/*
*******************************************************************************
*
-* Copyright (C) 2002-2005, International Business Machines
+* Copyright (C) 2002-2013, International Business Machines
* Corporation and others. All Rights Reserved.
*
*******************************************************************************
#include "unicode/ubrk.h"
#include "unicode/unistr.h"
#include "unicode/ucasemap.h"
+#include "ucase.h"
#include "ustrtest.h"
#include "unicode/tstdtmod.h"
+#define LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0]))
+
StringCaseTest::~StringCaseTest() {}
void
StringCaseTest::runIndexedTest(int32_t index, UBool exec, const char *&name, char * /*par*/) {
- if (exec) logln("TestSuite StringCaseTest: ");
- switch (index) {
- case 0: name = "TestCaseConversion"; if (exec) TestCaseConversion(); break;
- case 1:
- name = "TestCasing";
-#if !UCONFIG_NO_BREAK_ITERATION
- if(exec) TestCasing();
-#endif
- break;
-
- default: name = ""; break; //needed to end loop
+ if(exec) {
+ logln("TestSuite StringCaseTest: ");
}
+ TESTCASE_AUTO_BEGIN;
+ TESTCASE_AUTO(TestCaseConversion);
+#if !UCONFIG_NO_BREAK_ITERATION && !UCONFIG_NO_FILE_IO && !UCONFIG_NO_LEGACY_CONVERSION
+ TESTCASE_AUTO(TestCasing);
+#endif
+ TESTCASE_AUTO(TestFullCaseFoldingIterator);
+ TESTCASE_AUTO_END;
}
void
enum {
TEST_LOWER,
TEST_UPPER,
-#if !UCONFIG_NO_BREAK_ITERATION
TEST_TITLE,
-#endif
+ TEST_FOLD,
TEST_COUNT
};
static const char *const dataNames[TEST_COUNT+1]={
"lowercasing",
"uppercasing",
-#if !UCONFIG_NO_BREAK_ITERATION
"titlecasing",
-#endif
+ "casefolding",
""
};
StringCaseTest::TestCasingImpl(const UnicodeString &input,
const UnicodeString &output,
int32_t whichCase,
- const char *localeID, uint32_t options) {
+ void *iter, const char *localeID, uint32_t options) {
// UnicodeString
UnicodeString result;
const char *name;
+ Locale locale(localeID);
result=input;
switch(whichCase) {
case TEST_LOWER:
name="toLower";
- result.toLower(Locale(localeID));
+ result.toLower(locale);
break;
case TEST_UPPER:
name="toUpper";
- result.toUpper(Locale(localeID));
+ result.toUpper(locale);
+ break;
+#if !UCONFIG_NO_BREAK_ITERATION
+ case TEST_TITLE:
+ name="toTitle";
+ result.toTitle((BreakIterator *)iter, locale, options);
+ break;
+#endif
+ case TEST_FOLD:
+ name="foldCase";
+ result.foldCase(options);
break;
default:
name="";
break; // won't happen
}
if(result!=output) {
- errln("error: UnicodeString.%s() got a wrong result for a test case from casing.res", name);
+ dataerrln("error: UnicodeString.%s() got a wrong result for a test case from casing.res", name);
+ }
+#if !UCONFIG_NO_BREAK_ITERATION
+ if(whichCase==TEST_TITLE && options==0) {
+ result=input;
+ result.toTitle((BreakIterator *)iter, locale);
+ if(result!=output) {
+ dataerrln("error: UnicodeString.toTitle(options=0) got a wrong result for a test case from casing.res");
+ }
}
+#endif
// UTF-8
char utf8In[100], utf8Out[100];
int32_t utf8InLength, utf8OutLength, resultLength;
UChar *buffer;
- UCaseMap *csm;
- UErrorCode errorCode;
-
- errorCode=U_ZERO_ERROR;
- csm=ucasemap_open(localeID, options, &errorCode);
+ IcuTestErrorCode errorCode(*this, "TestCasingImpl");
+ LocalUCaseMapPointer csm(ucasemap_open(localeID, options, errorCode));
+#if !UCONFIG_NO_BREAK_ITERATION
+ if(iter!=NULL) {
+ // Clone the break iterator so that the UCaseMap can safely adopt it.
