Line | Count | Source |
1 | | /* GLIB - Library of useful routines for C programming |
2 | | * Copyright (C) 1995-1997 Peter Mattis, Spencer Kimball and Josh MacDonald |
3 | | * |
4 | | * SPDX-License-Identifier: LGPL-2.1-or-later |
5 | | * |
6 | | * This library is free software; you can redistribute it and/or |
7 | | * modify it under the terms of the GNU Lesser General Public |
8 | | * License as published by the Free Software Foundation; either |
9 | | * version 2.1 of the License, or (at your option) any later version. |
10 | | * |
11 | | * This library 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 GNU |
14 | | * Lesser General Public License for more details. |
15 | | * |
16 | | * You should have received a copy of the GNU Lesser General Public |
17 | | * License along with this library; if not, see <http://www.gnu.org/licenses/>. |
18 | | */ |
19 | | |
20 | | /* |
21 | | * Modified by the GLib Team and others 1997-2000. See the AUTHORS |
22 | | * file for a list of people on the GLib Team. See the ChangeLog |
23 | | * files for a list of changes. These files are distributed with |
24 | | * GLib at ftp://ftp.gtk.org/pub/gtk/. |
25 | | */ |
26 | | |
27 | | /* |
28 | | * MT safe |
29 | | */ |
30 | | |
31 | | #include "config.h" |
32 | | |
33 | | #include <string.h> /* memset */ |
34 | | |
35 | | #include "ghash.h" |
36 | | #include "gmacros.h" |
37 | | #include "glib-private.h" |
38 | | #include "gstrfuncs.h" |
39 | | #include "gatomic.h" |
40 | | #include "gtestutils.h" |
41 | | #include "gslice.h" |
42 | | #include "grefcount.h" |
43 | | #include "gvalgrind.h" |
44 | | |
45 | | /* The following #pragma is here so we can do this... |
46 | | * |
47 | | * #ifndef USE_SMALL_ARRAYS |
48 | | * is_big = TRUE; |
49 | | * #endif |
50 | | * return is_big ? *(((gpointer *) a) + index) : GUINT_TO_POINTER (*(((guint *) a) + index)); |
51 | | * |
52 | | * ...instead of this... |
53 | | * |
54 | | * #ifndef USE_SMALL_ARRAYS |
55 | | * return *(((gpointer *) a) + index); |
56 | | * #else |
57 | | * return is_big ? *(((gpointer *) a) + index) : GUINT_TO_POINTER (*(((guint *) a) + index)); |
58 | | * #endif |
59 | | * |
60 | | * ...and still compile successfully when -Werror=duplicated-branches is passed. */ |
61 | | |
62 | | #if defined(__GNUC__) && __GNUC__ > 6 |
63 | | #pragma GCC diagnostic ignored "-Wduplicated-branches" |
64 | | #endif |
65 | | |
66 | | /** |
67 | | * SECTION:hash_tables |
68 | | * @title: Hash Tables |
69 | | * @short_description: associations between keys and values so that |
70 | | * given a key the value can be found quickly |
71 | | * |
72 | | * A #GHashTable provides associations between keys and values which is |
73 | | * optimized so that given a key, the associated value can be found, |
74 | | * inserted or removed in amortized O(1). All operations going through |
75 | | * each element take O(n) time (list all keys/values, table resize, etc.). |
76 | | * |
77 | | * Note that neither keys nor values are copied when inserted into the |
78 | | * #GHashTable, so they must exist for the lifetime of the #GHashTable. |
79 | | * This means that the use of static strings is OK, but temporary |
80 | | * strings (i.e. those created in buffers and those returned by GTK |
81 | | * widgets) should be copied with g_strdup() before being inserted. |
82 | | * |
83 | | * If keys or values are dynamically allocated, you must be careful to |
84 | | * ensure that they are freed when they are removed from the |
85 | | * #GHashTable, and also when they are overwritten by new insertions |
86 | | * into the #GHashTable. It is also not advisable to mix static strings |
87 | | * and dynamically-allocated strings in a #GHashTable, because it then |
88 | | * becomes difficult to determine whether the string should be freed. |
89 | | * |
90 | | * To create a #GHashTable, use g_hash_table_new(). |
91 | | * |
92 | | * To insert a key and value into a #GHashTable, use |
93 | | * g_hash_table_insert(). |
94 | | * |
95 | | * To look up a value corresponding to a given key, use |
96 | | * g_hash_table_lookup() and g_hash_table_lookup_extended(). |
97 | | * |
98 | | * g_hash_table_lookup_extended() can also be used to simply |
99 | | * check if a key is present in the hash table. |
100 | | * |
101 | | * To remove a key and value, use g_hash_table_remove(). |
102 | | * |
103 | | * To call a function for each key and value pair use |
104 | | * g_hash_table_foreach() or use an iterator to iterate over the |
105 | | * key/value pairs in the hash table, see #GHashTableIter. The iteration order |
106 | | * of a hash table is not defined, and you must not rely on iterating over |
107 | | * keys/values in the same order as they were inserted. |
108 | | * |
109 | | * To destroy a #GHashTable use g_hash_table_destroy(). |
110 | | * |
111 | | * A common use-case for hash tables is to store information about a |
112 | | * set of keys, without associating any particular value with each |
113 | | * key. GHashTable optimizes one way of doing so: If you store only |
114 | | * key-value pairs where key == value, then GHashTable does not |
115 | | * allocate memory to store the values, which can be a considerable |
116 | | * space saving, if your set is large. The functions |
117 | | * g_hash_table_add() and g_hash_table_contains() are designed to be |
118 | | * used when using #GHashTable this way. |
119 | | * |
120 | | * #GHashTable is not designed to be statically initialised with keys and |
121 | | * values known at compile time. To build a static hash table, use a tool such |
122 | | * as [gperf](https://www.gnu.org/software/gperf/). |
123 | | */ |
124 | | |
125 | | /** |
126 | | * GHashTable: |
127 | | * |
128 | | * The #GHashTable struct is an opaque data structure to represent a |
129 | | * [Hash Table][glib-Hash-Tables]. It should only be accessed via the |
130 | | * following functions. |
131 | | */ |
132 | | |
133 | | /** |
134 | | * GHashFunc: |
135 | | * @key: a key |
136 | | * |
137 | | * Specifies the type of the hash function which is passed to |
138 | | * g_hash_table_new() when a #GHashTable is created. |
139 | | * |
140 | | * The function is passed a key and should return a #guint hash value. |
141 | | * The functions g_direct_hash(), g_int_hash() and g_str_hash() provide |
142 | | * hash functions which can be used when the key is a #gpointer, #gint*, |
143 | | * and #gchar* respectively. |
144 | | * |
145 | | * g_direct_hash() is also the appropriate hash function for keys |
146 | | * of the form `GINT_TO_POINTER (n)` (or similar macros). |
147 | | * |
148 | | * A good hash functions should produce |
149 | | * hash values that are evenly distributed over a fairly large range. |
150 | | * The modulus is taken with the hash table size (a prime number) to |
151 | | * find the 'bucket' to place each key into. The function should also |
152 | | * be very fast, since it is called for each key lookup. |
153 | | * |
154 | | * Note that the hash functions provided by GLib have these qualities, |
155 | | * but are not particularly robust against manufactured keys that |
156 | | * cause hash collisions. Therefore, you should consider choosing |
157 | | * a more secure hash function when using a GHashTable with keys |
158 | | * that originate in untrusted data (such as HTTP requests). |
159 | | * Using g_str_hash() in that situation might make your application |
160 | | * vulnerable to |
161 | | * [Algorithmic Complexity Attacks](https://lwn.net/Articles/474912/). |
162 | | * |
163 | | * The key to choosing a good hash is unpredictability. Even |
164 | | * cryptographic hashes are very easy to find collisions for when the |
165 | | * remainder is taken modulo a somewhat predictable prime number. There |
166 | | * must be an element of randomness that an attacker is unable to guess. |
167 | | * |
168 | | * Returns: the hash value corresponding to the key |
169 | | */ |
170 | | |
171 | | /** |
172 | | * GHFunc: |
173 | | * @key: a key |
174 | | * @value: the value corresponding to the key |
175 | | * @user_data: user data passed to g_hash_table_foreach() |
176 | | * |
177 | | * Specifies the type of the function passed to g_hash_table_foreach(). |
178 | | * It is called with each key/value pair, together with the @user_data |
179 | | * parameter which is passed to g_hash_table_foreach(). |
180 | | */ |
181 | | |
182 | | /** |
183 | | * GHRFunc: |
184 | | * @key: a key |
185 | | * @value: the value associated with the key |
186 | | * @user_data: user data passed to g_hash_table_remove() |
187 | | * |
188 | | * Specifies the type of the function passed to |
189 | | * g_hash_table_foreach_remove(). It is called with each key/value |
190 | | * pair, together with the @user_data parameter passed to |
191 | | * g_hash_table_foreach_remove(). It should return %TRUE if the |
192 | | * key/value pair should be removed from the #GHashTable. |
193 | | * |
194 | | * Returns: %TRUE if the key/value pair should be removed from the |
195 | | * #GHashTable |
196 | | */ |
197 | | |
198 | | /** |
199 | | * GEqualFunc: |
200 | | * @a: a value |
201 | | * @b: a value to compare with |
202 | | * |
203 | | * Specifies the type of a function used to test two values for |
204 | | * equality. The function should return %TRUE if both values are equal |
205 | | * and %FALSE otherwise. |
206 | | * |
207 | | * Returns: %TRUE if @a = @b; %FALSE otherwise |
208 | | */ |
209 | | |
210 | | /** |
211 | | * GHashTableIter: |
212 | | * |
213 | | * A GHashTableIter structure represents an iterator that can be used |
214 | | * to iterate over the elements of a #GHashTable. GHashTableIter |
215 | | * structures are typically allocated on the stack and then initialized |
216 | | * with g_hash_table_iter_init(). |
217 | | * |
218 | | * The iteration order of a #GHashTableIter over the keys/values in a hash |
219 | | * table is not defined. |
220 | | */ |
221 | | |
222 | | /** |
223 | | * g_hash_table_freeze: |
224 | | * @hash_table: a #GHashTable |
225 | | * |
226 | | * This function is deprecated and will be removed in the next major |
227 | | * release of GLib. It does nothing. |
228 | | */ |
229 | | |
230 | | /** |
231 | | * g_hash_table_thaw: |
232 | | * @hash_table: a #GHashTable |
233 | | * |
234 | | * This function is deprecated and will be removed in the next major |
235 | | * release of GLib. It does nothing. |
236 | | */ |
237 | | |
238 | 36.1k | #define HASH_TABLE_MIN_SHIFT 3 /* 1 << 3 == 8 buckets */ |
239 | | |
240 | 194k | #define UNUSED_HASH_VALUE 0 |
241 | 22.8k | #define TOMBSTONE_HASH_VALUE 1 |
242 | 194k | #define HASH_IS_UNUSED(h_) ((h_) == UNUSED_HASH_VALUE) |
243 | 35.1k | #define HASH_IS_TOMBSTONE(h_) ((h_) == TOMBSTONE_HASH_VALUE) |
244 | 164k | #define HASH_IS_REAL(h_) ((h_) >= 2) |
245 | | |
246 | | /* If int is smaller than void * on our arch, we start out with |
247 | | * int-sized keys and values and resize to pointer-sized entries as |
248 | | * needed. This saves a good amount of memory when the HT is being |
249 | | * used with e.g. GUINT_TO_POINTER(). */ |
250 | | |
251 | 27.9k | #define BIG_ENTRY_SIZE (SIZEOF_VOID_P) |
252 | 27.7k | #define SMALL_ENTRY_SIZE (SIZEOF_INT) |
253 | | |
254 | | /* NB: The USE_SMALL_ARRAYS code assumes pointers are at most 8 bytes. */ |
255 | | #if SMALL_ENTRY_SIZE < BIG_ENTRY_SIZE && BIG_ENTRY_SIZE <= 8 |
256 | | # define USE_SMALL_ARRAYS |
257 | | #endif |
258 | | |
259 | | struct _GHashTable |
260 | | { |
261 | | gsize size; |
262 | | gint mod; |
263 | | guint mask; |
264 | | guint nnodes; |
265 | | guint noccupied; /* nnodes + tombstones */ |
266 | | |
267 | | guint have_big_keys : 1; |
268 | | guint have_big_values : 1; |
269 | | |
270 | | gpointer keys; |
271 | | guint *hashes; |
272 | | gpointer values; |
273 | | |
274 | | GHashFunc hash_func; |
275 | | GEqualFunc key_equal_func; |
276 | | gatomicrefcount ref_count; |
277 | | #ifndef G_DISABLE_ASSERT |
278 | | /* |
279 | | * Tracks the structure of the hash table, not its contents: is only |
280 | | * incremented when a node is added or removed (is not incremented |
281 | | * when the key or data of a node is modified). |
282 | | */ |
283 | | int version; |
284 | | #endif |
285 | | GDestroyNotify key_destroy_func; |
286 | | GDestroyNotify value_destroy_func; |
287 | | }; |
288 | | |
289 | | typedef struct |
290 | | { |
291 | | GHashTable *hash_table; |
292 | | gpointer dummy1; |
293 | | gpointer dummy2; |
294 | | gint position; |
295 | | gboolean dummy3; |
296 | | gintptr version; |
297 | | } RealIter; |
298 | | |
299 | | G_STATIC_ASSERT (sizeof (GHashTableIter) == sizeof (RealIter)); |
300 | | G_STATIC_ASSERT (G_ALIGNOF (GHashTableIter) >= G_ALIGNOF (RealIter)); |
301 | | |
302 | | /* Each table size has an associated prime modulo (the first prime |
303 | | * lower than the table size) used to find the initial bucket. Probing |
304 | | * then works modulo 2^n. The prime modulo is necessary to get a |
305 | | * good distribution with poor hash functions. |
306 | | */ |
307 | | static const gint prime_mod [] = |
308 | | { |
309 | | 1, /* For 1 << 0 */ |
310 | | 2, |
311 | | 3, |
312 | | 7, |
313 | | 13, |
314 | | 31, |
315 | | 61, |
316 | | 127, |
317 | | 251, |
318 | | 509, |
319 | | 1021, |
320 | | 2039, |
321 | | 4093, |
322 | | 8191, |
323 | | 16381, |
324 | | 32749, |
325 | | 65521, /* For 1 << 16 */ |
326 | | 131071, |
327 | | 262139, |
328 | | 524287, |
329 | | 1048573, |
330 | | 2097143, |
331 | | 4194301, |
332 | | 8388593, |
333 | | 16777213, |
334 | | 33554393, |
335 | | 67108859, |
336 | | 134217689, |
337 | | 268435399, |
338 | | 536870909, |
339 | | 1073741789, |
340 | | 2147483647 /* For 1 << 31 */ |
