2 Copyright (C) 2000, 2001 Free Software Foundation, Inc.
4 This file is part of GNU Wget.
6 GNU Wget is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 2 of the License, or (at
9 your option) any later version.
11 GNU Wget is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
16 You should have received a copy of the GNU General Public License
17 along with Wget; if not, write to the Free Software
18 Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
20 In addition, as a special exception, the Free Software Foundation
21 gives permission to link the code of its release of Wget with the
22 OpenSSL project's "OpenSSL" library (or with modified versions of it
23 that use the same license as the "OpenSSL" library), and distribute
24 the linked executables. You must obey the GNU General Public License
25 in all respects for all of the code used other than "OpenSSL". If you
26 modify this file, you may extend this exception to your version of the
27 file, but you are not obligated to do so. If you do not wish to do
28 so, delete this exception statement from your version. */
30 /* With -DSTANDALONE, this file can be compiled outside Wget source
31 tree. To test, also use -DTEST. */
44 /* Get Wget's utility headers. */
48 /* Make do without them. */
49 # define xnew(x) xmalloc (sizeof (x))
50 # define xnew_array(type, x) xmalloc (sizeof (type) * (x))
51 # define xmalloc malloc /* or something that exits
52 if not enough memory */
54 # define countof(x) (sizeof (x) / sizeof ((x)[0]))
55 # define TOLOWER(x) ('A' <= (x) && (x) <= 'Z' ? (x) - 32 : (x))
62 Hash tables are a technique used to implement mapping between
63 objects with near-constant-time access and storage. The table
64 associates keys to values, and a value can be very quickly
65 retrieved by providing the key. Fast lookup tables are typically
66 implemented as hash tables.
69 hash_table_new -- creates the table.
70 hash_table_destroy -- destroys the table.
71 hash_table_put -- establishes or updates key->value mapping.
72 hash_table_get -- retrieves value of key.
73 hash_table_get_pair -- get key/value pair for key.
74 hash_table_contains -- test whether the table contains key.
75 hash_table_remove -- remove the key->value mapping for key.
76 hash_table_map -- iterate through table mappings.
77 hash_table_clear -- clear hash table contents.
78 hash_table_count -- return the number of entries in the table.
80 The hash table grows internally as new entries are added and is not
81 limited in size, except by available memory. The table doubles
82 with each resize, which ensures that the amortized time per
83 operation remains constant.
85 By default, tables created by hash_table_new consider the keys to
86 be equal if their pointer values are the same. You can use
87 make_string_hash_table to create tables whose keys are considered
88 equal if their string contents are the same. In the general case,
89 the criterion of equality used to compare keys is specified at
90 table creation time with two callback functions, "hash" and "test".
91 The hash function transforms the key into an arbitrary number that
92 must be the same for two equal keys. The test function accepts two
93 keys and returns non-zero if they are to be considered equal.
95 Note that neither keys nor values are copied when inserted into the
96 hash table, so they must exist for the lifetime of the table. This
97 means that e.g. the use of static strings is OK, but objects with a
98 shorter life-time need to be copied (with strdup() or the like in
99 the case of strings) before being inserted. */
103 The hash table is implemented as an open-addressed table with
104 linear probing collision resolution.
106 The above means that all the hash entries (pairs of pointers, key
107 and value) are stored in a contiguous array. The position of each
108 mapping is determined by the hash value of its key and the size of
109 the table: location := hash(key) % size. If two different keys end
110 up on the same position (collide), the one that came second is
111 placed at the next empty position following the occupied place.
112 This collision resolution technique is called "linear probing".
114 There are more advanced collision resolution methods (quadratic
115 probing, double hashing), but we don't use them because they incur
116 more non-sequential access to the array, which results in worse CPU
117 cache behavior. Linear probing works well as long as the
118 count/size ratio (fullness) is kept below 75%. We make sure to
119 grow and rehash the table whenever this threshold is exceeded.
121 Collisions make deletion tricky because clearing a position
122 followed by a colliding entry would make the position seem empty
123 and the colliding entry not found. One solution is to leave a
124 "tombstone" instead of clearing the entry, and another is to
125 carefully rehash the entries immediately following the deleted one.
