hash.c revision 1.32
1 /* $NetBSD: hash.c,v 1.32 2020/09/13 15:15:51 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 /* hash.c -- 73 * 74 * This module contains routines to manipulate a hash table. 75 * See hash.h for a definition of the structure of the hash 76 * table. Hash tables grow automatically as the amount of 77 * information increases. 78 */ 79 #include "make.h" 80 81 /* "@(#)hash.c 8.1 (Berkeley) 6/6/93" */ 82 MAKE_RCSID("$NetBSD: hash.c,v 1.32 2020/09/13 15:15:51 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 /* Searches the hash table for an entry corresponding to the key. 192 * If no entry is found, then one is created. 193 * 194 * Input: 195 * t Hash table to search. 196 * key A hash key. 197 * newPtr Filled with TRUE if new entry created, 198 * FALSE otherwise. 199 */ 200 Hash_Entry * 201 Hash_CreateEntry(Hash_Table *t, const char *key, Boolean *newPtr) 202 { 203 Hash_Entry *e; 204 unsigned h; 205 const char *p; 206 int keylen; 207 int chainlen; 208 struct Hash_Entry **hp; 209 210 /* 211 * Hash the key. As a side effect, save the length (strlen) of the 212 * key in case we need to create the entry. 213 */ 214 HASH(h, key, p); 215 keylen = p - key; 216 p = key; 217 chainlen = 0; 218 #ifdef DEBUG_HASH_LOOKUP 219 if (DEBUG(HASH)) 220 fprintf(debug_file, "%s: %p h=%x key=%s\n", __func__, 221 t, h, key); 222 #endif 223 for (e = t->buckets[h & t->bucketsMask]; e != NULL; e = e->next) { 224 chainlen++; 225 if (e->namehash == h && strcmp(e->name, p) == 0) { 226 if (newPtr != NULL) 227 *newPtr = FALSE; 228 break; 229 } 230 } 231 if (chainlen > t->maxchain) 232 t->maxchain = chainlen; 233 if (e) 234 return e; 235 236 /* 237 * The desired entry isn't there. Before allocating a new entry, 238 * expand the table if necessary (and this changes the resulting 239 * bucket chain). 240 */ 241 if (t->numEntries >= rebuildLimit * t->bucketsSize) 242 RebuildTable(t); 243 e = bmake_malloc(sizeof(*e) + keylen); 244 hp = &t->buckets[h & t->bucketsMask]; 245 e->next = *hp; 246 *hp = e; 247 Hash_SetValue(e, NULL); 248 e->namehash = h; 249 (void)strcpy(e->name, p); 250 t->numEntries++; 251 252 if (newPtr != NULL) 253 *newPtr = TRUE; 254 return e; 255 } 256 257 /* Delete the given hash table entry and free memory associated with it. */ 258 void 259 Hash_DeleteEntry(Hash_Table *t, Hash_Entry *e) 260 { 261 Hash_Entry **hp, *p; 262 263 if (e == NULL) 264 return; 265 for (hp = &t->buckets[e->namehash & t->bucketsMask]; 266 (p = *hp) != NULL; hp = &p->next) { 267 if (p == e) { 268 *hp = p->next; 269 free(p); 270 t->numEntries--; 271 return; 272 } 273 } 274 (void)write(2, "bad call to Hash_DeleteEntry\n", 29); 275 abort(); 276 } 277 278 /* Sets things up for enumerating all entries in the hash table. 279 * 280 * Input: 281 * t Table to be searched. 282 * searchPtr Area in which to keep state about search. 283 * 284 * Results: 285 * The return value is the address of the first entry in 286 * the hash table, or NULL if the table is empty. 287 */ 288 Hash_Entry * 289 Hash_EnumFirst(Hash_Table *t, Hash_Search *searchPtr) 290 { 291 searchPtr->table = t; 292 searchPtr->nextBucket = 0; 293 searchPtr->entry = NULL; 294 return Hash_EnumNext(searchPtr); 295 } 296 297 /* Returns the next entry in the hash table, or NULL if the end of the table 298 * is reached. 299 * 300 * Input: 301 * searchPtr Area used to keep state about search. 302 */ 303 Hash_Entry * 304 Hash_EnumNext(Hash_Search *searchPtr) 305 { 306 Hash_Entry *e; 307 Hash_Table *t = searchPtr->table; 308 309 /* 310 * The entry field points to the most recently returned 311 * entry, or is NULL if we are starting up. If not NULL, we have 312 * to start at the next one in the chain. 313 */ 314 e = searchPtr->entry; 315 if (e != NULL) 316 e = e->next; 317 /* 318 * If the chain ran out, or if we are starting up, we need to 319 * find the next nonempty chain. 320 */ 321 while (e == NULL) { 322 if (searchPtr->nextBucket >= t->bucketsSize) 323 return NULL; 324 e = t->buckets[searchPtr->nextBucket++]; 325 } 326 searchPtr->entry = e; 327 return e; 328 } 329 330 /* Makes a new hash table that is larger than the old one. The entire hash 331 * table is moved, so any bucket numbers from the old table become invalid. */ 332 static void 333 RebuildTable(Hash_Table *t) 334 { 335 Hash_Entry *e, *next = NULL, **hp, **xp; 336 int i, mask; 337 Hash_Entry **oldhp; 338 int oldsize; 339 340 oldhp = t->buckets; 341 oldsize = i = t->bucketsSize; 342 i <<= 1; 343 t->bucketsSize = i; 344 t->bucketsMask = mask = i - 1; 345 t->buckets = hp = bmake_malloc(sizeof(*hp) * i); 346 while (--i >= 0) 347 *hp++ = NULL; 348 for (hp = oldhp, i = oldsize; --i >= 0;) { 349 for (e = *hp++; e != NULL; e = next) { 350 next = e->next; 351 xp = &t->buckets[e->namehash & mask]; 352 e->next = *xp; 353 *xp = e; 354 } 355 } 356 free(oldhp); 357 if (DEBUG(HASH)) 358 fprintf(debug_file, "%s: %p size=%d entries=%d maxchain=%d\n", 359 __func__, t, t->bucketsSize, t->numEntries, t->maxchain); 360 t->maxchain = 0; 361 } 362 363 void 364 Hash_ForEach(Hash_Table *t, void (*action)(void *, void *), void *data) 365 { 366 Hash_Search search; 367 Hash_Entry *e; 368 369 for (e = Hash_EnumFirst(t, &search); 370 e != NULL; 371 e = Hash_EnumNext(&search)) 372 action(Hash_GetValue(e), data); 373 } 374 375 void 376 Hash_DebugStats(Hash_Table *t, const char *name) 377 { 378 if (DEBUG(HASH)) 379 fprintf(debug_file, "Hash_Table %s: size=%d numEntries=%d maxchain=%d\n", 380 name, t->bucketsSize, t->numEntries, t->maxchain); 381 } 382
Indexes created Thu Sep 25 16:09:42 GMT 2025