/*
* This is an example program that reads words from a text-file (a
* book or something like that) and uses those as keys in a hash table
* (the actual data stored is unimportant). The words are case
* sensitive.
*
* After this, the program opens another text-file and tries to match
* the words in that with the words stored in the table.
*
* The meaning with this program is to test the speed of the table and
* to provide an example of its use.
*/
int main(int argc, char *argv[])
{
FILE *p_file;
char *p_dict_name;
char *p_text_name;
ght_hash_table_t *p_table;
ght_iterator_t iterator;
struct stat stat_struct;
int i_found;
int i_cnt;
char *p_buf;
char *p_tok;
const void *p_key;
void *p_e;
/* Create a new hash table */
if ( !(p_table = ght_create(1000)) )
{
return 1;
}
ght_set_rehash(p_table, TRUE);
/* Parse the arguments */
if (argc < 3)
{
printf("Usage: dict_example [-m|-t|-b] dictfile textfile\n\n"
"Reads words from `dictfile' and looks up these words in `textfile'.\n"
"Options:\n"
" -m Use move-to-front heuristics\n"
" -t Use transpose heuristics\n"
" -b Use bounded buckets (use the hash table as a cache)\n"
);
return 0;
}
else if (argc > 3)
{
if(strcmp(argv[1], "-m") == 0)
ght_set_heuristics(p_table, GHT_HEURISTICS_MOVE_TO_FRONT);
else if(strcmp(argv[1], "-t") == 0)
ght_set_heuristics(p_table, GHT_HEURISTICS_TRANSPOSE);
else if (strcmp(argv[1], "-b") == 0)
{
/* Rehashing makes no sense in "cache" mode */
ght_set_rehash(p_table, FALSE);
ght_set_bounded_buckets(p_table, 3, bucket_free_callback);
}
p_dict_name = argv[2];
p_text_name = argv[3];
}
else
{
/* 2 arguments */
p_dict_name = argv[1];
p_text_name = argv[2];
}
/* Open the dictionary file (first check its size) */
if (stat(p_dict_name, &stat_struct) < 0)
{
perror("stat");
return 1;
}
if (!(p_buf = (char*)malloc(stat_struct.st_size)))
{
perror("malloc");
return 1;
}
/* Open the dictionary file and read that into the buffer. */
if (! (p_file = fopen(p_dict_name, "r")) )
{
perror("fopen");
return 1;
}
fread(p_buf, sizeof(char), stat_struct.st_size, p_file);
fclose(p_file);
/* For each word in the dictionary file, insert it into the hash table. */
p_tok = strtok(p_buf, DELIMS);
i_cnt = 0;
i_found = 0;
while (p_tok)
{
int *p_data;
if ( !(p_data = (int*)malloc(sizeof(int))) )
{
perror("malloc");
return 1;
}
*p_data = i_cnt++;
/* Insert the word into the table */
if (ght_insert(p_table,
p_data,
strlen(p_tok), p_tok) < 0)
free(p_data); /* Could not insert the item (already exists), free it. */
else
i_found++;
p_tok = strtok(NULL, DELIMS);
}
printf("Done reading %d unique words from the wordlist.\n"
"Total number of words is %d.\n\n", i_found, i_cnt);
free(p_buf);
/* Check the size of the text file. */
if (stat(p_text_name, &stat_struct) < 0)
{
perror("stat");
return 1;
}
if (!(p_buf = (char*)malloc(stat_struct.st_size)))
{
perror("malloc");
return 1;
}
/* Open the text file and read that into the buffer. */
if (! (p_file = fopen(p_text_name, "r")) )
{
perror("fopen");
return 1;
}
fread(p_buf, sizeof(char), stat_struct.st_size, p_file);
fclose(p_file);
/* For each word in the text file, check if it exists in the hash
table. */
p_tok = strtok(p_buf, DELIMS);
i_cnt = 0;
i_found = 0;
while (p_tok)
{
printf(" searching %s ;", p_tok);
if (ght_get(p_table,
strlen(p_tok), p_tok))
{
i_found++;
}
i_cnt++;
p_tok = strtok(NULL, DELIMS);
}
free(p_buf);
printf("Found %d words out of %d words\n", i_found, i_cnt);
/* Free the entry data in the table */
for(p_e = ght_first(p_table, &iterator, &p_key); p_e; p_e = ght_next(p_table, &iterator, &p_key))
{
free(p_e);
}
/* Free the table */
ght_finalize(p_table);
return 0;
}
/* Rehash the hash table (i.e. change its size and reinsert all
* items). This operation is slow and should not be used frequently.
*/
void ght_rehash(ght_hash_table_t *p_ht, unsigned int i_size)
{
ght_hash_table_t *p_tmp;
ght_iterator_t iterator;
const void *p_key;
void *p;
unsigned int i;
DEBUG_ASSERT(p_ht);
/* Recreate the hash table with the new size */
p_tmp = ght_create(i_size);
DEBUG_ASSERT(p_tmp);
/* Set the flags for the new hash table */
ght_set_hash(p_tmp, p_ht->fn_hash);
ght_set_alloc(p_tmp, p_ht->fn_alloc, p_ht->fn_free);
ght_set_heuristics(p_tmp, GHT_HEURISTICS_NONE);
ght_set_rehash(p_tmp, FALSE);
/* Walk through all elements in the table and insert them into the temporary one. */
for (p = ght_first(p_ht, &iterator, &p_key); p; p = ght_next(p_ht, &iterator, &p_key))
{
DEBUG_ASSERT(iterator.p_entry);
/* Insert the entry into the new table */
if (ght_insert(p_tmp,
iterator.p_entry->p_data,
iterator.p_entry->key.i_size, iterator.p_entry->key.p_key) < 0)
{
LOG_ERROR("Out of memory error or entry already in hash table when rehashing (internal error)\n");
}
}
/* Remove the old table... */
for (i=0; i<p_ht->i_size; i++)
{
if (p_ht->pp_entries[i])
{
/* Delete the entries in the bucket */
free_entry_chain (p_ht, p_ht->pp_entries[i]);
p_ht->pp_entries[i] = NULL;
}
}
DefaultFreeFunction (p_ht->pp_entries);
DefaultFreeFunction (p_ht->p_nr);
/* ... and replace it with the new */
p_ht->i_size = p_tmp->i_size;
p_ht->i_size_mask = p_tmp->i_size_mask;
p_ht->i_items = p_tmp->i_items;
p_ht->pp_entries = p_tmp->pp_entries;
p_ht->p_nr = p_tmp->p_nr;
p_ht->p_oldest = p_tmp->p_oldest;
p_ht->p_newest = p_tmp->p_newest;
/* Clean up */
p_tmp->pp_entries = NULL;
p_tmp->p_nr = NULL;
DefaultFreeFunction (p_tmp);
}