126 We use the latter method because it results in less bookkeeping and
127 faster retrieval at the (slight) expense of deletion. */
129 /* Maximum allowed fullness: when hash table's fullness exceeds this
130 value, the table is resized. */
131 #define HASH_MAX_FULLNESS 0.75
133 /* The hash table size is multiplied by this factor (and then rounded
134 to the next prime) with each resize. This guarantees infrequent
136 #define HASH_RESIZE_FACTOR 2
143 typedef unsigned long (*hashfun_t) (const void *);
144 typedef int (*testfun_t) (const void *, const void *);
147 hashfun_t hash_function;
148 testfun_t test_function;
150 struct mapping *mappings; /* pointer to the table entries. */
151 int size; /* size of the array. */
153 int count; /* number of non-empty entries. */
154 int resize_threshold; /* after size exceeds this number of
155 entries, resize the table. */
156 int prime_offset; /* the offset of the current prime in
160 /* We use the all-bits-set constant (INVALID_PTR) marker to mean that
161 a mapping is empty. It is unaligned and therefore illegal as a
162 pointer. INVALID_PTR_BYTE (0xff) is the one-byte value used to
163 initialize the mappings array as empty.
165 The all-bits-set value is a better choice than NULL because it
166 allows the use of NULL/0 keys. Since the keys are either integers
167 or pointers, the only key that cannot be used is the integer value
168 -1. This is acceptable because it still allows the use of
169 nonnegative integer keys. */
171 #define INVALID_PTR ((void *) ~0UL)
173 # define UCHAR_MAX 0xff
175 #define INVALID_PTR_BYTE UCHAR_MAX
177 #define NON_EMPTY(mp) ((mp)->key != INVALID_PTR)
178 #define MARK_AS_EMPTY(mp) ((mp)->key = INVALID_PTR)
180 /* "Next" mapping is the mapping after MP, but wrapping back to
181 MAPPINGS when MP would reach MAPPINGS+SIZE. */
182 #define NEXT_MAPPING(mp, mappings, size) (mp != mappings + (size - 1) \
185 /* Loop over non-empty mappings starting at MP. */
186 #define LOOP_NON_EMPTY(mp, mappings, size) \
187 for (; NON_EMPTY (mp); mp = NEXT_MAPPING (mp, mappings, size))
189 /* Return the position of KEY in hash table SIZE large, hash function
191 #define HASH_POSITION(key, hashfun, size) ((hashfun) (key) % size)
193 /* Find a prime near, but greather than or equal to SIZE. The primes
194 are looked up from a table with a selection of primes convenient
197 PRIME_OFFSET is a minor optimization: it specifies start position
198 for the search for the large enough prime. The final offset is
199 stored in the same variable. That way the list of primes does not
200 have to be scanned from the beginning each time around. */
203 prime_size (int size, int *prime_offset)
205 static const int primes[] = {
206 13, 19, 29, 41, 59, 79, 107, 149, 197, 263, 347, 457, 599, 787, 1031,
207 1361, 1777, 2333, 3037, 3967, 5167, 6719, 8737, 11369, 14783,
208 19219, 24989, 32491, 42257, 54941, 71429, 92861, 120721, 156941,
209 204047, 265271, 344857, 448321, 582821, 757693, 985003, 1280519,
210 1664681, 2164111, 2813353, 3657361, 4754591, 6180989, 8035301,
211 10445899, 13579681, 17653589, 22949669, 29834603, 38784989,
212 50420551, 65546729, 85210757, 110774011, 144006217, 187208107,
213 243370577, 316381771, 411296309, 534685237, 695090819, 903618083,
214 1174703521, 1527114613, 1837299131, 2147483647
218 for (i = *prime_offset; i < countof (primes); i++)
219 if (primes[i] >= size)
221 /* Set the offset to the next prime. That is safe because,
222 next time we are called, it will be with a larger SIZE,
223 which means we could never return the same prime anyway.
