void
IntlTestCollator::doTestVariant(Collator* col, const UnicodeString &source, const UnicodeString &target, Collator::EComparisonResult result)
{
UErrorCode status = U_ZERO_ERROR;
UCollator *myCollation = col->toUCollator();
Collator::EComparisonResult compareResult = col->compare(source, target);
CollationKey srckey, tgtkey;
col->getCollationKey(source, srckey, status);
col->getCollationKey(target, tgtkey, status);
if (U_FAILURE(status)){
errln("Creation of collation keys failed\n");
}
Collator::EComparisonResult keyResult = srckey.compareTo(tgtkey);
reportCResult(source, target, srckey, tgtkey, compareResult, keyResult, result, result);
UColAttributeValue norm = ucol_getAttribute(myCollation, UCOL_NORMALIZATION_MODE, &status);
int32_t sLen = source.length(), tLen = target.length();
const UChar* src = source.getBuffer();
const UChar* trg = target.getBuffer();
UCollationResult compareResultIter = (UCollationResult)result;
{
UCharIterator sIter, tIter;
uiter_setString(&sIter, src, sLen);
uiter_setString(&tIter, trg, tLen);
compareResultIter = ucol_strcollIter(myCollation, &sIter, &tIter, &status);
if(compareResultIter != (UCollationResult)result) {
errln("Different result for iterative comparison "+source+" "+target);
}
}
/* convert the strings to UTF-8 and do try comparing with char iterator */
if(!quick) { /*!QUICK*/
char utf8Source[256], utf8Target[256];
int32_t utf8SourceLen = 0, utf8TargetLen = 0;
u_strToUTF8(utf8Source, 256, &utf8SourceLen, src, sLen, &status);
if(U_FAILURE(status)) { /* probably buffer is not big enough */
log("Src UTF-8 buffer too small! Will not compare!\n");
} else {
u_strToUTF8(utf8Target, 256, &utf8TargetLen, trg, tLen, &status);
if(U_SUCCESS(status)) { /* probably buffer is not big enough */
UCollationResult compareResultUTF8 = (UCollationResult)result, compareResultUTF8Norm = (UCollationResult)result;
UCharIterator sIter, tIter;
/*log_verbose("Strings converted to UTF-8:%s, %s\n", aescstrdup(source,-1), aescstrdup(target,-1));*/
uiter_setUTF8(&sIter, utf8Source, utf8SourceLen);
uiter_setUTF8(&tIter, utf8Target, utf8TargetLen);
/*uiter_setString(&sIter, source, sLen);
uiter_setString(&tIter, target, tLen);*/
compareResultUTF8 = ucol_strcollIter(myCollation, &sIter, &tIter, &status);
ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, UCOL_ON, &status);
sIter.move(&sIter, 0, UITER_START);
tIter.move(&tIter, 0, UITER_START);
compareResultUTF8Norm = ucol_strcollIter(myCollation, &sIter, &tIter, &status);
ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, norm, &status);
if(compareResultUTF8 != compareResultIter) {
errln("different results in iterative comparison for UTF-16 and UTF-8 encoded strings. "+source+", "+target);
}
if(compareResultUTF8 != compareResultUTF8Norm) {
errln("different results in iterative when normalization is turned on with UTF-8 strings. "+source+", "+target);
}
} else {
log("Target UTF-8 buffer too small! Did not compare!\n");
}
if(U_FAILURE(status)) {
log("UTF-8 strcoll failed! Ignoring result\n");
}
}
}
/* testing the partial sortkeys */
{ /*!QUICK*/
int32_t partialSizes[] = { 3, 1, 2, 4, 8, 20, 80 }; /* just size 3 in the quick mode */
int32_t partialSizesSize = 1;
if(!quick) {
partialSizesSize = 7;
}
int32_t i = 0;
log("partial sortkey test piecesize=");
for(i = 0; i < partialSizesSize; i++) {
UCollationResult partialSKResult = (UCollationResult)result, partialNormalizedSKResult = (UCollationResult)result;
log("%i ", partialSizes[i]);
partialSKResult = compareUsingPartials(myCollation, src, sLen, trg, tLen, partialSizes[i], status);
if(partialSKResult != (UCollationResult)result) {
