static intptr_t hashChunk(const UCollator *Collator, intptr_t HashState,
const uint16_t *Str, uint32_t Length,
UErrorCode *ErrorCode) {
#if defined(__CYGWIN__) || defined(_MSC_VER) || defined(__MINGW32__)
UCollationElements *CollationIterator = ucol_openElements(
Collator, reinterpret_cast<const UChar *>(Str), Length, ErrorCode);
#else
UCollationElements *CollationIterator = ucol_openElements(
Collator, Str, Length, ErrorCode);
#endif
while (U_SUCCESS(*ErrorCode)) {
intptr_t Elem = ucol_next(CollationIterator, ErrorCode);
// Ignore zero valued collation elements. They don't participate in the
// ordering relation.
if (Elem == 0)
continue;
if (Elem != UCOL_NULLORDER) {
Elem *= HASH_M;
Elem ^= Elem >> HASH_R;
Elem *= HASH_M;
HashState *= HASH_M;
HashState ^= Elem;
} else {
break;
/*
Return value is a "Win32 BOOL" (1 = true, 0 = false)
*/
extern "C" int32_t StartsWith(
const char* lpLocaleName, const UChar* lpTarget, int32_t cwTargetLength, const UChar* lpSource, int32_t cwSourceLength, int32_t options)
{
int32_t result = FALSE;
UErrorCode err = U_ZERO_ERROR;
UCollator* pColl = GetCollatorForLocaleAndOptions(lpLocaleName, options, &err);
if (U_SUCCESS(err))
{
UStringSearch* pSearch = usearch_openFromCollator(lpTarget, cwTargetLength, lpSource, cwSourceLength, pColl, nullptr, &err);
int32_t idx = USEARCH_DONE;
if (U_SUCCESS(err))
{
idx = usearch_first(pSearch, &err);
if (idx != USEARCH_DONE)
{
if (idx == 0)
{
result = TRUE;
}
else
{
UCollationElements* pCollElem = ucol_openElements(pColl, lpSource, idx, &err);
if (U_SUCCESS(err))
{
int32_t curCollElem = UCOL_NULLORDER;
result = TRUE;
while ((curCollElem = ucol_next(pCollElem, &err)) != UCOL_NULLORDER)
{
if (curCollElem != 0)
{
// Non ignorable collation element found between start of the
// string and the first match for lpTarget.
result = FALSE;
break;
}
}
if (U_FAILURE(err))
{
result = FALSE;
}
ucol_closeElements(pCollElem);
}
}
}
usearch_close(pSearch);
}
ucol_close(pColl);
}
return result;
}
// Collator.startswith {{{
static PyObject *
icu_Collator_collation_order(icu_Collator *self, PyObject *args, PyObject *kwargs) {
PyObject *a_;
int32_t asz;
int32_t actual_a;
UChar *a;
wchar_t *aw;
UErrorCode status = U_ZERO_ERROR;
UCollationElements *iter = NULL;
int order = 0, len = -1;
if (!PyArg_ParseTuple(args, "U", &a_)) return NULL;
asz = (int32_t)PyUnicode_GetSize(a_);
a = (UChar*)calloc(asz*4 + 2, sizeof(UChar));
aw = (wchar_t*)calloc(asz*4 + 2, sizeof(wchar_t));
if (a == NULL || aw == NULL ) return PyErr_NoMemory();
actual_a = (int32_t)PyUnicode_AsWideChar((PyUnicodeObject*)a_, aw, asz*4+1);
if (actual_a > -1) {
u_strFromWCS(a, asz*4 + 1, &actual_a, aw, -1, &status);
iter = ucol_openElements(self->collator, a, actual_a, &status);
if (iter != NULL && U_SUCCESS(status)) {
order = ucol_next(iter, &status);
len = ucol_getOffset(iter);
ucol_closeElements(iter); iter = NULL;
}
}
free(a); free(aw);
return Py_BuildValue("ii", order, len);
} // }}}
/// Construct the ASCII collation table.