+ UBreakIterator *clone=ubrk_safeClone((UBreakIterator *)iter, NULL, NULL, errorCode);
+ ucasemap_setBreakIterator(csm.getAlias(), clone, errorCode);
+ }
+#endif
- u_strToUTF8(utf8In, (int32_t)sizeof(utf8In), &utf8InLength, input.getBuffer(), input.length(), &errorCode);
+ u_strToUTF8(utf8In, (int32_t)sizeof(utf8In), &utf8InLength, input.getBuffer(), input.length(), errorCode);
switch(whichCase) {
case TEST_LOWER:
name="ucasemap_utf8ToLower";
- utf8OutLength=ucasemap_utf8ToLower(csm,
+ utf8OutLength=ucasemap_utf8ToLower(csm.getAlias(),
utf8Out, (int32_t)sizeof(utf8Out),
- utf8In, utf8InLength, &errorCode);
+ utf8In, utf8InLength, errorCode);
break;
case TEST_UPPER:
name="ucasemap_utf8ToUpper";
- utf8OutLength=ucasemap_utf8ToUpper(csm,
+ utf8OutLength=ucasemap_utf8ToUpper(csm.getAlias(),
+ utf8Out, (int32_t)sizeof(utf8Out),
+ utf8In, utf8InLength, errorCode);
+ break;
+#if !UCONFIG_NO_BREAK_ITERATION
+ case TEST_TITLE:
+ name="ucasemap_utf8ToTitle";
+ utf8OutLength=ucasemap_utf8ToTitle(csm.getAlias(),
utf8Out, (int32_t)sizeof(utf8Out),
- utf8In, utf8InLength, &errorCode);
+ utf8In, utf8InLength, errorCode);
+ break;
+#endif
+ case TEST_FOLD:
+ name="ucasemap_utf8FoldCase";
+ utf8OutLength=ucasemap_utf8FoldCase(csm.getAlias(),
+ utf8Out, (int32_t)sizeof(utf8Out),
+ utf8In, utf8InLength, errorCode);
break;
default:
name="";
break; // won't happen
}
buffer=result.getBuffer(utf8OutLength);
- u_strFromUTF8(buffer, result.getCapacity(), &resultLength, utf8Out, utf8OutLength, &errorCode);
- result.releaseBuffer(U_SUCCESS(errorCode) ? resultLength : 0);
+ u_strFromUTF8(buffer, result.getCapacity(), &resultLength, utf8Out, utf8OutLength, errorCode);
+ result.releaseBuffer(errorCode.isSuccess() ? resultLength : 0);
- if(U_FAILURE(errorCode)) {
- errln("error: %s() got an error for a test case from casing.res - %s", name, u_errorName(errorCode));
+ if(errorCode.isFailure()) {
+ errcheckln(errorCode, "error: %s() got an error for a test case from casing.res - %s", name, u_errorName(errorCode));
+ errorCode.reset();
} else if(result!=output) {
errln("error: %s() got a wrong result for a test case from casing.res", name);
- }
- ucasemap_close(csm);
-}
-
-#if !UCONFIG_NO_BREAK_ITERATION
-
-void
-StringCaseTest::TestTitleCasing(const UnicodeString &input,
- const UnicodeString &output,
- const char *localeID,
- UBreakIterator *iter) {
- UnicodeString result;
-
- result=input;
- result.toTitle((BreakIterator *)iter, Locale(localeID));
- if(result!=output) {
- errln("error: UnicodeString.toTitle() got a wrong result for a test case from casing.res");
+ errln("expected \"" + output + "\" got \"" + result + "\"" );
}
}
-#endif
-
void
StringCaseTest::TestCasing() {
UErrorCode status = U_ZERO_ERROR;
#if !UCONFIG_NO_BREAK_ITERATION
- UBreakIterator *iter;
+ LocalUBreakIteratorPointer iter;
#endif
char cLocaleID[100];
- UnicodeString locale, input, output, result;
+ UnicodeString locale, input, output, optionsString, result;
+ uint32_t options;
int32_t whichCase, type;
- TestDataModule *driver = TestDataModule::getTestDataModule("casing", *this, status);
+ LocalPointer<TestDataModule> driver(TestDataModule::getTestDataModule("casing", *this, status));
if(U_SUCCESS(status)) {
for(whichCase=0; whichCase<TEST_COUNT; ++whichCase) {
- TestData *casingTest = driver->createTestData(dataNames[whichCase], status);
+#if UCONFIG_NO_BREAK_ITERATION
+ if(whichCase==TEST_TITLE) {
+ continue;
+ }
+#endif
+ LocalPointer<TestData> casingTest(driver->createTestData(dataNames[whichCase], status));
if(U_FAILURE(status)) {
errln("TestCasing failed to createTestData(%s) - %s", dataNames[whichCase], u_errorName(status));
break;
}
const DataMap *myCase = NULL;
while(casingTest->nextCase(myCase, status)) {
- locale = myCase->getString("Locale", status);
- locale.extract(0, 0x7fffffff, cLocaleID, sizeof(cLocaleID), "");
-
input = myCase->getString("Input", status);
output = myCase->getString("Output", status);
+ if(whichCase!=TEST_FOLD) {
+ locale = myCase->getString("Locale", status);
+ }
+ locale.extract(0, 0x7fffffff, cLocaleID, sizeof(cLocaleID), "");
+
#if !UCONFIG_NO_BREAK_ITERATION
- iter=NULL;
if(whichCase==TEST_TITLE) {
type = myCase->getInt("Type", status);
if(type>=0) {
- iter=ubrk_open((UBreakIteratorType)type, cLocaleID, NULL, 0, &status);
+ iter.adoptInstead(ubrk_open((UBreakIteratorType)type, cLocaleID, NULL, 0, &status));
+ } else if(type==-2) {
+ // Open a trivial break iterator that only delivers { 0, length }
+ // or even just { 0 } as boundaries.
