rename dir
This commit is contained in:
ranwenjie
@@ -0,0 +1,111 @@
|
||||
|
||||
#include "stdio.h"
|
||||
|
||||
|
||||
|
||||
typedef char * Key;
|
||||
typedef int Value;
|
||||
typedef unsigned int Hash;
|
||||
|
||||
|
||||
|
||||
typedef struct Entry
|
||||
{
|
||||
struct Entry *next;
|
||||
Hash hash; // key 对应的hash值
|
||||
Key key;
|
||||
Value value;
|
||||
}Entry;
|
||||
|
||||
|
||||
// 拉链法实现，也就是链表的数组.也是 数组和链表优势的结合
|
||||
typedef struct HashTable{
|
||||
|
||||
Entry **head;
|
||||
|
||||
unsigned int size;
|
||||
unsigned int usage;
|
||||
|
||||
}HashTable;
|
||||
|
||||
|
||||
HashTable *create(unsigned int);
|
||||
HashTable *put(HashTable *,Key,Value);
|
||||
HashTable *putInHeads(HashTable *,Key,Value); // 数组
|
||||
HashTable *putInLists(HashTable *,Key,Value); // 链表
|
||||
Value find(HashTable *,Key );
|
||||
HashTable*delete(HashTable *,Key );
|
||||
void print(HashTable *);
|
||||
|
||||
Hash hashCode(Key); // 哈希函数
|
||||
|
||||
|
||||
|
||||
HashTable *create(unsigned int size){
|
||||
|
||||
HashTable *hashTable = malloc(sizeof(HashTable));
|
||||
if(hashTable==NULL){
|
||||
printf("malloc error\n");
|
||||
return NULL;
|
||||
}
|
||||
|
||||
hashTable->size=size;
|
||||
hashTable->usage = 0;
|
||||
hashTable->head = calloc(size,sizeof(Entry *));
|
||||
|
||||
return hashTable;
|
||||
}
|
||||
|
||||
|
||||
HashTable *put(HashTable *hashTable,Key key,Value value){
|
||||
|
||||
if (key==NULL)
|
||||
{
|
||||
return hashTable;
|
||||
}
|
||||
|
||||
|
||||
|
||||
}
|
||||
|
||||
Value find(HashTable *hashTable,Key key){
|
||||
|
||||
|
||||
}
|
||||
|
||||
void print(HashTable *hashTable){
|
||||
|
||||
Entry *entry;
|
||||
for (int i = 0; i < hashTable->size; ++i)
|
||||
{
|
||||
entry = hashTable->head[i];
|
||||
while(entry!=NULL){
|
||||
printf("%s=%d\n",entry->key,entry->value);
|
||||
entry=entry->next;
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
|
||||
// ?????
|
||||
Hash hashCode(char *key){
|
||||
|
||||
int offset = 5;
|
||||
Hash hashCode = 0;
|
||||
while(*key){
|
||||
hashCode = (hashCode << offset) + (*key++);
|
||||
}
|
||||
|
||||
return hashCode;
|
||||
|
||||
}
|
||||
|
||||
|
||||
int main(){
|
||||
|
||||
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
@@ -0,0 +1,223 @@
|
||||
|
||||
|
||||
## 散列表
|
||||
|
||||
本节围绕以下内容展开：
|
||||
|
||||
* 散列表
|
||||
* 散列函数设计
|
||||
* 冲突处理
|
||||
* hashmap数据结构
|
||||
* Java 中HashMap实现
|
||||
|
||||
|
||||
散列表使用某种算法操作(散列函数)将键转化为数组的索引来访问数组中的数据，这样可以通过Key-value的方式来访问数据，达到常数级别的存取效率。现在的nosql数据库都是采用key-value的方式来访问存储数据。
|
||||
|
||||
散列表是算法在时间和空间上做出权衡的经典例子。通过一个散列函数，将键值key映射到记录的访问地址，达到快速查找的目的。如果没有内存限制，我们可以直接将键作为数组的索引，所有的操作操作只需要一次访问内存就可以完成。但这种情况不太现实。
