/********************************************************************
* COPYRIGHT:
- * Copyright (c) 1999-2008, International Business Machines Corporation and
+ * Copyright (c) 1999-2013, International Business Machines Corporation and
* others. All Rights Reserved.
********************************************************************/
/************************************************************************
* 01/12/2000 Madhu Updated for changed API and added new tests
************************************************************************/
+#include "utypeinfo.h" // for 'typeid' to work
+
#include "unicode/utypes.h"
#if !UCONFIG_NO_BREAK_ITERATION
#include "unicode/ucnv.h"
#include "unicode/schriter.h"
#include "unicode/uniset.h"
-#include "unicode/regex.h" // TODO: make conditional on regexp being built.
+#if !UCONFIG_NO_REGULAR_EXPRESSIONS
+#include "unicode/regex.h"
+#endif
#include "unicode/ustring.h"
#include "unicode/utext.h"
#include "intltest.h"
#include <string.h>
#include "uvector.h"
#include "uvectr32.h"
-#include "triedict.h"
#include <string.h>
#include <stdio.h>
#include <stdlib.h>
+#include "unicode/numfmt.h"
+#include "unicode/uscript.h"
#define TEST_ASSERT(x) {if (!(x)) { \
errln("Failure in file %s, line %d", __FILE__, __LINE__);}}
#define TEST_ASSERT_SUCCESS(errcode) { if (U_FAILURE(errcode)) { \
- errln("Failure in file %s, line %d, status = \"%s\"", __FILE__, __LINE__, u_errorName(errcode));}}
+ errcheckln(errcode, "Failure in file %s, line %d, status = \"%s\"", __FILE__, __LINE__, u_errorName(errcode));}}
//---------------------------------------------
// runIndexedTest
//---------------------------------------------
+
+// Note: Before adding new tests to this file, check whether the desired test data can
+// simply be added to the file testdata/rbbitest.txt. In most cases it can,
+// it's much less work than writing a new test, diagnostic output in the event of failures
+// is good, and the test data file will is shared with ICU4J, so eventually the test
+// will run there as well, without additional effort.
+
void RBBITest::runIndexedTest( int32_t index, UBool exec, const char* &name, char* params )
{
if (exec) logln("TestSuite RuleBasedBreakIterator: ");
switch (index) {
+#if !UCONFIG_NO_FILE_IO
case 0: name = "TestBug4153072";
if(exec) TestBug4153072(); break;
- case 1: name = "TestJapaneseLineBreak";
- if(exec) TestJapaneseLineBreak(); break;
+#else
+ case 0: name = "skip";
+ break;
+#endif
+
+ case 1: name = "skip";
+ break;
case 2: name = "TestStatusReturn";
if(exec) TestStatusReturn(); break;
+
+#if !UCONFIG_NO_FILE_IO
case 3: name = "TestUnicodeFiles";
if(exec) TestUnicodeFiles(); break;
case 4: name = "TestEmptyString";
if(exec) TestEmptyString(); break;
+#else
+ case 3: case 4: name = "skip";
+ break;
+#endif
case 5: name = "TestGetAvailableLocales";
if(exec) TestGetAvailableLocales(); break;
case 6: name = "TestGetDisplayName";
if(exec) TestGetDisplayName(); break;
+#if !UCONFIG_NO_FILE_IO
case 7: name = "TestEndBehaviour";
if(exec) TestEndBehaviour(); break;
- case 8: name = "TestMixedThaiLineBreak";
- if(exec) TestMixedThaiLineBreak(); break;
- case 9: name = "TestThaiLineBreak";
- if(exec) TestThaiLineBreak(); break;
- case 10: name = "TestMaiyamok";
- if(exec) TestMaiyamok(); break;
+ case 8: case 9: case 10: name = "skip";
+ break;
case 11: name = "TestWordBreaks";
if(exec) TestWordBreaks(); break;
case 12: name = "TestWordBoundary";
if(exec) TestSentBreaks(); break;
case 15: name = "TestExtended";
if(exec) TestExtended(); break;
- case 16: name = "TestMonkey";
- if(exec) {
- #if !UCONFIG_NO_REGULAR_EXPRESSIONS
- TestMonkey(params);
- #else
- logln("skipping TestMonkey (UCONFIG_NO_REGULAR_EXPRESSIONS)");
- #endif
- }
- break;
+#else
+ case 7: case 8: case 9: case 10: case 11: case 12: case 13: case 14: case 15: name = "skip";
+ break;
+#endif
+
+#if !UCONFIG_NO_REGULAR_EXPRESSIONS && !UCONFIG_NO_FILE_IO
+ case 16:
+ name = "TestMonkey"; if(exec) TestMonkey(params); break;
+#else
+ case 16:
+ name = "skip"; break;
+#endif
+
+#if !UCONFIG_NO_FILE_IO
case 17: name = "TestBug3818";
if(exec) TestBug3818(); break;
- case 18: name = "TestJapaneseWordBreak";
- if(exec) TestJapaneseWordBreak(); break;
+#else
+ case 17: name = "skip";
+ break;
+#endif
+
+ case 18: name = "skip";
+ break;
case 19: name = "TestDebug";
if(exec) TestDebug(); break;
- case 20: name = "TestTrieDict";
- if(exec) TestTrieDict(); break;
+ case 20: name = "skip";
+ break;
+
+#if !UCONFIG_NO_FILE_IO
case 21: name = "TestBug5775";
- if (exec) TestBug5775(); break;
- case 22: name = "TestThaiBreaks";
- if (exec) TestThaiBreaks(); break;
+ if (exec) TestBug5775(); break;
+#else
+ case 21: name = "skip";
+ break;
+#endif
+ case 22: name = "TestBug9983";
+ if (exec) TestBug9983(); break;
+ case 23: name = "TestDictRules";
+ if (exec) TestDictRules(); break;
+ case 24: name = "TestBug5532";
+ if (exec) TestBug5532(); break;
default: name = ""; break; //needed to end loop
}
}
}
-//-----------------------------------------------------------------------------------
-//
-// Cannned Test Characters
-//
-//-----------------------------------------------------------------------------------
-
-static const UChar cannedTestArray[] = {
- 0x0001, 0x0002, 0x0003, 0x0004, 0x0020, 0x0021, '\\', 0x0022, 0x0023, 0x0024, 0x0025, 0x0026, 0x0028, 0x0029, 0x002b, 0x002d, 0x0030, 0x0031,
- 0x0032, 0x0033, 0x0034, 0x003c, 0x003d, 0x003e, 0x0041, 0x0042, 0x0043, 0x0044, 0x0045, 0x005b, 0x005d, 0x005e, 0x005f, 0x0060, 0x0061, 0x0062, 0x0063, 0x0064, 0x0065, 0x007b,
- 0x007d, 0x007c, 0x002c, 0x00a0, 0x00a2,
- 0x00a3, 0x00a4, 0x00a5, 0x00a6, 0x00a7, 0x00a8, 0x00a9, 0x00ab, 0x00ad, 0x00ae, 0x00af, 0x00b0, 0x00b2, 0x00b3,
- 0x00b4, 0x00b9, 0x00bb, 0x00bc, 0x00bd, 0x02b0, 0x02b1, 0x02b2, 0x02b3, 0x02b4, 0x0300, 0x0301, 0x0302, 0x0303,
- 0x0304, 0x05d0, 0x05d1, 0x05d2, 0x05d3, 0x05d4, 0x0903, 0x093e, 0x093f, 0x0940, 0x0949, 0x0f3a, 0x0f3b, 0x2000,
- 0x2001, 0x2002, 0x200c, 0x200d, 0x200e, 0x200f, 0x2010, 0x2011, 0x2012, 0x2028, 0x2029, 0x202a, 0x203e, 0x203f,
- 0x2040, 0x20dd, 0x20de, 0x20df, 0x20e0, 0x2160, 0x2161, 0x2162, 0x2163, 0x2164, 0x0000
-};
-
-static UnicodeString* cannedTestChars = 0;
-
-#define halfNA "\\u0928\\u094d\\u200d"
-#define halfSA "\\u0938\\u094d\\u200d"
-#define halfCHA "\\u091a\\u094d\\u200d"
-#define halfKA "\\u0915\\u094d\\u200d"
-#define deadTA "\\u0924\\u094d"
-
//--------------------------------------------------------------------------------------
//
// RBBITest constructor and destructor
//--------------------------------------------------------------------------------------
RBBITest::RBBITest() {
- UnicodeString temp(cannedTestArray);
- cannedTestChars = new UnicodeString();
- *cannedTestChars += (UChar)0x0000;
- *cannedTestChars += temp;
}
RBBITest::~RBBITest() {
- delete cannedTestChars;
}
-
-static const int T_NUMBER = 100;
-static const int T_LETTER = 200;
-static const int T_H_OR_K = 300;
-static const int T_IDEO = 400;
-
-
-
-
-
-
-//--------------------------------------------------------------------
-//Testing the BreakIterator for devanagari script
-//--------------------------------------------------------------------
-
-#define deadRA "\\u0930\\u094d" /*deadform RA = devanagari RA + virama*/
-#define deadPHA "\\u092b\\u094d" /*deadform PHA = devanagari PHA + virama*/
-#define deadTTHA "\\u0920\\u094d"
-#define deadPA "\\u092a\\u094d"
-#define deadSA "\\u0938\\u094d"
-#define visarga "\\u0903" /*devanagari visarga looks like a english colon*/
-
-
-
-
-
-
//-----------------------------------------------------------------------------------
//
// Test for status {tag} return value from break rules.