errln("Partial sortkey comparison returned wrong result: "+source+", "+target+" (size "+partialSizes[i]+")");
}
if(norm != UCOL_ON && !quick) {
log("N ");
ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, UCOL_ON, &status);
partialNormalizedSKResult = compareUsingPartials(myCollation, src, sLen, trg, tLen, partialSizes[i], status);
ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, norm, &status);
if(partialSKResult != partialNormalizedSKResult) {
errln("Partial sortkey comparison gets different result when normalization is on: "+source+", "+target+" (size "+partialSizes[i]+")");
}
}
}
log("\n");
}
/*
if (compareResult != result) {
errln("String comparison failed in variant test\n");
}
if (keyResult != result) {
errln("Collation key comparison failed in variant test\n");
}
*/
}
static void doTestVariant(UCollator* myCollation, const UChar source[], const UChar target[], UCollationResult result)
{
int32_t sortklen1, sortklen2, sortklenmax, sortklenmin;
int temp=0, gSortklen1=0,gSortklen2=0;
UCollationResult compareResult, compareResulta, keyResult, compareResultIter = result;
uint8_t *sortKey1, *sortKey2, *sortKey1a, *sortKey2a;
uint32_t sLen = u_strlen(source);
uint32_t tLen = u_strlen(target);
char buffer[256];
uint32_t len;
UErrorCode status = U_ZERO_ERROR;
UColAttributeValue norm = ucol_getAttribute(myCollation, UCOL_NORMALIZATION_MODE, &status);
UCharIterator sIter, tIter;
uiter_setString(&sIter, source, sLen);
uiter_setString(&tIter, target, tLen);
compareResultIter = ucol_strcollIter(myCollation, &sIter, &tIter, &status);
if(compareResultIter != result) {
log_err("different results in iterative comparison for UTF-16 encoded strings. %s, %s\n", aescstrdup(source,-1), aescstrdup(target,-1));
}
/* convert the strings to UTF-8 and do try comparing with char iterator */
if(QUICK <= 0) { /*!QUICK*/
char utf8Source[256], utf8Target[256];
int32_t utf8SourceLen = 0, utf8TargetLen = 0;
u_strToUTF8(utf8Source, 256, &utf8SourceLen, source, sLen, &status);
if(U_FAILURE(status)) { /* probably buffer is not big enough */
log_verbose("Src UTF-8 buffer too small! Will not compare!\n");
} else {
u_strToUTF8(utf8Target, 256, &utf8TargetLen, target, tLen, &status);
if(U_SUCCESS(status)) { /* probably buffer is not big enough */
UCollationResult compareResultUTF8 = result, compareResultUTF8Norm = result;
/*UCharIterator sIter, tIter;*/
/*log_verbose("Strings converted to UTF-8:%s, %s\n", aescstrdup(source,-1), aescstrdup(target,-1));*/
uiter_setUTF8(&sIter, utf8Source, utf8SourceLen);
uiter_setUTF8(&tIter, utf8Target, utf8TargetLen);
/*uiter_setString(&sIter, source, sLen);
uiter_setString(&tIter, target, tLen);*/
compareResultUTF8 = ucol_strcollIter(myCollation, &sIter, &tIter, &status);
ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, UCOL_ON, &status);
sIter.move(&sIter, 0, UITER_START);
tIter.move(&tIter, 0, UITER_START);
compareResultUTF8Norm = ucol_strcollIter(myCollation, &sIter, &tIter, &status);
ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, norm, &status);
if(compareResultUTF8 != compareResultIter) {
log_err("different results in iterative comparison for UTF-16 and UTF-8 encoded strings. %s, %s\n", aescstrdup(source,-1), aescstrdup(target,-1));
}
if(compareResultUTF8 != compareResultUTF8Norm) {
log_err("different results in iterative when normalization is turned on with UTF-8 strings. %s, %s\n", aescstrdup(source,-1), aescstrdup(target,-1));
}
} else {
log_verbose("Target UTF-8 buffer too small! Did not compare!\n");
}
if(U_FAILURE(status)) {
log_verbose("UTF-8 strcoll failed! Ignoring result\n");
}
}
}
/* testing the partial sortkeys */