341 | | }; |
342 | | |
343 | | static void |
344 | | g_hash_table_set_shift (GHashTable *hash_table, gint shift) |
345 | 10.3k | { |
346 | 10.3k | hash_table->size = 1 << shift; |
347 | 10.3k | hash_table->mod = prime_mod [shift]; |
348 | | |
349 | | /* hash_table->size is always a power of two, so we can calculate the mask |
350 | | * by simply subtracting 1 from it. The leading assertion ensures that |
351 | | * we're really dealing with a power of two. */ |
352 | | |
353 | 10.3k | g_assert ((hash_table->size & (hash_table->size - 1)) == 0); |
354 | 10.3k | hash_table->mask = hash_table->size - 1; |
355 | 10.3k | } |
356 | | |
357 | | static gint |
358 | | g_hash_table_find_closest_shift (gint n) |
359 | 26 | { |
360 | 26 | gint i; |
361 | | |
362 | 160 | for (i = 0; n; i++) |
363 | 134 | n >>= 1; |
364 | | |
365 | 26 | return i; |
366 | 26 | } |
367 | | |
368 | | static void |
369 | | g_hash_table_set_shift_from_size (GHashTable *hash_table, gint size) |
370 | 26 | { |
371 | 26 | gint shift; |
372 | | |
373 | 26 | shift = g_hash_table_find_closest_shift (size); |
374 | 26 | shift = MAX (shift, HASH_TABLE_MIN_SHIFT); |
375 | | |
376 | 26 | g_hash_table_set_shift (hash_table, shift); |
377 | 26 | } |
378 | | |
379 | | static inline gpointer |
380 | | g_hash_table_realloc_key_or_value_array (gpointer a, guint size, G_GNUC_UNUSED gboolean is_big) |
381 | 10.3k | { |
382 | 10.3k | #ifdef USE_SMALL_ARRAYS |
383 | 10.3k | return g_realloc (a, size * (is_big ? BIG_ENTRY_SIZE : SMALL_ENTRY_SIZE)); |
384 | | #else |
385 | | return g_renew (gpointer, a, size); |
386 | | #endif |
387 | 10.3k | } |
388 | | |
389 | | static inline gpointer |
390 | | g_hash_table_fetch_key_or_value (gpointer a, guint index, gboolean is_big) |
391 | 89.6k | { |
392 | | #ifndef USE_SMALL_ARRAYS |
393 | | is_big = TRUE; |
394 | | #endif |
395 | 89.6k | return is_big ? *(((gpointer *) a) + index) : GUINT_TO_POINTER (*(((guint *) a) + index)); |
396 | 89.6k | } |
397 | | |
398 | | static inline void |
399 | | g_hash_table_assign_key_or_value (gpointer a, guint index, gboolean is_big, gpointer v) |
400 | 39.6k | { |
401 | | #ifndef USE_SMALL_ARRAYS |
402 | | is_big = TRUE; |
403 | | #endif |
404 | 39.6k | if (is_big) |
405 | 39.1k | *(((gpointer *) a) + index) = v; |
406 | 485 | else |
407 | 485 | *(((guint *) a) + index) = GPOINTER_TO_UINT (v); |
408 | 39.6k | } |
409 | | |
410 | | static inline gpointer |
411 | | g_hash_table_evict_key_or_value (gpointer a, guint index, gboolean is_big, gpointer v) |
412 | 1.12k | { |
413 | | #ifndef USE_SMALL_ARRAYS |
414 | | is_big = TRUE; |
415 | | #endif |
416 | 1.12k | if (is_big) |
417 | 782 | { |
418 | 782 | gpointer r = *(((gpointer *) a) + index); |
419 | 782 | *(((gpointer *) a) + index) = v; |
420 | 782 | return r; |
421 | 782 | } |
422 | 342 | else |
423 | 342 | { |
424 | 342 | gpointer r = GUINT_TO_POINTER (*(((guint *) a) + index)); |
425 | 342 | *(((guint *) a) + index) = GPOINTER_TO_UINT (v); |
426 | 342 | return r; |
427 | 342 | } |
428 | 1.12k | } |
429 | | |
430 | | static inline guint |
431 | | g_hash_table_hash_to_index (GHashTable *hash_table, guint hash) |
432 | 162k | { |
433 | | /* Multiply the hash by a small prime before applying the modulo. This |
434 | | * prevents the table from becoming densely packed, even with a poor hash |
435 | | * function. A densely packed table would have poor performance on |
436 | | * workloads with many failed lookups or a high degree of churn. */ |
437 | 162k | return (hash * 11) % hash_table->mod; |
438 | 162k | } |
439 | | |
440 | | /* |
441 | | * g_hash_table_lookup_node: |
442 | | * @hash_table: our #GHashTable |
443 | | * @key: the key to look up against |
444 | | * @hash_return: key hash return location |
445 | | * |
446 | | * Performs a lookup in the hash table, preserving extra information |
447 | | * usually needed for insertion. |
448 | | * |
449 | | * This function first computes the hash value of the key using the |
450 | | * user's hash function. |
451 | | * |
452 | | * If an entry in the table matching @key is found then this function |
453 | | * returns the index of that entry in the table, and if not, the |
454 | | * index of an unused node (empty or tombstone) where the key can be |
455 | | * inserted. |
456 | | * |
457 | | * The computed hash value is returned in the variable pointed to |
458 | | * by @hash_return. This is to save insertions from having to compute |
459 | | * the hash record again for the new record. |
460 | | * |
461 | | * Returns: index of the described node |
462 | | */ |
463 | | static inline guint |
464 | | g_hash_table_lookup_node (GHashTable *hash_table, |
465 | | gconstpointer key, |
466 | | guint *hash_return) |
467 | 162k | { |
468 | 162k | guint node_index; |
469 | 162k | guint node_hash; |
470 | 162k | guint hash_value; |
471 | 162k | guint first_tombstone = 0; |
472 | 162k | gboolean have_tombstone = FALSE; |
473 | 162k | guint step = 0; |
474 | | |
475 | 162k | hash_value = hash_table->hash_func (key); |
476 | 162k | if (G_UNLIKELY (!HASH_IS_REAL (hash_value))) |
477 | 41 | hash_value = 2; |
478 | | |
479 | 162k | *hash_return = hash_value; |
480 | | |
481 | 162k | node_index = g_hash_table_hash_to_index (hash_table, hash_value); |
482 | 162k | node_hash = hash_table->hashes[node_index]; |
483 | | |
484 | 180k | while (!HASH_IS_UNUSED (node_hash)) |
485 | 59.7k | { |
486 | | /* We first check if our full hash values |
487 | | * are equal so we can avoid calling the full-blown |
488 | | * key equality function in most cases. |
489 | | */ |
490 | 59.7k | if (node_hash == hash_value) |
491 | 42.1k | { |
492 | 42.1k | gpointer node_key = g_hash_table_fetch_key_or_value (hash_table->keys, node_index, hash_table->have_big_keys); |
493 | | |
494 | 42.1k | if (hash_table->key_equal_func) |
495 | 42.1k | { |
496 | 42.1k | if (hash_table->key_equal_func (node_key, key)) |
497 | 42.0k | return node_index; |
498 | 42.1k | } |
499 | 0 | else if (node_key == key) |
500 | 0 | { |
501 | 0 | return node_index; |
502 | 0 | } |
503 | 42.1k | } |
504 | 17.5k | else if (HASH_IS_TOMBSTONE (node_hash) && !have_tombstone) |
505 | 0 | { |
506 | 0 | first_tombstone = node_index; |
507 | 0 | have_tombstone = TRUE; |
508 | 0 | } |
509 | | |
510 | 17.7k | step++; |
511 | 17.7k | node_index += step; |
512 | 17.7k | node_index &= hash_table->mask; |
513 | 17.7k | node_hash = hash_table->hashes[node_index]; |
514 | 17.7k | } |
515 | | |
516 | 120k | if (have_tombstone) |
517 | 0 | return first_tombstone; |
518 | | |
519 | 120k | return node_index; |
520 | 120k | } |
521 | | |
522 | | /* |
523 | | * g_hash_table_remove_node: |
524 | | * @hash_table: our #GHashTable |
525 | | * @node: pointer to node to remove |
526 | | * @notify: %TRUE if the destroy notify handlers are to be called |
527 | | * |
528 | | * Removes a node from the hash table and updates the node count. |
529 | | * The node is replaced by a tombstone. No table resize is performed. |
530 | | * |
531 | | * If @notify is %TRUE then the destroy notify functions are called |
532 | | * for the key and value of the hash node. |
533 | | */ |
534 | | static void |
535 | | g_hash_table_remove_node (GHashTable *hash_table, |
536 | | gint i, |
537 | | gboolean notify) |
538 | 5.25k | { |
539 | 5.25k | gpointer key; |
540 | 5.25k | gpointer value; |
541 | | |
542 | 5.25k | key = g_hash_table_fetch_key_or_value (hash_table->keys, i, hash_table->have_big_keys); |
543 | 5.25k | value = g_hash_table_fetch_key_or_value (hash_table->values, i, hash_table->have_big_values); |
544 | | |
545 | | /* Erect tombstone */ |
546 | 5.25k | hash_table->hashes[i] = TOMBSTONE_HASH_VALUE; |
547 | | |
548 | | /* Be GC friendly */ |
549 | 5.25k | g_hash_table_assign_key_or_value (hash_table->keys, i, hash_table->have_big_keys, NULL); |
550 | 5.25k | g_hash_table_assign_key_or_value (hash_table->values, i, hash_table->have_big_values, NULL); |
551 | | |
552 | 5.25k | g_assert (hash_table->nnodes > 0); |
553 | 5.25k | hash_table->nnodes--; |
554 | | |
555 | 5.25k | if (notify && hash_table->key_destroy_func) |
556 | 0 | hash_table->key_destroy_func (key); |
557 | | |
558 | 5.25k | if (notify && hash_table->value_destroy_func) |
559 | 0 | hash_table->value_destroy_func (value); |
560 | | |
561 | 5.25k | } |
562 | | |
563 | | /* |
564 | | * g_hash_table_setup_storage: |
565 | | * @hash_table: our #GHashTable |
566 | | * |
567 | | * Initialise the hash table size, mask, mod, and arrays. |
568 | | */ |
569 | | static void |
570 | | g_hash_table_setup_storage (GHashTable *hash_table) |
571 | 10.3k | { |
572 | 10.3k | gboolean small = FALSE; |
573 | | |
574 | | /* We want to use small arrays only if: |
575 | | * - we are running on a system where that makes sense (64 bit); and |
576 | | * - we are not running under valgrind. |
577 | | */ |
578 | | |
579 | 10.3k | #ifdef USE_SMALL_ARRAYS |
580 | 10.3k | small = TRUE; |
581 | | |
582 | 10.3k | # ifdef ENABLE_VALGRIND |
583 | 10.3k | if (RUNNING_ON_VALGRIND) |
584 | 0 | small = FALSE; |
585 | 10.3k | # endif |
586 | 10.3k | #endif |
587 | | |
588 | 10.3k | g_hash_table_set_shift (hash_table, HASH_TABLE_MIN_SHIFT); |
589 | | |
590 | 10.3k | hash_table->have_big_keys = !small; |
591 | 10.3k | hash_table->have_big_values = !small; |
592 | | |
593 | 10.3k | hash_table->keys = g_hash_table_realloc_key_or_value_array (NULL, hash_table->size, hash_table->have_big_keys); |
594 | 10.3k | hash_table->values = hash_table->keys; |
595 | 10.3k | hash_table->hashes = g_new0 (guint, hash_table->size); |
596 | 10.3k | } |
597 | | |
598 | | /* |
599 | | * g_hash_table_remove_all_nodes: |
600 | | * @hash_table: our #GHashTable |
601 | | * @notify: %TRUE if the destroy notify handlers are to be called |
602 | | * |
603 | | * Removes all nodes from the table. |
604 | | * |
605 | | * If @notify is %TRUE then the destroy notify functions are called |
606 | | * for the key and value of the hash node. |
607 | | * |
608 | | * Since this may be a precursor to freeing the table entirely, we'd |
609 | | * ideally perform no resize, and we can indeed avoid that in some |
610 | | * cases. However: in the case that we'll be making callbacks to user |
611 | | * code (via destroy notifies) we need to consider that the user code |
612 | | * might call back into the table again. In this case, we setup a new |
613 | | * set of arrays so that any callers will see an empty (but valid) |
614 | | * table. |
615 | | */ |
616 | | static void |
617 | | g_hash_table_remove_all_nodes (GHashTable *hash_table, |
618 | | gboolean notify, |
619 | | gboolean destruction) |
620 | 22.2k | { |
621 | 22.2k | int i; |
622 | 22.2k | gpointer key; |
623 | 22.2k | gpointer value; |
624 | 22.2k | gint old_size; |
625 | 22.2k | gpointer *old_keys; |
626 | 22.2k | gpointer *old_values; |
627 | 22.2k | guint *old_hashes; |
628 | 22.2k | gboolean old_have_big_keys; |
629 | 22.2k | gboolean old_have_big_values; |
630 | | |
631 | | /* If the hash table is already empty, there is nothing to be done. */ |
632 | 22.2k | if (hash_table->nnodes == 0) |
633 | 16.8k | return; |
634 | | |
635 | 5.34k | hash_table->nnodes = 0; |
636 | 5.34k | hash_table->noccupied = 0; |
637 | | |
638 | | /* Easy case: no callbacks, so we just zero out the arrays */ |
639 | 5.34k | if (!notify || |
640 | 5.34k | (hash_table->key_destroy_func == NULL && |
641 | 5.34k | hash_table->value_destroy_func == NULL)) |
642 | 5.23k | { |
643 | 5.23k | if (!destruction) |
644 | 5.23k | { |
645 | 5.23k | memset (hash_table->hashes, 0, hash_table->size * sizeof (guint)); |
646 | | |
647 | 5.23k | #ifdef USE_SMALL_ARRAYS |
648 | 5.23k | memset (hash_table->keys, 0, hash_table->size * (hash_table->have_big_keys ? BIG_ENTRY_SIZE : SMALL_ENTRY_SIZE)); |
649 | 5.23k | memset (hash_table->values, 0, hash_table->size * (hash_table->have_big_values ? BIG_ENTRY_SIZE : SMALL_ENTRY_SIZE)); |
650 | | #else |
651 | | memset (hash_table->keys, 0, hash_table->size * sizeof (gpointer)); |
652 | | memset (hash_table->values, 0, hash_table->size * sizeof (gpointer)); |
653 | | #endif |
654 | 5.23k | } |
655 | | |
656 | 5.23k | return; |
657 | 5.23k | } |
658 | | |
659 | | /* Hard case: we need to do user callbacks. There are two |
660 | | * possibilities here: |
661 | | * |
662 | | * 1) there are no outstanding references on the table and therefore |
663 | | * nobody should be calling into it again (destroying == true) |
664 | | * |
665 | | * 2) there are outstanding references, and there may be future |
666 | | * calls into the table, either after we return, or from the destroy |
667 | | * notifies that we're about to do (destroying == false) |
668 | | * |
669 | | * We handle both cases by taking the current state of the table into |
670 | | * local variables and replacing it with something else: in the "no |
671 | | * outstanding references" cases we replace it with a bunch of |
672 | | * null/zero values so that any access to the table will fail. In the |
673 | | * "may receive future calls" case, we reinitialise the struct to |
674 | | * appear like a newly-created empty table. |
675 | | * |
676 | | * In both cases, we take over the references for the current state, |
677 | | * freeing them below. |
678 | | */ |
679 | 103 | old_size = hash_table->size; |
680 | 103 | old_have_big_keys = hash_table->have_big_keys; |
681 | 103 | old_have_big_values = hash_table->have_big_values; |
682 | 103 | old_keys = g_steal_pointer (&hash_table->keys); |
683 | 103 | old_values = g_steal_pointer (&hash_table->values); |
684 | 103 | old_hashes = g_steal_pointer (&hash_table->hashes); |
685 | | |
686 | 103 | if (!destruction) |
687 | | /* Any accesses will see an empty table */ |
