/*******************************************************************************
*
* Project: seft (search engine for text)
*
* File: hash.c
*
* Author: Owen de Kretser ([email protected])
*
* Organisation: Dept. of CS&SE, University of Melbourne
*
* Date: April 1999
*
* Purpose: hash table functions
*
*******************************************************************************/
/***** #includes **************************************************************/
#include <string.h>
#include "hash.h"
#include "util.h"
#include "local_strings.h"
/*******************************************************************************
*
* Function: hash_init
*
* Purpose: dynamically allocate memory for hash table (of size HASH_SIZE)
*
*******************************************************************************/
void
hash_init(HASH* hash_table)
{
int i;
*hash_table = (LLIST**)safe_malloc(HASH_SIZE*sizeof(LLIST*));
for (i=0; i < HASH_SIZE; i++)
{
(*hash_table)[i] = NULL;
}
}
/*******************************************************************************
*
* Function: hash
*
* Purpose: calculate the hash key corresponding to a given subject code
*
*******************************************************************************/
unsigned int
hash(u_char* Word)
{
int h=0;
int i;
int length = *Word;
for (i=1; i <= length; i++)
{
h = ((h<<6) + Word[i]) % HASH_SIZE;
}
return (h);
}
/*******************************************************************************
*
* Function: hash_insert
*
* Purpose: insert data item into hash table (uses separate chaining)
*
*******************************************************************************/
void
hash_insert(HASH* hash_table, u_char* Word, int* no_terms)
{
int key = hash(Word);
list_insert(&((*hash_table)[key]), Word, no_terms);
}
/*******************************************************************************
*
* Function: list_insert
*
* Purpose: insert data item into sorted linked list
*
*******************************************************************************/
void
list_insert(LLIST** hash_entry, u_char* Word, int* no_terms)
{
LLIST* new_node;
LLIST* curr = *hash_entry;
LLIST* prev = curr;
new_node = (LLIST*)safe_malloc(sizeof(LLIST));
(*no_terms)++;
new_node->term.Word = copy_string(Word);
new_node->term.fqt = 1;
new_node->term.ft = 0;
new_node->term.term_num = *no_terms;
new_node->next = NULL;
if (*hash_entry == NULL) /* first entry */
{
*hash_entry = new_node;
}
else /* insert */
{
while ((curr) && (compare(curr->term.Word,Word) > 0))
{
prev = curr;
curr = curr->next;
}
if (curr == NULL)
{
prev->next = new_node;
}
else if (compare(curr->term.Word,Word) == 0)
{
curr->term.fqt++;
(*no_terms)--;
free(new_node->term.Word);
free(new_node);
}
else if (*hash_entry == curr)
{
new_node->next = curr;
*hash_entry = new_node;
}
else
{
prev->next = new_node;
new_node->next = curr;
}
}
}
/*******************************************************************************
*
* Function: hash_lookup
*
* Purpose: lookup up given data item in hash table
*
*******************************************************************************/
LLIST*
hash_lookup(LLIST** hash_table, u_char* Word)
{
int key;
LLIST* curr;
int cmp;
key = hash(Word);
curr = hash_table[key];
while ((curr != NULL) && ((cmp=compare(curr->term.Word,Word)) > 0))
{
curr = curr->next;
}
if (curr == NULL)
{
return (NULL);
}
else if (cmp == 0)
{
return (curr);
}
else
{
return (NULL);
}
}
|