RuleBasedBreakIterator *bi = new RuleBasedBreakIterator(rulesString1, parseError, status);
if(U_FAILURE(status)) {
- errln("FAIL : in construction");
+ dataerrln("FAIL : in construction - %s", u_errorName(status));
} else {
int32_t pos;
int32_t i = 0;
}
}
-void RBBITest::TestThaiLineBreak() {
- UErrorCode status = U_ZERO_ERROR;
- BITestData thaiLineSelection(status);
-
- // \u0e2f-- the Thai paiyannoi character-- isn't a letter. It's a symbol that
- // represents elided letters at the end of a long word. It should be bound to
- // the end of the word and not treated as an independent punctuation mark.
-
-
- ADD_DATACHUNK(thaiLineSelection, NULL, 0, status); // Break at start of data
- ADD_DATACHUNK(thaiLineSelection, "\\u0e2a\\u0e16\\u0e32\\u0e19\\u0e35\\u0e2f", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e08\\u0e30", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e23\\u0e30\\u0e14\\u0e21", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e40\\u0e08\\u0e49\\u0e32", 0, status);
-// ADD_DATACHUNK(thaiLineSelection, "\\u0e2b\\u0e19\\u0e49\\u0e32", 0, status);
-// ADD_DATACHUNK(thaiLineSelection, "\\u0e17\\u0e35\\u0e48", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e2b\\u0e19\\u0e49\\u0e32\\u0e17\\u0e35\\u0e48", 0, status);
- // the commented-out lines (I think) are the preferred result; this line is what our current dictionary is giving us
- ADD_DATACHUNK(thaiLineSelection, "\\u0e2d\\u0e2d\\u0e01", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e21\\u0e32", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e40\\u0e23\\u0e48\\u0e07", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e23\\u0e30\\u0e1a\\u0e32\\u0e22", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e2d\\u0e22\\u0e48\\u0e32\\u0e07", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e40\\u0e15\\u0e47\\u0e21", 0, status);
-
- // the one time where the paiyannoi occurs somewhere other than at the end
- // of a word is in the Thai abbrevation for "etc.", which both begins and
- // ends with a paiyannoi
- ADD_DATACHUNK(thaiLineSelection, "\\u0e2f\\u0e25\\u0e2f", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e17\\u0e35\\u0e48", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e19\\u0e31\\u0e49\\u0e19", 0, status);
-
- RuleBasedBreakIterator* e = (RuleBasedBreakIterator *)BreakIterator::createLineInstance(
- Locale("th"), status);
- if (U_FAILURE(status))
- {
- errln("Failed to create the BreakIterator for Thai locale in TestThaiLineBreak.\n");
- return;
- }
-
- generalIteratorTest(*e, thaiLineSelection);
- delete e;
-}
-
-
-
-void RBBITest::TestMixedThaiLineBreak()
-{
- UErrorCode status = U_ZERO_ERROR;
- BITestData thaiLineSelection(status);
-
- ADD_DATACHUNK(thaiLineSelection, NULL, 0, status); // Break at start of data
-
-
- // @suwit -- Test Arabic numerals, Thai numerals, Punctuation and English characters
- // start
-
- ADD_DATACHUNK(thaiLineSelection, "\\u0E1B\\u0E35", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E1E\\u0E38\\u0E17\\u0E18\\u0E28\\u0E31\\u0E01\\u0E23\\u0E32\\u0E0A ", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "2545 ", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E40\\u0E1B\\u0E47\\u0E19", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E1B\\u0E35", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E09\\u0E25\\u0E2D\\u0E07", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E04\\u0E23\\u0E1A", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E23\\u0E2D\\u0E1A ", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\"\\u0E52\\u0E52\\u0E50 ", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E1b\\u0E35\" ", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E02\\u0E2d\\u0E07", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E01\\u0E23\\u0E38\\u0E07", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E23\\u0E31\\u0E15\\u0E19\\u0E42\\u0E01\\u0E2A\\u0E34\\u0E19\\u0E17\\u0E23\\u0E4C ", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "(\\u0E01\\u0E23\\u0E38\\u0E07\\u0E40\\u0E17\\u0E1e\\u0E2F", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0E2B\\u0E23\\u0E37\\u0E2D ", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "Bangkok)", 0, status);
-
- // @suwit - end of changes
-
-
- RuleBasedBreakIterator* e = (RuleBasedBreakIterator *)BreakIterator::createLineInstance(Locale("th"), status);
- if (U_FAILURE(status))
- {
- errln("Failed to create the BreakIterator for Thai locale in TestMixedThaiLineBreak.\n");
- return;
- }
-
-
- generalIteratorTest(*e, thaiLineSelection);
- delete e;
-}
-
-
-void RBBITest::TestMaiyamok()
-{
- UErrorCode status = U_ZERO_ERROR;
- BITestData thaiLineSelection(status);
- ADD_DATACHUNK(thaiLineSelection, NULL, 0, status); // Break at start of data
- // the Thai maiyamok character is a shorthand symbol that means "repeat the previous
- // word". Instead of appearing as a word unto itself, however, it's kept together
- // with the word before it
- ADD_DATACHUNK(thaiLineSelection, "\\u0e44\\u0e1b\\u0e46", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e21\\u0e32\\u0e46", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e23\\u0e30\\u0e2b\\u0e27\\u0e48\\u0e32\\u0e07", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e01\\u0e23\\u0e38\\u0e07", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e40\\u0e17\\u0e1e", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e41\\u0e25\\u0e30", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e40\\u0e03\\u0e35", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e22\\u0e07", 0, status);
- ADD_DATACHUNK(thaiLineSelection, "\\u0e43\\u0e2b\\u0e21\\u0e48", 0, status);
-
- RuleBasedBreakIterator* e = (RuleBasedBreakIterator *)BreakIterator::createLineInstance(
- Locale("th"), status);
-
- if (U_FAILURE(status))
- {
- errln("Failed to create the BreakIterator for Thai locale in TestMaiyamok.\n");
- return;
- }
- generalIteratorTest(*e, thaiLineSelection);
- delete e;
-}
-
-
void RBBITest::TestBug3818() {
UErrorCode status = U_ZERO_ERROR;
RuleBasedBreakIterator* bi =
(RuleBasedBreakIterator *)BreakIterator::createWordInstance(Locale("th"), status);
if (U_FAILURE(status) || bi == NULL) {
- errln("Fail at file %s, line %d, status = %s", __FILE__, __LINE__, u_errorName(status));
+ errcheckln(status, "Fail at file %s, line %d, status = %s", __FILE__, __LINE__, u_errorName(status));
return;
}
bi->setText(thaiStr);
delete bi;
}
-
-void RBBITest::TestJapaneseWordBreak() {
- UErrorCode status = U_ZERO_ERROR;
- BITestData japaneseWordSelection(status);
-
- ADD_DATACHUNK(japaneseWordSelection, NULL, 0, status); // Break at start of data
- ADD_DATACHUNK(japaneseWordSelection, "\\u4ECA\\u65E5", 400, status); //2
- ADD_DATACHUNK(japaneseWordSelection, "\\u306F\\u3044\\u3044", 300, status); //5
- ADD_DATACHUNK(japaneseWordSelection, "\\u5929\\u6C17", 400, status); //7
- ADD_DATACHUNK(japaneseWordSelection, "\\u3067\\u3059\\u306D", 300, status); //10
- ADD_DATACHUNK(japaneseWordSelection, "\\u3002", 0, status); //11
- ADD_DATACHUNK(japaneseWordSelection, "\\u000D\\u000A", 0, status); //12
-
- RuleBasedBreakIterator* e = (RuleBasedBreakIterator *)BreakIterator::createWordInstance(
- Locale("ja"), status);
- if (U_FAILURE(status))
- {
- errln("Failed to create the BreakIterator for Japanese locale in TestJapaneseWordBreak.\n");
- return;
- }
-
- generalIteratorTest(*e, japaneseWordSelection);
- delete e;
-}
-
-void RBBITest::TestTrieDict() {
- UErrorCode status = U_ZERO_ERROR;
-
- //
- // Open and read the test data file.
- //
- const char *testDataDirectory = IntlTest::getSourceTestData(status);
- char testFileName[1000];
- if (testDataDirectory == NULL || strlen(testDataDirectory) + strlen("riwords.txt") + 10 >= sizeof(testFileName)) {
- errln("Can't open test data. Path too long.");
- return;
- }
- strcpy(testFileName, testDataDirectory);
- strcat(testFileName, "riwords.txt");
-
- // Items needing deleting at the end
- MutableTrieDictionary *mutableDict = NULL;
- CompactTrieDictionary *compactDict = NULL;
- UnicodeSet *breaks = NULL;
- UChar *testFile = NULL;
- StringEnumeration *enumer1 = NULL;
- StringEnumeration *enumer2 = NULL;
- MutableTrieDictionary *mutable2 = NULL;
- StringEnumeration *cloneEnum = NULL;
- CompactTrieDictionary *compact2 = NULL;
-
-
- const UnicodeString *originalWord = NULL;
- const UnicodeString *cloneWord = NULL;
- UChar *current;
- UChar *word;
- UChar uc;
- int32_t wordLen;
- int32_t wordCount;
- int32_t testCount;
-
- int len;
- testFile = ReadAndConvertFile(testFileName, len, NULL, status);
- if (U_FAILURE(status)) {
- goto cleanup; /* something went wrong, error already output */
- }
-
- mutableDict = new MutableTrieDictionary(0x0E1C, status);
- if (U_FAILURE(status)) {
- errln("Error creating MutableTrieDictionary: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- breaks = new UnicodeSet;
- breaks->add(0x000A); // Line Feed
- breaks->add(0x000D); // Carriage Return
- breaks->add(0x2028); // Line Separator
- breaks->add(0x2029); // Paragraph Separator
-
- // Now add each non-comment line of the file as a word.
