UNormalizationCheckResult
FilteredNormalizer2::quickCheck(const UnicodeString &s, UErrorCode &errorCode) const {
uprv_checkCanGetBuffer(s, errorCode);
if(U_FAILURE(errorCode)) {
return UNORM_MAYBE;
}
UNormalizationCheckResult result=UNORM_YES;
USetSpanCondition spanCondition=USET_SPAN_SIMPLE;
for(int32_t prevSpanLimit=0; prevSpanLimit<s.length();) {
int32_t spanLimit=set.span(s, prevSpanLimit, spanCondition);
if(spanCondition==USET_SPAN_NOT_CONTAINED) {
spanCondition=USET_SPAN_SIMPLE;
} else {
UNormalizationCheckResult qcResult=
norm2.quickCheck(s.tempSubStringBetween(prevSpanLimit, spanLimit), errorCode);
if(U_FAILURE(errorCode) || qcResult==UNORM_NO) {
return qcResult;
} else if(qcResult==UNORM_MAYBE) {
result=qcResult;
}
spanCondition=USET_SPAN_NOT_CONTAINED;
}
prevSpanLimit=spanLimit;
}
return result;
}
int32_t
FilteredNormalizer2::spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const {
uprv_checkCanGetBuffer(s, errorCode);
if(U_FAILURE(errorCode)) {
return 0;
}
USetSpanCondition spanCondition=USET_SPAN_SIMPLE;
for(int32_t prevSpanLimit=0; prevSpanLimit<s.length();) {
int32_t spanLimit=set.span(s, prevSpanLimit, spanCondition);
if(spanCondition==USET_SPAN_NOT_CONTAINED) {
spanCondition=USET_SPAN_SIMPLE;
} else {
int32_t yesLimit=
prevSpanLimit+
norm2.spanQuickCheckYes(
s.tempSubStringBetween(prevSpanLimit, spanLimit), errorCode);
if(U_FAILURE(errorCode) || yesLimit<spanLimit) {
return yesLimit;
}
spanCondition=USET_SPAN_NOT_CONTAINED;
}
prevSpanLimit=spanLimit;
}
return s.length();
}
UBool
FilteredNormalizer2::isNormalized(const UnicodeString & s, UErrorCode & errorCode) const
{
uprv_checkCanGetBuffer(s, errorCode);
if (U_FAILURE(errorCode))
{
return FALSE;
}
USetSpanCondition spanCondition = USET_SPAN_SIMPLE;
for (int32_t prevSpanLimit = 0; prevSpanLimit < s.length();)
{
int32_t spanLimit = set.span(s, prevSpanLimit, spanCondition);
if (spanCondition == USET_SPAN_NOT_CONTAINED)
{
spanCondition = USET_SPAN_SIMPLE;
}
else
{
if (!norm2.isNormalized(s.tempSubStringBetween(prevSpanLimit, spanLimit), errorCode) ||
U_FAILURE(errorCode)
)
{
return FALSE;
}
spanCondition = USET_SPAN_NOT_CONTAINED;
}
prevSpanLimit = spanLimit;
}
return TRUE;
}
void CollationRegressionTest::Test4179216() {
// you can position a CollationElementIterator in the middle of
// a contracting character sequence, yielding a bogus collation
// element
IcuTestErrorCode errorCode(*this, "Test4179216");
RuleBasedCollator coll(en_us->getRules() + " & C < ch , cH , Ch , CH < cat < crunchy", errorCode);
UnicodeString testText = "church church catcatcher runcrunchynchy";
CollationElementIterator *iter = coll.createCollationElementIterator(testText);
// test that the "ch" combination works properly
iter->setOffset(4, errorCode);
int32_t elt4 = CollationElementIterator::primaryOrder(iter->next(errorCode));
iter->reset();
int32_t elt0 = CollationElementIterator::primaryOrder(iter->next(errorCode));
iter->setOffset(5, errorCode);
int32_t elt5 = CollationElementIterator::primaryOrder(iter->next(errorCode));
// Compares and prints only 16-bit primary weights.
if (elt4 != elt0 || elt5 != elt0) {
errln("The collation elements at positions 0 (0x%04x), "
"4 (0x%04x), and 5 (0x%04x) don't match.",
elt0, elt4, elt5);
}
// test that the "cat" combination works properly
iter->setOffset(14, errorCode);
int32_t elt14 = CollationElementIterator::primaryOrder(iter->next(errorCode));
iter->setOffset(15, errorCode);
int32_t elt15 = CollationElementIterator::primaryOrder(iter->next(errorCode));
iter->setOffset(16, errorCode);
int32_t elt16 = CollationElementIterator::primaryOrder(iter->next(errorCode));
iter->setOffset(17, errorCode);
int32_t elt17 = CollationElementIterator::primaryOrder(iter->next(errorCode));
iter->setOffset(18, errorCode);
int32_t elt18 = CollationElementIterator::primaryOrder(iter->next(errorCode));
iter->setOffset(19, errorCode);
int32_t elt19 = CollationElementIterator::primaryOrder(iter->next(errorCode));
// Compares and prints only 16-bit primary weights.
if (elt14 != elt15 || elt14 != elt16 || elt14 != elt17
|| elt14 != elt18 || elt14 != elt19) {
errln("\"cat\" elements don't match: elt14 = 0x%04x, "
"elt15 = 0x%04x, elt16 = 0x%04x, elt17 = 0x%04x, "
"elt18 = 0x%04x, elt19 = 0x%04x",
elt14, elt15, elt16, elt17, elt18, elt19);
}
// now generate a complete list of the collation elements,
// first using next() and then using setOffset(), and
// make sure both interfaces return the same set of elements
iter->reset();
int32_t elt = iter->next(errorCode);
int32_t count = 0;
while (elt != CollationElementIterator::NULLORDER) {
++count;
elt = iter->next(errorCode);
}
LocalArray<UnicodeString> nextElements(new UnicodeString[count]);
LocalArray<UnicodeString> setOffsetElements(new UnicodeString[count]);
int32_t lastPos = 0;
iter->reset();
elt = iter->next(errorCode);
count = 0;
while (elt != CollationElementIterator::NULLORDER) {
nextElements[count++] = testText.tempSubStringBetween(lastPos, iter->getOffset());
lastPos = iter->getOffset();
elt = iter->next(errorCode);
}
int32_t nextElementsLength = count;
count = 0;
for (int32_t i = 0; i < testText.length(); ) {
iter->setOffset(i, errorCode);
lastPos = iter->getOffset();
elt = iter->next(errorCode);
setOffsetElements[count++] = testText.tempSubStringBetween(lastPos, iter->getOffset());
i = iter->getOffset();
}
for (int32_t i = 0; i < nextElementsLength; i++) {
if (nextElements[i] == setOffsetElements[i]) {
logln(nextElements[i]);
} else {
errln(UnicodeString("Error: next() yielded ") + nextElements[i] +
", but setOffset() yielded " + setOffsetElements[i]);
}
}
delete iter;
}
