UChar32
UCharCharacterIterator::current32() const {
if (pos >= begin && pos < end) {
UChar32 c;
UTF_GET_CHAR(text, begin, pos, end, c);
return c;
} else {
return DONE;
}
}
/**
* Implements {@link Transliterator#handleTransliterate}.
*/
void UppercaseTransliterator::handleTransliterate(Replaceable& text,
UTransPosition& offsets,
UBool isIncremental) const {
int32_t textPos = offsets.start;
if (textPos >= offsets.limit) return;
// get string for context
UnicodeString original;
text.extractBetween(offsets.contextStart, offsets.contextLimit, original);
UCharIterator iter;
uiter_setReplaceable(&iter, &text);
iter.start = offsets.contextStart;
iter.limit = offsets.contextLimit;
// Walk through original string
// If there is a case change, modify corresponding position in replaceable
int32_t i = textPos - offsets.contextStart;
int32_t limit = offsets.limit - offsets.contextStart;
UChar32 cp;
int32_t oldLen;
for (; i < limit; ) {
UTF_GET_CHAR(original.getBuffer(), 0, i, original.length(), cp);
oldLen = UTF_CHAR_LENGTH(cp);
i += oldLen;
iter.index = i; // Point _past_ current char
int32_t newLen = u_internalToUpper(cp, &iter, buffer, u_getMaxCaseExpansion(), loc.getName());
if (newLen >= 0) {
UnicodeString temp(buffer, newLen);
text.handleReplaceBetween(textPos, textPos + oldLen, temp);
if (newLen != oldLen) {
textPos += newLen;
offsets.limit += newLen - oldLen;
offsets.contextLimit += newLen - oldLen;
continue;
}
}
textPos += oldLen;
}
offsets.start = offsets.limit;
}
Hashtable *CanonicalIterator::getEquivalents2(Hashtable *fillinResult, const UChar *segment, int32_t segLen, UErrorCode &status) {
if (U_FAILURE(status)) {
return NULL;
}
//if (PROGRESS) printf("Adding: %s\n", UToS(Tr(segment)));
UnicodeString toPut(segment, segLen);
fillinResult->put(toPut, new UnicodeString(toPut), status);
UnicodeSet starts;
// cycle through all the characters
UChar32 cp;
for (int32_t i = 0; i < segLen; i += UTF16_CHAR_LENGTH(cp)) {
// see if any character is at the start of some decomposition
UTF_GET_CHAR(segment, 0, i, segLen, cp);
if (!nfcImpl.getCanonStartSet(cp, starts)) {
continue;
}
// if so, see which decompositions match
UnicodeSetIterator iter(starts);
while (iter.next()) {
UChar32 cp2 = iter.getCodepoint();
Hashtable remainder(status);
remainder.setValueDeleter(uhash_deleteUnicodeString);
if (extract(&remainder, cp2, segment, segLen, i, status) == NULL) {
continue;
}
// there were some matches, so add all the possibilities to the set.
UnicodeString prefix(segment, i);
prefix += cp2;
int32_t el = -1;
const UHashElement *ne = remainder.nextElement(el);
while (ne != NULL) {
UnicodeString item = *((UnicodeString *)(ne->value.pointer));
UnicodeString *toAdd = new UnicodeString(prefix);
/* test for NULL */
if (toAdd == 0) {
status = U_MEMORY_ALLOCATION_ERROR;
return NULL;
}
*toAdd += item;
fillinResult->put(*toAdd, toAdd, status);
//if (PROGRESS) printf("Adding: %s\n", UToS(Tr(*toAdd)));
ne = remainder.nextElement(el);
}
}
}
/* Test for buffer overflows */
if(U_FAILURE(status)) {
return NULL;
}
return fillinResult;
}
/**
* Implements {@link Transliterator#handleTransliterate}.
*/
void TitlecaseTransliterator::handleTransliterate(
Replaceable& text, UTransPosition& offsets,
UBool isIncremental) const {
if (SKIP == NULL) {
return;
}
// Our mode; we are either converting letter toTitle or
// toLower.
UBool doTitle = TRUE;
// Determine if there is a preceding context of CASED SKIP*,
// in which case we want to start in toLower mode. If the
// prior context is anything else (including empty) then start
// in toTitle mode.
UChar32 c;
int32_t start;
for (start = offsets.start - 1; start >= offsets.contextStart; start -= UTF_CHAR_LENGTH(c)) {
c = text.char32At(start);
if (SKIP->contains(c)) {
continue;
}
doTitle = !CASED->contains(c);
break;
}
// Convert things after a CASED character toLower; things
// after a non-CASED, non-SKIP character toTitle. SKIP
// characters are copied directly and do not change the mode.
int32_t textPos = offsets.start;
if (textPos >= offsets.limit) return;
UnicodeString original;
text.extractBetween(offsets.contextStart, offsets.contextLimit, original);
UCharIterator iter;
uiter_setReplaceable(&iter, &text);
iter.start = offsets.contextStart;
iter.limit = offsets.contextLimit;
// Walk through original string
// If there is a case change, modify corresponding position in replaceable
int32_t i = textPos - offsets.contextStart;
int32_t limit = offsets.limit - offsets.contextStart;
UChar32 cp;
int32_t oldLen;
int32_t newLen;
for (; i < limit; ) {
UTF_GET_CHAR(original.getBuffer(), 0, i, original.length(), cp);
oldLen = UTF_CHAR_LENGTH(cp);
i += oldLen;
iter.index = i; // Point _past_ current char
if (!SKIP->contains(cp)) {
if (doTitle) {
newLen = u_internalToTitle(cp, &iter, buffer, u_getMaxCaseExpansion(), loc.getName());
} else {
newLen = u_internalToLower(cp, &iter, buffer, u_getMaxCaseExpansion(), loc.getName());
}
doTitle = !CASED->contains(cp);
if (newLen >= 0) {
UnicodeString temp(buffer, newLen);
text.handleReplaceBetween(textPos, textPos + oldLen, temp);
if (newLen != oldLen) {
textPos += newLen;
offsets.limit += newLen - oldLen;
offsets.contextLimit += newLen - oldLen;
continue;
}
}
}
textPos += oldLen;
}
offsets.start = offsets.limit;
}