- current = testFile;
- word = current;
- uc = *current++;
- wordLen = 0;
- wordCount = 0;
-
- while (uc) {
- if (uc == 0x0023) { // #comment line, skip
- while (uc && !breaks->contains(uc)) {
- uc = *current++;
- }
- }
- else while (uc && !breaks->contains(uc)) {
- ++wordLen;
- uc = *current++;
- }
- if (wordLen > 0) {
- mutableDict->addWord(word, wordLen, status);
- if (U_FAILURE(status)) {
- errln("Could not add word to mutable dictionary; status %s\n", u_errorName(status));
- goto cleanup;
- }
- wordCount += 1;
- }
-
- // Find beginning of next line
- while (uc && breaks->contains(uc)) {
- uc = *current++;
- }
- word = current-1;
- wordLen = 0;
- }
-
- if (wordCount < 50) {
- errln("Word count (%d) unreasonably small\n", wordCount);
- goto cleanup;
- }
-
- enumer1 = mutableDict->openWords(status);
- if (U_FAILURE(status)) {
- errln("Could not open mutable dictionary enumerator: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- testCount = 0;
- if (wordCount != (testCount = enumer1->count(status))) {
- errln("MutableTrieDictionary word count (%d) differs from file word count (%d), with status %s\n",
- testCount, wordCount, u_errorName(status));
- goto cleanup;
- }
-
- // Now compact it
- compactDict = new CompactTrieDictionary(*mutableDict, status);
- if (U_FAILURE(status)) {
- errln("Failed to create CompactTrieDictionary: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- enumer2 = compactDict->openWords(status);
- if (U_FAILURE(status)) {
- errln("Could not open compact trie dictionary enumerator: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- if (wordCount != (testCount = enumer2->count(status))) {
- errln("CompactTrieDictionary word count (%d) differs from file word count (%d), with status %s\n",
- testCount, wordCount, u_errorName(status));
- goto cleanup;
- }
-
- if (enumer1->getDynamicClassID() == enumer2->getDynamicClassID()) {
- errln("CompactTrieEnumeration and MutableTrieEnumeration ClassIDs are the same");
- }
- delete enumer1;
- enumer1 = NULL;
- delete enumer2;
- enumer2 = NULL;
-
- // Now un-compact it
- mutable2 = compactDict->cloneMutable(status);
- if (U_FAILURE(status)) {
- errln("Could not clone CompactTrieDictionary to MutableTrieDictionary: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- cloneEnum = mutable2->openWords(status);
- if (U_FAILURE(status)) {
- errln("Could not create cloned mutable enumerator: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- if (wordCount != (testCount = cloneEnum->count(status))) {
- errln("Cloned MutableTrieDictionary word count (%d) differs from file word count (%d), with status %s\n",
- testCount, wordCount, u_errorName(status));
- goto cleanup;
- }
-
- // Compact original dictionary to clone. Note that we can only compare the same kind of
- // dictionary as the order of the enumerators is not guaranteed to be the same between
- // different kinds
- enumer1 = mutableDict->openWords(status);
- if (U_FAILURE(status)) {
- errln("Could not re-open mutable dictionary enumerator: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- originalWord = enumer1->snext(status);
- cloneWord = cloneEnum->snext(status);
- while (U_SUCCESS(status) && originalWord != NULL && cloneWord != NULL) {
- if (*originalWord != *cloneWord) {
- errln("Original and cloned MutableTrieDictionary word mismatch\n");
- goto cleanup;
- }
- originalWord = enumer1->snext(status);
- cloneWord = cloneEnum->snext(status);
- }
-
- if (U_FAILURE(status)) {
- errln("Enumeration failed: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- if (originalWord != cloneWord) {
- errln("Original and cloned MutableTrieDictionary ended enumeration at different points\n");
- goto cleanup;
- }
-
- // Test the data copying constructor for CompactTrieDict, and the data access APIs.
- compact2 = new CompactTrieDictionary(compactDict->data(), status);
- if (U_FAILURE(status)) {
- errln("CompactTrieDictionary(const void *,...) failed\n");
- goto cleanup;
- }
-
- if (compact2->dataSize() == 0) {
- errln("CompactTrieDictionary->dataSize() == 0\n");
- goto cleanup;
- }
-
- // Now count the words via the second dictionary
- delete enumer1;
- enumer1 = compact2->openWords(status);
- if (U_FAILURE(status)) {
- errln("Could not open compact trie dictionary 2 enumerator: %s\n", u_errorName(status));
- goto cleanup;
- }
-
- if (wordCount != (testCount = enumer1->count(status))) {
- errln("CompactTrieDictionary 2 word count (%d) differs from file word count (%d), with status %s\n",
- testCount, wordCount, u_errorName(status));
- goto cleanup;
- }
-
-cleanup:
- delete compactDict;
- delete mutableDict;
- delete breaks;
- delete[] testFile;
- delete enumer1;
- delete mutable2;
- delete cloneEnum;
- delete compact2;
-}
-
-
//----------------------------------------------------------------------------
//
// generalIteratorTest Given a break iterator and a set of test data,
RuleBasedBreakIterator* bi = (RuleBasedBreakIterator *)BreakIterator::createLineInstance(Locale::getDefault(), status);
if (U_FAILURE(status))
{
- errln("Failed to create the BreakIterator for default locale in TestEmptyString.\n");
+ errcheckln(status, "Failed to create the BreakIterator for default locale in TestEmptyString. - %s", u_errorName(status));
return;
}
generalIteratorTest(*bi, x);
const Locale* locList = BreakIterator::getAvailableLocales(locCount);
if (locCount == 0)
- errln("getAvailableLocales() returned an empty list!");
+ dataerrln("getAvailableLocales() returned an empty list!");
// Just make sure that it's returning good memory.
int32_t i;
for (i = 0; i < locCount; ++i) {
BreakIterator::getDisplayName(Locale::getUS(), result);
if (Locale::getDefault() == Locale::getUS() && result != "English (United States)")
- errln("BreakIterator::getDisplayName() failed: expected \"English (United States)\", got \""
+ dataerrln("BreakIterator::getDisplayName() failed: expected \"English (United States)\", got \""
+ result);
BreakIterator::getDisplayName(Locale::getFrance(), Locale::getUS(), result);
if (result != "French (France)")
- errln("BreakIterator::getDisplayName() failed: expected \"French (France)\", got \""
+ dataerrln("BreakIterator::getDisplayName() failed: expected \"French (France)\", got \""
+ result);
}
/**
BreakIterator *wb = BreakIterator::createWordInstance(Locale::getDefault(), status);
if (U_FAILURE(status))
{
- errln("Failed to create the BreakIterator for default locale in TestEndBehaviour.\n");
+ errcheckln(status, "Failed to create the BreakIterator for default locale in TestEndBehaviour. - %s", u_errorName(status));
return;
}
wb->setText(testString);
BreakIterator *iter = BreakIterator::createWordInstance(Locale::getDefault(), status);
if (U_FAILURE(status))
{
- errln("Failed to create the BreakIterator for default locale in TestBug4153072\n");
+ errcheckln(status, "Failed to create the BreakIterator for default locale in TestBug4153072 - %s", u_errorName(status));
return;
}
UnicodeString str("...Hello, World!...");
UErrorCode status = U_ZERO_ERROR;
BreakIterator *bi = BreakIterator::createSentenceInstance(Locale::getEnglish(), status);
TEST_ASSERT_SUCCESS(status);
+ if (U_FAILURE(status)) {
+ return;
+ }
+// Check for status first for better handling of no data errors.
TEST_ASSERT(bi != NULL);
-
- if (U_FAILURE(status) || bi == NULL) {
- // TEST_ASSERT already printed error message.
+ if (bi == NULL) {
return;
}
-
+
UnicodeString s("One.\\u00ad Two.", -1, US_INV);
// 01234 56789
s = s.unescape();
-/**
- * Test Japanese Line Break
- * @bug 4095322
- */
-void RBBITest::TestJapaneseLineBreak()
-{
-#if 0
- // Test needs updating some more... Dump it for now.
-
-
- // Change for Unicode TR 14: Punctuation characters with categories Pi and Pf do not count
- // as opening and closing punctuation for line breaking.