ASCIICollation() {
const UCollator *Collator = GetRootCollator();
for (unsigned char c = 0; c < 128; ++c) {
UErrorCode ErrorCode = U_ZERO_ERROR;
intptr_t NumCollationElts = 0;
#if defined(__CYGWIN__) || defined(_MSC_VER) || defined(__MINGW32__)
UChar Buffer[1];
#else
uint16_t Buffer[1];
#endif
Buffer[0] = c;
UCollationElements *CollationIterator =
ucol_openElements(Collator, Buffer, 1, &ErrorCode);
while (U_SUCCESS(ErrorCode)) {
intptr_t Elem = ucol_next(CollationIterator, &ErrorCode);
if (Elem != UCOL_NULLORDER) {
CollationTable[c] = Elem;
++NumCollationElts;
} else {
break;
}
}
ucol_closeElements(CollationIterator);
if (U_FAILURE(ErrorCode) || NumCollationElts != 1) {
swift::crash("Error setting up the ASCII collation table");
}
}
}
//static jint NativeCollation_next(JNIEnv* env, jclass, jint address) {
JNIEXPORT jint JNICALL
Java_com_ibm_icu4jni_text_NativeCollation_next(JNIEnv* env, jclass,
jint address) {
UErrorCode status = U_ZERO_ERROR;
jint result = ucol_next(toCollationElements(address), &status);
icu4jni_error(env, status);
return result;
}
CEList::CEList(UCollator *coll, const UnicodeString &string, UErrorCode &status)
: ces(NULL), listMax(CELIST_BUFFER_SIZE), listSize(0)
{
UCollationElements *elems = ucol_openElements(coll, string.getBuffer(), string.length(), &status);
UCollationStrength strength = ucol_getStrength(coll);
UBool toShift = ucol_getAttribute(coll, UCOL_ALTERNATE_HANDLING, &status) == UCOL_SHIFTED;
uint32_t variableTop = ucol_getVariableTop(coll, &status);
uint32_t strengthMask = 0;
int32_t order;
if (U_FAILURE(status)) {
return;
}
// **** only set flag if string has Han(gul) ****
// ucol_forceHanImplicit(elems, &status); -- removed for ticket #10476
switch (strength)
{
default:
strengthMask |= UCOL_TERTIARYORDERMASK;
U_FALLTHROUGH;
case UCOL_SECONDARY:
strengthMask |= UCOL_SECONDARYORDERMASK;
U_FALLTHROUGH;
case UCOL_PRIMARY:
strengthMask |= UCOL_PRIMARYORDERMASK;
}
ces = ceBuffer;
while ((order = ucol_next(elems, &status)) != UCOL_NULLORDER) {
UBool cont = isContinuation(order);
order &= strengthMask;
if (toShift && variableTop > (uint32_t)order && (order & UCOL_PRIMARYORDERMASK) != 0) {
if (strength >= UCOL_QUATERNARY) {
order &= UCOL_PRIMARYORDERMASK;
} else {
order = UCOL_IGNORABLE;
}
}
if (order == UCOL_IGNORABLE) {
continue;
}
if (cont) {
order |= UCOL_CONTINUATION_MARKER;
}
add(order, status);
}
ucol_closeElements(elems);
}
__swift_int32_t swift::_swift_stdlib_unicodeCollationIterator_next(
void *CollationIterator, bool *HitEnd) {
UErrorCode ErrorCode = U_ZERO_ERROR;
auto Result = ucol_next(
static_cast<UCollationElements *>(CollationIterator), &ErrorCode);
if (U_FAILURE(ErrorCode)) {
swift::crash("_swift_stdlib_unicodeCollationIterator_next: ucol_next() failed.");
}
*HitEnd = (Result == UCOL_NULLORDER);
return Result;
}
static void assertEqual(UCollationElements *i1, UCollationElements *i2)
{
int32_t c1, c2;
int32_t count = 0;
UErrorCode status = U_ZERO_ERROR;
do
{
c1 = ucol_next(i1, &status);
c2 = ucol_next(i2, &status);
if (c1 != c2)
{
log_err("Error in iteration %d assetEqual between\n %d and %d, they are not equal\n", count, c1, c2);
break;
}
count += 1;
}
while (c1 != UCOL_NULLORDER);
}
const CEI *Target::nextCE(int32_t offset)
{
UErrorCode status = U_ZERO_ERROR;
int32_t low = -1, high = -1;
uint32_t order;
UBool cont = FALSE;
if (offset >= bufferMin && offset < bufferMax) {
return &ceb[offset];
}
if (bufferMax >= bufferSize || offset != bufferMax) {
return NULL;
}
do {
low = ucol_getOffset(elements);
order = ucol_next(elements, &status);
high = ucol_getOffset(elements);
if (order == (uint32_t)UCOL_NULLORDER) {
//high = low = -1;
break;
}
cont = isContinuation(order);
order &= strengthMask;
if (toShift && variableTop > order && (order & UCOL_PRIMARYORDERMASK) != 0) {
if (strength >= UCOL_QUATERNARY) {
order &= UCOL_PRIMARYORDERMASK;
} else {
order = UCOL_IGNORABLE;
}
}
