2 ******************************************************************************
3 * Copyright (C) 1997-2010, International Business Machines
4 * Corporation and others. All Rights Reserved.
5 ******************************************************************************
6 * Date Name Description
7 * 03/22/00 aliu Adapted from original C++ ICU Hashtable.
8 * 07/06/01 aliu Modified to support int32_t keys on
9 * platforms with sizeof(void*) < 32.
10 ******************************************************************************
16 #include "unicode/utypes.h"
19 * UHashtable stores key-value pairs and does moderately fast lookup
20 * based on keys. It provides a good tradeoff between access time and
21 * storage space. As elements are added to it, it grows to accomodate
22 * them. By default, the table never shrinks, even if all elements
23 * are removed from it.
25 * Keys and values are stored as void* pointers. These void* pointers
26 * may be actual pointers to strings, objects, or any other structure
27 * in memory, or they may simply be integral values cast to void*.
28 * UHashtable doesn't care and manipulates them via user-supplied
29 * functions. These functions hash keys, compare keys, delete keys,
30 * and delete values. Some function pointers are optional (may be
31 * NULL); others must be supplied. Several prebuilt functions exist
32 * to handle common key types.
34 * UHashtable ownership of keys and values is flexible, and controlled
35 * by whether or not the key deleter and value deleter functions are
36 * set. If a void* key is actually a pointer to a deletable object,
37 * then UHashtable can be made to delete that object by setting the
38 * key deleter function pointer to a non-NULL value. If this is done,
39 * then keys passed to uhash_put() are owned by the hashtable and will
40 * be deleted by it at some point, either as keys are replaced, or
41 * when uhash_close() is finally called. The same is true of values
42 * and the value deleter function pointer. Keys passed to methods
43 * other than uhash_put() are never owned by the hashtable.
45 * NULL values are not allowed. uhash_get() returns NULL to indicate
46 * a key that is not in the table, and having a NULL value in the
47 * table would generate an ambiguous result. If a key and a NULL
48 * value is passed to uhash_put(), this has the effect of doing a
49 * uhash_remove() on that key. This keeps uhash_get(), uhash_count(),
50 * and uhash_nextElement() consistent with one another.
52 * To see everything in a hashtable, use uhash_nextElement() to
53 * iterate through its contents. Each call to this function returns a
54 * UHashElement pointer. A hash element contains a key, value, and
55 * hashcode. During iteration an element may be deleted by calling
56 * uhash_removeElement(); iteration may safely continue thereafter.
57 * The uhash_remove() function may also be safely called in
58 * mid-iteration. However, if uhash_put() is called during iteration
59 * then the iteration will be out of sync. Under no circumstances
60 * should the UHashElement returned by uhash_nextElement be modified
63 * By default, the hashtable grows when necessary, but never shrinks,
64 * even if all items are removed. For most applications this is
65 * optimal. However, in a highly dynamic usage where memory is at a
66 * premium, the table can be set to both grow and shrink by calling
67 * uhash_setResizePolicy() with the policy U_GROW_AND_SHRINK. In a
68 * situation where memory is critical and the client wants a table
69 * that does not grow at all, the constant U_FIXED can be used.
72 /********************************************************************
74 ********************************************************************/
79 * A key or value within the hashtable. It may be either a 32-bit
80 * integral value or an opaque void* pointer. The void* pointer may
81 * be smaller than 32 bits (e.g. 24 bits) or may be larger (e.g. 64
82 * bits). The hashing and comparison functions take a pointer to a
83 * UHashTok, but the deleter receives the void* pointer within it.
85 * Because a UHashTok is the size of a native pointer or a 32-bit
86 * integer, we pass it around by value.
92 typedef union UHashTok UHashTok
;
95 * This is a single hash element.
98 /* Reorder these elements to pack nicely if necessary */
103 typedef struct UHashElement UHashElement
;
106 * A hashing function.
107 * @param key A key stored in a hashtable
108 * @return A NON-NEGATIVE hash code for parm.
110 typedef int32_t U_CALLCONV
UHashFunction(const UHashTok key
);
113 * A key comparison function.
114 * @param key1 A key stored in a hashtable
115 * @param key2 A key stored in a hashtable
116 * @return TRUE if the two keys are equal.