- // Also, \u30fc and \u30fe are not counted as hyphens. Remove these chars
- // from these tests. 6-13-2002
- //
- UErrorCode status = U_ZERO_ERROR;
- UnicodeString testString = CharsToUnicodeString("\\u4e00x\\u4e8c");
- UnicodeString precedingChars = CharsToUnicodeString(
- //"([{\\u00ab$\\u00a5\\u00a3\\u00a4\\u2018\\u201a\\u201c\\u201e\\u201b\\u201f");
- "([{$\\u00a5\\u00a3\\u00a4\\u201a\\u201e");
- UnicodeString followingChars = CharsToUnicodeString(
- // ")]}\\u00bb!%,.\\u3001\\u3002\\u3063\\u3083\\u3085\\u3087\\u30c3\\u30e3\\u30e5\\u30e7\\u30fc"
- ")]}!%,.\\u3001\\u3002\\u3063\\u3083\\u3085\\u3087\\u30c3\\u30e3\\u30e5\\u30e7"
- // ":;\\u309b\\u309c\\u3005\\u309d\\u309e\\u30fd\\u30fe\\u2019\\u201d\\u00b0\\u2032\\u2033\\u2034"
- ":;\\u309b\\u309c\\u3005\\u309d\\u309e\\u30fd\\u00b0\\u2032\\u2033\\u2034"
- "\\u2030\\u2031\\u2103\\u2109\\u00a2\\u0300\\u0301\\u0302");
- BreakIterator *iter = BreakIterator::createLineInstance(Locale::getJapan(), status);
-
- int32_t i;
- if (U_FAILURE(status))
- {
- errln("Failed to create the BreakIterator for Japanese locale in TestJapaneseLineBreak.\n");
- return;
- }
-
- for (i = 0; i < precedingChars.length(); i++) {
- testString.setCharAt(1, precedingChars[i]);
- iter->setText(testString);
- int32_t j = iter->first();
- if (j != 0)
- errln("ja line break failure: failed to start at 0");
- j = iter->next();
- if (j != 1)
- errln("ja line break failure: failed to stop before '" + UCharToUnicodeString(precedingChars[i])
- + "' (" + ((int)(precedingChars[i])) + ")");
- j = iter->next();
- if (j != 3)
- errln("ja line break failure: failed to skip position after '" + UCharToUnicodeString(precedingChars[i])
- + "' (" + ((int)(precedingChars[i])) + ")");
- }
-
- for (i = 0; i < followingChars.length(); i++) {
- testString.setCharAt(1, followingChars[i]);
- iter->setText(testString);
- int j = iter->first();
- if (j != 0)
- errln("ja line break failure: failed to start at 0");
- j = iter->next();
- if (j != 2)
- errln("ja line break failure: failed to skip position before '" + UCharToUnicodeString(followingChars[i])
- + "' (" + ((int)(followingChars[i])) + ")");
- j = iter->next();
- if (j != 3)
- errln("ja line break failure: failed to stop after '" + UCharToUnicodeString(followingChars[i])
- + "' (" + ((int)(followingChars[i])) + ")");
- }
- delete iter;
-#endif
-}
-
-
//------------------------------------------------------------------------------
//
// RBBITest::Extended Run RBBI Tests from an external test data file
i, t->srcLine->elementAti(i), t->srcCol->elementAti(i));
}
}
+
+ // Check isBoundary()
+ for (i=0; i<t->expectedBreaks->size(); i++) {
+ UBool boundaryExpected = (t->expectedBreaks->elementAti(i) != 0);
+ UBool boundaryFound = t->bi->isBoundary(i);
+ if (boundaryExpected != boundaryFound) {
+ errln("isBoundary(%d) incorrect. File line,col= %4d,%4d\n"
+ " Expected, Actual= %s, %s",
+ i, t->srcLine->elementAti(i), t->srcCol->elementAti(i),
+ boundaryExpected ? "true":"false", boundaryFound? "true" : "false");
+ }
+ }
+
+ // Check following()
+ for (i=0; i<t->expectedBreaks->size(); i++) {
+ int32_t actualBreak = t->bi->following(i);
+ int32_t expectedBreak = BreakIterator::DONE;
+ for (int32_t j=i+1; j < t->expectedBreaks->size(); j++) {
+ if (t->expectedBreaks->elementAti(j) != 0) {
+ expectedBreak = j;
+ break;
+ }
+ }
+ if (expectedBreak != actualBreak) {
+ errln("following(%d) incorrect. File line,col= %4d,%4d\n"
+ " Expected, Actual= %d, %d",
+ i, t->srcLine->elementAti(i), t->srcCol->elementAti(i), expectedBreak, actualBreak);
+ }
+ }
+
+ // Check preceding()
+ for (i=t->expectedBreaks->size(); i>=0; i--) {
+ int32_t actualBreak = t->bi->preceding(i);
+ int32_t expectedBreak = BreakIterator::DONE;
+
+ for (int32_t j=i-1; j >= 0; j--) {
+ if (t->expectedBreaks->elementAti(j) != 0) {
+ expectedBreak = j;
+ break;
+ }
+ }
+ if (expectedBreak != actualBreak) {
+ errln("preceding(%d) incorrect. File line,col= %4d,%4d\n"
+ " Expected, Actual= %d, %d",
+ i, t->srcLine->elementAti(i), t->srcCol->elementAti(i), expectedBreak, actualBreak);
+ }
+ }
}
tp.srcCol = new UVector32(status);
RegexMatcher localeMatcher(UNICODE_STRING_SIMPLE("<locale *([\\p{L}\\p{Nd}_]*) *>"), 0, status);
- TEST_ASSERT_SUCCESS(status);
+ if (U_FAILURE(status)) {
+ dataerrln("Failure in file %s, line %d, status = \"%s\"", __FILE__, __LINE__, u_errorName(status));
+ }
//
char localeName8[100];
localeName.extract(0, localeName.length(), localeName8, sizeof(localeName8), 0);
locale = Locale::createFromName(localeName8);
- charIdx += localeMatcher.group(0, status).length();
+ charIdx += localeMatcher.group(0, status).length() - 1;
TEST_ASSERT_SUCCESS(status);
break;
}
if (U_FAILURE(status)) {
- errln("ICU Error %s while parsing test file at line %d.",
+ dataerrln("ICU Error %s while parsing test file at line %d.",
u_errorName(status), lineNum);
status = U_ZERO_ERROR;
goto end_test; // Stop the test
#endif
}
-void RBBITest::TestThaiBreaks() {
- UErrorCode status=U_ZERO_ERROR;
- BreakIterator* b;
- Locale locale = Locale("th");
- int32_t p, index;
- UChar c[]= {
- 0x0E01, 0x0E39, 0x0020, 0x0E01, 0x0E34, 0x0E19, 0x0E01, 0x0E38, 0x0E49, 0x0E07, 0x0020, 0x0E1B,
- 0x0E34, 0x0E49, 0x0E48, 0x0E07, 0x0E2D, 0x0E22, 0x0E39, 0x0E48, 0x0E43, 0x0E19,
- 0x0E16, 0x0E49, 0x0E33
- };
- int32_t expectedWordResult[] = {
- 2, 3, 6, 10, 11, 15, 17, 20, 22
- };
- int32_t expectedLineResult[] = {
- 3, 6, 11, 15, 17, 20, 22
- };
- int32_t size = sizeof(c)/sizeof(UChar);
- UnicodeString text=UnicodeString(c);
-
- b = BreakIterator::createWordInstance(locale, status);
- if (U_FAILURE(status)) {
- errln("Unable to create thai word break iterator.\n");
- return;
- }
- b->setText(text);
- p = index = 0;
- while ((p=b->next())!=BreakIterator::DONE && p < size) {
- if (p != expectedWordResult[index++]) {
- errln("Incorrect break given by thai word break iterator. Expected: %d Got: %d", expectedWordResult[index-1], p);
- }
- }
- delete b;
-
- b = BreakIterator::createLineInstance(locale, status);
- if (U_FAILURE(status)) {
- printf("Unable to create thai line break iterator.\n");
- return;
- }
- b->setText(text);
- p = index = 0;
- while ((p=b->next())!=BreakIterator::DONE && p < size) {
- if (p != expectedLineResult[index++]) {
- errln("Incorrect break given by thai line break iterator. Expected: %d Got: %d", expectedLineResult[index-1], p);
+
+//-------------------------------------------------------------------------------
+//
+// TestDictRules create a break iterator from source rules that includes a
+// dictionary range. Regression for bug #7130. Source rules
+// do not declare a break iterator type (word, line, sentence, etc.
+// but the dictionary code, without a type, would loop.