|
||||
|
||||
|
||||
### 散列函数
|
||||
|
||||
散列函数就是将键转化为数组索引的过程。且这个函数应该易于计算且能够均与分布所有的键。
|
||||
|
||||
散列函数最常用的方法是`除留余数法`。这时候被除数应该选用`素数`，这样才能保证键值的均匀散步。
|
||||
|
||||
散列函数和键的类型有关，每种数据类型都需要相应的散列函数；比如键的类型是整数，那我们可以直接使用`除留余数法`；这里特别说明下，大多数情况下，键的类型都是字符串，这个时候我们任然可以使用`除留余数法`，将字符串当做一个特别大的整数。
|
||||
|
||||
```
|
||||
int hash = 0;
|
||||
for (int i=0;i<s.length();i++){
|
||||
hash = (R*hash +s.charAt(i)%M);
|
||||
}
|
||||
```
|
||||
|
||||
还有，比如下面的：
|
||||
|
||||
```
|
||||
Hash hashCode(char *key){
|
||||
int offset = 5;
|
||||
Hash hashCode = 0;
|
||||
while(*key){
|
||||
hashCode = (hashCode << offset) + *key++;
|
||||
}
|
||||
return hashCode;
|
||||
}
|
||||
```
|
||||
|
||||
使用时 `hashCode(key) & (size-1) ` 就可以得到一个 `size-1` 范围内的hash值
|
||||
|
||||
|
||||
当然，还有其他的散列函数，如`平方取中法`, `随机数法`等。
|
||||
|
||||
|
||||
### 碰撞解决
|
||||
|
||||
不同的关键字得到同一个散列地址` f(key1)=f(key2) `，即为碰撞 。这是我们需要尽量避免的情况。常见的处理方法有：
|
||||
|
||||
1. 拉链法
|
||||
2. 线性探测法
|
||||
|
||||
|
||||
#### 拉链法
|
||||
|
||||
将大小为M的数组中的每个元素指向一条链表，链表中的每个节点都存储了散列值为该元素索引的键值对。每条链表的平均长度是N/M，N是键值对的总个数。如下图：
|
||||
|
||||
|
||||
|
||||
添加操作：
|
||||
|
||||
1. 通过hash函数得到hashCode
|
||||
2. 通过hashcode得到index
|
||||
3. 如果index处没有链表，建立好新结点，作为新链表的首结点
|
||||
4. 如果index处已经有链表，先要遍历看key是否已经存在，如果存在直接返回，如果不存在，加入链表头部
|
||||
|
||||
|
||||
删除操作：
|
||||
|
||||
1. 通过hash函数得到hashCode
|
||||
2. 通过hashcode得到index
|
||||
3. 遍历链表，删除结点
|
||||
|
||||
|
||||
#### 线性探测法
|
||||
|
||||
使用大小为M的数组保存N个键值对，当碰撞发生时，直接检查散列表中的下一个位置。
|
||||
|
||||
|
||||
### 数据结构和算法
|
||||
|
||||
这里给出拉链法构造的hashmap的算法，表示如下：
|
||||
|
||||
```
|
||||
typedef char * Key;
|
||||
typedef int value;
|
||||
typedef unsigned int Hash;
|
||||
|
||||
/*每个节点表示*/
|
||||
typedef struct Entry{
|
||||
Hash hash;
|
||||
Key key;
|
||||
Value value;
|
||||
Entry *next;
|
||||
}Entry;
|
||||
|
||||
typedef struct HashMap{
|
||||
Entry **heads;
|
||||
unsigned int size; /* 数组大小*/
|
||||
unsigned int usage;/* 键值对的个数*/
|
||||
}HashMap;
|
||||
|
||||
Hash hashCode(Key);
|
||||
HashMap *create(unsigned int size);
|
||||
HashMap *put(HashMap *,Key ,Value);
|
||||
int get(HashMap *,Key);
|
||||
|
||||
HashMap *_putInHead(HashMap *,int index,Key,Value);
|
||||
HashMap *_putInList(HashMap *,int index,Key,Value);
|
||||
|
||||
Hash hashCode(char *key){
|
||||
int offset = 5;
|
||||
Hash hashCode = 0;
|
||||
while(*key){
|
||||
hashCode = (hashCode << offset) + *key++;
|
||||
}
|
||||
return hashCode;
|
||||
}
|
||||
|
||||
HashMap *create(unsigned int size){
|
||||
HashMap *hashMap = malloc(sizeof(HashMap));
|
||||
hashMap->size = size;
|
||||
hashMap->usage = 0;
|
||||
hashMap->heads = calloc(size,sizeof(Entry *));
|
||||
|
||||
return hashMap;