} while (order == UCOL_IGNORABLE);
if (cont) {
order |= UCOL_CONTINUATION_MARKER;
}
ceb[offset].order = order;
ceb[offset].lowOffset = low;
ceb[offset].highOffset = high;
bufferMax += 1;
return &ceb[offset];
}
/**
* Return an integer array containing all of the collation orders
* returned by calls to next on the specified iterator
*/
OrderAndOffset* getOrders(UCollationElements *iter, int32_t *orderLength)
{
UErrorCode status;
int32_t order;
int32_t maxSize = 100;
int32_t size = 0;
int32_t offset = ucol_getOffset(iter);
OrderAndOffset *temp;
OrderAndOffset *orders =(OrderAndOffset *)malloc(sizeof(OrderAndOffset) * maxSize);
status= U_ZERO_ERROR;
while ((order=ucol_next(iter, &status)) != UCOL_NULLORDER)
{
if (size == maxSize)
{
maxSize *= 2;
temp = (OrderAndOffset *)malloc(sizeof(OrderAndOffset) * maxSize);
memcpy(temp, orders, size * sizeof(OrderAndOffset));
free(orders);
orders = temp;
}
orders[size].order = order;
orders[size].offset = offset;
offset = ucol_getOffset(iter);
size += 1;
}
if (maxSize > size && size > 0)
{
temp = (OrderAndOffset *)malloc(sizeof(OrderAndOffset) * size);
memcpy(temp, orders, size * sizeof(OrderAndOffset));
free(orders);
orders = temp;
}
*orderLength = size;
return orders;
}
OrderList::OrderList(UCollator *coll, const UnicodeString &string, int32_t stringOffset)
: list(NULL), listMax(16), listSize(0)
{
UErrorCode status = U_ZERO_ERROR;
UCollationElements *elems = ucol_openElements(coll, string.getBuffer(), string.length(), &status);
uint32_t strengthMask = 0;
int32_t order, low, high;
switch (ucol_getStrength(coll))
{
default:
strengthMask |= UCOL_TERTIARYORDERMASK;
/* fall through */
case UCOL_SECONDARY:
strengthMask |= UCOL_SECONDARYORDERMASK;
/* fall through */
case UCOL_PRIMARY:
strengthMask |= UCOL_PRIMARYORDERMASK;
}
list = new Order[listMax];
ucol_setOffset(elems, stringOffset, &status);
do {
low = ucol_getOffset(elems);
order = ucol_next(elems, &status);
high = ucol_getOffset(elems);
if (order != UCOL_NULLORDER) {
order &= strengthMask;
}
if (order != UCOL_IGNORABLE) {
add(order, low, high);
}
} while (order != UCOL_NULLORDER);
ucol_closeElements(elems);
}
static intptr_t hashChunk(const UCollator *Collator, intptr_t HashState,
const uint16_t *Str, uint32_t Length,
UErrorCode *ErrorCode) {
UCollationElements *CollationIterator = ucol_openElements(
Collator, Str, Length, ErrorCode);
while (U_SUCCESS(*ErrorCode)) {
intptr_t Elem = ucol_next(CollationIterator, ErrorCode);
// Ignore zero valued collation elements. They don't participate in the
// ordering relation.
if (Elem == 0)
continue;
if (Elem != UCOL_NULLORDER) {
Elem *= HASH_M;
Elem ^= Elem >> HASH_R;
Elem *= HASH_M;
HashState *= HASH_M;
HashState ^= Elem;
} else {
break;
// Collator.collation_order {{{
static PyObject *
icu_Collator_collation_order(icu_Collator *self, PyObject *a_) {
int32_t asz = 0;
UChar *a = NULL;
UErrorCode status = U_ZERO_ERROR;
UCollationElements *iter = NULL;
int order = 0, len = -1;
a = python_to_icu(a_, &asz);
if (a == NULL) goto end;
iter = ucol_openElements(self->collator, a, asz, &status);
if (U_FAILURE(status)) { PyErr_SetString(PyExc_ValueError, u_errorName(status)); goto end; }
order = ucol_next(iter, &status);
len = ucol_getOffset(iter);
end:
if (iter != NULL) ucol_closeElements(iter); iter = NULL;
if (a != NULL) free(a);
if (PyErr_Occurred()) return NULL;
return Py_BuildValue("ii", order, len);
} // }}}
/**
* Testing the discontigous contractions
*/
static void TestDiscontiguos() {
const char *rulestr =
"&z < AB < X\\u0300 < ABC < X\\u0300\\u0315";
UChar rule[50];
int rulelen = u_unescape(rulestr, rule, 50);
const char *src[] = {
"ADB", "ADBC", "A\\u0315B", "A\\u0315BC",
/* base character blocked */
"XD\\u0300", "XD\\u0300\\u0315",
/* non blocking combining character */
"X\\u0319\\u0300", "X\\u0319\\u0300\\u0315",
/* blocking combining character */
"X\\u0314\\u0300", "X\\u0314\\u0300\\u0315",