+//
+//-------------------------------------------------------------------------------
+void RBBITest::TestDictRules() {
+ const char *rules = "$dictionary = [a-z]; \n"
+ "!!forward; \n"
+ "$dictionary $dictionary; \n"
+ "!!reverse; \n"
+ "$dictionary $dictionary; \n";
+ const char *text = "aa";
+ UErrorCode status = U_ZERO_ERROR;
+ UParseError parseError;
+
+ RuleBasedBreakIterator bi(rules, parseError, status);
+ if (U_SUCCESS(status)) {
+ UnicodeString utext = text;
+ bi.setText(utext);
+ int32_t position;
+ int32_t loops;
+ for (loops = 0; loops<10; loops++) {
+ position = bi.next();
+ if (position == RuleBasedBreakIterator::DONE) {
+ break;
+ }
}
+ TEST_ASSERT(loops == 1);
+ } else {
+ dataerrln("Error creating RuleBasedBreakIterator: %s", u_errorName(status));
}
-
- delete b;
}
+
//-------------------------------------------------------------------------------
//
// ReadAndConvertFile Read a text data file, convert it to UChars, and
//
f = fopen(fileName, "rb");
if (f == 0) {
- dataerrln("[DATA] Error opening test data file %s\n", fileName);
+ dataerrln("Error opening test data file %s\n", fileName);
status = U_FILE_ACCESS_ERROR;
return NULL;
}
ucnv_close(conv);
if (U_FAILURE(status)) {
errln("ucnv_toUChars: ICU Error \"%s\"\n", u_errorName(status));
- delete retPtr;
+ delete []retPtr;
retPtr = 0;
ulen = 0;
};
RuleBasedBreakIterator *bi;
UErrorCode status = U_ZERO_ERROR;
- bi = (RuleBasedBreakIterator *)BreakIterator::createCharacterInstance(Locale::getDefault(), status);
+ bi = (RuleBasedBreakIterator *)BreakIterator::createCharacterInstance(Locale::getEnglish(), status);
TEST_ASSERT_SUCCESS(status);
if (U_SUCCESS(status)) {
runUnicodeTestData("GraphemeBreakTest.txt", bi);
}
delete bi;
- bi = (RuleBasedBreakIterator *)BreakIterator::createWordInstance(Locale::getDefault(), status);
+ bi = (RuleBasedBreakIterator *)BreakIterator::createWordInstance(Locale::getEnglish(), status);
TEST_ASSERT_SUCCESS(status);
if (U_SUCCESS(status)) {
runUnicodeTestData("WordBreakTest.txt", bi);
}
delete bi;
- bi = (RuleBasedBreakIterator *)BreakIterator::createSentenceInstance(Locale::getDefault(), status);
+ bi = (RuleBasedBreakIterator *)BreakIterator::createSentenceInstance(Locale::getEnglish(), status);
TEST_ASSERT_SUCCESS(status);
if (U_SUCCESS(status)) {
runUnicodeTestData("SentenceBreakTest.txt", bi);
}
delete bi;
- bi = (RuleBasedBreakIterator *)BreakIterator::createLineInstance(Locale::getDefault(), status);
+ bi = (RuleBasedBreakIterator *)BreakIterator::createLineInstance(Locale::getEnglish(), status);
TEST_ASSERT_SUCCESS(status);
if (U_SUCCESS(status)) {
runUnicodeTestData("LineBreakTest.txt", bi);
//-------------------------------------------------------------------------------------------
void RBBITest::runUnicodeTestData(const char *fileName, RuleBasedBreakIterator *bi) {
#if !UCONFIG_NO_REGULAR_EXPRESSIONS
+ // TODO(andy): Match line break behavior to Unicode 6.0 and remove this time bomb. Ticket #7270
+ UBool isTicket7270Fixed = isICUVersionAtLeast(52, 1);
+ UBool isLineBreak = 0 == strcmp(fileName, "LineBreakTest.txt");
UErrorCode status = U_ZERO_ERROR;
//
const char *testDataDirectory = IntlTest::getSourceTestData(status);
char testFileName[1000];
if (testDataDirectory == NULL || strlen(testDataDirectory) >= sizeof(testFileName)) {
- dataerrln("[DATA] Can't open test data. Path too long.");
+ dataerrln("Can't open test data. Path too long.");
return;
}
strcpy(testFileName, testDataDirectory);
// Scan through each test case, building up the string to be broken in testString,
// and the positions that should be boundaries in the breakPositions vector.
//
+ int spin = 0;
while (tokenMatcher.find()) {
+ if(tokenMatcher.hitEnd()) {
+ /* Shouldnt Happen(TM). This means we didn't find the symbols we were looking for.
+ This occurred when the text file was corrupt (wasn't marked as UTF-8)
+ and caused an infinite loop here on EBCDIC systems!
+ */
+ fprintf(stderr,"FAIL: hit end of file %s for the %8dth time- corrupt data file?\r", fileName, ++spin);
+ // return;
+ }
if (tokenMatcher.start(1, status) >= 0) {
// Scanned a divide sign, indicating a break position in the test data.
if (testString.length()>0) {
// If the line from the file contained test data, run the test now.
//
if (testString.length() > 0) {
+// TODO(andy): Remove this time bomb code. Note: Failing line numbers may change when updating to new Unicode data.
+// Rule 8
+// ZW SP* <break>
+// is not yet implemented.
+if (!(isLineBreak && !isTicket7270Fixed && (5198 == lineNumber ||
+ 5202 == lineNumber ||
+ 5214 == lineNumber ||
+ 5246 == lineNumber ||
+ 5298 == lineNumber ||
+ 5302 == lineNumber ))) {
checkUnicodeTestCase(fileName, lineNumber, testString, &breakPositions, bi);
+}
}
// Clear out this test case.
UnicodeSet *fCRLFSet;
UnicodeSet *fControlSet;
UnicodeSet *fExtendSet;
+ UnicodeSet *fRegionalIndicatorSet;
UnicodeSet *fPrependSet;
UnicodeSet *fSpacingSet;
UnicodeSet *fLSet;
fCRLFSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\r\\n]"), status);
fControlSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Grapheme_Cluster_Break = Control}]"), status);
fExtendSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Grapheme_Cluster_Break = Extend}]"), status);
+ fRegionalIndicatorSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Grapheme_Cluster_Break = Regional_Indicator}]"), status);
fPrependSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Grapheme_Cluster_Break = Prepend}]"), status);
fSpacingSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Grapheme_Cluster_Break = SpacingMark}]"), status);
fLSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Grapheme_Cluster_Break = L}]"), status);
fHangulSet->addAll(*fTSet);
fHangulSet->addAll(*fLVSet);
fHangulSet->addAll(*fLVTSet);
- fAnySet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\u0000-\\U0010ffff]"), status);
-
+ fAnySet = new UnicodeSet(0, 0x10ffff);
+
fSets = new UVector(status);
fSets->addElement(fCRLFSet, status);
fSets->addElement(fControlSet, status);
fSets->addElement(fExtendSet, status);
- fSets->addElement(fPrependSet, status);
+ fSets->addElement(fRegionalIndicatorSet, status);
+ if (!fPrependSet->isEmpty()) {
+ fSets->addElement(fPrependSet, status);
+ }
fSets->addElement(fSpacingSet, status);
fSets->addElement(fHangulSet, status);
fSets->addElement(fAnySet, status);
continue;
}
+ // Just adding extra Apple rule does here not work, behavior depends on arbitrary context
+
+ // Rule (GB8a) Regional_Indicator x Regional_Indicator
+ if (fRegionalIndicatorSet->contains(c1) && fRegionalIndicatorSet->contains(c2)) {
+ continue;
+ }
+
// Rule (GB9) Numeric x ALetter
if (fExtendSet->contains(c2)) {
continue;
delete fCRLFSet;
delete fControlSet;
delete fExtendSet;
+ delete fRegionalIndicatorSet;
delete fPrependSet;
delete fSpacingSet;
delete fLSet;
UnicodeSet *fNewlineSet;
UnicodeSet *fKatakanaSet;
UnicodeSet *fALetterSet;
+ // TODO(jungshik): Do we still need this change?
+ // UnicodeSet *fALetterSet; // matches ALetterPlus in word.txt
UnicodeSet *fMidNumLetSet;
UnicodeSet *fMidLetterSet;
UnicodeSet *fMidNumSet;
UnicodeSet *fOtherSet;
UnicodeSet *fExtendSet;
UnicodeSet *fExtendNumLetSet;
+ UnicodeSet *fRegionalIndicatorSet;
+ UnicodeSet *fDictionaryCjkSet;
RegexMatcher *fMatcher;
fCRSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = CR}]"), status);
fLFSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = LF}]"), status);
fNewlineSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = Newline}]"), status);
- fALetterSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = ALetter}]"), status);
+ fDictionaryCjkSet= new UnicodeSet("[[\\uac00-\\ud7a3][:Han:][:Hiragana:][:Katakana:]]", status);
+ // Exclude Hangul syllables from ALetterSet during testing.
+ // Leave CJK dictionary characters out from the monkey tests!
+#if 0
+ fALetterSet = new UnicodeSet("[\\p{Word_Break = ALetter}"
+ "[\\p{Line_Break = Complex_Context}"
+ "-\\p{Grapheme_Cluster_Break = Extend}"
+ "-\\p{Grapheme_Cluster_Break = Control}"
+ "]]",
+ status);
+#endif
+ fALetterSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = ALetter}]"), status);
+ fALetterSet->removeAll(*fDictionaryCjkSet);
fKatakanaSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = Katakana}]"), status);
fMidNumLetSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = MidNumLet}]"), status);
fMidLetterSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = MidLetter}]"), status);
fMidNumSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = MidNum}]"), status);
+ // TODO: this set used to contain [\\uff10-\\uff19] (fullwidth digits), but this breaks the test
+ // we should figure out why
fNumericSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = Numeric}]"), status);
fFormatSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = Format}]"), status);
fExtendNumLetSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = ExtendNumLet}]"), status);
fExtendSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = Extend}]"), status);
+ fRegionalIndicatorSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Word_Break = Regional_Indicator}]"), status);
fOtherSet = new UnicodeSet();
if(U_FAILURE(status)) {
fOtherSet->removeAll(*fExtendNumLetSet);
fOtherSet->removeAll(*fFormatSet);
fOtherSet->removeAll(*fExtendSet);
+ fOtherSet->removeAll(*fRegionalIndicatorSet);
// Inhibit dictionary characters from being tested at all.