|
||||
}
|
||||
|
||||
HashMap *put(HashMap *hashMap,Key key,Value value){
|
||||
if (key == NULL){
|
||||
return hashMap;
|
||||
}
|
||||
Hash hash = hashCode(key);
|
||||
int index = hash & (size-1);/* */
|
||||
if (hashMap->heads[index] == NULL){
|
||||
_putInHead(hashMap,index,key,value);
|
||||
}else{
|
||||
_putInList(hashMap,index,key,value);
|
||||
}
|
||||
}
|
||||
|
||||
Value get(HashMap hashMap*,Key key){
|
||||
if (key == NULL){
|
||||
return hashMap;
|
||||
}
|
||||
Hash hash = hashCode(key);
|
||||
int index = hash & (size-1);/* */
|
||||
|
||||
Entry *entry = hashMap->heads[index];
|
||||
while(entry != NULL){
|
||||
if (entry->hash == hash){
|
||||
return entry->value;
|
||||
}
|
||||
entry = entry->next;
|
||||
}
|
||||
return NULL;
|
||||
}
|
||||
|
||||
HashMap *_putInHead(HashMap *hashMap,int index,Key key,Value value){
|
||||
Entry *newHead = malloc(sizeof(Entry));
|
||||
newHead->hash = hash;
|
||||
newHead->key = key;
|
||||
newHead->value = value;
|
||||
|
||||
hashMap->heads[index] = newHead;
|
||||
(hashMap->usage)++;
|
||||
return hashMap;
|
||||
}
|
||||
|
||||
HashMap *_putInList(HashMap *hashMap,int index,Key key,Value value){
|
||||
Entry *lastEntry = hashMap->heads[index];
|
||||
while(lastEntry != NULL){
|
||||
if (lastEntry->hash == hash){
|
||||
return hashMap;
|
||||
}else{
|
||||
lastEntry = lastEntry->next;
|
||||
}
|
||||
}
|
||||
|
||||
lastEntry = malloc(sizeof(Entry));
|
||||
lastEntry->hash = hash;
|
||||
lastEntry->key = key;
|
||||
lastEntry->value = value;
|
||||
lastEntry->next = hashMap->heads[index];
|
||||
|
||||
hashMap->heads[index] = lastEntry;
|
||||
(hashMap->usage)++;
|
||||
return hashMap;
|
||||
}
|
||||
|
||||
```
|
||||
|
||||
|
||||
### Hashmap应用
|
||||
|
||||
1. cocos2d 游戏引擎 CCScheduler
|
||||
2. linux 内核bcache。 缓存加速技术，使用SSD固态硬盘作为高速缓存，提高慢速存储设备HDD机械硬盘的性能
|
||||
3. hash表在海量数据处理中有广泛应用。如海量日志中，提取出某日访问百度次数最多的IP
|
||||
|
||||
|
||||
## Java 中HashMap实现
|
||||
|
||||
编程语言中HashMap是如何实现的呢? 说说 Java ， Golang
|
||||
|
||||
|
||||
|
||||
## 参考
|
||||
|
||||
《Algorithms》
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
@@ -0,0 +1,186 @@
|
||||
#include "stdio.h"
|
||||
#include "stdlib.h"
|
||||
#include "string.h"
|
||||
#include "time.h"
|
||||
|
||||
#define DEFAULT_MAP_SIZE 1<<5
|
||||
#define MAX_MAP_SIZE 1<<30
|
||||
#define CAPACITY_FACTOR 0.6f
|
||||
#define MAX_KEY_LENGTH 1<<3
|
||||
|
||||
typedef unsigned int Hash;
|
||||
typedef char* Key;
|
||||
typedef int Value;
|
||||
typedef struct Entry{
|
||||
struct Entry *next;
|
||||
Hash hash;
|
||||
Value value;
|
||||
Key key;
|
||||
}Entry;
|
||||
|
||||
typedef struct HashMap{
|
||||
Entry** heads;
