bool CubeUtils::IsCaseInvariant(const char_32 *str32, CharSet *char_set) {
bool all_one_case = true;
bool capitalized;
bool prev_upper;
bool prev_lower;
bool first_upper;
bool first_lower;
bool cur_upper;
bool cur_lower;
string str8;
if (!char_set) {
// If cube char_set is missing, use C-locale-dependent functions
// on UTF8 characters to determine case properties.
first_upper = isupper(str32[0]);
first_lower = islower(str32[0]);
if (first_upper)
capitalized = true;
prev_upper = first_upper;
prev_lower = islower(str32[0]);
for (int c = 1; str32[c] != 0; ++c) {
cur_upper = isupper(str32[c]);
cur_lower = islower(str32[c]);
if ((prev_upper && cur_lower) || (prev_lower && cur_upper))
all_one_case = false;
if (cur_upper)
capitalized = false;
prev_upper = cur_upper;
prev_lower = cur_lower;
}
} else {
UNICHARSET *unicharset = char_set->InternalUnicharset();
// Use UNICHARSET functions to determine case properties
first_upper = unicharset->get_isupper(char_set->ClassID(str32[0]));
first_lower = unicharset->get_islower(char_set->ClassID(str32[0]));
if (first_upper)
capitalized = true;
prev_upper = first_upper;
prev_lower = unicharset->get_islower(char_set->ClassID(str32[0]));
for (int c = 1; c < StrLen(str32); ++c) {
cur_upper = unicharset->get_isupper(char_set->ClassID(str32[c]));
cur_lower = unicharset->get_islower(char_set->ClassID(str32[c]));
if ((prev_upper && cur_lower) || (prev_lower && cur_upper))
all_one_case = false;
if (cur_upper)
capitalized = false;
prev_upper = cur_upper;
prev_lower = cur_lower;
}
}
return all_one_case || capitalized;
}
char_32 *CubeUtils::ToLower(const char_32 *str32, CharSet *char_set) {
if (!char_set) {
return NULL;
}
UNICHARSET *unicharset = char_set->InternalUnicharset();
int len = StrLen(str32);
char_32 *lower = new char_32[len + 1];
if (!lower)
return NULL;
for (int i = 0; i < len; ++i) {
char_32 ch = str32[i];
if (ch == INVALID_UNICHAR_ID) {
delete[] lower;
return NULL;
}
// convert upper-case characters to lower-case
if (unicharset->get_isupper(char_set->ClassID(ch))) {
UNICHAR_ID uid_lower = unicharset->get_other_case(char_set->ClassID(ch));
const char_32 *str32_lower = char_set->ClassString(uid_lower);
// expect lower-case version of character to be a single character
if (!str32_lower || StrLen(str32_lower) != 1) {
delete[] lower;
return NULL;
}
lower[i] = str32_lower[0];
} else {
lower[i] = ch;
}
}
lower[len] = 0;
return lower;
}
int Dict::case_ok(const WERD_CHOICE &word, const UNICHARSET &unicharset) {
int state = 0;
int x;
for (x = 0; x < word.length(); ++x) {
UNICHAR_ID ch_id = word.unichar_id(x);
if (unicharset.get_isupper(ch_id))
state = case_state_table[state][1];
else if (unicharset.get_islower(ch_id))
state = case_state_table[state][2];
else if (unicharset.get_isdigit(ch_id))
state = case_state_table[state][3];
else
state = case_state_table[state][0];
if (state == -1) return false;
}
return state != 5; // single lower is bad
}
BOOL8 Tesseract::non_O_upper(const UNICHARSET& ch_set, UNICHAR_ID unichar_id) {
return ch_set.get_isupper(unichar_id) && !ch_set.eq(unichar_id, "O");
}