688 | 103 | g_hash_table_setup_storage (hash_table); |
689 | 0 | else |
690 | | /* Will cause a quick crash on any attempted access */ |
691 | 0 | hash_table->size = hash_table->mod = hash_table->mask = 0; |
692 | | |
693 | | /* Now do the actual destroy notifies */ |
694 | 927 | for (i = 0; i < old_size; i++) |
695 | 824 | { |
696 | 824 | if (HASH_IS_REAL (old_hashes[i])) |
697 | 103 | { |
698 | 103 | key = g_hash_table_fetch_key_or_value (old_keys, i, old_have_big_keys); |
699 | 103 | value = g_hash_table_fetch_key_or_value (old_values, i, old_have_big_values); |
700 | | |
701 | 103 | old_hashes[i] = UNUSED_HASH_VALUE; |
702 | | |
703 | 103 | g_hash_table_assign_key_or_value (old_keys, i, old_have_big_keys, NULL); |
704 | 103 | g_hash_table_assign_key_or_value (old_values, i, old_have_big_values, NULL); |
705 | | |
706 | 103 | if (hash_table->key_destroy_func != NULL) |
707 | 0 | hash_table->key_destroy_func (key); |
708 | | |
709 | 103 | if (hash_table->value_destroy_func != NULL) |
710 | 103 | hash_table->value_destroy_func (value); |
711 | 103 | } |
712 | 824 | } |
713 | | |
714 | | /* Destroy old storage space. */ |
715 | 103 | if (old_keys != old_values) |
716 | 103 | g_free (old_values); |
717 | | |
718 | 103 | g_free (old_keys); |
719 | 103 | g_free (old_hashes); |
720 | 103 | } |
721 | | |
722 | | static void |
723 | | realloc_arrays (GHashTable *hash_table, gboolean is_a_set) |
724 | 26 | { |
725 | 26 | hash_table->hashes = g_renew (guint, hash_table->hashes, hash_table->size); |
726 | 26 | hash_table->keys = g_hash_table_realloc_key_or_value_array (hash_table->keys, hash_table->size, hash_table->have_big_keys); |
727 | | |
728 | 26 | if (is_a_set) |
729 | 0 | hash_table->values = hash_table->keys; |
730 | 26 | else |
731 | 26 | hash_table->values = g_hash_table_realloc_key_or_value_array (hash_table->values, hash_table->size, hash_table->have_big_values); |
732 | 26 | } |
733 | | |
734 | | /* When resizing the table in place, we use a temporary bit array to keep |
735 | | * track of which entries have been assigned a proper location in the new |
736 | | * table layout. |
737 | | * |
738 | | * Each bit corresponds to a bucket. A bit is set if an entry was assigned |
739 | | * its corresponding location during the resize and thus should not be |
740 | | * evicted. The array starts out cleared to zero. */ |
741 | | |
742 | | static inline gboolean |
743 | | get_status_bit (const guint32 *bitmap, guint index) |
744 | 1.45k | { |
745 | 1.45k | return (bitmap[index / 32] >> (index % 32)) & 1; |
746 | 1.45k | } |
747 | | |
748 | | static inline void |
749 | | set_status_bit (guint32 *bitmap, guint index) |
750 | 562 | { |
751 | 562 | bitmap[index / 32] |= 1U << (index % 32); |
752 | 562 | } |
753 | | |
754 | | /* By calling dedicated resize functions for sets and maps, we avoid 2x |
755 | | * test-and-branch per key in the inner loop. This yields a small |
756 | | * performance improvement at the cost of a bit of macro gunk. */ |
757 | | |
758 | | #define DEFINE_RESIZE_FUNC(fname) \ |
759 | 26 | static void fname (GHashTable *hash_table, guint old_size, guint32 *reallocated_buckets_bitmap) \ |
760 | 26 | { \ |
761 | 26 | guint i; \ |
762 | 26 | \ |
763 | 602 | for (i = 0; i < old_size; i++) \ |
764 | 576 | { \ |
765 | 576 | guint node_hash = hash_table->hashes[i]; \ |
766 | 576 | gpointer key, value G_GNUC_UNUSED; \ |
767 | 576 | \ |
768 | 576 | if (!HASH_IS_REAL (node_hash)) \ |
769 | 576 | { \ |
770 | 12 | /* Clear tombstones */ \ |
771 | 12 | hash_table->hashes[i] = UNUSED_HASH_VALUE; \ |
772 | 12 | continue; \ |
773 | 12 | } \ |
774 | 576 | \ |
775 | 576 | /* Skip entries relocated through eviction */ \ |
776 | 576 | if (get_status_bit (reallocated_buckets_bitmap, i)) \ |
777 | 564 | continue; \ |
778 | 564 | \ |
779 | 564 | hash_table->hashes[i] = UNUSED_HASH_VALUE; \ |
780 | 405 | EVICT_KEYVAL (hash_table, i, NULL, NULL, key, value); \ |
781 | 405 | \ |
782 | 405 | for (;;) \ |
783 | 562 | { \ |
784 | 562 | guint hash_val; \ |
785 | 562 | guint replaced_hash; \ |
786 | 562 | guint step = 0; \ |
787 | 562 | \ |
788 | 562 | hash_val = g_hash_table_hash_to_index (hash_table, node_hash); \ |
789 | 562 | \ |
790 | 889 | while (get_status_bit (reallocated_buckets_bitmap, hash_val)) \ |
791 | 562 | { \ |
792 | 327 | step++; \ |
793 | 327 | hash_val += step; \ |
794 | 327 | hash_val &= hash_table->mask; \ |
795 | 327 | } \ |
796 | 562 | \ |
797 | 562 | set_status_bit (reallocated_buckets_bitmap, hash_val); \ |
798 | 562 | \ |
799 | 562 | replaced_hash = hash_table->hashes[hash_val]; \ |
800 | 562 | hash_table->hashes[hash_val] = node_hash; \ |
801 | 562 | if (!HASH_IS_REAL (replaced_hash)) \ |
802 | 562 | { \ |
803 | 405 | ASSIGN_KEYVAL (hash_table, hash_val, key, value); \ |
804 | 405 | break; \ |
805 | 405 | } \ |
806 | 562 | \ |
807 | 562 | node_hash = replaced_hash; \ |
808 | 157 | EVICT_KEYVAL (hash_table, hash_val, key, value, key, value); \ |
809 | 157 | } \ |
810 | 405 | } \ |
811 | 26 | } Unexecuted instantiation: ghash.c:resize_set Line | Count | Source | 759 | 26 | static void fname (GHashTable *hash_table, guint old_size, guint32 *reallocated_buckets_bitmap) \ | 760 | 26 | { \ | 761 | 26 | guint i; \ | 762 | 26 | \ | 763 | 602 | for (i = 0; i < old_size; i++) \ | 764 | 576 | { \ | 765 | 576 | guint node_hash = hash_table->hashes[i]; \ | 766 | 576 | gpointer key, value G_GNUC_UNUSED; \ | 767 | 576 | \ | 768 | 576 | if (!HASH_IS_REAL (node_hash)) \ | 769 | 576 | { \ | 770 | 12 | /* Clear tombstones */ \ | 771 | 12 | hash_table->hashes[i] = UNUSED_HASH_VALUE; \ | 772 | 12 | continue; \ | 773 | 12 | } \ | 774 | 576 | \ | 775 | 576 | /* Skip entries relocated through eviction */ \ | 776 | 576 | if (get_status_bit (reallocated_buckets_bitmap, i)) \ | 777 | 564 | continue; \ | 778 | 564 | \ | 779 | 564 | hash_table->hashes[i] = UNUSED_HASH_VALUE; \ | 780 | 405 | EVICT_KEYVAL (hash_table, i, NULL, NULL, key, value); \ | 781 | 405 | \ | 782 | 405 | for (;;) \ | 783 | 562 | { \ | 784 | 562 | guint hash_val; \ | 785 | 562 | guint replaced_hash; \ | 786 | 562 | guint step = 0; \ | 787 | 562 | \ | 788 | 562 | hash_val = g_hash_table_hash_to_index (hash_table, node_hash); \ | 789 | 562 | \ | 790 | 889 | while (get_status_bit (reallocated_buckets_bitmap, hash_val)) \ | 791 | 562 | { \ | 792 | 327 | step++; \ | 793 | 327 | hash_val += step; \ | 794 | 327 | hash_val &= hash_table->mask; \ | 795 | 327 | } \ | 796 | 562 | \ | 797 | 562 | set_status_bit (reallocated_buckets_bitmap, hash_val); \ | 798 | 562 | \ | 799 | 562 | replaced_hash = hash_table->hashes[hash_val]; \ | 800 | 562 | hash_table->hashes[hash_val] = node_hash; \ | 801 | 562 | if (!HASH_IS_REAL (replaced_hash)) \ | 802 | 562 | { \ | 803 | 405 | ASSIGN_KEYVAL (hash_table, hash_val, key, value); \ | 804 | 405 | break; \ | 805 | 405 | } \ | 806 | 562 | \ | 807 | 562 | node_hash = replaced_hash; \ | 808 | 157 | EVICT_KEYVAL (hash_table, hash_val, key, value, key, value); \ | 809 | 157 | } \ | 810 | 405 | } \ | 811 | 26 | } |
|
812 | | |
813 | 405 | #define ASSIGN_KEYVAL(ht, index, key, value) G_STMT_START{ \ |
814 | 405 | g_hash_table_assign_key_or_value ((ht)->keys, (index), (ht)->have_big_keys, (key)); \ |
815 | 405 | g_hash_table_assign_key_or_value ((ht)->values, (index), (ht)->have_big_values, (value)); \ |
816 | 405 | }G_STMT_END |
817 | | |
818 | 562 | #define EVICT_KEYVAL(ht, index, key, value, outkey, outvalue) G_STMT_START{ \ |
819 | 562 | (outkey) = g_hash_table_evict_key_or_value ((ht)->keys, (index), (ht)->have_big_keys, (key)); \ |
820 | 562 | (outvalue) = g_hash_table_evict_key_or_value ((ht)->values, (index), (ht)->have_big_values, (value)); \ |
821 | 562 | }G_STMT_END |
822 | | |
823 | | DEFINE_RESIZE_FUNC (resize_map) |
824 | | |
825 | | #undef ASSIGN_KEYVAL |
826 | | #undef EVICT_KEYVAL |
827 | | |
828 | 0 | #define ASSIGN_KEYVAL(ht, index, key, value) G_STMT_START{ \ |
829 | 0 | g_hash_table_assign_key_or_value ((ht)->keys, (index), (ht)->have_big_keys, (key)); \ |
830 | 0 | }G_STMT_END |
831 | | |
832 | 0 | #define EVICT_KEYVAL(ht, index, key, value, outkey, outvalue) G_STMT_START{ \ |
833 | 0 | (outkey) = g_hash_table_evict_key_or_value ((ht)->keys, (index), (ht)->have_big_keys, (key)); \ |
834 | 0 | }G_STMT_END |
835 | | |
836 | | DEFINE_RESIZE_FUNC (resize_set) |
837 | | |
838 | | #undef ASSIGN_KEYVAL |
839 | | #undef EVICT_KEYVAL |
840 | | |
841 | | /* |
842 | | * g_hash_table_resize: |
843 | | * @hash_table: our #GHashTable |
844 | | * |
845 | | * Resizes the hash table to the optimal size based on the number of |
846 | | * nodes currently held. If you call this function then a resize will |
847 | | * occur, even if one does not need to occur. |
848 | | * Use g_hash_table_maybe_resize() instead. |
849 | | * |
850 | | * This function may "resize" the hash table to its current size, with |
851 | | * the side effect of cleaning up tombstones and otherwise optimizing |
852 | | * the probe sequences. |
853 | | */ |
854 | | static void |
855 | | g_hash_table_resize (GHashTable *hash_table) |
856 | 26 | { |
857 | 26 | guint32 *reallocated_buckets_bitmap; |
858 | 26 | gsize old_size; |
859 | 26 | gboolean is_a_set; |
860 | | |
861 | 26 | old_size = hash_table->size; |
862 | 26 | is_a_set = hash_table->keys == hash_table->values; |
863 | | |
864 | | /* The outer checks in g_hash_table_maybe_resize() will only consider |
865 | | * cleanup/resize when the load factor goes below .25 (1/4, ignoring |
866 | | * tombstones) or above .9375 (15/16, including tombstones). |
867 | | * |
868 | | * Once this happens, tombstones will always be cleaned out. If our |
869 | | * load sans tombstones is greater than .75 (1/1.333, see below), we'll |
870 | | * take this opportunity to grow the table too. |
871 | | * |
872 | | * Immediately after growing, the load factor will be in the range |
873 | | * .375 .. .469. After shrinking, it will be exactly .5. */ |
874 | | |
875 | 26 | g_hash_table_set_shift_from_size (hash_table, hash_table->nnodes * 1.333); |
876 | | |
877 | 26 | if (hash_table->size > old_size) |
878 | 26 | { |
879 | 26 | realloc_arrays (hash_table, is_a_set); |
880 | 26 | memset (&hash_table->hashes[old_size], 0, (hash_table->size - old_size) * sizeof (guint)); |
881 | | |
882 | 26 | reallocated_buckets_bitmap = g_new0 (guint32, (hash_table->size + 31) / 32); |
883 | 26 | } |
884 | 0 | else |
885 | 0 | { |
886 | 0 | reallocated_buckets_bitmap = g_new0 (guint32, (old_size + 31) / 32); |
887 | 0 | } |
888 | | |
889 | 26 | if (is_a_set) |
890 | 0 | resize_set (hash_table, old_size, reallocated_buckets_bitmap); |
891 | 26 | else |
892 | 26 | resize_map (hash_table, old_size, reallocated_buckets_bitmap); |
893 | | |
894 | 26 | g_free (reallocated_buckets_bitmap); |
895 | | |
896 | 26 | if (hash_table->size < old_size) |
897 | 0 | realloc_arrays (hash_table, is_a_set); |
898 | | |
899 | 26 | hash_table->noccupied = hash_table->nnodes; |
900 | 26 | } |
901 | | |
902 | | /* |
903 | | * g_hash_table_maybe_resize: |
904 | | * @hash_table: our #GHashTable |
905 | | * |
906 | | * Resizes the hash table, if needed. |
907 | | * |
908 | | * Essentially, calls g_hash_table_resize() if the table has strayed |
909 | | * too far from its ideal size for its number of nodes. |
910 | | */ |
911 | | static inline void |
912 | | g_hash_table_maybe_resize (GHashTable *hash_table) |
913 | 31.3k | { |
914 | 31.3k | gsize noccupied = hash_table->noccupied; |
915 | 31.3k | gsize size = hash_table->size; |
916 | | |
917 | 31.3k | if ((size > hash_table->nnodes * 4 && size > 1 << HASH_TABLE_MIN_SHIFT) || |
918 | 31.3k | (size <= noccupied + (noccupied / 16))) |
919 | 26 | g_hash_table_resize (hash_table); |
920 | 31.3k | } |
921 | | |
922 | | #ifdef USE_SMALL_ARRAYS |
923 | | |
924 | | static inline gboolean |
925 | | entry_is_big (gpointer v) |
926 | 17.4k | { |
927 | 17.4k | return (((guintptr) v) >> ((BIG_ENTRY_SIZE - SMALL_ENTRY_SIZE) * 8)) != 0; |
928 | 17.4k | } |
929 | | |
930 | | static inline gboolean |
931 | | g_hash_table_maybe_make_big_keys_or_values (gpointer *a_p, gpointer v, gint ht_size) |
932 | 17.4k | { |
933 | 17.4k | if (entry_is_big (v)) |
934 | 17.1k | { |
935 | 17.1k | guint *a = (guint *) *a_p; |
936 | 17.1k | gpointer *a_new; |
937 | 17.1k | gint i; |
938 | | |
939 | 17.1k | a_new = g_new (gpointer, ht_size); |
940 | | |
941 | 154k | for (i = 0; i < ht_size; i++) |
942 | 137k | { |
943 | 137k | a_new[i] = GUINT_TO_POINTER (a[i]); |
944 | 137k | } |
945 | | |
946 | 17.1k | g_free (a); |
947 | 17.1k | *a_p = a_new; |
948 | 17.1k | return TRUE; |
949 | 17.1k | } |
950 | | |
951 | 240 | return FALSE; |
952 | 17.4k | } |
953 | | |
954 | | #endif |
955 | | |
956 | | static inline void |
957 | | g_hash_table_ensure_keyval_fits (GHashTable *hash_table, gpointer key, gpointer value) |
958 | 14.0k | { |
959 | 14.0k | gboolean is_a_set = (hash_table->keys == hash_table->values); |
960 | | |
961 | 14.0k | #ifdef USE_SMALL_ARRAYS |
962 | | |
963 | | /* Convert from set to map? */ |
964 | 14.0k | if (is_a_set) |
965 | 8.60k | { |
966 | 8.60k | if (hash_table->have_big_keys) |
967 | 6 | { |
968 | 6 | if (key != value) |
969 | 0 | hash_table->values = g_memdup2 (hash_table->keys, sizeof (gpointer) * hash_table->size); |
970 | | /* Keys and values are both big now, so no need for further checks */ |
971 | 6 | return; |
972 | 6 | } |
973 | 8.59k | else |
974 | 8.59k | { |
975 | 8.59k | if (key != value) |
976 | 8.59k | { |
977 | 8.59k | hash_table->values = g_memdup2 (hash_table->keys, sizeof (guint) * hash_table->size); |
978 | 8.59k | is_a_set = FALSE; |
979 | 8.59k | } |
980 | 8.59k | } |
981 | 8.60k | } |
982 | | |