|
||||
Value nul;
|
||||
unsigned int size;
|
||||
unsigned int usage;
|
||||
unsigned int realCapacity; //size * CAPACITY_FACTOR
|
||||
unsigned int sizeForIndex; //size - 1
|
||||
}HashMap;
|
||||
|
||||
HashMap* create(unsigned int);
|
||||
HashMap* transfer(HashMap*, HashMap*);
|
||||
Hash hash(char[]);
|
||||
HashMap* put(HashMap*, Key, Value);
|
||||
HashMap* putInHeads(HashMap*, int, Hash, Key, Value);
|
||||
HashMap* putInList(HashMap*, int, Hash, Key, Value);
|
||||
HashMap* resize(HashMap*);
|
||||
int find(char[]);
|
||||
void print(HashMap*);
|
||||
|
||||
HashMap* create(unsigned int size){
|
||||
HashMap* hashMap = malloc(sizeof(HashMap));
|
||||
int i;
|
||||
hashMap->size = size;
|
||||
hashMap->sizeForIndex = size - 1;
|
||||
hashMap->realCapacity = (unsigned int)(size * CAPACITY_FACTOR);
|
||||
hashMap->usage = 0;
|
||||
hashMap->heads = calloc(size, sizeof(Entry*));
|
||||
hashMap->nul = 0;
|
||||
return hashMap;
|
||||
}
|
||||
|
||||
HashMap* transfer(HashMap* newHashMap, HashMap* oldHashMap){
|
||||
unsigned int oldIndex, newIndex;
|
||||
Entry *oldEntry, *oldNextEntry;
|
||||
for (oldIndex = 0; oldIndex < oldHashMap->size; oldIndex++){
|
||||
if (oldHashMap->heads[oldIndex] != NULL){
|
||||
oldEntry = oldHashMap->heads[oldIndex];
|
||||
do{
|
||||
newIndex = (oldEntry->hash) & (newHashMap->sizeForIndex);
|
||||
oldNextEntry = oldEntry->next;
|
||||
oldEntry->next = newHashMap->heads[newIndex];
|
||||
newHashMap->heads[newIndex] = oldEntry;
|
||||
oldEntry = oldNextEntry;
|
||||
} while (oldEntry != NULL);
|
||||
}
|
||||
}
|
||||
newHashMap->usage = oldHashMap->usage;
|
||||
newHashMap->nul = oldHashMap->nul;
|
||||
free(oldHashMap->heads);
|
||||
oldHashMap->heads = NULL;
|
||||
free(oldHashMap);
|
||||
oldHashMap = NULL;
|
||||
return newHashMap;
|
||||
}
|
||||
|
||||
Hash hashCode(char* key){
|
||||
int offset = 5;
|
||||
Hash hashCode = 0;
|
||||
while (*key) hashCode = (hashCode << offset) + (*key++);
|
||||
return hashCode;
|
||||
}
|
||||
|
||||
HashMap* put(HashMap* hashMap, Key key, Value value){
|
||||
if (key == NULL){
|
||||
hashMap->nul = value;
|
||||
return hashMap;
|
||||
}
|
||||
Hash hash = hashCode(key);
|
||||
int index = hash & (hashMap->sizeForIndex);
|
||||
|
||||
if (hashMap->heads[index] == NULL){
|
||||
return putInHeads(hashMap, index, hash, key, value);
|
||||
} else {
|
||||
return putInList(hashMap, index, hash, key, value);
|
||||
}
|
||||
}
|
||||
|
||||
HashMap* putInHeads(HashMap* hashMap, int index, Hash hash, Key key, Value value){
|
||||
Entry* newHead = malloc(sizeof(Entry));
|
||||
newHead->hash = hash;
|
||||
newHead->key = _strdup(key);
|