/* contraction prefix */
"ABDC", "AB\\u0315C","X\\u0300D\\u0315", "X\\u0300\\u0319\\u0315",
"X\\u0300\\u031A\\u0315",
/* ends not with a contraction character */
"X\\u0319\\u0300D", "X\\u0319\\u0300\\u0315D", "X\\u0300D\\u0315D",
"X\\u0300\\u0319\\u0315D", "X\\u0300\\u031A\\u0315D"
};
const char *tgt[] = {
/* non blocking combining character */
"A D B", "A D BC", "A \\u0315 B", "A \\u0315 BC",
/* base character blocked */
"X D \\u0300", "X D \\u0300\\u0315",
/* non blocking combining character */
"X\\u0300 \\u0319", "X\\u0300\\u0315 \\u0319",
/* blocking combining character */
"X \\u0314 \\u0300", "X \\u0314 \\u0300\\u0315",
/* contraction prefix */
"AB DC", "AB \\u0315 C","X\\u0300 D \\u0315", "X\\u0300\\u0315 \\u0319",
"X\\u0300 \\u031A \\u0315",
/* ends not with a contraction character */
"X\\u0300 \\u0319D", "X\\u0300\\u0315 \\u0319D", "X\\u0300 D\\u0315D",
"X\\u0300\\u0315 \\u0319D", "X\\u0300 \\u031A\\u0315D"
};
int size = 20;
UCollator *coll;
UErrorCode status = U_ZERO_ERROR;
int count = 0;
UCollationElements *iter;
UCollationElements *resultiter;
coll = ucol_openRules(rule, rulelen, UCOL_OFF, UCOL_DEFAULT_STRENGTH,NULL, &status);
iter = ucol_openElements(coll, rule, 1, &status);
resultiter = ucol_openElements(coll, rule, 1, &status);
if (U_FAILURE(status)) {
log_err_status(status, "Error opening collation rules -> %s\n", u_errorName(status));
return;
}
while (count < size) {
UChar str[20];
UChar tstr[20];
int strLen = u_unescape(src[count], str, 20);
UChar *s;
ucol_setText(iter, str, strLen, &status);
if (U_FAILURE(status)) {
log_err("Error opening collation iterator\n");
return;
}
u_unescape(tgt[count], tstr, 20);
s = tstr;
log_verbose("count %d\n", count);
for (;;) {
uint32_t ce;
UChar *e = u_strchr(s, 0x20);
if (e == 0) {
e = u_strchr(s, 0);
}
ucol_setText(resultiter, s, (int32_t)(e - s), &status);
ce = ucol_next(resultiter, &status);
if (U_FAILURE(status)) {
log_err("Error manipulating collation iterator\n");
return;
}
while (ce != UCOL_NULLORDER) {
if (ce != (uint32_t)ucol_next(iter, &status) ||
U_FAILURE(status)) {
log_err("Discontiguos contraction test mismatch\n");
return;
}
ce = ucol_next(resultiter, &status);
if (U_FAILURE(status)) {
log_err("Error getting next collation element\n");
return;
}
}
s = e + 1;
if (*e == 0) {
break;
}
}
ucol_reset(iter);
backAndForth(iter);
count ++;
}
ucol_closeElements(resultiter);
ucol_closeElements(iter);
ucol_close(coll);
}
/**
* Test for getOffset() and setOffset()
*/
static void TestOffset()
{
UErrorCode status= U_ZERO_ERROR;
UCollator *en_us=NULL;
UCollationElements *iter, *pristine;
int32_t offset;
OrderAndOffset *orders;
int32_t orderLength=0;
int count = 0;
UChar test1[50];
UChar test2[50];
u_uastrcpy(test1, "What subset of all possible test cases?");
u_uastrcpy(test2, "has the highest probability of detecting");
en_us = ucol_open("en_US", &status);
log_verbose("Testing getOffset and setOffset for collations\n");
iter = ucol_openElements(en_us, test1, u_strlen(test1), &status);
if(U_FAILURE(status)){
log_err_status(status, "ERROR: in creation of collation element iterator using ucol_openElements()\n %s\n",
myErrorName(status));
ucol_close(en_us);
return;
}
/* testing boundaries */
ucol_setOffset(iter, 0, &status);
if (U_FAILURE(status) || ucol_previous(iter, &status) != UCOL_NULLORDER) {
log_err("Error: After setting offset to 0, we should be at the end "
"of the backwards iteration");
}
ucol_setOffset(iter, u_strlen(test1), &status);
if (U_FAILURE(status) || ucol_next(iter, &status) != UCOL_NULLORDER) {
log_err("Error: After setting offset to end of the string, we should "
"be at the end of the backwards iteration");
}
/* Run all the way through the iterator, then get the offset */
orders = getOrders(iter, &orderLength);
offset = ucol_getOffset(iter);
if (offset != u_strlen(test1))
{
log_err("offset at end != length %d vs %d\n", offset,
u_strlen(test1) );
}