983 | | /* Make keys big? */ |
984 | 14.0k | if (!hash_table->have_big_keys) |
985 | 8.59k | { |
986 | 8.59k | hash_table->have_big_keys = g_hash_table_maybe_make_big_keys_or_values (&hash_table->keys, key, hash_table->size); |
987 | | |
988 | 8.59k | if (is_a_set) |
989 | 2 | { |
990 | 2 | hash_table->values = hash_table->keys; |
991 | 2 | hash_table->have_big_values = hash_table->have_big_keys; |
992 | 2 | } |
993 | 8.59k | } |
994 | | |
995 | | /* Make values big? */ |
996 | 14.0k | if (!is_a_set && !hash_table->have_big_values) |
997 | 8.82k | { |
998 | 8.82k | hash_table->have_big_values = g_hash_table_maybe_make_big_keys_or_values (&hash_table->values, value, hash_table->size); |
999 | 8.82k | } |
1000 | | |
1001 | | #else |
1002 | | |
1003 | | /* Just split if necessary */ |
1004 | | if (is_a_set && key != value) |
1005 | | hash_table->values = g_memdup2 (hash_table->keys, sizeof (gpointer) * hash_table->size); |
1006 | | |
1007 | | #endif |
1008 | 14.0k | } |
1009 | | |
1010 | | /** |
1011 | | * g_hash_table_new: |
1012 | | * @hash_func: a function to create a hash value from a key |
1013 | | * @key_equal_func: a function to check two keys for equality |
1014 | | * |
1015 | | * Creates a new #GHashTable with a reference count of 1. |
1016 | | * |
1017 | | * Hash values returned by @hash_func are used to determine where keys |
1018 | | * are stored within the #GHashTable data structure. The g_direct_hash(), |
1019 | | * g_int_hash(), g_int64_hash(), g_double_hash() and g_str_hash() |
1020 | | * functions are provided for some common types of keys. |
1021 | | * If @hash_func is %NULL, g_direct_hash() is used. |
1022 | | * |
1023 | | * @key_equal_func is used when looking up keys in the #GHashTable. |
1024 | | * The g_direct_equal(), g_int_equal(), g_int64_equal(), g_double_equal() |
1025 | | * and g_str_equal() functions are provided for the most common types |
1026 | | * of keys. If @key_equal_func is %NULL, keys are compared directly in |
1027 | | * a similar fashion to g_direct_equal(), but without the overhead of |
1028 | | * a function call. @key_equal_func is called with the key from the hash table |
1029 | | * as its first parameter, and the user-provided key to check against as |
1030 | | * its second. |
1031 | | * |
1032 | | * Returns: a new #GHashTable |
1033 | | */ |
1034 | | GHashTable * |
1035 | | g_hash_table_new (GHashFunc hash_func, |
1036 | | GEqualFunc key_equal_func) |
1037 | 8.51k | { |
1038 | 8.51k | return g_hash_table_new_full (hash_func, key_equal_func, NULL, NULL); |
1039 | 8.51k | } |
1040 | | |
1041 | | |
1042 | | /** |
1043 | | * g_hash_table_new_full: |
1044 | | * @hash_func: a function to create a hash value from a key |
1045 | | * @key_equal_func: a function to check two keys for equality |
1046 | | * @key_destroy_func: (nullable): a function to free the memory allocated for the key |
1047 | | * used when removing the entry from the #GHashTable, or %NULL |
1048 | | * if you don't want to supply such a function. |
1049 | | * @value_destroy_func: (nullable): a function to free the memory allocated for the |
1050 | | * value used when removing the entry from the #GHashTable, or %NULL |
1051 | | * if you don't want to supply such a function. |
1052 | | * |
1053 | | * Creates a new #GHashTable like g_hash_table_new() with a reference |
1054 | | * count of 1 and allows to specify functions to free the memory |
1055 | | * allocated for the key and value that get called when removing the |
1056 | | * entry from the #GHashTable. |
1057 | | * |
1058 | | * Since version 2.42 it is permissible for destroy notify functions to |
1059 | | * recursively remove further items from the hash table. This is only |
1060 | | * permissible if the application still holds a reference to the hash table. |
1061 | | * This means that you may need to ensure that the hash table is empty by |
1062 | | * calling g_hash_table_remove_all() before releasing the last reference using |
1063 | | * g_hash_table_unref(). |
1064 | | * |
1065 | | * Returns: a new #GHashTable |
1066 | | */ |
1067 | | GHashTable * |
1068 | | g_hash_table_new_full (GHashFunc hash_func, |
1069 | | GEqualFunc key_equal_func, |
1070 | | GDestroyNotify key_destroy_func, |
1071 | | GDestroyNotify value_destroy_func) |
1072 | 10.2k | { |
1073 | 10.2k | GHashTable *hash_table; |
1074 | | |
1075 | 10.2k | hash_table = g_slice_new (GHashTable); |
1076 | 10.2k | g_atomic_ref_count_init (&hash_table->ref_count); |
1077 | 10.2k | hash_table->nnodes = 0; |
1078 | 10.2k | hash_table->noccupied = 0; |
1079 | 10.2k | hash_table->hash_func = hash_func ? hash_func : g_direct_hash; |
1080 | 10.2k | hash_table->key_equal_func = key_equal_func; |
1081 | 10.2k | #ifndef G_DISABLE_ASSERT |
1082 | 10.2k | hash_table->version = 0; |
1083 | 10.2k | #endif |
1084 | 10.2k | hash_table->key_destroy_func = key_destroy_func; |
1085 | 10.2k | hash_table->value_destroy_func = value_destroy_func; |
1086 | | |
1087 | 10.2k | g_hash_table_setup_storage (hash_table); |
1088 | | |
1089 | 10.2k | return hash_table; |
1090 | 10.2k | } |
1091 | | |
1092 | | /** |
1093 | | * g_hash_table_new_similar: |
1094 | | * @other_hash_table: (not nullable) (transfer none): Another #GHashTable |
1095 | | * |
1096 | | * Creates a new #GHashTable like g_hash_table_new_full() with a reference |
1097 | | * count of 1. |
1098 | | * |
1099 | | * It inherits the hash function, the key equal function, the key destroy function, |
1100 | | * as well as the value destroy function, from @other_hash_table. |
1101 | | * |
1102 | | * The returned hash table will be empty; it will not contain the keys |
1103 | | * or values from @other_hash_table. |
1104 | | * |
1105 | | * Returns: (transfer full) (not nullable): a new #GHashTable |
1106 | | * Since: 2.72 |
1107 | | */ |
1108 | | GHashTable * |
1109 | | g_hash_table_new_similar (GHashTable *other_hash_table) |
1110 | 0 | { |
1111 | 0 | g_return_val_if_fail (other_hash_table, NULL); |
1112 | | |
1113 | 0 | return g_hash_table_new_full (other_hash_table->hash_func, |
1114 | 0 | other_hash_table->key_equal_func, |
1115 | 0 | other_hash_table->key_destroy_func, |
1116 | 0 | other_hash_table->value_destroy_func); |
1117 | 0 | } |
1118 | | |
1119 | | /** |
1120 | | * g_hash_table_iter_init: |
1121 | | * @iter: an uninitialized #GHashTableIter |
1122 | | * @hash_table: a #GHashTable |
1123 | | * |
1124 | | * Initializes a key/value pair iterator and associates it with |
1125 | | * @hash_table. Modifying the hash table after calling this function |
1126 | | * invalidates the returned iterator. |
1127 | | * |
1128 | | * The iteration order of a #GHashTableIter over the keys/values in a hash |
1129 | | * table is not defined. |
1130 | | * |
1131 | | * |[<!-- language="C" --> |
1132 | | * GHashTableIter iter; |
1133 | | * gpointer key, value; |
1134 | | * |
1135 | | * g_hash_table_iter_init (&iter, hash_table); |
1136 | | * while (g_hash_table_iter_next (&iter, &key, &value)) |
1137 | | * { |
1138 | | * // do something with key and value |
1139 | | * } |
1140 | | * ]| |
1141 | | * |
1142 | | * Since: 2.16 |
1143 | | */ |
1144 | | void |
1145 | | g_hash_table_iter_init (GHashTableIter *iter, |
1146 | | GHashTable *hash_table) |
1147 | 0 | { |
1148 | 0 | RealIter *ri = (RealIter *) iter; |
1149 | |
|
1150 | 0 | g_return_if_fail (iter != NULL); |
1151 | 0 | g_return_if_fail (hash_table != NULL); |
1152 | | |
1153 | 0 | ri->hash_table = hash_table; |
1154 | 0 | ri->position = -1; |
1155 | 0 | #ifndef G_DISABLE_ASSERT |
1156 | 0 | ri->version = hash_table->version; |
1157 | 0 | #endif |
1158 | 0 | } |
1159 | | |
1160 | | /** |
1161 | | * g_hash_table_iter_next: |
1162 | | * @iter: an initialized #GHashTableIter |
1163 | | * @key: (out) (optional): a location to store the key |
1164 | | * @value: (out) (optional) (nullable): a location to store the value |
1165 | | * |
1166 | | * Advances @iter and retrieves the key and/or value that are now |
1167 | | * pointed to as a result of this advancement. If %FALSE is returned, |
1168 | | * @key and @value are not set, and the iterator becomes invalid. |
1169 | | * |
1170 | | * Returns: %FALSE if the end of the #GHashTable has been reached. |
1171 | | * |
1172 | | * Since: 2.16 |
1173 | | */ |
1174 | | gboolean |
1175 | | g_hash_table_iter_next (GHashTableIter *iter, |
1176 | | gpointer *key, |
1177 | | gpointer *value) |
1178 | 0 | { |
1179 | 0 | RealIter *ri = (RealIter *) iter; |
1180 | 0 | gint position; |
1181 | |
|
1182 | 0 | g_return_val_if_fail (iter != NULL, FALSE); |
1183 | 0 | #ifndef G_DISABLE_ASSERT |
1184 | 0 | g_return_val_if_fail (ri->version == ri->hash_table->version, FALSE); |
1185 | 0 | #endif |
1186 | 0 | g_return_val_if_fail (ri->position < (gssize) ri->hash_table->size, FALSE); |
1187 | | |
1188 | 0 | position = ri->position; |
1189 | |
|
1190 | 0 | do |
1191 | 0 | { |
1192 | 0 | position++; |
1193 | 0 | if (position >= (gssize) ri->hash_table->size) |
1194 | 0 | { |
1195 | 0 | ri->position = position; |
1196 | 0 | return FALSE; |
1197 | 0 | } |
1198 | 0 | } |
1199 | 0 | while (!HASH_IS_REAL (ri->hash_table->hashes[position])); |
1200 | | |
1201 | 0 | if (key != NULL) |
1202 | 0 | *key = g_hash_table_fetch_key_or_value (ri->hash_table->keys, position, ri->hash_table->have_big_keys); |
1203 | 0 | if (value != NULL) |
1204 | 0 | *value = g_hash_table_fetch_key_or_value (ri->hash_table->values, position, ri->hash_table->have_big_values); |
1205 | |
|
1206 | 0 | ri->position = position; |
1207 | 0 | return TRUE; |
1208 | 0 | } |
1209 | | |
1210 | | /** |
1211 | | * g_hash_table_iter_get_hash_table: |
1212 | | * @iter: an initialized #GHashTableIter |
1213 | | * |
1214 | | * Returns the #GHashTable associated with @iter. |
1215 | | * |
1216 | | * Returns: the #GHashTable associated with @iter. |
1217 | | * |
1218 | | * Since: 2.16 |
1219 | | */ |
1220 | | GHashTable * |
1221 | | g_hash_table_iter_get_hash_table (GHashTableIter *iter) |
1222 | 0 | { |
1223 | 0 | g_return_val_if_fail (iter != NULL, NULL); |
1224 | | |
1225 | 0 | return ((RealIter *) iter)->hash_table; |
1226 | 0 | } |
1227 | | |
1228 | | static void |
1229 | | iter_remove_or_steal (RealIter *ri, gboolean notify) |
1230 | 0 | { |
1231 | 0 | g_return_if_fail (ri != NULL); |
1232 | 0 | #ifndef G_DISABLE_ASSERT |
1233 | 0 | g_return_if_fail (ri->version == ri->hash_table->version); |
1234 | 0 | #endif |
1235 | 0 | g_return_if_fail (ri->position >= 0); |
1236 | 0 | g_return_if_fail ((gsize) ri->position < ri->hash_table->size); |
1237 | | |
1238 | 0 | g_hash_table_remove_node (ri->hash_table, ri->position, notify); |
1239 | |
|
1240 | 0 | #ifndef G_DISABLE_ASSERT |
1241 | 0 | ri->version++; |
1242 | 0 | ri->hash_table->version++; |
1243 | 0 | #endif |
1244 | 0 | } |
1245 | | |
1246 | | /** |
1247 | | * g_hash_table_iter_remove: |
1248 | | * @iter: an initialized #GHashTableIter |
1249 | | * |
1250 | | * Removes the key/value pair currently pointed to by the iterator |
1251 | | * from its associated #GHashTable. Can only be called after |
1252 | | * g_hash_table_iter_next() returned %TRUE, and cannot be called |
1253 | | * more than once for the same key/value pair. |
1254 | | * |
1255 | | * If the #GHashTable was created using g_hash_table_new_full(), |
1256 | | * the key and value are freed using the supplied destroy functions, |
1257 | | * otherwise you have to make sure that any dynamically allocated |
1258 | | * values are freed yourself. |
1259 | | * |
1260 | | * It is safe to continue iterating the #GHashTable afterward: |
1261 | | * |[<!-- language="C" --> |
1262 | | * while (g_hash_table_iter_next (&iter, &key, &value)) |
1263 | | * { |
1264 | | * if (condition) |
1265 | | * g_hash_table_iter_remove (&iter); |
1266 | | * } |
1267 | | * ]| |
1268 | | * |
1269 | | * Since: 2.16 |
1270 | | */ |
1271 | | void |
1272 | | g_hash_table_iter_remove (GHashTableIter *iter) |
1273 | 0 | { |
1274 | 0 | iter_remove_or_steal ((RealIter *) iter, TRUE); |
1275 | 0 | } |
1276 | | |
1277 | | /* |
1278 | | * g_hash_table_insert_node: |
1279 | | * @hash_table: our #GHashTable |
1280 | | * @node_index: pointer to node to insert/replace |
1281 | | * @key_hash: key hash |
1282 | | * @key: (nullable): key to replace with, or %NULL |
1283 | | * @value: value to replace with |
1284 | | * @keep_new_key: whether to replace the key in the node with @key |
1285 | | * @reusing_key: whether @key was taken out of the existing node |
1286 | | * |
1287 | | * Inserts a value at @node_index in the hash table and updates it. |
1288 | | * |
1289 | | * If @key has been taken out of the existing node (ie it is not |
1290 | | * passed in via a g_hash_table_insert/replace) call, then @reusing_key |
1291 | | * should be %TRUE. |
1292 | | * |
1293 | | * Returns: %TRUE if the key did not exist yet |
1294 | | */ |
1295 | | static gboolean |
1296 | | g_hash_table_insert_node (GHashTable *hash_table, |
1297 | | guint node_index, |
1298 | | guint key_hash, |
1299 | | gpointer new_key, |
1300 | | gpointer new_value, |
1301 | | gboolean keep_new_key, |
1302 | | gboolean reusing_key) |
1303 | 14.0k | { |
1304 | 14.0k | gboolean already_exists; |
1305 | 14.0k | guint old_hash; |
1306 | 14.0k | gpointer key_to_free = NULL; |
1307 | 14.0k | gpointer key_to_keep = NULL; |
1308 | 14.0k | gpointer value_to_free = NULL; |
1309 | | |
1310 | 14.0k | old_hash = hash_table->hashes[node_index]; |
1311 | 14.0k | already_exists = HASH_IS_REAL (old_hash); |
1312 | | |
1313 | | /* Proceed in three steps. First, deal with the key because it is the |
1314 | | * most complicated. Then consider if we need to split the table in |
1315 | | * two (because writing the value will result in the set invariant |
