#include "unicode/utypes.h"
#if !UCONFIG_NO_COLLATION
#include "thcoll.h"
#include "unicode/utypes.h"
#include "unicode/coll.h"
#include "unicode/localpointer.h"
#include "unicode/sortkey.h"
#include "unicode/tblcoll.h"
#include "unicode/ustring.h"
#include "cstring.h"
#include "filestrm.h"
#include "textfile.h"
#define MAX_FAILURES_TO_SHOW -1
CollationThaiTest::CollationThaiTest() {
UErrorCode status = U_ZERO_ERROR;
coll = Collator::createInstance(Locale("th", "TH", ""), status);
if (coll && U_SUCCESS(status)) {
} else {
delete coll;
coll = 0;
}
}
CollationThaiTest::~CollationThaiTest() {
delete coll;
}
void CollationThaiTest::runIndexedTest(int32_t index, UBool exec, const char* &name,
char* ) {
if((!coll) && exec) {
dataerrln(__FILE__ " cannot test - failed to create collator.");
name = "some test";
return;
}
switch (index) {
TESTCASE(0,TestDictionary);
TESTCASE(1,TestCornerCases);
TESTCASE(2,TestNamesList);
TESTCASE(3,TestInvalidThai);
TESTCASE(4,TestReordering);
default: name = ""; break;
}
}
void CollationThaiTest::TestNamesList(void) {
if (coll == 0) {
errln("Error: could not construct Thai collator");
return;
}
UErrorCode ec = U_ZERO_ERROR;
TextFile names("TestNames_Thai.txt", "UTF16LE", ec);
if (U_FAILURE(ec)) {
logln("Can't open TestNames_Thai.txt: %s; skipping test",
u_errorName(ec));
return;
}
UnicodeString lastWord, word;
int32_t wordCount = 0;
while (names.readLineSkippingComments(word, ec, FALSE) && U_SUCCESS(ec)) {
++wordCount;
if (wordCount <= 8) {
UnicodeString str;
logln((UnicodeString)"Word " + wordCount + ": " + IntlTest::prettify(word, str));
}
if (lastWord.length() > 0) {
Collator::EComparisonResult result = coll->compare(lastWord, word);
doTest(coll, lastWord, word, result);
}
lastWord = word;
}
assertSuccess("readLine", ec);
logln((UnicodeString)"Words checked: " + wordCount);
}
void CollationThaiTest::TestDictionary(void) {
if (coll == 0) {
errln("Error: could not construct Thai collator");
return;
}
UErrorCode ec = U_ZERO_ERROR;
TextFile riwords("riwords.txt", "UTF8", ec);
if (U_FAILURE(ec)) {
logln("Can't open riwords.txt: %s; skipping test",
u_errorName(ec));
return;
}
UnicodeString lastWord, word;
int32_t failed = 0;
int32_t wordCount = 0;
while (riwords.readLineSkippingComments(word, ec, FALSE) && U_SUCCESS(ec)) {
++wordCount;
if (wordCount <= 8) {
UnicodeString str;
logln((UnicodeString)"Word " + wordCount + ": " + IntlTest::prettify(word, str));
}
if (lastWord.length() > 0) {
int32_t result = coll->compare(lastWord, word);
if (result > 0) {
failed++;
if (MAX_FAILURES_TO_SHOW < 0 || failed <= MAX_FAILURES_TO_SHOW) {
UnicodeString str;
UnicodeString msg =
UnicodeString("--------------------------------------------\n")
+ riwords.getLineNumber()
+ " compare(" + IntlTest::prettify(lastWord, str);
msg += UnicodeString(", ")
+ IntlTest::prettify(word, str) + ") returned " + result
+ ", expected -1\n";
UErrorCode status = U_ZERO_ERROR;
CollationKey k1, k2;
coll->getCollationKey(lastWord, k1, status);
coll->getCollationKey(word, k2, status);
if (U_FAILURE(status)) {
errln((UnicodeString)"Fail: getCollationKey returned " + u_errorName(status));
return;
}
msg.append("key1: ").append(prettify(k1, str)).append("\n");
msg.append("key2: ").append(prettify(k2, str));
errln(msg);
}
}
}
lastWord = word;
}
assertSuccess("readLine", ec);
if (failed != 0) {
if (failed > MAX_FAILURES_TO_SHOW) {