/* Now set the offset back to the beginning and see if it works */
pristine=ucol_openElements(en_us, test1, u_strlen(test1), &status);
if(U_FAILURE(status)){
log_err("ERROR: in creation of collation element iterator using ucol_openElements()\n %s\n",
myErrorName(status));
ucol_close(en_us);
return;
}
status = U_ZERO_ERROR;
ucol_setOffset(iter, 0, &status);
if (U_FAILURE(status))
{
log_err("setOffset failed. %s\n", myErrorName(status));
}
else
{
assertEqual(iter, pristine);
}
ucol_closeElements(pristine);
ucol_closeElements(iter);
free(orders);
/* testing offsets in normalization buffer */
test1[0] = 0x61;
test1[1] = 0x300;
test1[2] = 0x316;
test1[3] = 0x62;
test1[4] = 0;
ucol_setAttribute(en_us, UCOL_NORMALIZATION_MODE, UCOL_ON, &status);
iter = ucol_openElements(en_us, test1, 4, &status);
if(U_FAILURE(status)){
log_err("ERROR: in creation of collation element iterator using ucol_openElements()\n %s\n",
myErrorName(status));
ucol_close(en_us);
return;
}
count = 0;
while (ucol_next(iter, &status) != UCOL_NULLORDER &&
U_SUCCESS(status)) {
switch (count) {
case 0:
if (ucol_getOffset(iter) != 1) {
log_err("ERROR: Offset of iteration should be 1\n");
}
break;
case 3:
if (ucol_getOffset(iter) != 4) {
log_err("ERROR: Offset of iteration should be 4\n");
}
break;
default:
if (ucol_getOffset(iter) != 3) {
log_err("ERROR: Offset of iteration should be 3\n");
}
}
count ++;
}
ucol_reset(iter);
count = 0;
while (ucol_previous(iter, &status) != UCOL_NULLORDER &&
U_SUCCESS(status)) {
switch (count) {
case 0:
case 1:
if (ucol_getOffset(iter) != 3) {
log_err("ERROR: Offset of iteration should be 3\n");
}
break;
case 2:
if (ucol_getOffset(iter) != 1) {
log_err("ERROR: Offset of iteration should be 1\n");
}
break;
default:
if (ucol_getOffset(iter) != 0) {
log_err("ERROR: Offset of iteration should be 0\n");
}
}
count ++;
}
if(U_FAILURE(status)){
log_err("ERROR: in iterating collation elements %s\n",
myErrorName(status));
}
ucol_closeElements(iter);
ucol_close(en_us);
}
static void TestBug672() {
UErrorCode status = U_ZERO_ERROR;
UChar pattern[20];
UChar text[50];
int i;
int result[3][3];
u_uastrcpy(pattern, "resume");
u_uastrcpy(text, "Time to resume updating my resume.");
for (i = 0; i < 3; ++ i) {
UCollator *coll = ucol_open(LOCALES[i], &status);
UCollationElements *pitr = ucol_openElements(coll, pattern, -1,
&status);
UCollationElements *titer = ucol_openElements(coll, text, -1,
&status);
if (U_FAILURE(status)) {
log_err_status(status, "ERROR: in creation of either the collator or the collation iterator :%s\n",
myErrorName(status));
return;
}
log_verbose("locale tested %s\n", LOCALES[i]);
while (ucol_next(pitr, &status) != UCOL_NULLORDER &&
U_SUCCESS(status)) {
}
if (U_FAILURE(status)) {
log_err("ERROR: reversing collation iterator :%s\n",
myErrorName(status));
return;
}
ucol_reset(pitr);
ucol_setOffset(titer, u_strlen(pattern), &status);
if (U_FAILURE(status)) {
log_err("ERROR: setting offset in collator :%s\n",
myErrorName(status));
return;
}
result[i][0] = ucol_getOffset(titer);
log_verbose("Text iterator set to offset %d\n", result[i][0]);
/* Use previous() */
ucol_previous(titer, &status);
result[i][1] = ucol_getOffset(titer);
log_verbose("Current offset %d after previous\n", result[i][1]);
/* Add one to index */
log_verbose("Adding one to current offset...\n");
ucol_setOffset(titer, ucol_getOffset(titer) + 1, &status);
if (U_FAILURE(status)) {
log_err("ERROR: setting offset in collator :%s\n",
myErrorName(status));
return;
}
result[i][2] = ucol_getOffset(titer);
log_verbose("Current offset in text = %d\n", result[i][2]);
ucol_closeElements(pitr);
ucol_closeElements(titer);
ucol_close(coll);
}
if (uprv_memcmp(result[0], result[1], 3) != 0 ||
uprv_memcmp(result[1], result[2], 3) != 0) {
log_err("ERROR: Different locales have different offsets at the same character\n");
}
}
/**
* Get the ordering priority of the next character in the string.
