bool Dawg::match_words(WERD_CHOICE *word, inT32 index,
NODE_REF node, UNICHAR_ID wildcard) const {
EDGE_REF edge;
inT32 word_end;
if (wildcard != INVALID_UNICHAR_ID && word->unichar_id(index) == wildcard) {
bool any_matched = false;
NodeChildVector vec;
this->unichar_ids_of(node, &vec);
for (int i = 0; i < vec.size(); ++i) {
word->set_unichar_id(vec[i].unichar_id, index);
if (match_words(word, index, node, wildcard))
any_matched = true;
}
word->set_unichar_id(wildcard, index);
return any_matched;
} else {
word_end = index == word->length() - 1;
edge = edge_char_of(node, word->unichar_id(index), word_end);
if (edge != NO_EDGE) { // normal edge in DAWG
node = next_node(edge);
if (word_end) {
if (debug_level_ > 1) word->print("match_words() found: ");
return true;
} else if (node != 0) {
return match_words(word, index+1, node, wildcard);
}
}
}
return false;
}
/* Returns the basename of $fname if calls to it are to be reported. */
static char const *report_dso(char const *fname)
{
static int report_all = 0;
static struct word_st const *flist = NULL;
static int is_whitelist;
char const *base;
/* Read the environment if we haven't. */
if (!report_all && !flist)
{
char const *env;
if ((env=getenv("TRACY_INLIBS")) && (flist=mkwords(env)))
is_whitelist = 1;
else if ((env=getenv("TRACY_EXLIBS")) && (flist=mkwords(env)))
is_whitelist = 0;
else
report_all = 1;
}
/* Match against $flist if we have one. */
if (!report_all)
{
if ((base = match_words(flist, fname)) != NULL)
return is_whitelist ? base : NULL;
else if (is_whitelist)
return NULL;
}
/* Return the basename of $fname. */
return (!(base = strrchr(fname, '/'))) ? fname : base + 1;
} /* report_dso */
int Dawg::check_for_words(const char *filename,
const UNICHARSET &unicharset,
bool enable_wildcard) const {
if (filename == nullptr) return 0;
FILE *word_file;
char string [CHARS_PER_LINE];
int misses = 0;
UNICHAR_ID wildcard = unicharset.unichar_to_id(kWildcard);
word_file = fopen(filename, "r");
if (word_file == nullptr) {
tprintf("Error: Could not open file %s\n", filename);
ASSERT_HOST(word_file);
}
while (fgets (string, CHARS_PER_LINE, word_file) != nullptr) {
chomp_string(string); // remove newline
WERD_CHOICE word(string, unicharset);
if (word.length() > 0 &&
!word.contains_unichar_id(INVALID_UNICHAR_ID)) {
if (!match_words(&word, 0, 0,
enable_wildcard ? wildcard : INVALID_UNICHAR_ID)) {
tprintf("Missing word: %s\n", string);
++misses;
}
} else {
tprintf("Failed to create a valid word from %s\n", string);
}
}
fclose (word_file);
// Make sure the user sees this with fprintf instead of tprintf.
if (debug_level_) tprintf("Number of lost words=%d\n", misses);
return misses;
}
int search(const word_t *pre_words, const char **pat_words, int pat_count, int **documents) {
// find common words
word_t **words;
int num_of_words = match_words(pre_words, pat_words, pat_count, &words);
// find priorities of these words
double *priorities = NULL;
get_priorities((const word_t **)words, num_of_words, &priorities);
// get used documents
int num_of_doc = get_documents((const word_t **)words, num_of_words, documents);
int *measures = (int *)malloc(sizeof(int) * num_of_doc);
int *step_words = NULL;
// i.e. initializing with 0s
int *step_info_position = (int *)calloc(sizeof(int), num_of_words);
int cur_doc_num = -1;
int cur_doc = -1;
int num_step_words = 0;
while((num_step_words = get_step_words((const word_t **)words, num_of_words,
cur_doc, step_info_position, &step_words)) != -1) {
int measure = get_measure((const word_t **)words, step_info_position,
priorities, step_words, num_step_words);
++cur_doc_num;
cur_doc = (*documents)[cur_doc_num];
for(int i = 0; i < num_of_doc; i++)
if(cur_doc == (*documents)[i]) {
measures[i] = measure;
break;
}
if(cur_doc_num == num_of_doc - 1)
break;
}
for(int i = 0; i < num_of_doc - 1; i++) {
int max = 0;
for(int j = 1; j < num_of_doc - i; j++) {
if(measures[j] < measures[max]) {
max = j;
}
}
int temp = measures[num_of_doc - i - 1];
measures[num_of_doc - i - 1] = measures[max];
measures[max] = temp;
temp = (*documents)[num_of_doc - i - 1];
(*documents)[num_of_doc - i - 1] = (*documents)[max];
(*documents)[max] = temp;
}
free(step_info_position);
free(priorities);
free(step_words);
for(int i = 0; i < num_of_doc; i++)
printf("%d ", measures[i]);
printf("\n");
free (measures);
free(words);
return num_of_doc;
}
