/* 从文件中读取数据，存入数组中
* 文件中的数据是按照每行一个数字存放的
*/
int main(int argc, char *argv[]){
setenv("MALLOC_TRACE", "memoryTraceResult.txt", 1);//trace_result是保存检测结果的文件
mtrace();
if(argc < 2){
printf("Usage : %s <filename>\n", argv[0]);
exit(EXIT_SUCCESS);
}
char *filename = argv[1];
FILE *fp;
fp = fopen_file(filename, "r");
int data_length = cal_data_length(fp);
int *arr = malloc_space(data_length*sizeof(int));
assign_arr(fp, arr);
fclose(fp);
bucket_sort(arr, data_length);
int i;
for(i = 0; i < data_length; i++){
printf("%d, ", arr[i]);
}
printf("\n");
free(arr);
muntrace();
return 0;
}
int init_lda(Lda *lda) {
FILE *fp = NULL;
if (NULL == (fp = fopen(lda->p.in_file, "r"))) {
fprintf(stderr, "can not open file \"%s\"\n", lda->p.in_file);
return -1;
}
char buffer[LDA_LINE_LEN];
char **str_array = NULL;
int count = 0, token_size = 0;
IdMap *uidmap = idmap_create();
IdMap *vidmap = idmap_create();
while (NULL != fgets(buffer, LDA_LINE_LEN, fp)) {
str_array = split(trim(buffer, 3), '\t', &count);
if (count < 2) {
goto free_str;
}
if (-1 == idmap_get_value(uidmap, str_array[0])) {
idmap_add(uidmap, dupstr(str_array[0]), idmap_size(uidmap));
}
if (-1 == idmap_get_value(vidmap, str_array[1])) {
idmap_add(vidmap, dupstr(str_array[1]), idmap_size(vidmap));
}
token_size += 1;
free_str:
free(str_array[0]);
free(str_array);
}
lda->d = idmap_size(uidmap);
lda->t = token_size;
lda->v = idmap_size(vidmap);
malloc_space(lda);
rewind(fp);
int uid = -1, vid = -1, tid = -1;
int token_index = 0;
while (NULL != fgets(buffer, LDA_LINE_LEN, fp)) {
str_array = split(trim(buffer, 3), '\t', &count);
if (count < 2) {
goto str_free;
}
uid = idmap_get_value(uidmap, str_array[0]);
strncpy(lda->id_doc_map[uid], str_array[0], KEY_SIZE - 1);
lda->tokens[token_index][0] = uid;
vid = idmap_get_value(vidmap, str_array[1]);
strncpy(lda->id_v_map[vid], str_array[1], KEY_SIZE - 1);
lda->tokens[token_index][1] = vid;
tid = (int) ((1.0 + rand()) / (1.0 + RAND_MAX) * (lda->p.k));
if (count == 3){
tid = atoi(str_array[2]);
}
lda->tokens[token_index][2] = tid;
token_index += 1;
str_free:
free(str_array[0]);
free(str_array);
}
fclose(fp);
idmap_free(uidmap); uidmap = NULL;
idmap_free(vidmap); vidmap = NULL;
return 0;
}
int init_lda(Lda *lda) {
FILE *fp = NULL;
if (NULL == (fp = fopen(lda->p.in_file, "r"))) {
fprintf(stderr, "can not open file \"%s\"\n", lda->p.in_file);
return -1;
}
char buffer[LDA_LINE_LEN] = {'\0'};
char *string = NULL, *token = NULL;
int token_size = 0;
Hash * uhs = hash_create(1<<20, STRING);
Hash * vhs = hash_create(1<<20, STRING);
// first scan to generate uniq doc & word set
while (NULL != fgets(buffer, LDA_LINE_LEN, fp)) {
string = trim(buffer, 3);
token = strsep(&string, "\t");
hash_add(uhs, token);
token = strsep(&string, "\t");
hash_add(vhs, token);
token_size += 1;
}
lda->d = hash_cnt(uhs);
lda->v = hash_cnt(vhs);
lda->t = token_size;
// malloc the model space for iteration
malloc_space(lda);
// re scan the input file to load data
rewind(fp);
int uid = -1, vid = -1, tid = -1;
int token_index = 0;
while (NULL != fgets(buffer, LDA_LINE_LEN, fp)) {
string = trim(buffer, 3);
// read doc
token = strsep(&string, "\t");
uid = hash_find(uhs, token);
if (lda->id_doc_map[uid][0] == '\0') {
strncpy(lda->id_doc_map[uid], token, KEY_SIZE - 1);
}
lda->tokens[token_index][0] = uid;
// read word
token = strsep(&string, "\t");
vid = hash_find(vhs, token);
if (lda->id_v_map[vid][0] == '\0') {
strncpy(lda->id_v_map[vid], token, KEY_SIZE - 1);
}
lda->tokens[token_index][1] = vid;
// read tid
tid = (int) ((0.1 + rand()) / (0.1 + RAND_MAX) * (lda->p.k));
tid += 1;
token = strsep(&string, "\t");
if (token) {
tid = atoi(token);
}
lda->tokens[token_index][2] = tid;
// link the tokens for current doc
lda->tokens[token_index][3] = lda->doc_entry[uid];
lda->doc_entry[uid] = token_index;
// token_index ++
token_index += 1;
}
fclose(fp);
hash_free(uhs);
uhs = NULL;
hash_free(vhs);
vhs = NULL;
return 0;
}