* @return the next character's ordering. Returns NULLORDER if an error has
* occured or if the end of string has been reached
*/
int32_t CollationElementIterator::next(UErrorCode& status)
{
return ucol_next(m_data_, &status);
}
static void TestJB1401(void)
{
UCollator *myCollator = 0;
UErrorCode status = U_ZERO_ERROR;
static UChar NFD_UnsafeStartChars[] = {
0x0f73, /* Tibetan Vowel Sign II */
0x0f75, /* Tibetan Vowel Sign UU */
0x0f81, /* Tibetan Vowel Sign Reversed II */
0
};
int i;
myCollator = ucol_open("en_US", &status);
if (U_FAILURE(status)){
int32_t bufferLen = 0;
UChar dispName [100];
bufferLen = uloc_getDisplayName("en_US", 0, dispName, 100, &status);
/*Report the error with display name... */
log_err("ERROR: Failed to create the collator for : \"%s\"\n", dispName);
return;
}
ucol_setAttribute(myCollator, UCOL_NORMALIZATION_MODE, UCOL_ON, &status);
if (U_FAILURE(status)){
log_err("ERROR: Failed to set normalization mode ON for collator.\n");
return;
}
for (i=0; ; i++) {
UChar c;
UChar X[4];
UChar Y[20];
UChar Z[20];
/* Get the next funny character to be tested, and set up the
* three test strings X, Y, Z, consisting of an A-grave + test char,
* in original form, NFD, and then NFC form.
*/
c = NFD_UnsafeStartChars[i];
if (c==0) {break;}
X[0]=0xC0; X[1]=c; X[2]=0; /* \u00C0 is A Grave*/
unorm_normalize(X, -1, UNORM_NFD, 0, Y, 20, &status);
unorm_normalize(Y, -1, UNORM_NFC, 0, Z, 20, &status);
if (U_FAILURE(status)){
log_err("ERROR: Failed to normalize test of character %x\n", c);
return;
}
/* Collation test. All three strings should be equal.
* doTest does both strcoll and sort keys, with params in both orders.
*/
doTest(myCollator, X, Y, UCOL_EQUAL);
doTest(myCollator, X, Z, UCOL_EQUAL);
doTest(myCollator, Y, Z, UCOL_EQUAL);
/* Run collation element iterators over the three strings. Results should be same for each.
*/
{
UCollationElements *ceiX, *ceiY, *ceiZ;
int32_t ceX, ceY, ceZ;
int j;
ceiX = ucol_openElements(myCollator, X, -1, &status);
ceiY = ucol_openElements(myCollator, Y, -1, &status);
ceiZ = ucol_openElements(myCollator, Z, -1, &status);
if (U_FAILURE(status)) {
log_err("ERROR: uucol_openElements failed.\n");
return;
}
for (j=0;; j++) {
ceX = ucol_next(ceiX, &status);
ceY = ucol_next(ceiY, &status);
ceZ = ucol_next(ceiZ, &status);
if (U_FAILURE(status)) {
log_err("ERROR: ucol_next failed for iteration #%d.\n", j);
break;
}
if (ceX != ceY || ceY != ceZ) {
log_err("ERROR: ucol_next failed for iteration #%d.\n", j);
break;
}
if (ceX == UCOL_NULLORDER) {
break;
}
}
ucol_closeElements(ceiX);
ucol_closeElements(ceiY);
ucol_closeElements(ceiZ);
}
}
ucol_close(myCollator);
}
void
backAndForth(UCollationElements *iter)
{
/* Run through the iterator forwards and stick it into an array */
int32_t index, o;
UErrorCode status = U_ZERO_ERROR;
int32_t orderLength = 0;
int32_t *orders;
orders= getOrders(iter, &orderLength);
/* Now go through it backwards and make sure we get the same values */
index = orderLength;
ucol_reset(iter);
/* synwee : changed */
while ((o = ucol_previous(iter, &status)) != UCOL_NULLORDER)
{
if (o != orders[-- index])
{
if (o == 0)
index ++;
else
{
while (index > 0 && orders[-- index] == 0)
{
}
if (o != orders[index])
{
log_err("Mismatch at index : 0x%x\n", index);
return;
}
}
}
}
while (index != 0 && orders[index - 1] == 0) {
index --;
}
if (index != 0)
{
log_err("Didn't get back to beginning - index is %d\n", index);
ucol_reset(iter);
log_err("\nnext: ");
if ((o = ucol_next(iter, &status)) != UCOL_NULLORDER)
{
log_err("Error at %x\n", o);
}
log_err("\nprev: ");
if ((o = ucol_previous(iter, &status)) != UCOL_NULLORDER)
{
log_err("Error at %x\n", o);
}
log_verbose("\n");
}
free(orders);
}
/**
* Test for setText()
*/
static void TestSetText()
{
int32_t c,i;
UErrorCode status = U_ZERO_ERROR;
UCollator *en_us=NULL;
UCollationElements *iter1, *iter2;
UChar test1[50];
UChar test2[50];