224 (If that is not the case, the caller can simply reset
226 *prime_offset = i + 1;
233 static int cmp_pointer (const void *, const void *);
235 /* Create a hash table with hash function HASH_FUNCTION and test
236 function TEST_FUNCTION. The table is empty (its count is 0), but
237 pre-allocated to store at least ITEMS items.
239 ITEMS is the number of items that the table can accept without
240 needing to resize. It is useful when creating a table that is to
241 be immediately filled with a known number of items. In that case,
242 the regrows are a waste of time, and specifying ITEMS correctly
243 will avoid them altogether.
245 Note that hash tables grow dynamically regardless of ITEMS. The
246 only use of ITEMS is to preallocate the table and avoid unnecessary
247 dynamic regrows. Don't bother making ITEMS prime because it's not
248 used as size unchanged. To start with a small table that grows as
249 needed, simply specify zero ITEMS.
251 If hash and test callbacks are not specified, identity mapping is
252 assumed, i.e. pointer values are used for key comparison. (Common
253 Lisp calls such tables EQ hash tables, and Java calls them
254 IdentityHashMaps.) If your keys require different comparison,
255 specify hash and test functions. For easy use of C strings as hash
256 keys, you can use the convenience functions make_string_hash_table
257 and make_nocase_string_hash_table. */
260 hash_table_new (int items,
261 unsigned long (*hash_function) (const void *),
262 int (*test_function) (const void *, const void *))
265 struct hash_table *ht = xnew (struct hash_table);
267 ht->hash_function = hash_function ? hash_function : hash_pointer;
268 ht->test_function = test_function ? test_function : cmp_pointer;
270 /* If the size of struct hash_table ever becomes a concern, this
271 field can go. (Wget doesn't create many hashes.) */
272 ht->prime_offset = 0;
274 /* Calculate the size that ensures that the table will store at
275 least ITEMS keys without the need to resize. */
276 size = 1 + items / HASH_MAX_FULLNESS;
277 size = prime_size (size, &ht->prime_offset);
279 ht->resize_threshold = size * HASH_MAX_FULLNESS;
280 /*assert (ht->resize_threshold >= items);*/
282 ht->mappings = xnew_array (struct mapping, ht->size);
284 /* Mark mappings as empty. We use 0xff rather than 0 to mark empty
285 keys because it allows us to use NULL/0 as keys. */
286 memset (ht->mappings, INVALID_PTR_BYTE, size * sizeof (struct mapping));
293 /* Free the data associated with hash table HT. */
296 hash_table_destroy (struct hash_table *ht)
298 xfree (ht->mappings);
302 /* The heart of most functions in this file -- find the mapping whose
303 KEY is equal to key, using linear probing. Returns the mapping
304 that matches KEY, or the first empty mapping if none matches. */
306 static inline struct mapping *
307 find_mapping (const struct hash_table *ht, const void *key)
309 struct mapping *mappings = ht->mappings;
311 struct mapping *mp = mappings + HASH_POSITION (key, ht->hash_function, size);
312 testfun_t equals = ht->test_function;
314 LOOP_NON_EMPTY (mp, mappings, size)
315 if (equals (key, mp->key))
320 /* Get the value that corresponds to the key KEY in the hash table HT.