errln((UnicodeString)"Too many failures; only the first " +
MAX_FAILURES_TO_SHOW + " failures were shown");
}
errln((UnicodeString)"Summary: " + failed + " of " + (riwords.getLineNumber() - 1) +
" comparisons failed");
}
logln((UnicodeString)"Words checked: " + wordCount);
}
void CollationThaiTest::TestCornerCases(void) {
const char* TESTS[] = {
"\\u0e01", "<", "\\u0e01\\u0e01",
"\\u0e01\\u0e32", "<", "\\u0e01\\u0e49\\u0e32",
"\\u0e01\\u0e32", "<", "\\u0e01\\u0e32\\u0e4c",
"\\u0e01\\u0e32\\u0e01\\u0e49\\u0e32", "<", "\\u0e01\\u0e48\\u0e32\\u0e01\\u0e49\\u0e32",
"\\u0e01\\u0e32", "=", "\\u0e01\\u0e32-",
"\\u0e01\\u0e32-", "<", "\\u0e01\\u0e32\\u0e01\\u0e32",
"\\u0e01\\u0e32", "=", "\\u0e01\\u0e32\\u0e46",
"\\u0e01\\u0e32\\u0e46", "<", "\\u0e01\\u0e32\\u0e01\\u0e32",
"\\u0e24\\u0e29\\u0e35", "<", "\\u0e24\\u0e45\\u0e29\\u0e35",
"\\u0e26\\u0e29\\u0e35", "<", "\\u0e26\\u0e45\\u0e29\\u0e35",
"\\u0e40\\u0e01\\u0e2d", "<", "\\u0e40\\u0e01\\u0e34",
"\\u0e01\\u0e32\\u0e01\\u0e48\\u0e32", "<", "\\u0e01\\u0e49\\u0e32\\u0e01\\u0e32",
"\\u0e01.\\u0e01.", "<", "\\u0e01\\u0e32",
};
const int32_t TESTS_length = (int32_t)(sizeof(TESTS)/sizeof(TESTS[0]));
if (coll == 0) {
errln("Error: could not construct Thai collator");
return;
}
compareArray(*coll, TESTS, TESTS_length);
}
void CollationThaiTest::compareArray(Collator& c, const char* tests[],
int32_t testsLength) {
for (int32_t i = 0; i < testsLength; i += 3) {
Collator::EComparisonResult expect;
if (tests[i+1][0] == '<') {
expect = Collator::LESS;
} else if (tests[i+1][0] == '>') {
expect = Collator::GREATER;
} else if (tests[i+1][0] == '=') {
expect = Collator::EQUAL;
} else {
errln((UnicodeString)"Error: unknown operator " + tests[i+1]);
return;
}
UnicodeString s1, s2;
parseChars(s1, tests[i]);
parseChars(s2, tests[i+2]);
doTest(&c, s1, s2, expect);
#if 0
UErrorCode status = U_ZERO_ERROR;
int32_t result = c.compare(s1, s2);
if (sign(result) != sign(expect))
{
UnicodeString t1, t2;
errln(UnicodeString("") +
i/3 + ": compare(" + IntlTest::prettify(s1, t1)
+ " , " + IntlTest::prettify(s2, t2)
+ ") got " + result + "; expected " + expect);
CollationKey k1, k2;
c.getCollationKey(s1, k1, status);
c.getCollationKey(s2, k2, status);
if (U_FAILURE(status)) {
errln((UnicodeString)"Fail: getCollationKey returned " + u_errorName(status));
return;
}
errln((UnicodeString)" key1: " + prettify(k1, t1) );
errln((UnicodeString)" key2: " + prettify(k2, t2) );
}
else
{
CollationKey k1, k2;
c.getCollationKey(s1, k1, status);
c.getCollationKey(s2, k2, status);
if (U_FAILURE(status)) {
errln((UnicodeString)"Fail: getCollationKey returned " + u_errorName(status));
return;
}
result = k1.compareTo(k2);
if (sign(result) != sign(expect)) {
UnicodeString t1, t2;
errln(UnicodeString("") +
i/3 + ": key(" + IntlTest::prettify(s1, t1)
+ ").compareTo(key(" + IntlTest::prettify(s2, t2)
+ ")) got " + result + "; expected " + expect);
errln((UnicodeString)" " + prettify(k1, t1) + " vs. " + prettify(k2, t2));
}
}
#endif
}
}
int8_t CollationThaiTest::sign(int32_t i) {
if (i < 0) return -1;
if (i > 0) return 1;
return 0;
}
UnicodeString& CollationThaiTest::parseChars(UnicodeString& result,
const char* chars) {
return result = CharsToUnicodeString(chars);
}
UCollator *thaiColl = NULL;
U_CDECL_BEGIN
static int U_CALLCONV
StrCmp(const void *p1, const void *p2) {
return ucol_strcoll(thaiColl, *(UChar **) p1, -1, *(UChar **)p2, -1);
}
U_CDECL_END
#define LINES 6