1316 | | * becoming broken). Then deal with the value. |
1317 | | * |
1318 | | * There are three cases for the key: |
1319 | | * |
1320 | | * - entry already exists in table, reusing key: |
1321 | | * free the just-passed-in new_key and use the existing value |
1322 | | * |
1323 | | * - entry already exists in table, not reusing key: |
1324 | | * free the entry in the table, use the new key |
1325 | | * |
1326 | | * - entry not already in table: |
1327 | | * use the new key, free nothing |
1328 | | * |
1329 | | * We update the hash at the same time... |
1330 | | */ |
1331 | 14.0k | if (already_exists) |
1332 | 0 | { |
1333 | | /* Note: we must record the old value before writing the new key |
1334 | | * because we might change the value in the event that the two |
1335 | | * arrays are shared. |
1336 | | */ |
1337 | 0 | value_to_free = g_hash_table_fetch_key_or_value (hash_table->values, node_index, hash_table->have_big_values); |
1338 | |
|
1339 | 0 | if (keep_new_key) |
1340 | 0 | { |
1341 | 0 | key_to_free = g_hash_table_fetch_key_or_value (hash_table->keys, node_index, hash_table->have_big_keys); |
1342 | 0 | key_to_keep = new_key; |
1343 | 0 | } |
1344 | 0 | else |
1345 | 0 | { |
1346 | 0 | key_to_free = new_key; |
1347 | 0 | key_to_keep = g_hash_table_fetch_key_or_value (hash_table->keys, node_index, hash_table->have_big_keys); |
1348 | 0 | } |
1349 | 0 | } |
1350 | 14.0k | else |
1351 | 14.0k | { |
1352 | 14.0k | hash_table->hashes[node_index] = key_hash; |
1353 | 14.0k | key_to_keep = new_key; |
1354 | 14.0k | } |
1355 | | |
1356 | | /* Resize key/value arrays and split table as necessary */ |
1357 | 14.0k | g_hash_table_ensure_keyval_fits (hash_table, key_to_keep, new_value); |
1358 | 14.0k | g_hash_table_assign_key_or_value (hash_table->keys, node_index, hash_table->have_big_keys, key_to_keep); |
1359 | | |
1360 | | /* Step 3: Actually do the write */ |
1361 | 14.0k | g_hash_table_assign_key_or_value (hash_table->values, node_index, hash_table->have_big_values, new_value); |
1362 | | |
1363 | | /* Now, the bookkeeping... */ |
1364 | 14.0k | if (!already_exists) |
1365 | 14.0k | { |
1366 | 14.0k | hash_table->nnodes++; |
1367 | | |
1368 | 14.0k | if (HASH_IS_UNUSED (old_hash)) |
1369 | 14.0k | { |
1370 | | /* We replaced an empty node, and not a tombstone */ |
1371 | 14.0k | hash_table->noccupied++; |
1372 | 14.0k | g_hash_table_maybe_resize (hash_table); |
1373 | 14.0k | } |
1374 | | |
1375 | 14.0k | #ifndef G_DISABLE_ASSERT |
1376 | 14.0k | hash_table->version++; |
1377 | 14.0k | #endif |
1378 | 14.0k | } |
1379 | | |
1380 | 14.0k | if (already_exists) |
1381 | 0 | { |
1382 | 0 | if (hash_table->key_destroy_func && !reusing_key) |
1383 | 0 | (* hash_table->key_destroy_func) (key_to_free); |
1384 | 0 | if (hash_table->value_destroy_func) |
1385 | 0 | (* hash_table->value_destroy_func) (value_to_free); |
1386 | 0 | } |
1387 | | |
1388 | 14.0k | return !already_exists; |
1389 | 14.0k | } |
1390 | | |
1391 | | /** |
1392 | | * g_hash_table_iter_replace: |
1393 | | * @iter: an initialized #GHashTableIter |
1394 | | * @value: the value to replace with |
1395 | | * |
1396 | | * Replaces the value currently pointed to by the iterator |
1397 | | * from its associated #GHashTable. Can only be called after |
1398 | | * g_hash_table_iter_next() returned %TRUE. |
1399 | | * |
1400 | | * If you supplied a @value_destroy_func when creating the |
1401 | | * #GHashTable, the old value is freed using that function. |
1402 | | * |
1403 | | * Since: 2.30 |
1404 | | */ |
1405 | | void |
1406 | | g_hash_table_iter_replace (GHashTableIter *iter, |
1407 | | gpointer value) |
1408 | 0 | { |
1409 | 0 | RealIter *ri; |
1410 | 0 | guint node_hash; |
1411 | 0 | gpointer key; |
1412 | |
|
1413 | 0 | ri = (RealIter *) iter; |
1414 | |
|
1415 | 0 | g_return_if_fail (ri != NULL); |
1416 | 0 | #ifndef G_DISABLE_ASSERT |
1417 | 0 | g_return_if_fail (ri->version == ri->hash_table->version); |
1418 | 0 | #endif |
1419 | 0 | g_return_if_fail (ri->position >= 0); |
1420 | 0 | g_return_if_fail ((gsize) ri->position < ri->hash_table->size); |
1421 | | |
1422 | 0 | node_hash = ri->hash_table->hashes[ri->position]; |
1423 | |
|
1424 | 0 | key = g_hash_table_fetch_key_or_value (ri->hash_table->keys, ri->position, ri->hash_table->have_big_keys); |
1425 | |
|
1426 | 0 | g_hash_table_insert_node (ri->hash_table, ri->position, node_hash, key, value, TRUE, TRUE); |
1427 | |
|
1428 | 0 | #ifndef G_DISABLE_ASSERT |
1429 | 0 | ri->version++; |
1430 | 0 | ri->hash_table->version++; |
1431 | 0 | #endif |
1432 | 0 | } |
1433 | | |
1434 | | /** |
1435 | | * g_hash_table_iter_steal: |
1436 | | * @iter: an initialized #GHashTableIter |
1437 | | * |
1438 | | * Removes the key/value pair currently pointed to by the |
1439 | | * iterator from its associated #GHashTable, without calling |
1440 | | * the key and value destroy functions. Can only be called |
1441 | | * after g_hash_table_iter_next() returned %TRUE, and cannot |
1442 | | * be called more than once for the same key/value pair. |
1443 | | * |
1444 | | * Since: 2.16 |
1445 | | */ |
1446 | | void |
1447 | | g_hash_table_iter_steal (GHashTableIter *iter) |
1448 | 0 | { |
1449 | 0 | iter_remove_or_steal ((RealIter *) iter, FALSE); |
1450 | 0 | } |
1451 | | |
1452 | | |
1453 | | /** |
1454 | | * g_hash_table_ref: |
1455 | | * @hash_table: a valid #GHashTable |
1456 | | * |
1457 | | * Atomically increments the reference count of @hash_table by one. |
1458 | | * This function is MT-safe and may be called from any thread. |
1459 | | * |
1460 | | * Returns: the passed in #GHashTable |
1461 | | * |
1462 | | * Since: 2.10 |
1463 | | */ |
1464 | | GHashTable * |
1465 | | g_hash_table_ref (GHashTable *hash_table) |
1466 | 0 | { |
1467 | 0 | g_return_val_if_fail (hash_table != NULL, NULL); |
1468 | | |
1469 | 0 | g_atomic_ref_count_inc (&hash_table->ref_count); |
1470 | |
|
1471 | 0 | return hash_table; |
1472 | 0 | } |
1473 | | |
1474 | | /** |
1475 | | * g_hash_table_unref: |
1476 | | * @hash_table: a valid #GHashTable |
1477 | | * |
1478 | | * Atomically decrements the reference count of @hash_table by one. |
1479 | | * If the reference count drops to 0, all keys and values will be |
1480 | | * destroyed, and all memory allocated by the hash table is released. |
1481 | | * This function is MT-safe and may be called from any thread. |
1482 | | * |
1483 | | * Since: 2.10 |
1484 | | */ |
1485 | | void |
1486 | | g_hash_table_unref (GHashTable *hash_table) |
1487 | 10.1k | { |
1488 | 10.1k | g_return_if_fail (hash_table != NULL); |
1489 | | |
1490 | 10.1k | if (g_atomic_ref_count_dec (&hash_table->ref_count)) |
1491 | 10.1k | { |
1492 | 10.1k | g_hash_table_remove_all_nodes (hash_table, TRUE, TRUE); |
1493 | 10.1k | if (hash_table->keys != hash_table->values) |
1494 | 8.48k | g_free (hash_table->values); |
1495 | 10.1k | g_free (hash_table->keys); |
1496 | 10.1k | g_free (hash_table->hashes); |
1497 | 10.1k | g_slice_free (GHashTable, hash_table); |
1498 | 10.1k | } |
1499 | 10.1k | } |
1500 | | |
1501 | | /** |
1502 | | * g_hash_table_destroy: |
1503 | | * @hash_table: a #GHashTable |
1504 | | * |
1505 | | * Destroys all keys and values in the #GHashTable and decrements its |
1506 | | * reference count by 1. If keys and/or values are dynamically allocated, |
1507 | | * you should either free them first or create the #GHashTable with destroy |
1508 | | * notifiers using g_hash_table_new_full(). In the latter case the destroy |
1509 | | * functions you supplied will be called on all keys and values during the |
1510 | | * destruction phase. |
1511 | | */ |
1512 | | void |
1513 | | g_hash_table_destroy (GHashTable *hash_table) |
1514 | 9.33k | { |
1515 | 9.33k | g_return_if_fail (hash_table != NULL); |
1516 | | |
1517 | 9.33k | g_hash_table_remove_all (hash_table); |
1518 | 9.33k | g_hash_table_unref (hash_table); |
1519 | 9.33k | } |
1520 | | |
1521 | | /** |
1522 | | * g_hash_table_lookup: |
1523 | | * @hash_table: a #GHashTable |
1524 | | * @key: the key to look up |
1525 | | * |
1526 | | * Looks up a key in a #GHashTable. Note that this function cannot |
1527 | | * distinguish between a key that is not present and one which is present |
1528 | | * and has the value %NULL. If you need this distinction, use |
1529 | | * g_hash_table_lookup_extended(). |
1530 | | * |
1531 | | * Returns: (nullable): the associated value, or %NULL if the key is not found |
1532 | | */ |
1533 | | gpointer |
1534 | | g_hash_table_lookup (GHashTable *hash_table, |
1535 | | gconstpointer key) |
1536 | 142k | { |
1537 | 142k | guint node_index; |
1538 | 142k | guint node_hash; |
1539 | | |
1540 | 142k | g_return_val_if_fail (hash_table != NULL, NULL); |
1541 | | |
1542 | 142k | node_index = g_hash_table_lookup_node (hash_table, key, &node_hash); |
1543 | | |
1544 | 142k | return HASH_IS_REAL (hash_table->hashes[node_index]) |
1545 | 142k | ? g_hash_table_fetch_key_or_value (hash_table->values, node_index, hash_table->have_big_values) |
1546 | 142k | : NULL; |
1547 | 142k | } |
1548 | | |
1549 | | /** |
1550 | | * g_hash_table_lookup_extended: |
1551 | | * @hash_table: a #GHashTable |
1552 | | * @lookup_key: the key to look up |
1553 | | * @orig_key: (out) (optional): return location for the original key |
1554 | | * @value: (out) (optional) (nullable): return location for the value associated |
1555 | | * with the key |
1556 | | * |
1557 | | * Looks up a key in the #GHashTable, returning the original key and the |
1558 | | * associated value and a #gboolean which is %TRUE if the key was found. This |
1559 | | * is useful if you need to free the memory allocated for the original key, |
1560 | | * for example before calling g_hash_table_remove(). |
1561 | | * |
1562 | | * You can actually pass %NULL for @lookup_key to test |
1563 | | * whether the %NULL key exists, provided the hash and equal functions |
1564 | | * of @hash_table are %NULL-safe. |
1565 | | * |
1566 | | * Returns: %TRUE if the key was found in the #GHashTable |
1567 | | */ |
1568 | | gboolean |
1569 | | g_hash_table_lookup_extended (GHashTable *hash_table, |
1570 | | gconstpointer lookup_key, |
1571 | | gpointer *orig_key, |
1572 | | gpointer *value) |
1573 | 0 | { |
1574 | 0 | guint node_index; |
1575 | 0 | guint node_hash; |
1576 | |
|
1577 | 0 | g_return_val_if_fail (hash_table != NULL, FALSE); |
1578 | | |
1579 | 0 | node_index = g_hash_table_lookup_node (hash_table, lookup_key, &node_hash); |
1580 | |
|
1581 | 0 | if (!HASH_IS_REAL (hash_table->hashes[node_index])) |
1582 | 0 | { |
1583 | 0 | if (orig_key != NULL) |
1584 | 0 | *orig_key = NULL; |
1585 | 0 | if (value != NULL) |
1586 | 0 | *value = NULL; |
1587 | |
|
1588 | 0 | return FALSE; |
1589 | 0 | } |
1590 | | |
1591 | 0 | if (orig_key) |
1592 | 0 | *orig_key = g_hash_table_fetch_key_or_value (hash_table->keys, node_index, hash_table->have_big_keys); |
1593 | |
|
1594 | 0 | if (value) |
1595 | 0 | *value = g_hash_table_fetch_key_or_value (hash_table->values, node_index, hash_table->have_big_values); |
1596 | |
|
1597 | 0 | return TRUE; |
1598 | 0 | } |
1599 | | |
1600 | | /* |
1601 | | * g_hash_table_insert_internal: |
1602 | | * @hash_table: our #GHashTable |
1603 | | * @key: the key to insert |
1604 | | * @value: the value to insert |
1605 | | * @keep_new_key: if %TRUE and this key already exists in the table |
1606 | | * then call the destroy notify function on the old key. If %FALSE |
1607 | | * then call the destroy notify function on the new key. |
1608 | | * |
1609 | | * Implements the common logic for the g_hash_table_insert() and |
1610 | | * g_hash_table_replace() functions. |
1611 | | * |
1612 | | * Do a lookup of @key. If it is found, replace it with the new |
1613 | | * @value (and perhaps the new @key). If it is not found, create |
1614 | | * a new node. |
1615 | | * |
1616 | | * Returns: %TRUE if the key did not exist yet |
1617 | | */ |
1618 | | static gboolean |
1619 | | g_hash_table_insert_internal (GHashTable *hash_table, |
1620 | | gpointer key, |
1621 | | gpointer value, |
1622 | | gboolean keep_new_key) |
1623 | 14.0k | { |
1624 | 14.0k | guint key_hash; |
1625 | 14.0k | guint node_index; |
1626 | | |
1627 | 14.0k | g_return_val_if_fail (hash_table != NULL, FALSE); |
1628 | | |
1629 | 14.0k | node_index = g_hash_table_lookup_node (hash_table, key, &key_hash); |
1630 | | |
1631 | 14.0k | return g_hash_table_insert_node (hash_table, node_index, key_hash, key, value, keep_new_key, FALSE); |
1632 | 14.0k | } |
1633 | | |
1634 | | /** |
1635 | | * g_hash_table_insert: |
1636 | | * @hash_table: a #GHashTable |
1637 | | * @key: a key to insert |
1638 | | * @value: the value to associate with the key |
1639 | | * |
1640 | | * Inserts a new key and value into a #GHashTable. |
1641 | | * |
1642 | | * If the key already exists in the #GHashTable its current |
1643 | | * value is replaced with the new value. If you supplied a |
1644 | | * @value_destroy_func when creating the #GHashTable, the old |
1645 | | * value is freed using that function. If you supplied a |
1646 | | * @key_destroy_func when creating the #GHashTable, the passed |
1647 | | * key is freed using that function. |
1648 | | * |
1649 | | * Starting from GLib 2.40, this function returns a boolean value to |
1650 | | * indicate whether the newly added value was already in the hash table |
1651 | | * or not. |
1652 | | * |
1653 | | * Returns: %TRUE if the key did not exist yet |
1654 | | */ |
1655 | | gboolean |
1656 | | g_hash_table_insert (GHashTable *hash_table, |
1657 | | gpointer key, |
1658 | | gpointer value) |
1659 | 5.67k | { |
1660 | 5.67k | return g_hash_table_insert_internal (hash_table, key, value, FALSE); |
