C实现的hash table动态关联数组
2019-04-21 本文已影响0人
山中散人的博客
动态关联数组中的元素是一系列键值对,实现关联数组的数据结构有binary search tree和hash table。在大多数高级语言中,都有现成的关联数组, 如c++和golang中的std::map和map。这里在c中实现hash table和char* 到 char* 的关联数组。
- 添加文件hash_tbl.h和hash_tbl.c
//hash_tbl.h
#ifndef HASH_TBL_H
#define HASH_TBL_H
#define INIT_BUCKET_NUM 40
typedef struct hash_node{
char *key;
char *val;
struct hash_node *next;
}hash_node;
typedef struct hash_tbl{
int bucket_num;
int node_num;
hash_node **buckets;
}hash_tbl;
void init_hash_tbl(hash_tbl *tbl);
#endif
//hash_tbl.c
#include "hash_tbl.h"
#include "comm.h"
void init_hash_tbl(hash_tbl *tbl){
tbl->bucket_num = INIT_BUCKET_NUM;
tbl->node_num = 0;
tbl->buckets = calloc(tbl->bucket_num, sizeof(hash_node*));
if(!tbl->buckets)
err_exit(MEM_ERR);
}
- (1)哈希节点定义了哈希表中的元素
- (2)利用链表来解决哈希寻址冲突的问题,即将冲突的哈希节点挂在链表的尾部
- (3)为哈希表分配了40个桶,每一个桶对应一个哈希节点链表
2. 向哈希表中插入节点
//hash_tbl.c
void insert_hash_tbl(hash_tbl *tbl, const char *k,
const char *v){
hash_node *item = new_hash_node(k, v);
int ix = hash_func(k, HT_BASE, tbl->bucket_num);
if(!tbl->buckets[ix]){
tbl->buckets[ix] = item;
}else{
hash_node dummy = {NULL, NULL, NULL};
dummy.next = tbl->buckets[ix];
hash_node *prev = &dummy;
while(prev && prev->next){
//update node if key exists
if(prev->next->key &&
strcmp(prev->next->key, k) == 0){
item->next = prev->next->next;
del_hash_node(&(prev->next));
prev->next = item;
return;
}
prev = prev->next;
}
//link the new node
prev->next = item;
}
tbl->node_num++;
}
- (1)通过哈希函数计算得出桶的索引,新建哈希节点用以描述键值对
- (2)若桶对应的链表为空,用新哈希节点作为链表的头
- (3)若桶不为空,遍历链表,如存在重复键,替换哈希节点,否则节点添加到链表尾部
3. 在哈希表中搜索键,返回值
//vector.c
char *search_hash_tbl(hash_tbl *tbl, const char *k){
int ix = hash_func(k, HT_BASE, tbl->bucket_num);
hash_node *prev = tbl->buckets[ix];
while(prev){
if(prev->key && strcmp(prev->key, k) == 0){
return prev->key;
}
prev = prev->next;
}
return NULL;
}
- (1)搜索的过程与插入过程类型,若存在键,返回对应的值,若不存在,返回空指针
4. 从哈希表中删除键值对
//vector.c
void del_key_hash_tbl(hash_tbl *tbl, const char *k){
int ix = hash_func(k, HT_BASE, tbl->bucket_num);
hash_node dummy = {NULL, NULL, NULL};
dummy.next = tbl->buckets[ix];
hash_node *prev = &dummy;
while(prev && prev->next){
if(prev->next->key &&
strcmp(prev->next->key, k) == 0){
hash_node *next = prev->next->next;
del_hash_node(&(prev->next));
prev->next = next;
tbl->node_num--;
break;
}
prev = prev->next;
}
tbl->buckets[ix] = dummy.next;
}
- (1)删除键值对的过程类似于从链表中删除一个节点
5. 添加测试用例
//vector.c
void test_hash_tbl(){
hash_tbl tbl;
init_hash_tbl(&tbl);
char key[125], val[125];
for(char c = 'a'; c <= 'z'; c++){
sprintf(key, "%c", c);
sprintf(val, "%c", c + 'A' - 'a');
//printf("insert %s-%s\n", key, val);
insert_hash_tbl(&tbl, key, val);
}
for(char c = 'a'; c <= 'f'; c++){
sprintf(key, "%c", c);
if(search_hash_tbl(&tbl, key)){
sprintf(val, "%s", search_hash_tbl(&tbl, key));
printf("%s-%s\n", key, val);
}
del_key_hash_tbl(&tbl, key);
}
reset_hash_tbl(&tbl);
}
- (1)建立大小写字符串键值表
- (2)在哈希表中检索部分键,同时进行删除