void CollationThaiTest::TestInvalidThai(void) {
const char *tests[LINES] = {
"\\u0E44\\u0E01\\u0E44\\u0E01",
"\\u0E44\\u0E01\\u0E01\\u0E44",
"\\u0E01\\u0E44\\u0E01\\u0E44",
"\\u0E01\\u0E01\\u0E44\\u0E44",
"\\u0E44\\u0E44\\u0E01\\u0E01",
"\\u0E01\\u0E44\\u0E44\\u0E01",
};
UChar strings[LINES][20];
UChar *toSort[LINES];
int32_t i = 0, j = 0, len = 0;
UErrorCode coll_status = U_ZERO_ERROR;
UnicodeString iteratorText;
thaiColl = ucol_open ("th_TH", &coll_status);
if (U_FAILURE(coll_status)) {
errln("Error opening Thai collator: %s", u_errorName(coll_status));
return;
}
CollationElementIterator* c = ((RuleBasedCollator *)coll)->createCollationElementIterator( iteratorText );
for(i = 0; i < (int32_t)(sizeof(tests)/sizeof(tests[0])); i++) {
len = u_unescape(tests[i], strings[i], 20);
strings[i][len] = 0;
toSort[i] = strings[i];
}
qsort (toSort, LINES, sizeof (UChar *), StrCmp);
for (i=0; i < LINES; i++)
{
logln("%i", i);
for (j=i+1; j < LINES; j++) {
if (ucol_strcoll (thaiColl, toSort[i], -1, toSort[j], -1) == UCOL_GREATER)
{
errln("Inconsistent ordering between strings %i and %i", i, j);
}
}
iteratorText.setTo(toSort[i]);
c->setText(iteratorText, coll_status);
backAndForth(*c);
}
ucol_close(thaiColl);
delete c;
}
void CollationThaiTest::TestReordering(void) {
const char *tests[] = {
"\\u0E41c\\u0301", "=", "\\u0E41\\u0107", "\\u0E41\\U0001D7CE", "<", "\\u0E41\\U0001D7CF", "\\u0E41\\U0001D15F", "=", "\\u0E41\\U0001D158\\U0001D165", "\\u0E41\\U0002F802", "=", "\\u0E41\\u4E41", "\\u0E41\\u0301", "=", "\\u0E41\\u0301", "\\u0E41\\u0301\\u0316", "=", "\\u0E41\\u0316\\u0301",
"\\u0e24\\u0e41", "=", "\\u0e41\\u0e24", "\\u0e3f\\u0e3f\\u0e24\\u0e41", "=", "\\u0e3f\\u0e3f\\u0e41\\u0e24",
"abc\\u0E41c\\u0301", "=", "abc\\u0E41\\u0107", "abc\\u0E41\\U0001D000", "<", "abc\\u0E41\\U0001D001", "abc\\u0E41\\U0001D15F", "=", "abc\\u0E41\\U0001D158\\U0001D165", "abc\\u0E41\\U0002F802", "=", "abc\\u0E41\\u4E41", "abc\\u0E41\\u0301", "=", "abc\\u0E41\\u0301", "abc\\u0E41\\u0301\\u0316", "=", "abc\\u0E41\\u0316\\u0301",
"\\u0E41c\\u0301abc", "=", "\\u0E41\\u0107abc", "\\u0E41\\U0001D000abc", "<", "\\u0E41\\U0001D001abc", "\\u0E41\\U0001D15Fabc", "=", "\\u0E41\\U0001D158\\U0001D165abc", "\\u0E41\\U0002F802abc", "=", "\\u0E41\\u4E41abc", "\\u0E41\\u0301abc", "=", "\\u0E41\\u0301abc", "\\u0E41\\u0301\\u0316abc", "=", "\\u0E41\\u0316\\u0301abc",
"abc\\u0E41c\\u0301abc", "=", "abc\\u0E41\\u0107abc", "abc\\u0E41\\U0001D000abc", "<", "abc\\u0E41\\U0001D001abc", "abc\\u0E41\\U0001D15Fabc", "=", "abc\\u0E41\\U0001D158\\U0001D165abc", "abc\\u0E41\\U0002F802abc", "=", "abc\\u0E41\\u4E41abc", "abc\\u0E41\\u0301abc", "=", "abc\\u0E41\\u0301abc", "abc\\u0E41\\u0301\\u0316abc", "=", "abc\\u0E41\\u0316\\u0301abc",
};
LocalPointer<Collator> coll2(coll->clone());
UErrorCode status = U_ZERO_ERROR;
coll2->setAttribute(UCOL_STRENGTH, UCOL_SECONDARY, status);
if(U_FAILURE(status)) {
errln("Unable to set the Thai collator clone to secondary strength");
return;
}
compareArray(*coll2, tests, sizeof(tests)/sizeof(tests[0]));
const char *rule = "& c < ab";
const char *testcontraction[] = { "\\u0E41ab", ">", "\\u0E41c"}; UnicodeString rules;
parseChars(rules, rule);
LocalPointer<RuleBasedCollator> rcoll(new RuleBasedCollator(rules, status), status);
if(U_SUCCESS(status)) {
compareArray(*rcoll, testcontraction, 3);
} else {
errln("Couldn't instantiate collator from rules");
}
}
#endif