if(1) { /*!QUICK*/
int32_t i = 0;
int32_t partialSizes[] = { 3, 1, 2, 4, 8, 20, 80 }; /* just size 3 in the quick mode */
int32_t partialSizesSize = 1;
if(QUICK <= 0) {
partialSizesSize = 7;
}
/*log_verbose("partial sortkey test piecesize=");*/
for(i = 0; i < partialSizesSize; i++) {
UCollationResult partialSKResult = result, partialNormalizedSKResult = result;
/*log_verbose("%i ", partialSizes[i]);*/
partialSKResult = compareUsingPartials(myCollation, source, sLen, target, tLen, partialSizes[i], &status);
if(partialSKResult != result) {
log_err("Partial sortkey comparison returned wrong result (%i exp. %i): %s, %s (size %i)\n",
partialSKResult, result,
aescstrdup(source,-1), aescstrdup(target,-1), partialSizes[i]);
}
if(QUICK <= 0 && norm != UCOL_ON) {
/*log_verbose("N ");*/
ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, UCOL_ON, &status);
partialNormalizedSKResult = compareUsingPartials(myCollation, source, sLen, target, tLen, partialSizes[i], &status);
ucol_setAttribute(myCollation, UCOL_NORMALIZATION_MODE, norm, &status);
if(partialSKResult != partialNormalizedSKResult) {
log_err("Partial sortkey comparison gets different result when normalization is on: %s, %s (size %i)\n",
aescstrdup(source,-1), aescstrdup(target,-1), partialSizes[i]);
}
}
}
/*log_verbose("\n");*/
}
compareResult = ucol_strcoll(myCollation, source, sLen, target, tLen);
compareResulta = ucol_strcoll(myCollation, source, -1, target, -1);
if (compareResult != compareResulta) {
log_err("ucol_strcoll result from null terminated and explicit length strings differs.\n");
}
sortklen1=ucol_getSortKey(myCollation, source, sLen, NULL, 0);
sortklen2=ucol_getSortKey(myCollation, target, tLen, NULL, 0);
sortklenmax = (sortklen1>sortklen2?sortklen1:sortklen2);
sortklenmin = (sortklen1<sortklen2?sortklen1:sortklen2);
sortKey1 =(uint8_t*)malloc(sizeof(uint8_t) * (sortklenmax+1));
sortKey1a=(uint8_t*)malloc(sizeof(uint8_t) * (sortklenmax+1));
ucol_getSortKey(myCollation, source, sLen, sortKey1, sortklen1+1);
ucol_getSortKey(myCollation, source, -1, sortKey1a, sortklen1+1);
sortKey2 =(uint8_t*)malloc(sizeof(uint8_t) * (sortklenmax+1));
sortKey2a=(uint8_t*)malloc(sizeof(uint8_t) * (sortklenmax+1));
ucol_getSortKey(myCollation, target, tLen, sortKey2, sortklen2+1);
ucol_getSortKey(myCollation, target, -1, sortKey2a, sortklen2+1);
/* Check that sort key generated with null terminated string is identical */
/* to that generted with a length specified. */
if (uprv_strcmp((const char *)sortKey1, (const char *)sortKey1a) != 0 ||
uprv_strcmp((const char *)sortKey2, (const char *)sortKey2a) != 0 ) {
log_err("Sort Keys from null terminated and explicit length strings differ.\n");
}
/*memcmp(sortKey1, sortKey2,sortklenmax);*/
temp= uprv_strcmp((const char *)sortKey1, (const char *)sortKey2);
gSortklen1 = uprv_strlen((const char *)sortKey1)+1;
gSortklen2 = uprv_strlen((const char *)sortKey2)+1;
if(sortklen1 != gSortklen1){
log_err("SortKey length does not match Expected: %i Got: %i\n",sortklen1, gSortklen1);
log_verbose("Generated sortkey: %s\n", sortKeyToString(myCollation, sortKey1, buffer, &len));
}
if(sortklen2!= gSortklen2){
log_err("SortKey length does not match Expected: %i Got: %i\n", sortklen2, gSortklen2);
log_verbose("Generated sortkey: %s\n", sortKeyToString(myCollation, sortKey2, buffer, &len));
}
if(temp < 0) {
keyResult=UCOL_LESS;
}
else if(temp > 0) {
keyResult= UCOL_GREATER;
}
else {
keyResult = UCOL_EQUAL;
}
reportCResult( source, target, sortKey1, sortKey2, compareResult, keyResult, compareResultIter, result );
free(sortKey1);
free(sortKey2);
free(sortKey1a);
free(sortKey2a);
}