+ fOtherSet->removeAll(*fDictionaryCjkSet);
fOtherSet->removeAll(UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{LineBreak = Complex_Context}]"), status));
fSets->addElement(fCRSet, status);
fSets->addElement(fLFSet, status);
fSets->addElement(fNewlineSet, status);
fSets->addElement(fALetterSet, status);
- fSets->addElement(fKatakanaSet, status);
+ //fSets->addElement(fKatakanaSet, status); //TODO: work out how to test katakana
fSets->addElement(fMidLetterSet, status);
fSets->addElement(fMidNumLetSet, status);
fSets->addElement(fMidNumSet, status);
fSets->addElement(fExtendSet, status);
fSets->addElement(fOtherSet, status);
fSets->addElement(fExtendNumLetSet, status);
+ fSets->addElement(fRegionalIndicatorSet, status);
if (U_FAILURE(status)) {
deferredStatus = status;
fKatakanaSet->contains(c1) || fExtendNumLetSet->contains(c1)) &&
fExtendNumLetSet->contains(c2)) {
continue;
- }
+ }
// Rule 13b
if (fExtendNumLetSet->contains(c1) &&
(fALetterSet->contains(c2) || fNumericSet->contains(c2) ||
fKatakanaSet->contains(c2))) {
continue;
- }
+ }
+
+ // Rule 13c
+ if (fRegionalIndicatorSet->contains(c1) && fRegionalIndicatorSet->contains(c2)) {
+ continue;
+ }
// Rule 14. Break found here.
break;
delete fFormatSet;
delete fExtendSet;
delete fExtendNumLetSet;
+ delete fRegionalIndicatorSet;
+ delete fDictionaryCjkSet;
delete fOtherSet;
}
UnicodeSet *fH2;
UnicodeSet *fH3;
UnicodeSet *fCL;
+ UnicodeSet *fCP;
UnicodeSet *fEX;
UnicodeSet *fIN;
UnicodeSet *fJL;
UnicodeSet *fSY;
UnicodeSet *fAI;
UnicodeSet *fAL;
+ UnicodeSet *fCJ;
+ UnicodeSet *fHL;
UnicodeSet *fID;
+ UnicodeSet *fRI;
UnicodeSet *fSA;
UnicodeSet *fXX;
fH2 = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=H2}]"), status);
fH3 = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=H3}]"), status);
fCL = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=CL}]"), status);
+ fCP = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=CP}]"), status);
fEX = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=EX}]"), status);
fIN = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=IN}]"), status);
fJL = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=JL}]"), status);
fSY = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=SY}]"), status);
fAI = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=AI}]"), status);
fAL = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=AL}]"), status);
+ fCJ = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=CJ}]"), status);
+ fHL = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=HL}]"), status);
fID = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=ID}]"), status);
+ fRI = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=RI}]"), status);
fSA = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=SA}]"), status);
fSG = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\ud800-\\udfff]"), status);
fXX = new UnicodeSet(UNICODE_STRING_SIMPLE("[\\p{Line_break=XX}]"), status);
fAL->addAll(*fSA); // Default behavior for SA is XX, which defaults to AL
fAL->addAll(*fSG); // Default behavior for SG is identical to AL.
+ fNS->addAll(*fCJ); // Default behavior for CJ is identical to NS.
+
fSets->addElement(fBK, status);
fSets->addElement(fCR, status);
fSets->addElement(fLF, status);
fSets->addElement(fH2, status);
fSets->addElement(fH3, status);
fSets->addElement(fCL, status);
+ fSets->addElement(fCP, status);
fSets->addElement(fEX, status);
fSets->addElement(fIN, status);
fSets->addElement(fJL, status);
fSets->addElement(fSY, status);
fSets->addElement(fAI, status);
fSets->addElement(fAL, status);
+ fSets->addElement(fHL, status);
fSets->addElement(fID, status);
fSets->addElement(fWJ, status);
+ fSets->addElement(fRI, status);
fSets->addElement(fSA, status);
fSets->addElement(fSG, status);
"((\\p{Line_Break=OP}|\\p{Line_Break=HY})\\p{Line_Break=CM}*)?"
"\\p{Line_Break=NU}\\p{Line_Break=CM}*"
"((\\p{Line_Break=NU}|\\p{Line_Break=IS}|\\p{Line_Break=SY})\\p{Line_Break=CM}*)*"
- "(\\p{Line_Break=CL}\\p{Line_Break=CM}*)?"
+ "((\\p{Line_Break=CL}|\\p{Line_Break=CP})\\p{Line_Break=CM}*)?"
"((\\p{Line_Break=PR}|\\p{Line_Break=PO})\\p{Line_Break=CM}*)?";
fNumberMatcher = new RegexMatcher(
// and thisChar may not be adjacent because combining
// characters between them will be ignored.
+ int32_t prevPosX2; // Second previous character. Wider context for LB21a.
+ UChar32 prevCharX2;
+
int32_t nextPos; // Index of the next character following pos.
// Usually skips over combining marks.
int32_t nextCPPos; // Index of the code point following "pos."
// Initial values for loop. Loop will run the first time without finding breaks,
// while the invalid values shift out and the "this" and
// "prev" positions are filled in with good values.
- pos = prevPos = -1; // Invalid value, serves as flag for initial loop iteration.
- thisChar = prevChar = 0;
+ pos = prevPos = prevPosX2 = -1; // Invalid value, serves as flag for initial loop iteration.
+ thisChar = prevChar = prevCharX2 = 0;
nextPos = nextCPPos = startPos;
// Loop runs once per position in the test text, until a break position
// is found.
for (;;) {
+ prevPosX2 = prevPos;
+ prevCharX2 = prevChar;
+
prevPos = pos;
prevChar = thisChar;
// LB 13 Don't break before closings.
- // NU x CL and NU x IS are not matched here so that they will
+ // NU x CL, NU x CP and NU x IS are not matched here so that they will
// fall into LB 17 and the more general number regular expression.
//
- if (!fNU->contains(prevChar) && fCL->contains(thisChar) ||
- fEX->contains(thisChar) ||
- !fNU->contains(prevChar) && fIS->contains(thisChar) ||
- !fNU->contains(prevChar) && fSY->contains(thisChar)) {
+ if ((!fNU->contains(prevChar) && fCL->contains(thisChar)) ||
+ (!fNU->contains(prevChar) && fCP->contains(thisChar)) ||
+ fEX->contains(thisChar) ||
+ (!fNU->contains(prevChar) && fIS->contains(thisChar)) ||
+ (!fNU->contains(prevChar) && fSY->contains(thisChar))) {
continue;
}
- // LB 16 CL SP* x NS
- // Scan backwards for SP* CM* CL
+ // LB 16 (CL | CP) SP* x NS
+ // Scan backwards for SP* CM* (CL | CP)
if (fNS->contains(thisChar)) {
int tPos = prevPos;
while (tPos>0 && fSP->contains(fText->char32At(tPos))) {
while (tPos>0 && fCM->contains(fText->char32At(tPos))) {
tPos = fText->moveIndex32(tPos, -1);
}
- if (fCL->contains(fText->char32At(tPos))) {
+ if (fCL->contains(fText->char32At(tPos)) || fCP->contains(fText->char32At(tPos))) {
continue;
}
}
continue;
}
+ // LB 21a
+ // HL (HY | BA) x
+ if (fHL->contains(prevCharX2) &&
+ (fHY->contains(prevChar) || fBA->contains(prevChar))) {
+ continue;
+ }
+
+ // LB 21b - Added for Apple 13927604
+ if (fSY->contains(prevChar) && fHL->contains(thisChar)) {
+ continue;
+ }
+
// LB 22
- if (fAL->contains(prevChar) && fIN->contains(thisChar) ||
- fID->contains(prevChar) && fIN->contains(thisChar) ||
- fIN->contains(prevChar) && fIN->contains(thisChar) ||
- fNU->contains(prevChar) && fIN->contains(thisChar) ) {
+ if ((fAL->contains(prevChar) && fIN->contains(thisChar)) ||
+ (fHL->contains(prevChar) && fIN->contains(thisChar)) ||
+ (fID->contains(prevChar) && fIN->contains(thisChar)) ||
+ (fIN->contains(prevChar) && fIN->contains(thisChar)) ||
+ (fNU->contains(prevChar) && fIN->contains(thisChar)) ) {
continue;
}
// LB 23 ID x PO
// AL x NU
+ // HL x NU
// NU x AL
- if (fID->contains(prevChar) && fPO->contains(thisChar) ||
- fAL->contains(prevChar) && fNU->contains(thisChar) ||
- fNU->contains(prevChar) && fAL->contains(thisChar) ) {
+ if ((fID->contains(prevChar) && fPO->contains(thisChar)) ||
+ (fAL->contains(prevChar) && fNU->contains(thisChar)) ||
+ (fHL->contains(prevChar) && fNU->contains(thisChar)) ||
+ (fNU->contains(prevChar) && fAL->contains(thisChar)) ||
+ (fNU->contains(prevChar) && fHL->contains(thisChar)) ) {
continue;
}
// LB 24 Do not break between prefix and letters or ideographs.