u_uastrcpy(test1, "What subset of all possible test cases?");
u_uastrcpy(test2, "has the highest probability of detecting");
en_us = ucol_open("en_US", &status);
log_verbose("testing setText for Collation elements\n");
iter1=ucol_openElements(en_us, test1, u_strlen(test1), &status);
if(U_FAILURE(status)){
log_err_status(status, "ERROR: in creation of collation element iterator1 using ucol_openElements()\n %s\n",
myErrorName(status));
ucol_close(en_us);
return;
}
iter2=ucol_openElements(en_us, test2, u_strlen(test2), &status);
if(U_FAILURE(status)){
log_err("ERROR: in creation of collation element iterator2 using ucol_openElements()\n %s\n",
myErrorName(status));
ucol_close(en_us);
return;
}
/* Run through the second iterator just to exercise it */
c = ucol_next(iter2, &status);
i = 0;
while ( ++i < 10 && (c != UCOL_NULLORDER))
{
if (U_FAILURE(status))
{
log_err("iter2->next() returned an error. %s\n", myErrorName(status));
ucol_closeElements(iter2);
ucol_closeElements(iter1);
ucol_close(en_us);
return;
}
c = ucol_next(iter2, &status);
}
/* Now set it to point to the same string as the first iterator */
ucol_setText(iter2, test1, u_strlen(test1), &status);
if (U_FAILURE(status))
{
log_err("call to iter2->setText(test1) failed. %s\n", myErrorName(status));
}
else
{
assertEqual(iter1, iter2);
}
/* Now set it to point to a null string with fake length*/
ucol_setText(iter2, NULL, 2, &status);
if (status != U_ILLEGAL_ARGUMENT_ERROR)
{
log_err("call to iter2->setText(null, 2) should yield an illegal-argument-error - %s\n",
myErrorName(status));
}
ucol_closeElements(iter2);
ucol_closeElements(iter1);
ucol_close(en_us);
}
/**
* Testing iterators with extremely small buffers
*/
static void TestSmallBuffer()
{
UErrorCode status = U_ZERO_ERROR;
UCollator *coll;
UCollationElements *testiter,
*iter;
int32_t count = 0;
OrderAndOffset *testorders,
*orders;
UChar teststr[500];
UChar str[] = {0x300, 0x31A, 0};
/*
creating a long string of decomposable characters,
since by default the writable buffer is of size 256
*/
while (count < 500) {
if ((count & 1) == 0) {
teststr[count ++] = 0x300;
}
else {
teststr[count ++] = 0x31A;
}
}
coll = ucol_open("th_TH", &status);
if(U_SUCCESS(status) && coll) {
testiter = ucol_openElements(coll, teststr, 500, &status);
iter = ucol_openElements(coll, str, 2, &status);
orders = getOrders(iter, &count);
if (count != 2) {
log_err("Error collation elements size is not 2 for \\u0300\\u031A\n");
}
/*
this will rearrange the string data to 250 characters of 0x300 first then
250 characters of 0x031A
*/
testorders = getOrders(testiter, &count);
if (count != 500) {
log_err("Error decomposition does not give the right sized collation elements\n");
}
while (count != 0) {
/* UCA collation element for 0x0F76 */
if ((count > 250 && testorders[-- count].order != orders[1].order) ||
(count <= 250 && testorders[-- count].order != orders[0].order)) {
log_err("Error decomposition does not give the right collation element at %d count\n", count);
break;
}
}
free(testorders);
free(orders);
ucol_reset(testiter);
/* ensures closing of elements done properly to clear writable buffer */
ucol_next(testiter, &status);
ucol_next(testiter, &status);
ucol_closeElements(testiter);
ucol_closeElements(iter);
ucol_close(coll);
} else {
log_err_status(status, "Couldn't open collator -> %s\n", u_errorName(status));
}
}
static jint NativeCollation_next(JNIEnv* env, jclass, jint address) {
UErrorCode status = U_ZERO_ERROR;
jint result = ucol_next(toCollationElements(address), &status);
icu4jni_error(env, status);
return result;
}
static jint NativeCollation_next(JNIEnv* env, jclass, jlong address) {
UErrorCode status = U_ZERO_ERROR;
jint result = ucol_next(toCollationElements(address), &status);
maybeThrowIcuException(env, "ucol_next", status);
return result;
}
void
backAndForth(UCollationElements *iter)
{
/* Run through the iterator forwards and stick it into an array */
