketsMask]; 274 (p = *hp) != NULL; hp = &p->next) { 275 if (p == e) { 276 *hp = p->next; 277 free(p); 278 t->numEntries--; 279 return; 280 } 281 } 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 void 338 Hash_ForEach(Hash_Table *t, void (*action)(void *, void *), void *data) 339 { 340 Hash_Search search; 341 Hash_Entry *e; 342 343 for (e = Hash_EnumFirst(t, &search); 344 e != NULL; 345 e = Hash_EnumNext(&search)) 346 action(Hash_GetValue(e), data); 347 } 348 349 void 350 Hash_DebugStats(Hash_Table *t, const char *name) 351 { 352 DEBUG4(HASH, "Hash_Table %s: size=%d numEntries=%d maxchain=%d\n", 353 name, t->bucketsSize, t->numEntries, t->maxchain); 354 } 355
Indexes created Mon Sep 22 13:09:51 GMT 2025