+ static const UChar rules[] = { 0x2e, 0x2a, 0x3b }; // ".*;"
+ UParseError parseError;
+ iter.adoptInstead(ubrk_openRules(rules, LENGTHOF(rules), NULL, 0, &parseError, &status));
}
}
#endif
+ options = 0;
+ if(whichCase==TEST_TITLE || whichCase==TEST_FOLD) {
+ optionsString = myCase->getString("Options", status);
+ if(optionsString.indexOf((UChar)0x54)>=0) { // T
+ options|=U_FOLD_CASE_EXCLUDE_SPECIAL_I;
+ }
+ if(optionsString.indexOf((UChar)0x4c)>=0) { // L
+ options|=U_TITLECASE_NO_LOWERCASE;
+ }
+ if(optionsString.indexOf((UChar)0x41)>=0) { // A
+ options|=U_TITLECASE_NO_BREAK_ADJUSTMENT;
+ }
+ }
if(U_FAILURE(status)) {
- errln("error: TestCasing() setup failed for %s test case from casing.res: %s", dataNames[whichCase], u_errorName(status));
+ dataerrln("error: TestCasing() setup failed for %s test case from casing.res: %s", dataNames[whichCase], u_errorName(status));
status = U_ZERO_ERROR;
} else {
- switch(whichCase) {
- case TEST_LOWER:
- case TEST_UPPER:
- TestCasingImpl(input, output, whichCase, cLocaleID, 0);
- break;
-#if !UCONFIG_NO_BREAK_ITERATION
- case TEST_TITLE:
- TestTitleCasing(input, output, cLocaleID, iter);
- break;
+#if UCONFIG_NO_BREAK_ITERATION
+ LocalPointer<UMemory> iter;
#endif
- default:
- break; // won't happen
- }
+ TestCasingImpl(input, output, whichCase, iter.getAlias(), cLocaleID, options);
}
#if !UCONFIG_NO_BREAK_ITERATION
- if(iter!=NULL) {
- ubrk_close(iter);
- }
+ iter.adoptInstead(NULL);
#endif
}
- delete casingTest;
}
}
- delete driver;
#if !UCONFIG_NO_BREAK_ITERATION
// more tests for API coverage
input=UNICODE_STRING_SIMPLE("sTrA\\u00dfE").unescape();
(result=input).toTitle(NULL);
if(result!=UNICODE_STRING_SIMPLE("Stra\\u00dfe").unescape()) {
- errln("UnicodeString::toTitle(NULL) failed");
+ dataerrln("UnicodeString::toTitle(NULL) failed.");
}
#endif
}
+
+void
+StringCaseTest::TestFullCaseFoldingIterator() {
+ UnicodeString ffi=UNICODE_STRING_SIMPLE("ffi");
+ UnicodeString ss=UNICODE_STRING_SIMPLE("ss");
+ FullCaseFoldingIterator iter;
+ int32_t count=0;
+ int32_t countSpecific=0;
+ UChar32 c;
+ UnicodeString full;
+ while((c=iter.next(full))>=0) {
+ ++count;
+ // Check that the full Case_Folding has more than 1 code point.
+ if(!full.hasMoreChar32Than(0, 0x7fffffff, 1)) {
+ errln("error: FullCaseFoldingIterator.next()=U+%04lX full Case_Folding has at most 1 code point", (long)c);
+ continue;
+ }
+ // Check that full == Case_Folding(c).
+ UnicodeString cf(c);
+ cf.foldCase();
+ if(full!=cf) {
+ errln("error: FullCaseFoldingIterator.next()=U+%04lX full Case_Folding != cf(c)", (long)c);
+ continue;
+ }
+ // Spot-check a couple of specific cases.
+ if((full==ffi && c==0xfb03) || (full==ss && (c==0xdf || c==0x1e9e))) {
+ ++countSpecific;
+ }
+ }
+ if(countSpecific!=3) {
+ errln("error: FullCaseFoldingIterator did not yield exactly the expected specific cases");
+ }
+ if(count<70) {
+ errln("error: FullCaseFoldingIterator yielded only %d (cp, full) pairs", (int)count);
+ }
+}