118 typedef UBool U_CALLCONV
UKeyComparator(const UHashTok key1
,
119 const UHashTok key2
);
121 * A key comparison function.
122 * @param val1 A key stored in a hashtable
123 * @param val2 A key stored in a hashtable
124 * @return TRUE if the two keys are equal.
126 typedef UBool U_CALLCONV
UValueComparator(const UHashTok val1
,
127 const UHashTok val2
);
129 * A function called by <TT>uhash_remove</TT>,
130 * <TT>uhash_close</TT>, or <TT>uhash_put</TT> to delete
131 * an existing key or value.
132 * @param obj A key or value stored in a hashtable
133 * @see uhash_deleteUObject
135 typedef void U_CALLCONV
UObjectDeleter(void* obj
);
138 * This specifies whether or not, and how, the hastable resizes itself.
139 * See uhash_setResizePolicy().
141 enum UHashResizePolicy
{
142 U_GROW
, /* Grow on demand, do not shrink */
143 U_GROW_AND_SHRINK
, /* Grow and shrink on demand */
144 U_FIXED
/* Never change size */
148 * The UHashtable struct. Clients should treat this as an opaque data
149 * type and manipulate it only through the uhash_... API.
153 /* Main key-value pair storage array */
155 UHashElement
*elements
;
157 /* Function pointers */
159 UHashFunction
*keyHasher
; /* Computes hash from key.
161 UKeyComparator
*keyComparator
; /* Compares keys for equality.
163 UValueComparator
*valueComparator
; /* Compares the values for equality */
165 UObjectDeleter
*keyDeleter
; /* Deletes keys when required.
166 * If NULL won't do anything */
167 UObjectDeleter
*valueDeleter
; /* Deletes values when required.
168 * If NULL won't do anything */
170 /* Size parameters */
172 int32_t count
; /* The number of key-value pairs in this table.
173 * 0 <= count <= length. In practice we
174 * never let count == length (see code). */
175 int32_t length
; /* The physical size of the arrays hashes, keys
176 * and values. Must be prime. */
178 /* Rehashing thresholds */
180 int32_t highWaterMark
; /* If count > highWaterMark, rehash */
181 int32_t lowWaterMark
; /* If count < lowWaterMark, rehash */
182 float highWaterRatio
; /* 0..1; high water as a fraction of length */
183 float lowWaterRatio
; /* 0..1; low water as a fraction of length */
185 int8_t primeIndex
; /* Index into our prime table for length.
186 * length == PRIMES[primeIndex] */
187 UBool allocated
; /* Was this UHashtable allocated? */
189 typedef struct UHashtable UHashtable
;
193 /********************************************************************
195 ********************************************************************/
198 * Initialize a new UHashtable.
199 * @param keyHash A pointer to the key hashing function. Must not be
201 * @param keyComp A pointer to the function that compares keys. Must
203 * @param status A pointer to an UErrorCode to receive any errors.
204 * @return A pointer to a UHashtable, or 0 if an error occurred.
205 * @see uhash_openSize
207 U_CAPI UHashtable
* U_EXPORT2
208 uhash_open(UHashFunction
*keyHash
,
209 UKeyComparator
*keyComp
,
210 UValueComparator
*valueComp
,
214 * Initialize a new UHashtable with a given initial size.
215 * @param keyHash A pointer to the key hashing function. Must not be
217 * @param keyComp A pointer to the function that compares keys. Must
219 * @param size The initial capacity of this hash table.
220 * @param status A pointer to an UErrorCode to receive any errors.
221 * @return A pointer to a UHashtable, or 0 if an error occurred.
224 U_CAPI UHashtable
* U_EXPORT2
225 uhash_openSize(UHashFunction
*keyHash
,
226 UKeyComparator
*keyComp
,
227 UValueComparator
*valueComp
,
232 * Initialize an existing UHashtable.
233 * @param keyHash A pointer to the key hashing function. Must not be
235 * @param keyComp A pointer to the function that compares keys. Must
237 * @param status A pointer to an UErrorCode to receive any errors.
238 * @return A pointer to a UHashtable, or 0 if an error occurred.