321 If no value is found, return NULL. Note that NULL is a legal value
322 for value; if you are storing NULLs in your hash table, you can use
323 hash_table_contains to be sure that a (possibly NULL) value exists
324 in the table. Or, you can use hash_table_get_pair instead of this
328 hash_table_get (const struct hash_table *ht, const void *key)
330 struct mapping *mp = find_mapping (ht, key);
337 /* Like hash_table_get, but writes out the pointers to both key and
338 value. Returns non-zero on success. */
341 hash_table_get_pair (const struct hash_table *ht, const void *lookup_key,
342 void *orig_key, void *value)
344 struct mapping *mp = find_mapping (ht, lookup_key);
348 *(void **)orig_key = mp->key;
350 *(void **)value = mp->value;
357 /* Return 1 if HT contains KEY, 0 otherwise. */
360 hash_table_contains (const struct hash_table *ht, const void *key)
362 struct mapping *mp = find_mapping (ht, key);
363 return NON_EMPTY (mp);
366 /* Grow hash table HT as necessary, and rehash all the key-value
370 grow_hash_table (struct hash_table *ht)
372 hashfun_t hasher = ht->hash_function;
373 struct mapping *old_mappings = ht->mappings;
374 struct mapping *old_end = ht->mappings + ht->size;
375 struct mapping *mp, *mappings;
378 newsize = prime_size (ht->size * HASH_RESIZE_FACTOR, &ht->prime_offset);
380 printf ("growing from %d to %d; fullness %.2f%% to %.2f%%\n",
382 100.0 * ht->count / ht->size,
383 100.0 * ht->count / newsize);
387 ht->resize_threshold = newsize * HASH_MAX_FULLNESS;
389 mappings = xnew_array (struct mapping, newsize);
390 memset (mappings, INVALID_PTR_BYTE, newsize * sizeof (struct mapping));
391 ht->mappings = mappings;
393 for (mp = old_mappings; mp < old_end; mp++)
396 struct mapping *new_mp;
397 /* We don't need to test for uniqueness of keys because they
398 come from the hash table and are therefore known to be
400 new_mp = mappings + HASH_POSITION (mp->key, hasher, newsize);
401 LOOP_NON_EMPTY (new_mp, mappings, newsize)
406 xfree (old_mappings);
409 /* Put VALUE in the hash table HT under the key KEY. This regrows the
410 table if necessary. */
413 hash_table_put (struct hash_table *ht, const void *key, void *value)
415 struct mapping *mp = find_mapping (ht, key);
418 /* update existing item */
419 mp->key = (void *)key; /* const? */
424 /* If adding the item would make the table exceed max. fullness,
425 grow the table first. */
426 if (ht->count >= ht->resize_threshold)
428 grow_hash_table (ht);
429 mp = find_mapping (ht, key);
434 mp->key = (void *)key; /* const? */
438 /* Remove a mapping that matches KEY from HT. Return 0 if there was
439 no such entry; return 1 if an entry was removed. */
442 hash_table_remove (struct hash_table *ht, const void *key)
444 struct mapping *mp = find_mapping (ht, key);
450 struct mapping *mappings = ht->mappings;
451 hashfun_t hasher = ht->hash_function;
456 /* Rehash all the entries following MP. The alternative
457 approach is to mark the entry as deleted, i.e. create a
458 "tombstone". That speeds up removal, but leaves a lot of
459 garbage and slows down hash_table_get and hash_table_put. */
461 mp = NEXT_MAPPING (mp, mappings, size);
462 LOOP_NON_EMPTY (mp, mappings, size)
464 const void *key2 = mp->key;
465 struct mapping *mp_new;
467 /* Find the new location for the key. */
468 mp_new = mappings + HASH_POSITION (key2, hasher, size);
469 LOOP_NON_EMPTY (mp_new, mappings, size)
470 if (key2 == mp_new->key)
471 /* The mapping MP (key2) is already where we want it (in
472 MP_NEW's "chain" of keys.) */
485 /* Clear HT of all entries. After calling this function, the count
486 and the fullness of the hash table will be zero. The size will
490 hash_table_clear (struct hash_table *ht)
492 memset (ht->mappings, INVALID_PTR_BYTE, ht->size * sizeof (struct mapping));
496 /* Map MAPFUN over all the mappings in hash table HT. MAPFUN is
497 called with three arguments: the key, the value, and MAPARG.
499 It is undefined what happens if you add or remove entries in the
500 hash table while hash_table_map is running. The exception is the
501 entry you're currently mapping over; you may remove or change that
505 hash_table_map (struct hash_table *ht,
506 int (*mapfun) (void *, void *, void *),
509 struct mapping *mp = ht->mappings;
510 struct mapping *end = ht->mappings + ht->size;
512 for (; mp < end; mp++)
518 if (mapfun (key, mp->value, maparg))
520 /* hash_table_remove might have moved the adjacent
522 if (mp->key != key && NON_EMPTY (mp))
527 /* Return the number of elements in the hash table. This is not the
528 same as the physical size of the hash table, which is always
529 greater than the number of elements. */
532 hash_table_count (const struct hash_table *ht)
537 /* Functions from this point onward are meant for convenience and
538 don't strictly belong to this file. However, this is as good a
539 place for them as any. */
541 /* Guidelines for creating custom hash and test functions:
543 - The test function returns non-zero for keys that are considered
544 "equal", zero otherwise.
