void AlphabeticIndex::addIndexExemplars(const Locale &locale, UErrorCode &status) {
LocalULocaleDataPointer uld(ulocdata_open(locale.getName(), &status));
if (U_FAILURE(status)) {
return;
}
UnicodeSet exemplars;
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_INDEX, &status);
if (U_SUCCESS(status)) {
initialLabels_->addAll(exemplars);
return;
}
status = U_ZERO_ERROR; // Clear out U_MISSING_RESOURCE_ERROR
// The locale data did not include explicit Index characters.
// Synthesize a set of them from the locale's standard exemplar characters.
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_STANDARD, &status);
if (U_FAILURE(status)) {
return;
}
// question: should we add auxiliary exemplars?
if (exemplars.containsSome(0x61, 0x7A) /* a-z */ || exemplars.size() == 0) {
exemplars.add(0x61, 0x7A);
}
if (exemplars.containsSome(0xAC00, 0xD7A3)) { // Hangul syllables
// cut down to small list
exemplars.remove(0xAC00, 0xD7A3).
add(0xAC00).add(0xB098).add(0xB2E4).add(0xB77C).
add(0xB9C8).add(0xBC14).add(0xC0AC).add(0xC544).
add(0xC790).add(0xCC28).add(0xCE74).add(0xD0C0).
add(0xD30C).add(0xD558);
}
if (exemplars.containsSome(0x1200, 0x137F)) { // Ethiopic block
// cut down to small list
// make use of the fact that Ethiopic is allocated in 8's, where
// the base is 0 mod 8.
UnicodeSet ethiopic(
UNICODE_STRING_SIMPLE("[[:Block=Ethiopic:]&[:Script=Ethiopic:]]"), status);
UnicodeSetIterator it(ethiopic);
while (it.next() && !it.isString()) {
if ((it.getCodepoint() & 0x7) != 0) {
exemplars.remove(it.getCodepoint());
}
}
}
// Upper-case any that aren't already so.
// (We only do this for synthesized index characters.)
UnicodeSetIterator it(exemplars);
UnicodeString upperC;
while (it.next()) {
const UnicodeString &exemplarC = it.getString();
upperC = exemplarC;
upperC.toUpper(locale);
initialLabels_->add(upperC);
}
}
void AlphabeticIndex::addIndexExemplars(const Locale &locale, UErrorCode &status) {
if (U_FAILURE(status)) { return; }
// Chinese index characters, which are specific to each of the several Chinese tailorings,
// take precedence over the single locale data exemplar set per language.
const char *language = locale.getLanguage();
if (uprv_strcmp(language, "zh") == 0 || uprv_strcmp(language, "ja") == 0 ||
uprv_strcmp(language, "ko") == 0) {
// TODO: This should be done regardless of the language, but it's expensive.
// We should add a Collator function (can be @internal)
// to enumerate just the contractions that start with a given code point or string.
if (addChineseIndexCharacters(status) || U_FAILURE(status)) {
return;
}
}
LocalULocaleDataPointer uld(ulocdata_open(locale.getName(), &status));
if (U_FAILURE(status)) {
return;
}
UnicodeSet exemplars;
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_INDEX, &status);
if (U_SUCCESS(status)) {
initialLabels_->addAll(exemplars);
return;
}
status = U_ZERO_ERROR; // Clear out U_MISSING_RESOURCE_ERROR
// The locale data did not include explicit Index characters.
// Synthesize a set of them from the locale's standard exemplar characters.
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_STANDARD, &status);
if (U_FAILURE(status)) {
return;
}
// question: should we add auxiliary exemplars?
if (exemplars.containsSome(0x61, 0x7A) /* a-z */ || exemplars.size() == 0) {
exemplars.add(0x61, 0x7A);
}
if (exemplars.containsSome(0xAC00, 0xD7A3)) { // Hangul syllables
// cut down to small list
exemplars.remove(0xAC00, 0xD7A3).
add(0xAC00).add(0xB098).add(0xB2E4).add(0xB77C).
add(0xB9C8).add(0xBC14).add(0xC0AC).add(0xC544).
add(0xC790).add(0xCC28).add(0xCE74).add(0xD0C0).
add(0xD30C).add(0xD558);
}
if (exemplars.containsSome(0x1200, 0x137F)) { // Ethiopic block
// cut down to small list
// make use of the fact that Ethiopic is allocated in 8's, where
// the base is 0 mod 8.
UnicodeSet ethiopic(
UNICODE_STRING_SIMPLE("[[:Block=Ethiopic:]&[:Script=Ethiopic:]]"), status);
UnicodeSetIterator it(ethiopic);
while (it.next() && !it.isString()) {
if ((it.getCodepoint() & 0x7) != 0) {
exemplars.remove(it.getCodepoint());
}
}
}
// Upper-case any that aren't already so.
// (We only do this for synthesized index characters.)
UnicodeSetIterator it(exemplars);
UnicodeString upperC;
while (it.next()) {
const UnicodeString &exemplarC = it.getString();
upperC = exemplarC;
upperC.toUpper(locale);
initialLabels_->add(upperC);
}
}
void AlphabeticIndex::getIndexExemplars(UnicodeSet &dest, const Locale &locale, UErrorCode &status) {
if (U_FAILURE(status)) {
return;
}
LocalULocaleDataPointer uld(ulocdata_open(locale.getName(), &status));
UnicodeSet exemplars;
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_INDEX, &status);
if (U_SUCCESS(status)) {
dest.addAll(exemplars);
return;
}
status = U_ZERO_ERROR; // Clear out U_MISSING_RESOURCE_ERROR
// Locale data did not include explicit Index characters.
// Synthesize a set of them from the locale's standard exemplar characters.
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_STANDARD, &status);
if (U_FAILURE(status)) {
return;
}
// Upper-case any that aren't already so.
// (We only do this for synthesized index characters.)
UnicodeSetIterator it(exemplars);
UnicodeString upperC;
UnicodeSet lowersToRemove;
UnicodeSet uppersToAdd;
while (it.next()) {
const UnicodeString &exemplarC = it.getString();
upperC = exemplarC;
upperC.toUpper(locale);
if (exemplarC != upperC) {
lowersToRemove.add(exemplarC);
uppersToAdd.add(upperC);
}
}
exemplars.removeAll(lowersToRemove);
exemplars.addAll(uppersToAdd);
// get the exemplars, and handle special cases
// question: should we add auxiliary exemplars?
if (exemplars.containsSome(*CORE_LATIN)) {
exemplars.addAll(*CORE_LATIN);
}
if (exemplars.containsSome(*HANGUL)) {
// cut down to small list
UnicodeSet BLOCK_HANGUL_SYLLABLES(UNICODE_STRING_SIMPLE("[:block=hangul_syllables:]"), status);
exemplars.removeAll(BLOCK_HANGUL_SYLLABLES);
exemplars.addAll(*HANGUL);
}
if (exemplars.containsSome(*ETHIOPIC)) {
// cut down to small list
// make use of the fact that Ethiopic is allocated in 8's, where
// the base is 0 mod 8.
UnicodeSetIterator it(*ETHIOPIC);
while (it.next() && !it.isString()) {
if ((it.getCodepoint() & 0x7) != 0) {
exemplars.remove(it.getCodepoint());
}
}
}
dest.addAll(exemplars);
}