||||
newHead->value = value;
|
||||
newHead->next = NULL;
|
||||
hashMap->heads[index] = newHead;
|
||||
(hashMap->usage)++;
|
||||
if (hashMap->usage > hashMap->realCapacity) return resize(hashMap);
|
||||
return hashMap;
|
||||
}
|
||||
HashMap* putInList(HashMap* hashMap, int index, Hash hash, Key key, Value value){
|
||||
Entry* lastEntry = hashMap->heads[index];
|
||||
while (lastEntry != NULL){
|
||||
if (lastEntry->hash == hash){
|
||||
lastEntry->value += value;
|
||||
return hashMap;
|
||||
}
|
||||
else lastEntry = lastEntry->next;
|
||||
}
|
||||
lastEntry = malloc(sizeof(Entry));
|
||||
lastEntry->hash = hash;
|
||||
lastEntry->key = _strdup(key);
|
||||
lastEntry->value = value;
|
||||
lastEntry->next = hashMap->heads[index];
|
||||
hashMap->heads[index] = lastEntry;
|
||||
(hashMap->usage)++;
|
||||
if (hashMap->usage > hashMap->realCapacity) return resize(hashMap);
|
||||
return hashMap;
|
||||
}
|
||||
|
||||
HashMap* resize(HashMap* hashMap){
|
||||
HashMap* newHashMap;
|
||||
if ((hashMap->size << 1) <= MAX_MAP_SIZE) {
|
||||
newHashMap = create((hashMap->size << 1));
|
||||
newHashMap = transfer(newHashMap, hashMap);
|
||||
return newHashMap;
|
||||
}
|
||||
return hashMap;
|
||||
}
|
||||
|
||||
Value get(HashMap* hashMap, Key key){
|
||||
Hash hash = hashCode(key);
|
||||
int index = hash & (hashMap->sizeForIndex);
|
||||
Entry* entry = hashMap->heads[index];
|
||||
while (entry != NULL){
|
||||
if (entry->hash == hash) return entry->value;
|
||||
entry = entry->next;
|
||||
}
|
||||
return NULL;
|
||||
}
|
||||
|
||||
void print(HashMap* hashMap){
|
||||
unsigned int i;
|
||||
Entry* entry;
|
||||
for (i = 0; i < hashMap->size; i++){
|
||||
entry = hashMap->heads[i];
|
||||
while (entry != NULL){
|
||||
printf("%s: %d\n", entry->key, entry->value);
|
||||
entry = entry->next;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
int main(){
|
||||
clock_t start, end;
|
||||
HashMap* hashMap = create(DEFAULT_MAP_SIZE);
|
||||
FILE* fp;
|
||||
char filePath[0xFF];
|
||||
char line[5];
|
||||
char key[MAX_KEY_LENGTH];
|
||||
|
||||
printf("input file path: ");
|
||||
scanf_s("%s", filePath, 0xFF);
|
||||
if (fopen_s(&fp, filePath, "r") == 0){
|
||||
start = clock();
|
||||
while (fgets(line, 5, fp) != NULL){
|
||||
line[3] = '\0';
|
||||
hashMap = put(hashMap, line, 1);
|
||||
}
|
||||
end = clock();
|
||||
printf("\n%fs", (float)(end - start)/1000);
|
||||
fclose(fp);
|
||||
print(hashMap);
|
||||
} else {
|
||||
printf("can't open file");
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
Binary file not shown.
|
After Width: | Height: | Size: 202 KiB |
Reference in New Issue
Block a user