546 - The hash function returns a number that represents the
547 "distinctness" of the object. In more precise terms, it means
548 that for any two objects that test "equal" under the test
549 function, the hash function MUST produce the same result.
551 This does not mean that all different objects must produce
552 different values (that would be "perfect" hashing), only that
553 non-distinct objects must produce the same values! For instance,
554 a hash function that returns 0 for any given object is a
555 perfectly valid (albeit extremely bad) hash function. A hash
556 function that hashes a string by adding up all its characters is
557 another example of a valid (but quite bad) hash function.
559 It is not hard to make hash and test functions agree about
560 equality. For example, if the test function compares strings
561 case-insensitively, the hash function can lower-case the
562 characters when calculating the hash value. That ensures that
563 two strings differing only in case will hash the same.
565 - If you care about performance, choose a hash function with as
566 good "spreading" as possible. A good hash function will use all
567 the bits of the input when calculating the hash, and will react
568 to even small changes in input with a completely different
569 output. Finally, don't make the hash function itself overly
570 slow, because you'll be incurring a non-negligible overhead to
571 all hash table operations. */
574 * Support for hash tables whose keys are strings.
578 /* 31 bit hash function. Taken from Gnome's glib, modified to use
581 We used to use the popular hash function from the Dragon Book, but
582 this one seems to perform much better. */
585 hash_string (const void *key)
591 for (p += 1; *p != '\0'; p++)
592 h = (h << 5) - h + *p;
597 /* Frontend for strcmp usable for hash tables. */
600 cmp_string (const void *s1, const void *s2)
602 return !strcmp ((const char *)s1, (const char *)s2);
605 /* Return a hash table of preallocated to store at least ITEMS items
606 suitable to use strings as keys. */
609 make_string_hash_table (int items)
611 return hash_table_new (items, hash_string, cmp_string);
615 * Support for hash tables whose keys are strings, but which are
616 * compared case-insensitively.
620 /* Like hash_string, but produce the same hash regardless of the case. */
623 hash_string_nocase (const void *key)
626 unsigned int h = TOLOWER (*p);
629 for (p += 1; *p != '\0'; p++)
630 h = (h << 5) - h + TOLOWER (*p);
635 /* Like string_cmp, but doing case-insensitive compareison. */
638 string_cmp_nocase (const void *s1, const void *s2)
640 return !strcasecmp ((const char *)s1, (const char *)s2);
643 /* Like make_string_hash_table, but uses string_hash_nocase and
644 string_cmp_nocase. */
647 make_nocase_string_hash_table (int items)
649 return hash_table_new (items, hash_string_nocase, string_cmp_nocase);
652 /* Hashing of numeric values, such as pointers and integers.
654 This implementation is the Robert Jenkins' 32 bit Mix Function,
655 with a simple adaptation for 64-bit values. It offers excellent
656 spreading of values and doesn't need to know the hash table size to
657 work (unlike the very popular Knuth's multiplication hash). */
660 hash_pointer (const void *ptr)
662 unsigned long key = (unsigned long)ptr;
685 cmp_pointer (const void *ptr1, const void *ptr2)
696 print_hash_table_mapper (void *key, void *value, void *count)
699 printf ("%s: %s\n", (const char *)key, (char *)value);
704 print_hash (struct hash_table *sht)
707 hash_table_map (sht, print_hash_table_mapper, &debug_count);
708 assert (debug_count == sht->count);
714 struct hash_table *ht = make_string_hash_table (0);
716 while ((fgets (line, sizeof (line), stdin)))
718 int len = strlen (line);
722 if (!hash_table_contains (ht, line))
723 hash_table_put (ht, strdup (line), "here I am!");
728 if (hash_table_get_pair (ht, line, &line_copy, NULL))
730 hash_table_remove (ht, line);
740 printf ("%d %d\n", ht->count, ht->size);