// PR x ID
- // PR x AL
- // PO x AL
- if (fPR->contains(prevChar) && fID->contains(thisChar) ||
- fPR->contains(prevChar) && fAL->contains(thisChar) ||
- fPO->contains(prevChar) && fAL->contains(thisChar) ) {
+ // PR x (AL | HL)
+ // PO x (AL | HL)
+ if ((fPR->contains(prevChar) && fID->contains(thisChar)) ||
+ (fPR->contains(prevChar) && (fAL->contains(thisChar) || fHL->contains(thisChar))) ||
+ (fPO->contains(prevChar) && (fAL->contains(thisChar) || fHL->contains(thisChar)))) {
continue;
}
// LB 28 Do not break between alphabetics ("at").
- if (fAL->contains(prevChar) && fAL->contains(thisChar)) {
+ if ((fAL->contains(prevChar) || fHL->contains(prevChar)) && (fAL->contains(thisChar) || fHL->contains(thisChar))) {
continue;
}
// LB 29 Do not break between numeric punctuation and alphabetics ("e.g.").
- if (fIS->contains(prevChar) && fAL->contains(thisChar)) {
+ if (fIS->contains(prevChar) && (fAL->contains(thisChar) || fHL->contains(thisChar))) {
+ continue;
+ }
+
+ // LB 30 Do not break between letters, numbers, or ordinary symbols and opening or closing punctuation.
+ // (AL | NU) x OP
+ // CP x (AL | NU)
+ if ((fAL->contains(prevChar) || fHL->contains(prevChar) || fNU->contains(prevChar)) && fOP->contains(thisChar)) {
+ continue;
+ }
+ if (fCP->contains(prevChar) && (fAL->contains(thisChar) || fHL->contains(thisChar) || fNU->contains(thisChar))) {
+ continue;
+ }
+
+ // LB30a Do not break between regional indicators.
+ // RI x RI
+ if (fRI->contains(prevChar) && fRI->contains(thisChar)) {
continue;
}
delete fH2;
delete fH3;
delete fCL;
+ delete fCP;
delete fEX;
delete fIN;
delete fJL;
delete fSY;
delete fAI;
delete fAL;
+ delete fCJ;
+ delete fHL;
delete fID;
+ delete fRI;
delete fSA;
delete fSG;
delete fXX;
}
#endif
+#if !UCONFIG_NO_REGULAR_EXPRESSIONS
static void testBreakBoundPreceding(RBBITest *test, UnicodeString ustr,
BreakIterator *bi,
int expected[],
for (i = bi->last(); i != BreakIterator::DONE; i = bi->previous()) {
count --;
if (forward[count] != i) {
+ printStringBreaks(ustr, expected, expectedcount);
test->errln("happy break test previous() failed: expected %d but got %d",
forward[count], i);
break;
}
}
}
+#endif
void RBBITest::TestWordBreaks(void)
{
UErrorCode status = U_ZERO_ERROR;
// BreakIterator *bi = BreakIterator::createCharacterInstance(locale, status);
BreakIterator *bi = BreakIterator::createWordInstance(locale, status);
+ // Replaced any C+J characters in a row with a random sequence of characters
+ // of the same length to make our C+J segmentation not get in the way.
static const char *strlist[] =
{
"\\U000e0032\\u0097\\u0f94\\uc2d8\\u05f4\\U000e0031\\u060d",
- "\\U000e0037\\u4666\\u1202\\u003a\\U000e0031\\u064d\\u0bea\\u591c\\U000e0040\\u003b",
+ "\\U000e0037\\u2666\\u1202\\u003a\\U000e0031\\u064d\\u0bea\\u091c\\U000e0040\\u003b",
"\\u0589\\u3e99\\U0001d7f3\\U000e0074\\u1810\\u200e\\U000e004b\\u0027\\U000e0061\\u003a",
"\\u398c\\U000104a5\\U0001d173\\u102d\\u002e\\uca3b\\u002e\\u002c\\u5622",
- "\\u90ca\\u3588\\u009c\\u0953\\u194b",
+ "\\uac00\\u3588\\u009c\\u0953\\u194b",
"\\u200e\\U000e0072\\u0a4b\\U000e003f\\ufd2b\\u2027\\u002e\\u002e",
"\\u0602\\u2019\\ua191\\U000e0063\\u0a4c\\u003a\\ub4b5\\u003a\\u827f\\u002e",
- "\\u7f1f\\uc634\\u65f8\\u0944\\u04f2\\uacdf\\u1f9c\\u05f4\\u002e",
+ "\\u2f1f\\u1634\\u05f8\\u0944\\u04f2\\u0cdf\\u1f9c\\u05f4\\u002e",
"\\U000e0042\\u002e\\u0fb8\\u09ef\\u0ed1\\u2044",
"\\u003b\\u024a\\u102e\\U000e0071\\u0600",
"\\u2027\\U000e0067\\u0a47\\u00b7",
"\\u1fcd\\u002c\\u07aa\\u0027\\u11b0",
"\\u002c\\U000e003c\\U0001d7f4\\u003a\\u0c6f\\u0027",
"\\u0589\\U000e006e\\u0a42\\U000104a5",
- "\\u4f66\\ub523\\u003a\\uacae\\U000e0047\\u003a",
+ "\\u0f66\\u2523\\u003a\\u0cae\\U000e0047\\u003a",
"\\u003a\\u0f21\\u0668\\u0dab\\u003a\\u0655\\u00b7",
"\\u0027\\u11af\\U000e0057\\u0602",
"\\U0001d7f2\\U000e007\\u0004\\u0589",
"\\u0be8\\u002e\\u0c68\\u066e\\u136d\\ufc99\\u59e7",
"\\u0233\\U000e0020\\u0a69\\u0d6a",
"\\u206f\\u0741\\ub3ab\\u2019\\ubcac\\u2019",
- "\\u58f4\\U000e0049\\u20e7\\u2027",
+ "\\u18f4\\U000e0049\\u20e7\\u2027",
"\\ub315\\U0001d7e5\\U000e0073\\u0c47\\u06f2\\u0c6a\\u0037\\u10fe",
"\\ua183\\u102d\\u0bec\\u003a",
"\\u17e8\\u06e7\\u002e\\u096d\\u003b",
"\\U000e005d\\u2044\\u0731\\u0650\\u0061",
"\\u003a\\u0664\\u00b7\\u1fba",
"\\u003b\\u0027\\u00b7\\u47a3",
- "\\u2027\\U000e0067\\u0a42\\u00b7\\ubddf\\uc26c\\u003a\\u4186\\u041b",
+ "\\u2027\\U000e0067\\u0a42\\u00b7\\u4edf\\uc26c\\u003a\\u4186\\u041b",
"\\u0027\\u003a\\U0001d70f\\U0001d7df\\ubf4a\\U0001d7f5\\U0001d177\\u003a\\u0e51\\u1058\\U000e0058\\u00b7\\u0673",
"\\uc30d\\u002e\\U000e002c\\u0c48\\u003a\\ub5a1\\u0661\\u002c",
};
int loop;
if (U_FAILURE(status)) {
- errln("Creation of break iterator failed %s", u_errorName(status));
+ errcheckln(status, "Creation of break iterator failed %s", u_errorName(status));
return;
}
for (loop = 0; loop < (int)(sizeof(strlist) / sizeof(char *)); loop ++) {
"\\U0001d7f2\\U000e007d\\u0004\\u0589",
"\\u82ab\\u17e8\\u0736\\u2019\\U0001d64d",
"\\u0e01\\ub55c\\u0a68\\U000e0037\\u0cd6\\u002c\\ub959",
- "\\U000e0065\\u302c\\uc986\\u09ee\\U000e0068",
+ "\\U000e0065\\u302c\\u09ee\\U000e0068",
"\\u0be8\\u002e\\u0c68\\u066e\\u136d\\ufc99\\u59e7",
"\\u0233\\U000e0020\\u0a69\\u0d6a",
"\\u206f\\u0741\\ub3ab\\u2019\\ubcac\\u2019",
"\\u58f4\\U000e0049\\u20e7\\u2027",
- "\\ub315\\U0001d7e5\\U000e0073\\u0c47\\u06f2\\u0c6a\\u0037\\u10fe",
+ "\\U0001d7e5\\U000e0073\\u0c47\\u06f2\\u0c6a\\u0037\\u10fe",
"\\ua183\\u102d\\u0bec\\u003a",
"\\u17e8\\u06e7\\u002e\\u096d\\u003b",
"\\u003a\\u0e57\\u0fad\\u002e",
};
int loop;
if (U_FAILURE(status)) {
- errln("Creation of break iterator failed %s", u_errorName(status));
+ errcheckln(status, "Creation of break iterator failed %s", u_errorName(status));
return;
}
for (loop = 0; loop < (int)(sizeof(strlist) / sizeof(char *)); loop ++) {
};
int loop;
if (U_FAILURE(status)) {
- errln("Creation of break iterator failed %s", u_errorName(status));
+ errcheckln(status, "Creation of break iterator failed %s", u_errorName(status));
return;
}
for (loop = 0; loop < (int)(sizeof(strlist) / sizeof(char *)); loop ++) {
}
}
else {
- errln("Creation of character break iterator failed %s", u_errorName(status));
+ errcheckln(status, "Creation of character break iterator failed %s", u_errorName(status));
}
delete bi;
}
RunMonkey(bi, m, "word", seed, loopCount, useUText);