1661 | 5.67k | } |
1662 | | |
1663 | | /** |
1664 | | * g_hash_table_replace: |
1665 | | * @hash_table: a #GHashTable |
1666 | | * @key: a key to insert |
1667 | | * @value: the value to associate with the key |
1668 | | * |
1669 | | * Inserts a new key and value into a #GHashTable similar to |
1670 | | * g_hash_table_insert(). The difference is that if the key |
1671 | | * already exists in the #GHashTable, it gets replaced by the |
1672 | | * new key. If you supplied a @value_destroy_func when creating |
1673 | | * the #GHashTable, the old value is freed using that function. |
1674 | | * If you supplied a @key_destroy_func when creating the |
1675 | | * #GHashTable, the old key is freed using that function. |
1676 | | * |
1677 | | * Starting from GLib 2.40, this function returns a boolean value to |
1678 | | * indicate whether the newly added value was already in the hash table |
1679 | | * or not. |
1680 | | * |
1681 | | * Returns: %TRUE if the key did not exist yet |
1682 | | */ |
1683 | | gboolean |
1684 | | g_hash_table_replace (GHashTable *hash_table, |
1685 | | gpointer key, |
1686 | | gpointer value) |
1687 | 8.38k | { |
1688 | 8.38k | return g_hash_table_insert_internal (hash_table, key, value, TRUE); |
1689 | 8.38k | } |
1690 | | |
1691 | | /** |
1692 | | * g_hash_table_add: |
1693 | | * @hash_table: a #GHashTable |
1694 | | * @key: (transfer full): a key to insert |
1695 | | * |
1696 | | * This is a convenience function for using a #GHashTable as a set. It |
1697 | | * is equivalent to calling g_hash_table_replace() with @key as both the |
1698 | | * key and the value. |
1699 | | * |
1700 | | * In particular, this means that if @key already exists in the hash table, then |
1701 | | * the old copy of @key in the hash table is freed and @key replaces it in the |
1702 | | * table. |
1703 | | * |
1704 | | * When a hash table only ever contains keys that have themselves as the |
1705 | | * corresponding value it is able to be stored more efficiently. See |
1706 | | * the discussion in the section description. |
1707 | | * |
1708 | | * Starting from GLib 2.40, this function returns a boolean value to |
1709 | | * indicate whether the newly added value was already in the hash table |
1710 | | * or not. |
1711 | | * |
1712 | | * Returns: %TRUE if the key did not exist yet |
1713 | | * |
1714 | | * Since: 2.32 |
1715 | | */ |
1716 | | gboolean |
1717 | | g_hash_table_add (GHashTable *hash_table, |
1718 | | gpointer key) |
1719 | 8 | { |
1720 | 8 | return g_hash_table_insert_internal (hash_table, key, key, TRUE); |
1721 | 8 | } |
1722 | | |
1723 | | /** |
1724 | | * g_hash_table_contains: |
1725 | | * @hash_table: a #GHashTable |
1726 | | * @key: a key to check |
1727 | | * |
1728 | | * Checks if @key is in @hash_table. |
1729 | | * |
1730 | | * Returns: %TRUE if @key is in @hash_table, %FALSE otherwise. |
1731 | | * |
1732 | | * Since: 2.32 |
1733 | | **/ |
1734 | | gboolean |
1735 | | g_hash_table_contains (GHashTable *hash_table, |
1736 | | gconstpointer key) |
1737 | 103 | { |
1738 | 103 | guint node_index; |
1739 | 103 | guint node_hash; |
1740 | | |
1741 | 103 | g_return_val_if_fail (hash_table != NULL, FALSE); |
1742 | | |
1743 | 103 | node_index = g_hash_table_lookup_node (hash_table, key, &node_hash); |
1744 | | |
1745 | 103 | return HASH_IS_REAL (hash_table->hashes[node_index]); |
1746 | 103 | } |
1747 | | |
1748 | | /* |
1749 | | * g_hash_table_remove_internal: |
1750 | | * @hash_table: our #GHashTable |
1751 | | * @key: the key to remove |
1752 | | * @notify: %TRUE if the destroy notify handlers are to be called |
1753 | | * Returns: %TRUE if a node was found and removed, else %FALSE |
1754 | | * |
1755 | | * Implements the common logic for the g_hash_table_remove() and |
1756 | | * g_hash_table_steal() functions. |
1757 | | * |
1758 | | * Do a lookup of @key and remove it if it is found, calling the |
1759 | | * destroy notify handlers only if @notify is %TRUE. |
1760 | | */ |
1761 | | static gboolean |
1762 | | g_hash_table_remove_internal (GHashTable *hash_table, |
1763 | | gconstpointer key, |
1764 | | gboolean notify) |
1765 | 5.25k | { |
1766 | 5.25k | guint node_index; |
1767 | 5.25k | guint node_hash; |
1768 | | |
1769 | 5.25k | g_return_val_if_fail (hash_table != NULL, FALSE); |
1770 | | |
1771 | 5.25k | node_index = g_hash_table_lookup_node (hash_table, key, &node_hash); |
1772 | | |
1773 | 5.25k | if (!HASH_IS_REAL (hash_table->hashes[node_index])) |
1774 | 0 | return FALSE; |
1775 | | |
1776 | 5.25k | g_hash_table_remove_node (hash_table, node_index, notify); |
1777 | 5.25k | g_hash_table_maybe_resize (hash_table); |
1778 | | |
1779 | 5.25k | #ifndef G_DISABLE_ASSERT |
1780 | 5.25k | hash_table->version++; |
1781 | 5.25k | #endif |
1782 | | |
1783 | 5.25k | return TRUE; |
1784 | 5.25k | } |
1785 | | |
1786 | | /** |
1787 | | * g_hash_table_remove: |
1788 | | * @hash_table: a #GHashTable |
1789 | | * @key: the key to remove |
1790 | | * |
1791 | | * Removes a key and its associated value from a #GHashTable. |
1792 | | * |
1793 | | * If the #GHashTable was created using g_hash_table_new_full(), the |
1794 | | * key and value are freed using the supplied destroy functions, otherwise |
1795 | | * you have to make sure that any dynamically allocated values are freed |
1796 | | * yourself. |
1797 | | * |
1798 | | * Returns: %TRUE if the key was found and removed from the #GHashTable |
1799 | | */ |
1800 | | gboolean |
1801 | | g_hash_table_remove (GHashTable *hash_table, |
1802 | | gconstpointer key) |
1803 | 5.25k | { |
1804 | 5.25k | return g_hash_table_remove_internal (hash_table, key, TRUE); |
1805 | 5.25k | } |
1806 | | |
1807 | | /** |
1808 | | * g_hash_table_steal: |
1809 | | * @hash_table: a #GHashTable |
1810 | | * @key: the key to remove |
1811 | | * |
1812 | | * Removes a key and its associated value from a #GHashTable without |
1813 | | * calling the key and value destroy functions. |
1814 | | * |
1815 | | * Returns: %TRUE if the key was found and removed from the #GHashTable |
1816 | | */ |
1817 | | gboolean |
1818 | | g_hash_table_steal (GHashTable *hash_table, |
1819 | | gconstpointer key) |
1820 | 0 | { |
1821 | 0 | return g_hash_table_remove_internal (hash_table, key, FALSE); |
1822 | 0 | } |
1823 | | |
1824 | | /** |
1825 | | * g_hash_table_steal_extended: |
1826 | | * @hash_table: a #GHashTable |
1827 | | * @lookup_key: the key to look up |
1828 | | * @stolen_key: (out) (optional) (transfer full): return location for the |
1829 | | * original key |
1830 | | * @stolen_value: (out) (optional) (nullable) (transfer full): return location |
1831 | | * for the value associated with the key |
1832 | | * |
1833 | | * Looks up a key in the #GHashTable, stealing the original key and the |
1834 | | * associated value and returning %TRUE if the key was found. If the key was |
1835 | | * not found, %FALSE is returned. |
1836 | | * |
1837 | | * If found, the stolen key and value are removed from the hash table without |
1838 | | * calling the key and value destroy functions, and ownership is transferred to |
1839 | | * the caller of this method, as with g_hash_table_steal(). That is the case |
1840 | | * regardless whether @stolen_key or @stolen_value output parameters are |
1841 | | * requested. |
1842 | | * |
1843 | | * You can pass %NULL for @lookup_key, provided the hash and equal functions |
1844 | | * of @hash_table are %NULL-safe. |
1845 | | * |
1846 | | * The dictionary implementation optimizes for having all values identical to |
1847 | | * their keys, for example by using g_hash_table_add(). When stealing both the |
1848 | | * key and the value from such a dictionary, the value will be %NULL. |
1849 | | * |
1850 | | * Returns: %TRUE if the key was found in the #GHashTable |
1851 | | * Since: 2.58 |
1852 | | */ |
1853 | | gboolean |
1854 | | g_hash_table_steal_extended (GHashTable *hash_table, |
1855 | | gconstpointer lookup_key, |
1856 | | gpointer *stolen_key, |
1857 | | gpointer *stolen_value) |
1858 | 0 | { |
1859 | 0 | guint node_index; |
1860 | 0 | guint node_hash; |
1861 | |
|
1862 | 0 | g_return_val_if_fail (hash_table != NULL, FALSE); |
1863 | | |
1864 | 0 | node_index = g_hash_table_lookup_node (hash_table, lookup_key, &node_hash); |
1865 | |
|
1866 | 0 | if (!HASH_IS_REAL (hash_table->hashes[node_index])) |
1867 | 0 | { |
1868 | 0 | if (stolen_key != NULL) |
1869 | 0 | *stolen_key = NULL; |
1870 | 0 | if (stolen_value != NULL) |
1871 | 0 | *stolen_value = NULL; |
1872 | 0 | return FALSE; |
1873 | 0 | } |
1874 | | |
1875 | 0 | if (stolen_key != NULL) |
1876 | 0 | { |
1877 | 0 | *stolen_key = g_hash_table_fetch_key_or_value (hash_table->keys, node_index, hash_table->have_big_keys); |
1878 | 0 | g_hash_table_assign_key_or_value (hash_table->keys, node_index, hash_table->have_big_keys, NULL); |
1879 | 0 | } |
1880 | |
|
1881 | 0 | if (stolen_value != NULL) |
1882 | 0 | { |
1883 | 0 | *stolen_value = g_hash_table_fetch_key_or_value (hash_table->values, node_index, hash_table->have_big_values); |
1884 | 0 | g_hash_table_assign_key_or_value (hash_table->values, node_index, hash_table->have_big_values, NULL); |
1885 | 0 | } |
1886 | |
|
1887 | 0 | g_hash_table_remove_node (hash_table, node_index, FALSE); |
1888 | 0 | g_hash_table_maybe_resize (hash_table); |
1889 | |
|
1890 | 0 | #ifndef G_DISABLE_ASSERT |
1891 | 0 | hash_table->version++; |
1892 | 0 | #endif |
1893 | |
|
1894 | 0 | return TRUE; |
1895 | 0 | } |
1896 | | |
1897 | | /** |
1898 | | * g_hash_table_remove_all: |
1899 | | * @hash_table: a #GHashTable |
1900 | | * |
1901 | | * Removes all keys and their associated values from a #GHashTable. |
1902 | | * |
1903 | | * If the #GHashTable was created using g_hash_table_new_full(), |
1904 | | * the keys and values are freed using the supplied destroy functions, |
1905 | | * otherwise you have to make sure that any dynamically allocated |
1906 | | * values are freed yourself. |
1907 | | * |
1908 | | * Since: 2.12 |
1909 | | */ |
1910 | | void |
1911 | | g_hash_table_remove_all (GHashTable *hash_table) |
1912 | 12.0k | { |
1913 | 12.0k | g_return_if_fail (hash_table != NULL); |
1914 | | |
1915 | 12.0k | #ifndef G_DISABLE_ASSERT |
1916 | 12.0k | if (hash_table->nnodes != 0) |
1917 | 5.34k | hash_table->version++; |
1918 | 12.0k | #endif |
1919 | | |
1920 | 12.0k | g_hash_table_remove_all_nodes (hash_table, TRUE, FALSE); |
1921 | 12.0k | g_hash_table_maybe_resize (hash_table); |
1922 | 12.0k | } |
1923 | | |
1924 | | /** |
1925 | | * g_hash_table_steal_all: |
1926 | | * @hash_table: a #GHashTable |
1927 | | * |
1928 | | * Removes all keys and their associated values from a #GHashTable |
1929 | | * without calling the key and value destroy functions. |
1930 | | * |
1931 | | * Since: 2.12 |
1932 | | */ |
1933 | | void |
1934 | | g_hash_table_steal_all (GHashTable *hash_table) |
1935 | 0 | { |
1936 | 0 | g_return_if_fail (hash_table != NULL); |
1937 | | |
1938 | 0 | #ifndef G_DISABLE_ASSERT |
1939 | 0 | if (hash_table->nnodes != 0) |
1940 | 0 | hash_table->version++; |
1941 | 0 | #endif |
1942 | |
|
1943 | 0 | g_hash_table_remove_all_nodes (hash_table, FALSE, FALSE); |
1944 | 0 | g_hash_table_maybe_resize (hash_table); |
1945 | 0 | } |
1946 | | |
1947 | | /** |
1948 | | * g_hash_table_steal_all_keys: (skip) |
1949 | | * @hash_table: a #GHashTable |
1950 | | * |
1951 | | * Removes all keys and their associated values from a #GHashTable |
1952 | | * without calling the key destroy functions, returning the keys |
1953 | | * as a #GPtrArray with the free func set to the @hash_table key |
1954 | | * destroy function. |
1955 | | * |
1956 | | * Returns: (transfer container): a #GPtrArray containing each key of |
1957 | | * the table. Unref with with g_ptr_array_unref() when done. |
1958 | | * |
1959 | | * Since: 2.76 |
1960 | | */ |
1961 | | GPtrArray * |
1962 | | g_hash_table_steal_all_keys (GHashTable *hash_table) |
1963 | 0 | { |
1964 | 0 | GPtrArray *array; |
1965 | 0 | GDestroyNotify key_destroy_func; |
1966 | |
|
1967 | 0 | g_return_val_if_fail (hash_table != NULL, NULL); |
1968 | | |
1969 | 0 | array = g_hash_table_get_keys_as_ptr_array (hash_table); |
1970 | | |
1971 | | /* Ignore the key destroy notify calls during removal, and use it for the |
1972 | | * array elements instead, but restore it after the hash table has been |
1973 | | * cleared, so that newly added keys will continue using it. |
1974 | | */ |
1975 | 0 | key_destroy_func = g_steal_pointer (&hash_table->key_destroy_func); |
1976 | 0 | g_ptr_array_set_free_func (array, key_destroy_func); |
1977 | |
|
1978 | 0 | g_hash_table_remove_all (hash_table); |
1979 | 0 | hash_table->key_destroy_func = g_steal_pointer (&key_destroy_func); |
1980 | |
|
1981 | 0 | return array; |
1982 | 0 | } |
1983 | | |
1984 | | /** |
1985 | | * g_hash_table_steal_all_values: (skip) |
1986 | | * @hash_table: a #GHashTable |
1987 | | * |
1988 | | * Removes all keys and their associated values from a #GHashTable |
1989 | | * without calling the value destroy functions, returning the values |
1990 | | * as a #GPtrArray with the free func set to the @hash_table value |
1991 | | * destroy function. |
1992 | | * |
1993 | | * Returns: (transfer container): a #GPtrArray containing each value of |
1994 | | * the table. Unref with with g_ptr_array_unref() when done. |
1995 | | * |
1996 | | * Since: 2.76 |
1997 | | */ |
1998 | | GPtrArray * |
1999 | | g_hash_table_steal_all_values (GHashTable *hash_table) |
2000 | 0 | { |
2001 | 0 | GPtrArray *array; |
2002 | 0 | GDestroyNotify value_destroy_func; |
2003 | |
|
2004 | 0 | g_return_val_if_fail (hash_table != NULL, NULL); |
