/* creates a new counting bloom filter from a given scaling bloom filter, with count and id */
counting_bloom_t *new_counting_bloom_from_scale(scaling_bloom_t *bloom)
{
int i;
long offset;
double error_rate;
counting_bloom_t *cur_bloom;
error_rate = bloom->error_rate * (pow(ERROR_TIGHTENING_RATIO, bloom->num_blooms + 1));
if ((bloom->blooms = (counting_bloom_t **)realloc(bloom->blooms, (bloom->num_blooms + 1) * sizeof(counting_bloom_t *))) == NULL) {
fprintf(stderr, "Error, could not realloc a new bloom filter\n");
return NULL;
}
cur_bloom = counting_bloom_init(bloom->capacity, error_rate, bloom->num_bytes);
bloom->blooms[bloom->num_blooms] = cur_bloom;
bloom->bitmap = bitmap_resize(bloom->bitmap, bloom->num_bytes, bloom->num_bytes + cur_bloom->num_bytes);
/* reset header pointer, as mmap may have moved */
bloom->header = (scaling_bloom_header_t *) bloom->bitmap->array;
/* Set the pointers for these header structs to the right location since mmap may have moved */
bloom->num_blooms++;
for (i = 0; i < bloom->num_blooms; i++) {
offset = bloom->blooms[i]->offset - sizeof(counting_bloom_header_t);
bloom->blooms[i]->header = (counting_bloom_header_t *) (bloom->bitmap->array + offset);
}
bloom->num_bytes += cur_bloom->num_bytes;
cur_bloom->bitmap = bloom->bitmap;
return cur_bloom;
}
counting_bloom_t *new_counting_bloom(unsigned int capacity, double error_rate)
{
counting_bloom_t *cur_bloom;
cur_bloom = counting_bloom_init(capacity, error_rate, 0);
cur_bloom->bitmap = new_bitmap(cur_bloom->num_bytes);
cur_bloom->header = (counting_bloom_header_t *)
(cur_bloom->bitmap->array);
return cur_bloom;
}
counting_bloom_t *new_counting_bloom(unsigned int capacity, double error_rate, const char *filename)
{
counting_bloom_t *cur_bloom;
int fd;
if ((fd = open(filename, O_RDWR | O_CREAT | O_TRUNC, (mode_t)0600)) < 0) {
perror("Error, Opening File Failed");
fprintf(stderr, " %s \n", filename);
return NULL;
}
cur_bloom = counting_bloom_init(capacity, error_rate, 0);
cur_bloom->bitmap = new_bitmap(fd, cur_bloom->num_bytes);
cur_bloom->header = (counting_bloom_header_t *)(cur_bloom->bitmap->array);
return cur_bloom;
}
counting_bloom_t *counting_bloom_from_file(unsigned capacity, double error_rate, const char *filename)
{
counting_bloom_t *cur_bloom;
int fd;
if ((fd = open(filename, O_RDWR, (mode_t)0600)) < 0) {
perror("Error, Opening File Failed");
fprintf(stderr, " %s \n", filename);
return NULL;
}
cur_bloom = counting_bloom_init(capacity, error_rate, 0);
cur_bloom->parent_bitmap = new_bitmap(fd, cur_bloom->num_bytes);
return cur_bloom;
}
int main(int argc, char **argv) {
printf("Testing Counting Bloom version %s\n", counting_bloom_get_version());
CountingBloom cb;
counting_bloom_init(&cb, 10, 0.01);
counting_bloom_add_string(&cb, "test");
counting_bloom_add_string(&cb, "out");
counting_bloom_add_string(&cb, "the");
counting_bloom_add_string(&cb, "counting");
counting_bloom_add_string(&cb, "bloom");
counting_bloom_add_string(&cb, "filter");
// we can add it multiple times!
counting_bloom_add_string(&cb, "test");
counting_bloom_add_string(&cb, "Test"); // should not be added to the 'test' strings
counting_bloom_add_string(&cb, "out");
counting_bloom_add_string(&cb, "test");
if (counting_bloom_check_string(&cb, "test") == COUNTING_BLOOM_SUCCESS) {
printf("'test' was found in the counting bloom with false positive rate of %f!\n", counting_bloom_current_false_positive_rate(&cb));
printf("'test' is in the counting bloom a maximum of %d times!\n", counting_bloom_get_max_insertions(&cb, "test"));
} else {
printf("'test' was not found in the counting bloom!\n");
}
printf("Export the Counting Bloom!\n");
counting_bloom_export(&cb, "./dist/test.cbm");
counting_bloom_stats(&cb);
counting_bloom_destroy(&cb);
printf("Exported and destroyed the original counting bloom!\n\n");
printf("Re-import the bloom filter!\n");
CountingBloom cb1;
counting_bloom_import(&cb1, "./dist/test.cbm");
if (counting_bloom_check_string(&cb1, "test") == COUNTING_BLOOM_SUCCESS) {
printf("'test' was found in the counting bloom with false positive rate of %f!\n", counting_bloom_current_false_positive_rate(&cb1));
printf("'test' is in the counting bloom a maximum of %d times!\n", counting_bloom_get_max_insertions(&cb1, "test"));
} else {
printf("'test' was not found in the counting bloom!\n");
}
if (counting_bloom_check_string(&cb1, "blah") == COUNTING_BLOOM_SUCCESS) {
printf("'blah' was found in the counting bloom with false positive rate of %f!\n", counting_bloom_current_false_positive_rate(&cb1));
printf("'blah' is in the counting bloom a maximum of %d times!\n", counting_bloom_get_max_insertions(&cb1, "test"));
} else {
printf("'blah' was not found in the counting bloom!\n");
}
counting_bloom_stats(&cb1);
counting_bloom_destroy(&cb1);
}
counting_bloom_t *new_counting_bloom_from_file(unsigned int capacity, double error_rate, const char *filename)
{
int fd;
off_t size;
counting_bloom_t *bloom;
if ((fd = open(filename, O_RDWR, (mode_t)0600)) < 0) {
fprintf(stderr, "Error, Could not open file %s: %s\n", filename, strerror(errno));
return NULL;
}
if ((size = lseek(fd, 0, SEEK_END)) < 0) {
perror("Error, calling lseek() to tell file size");
close(fd);
return NULL;
}
if (size == 0) {
fprintf(stderr, "Error, File size zero\n");
}
bloom = counting_bloom_init(capacity, error_rate, 0);
if (size != bloom->num_bytes) {
free_counting_bloom(bloom);
fprintf(stderr, "Error, Actual filesize and expected filesize are not equal\n");
return NULL;
}
if ((bloom->bitmap = new_bitmap(fd, size)) == NULL) {
fprintf(stderr, "Error, Could not create bitmap with file\n");
free_counting_bloom(bloom);
return NULL;
}
bloom->header = (counting_bloom_header_t *)(bloom->bitmap->array);
return bloom;
}
