File: hash.h

package info (click to toggle)
s3d 0.2.3-1
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid
  • size: 5,708 kB
  • sloc: ansic: 23,609; python: 488; perl: 98; makefile: 31; sh: 29
file content (80 lines) | stat: -rw-r--r-- 3,351 bytes parent folder | download | duplicates (4)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
/* SPDX-License-Identifier: LGPL-2.1-or-later */
/* SPDX-FileCopyrightText: 2004-2015  Simon Wunderlich <sw@simonwunderlich.de>
 * SPDX-FileCopyrightText: 2004-2015  Marek Lindner <mareklindner@neomailbox.ch>
 */

#ifndef _S3D_HASH_H
#define _S3D_HASH_H



typedef int (*hashdata_compare_cb)(const void *, const void *);
typedef int (*hashdata_choose_cb)(const void *, int);
typedef void (*hashdata_free_cb)(void *);

struct element_t {
	void *data;      /* pointer to the data */
	struct element_t *next;   /* overflow bucket pointer */
};

struct hash_it_t {
	int index;
	struct element_t *bucket;
	struct element_t *prev_bucket;
	struct element_t **first_bucket;
};

struct hashtable_t {
	struct element_t **table;     /* the hashtable itself, with the buckets */
	int elements;        /* number of elements registered */
	int size;         /* size of hashtable */
	hashdata_compare_cb compare;       /* callback to a compare function.
             * should compare 2 element datas for their keys,
             * return 0 if same and not 0 if not same */
	hashdata_choose_cb choose;     /* the hashfunction, should return an index based
             * on the key in the data of the first argument
             * and the size the second */
};

/* clears the hash */
void      _s3d_hash_init(struct hashtable_t *hash);

/* allocates and clears the hash */
struct hashtable_t *_s3d_hash_new(int size, hashdata_compare_cb compare, hashdata_choose_cb choose);

/* remove bucket (this might be used in hash_iterate() if you already found the bucket
 * you want to delete and don't need the overhead to find it again with hash_remove().
 * But usually, you don't want to use this function, as it fiddles with hash-internals. */
void     *_s3d_hash_remove_bucket(struct hashtable_t *hash, struct hash_it_t *hash_it_t);

/* remove the hash structure. if hashdata_free_cb != NULL,
 * this function will be called to remove the elements inside of the hash.
 * if you don't remove the elements, memory might be leaked. */
void      _s3d_hash_delete(struct hashtable_t *hash, hashdata_free_cb free_cb);

/* free only the hashtable and the hash itself. */
void      _s3d_hash_destroy(struct hashtable_t *hash);

/* adds data to the hashtable. returns 0 on success, -1 on error */
int      _s3d_hash_add(struct hashtable_t *hash, void *data);

/* removes data from hash, if found. returns pointer do data on success,
 * so you can remove the used structure yourself, or NULL on error .
 * data could be the structure you use with just the key filled,
 * we just need the key for comparing. */
void     *_s3d_hash_remove(struct hashtable_t *hash, void *data);

/* finds data, based on the key in keydata. returns the found data on success, or NULL on error */
void     *_s3d_hash_find(struct hashtable_t *hash, void *keydata);

/* resize the hash, returns the pointer to the new hash or NULL on error. removes the old hash on success */
struct hashtable_t *_s3d_hash_resize(struct hashtable_t *hash, int size);

/* print the hash table for debugging */
void      _s3d_hash_debug(struct hashtable_t *hash);

/* iterate though the hash. first element is selected with iter_in NULL.
 * use the returned iterator to access the elements until hash_it_t returns NULL. */
struct hash_it_t  *_s3d_hash_iterate(struct hashtable_t *hash, struct hash_it_t *iter_in);

#endif