2005 | | |
2006 | 0 | array = g_hash_table_get_values_as_ptr_array (hash_table); |
2007 | | |
2008 | | /* Ignore the value destroy notify calls during removal, and use it for the |
2009 | | * array elements instead, but restore it after the hash table has been |
2010 | | * cleared, so that newly added values will continue using it. |
2011 | | */ |
2012 | 0 | value_destroy_func = g_steal_pointer (&hash_table->value_destroy_func); |
2013 | 0 | g_ptr_array_set_free_func (array, value_destroy_func); |
2014 | |
|
2015 | 0 | g_hash_table_remove_all (hash_table); |
2016 | 0 | hash_table->value_destroy_func = g_steal_pointer (&value_destroy_func); |
2017 | |
|
2018 | 0 | return array; |
2019 | 0 | } |
2020 | | |
2021 | | /* |
2022 | | * g_hash_table_foreach_remove_or_steal: |
2023 | | * @hash_table: a #GHashTable |
2024 | | * @func: the user's callback function |
2025 | | * @user_data: data for @func |
2026 | | * @notify: %TRUE if the destroy notify handlers are to be called |
2027 | | * |
2028 | | * Implements the common logic for g_hash_table_foreach_remove() |
2029 | | * and g_hash_table_foreach_steal(). |
2030 | | * |
2031 | | * Iterates over every node in the table, calling @func with the key |
2032 | | * and value of the node (and @user_data). If @func returns %TRUE the |
2033 | | * node is removed from the table. |
2034 | | * |
2035 | | * If @notify is true then the destroy notify handlers will be called |
2036 | | * for each removed node. |
2037 | | */ |
2038 | | static guint |
2039 | | g_hash_table_foreach_remove_or_steal (GHashTable *hash_table, |
2040 | | GHRFunc func, |
2041 | | gpointer user_data, |
2042 | | gboolean notify) |
2043 | 0 | { |
2044 | 0 | guint deleted = 0; |
2045 | 0 | gsize i; |
2046 | 0 | #ifndef G_DISABLE_ASSERT |
2047 | 0 | gint version = hash_table->version; |
2048 | 0 | #endif |
2049 | |
|
2050 | 0 | for (i = 0; i < hash_table->size; i++) |
2051 | 0 | { |
2052 | 0 | guint node_hash = hash_table->hashes[i]; |
2053 | 0 | gpointer node_key = g_hash_table_fetch_key_or_value (hash_table->keys, i, hash_table->have_big_keys); |
2054 | 0 | gpointer node_value = g_hash_table_fetch_key_or_value (hash_table->values, i, hash_table->have_big_values); |
2055 | |
|
2056 | 0 | if (HASH_IS_REAL (node_hash) && |
2057 | 0 | (* func) (node_key, node_value, user_data)) |
2058 | 0 | { |
2059 | 0 | g_hash_table_remove_node (hash_table, i, notify); |
2060 | 0 | deleted++; |
2061 | 0 | } |
2062 | |
|
2063 | 0 | #ifndef G_DISABLE_ASSERT |
2064 | 0 | g_return_val_if_fail (version == hash_table->version, 0); |
2065 | 0 | #endif |
2066 | 0 | } |
2067 | | |
2068 | 0 | g_hash_table_maybe_resize (hash_table); |
2069 | |
|
2070 | 0 | #ifndef G_DISABLE_ASSERT |
2071 | 0 | if (deleted > 0) |
2072 | 0 | hash_table->version++; |
2073 | 0 | #endif |
2074 | |
|
2075 | 0 | return deleted; |
2076 | 0 | } |
2077 | | |
2078 | | /** |
2079 | | * g_hash_table_foreach_remove: |
2080 | | * @hash_table: a #GHashTable |
2081 | | * @func: the function to call for each key/value pair |
2082 | | * @user_data: user data to pass to the function |
2083 | | * |
2084 | | * Calls the given function for each key/value pair in the |
2085 | | * #GHashTable. If the function returns %TRUE, then the key/value |
2086 | | * pair is removed from the #GHashTable. If you supplied key or |
2087 | | * value destroy functions when creating the #GHashTable, they are |
2088 | | * used to free the memory allocated for the removed keys and values. |
2089 | | * |
2090 | | * See #GHashTableIter for an alternative way to loop over the |
2091 | | * key/value pairs in the hash table. |
2092 | | * |
2093 | | * Returns: the number of key/value pairs removed |
2094 | | */ |
2095 | | guint |
2096 | | g_hash_table_foreach_remove (GHashTable *hash_table, |
2097 | | GHRFunc func, |
2098 | | gpointer user_data) |
2099 | 0 | { |
2100 | 0 | g_return_val_if_fail (hash_table != NULL, 0); |
2101 | 0 | g_return_val_if_fail (func != NULL, 0); |
2102 | | |
2103 | 0 | return g_hash_table_foreach_remove_or_steal (hash_table, func, user_data, TRUE); |
2104 | 0 | } |
2105 | | |
2106 | | /** |
2107 | | * g_hash_table_foreach_steal: |
2108 | | * @hash_table: a #GHashTable |
2109 | | * @func: the function to call for each key/value pair |
2110 | | * @user_data: user data to pass to the function |
2111 | | * |
2112 | | * Calls the given function for each key/value pair in the |
2113 | | * #GHashTable. If the function returns %TRUE, then the key/value |
2114 | | * pair is removed from the #GHashTable, but no key or value |
2115 | | * destroy functions are called. |
2116 | | * |
2117 | | * See #GHashTableIter for an alternative way to loop over the |
2118 | | * key/value pairs in the hash table. |
2119 | | * |
2120 | | * Returns: the number of key/value pairs removed. |
2121 | | */ |
2122 | | guint |
2123 | | g_hash_table_foreach_steal (GHashTable *hash_table, |
2124 | | GHRFunc func, |
2125 | | gpointer user_data) |
2126 | 0 | { |
2127 | 0 | g_return_val_if_fail (hash_table != NULL, 0); |
2128 | 0 | g_return_val_if_fail (func != NULL, 0); |
2129 | | |
2130 | 0 | return g_hash_table_foreach_remove_or_steal (hash_table, func, user_data, FALSE); |
2131 | 0 | } |
2132 | | |
2133 | | /** |
2134 | | * g_hash_table_foreach: |
2135 | | * @hash_table: a #GHashTable |
2136 | | * @func: the function to call for each key/value pair |
2137 | | * @user_data: user data to pass to the function |
2138 | | * |
2139 | | * Calls the given function for each of the key/value pairs in the |
2140 | | * #GHashTable. The function is passed the key and value of each |
2141 | | * pair, and the given @user_data parameter. The hash table may not |
2142 | | * be modified while iterating over it (you can't add/remove |
2143 | | * items). To remove all items matching a predicate, use |
2144 | | * g_hash_table_foreach_remove(). |
2145 | | * |
2146 | | * The order in which g_hash_table_foreach() iterates over the keys/values in |
2147 | | * the hash table is not defined. |
2148 | | * |
2149 | | * See g_hash_table_find() for performance caveats for linear |
2150 | | * order searches in contrast to g_hash_table_lookup(). |
2151 | | */ |
2152 | | void |
2153 | | g_hash_table_foreach (GHashTable *hash_table, |
2154 | | GHFunc func, |
2155 | | gpointer user_data) |
2156 | 2 | { |
2157 | 2 | gsize i; |
2158 | 2 | #ifndef G_DISABLE_ASSERT |
2159 | 2 | gint version; |
2160 | 2 | #endif |
2161 | | |
2162 | 2 | g_return_if_fail (hash_table != NULL); |
2163 | 2 | g_return_if_fail (func != NULL); |
2164 | | |
2165 | 2 | #ifndef G_DISABLE_ASSERT |
2166 | 2 | version = hash_table->version; |
2167 | 2 | #endif |
2168 | | |
2169 | 18 | for (i = 0; i < hash_table->size; i++) |
2170 | 16 | { |
2171 | 16 | guint node_hash = hash_table->hashes[i]; |
2172 | 16 | gpointer node_key = g_hash_table_fetch_key_or_value (hash_table->keys, i, hash_table->have_big_keys); |
2173 | 16 | gpointer node_value = g_hash_table_fetch_key_or_value (hash_table->values, i, hash_table->have_big_values); |
2174 | | |
2175 | 16 | if (HASH_IS_REAL (node_hash)) |
2176 | 10 | (* func) (node_key, node_value, user_data); |
2177 | | |
2178 | 16 | #ifndef G_DISABLE_ASSERT |
2179 | 16 | g_return_if_fail (version == hash_table->version); |
2180 | 16 | #endif |
2181 | 16 | } |
2182 | 2 | } |
2183 | | |
2184 | | /** |
2185 | | * g_hash_table_find: |
2186 | | * @hash_table: a #GHashTable |
2187 | | * @predicate: function to test the key/value pairs for a certain property |
2188 | | * @user_data: user data to pass to the function |
2189 | | * |
2190 | | * Calls the given function for key/value pairs in the #GHashTable |
2191 | | * until @predicate returns %TRUE. The function is passed the key |
2192 | | * and value of each pair, and the given @user_data parameter. The |
2193 | | * hash table may not be modified while iterating over it (you can't |
2194 | | * add/remove items). |
2195 | | * |
2196 | | * Note, that hash tables are really only optimized for forward |
2197 | | * lookups, i.e. g_hash_table_lookup(). So code that frequently issues |
2198 | | * g_hash_table_find() or g_hash_table_foreach() (e.g. in the order of |
2199 | | * once per every entry in a hash table) should probably be reworked |
2200 | | * to use additional or different data structures for reverse lookups |
2201 | | * (keep in mind that an O(n) find/foreach operation issued for all n |
2202 | | * values in a hash table ends up needing O(n*n) operations). |
2203 | | * |
2204 | | * Returns: (nullable): The value of the first key/value pair is returned, |
2205 | | * for which @predicate evaluates to %TRUE. If no pair with the |
2206 | | * requested property is found, %NULL is returned. |
2207 | | * |
2208 | | * Since: 2.4 |
2209 | | */ |
2210 | | gpointer |
2211 | | g_hash_table_find (GHashTable *hash_table, |
2212 | | GHRFunc predicate, |
2213 | | gpointer user_data) |
2214 | 0 | { |
2215 | 0 | gsize i; |
2216 | 0 | #ifndef G_DISABLE_ASSERT |
2217 | 0 | gint version; |
2218 | 0 | #endif |
2219 | 0 | gboolean match; |
2220 | |
|
2221 | 0 | g_return_val_if_fail (hash_table != NULL, NULL); |
2222 | 0 | g_return_val_if_fail (predicate != NULL, NULL); |
2223 | | |
2224 | 0 | #ifndef G_DISABLE_ASSERT |
2225 | 0 | version = hash_table->version; |
2226 | 0 | #endif |
2227 | |
|
2228 | 0 | match = FALSE; |
2229 | |
|
2230 | 0 | for (i = 0; i < hash_table->size; i++) |
2231 | 0 | { |
2232 | 0 | guint node_hash = hash_table->hashes[i]; |
2233 | 0 | gpointer node_key = g_hash_table_fetch_key_or_value (hash_table->keys, i, hash_table->have_big_keys); |
2234 | 0 | gpointer node_value = g_hash_table_fetch_key_or_value (hash_table->values, i, hash_table->have_big_values); |
2235 | |
|
2236 | 0 | if (HASH_IS_REAL (node_hash)) |
2237 | 0 | match = predicate (node_key, node_value, user_data); |
2238 | |
|
2239 | 0 | #ifndef G_DISABLE_ASSERT |
2240 | 0 | g_return_val_if_fail (version == hash_table->version, NULL); |
2241 | 0 | #endif |
2242 | | |
2243 | 0 | if (match) |
2244 | 0 | return node_value; |
2245 | 0 | } |
2246 | | |
2247 | 0 | return NULL; |
2248 | 0 | } |
2249 | | |
2250 | | /** |
2251 | | * g_hash_table_size: |
2252 | | * @hash_table: a #GHashTable |
2253 | | * |
2254 | | * Returns the number of elements contained in the #GHashTable. |
2255 | | * |
2256 | | * Returns: the number of key/value pairs in the #GHashTable. |
2257 | | */ |
2258 | | guint |
2259 | | g_hash_table_size (GHashTable *hash_table) |
2260 | 0 | { |
2261 | 0 | g_return_val_if_fail (hash_table != NULL, 0); |
2262 | | |
2263 | 0 | return hash_table->nnodes; |
2264 | 0 | } |
2265 | | |
2266 | | /** |
2267 | | * g_hash_table_get_keys: |
2268 | | * @hash_table: a #GHashTable |
2269 | | * |
2270 | | * Retrieves every key inside @hash_table. The returned data is valid |
2271 | | * until changes to the hash release those keys. |
2272 | | * |
2273 | | * This iterates over every entry in the hash table to build its return value. |
2274 | | * To iterate over the entries in a #GHashTable more efficiently, use a |
2275 | | * #GHashTableIter. |
2276 | | * |
2277 | | * Returns: (transfer container): a #GList containing all the keys |
2278 | | * inside the hash table. The content of the list is owned by the |
2279 | | * hash table and should not be modified or freed. Use g_list_free() |
2280 | | * when done using the list. |
2281 | | * |
2282 | | * Since: 2.14 |
2283 | | */ |
2284 | | GList * |
2285 | | g_hash_table_get_keys (GHashTable *hash_table) |
2286 | 0 | { |
2287 | 0 | gsize i; |
2288 | 0 | GList *retval; |
2289 | |
|
2290 | 0 | g_return_val_if_fail (hash_table != NULL, NULL); |
2291 | | |
2292 | 0 | retval = NULL; |
2293 | 0 | for (i = 0; i < hash_table->size; i++) |
2294 | 0 | { |
2295 | 0 | if (HASH_IS_REAL (hash_table->hashes[i])) |
2296 | 0 | retval = g_list_prepend (retval, g_hash_table_fetch_key_or_value (hash_table->keys, i, hash_table->have_big_keys)); |
2297 | 0 | } |
2298 | |
|
2299 | 0 | return retval; |
2300 | 0 | } |
2301 | | |
2302 | | /** |
2303 | | * g_hash_table_get_keys_as_array: |
2304 | | * @hash_table: a #GHashTable |
2305 | | * @length: (out) (optional): the length of the returned array |
2306 | | * |
2307 | | * Retrieves every key inside @hash_table, as an array. |
2308 | | * |
2309 | | * The returned array is %NULL-terminated but may contain %NULL as a |
2310 | | * key. Use @length to determine the true length if it's possible that |
2311 | | * %NULL was used as the value for a key. |
2312 | | * |
2313 | | * Note: in the common case of a string-keyed #GHashTable, the return |
2314 | | * value of this function can be conveniently cast to (const gchar **). |
2315 | | * |
2316 | | * This iterates over every entry in the hash table to build its return value. |
2317 | | * To iterate over the entries in a #GHashTable more efficiently, use a |
2318 | | * #GHashTableIter. |
2319 | | * |
2320 | | * You should always free the return result with g_free(). In the |
2321 | | * above-mentioned case of a string-keyed hash table, it may be |
2322 | | * appropriate to use g_strfreev() if you call g_hash_table_steal_all() |
2323 | | * first to transfer ownership of the keys. |
2324 | | * |
2325 | | * Returns: (array length=length) (transfer container): a |
2326 | | * %NULL-terminated array containing each key from the table. |
2327 | | * |
2328 | | * Since: 2.40 |
2329 | | **/ |
2330 | | gpointer * |
2331 | | g_hash_table_get_keys_as_array (GHashTable *hash_table, |
2332 | | guint *length) |
2333 | 0 | { |
2334 | 0 | gpointer *result; |
2335 | 0 | gsize i, j = 0; |
2336 | |
|
2337 | 0 | result = g_new (gpointer, hash_table->nnodes + 1); |
2338 | 0 | for (i = 0; i < hash_table->size; i++) |
2339 | 0 | { |