239 * @see uhash_openSize
241 U_CAPI UHashtable
* U_EXPORT2
242 uhash_init(UHashtable
*hash
,
243 UHashFunction
*keyHash
,
244 UKeyComparator
*keyComp
,
245 UValueComparator
*valueComp
,
249 * Close a UHashtable, releasing the memory used.
250 * @param hash The UHashtable to close. If hash is NULL no operation is performed.
252 U_CAPI
void U_EXPORT2
253 uhash_close(UHashtable
*hash
);
258 * Set the function used to hash keys.
259 * @param hash The UHashtable to set
260 * @param fn the function to be used hash keys; must not be NULL
261 * @return the previous key hasher; non-NULL
263 U_CAPI UHashFunction
*U_EXPORT2
264 uhash_setKeyHasher(UHashtable
*hash
, UHashFunction
*fn
);
267 * Set the function used to compare keys. The default comparison is a
268 * void* pointer comparison.
269 * @param hash The UHashtable to set
270 * @param fn the function to be used compare keys; must not be NULL
271 * @return the previous key comparator; non-NULL
273 U_CAPI UKeyComparator
*U_EXPORT2
274 uhash_setKeyComparator(UHashtable
*hash
, UKeyComparator
*fn
);
277 * Set the function used to compare values. The default comparison is a
278 * void* pointer comparison.
279 * @param hash The UHashtable to set
280 * @param fn the function to be used compare keys; must not be NULL
281 * @return the previous key comparator; non-NULL
283 U_CAPI UValueComparator
*U_EXPORT2
284 uhash_setValueComparator(UHashtable
*hash
, UValueComparator
*fn
);
287 * Set the function used to delete keys. If this function pointer is
288 * NULL, this hashtable does not delete keys. If it is non-NULL, this
289 * hashtable does delete keys. This function should be set once
290 * before any elements are added to the hashtable and should not be
291 * changed thereafter.
292 * @param hash The UHashtable to set
293 * @param fn the function to be used delete keys, or NULL
294 * @return the previous key deleter; may be NULL
296 U_CAPI UObjectDeleter
*U_EXPORT2
297 uhash_setKeyDeleter(UHashtable
*hash
, UObjectDeleter
*fn
);
300 * Set the function used to delete values. If this function pointer
301 * is NULL, this hashtable does not delete values. If it is non-NULL,
302 * this hashtable does delete values. This function should be set
303 * once before any elements are added to the hashtable and should not
304 * be changed thereafter.
305 * @param hash The UHashtable to set
306 * @param fn the function to be used delete values, or NULL
307 * @return the previous value deleter; may be NULL
309 U_CAPI UObjectDeleter
*U_EXPORT2
310 uhash_setValueDeleter(UHashtable
*hash
, UObjectDeleter
*fn
);
313 * Specify whether or not, and how, the hastable resizes itself.
314 * By default, tables grow but do not shrink (policy U_GROW).
315 * See enum UHashResizePolicy.
316 * @param hash The UHashtable to set
317 * @param policy The way the hashtable resizes itself, {U_GROW, U_GROW_AND_SHRINK, U_FIXED}
319 U_CAPI
void U_EXPORT2
320 uhash_setResizePolicy(UHashtable
*hash
, enum UHashResizePolicy policy
);
323 * Get the number of key-value pairs stored in a UHashtable.
324 * @param hash The UHashtable to query.
325 * @return The number of key-value pairs stored in hash.
327 U_CAPI
int32_t U_EXPORT2
328 uhash_count(const UHashtable
*hash
);
331 * Put a (key=pointer, value=pointer) item in a UHashtable. If the
332 * keyDeleter is non-NULL, then the hashtable owns 'key' after this
333 * call. If the valueDeleter is non-NULL, then the hashtable owns
334 * 'value' after this call. Storing a NULL value is the same as
335 * calling uhash_remove().
336 * @param hash The target UHashtable.
337 * @param key The key to store.
338 * @param value The value to store, may be NULL (see above).
339 * @param status A pointer to an UErrorCode to receive any errors.
340 * @return The previous value, or NULL if none.
343 U_CAPI
void* U_EXPORT2
344 uhash_put(UHashtable
*hash
,
350 * Put a (key=integer, value=pointer) item in a UHashtable.