int32_t idx, o;
UErrorCode status = U_ZERO_ERROR;
int32_t orderLength = 0;
OrderAndOffset *orders = getOrders(iter, &orderLength);
/* Now go through it backwards and make sure we get the same values */
idx = orderLength;
ucol_reset(iter);
/* synwee : changed */
while ((o = ucol_previous(iter, &status)) != UCOL_NULLORDER) {
#if TEST_OFFSETS
int32_t offset =
#endif
ucol_getOffset(iter);
idx -= 1;
if (o != orders[idx].order) {
if (o == 0)
idx ++;
else {
while (idx > 0 && orders[-- idx].order == 0) {
/* nothing... */
}
if (o != orders[idx].order) {
log_err("Mismatched order at index %d: 0x%8.8X vs. 0x%8.8X\n", idx,
orders[idx].order, o);
goto bail;
}
}
}
#if TEST_OFFSETS
if (offset != orders[idx].offset) {
log_err("Mismatched offset at index %d: %d vs. %d\n", idx,
orders[idx].offset, offset);
goto bail;
}
#endif
}
while (idx != 0 && orders[idx - 1].order == 0) {
idx -= 1;
}
if (idx != 0) {
log_err("Didn't get back to beginning - index is %d\n", idx);
ucol_reset(iter);
log_err("\nnext: ");
if ((o = ucol_next(iter, &status)) != UCOL_NULLORDER) {
log_err("Error at %x\n", o);
}
log_err("\nprev: ");
if ((o = ucol_previous(iter, &status)) != UCOL_NULLORDER) {
log_err("Error at %x\n", o);
}
log_verbose("\n");
}
bail:
free(orders);
}
static void TestSearchCollatorElements(void)
{
const TSCEItem * tsceItemPtr;
for (tsceItemPtr = tsceItems; tsceItemPtr->locale != NULL; tsceItemPtr++) {
UErrorCode status = U_ZERO_ERROR;
UCollator* ucol = ucol_open(tsceItemPtr->locale, &status);
if ( U_SUCCESS(status) ) {
UCollationElements * uce = ucol_openElements(ucol, tsceText, kLen_tsceText, &status);
if ( U_SUCCESS(status) ) {
int32_t offset, element;
const int32_t * nextOffsetPtr;
const int32_t * limitOffsetPtr;
nextOffsetPtr = tsceItemPtr->offsets;
limitOffsetPtr = tsceItemPtr->offsets + tsceItemPtr->offsetsLen;
do {
offset = ucol_getOffset(uce);
element = ucol_next(uce, &status);
log_verbose("(%s) offset=%2d ce=%08x\n", tsceItemPtr->locale, offset, element);
if ( element == 0 ) {
log_err("error, locale %s, ucol_next returned element 0\n", tsceItemPtr->locale );
}
if ( nextOffsetPtr < limitOffsetPtr ) {
if (offset != *nextOffsetPtr) {
log_err("error, locale %s, expected ucol_next -> ucol_getOffset %d, got %d\n",
tsceItemPtr->locale, *nextOffsetPtr, offset );
nextOffsetPtr = limitOffsetPtr;
break;
}
nextOffsetPtr++;
} else {
log_err("error, locale %s, ucol_next returned more elements than expected\n", tsceItemPtr->locale );
}
} while ( U_SUCCESS(status) && element != UCOL_NULLORDER );
if ( nextOffsetPtr < limitOffsetPtr ) {
log_err("error, locale %s, ucol_next returned fewer elements than expected\n", tsceItemPtr->locale );
}
ucol_setOffset(uce, kLen_tsceText, &status);
status = U_ZERO_ERROR;
nextOffsetPtr = tsceItemPtr->offsets + tsceItemPtr->offsetsLen;
limitOffsetPtr = tsceItemPtr->offsets;
do {
offset = ucol_getOffset(uce);
element = ucol_previous(uce, &status);
if ( element == 0 ) {
log_err("error, locale %s, ucol_previous returned element 0\n", tsceItemPtr->locale );
}
if ( nextOffsetPtr > limitOffsetPtr ) {
nextOffsetPtr--;
if (offset != *nextOffsetPtr) {
log_err("error, locale %s, expected ucol_previous -> ucol_getOffset %d, got %d\n",
tsceItemPtr->locale, *nextOffsetPtr, offset );
nextOffsetPtr = limitOffsetPtr;
break;
}
} else {
log_err("error, locale %s, ucol_previous returned more elements than expected\n", tsceItemPtr->locale );
}
} while ( U_SUCCESS(status) && element != UCOL_NULLORDER );
if ( nextOffsetPtr > limitOffsetPtr ) {
log_err("error, locale %s, ucol_previous returned fewer elements than expected\n", tsceItemPtr->locale );
}
ucol_closeElements(uce);
} else {
log_err("error, locale %s, ucol_openElements failed: %s\n", tsceItemPtr->locale, u_errorName(status) );
}
ucol_close(ucol);
} else {
log_data_err("error, locale %s, ucol_open failed: %s\n", tsceItemPtr->locale, u_errorName(status) );
}
}
}