2340 | 0 | if (HASH_IS_REAL (hash_table->hashes[i])) |
2341 | 0 | result[j++] = g_hash_table_fetch_key_or_value (hash_table->keys, i, hash_table->have_big_keys); |
2342 | 0 | } |
2343 | 0 | g_assert (j == hash_table->nnodes); |
2344 | 0 | result[j] = NULL; |
2345 | |
|
2346 | 0 | if (length) |
2347 | 0 | *length = j; |
2348 | |
|
2349 | 0 | return result; |
2350 | 0 | } |
2351 | | |
2352 | | /** |
2353 | | * g_hash_table_get_keys_as_ptr_array: (skip) |
2354 | | * @hash_table: a #GHashTable |
2355 | | * |
2356 | | * Retrieves every key inside @hash_table, as a #GPtrArray. |
2357 | | * The returned data is valid until changes to the hash release those keys. |
2358 | | * |
2359 | | * This iterates over every entry in the hash table to build its return value. |
2360 | | * To iterate over the entries in a #GHashTable more efficiently, use a |
2361 | | * #GHashTableIter. |
2362 | | * |
2363 | | * You should always unref the returned array with g_ptr_array_unref(). |
2364 | | * |
2365 | | * Returns: (transfer container): a #GPtrArray containing each key from |
2366 | | * the table. Unref with with g_ptr_array_unref() when done. |
2367 | | * |
2368 | | * Since: 2.76 |
2369 | | **/ |
2370 | | GPtrArray * |
2371 | | g_hash_table_get_keys_as_ptr_array (GHashTable *hash_table) |
2372 | 0 | { |
2373 | 0 | GPtrArray *array; |
2374 | |
|
2375 | 0 | g_return_val_if_fail (hash_table != NULL, NULL); |
2376 | | |
2377 | 0 | array = g_ptr_array_sized_new (hash_table->size); |
2378 | 0 | for (gsize i = 0; i < hash_table->size; ++i) |
2379 | 0 | { |
2380 | 0 | if (HASH_IS_REAL (hash_table->hashes[i])) |
2381 | 0 | { |
2382 | 0 | g_ptr_array_add (array, g_hash_table_fetch_key_or_value ( |
2383 | 0 | hash_table->keys, i, hash_table->have_big_keys)); |
2384 | 0 | } |
2385 | 0 | } |
2386 | 0 | g_assert (array->len == hash_table->nnodes); |
2387 | | |
2388 | 0 | return array; |
2389 | 0 | } |
2390 | | |
2391 | | /** |
2392 | | * g_hash_table_get_values: |
2393 | | * @hash_table: a #GHashTable |
2394 | | * |
2395 | | * Retrieves every value inside @hash_table. The returned data |
2396 | | * is valid until @hash_table is modified. |
2397 | | * |
2398 | | * This iterates over every entry in the hash table to build its return value. |
2399 | | * To iterate over the entries in a #GHashTable more efficiently, use a |
2400 | | * #GHashTableIter. |
2401 | | * |
2402 | | * Returns: (transfer container): a #GList containing all the values |
2403 | | * inside the hash table. The content of the list is owned by the |
2404 | | * hash table and should not be modified or freed. Use g_list_free() |
2405 | | * when done using the list. |
2406 | | * |
2407 | | * Since: 2.14 |
2408 | | */ |
2409 | | GList * |
2410 | | g_hash_table_get_values (GHashTable *hash_table) |
2411 | 0 | { |
2412 | 0 | gsize i; |
2413 | 0 | GList *retval; |
2414 | |
|
2415 | 0 | g_return_val_if_fail (hash_table != NULL, NULL); |
2416 | | |
2417 | 0 | retval = NULL; |
2418 | 0 | for (i = 0; i < hash_table->size; i++) |
2419 | 0 | { |
2420 | 0 | if (HASH_IS_REAL (hash_table->hashes[i])) |
2421 | 0 | retval = g_list_prepend (retval, g_hash_table_fetch_key_or_value (hash_table->values, i, hash_table->have_big_values)); |
2422 | 0 | } |
2423 | |
|
2424 | 0 | return retval; |
2425 | 0 | } |
2426 | | |
2427 | | /** |
2428 | | * g_hash_table_get_values_as_ptr_array: (skip) |
2429 | | * @hash_table: a #GHashTable |
2430 | | * |
2431 | | * Retrieves every value inside @hash_table, as a #GPtrArray. |
2432 | | * The returned data is valid until changes to the hash release those values. |
2433 | | * |
2434 | | * This iterates over every entry in the hash table to build its return value. |
2435 | | * To iterate over the entries in a #GHashTable more efficiently, use a |
2436 | | * #GHashTableIter. |
2437 | | * |
2438 | | * You should always unref the returned array with g_ptr_array_unref(). |
2439 | | * |
2440 | | * Returns: (transfer container): a #GPtrArray containing each value from |
2441 | | * the table. Unref with with g_ptr_array_unref() when done. |
2442 | | * |
2443 | | * Since: 2.76 |
2444 | | **/ |
2445 | | GPtrArray * |
2446 | | g_hash_table_get_values_as_ptr_array (GHashTable *hash_table) |
2447 | 0 | { |
2448 | 0 | GPtrArray *array; |
2449 | |
|
2450 | 0 | g_return_val_if_fail (hash_table != NULL, NULL); |
2451 | | |
2452 | 0 | array = g_ptr_array_sized_new (hash_table->size); |
2453 | 0 | for (gsize i = 0; i < hash_table->size; ++i) |
2454 | 0 | { |
2455 | 0 | if (HASH_IS_REAL (hash_table->hashes[i])) |
2456 | 0 | { |
2457 | 0 | g_ptr_array_add (array, g_hash_table_fetch_key_or_value ( |
2458 | 0 | hash_table->values, i, hash_table->have_big_values)); |
2459 | 0 | } |
2460 | 0 | } |
2461 | 0 | g_assert (array->len == hash_table->nnodes); |
2462 | | |
2463 | 0 | return array; |
2464 | 0 | } |
2465 | | |
2466 | | /* Hash functions. |
2467 | | */ |
2468 | | |
2469 | | /** |
2470 | | * g_str_equal: |
2471 | | * @v1: (not nullable): a key |
2472 | | * @v2: (not nullable): a key to compare with @v1 |
2473 | | * |
2474 | | * Compares two strings for byte-by-byte equality and returns %TRUE |
2475 | | * if they are equal. It can be passed to g_hash_table_new() as the |
2476 | | * @key_equal_func parameter, when using non-%NULL strings as keys in a |
2477 | | * #GHashTable. |
2478 | | * |
2479 | | * This function is typically used for hash table comparisons, but can be used |
2480 | | * for general purpose comparisons of non-%NULL strings. For a %NULL-safe string |
2481 | | * comparison function, see g_strcmp0(). |
2482 | | * |
2483 | | * Returns: %TRUE if the two keys match |
2484 | | */ |
2485 | | gboolean |
2486 | | (g_str_equal) (gconstpointer v1, |
2487 | | gconstpointer v2) |
2488 | 37.3k | { |
2489 | 37.3k | const gchar *string1 = v1; |
2490 | 37.3k | const gchar *string2 = v2; |
2491 | | |
2492 | 37.3k | return strcmp (string1, string2) == 0; |
2493 | 37.3k | } |
2494 | | |
2495 | | /** |
2496 | | * g_str_hash: |
2497 | | * @v: (not nullable): a string key |
2498 | | * |
2499 | | * Converts a string to a hash value. |
2500 | | * |
2501 | | * This function implements the widely used "djb" hash apparently |
2502 | | * posted by Daniel Bernstein to comp.lang.c some time ago. The 32 |
2503 | | * bit unsigned hash value starts at 5381 and for each byte 'c' in |
2504 | | * the string, is updated: `hash = hash * 33 + c`. This function |
2505 | | * uses the signed value of each byte. |
2506 | | * |
2507 | | * It can be passed to g_hash_table_new() as the @hash_func parameter, |
2508 | | * when using non-%NULL strings as keys in a #GHashTable. |
2509 | | * |
2510 | | * Note that this function may not be a perfect fit for all use cases. |
2511 | | * For example, it produces some hash collisions with strings as short |
2512 | | * as 2. |
2513 | | * |
2514 | | * Returns: a hash value corresponding to the key |
2515 | | */ |
2516 | | guint |
2517 | | g_str_hash (gconstpointer v) |
2518 | 90.6k | { |
2519 | 90.6k | const signed char *p; |
2520 | 90.6k | guint32 h = 5381; |
2521 | | |
2522 | 521M | for (p = v; *p != '\0'; p++) |
2523 | 521M | h = (h << 5) + h + *p; |
2524 | | |
2525 | 90.6k | return h; |
2526 | 90.6k | } |
2527 | | |
2528 | | /** |
2529 | | * g_direct_hash: |
2530 | | * @v: (nullable): a #gpointer key |
2531 | | * |
2532 | | * Converts a gpointer to a hash value. |
2533 | | * It can be passed to g_hash_table_new() as the @hash_func parameter, |
2534 | | * when using opaque pointers compared by pointer value as keys in a |
2535 | | * #GHashTable. |
2536 | | * |
2537 | | * This hash function is also appropriate for keys that are integers |
2538 | | * stored in pointers, such as `GINT_TO_POINTER (n)`. |
2539 | | * |
2540 | | * Returns: a hash value corresponding to the key. |
2541 | | */ |
2542 | | guint |
2543 | | g_direct_hash (gconstpointer v) |
2544 | 66.8k | { |
2545 | 66.8k | return GPOINTER_TO_UINT (v); |
2546 | 66.8k | } |
2547 | | |
2548 | | /** |
2549 | | * g_direct_equal: |
2550 | | * @v1: (nullable): a key |
2551 | | * @v2: (nullable): a key to compare with @v1 |
2552 | | * |
2553 | | * Compares two #gpointer arguments and returns %TRUE if they are equal. |
2554 | | * It can be passed to g_hash_table_new() as the @key_equal_func |
2555 | | * parameter, when using opaque pointers compared by pointer value as |
2556 | | * keys in a #GHashTable. |
2557 | | * |
2558 | | * This equality function is also appropriate for keys that are integers |
2559 | | * stored in pointers, such as `GINT_TO_POINTER (n)`. |
2560 | | * |
2561 | | * Returns: %TRUE if the two keys match. |
2562 | | */ |
2563 | | gboolean |
2564 | | g_direct_equal (gconstpointer v1, |
2565 | | gconstpointer v2) |
2566 | 0 | { |
2567 | 0 | return v1 == v2; |
2568 | 0 | } |
2569 | | |
2570 | | /** |
2571 | | * g_int_equal: |
2572 | | * @v1: (not nullable): a pointer to a #gint key |
2573 | | * @v2: (not nullable): a pointer to a #gint key to compare with @v1 |
2574 | | * |
2575 | | * Compares the two #gint values being pointed to and returns |
2576 | | * %TRUE if they are equal. |
2577 | | * It can be passed to g_hash_table_new() as the @key_equal_func |
2578 | | * parameter, when using non-%NULL pointers to integers as keys in a |
2579 | | * #GHashTable. |
2580 | | * |
2581 | | * Note that this function acts on pointers to #gint, not on #gint |
2582 | | * directly: if your hash table's keys are of the form |
2583 | | * `GINT_TO_POINTER (n)`, use g_direct_equal() instead. |
2584 | | * |
2585 | | * Returns: %TRUE if the two keys match. |
2586 | | */ |
2587 | | gboolean |
2588 | | g_int_equal (gconstpointer v1, |
2589 | | gconstpointer v2) |
2590 | 0 | { |
2591 | 0 | return *((const gint*) v1) == *((const gint*) v2); |
2592 | 0 | } |
2593 | | |
2594 | | /** |
2595 | | * g_int_hash: |
2596 | | * @v: (not nullable): a pointer to a #gint key |
2597 | | * |
2598 | | * Converts a pointer to a #gint to a hash value. |
2599 | | * It can be passed to g_hash_table_new() as the @hash_func parameter, |
2600 | | * when using non-%NULL pointers to integer values as keys in a #GHashTable. |
2601 | | * |
2602 | | * Note that this function acts on pointers to #gint, not on #gint |
2603 | | * directly: if your hash table's keys are of the form |
2604 | | * `GINT_TO_POINTER (n)`, use g_direct_hash() instead. |
2605 | | * |
2606 | | * Returns: a hash value corresponding to the key. |
2607 | | */ |
2608 | | guint |
2609 | | g_int_hash (gconstpointer v) |
2610 | 0 | { |
2611 | 0 | return *(const gint*) v; |
2612 | 0 | } |
2613 | | |
2614 | | /** |
2615 | | * g_int64_equal: |
2616 | | * @v1: (not nullable): a pointer to a #gint64 key |
2617 | | * @v2: (not nullable): a pointer to a #gint64 key to compare with @v1 |
2618 | | * |
2619 | | * Compares the two #gint64 values being pointed to and returns |
2620 | | * %TRUE if they are equal. |
2621 | | * It can be passed to g_hash_table_new() as the @key_equal_func |
2622 | | * parameter, when using non-%NULL pointers to 64-bit integers as keys in a |
2623 | | * #GHashTable. |
2624 | | * |
2625 | | * Returns: %TRUE if the two keys match. |
2626 | | * |
2627 | | * Since: 2.22 |
2628 | | */ |
2629 | | gboolean |
2630 | | g_int64_equal (gconstpointer v1, |
2631 | | gconstpointer v2) |
2632 | 0 | { |
2633 | 0 | return *((const gint64*) v1) == *((const gint64*) v2); |
2634 | 0 | } |
2635 | | |
2636 | | /** |
2637 | | * g_int64_hash: |
2638 | | * @v: (not nullable): a pointer to a #gint64 key |
2639 | | * |
2640 | | * Converts a pointer to a #gint64 to a hash value. |
2641 | | * |
2642 | | * It can be passed to g_hash_table_new() as the @hash_func parameter, |
2643 | | * when using non-%NULL pointers to 64-bit integer values as keys in a |
2644 | | * #GHashTable. |
2645 | | * |
2646 | | * Returns: a hash value corresponding to the key. |
2647 | | * |
2648 | | * Since: 2.22 |
2649 | | */ |
2650 | | guint |
2651 | | g_int64_hash (gconstpointer v) |
2652 | 0 | { |
2653 | 0 | const guint64 *bits = v; |
2654 | |
|
2655 | 0 | return (guint) ((*bits >> 32) ^ (*bits & 0xffffffffU)); |
2656 | 0 | } |
2657 | | |
2658 | | /** |
2659 | | * g_double_equal: |
2660 | | * @v1: (not nullable): a pointer to a #gdouble key |
2661 | | * @v2: (not nullable): a pointer to a #gdouble key to compare with @v1 |
2662 | | * |
2663 | | * Compares the two #gdouble values being pointed to and returns |
2664 | | * %TRUE if they are equal. |
2665 | | * It can be passed to g_hash_table_new() as the @key_equal_func |
2666 | | * parameter, when using non-%NULL pointers to doubles as keys in a |
2667 | | * #GHashTable. |
2668 | | * |
2669 | | * Returns: %TRUE if the two keys match. |
2670 | | * |
2671 | | * Since: 2.22 |
2672 | | */ |
2673 | | gboolean |
2674 | | g_double_equal (gconstpointer v1, |
2675 | | gconstpointer v2) |
2676 | 0 | { |
2677 | 0 | return *((const gdouble*) v1) == *((const gdouble*) v2); |
2678 | 0 | } |
2679 | | |
2680 | | /** |
2681 | | * g_double_hash: |
2682 | | * @v: (not nullable): a pointer to a #gdouble key |
2683 | | * |
2684 | | * Converts a pointer to a #gdouble to a hash value. |
2685 | | * It can be passed to g_hash_table_new() as the @hash_func parameter, |
2686 | | * It can be passed to g_hash_table_new() as the @hash_func parameter, |
2687 | | * when using non-%NULL pointers to doubles as keys in a #GHashTable. |
2688 | | * |
2689 | | * Returns: a hash value corresponding to the key. |
2690 | | * |
2691 | | * Since: 2.22 |
2692 | | */ |
2693 | | guint |
2694 | | g_double_hash (gconstpointer v) |
2695 | 0 | { |
2696 | | /* Same as g_int64_hash() */ |
2697 | 0 | const guint64 *bits = v; |
2698 | |
|
2699 | 0 | return (guint) ((*bits >> 32) ^ (*bits & 0xffffffffU)); |
2700 | 0 | } |