351 * keyDeleter must be NULL. If the valueDeleter is non-NULL, then the
352 * hashtable owns 'value' after this call. Storing a NULL value is
353 * the same as calling uhash_remove().
354 * @param hash The target UHashtable.
355 * @param key The integer key to store.
356 * @param value The value to store, may be NULL (see above).
357 * @param status A pointer to an UErrorCode to receive any errors.
358 * @return The previous value, or NULL if none.
361 U_CAPI
void* U_EXPORT2
362 uhash_iput(UHashtable
*hash
,
368 * Put a (key=pointer, value=integer) item in a UHashtable. If the
369 * keyDeleter is non-NULL, then the hashtable owns 'key' after this
370 * call. valueDeleter must be NULL. Storing a 0 value is the same as
371 * calling uhash_remove().
372 * @param hash The target UHashtable.
373 * @param key The key to store.
374 * @param value The integer value to store.
375 * @param status A pointer to an UErrorCode to receive any errors.
376 * @return The previous value, or 0 if none.
379 U_CAPI
int32_t U_EXPORT2
380 uhash_puti(UHashtable
*hash
,
386 * Put a (key=integer, value=integer) item in a UHashtable. If the
387 * keyDeleter is non-NULL, then the hashtable owns 'key' after this
388 * call. valueDeleter must be NULL. Storing a 0 value is the same as
389 * calling uhash_remove().
390 * @param hash The target UHashtable.
391 * @param key The key to store.
392 * @param value The integer value to store.
393 * @param status A pointer to an UErrorCode to receive any errors.
394 * @return The previous value, or 0 if none.
397 U_CAPI
int32_t U_EXPORT2
398 uhash_iputi(UHashtable
*hash
,
404 * Retrieve a pointer value from a UHashtable using a pointer key,
405 * as previously stored by uhash_put().
406 * @param hash The target UHashtable.
407 * @param key A pointer key stored in a hashtable
408 * @return The requested item, or NULL if not found.
410 U_CAPI
void* U_EXPORT2
411 uhash_get(const UHashtable
*hash
,
415 * Retrieve a pointer value from a UHashtable using a integer key,
416 * as previously stored by uhash_iput().
417 * @param hash The target UHashtable.
418 * @param key An integer key stored in a hashtable
419 * @return The requested item, or NULL if not found.
421 U_CAPI
void* U_EXPORT2
422 uhash_iget(const UHashtable
*hash
,
426 * Retrieve an integer value from a UHashtable using a pointer key,
427 * as previously stored by uhash_puti().
428 * @param hash The target UHashtable.
429 * @param key A pointer key stored in a hashtable
430 * @return The requested item, or 0 if not found.
432 U_CAPI
int32_t U_EXPORT2
433 uhash_geti(const UHashtable
*hash
,
436 * Retrieve an integer value from a UHashtable using an integer key,
437 * as previously stored by uhash_iputi().
438 * @param hash The target UHashtable.
439 * @param key An integer key stored in a hashtable
440 * @return The requested item, or 0 if not found.
442 U_CAPI
int32_t U_EXPORT2
443 uhash_igeti(const UHashtable
*hash
,
447 * Remove an item from a UHashtable stored by uhash_put().
448 * @param hash The target UHashtable.
449 * @param key A key stored in a hashtable
450 * @return The item removed, or NULL if not found.
452 U_CAPI
void* U_EXPORT2
453 uhash_remove(UHashtable
*hash
,
457 * Remove an item from a UHashtable stored by uhash_iput().
458 * @param hash The target UHashtable.
459 * @param key An integer key stored in a hashtable
460 * @return The item removed, or NULL if not found.
462 U_CAPI
void* U_EXPORT2
463 uhash_iremove(UHashtable
*hash
,
467 * Remove an item from a UHashtable stored by uhash_puti().
468 * @param hash The target UHashtable.
469 * @param key An key stored in a hashtable
470 * @return The item removed, or 0 if not found.
472 U_CAPI
int32_t U_EXPORT2
473 uhash_removei(UHashtable
*hash
,
477 * Remove an item from a UHashtable stored by uhash_iputi().