}
else {
- errln("Creation of word break iterator failed %s", u_errorName(status));
+ errcheckln(status, "Creation of word break iterator failed %s", u_errorName(status));
}
delete bi;
}
RunMonkey(bi, m, "line", seed, loopCount, useUText);
}
else {
- errln("Creation of line break iterator failed %s", u_errorName(status));
+ errcheckln(status, "Creation of line break iterator failed %s", u_errorName(status));
}
delete bi;
}
RunMonkey(bi, m, "sentence", seed, loopCount, useUText);
}
else {
- errln("Creation of line break iterator failed %s", u_errorName(status));
+ errcheckln(status, "Creation of line break iterator failed %s", u_errorName(status));
}
delete bi;
}
if (breakPos <= i ||
breakPos < lastBreakPos ||
breakPos > testText.length() ||
- breakPos > lastBreakPos && lastBreakPos > i ) {
+ (breakPos > lastBreakPos && lastBreakPos > i)) {
+ UChar32 brkChar = testText.char32At(lastBreakPos);
+ if ((strcmp(name, "char") != 0 && strcmp(name, "word") != 0) || brkChar < 0x1F1E6 || brkChar > 0x1F1FF) { // Apple, skip RI char/word break monkey tests
errln("%s break monkey test: "
"Out of range value returned by BreakIterator::following().\n"
"Random seed=%d index=%d; following returned %d; lastbreak=%d",
name, seed, i, breakPos, lastBreakPos);
+ }
break;
}
followingBreaks[breakPos] = 1;
breakPos = bi->preceding(i);
if (breakPos >= i ||
breakPos > lastBreakPos ||
- breakPos < 0 && testText.getChar32Start(i)>0 ||
- breakPos < lastBreakPos && lastBreakPos < testText.getChar32Start(i) ) {
+ (breakPos < 0 && testText.getChar32Start(i)>0) ||
+ (breakPos < lastBreakPos && lastBreakPos < testText.getChar32Start(i)) ) {
+ UChar32 brkChar = testText.char32At(breakPos);
+ if ((strcmp(name, "char") != 0 && strcmp(name, "word") != 0) || brkChar < 0x1F1E6 || brkChar > 0x1F1FF) { // Apple, skip RI char/word break monkey tests
errln("%s break monkey test: "
"Out of range value returned by BreakIterator::preceding().\n"
"index=%d; prev returned %d; lastBreak=%d" ,
if (breakPos >= 0 && breakPos < (int32_t)sizeof(precedingBreaks)) {
precedingBreaks[i] = 2; // Forces an error.
}
+ }
} else {
if (breakPos >= 0) {
precedingBreaks[breakPos] = 1;
UErrorCode status = U_ZERO_ERROR;
errorText.extract(charErrorTxt, sizeof(charErrorTxt), NULL, status);
charErrorTxt[sizeof(charErrorTxt)-1] = 0;
- errln("%s break monkey test error. %s. Operation = %s; Random seed = %d; buf Idx = %d\n%s",
- name, (expectedBreaks[i]? "break expected but not found" : "break found but not expected"),
+ const char *badLocale = bi->getLocaleID(ULOC_ACTUAL_LOCALE, status);
+
+ UChar32 brkChar = testText.char32At(i);
+ if ((strcmp(name, "char") != 0 && strcmp(name, "word") != 0) || brkChar < 0x1F1E6 || brkChar > 0x1F1FF) { // Apple, skip RI char/word break monkey tests
+ errln("%s break monkey test error [%s]. %s. Operation = %s; Random seed = %d; buf Idx = %d\n%s",
+ name, badLocale, (expectedBreaks[i]? "break expected but not found" : "break found but not expected"),
errorType, seed, i, charErrorTxt);
+ }
break;
}
}
#endif
}
+
+// Bug 5532. UTF-8 based UText fails in dictionary code.
+// This test checks the initial patch,
+// which is to just keep it from crashing. Correct word boundaries
+// await a proper fix to the dictionary code.
+//
+void RBBITest::TestBug5532(void) {
+ // Text includes a mixture of Thai and Latin.
+ const unsigned char utf8Data[] = {
+ 0xE0u, 0xB8u, 0x82u, 0xE0u, 0xB8u, 0xB2u, 0xE0u, 0xB8u, 0xA2u, 0xE0u,
+ 0xB9u, 0x80u, 0xE0u, 0xB8u, 0x84u, 0xE0u, 0xB8u, 0xA3u, 0xE0u, 0xB8u,
+ 0xB7u, 0xE0u, 0xB9u, 0x88u, 0xE0u, 0xB8u, 0xADu, 0xE0u, 0xB8u, 0x87u,
+ 0xE0u, 0xB9u, 0x80u, 0xE0u, 0xB8u, 0xA5u, 0xE0u, 0xB9u, 0x88u, 0xE0u,
+ 0xB8u, 0x99u, 0xE0u, 0xB8u, 0x8Bu, 0xE0u, 0xB8u, 0xB5u, 0xE0u, 0xB8u,
+ 0x94u, 0xE0u, 0xB8u, 0xB5u, 0x20u, 0x73u, 0x69u, 0x6Du, 0x20u, 0x61u,
+ 0x75u, 0x64u, 0x69u, 0x6Fu, 0x2Fu, 0x20u, 0x4Du, 0x4Fu, 0x4Fu, 0x4Eu,
+ 0x20u, 0x65u, 0x63u, 0x6Cu, 0x69u, 0x70u, 0x73u, 0x65u, 0x20u, 0xE0u,
+ 0xB8u, 0xA3u, 0xE0u, 0xB8u, 0xB2u, 0xE0u, 0xB8u, 0x84u, 0xE0u, 0xB8u,
+ 0xB2u, 0x20u, 0x34u, 0x37u, 0x30u, 0x30u, 0x20u, 0xE0u, 0xB8u, 0xA2u,
+ 0xE0u, 0xB8u, 0xB9u, 0xE0u, 0xB9u, 0x82u, 0xE0u, 0xB8u, 0xA3u, 0x00};
+
+ UErrorCode status = U_ZERO_ERROR;
+ UText utext=UTEXT_INITIALIZER;
+ utext_openUTF8(&utext, (const char *)utf8Data, -1, &status);
+ TEST_ASSERT_SUCCESS(status);
+
+ BreakIterator *bi = BreakIterator::createWordInstance(Locale("th"), status);
+ TEST_ASSERT_SUCCESS(status);
+ if (U_SUCCESS(status)) {
+ bi->setText(&utext, status);
+ TEST_ASSERT_SUCCESS(status);
+
+ int32_t breakCount = 0;
+ int32_t previousBreak = -1;
+ for (bi->first(); bi->next() != BreakIterator::DONE; breakCount++) {
+ // For now, just make sure that the break iterator doesn't hang.
+ TEST_ASSERT(previousBreak < bi->current());
+ previousBreak = bi->current();
+ }
+ TEST_ASSERT(breakCount > 0);
+ }
+ delete bi;
+ utext_close(&utext);
+}
+
+
+void RBBITest::TestBug9983(void) {
+ UnicodeString text = UnicodeString("\\u002A" // * Other
+ "\\uFF65" // Other
+ "\\u309C" // Katakana
+ "\\uFF9F" // Extend
+ "\\uFF65" // Other
+ "\\u0020" // Other
+ "\\u0000").unescape();
+
+ UErrorCode status = U_ZERO_ERROR;
+ LocalPointer<RuleBasedBreakIterator> brkiter(static_cast<RuleBasedBreakIterator *>(
+ BreakIterator::createWordInstance(Locale::getRoot(), status)));
+ TEST_ASSERT_SUCCESS(status);
+ if (U_FAILURE(status)) {
+ return;
+ }
+ brkiter->setText(text);
+ int32_t offset, rstatus;
+ brkiter->last();
+ int32_t iterationCount = 0;
+ while ( (offset = brkiter->previous()) != UBRK_DONE ) {
+ iterationCount++;
+ rstatus = brkiter->getRuleStatus();
+ // printf(" %d(%d)", offset, rstatus);
+ if (iterationCount >= 10) {
+ break;
+ }
+ }
+ TEST_ASSERT(iterationCount == 6);
+}
+
+
//
// TestDebug - A place-holder test for debugging purposes.
// For putting in fragments of other tests that can be invoked
#endif
}
+void RBBITest::TestProperties() {
+ UErrorCode errorCode = U_ZERO_ERROR;
+ UnicodeSet prependSet(UNICODE_STRING_SIMPLE("[:GCB=Prepend:]"), errorCode);
+ if (!prependSet.isEmpty()) {
+ errln(
+ "[:GCB=Prepend:] is not empty any more. "
+ "Uncomment relevant lines in source/data/brkitr/char.txt and "
+ "change this test to the opposite condition.");
+ }
+}
+
#endif /* #if !UCONFIG_NO_BREAK_ITERATION */
projects / apple / icu.git / blobdiff