string LigatureTable::RemoveCustomLigatures(const string& str) const {
string result;
UNICHAR::const_iterator it_begin = UNICHAR::begin(str.c_str(), str.length());
UNICHAR::const_iterator it_end = UNICHAR::end(str.c_str(), str.length());
char tmp[5];
int len;
int norm_ind;
for (UNICHAR::const_iterator it = it_begin; it != it_end; ++it) {
len = it.get_utf8(tmp);
tmp[len] = '\0';
norm_ind = -1;
for (int i = 0;
UNICHARSET::kCustomLigatures[i][0] != nullptr && norm_ind < 0; ++i) {
if (!strcmp(tmp, UNICHARSET::kCustomLigatures[i][1])) {
norm_ind = i;
}
}
if (norm_ind >= 0) {
result += UNICHARSET::kCustomLigatures[norm_ind][0];
} else {
result += tmp;
}
}
return result;
}
unsigned int SpanUTF8NotWhitespace(const char* text) {
int n_notwhite = 0;
for (UNICHAR::const_iterator it = UNICHAR::begin(text, strlen(text));
it != UNICHAR::end(text, strlen(text)); ++it) {
if (IsWhitespace(*it)) break;
n_notwhite += it.utf8_len();
}
return n_notwhite;
}
string LigatureTable::RemoveLigatures(const string& str) const {
string result;
UNICHAR::const_iterator it_begin = UNICHAR::begin(str.c_str(), str.length());
UNICHAR::const_iterator it_end = UNICHAR::end(str.c_str(), str.length());
char tmp[5];
int len;
for (UNICHAR::const_iterator it = it_begin; it != it_end; ++it) {
len = it.get_utf8(tmp);
tmp[len] = '\0';
LigHash::const_iterator lig_it = lig_to_norm_table_.find(tmp);
if (lig_it != lig_to_norm_table_.end()) {
result += lig_it->second;
} else {
result += tmp;
}
}
return result;
}
bool PangoFontInfo::CoversUTF8Text(const char* utf8_text, int byte_length) const {
PangoFont* font = ToPangoFont();
PangoCoverage* coverage = pango_font_get_coverage(font, NULL);
for (UNICHAR::const_iterator it = UNICHAR::begin(utf8_text, byte_length);
it != UNICHAR::end(utf8_text, byte_length);
++it) {
if (IsWhitespace(*it) || pango_is_zero_width(*it))
continue;
if (pango_coverage_get(coverage, *it) != PANGO_COVERAGE_EXACT) {
char tmp[5];
int len = it.get_utf8(tmp);
tmp[len] = '\0';
tlog(2, "'%s' (U+%x) not covered by font\n", tmp, *it);
return false;
}
}
return true;
}
int PangoFontInfo::DropUncoveredChars(string* utf8_text) const {
PangoFont* font = ToPangoFont();
PangoCoverage* coverage = pango_font_get_coverage(font, NULL);
int num_dropped_chars = 0;
// Maintain two iterators that point into the string. For space efficiency, we
// will repeatedly copy one covered UTF8 character from one to the other, and
// at the end resize the string to the right length.
char* out = const_cast<char*>(utf8_text->c_str());
const UNICHAR::const_iterator it_begin =
UNICHAR::begin(utf8_text->c_str(), utf8_text->length());
const UNICHAR::const_iterator it_end =
UNICHAR::end(utf8_text->c_str(), utf8_text->length());
for (UNICHAR::const_iterator it = it_begin; it != it_end; ++it) {
// Skip bad utf-8.
if (!it.is_legal())
continue; // One suitable error message will still be issued.
if (!IsWhitespace(*it) && !pango_is_zero_width(*it) &&
pango_coverage_get(coverage, *it) != PANGO_COVERAGE_EXACT) {
if (TLOG_IS_ON(2)) {
char tmp[5];
int len = it.get_utf8(tmp);
tmp[len] = '\0';
tlog(2, "'%s' (U+%x) not covered by font\n", tmp, *it);
}
++num_dropped_chars;
continue;
}
strncpy(out, it.utf8_data(), it.utf8_len());
out += it.utf8_len();
}
utf8_text->resize(out - utf8_text->c_str());
return num_dropped_chars;
}