478 * @param hash The target UHashtable.
479 * @param key An integer key stored in a hashtable
480 * @return The item removed, or 0 if not found.
482 U_CAPI
int32_t U_EXPORT2
483 uhash_iremovei(UHashtable
*hash
,
487 * Remove all items from a UHashtable.
488 * @param hash The target UHashtable.
490 U_CAPI
void U_EXPORT2
491 uhash_removeAll(UHashtable
*hash
);
494 * Locate an element of a UHashtable. The caller must not modify the
495 * returned object. The primary use of this function is to obtain the
496 * stored key when it may not be identical to the search key. For
497 * example, if the compare function is a case-insensitive string
498 * compare, then the hash key may be desired in order to obtain the
499 * canonical case corresponding to a search key.
500 * @param hash The target UHashtable.
501 * @param key A key stored in a hashtable
502 * @return a hash element, or NULL if the key is not found.
504 U_CAPI
const UHashElement
* U_EXPORT2
505 uhash_find(const UHashtable
*hash
, const void* key
);
508 * Iterate through the elements of a UHashtable. The caller must not
509 * modify the returned object. However, uhash_removeElement() may be
510 * called during iteration to remove an element from the table.
511 * Iteration may safely be resumed afterwards. If uhash_put() is
512 * called during iteration the iteration will then be out of sync and
513 * should be restarted.
514 * @param hash The target UHashtable.
515 * @param pos This should be set to -1 initially, and left untouched
517 * @return a hash element, or NULL if no further key-value pairs
518 * exist in the table.
520 U_CAPI
const UHashElement
* U_EXPORT2
521 uhash_nextElement(const UHashtable
*hash
,
525 * Remove an element, returned by uhash_nextElement(), from the table.
526 * Iteration may be safely continued afterwards.
527 * @param hash The hashtable
528 * @param e The element, returned by uhash_nextElement(), to remove.
529 * Must not be NULL. Must not be an empty or deleted element (as long
530 * as this was returned by uhash_nextElement() it will not be empty or
531 * deleted). Note: Although this parameter is const, it will be
533 * @return the value that was removed.
535 U_CAPI
void* U_EXPORT2
536 uhash_removeElement(UHashtable
*hash
, const UHashElement
* e
);
538 /********************************************************************
539 * UHashTok convenience
540 ********************************************************************/
543 * Return a UHashTok for an integer.
544 * @param i The given integer
545 * @return a UHashTok for an integer.
547 /*U_CAPI UHashTok U_EXPORT2
548 uhash_toki(int32_t i);*/
551 * Return a UHashTok for a pointer.
552 * @param p The given pointer
553 * @return a UHashTok for a pointer.
555 /*U_CAPI UHashTok U_EXPORT2
556 uhash_tokp(void* p);*/
558 /********************************************************************
559 * UChar* and char* Support Functions
560 ********************************************************************/
563 * Generate a hash code for a null-terminated UChar* string. If the
564 * string is not null-terminated do not use this function. Use
565 * together with uhash_compareUChars.
566 * @param key The string (const UChar*) to hash.
567 * @return A hash code for the key.
569 U_CAPI
int32_t U_EXPORT2
570 uhash_hashUChars(const UHashTok key
);
573 * Generate a hash code for a null-terminated char* string. If the
574 * string is not null-terminated do not use this function. Use
575 * together with uhash_compareChars.
576 * @param key The string (const char*) to hash.
577 * @return A hash code for the key.
579 U_CAPI
int32_t U_EXPORT2
580 uhash_hashChars(const UHashTok key
);
582 /* Used by UnicodeString to compute its hashcode - Not public API. */
583 U_CAPI
int32_t U_EXPORT2
584 uhash_hashUCharsN(const UChar
*key
, int32_t length
);
587 * Generate a case-insensitive hash code for a null-terminated char*
588 * string. If the string is not null-terminated do not use this
589 * function. Use together with uhash_compareIChars.
590 * @param key The string (const char*) to hash.
591 * @return A hash code for the key.
593 U_CAPI
int32_t U_EXPORT2
594 uhash_hashIChars(const UHashTok key
);
597 * Comparator for null-terminated UChar* strings. Use together with
599 * @param key1 The string for comparison
600 * @param key2 The string for comparison
601 * @return true if key1 and key2 are equal, return false otherwise.
