void text_split(text_t * self){
char * part = strtok(self->text,".;?");
while (part!= NULL){
list_sentence_add(self->sentences,sentence_new(part));
part = strtok(NULL,".;");
}
}
void text_divide(text_t * self){
char * part = strtok(self->text,".?!");
while (part != NULL) {
list_push_back(self->sentencesList,sentence_new(part));
part = strtok(NULL,".?!");
}
}
static int read_conllx(FILE* fp)
{
int num_sent = 0;
unsigned int line_num = 0;
char buff[BUF_SIZE];
const char* sep = "\t";
char *t, *line, *tofree;
size_t len;
const char** seq = xmalloc(sizeof(char *) * CONLLX_TOKEN_NUM_FIELDS);
struct blocks* blocks = new_blocks(256);
struct sentence* sent = sentence_new();
while (fgets(buff, sizeof(buff), fp)) {
line_num++;
if (buff[0] == '\n') {
latex_print_dep_tree(sent);
sentence_destroy(sent);
sent = sentence_new();
num_sent++;
continue;
}
len = strnlen(buff, BUF_SIZE);
tofree = line = strndup(buff, len-1);
blocks_add_elem(blocks, tofree);
int i = 0;
while ((t = strsep(&line, sep)) != NULL) {
seq[i] = t;
i++;
}
sentence_add_token(sent, token_new(seq, CONLLX_TOKEN_NUM_FIELDS));
}
destroy_blocks(blocks);
sentence_destroy(sent);
free(seq);
fprintf(stderr, "INFO: Number of sentences = %d\n", num_sent);
return 0;
}
status_t
parse_article(const char* file_name, lang_t* lang, article_t* article)
{
string_t word, word_core, word_stem;
sentence_t* sentence;
word_t* word_entry;
stream_t* stream = &article->stream;
bool_t is_new, is_para_end = SMRZR_FALSE;
PROF_START;
if(SMRZR_OK != stream_create(file_name, stream))
ERROR_RET;
while(!STREAM_END(stream)) {
STREAM_FIND_WORD(stream);
if(STREAM_END(stream)) break;
sentence = sentence_new(&article->sentences, article->stream.curr);
assert(NULL != sentence);
if(SMRZR_TRUE == is_para_end) {
sentence->is_para_begin = SMRZR_TRUE;
is_para_end = SMRZR_FALSE;
}
while(!STREAM_END(stream)) {
STREAM_GET_WORD(stream, word, is_para_end);
sentence->num_words++;
if(NULL == (word_core = get_word_core(&article->stack, lang, word)))
ERROR_RET;
if(NULL == array_search(lang->exclude, word_core, comp_strings)) {
if(NULL == (word_stem = get_word_stem(&article->stack, lang,
word_core, SMRZR_TRUE)))
ERROR_RET;
if(NULL == (word_entry = array_search_or_alloc(&article->words,
word_stem, comp_word_by_stem, &is_new)))
ERROR_RET;
if(SMRZR_TRUE == is_new) {
word_entry->num_occ = 1;
word_entry->stem = word_stem;
} else {
++(word_entry->num_occ);
array_pop_free(article->stack, word_stem);
}
} else {
array_pop_free(article->stack, word_core);
}
if(end_of_line(lang, word)) {
sentence->end = word + strlen(word);
article->num_words += sentence->num_words;
break;
}
}
}
PROF_END("article parsing");
/*fprintf(stdout, "Number of sentences - %lu\n", ARR_SZ(article->sentences));
fprintf(stdout, "Number of words - %lu\n", ARR_SZ(article->words));*/
return(SMRZR_OK);
}