603 U_CAPI UBool U_EXPORT2
604 uhash_compareUChars(const UHashTok key1
, const UHashTok key2
);
607 * Comparator for null-terminated char* strings. Use together with
609 * @param key1 The string for comparison
610 * @param key2 The string for comparison
611 * @return true if key1 and key2 are equal, return false otherwise.
613 U_CAPI UBool U_EXPORT2
614 uhash_compareChars(const UHashTok key1
, const UHashTok key2
);
617 * Case-insensitive comparator for null-terminated char* strings. Use
618 * together with uhash_hashIChars.
619 * @param key1 The string for comparison
620 * @param key2 The string for comparison
621 * @return true if key1 and key2 are equal, return false otherwise.
623 U_CAPI UBool U_EXPORT2
624 uhash_compareIChars(const UHashTok key1
, const UHashTok key2
);
626 /********************************************************************
627 * UnicodeString Support Functions
628 ********************************************************************/
631 * Hash function for UnicodeString* keys.
632 * @param key The string (const char*) to hash.
633 * @return A hash code for the key.
635 U_CAPI
int32_t U_EXPORT2
636 uhash_hashUnicodeString(const UHashTok key
);
639 * Hash function for UnicodeString* keys (case insensitive).
640 * Make sure to use together with uhash_compareCaselessUnicodeString.
641 * @param key The string (const char*) to hash.
642 * @return A hash code for the key.
644 U_CAPI
int32_t U_EXPORT2
645 uhash_hashCaselessUnicodeString(const UHashTok key
);
648 * Comparator function for UnicodeString* keys.
649 * @param key1 The string for comparison
650 * @param key2 The string for comparison
651 * @return true if key1 and key2 are equal, return false otherwise.
653 U_CAPI UBool U_EXPORT2
654 uhash_compareUnicodeString(const UHashTok key1
, const UHashTok key2
);
657 * Comparator function for UnicodeString* keys (case insensitive).
658 * Make sure to use together with uhash_hashCaselessUnicodeString.
659 * @param key1 The string for comparison
660 * @param key2 The string for comparison
661 * @return true if key1 and key2 are equal, return false otherwise.
663 U_CAPI UBool U_EXPORT2
664 uhash_compareCaselessUnicodeString(const UHashTok key1
, const UHashTok key2
);
667 * Deleter function for UnicodeString* keys or values.
668 * @param obj The object to be deleted
670 U_CAPI
void U_EXPORT2
671 uhash_deleteUnicodeString(void *obj
);
673 /********************************************************************
674 * int32_t Support Functions
675 ********************************************************************/
678 * Hash function for 32-bit integer keys.
679 * @param key The string (const char*) to hash.
680 * @return A hash code for the key.
682 U_CAPI
int32_t U_EXPORT2
683 uhash_hashLong(const UHashTok key
);
686 * Comparator function for 32-bit integer keys.
687 * @param key1 The integer for comparison
688 * @param Key2 The integer for comparison
689 * @return true if key1 and key2 are equal, return false otherwise
691 U_CAPI UBool U_EXPORT2
692 uhash_compareLong(const UHashTok key1
, const UHashTok key2
);
694 /********************************************************************
695 * Other Support Functions
696 ********************************************************************/
699 * Deleter for Hashtable objects.
700 * @param obj The object to be deleted
702 U_CAPI
void U_EXPORT2
703 uhash_deleteHashtable(void *obj
);
706 * Deleter for UObject instances.
707 * @param obj The object to be deleted
709 U_CAPI
void U_EXPORT2
710 uhash_deleteUObject(void *obj
);
713 * Deleter for any key or value allocated using uprv_malloc. Calls
715 * @param obj The object to be deleted
717 U_CAPI
void U_EXPORT2
718 uhash_freeBlock(void *obj
);
721 * Checks if the given hash tables are equal or not.
724 * @return true if the hashtables are equal and false if not.
726 U_CAPI UBool U_EXPORT2
727 uhash_equals(const UHashtable
* hash1
, const UHashtable
* hash2
);