source/test/intltest/ucdtest.cpp

   1 // Copyright (C) 2016 and later: Unicode, Inc. and others.
   2 // License & terms of use: http://www.unicode.org/copyright.html
   3 /********************************************************************
   4  * Copyright (c) 1997-2016, International Business Machines Corporation and
   5  * others. All Rights Reserved.
   6  ********************************************************************/
   7
   8 #include "unicode/ustring.h"
   9 #include "unicode/uchar.h"
  10 #include "unicode/uniset.h"
  11 #include "unicode/putil.h"
  12 #include "unicode/uscript.h"
  13 #include "cstring.h"
  14 #include "hash.h"
  15 #include "patternprops.h"
  16 #include "normalizer2impl.h"
  17 #include "uparse.h"
  18 #include "ucdtest.h"
  19
  20 static const char *ignorePropNames[]={
  21     "FC_NFKC",
  22     "NFD_QC",
  23     "NFC_QC",
  24     "NFKD_QC",
  25     "NFKC_QC",
  26     "Expands_On_NFD",
  27     "Expands_On_NFC",
  28     "Expands_On_NFKD",
  29     "Expands_On_NFKC",
  30     "NFKC_CF"
  31 };
  32
  33 UnicodeTest::UnicodeTest()
  34 {
  35     UErrorCode errorCode=U_ZERO_ERROR;
  36     unknownPropertyNames=new U_NAMESPACE_QUALIFIER Hashtable(errorCode);
  37     if(U_FAILURE(errorCode)) {
  38         delete unknownPropertyNames;
  39         unknownPropertyNames=NULL;
  40     }
  41     // Ignore some property names altogether.
  42     for(int32_t i=0; i<UPRV_LENGTHOF(ignorePropNames); ++i) {
  43         unknownPropertyNames->puti(UnicodeString(ignorePropNames[i], -1, US_INV), 1, errorCode);
  44     }
  45 }
  46
  47 UnicodeTest::~UnicodeTest()
  48 {
  49     delete unknownPropertyNames;
  50 }
  51
  52 void UnicodeTest::runIndexedTest( int32_t index, UBool exec, const char* &name, char* /*par*/ )
  53 {
  54     if(exec) {
  55         logln("TestSuite UnicodeTest: ");
  56     }
  57     TESTCASE_AUTO_BEGIN;
  58     TESTCASE_AUTO(TestAdditionalProperties);
  59     TESTCASE_AUTO(TestBinaryValues);
  60     TESTCASE_AUTO(TestConsistency);
  61     TESTCASE_AUTO(TestPatternProperties);
  62     TESTCASE_AUTO(TestScriptMetadata);
  63     TESTCASE_AUTO(TestBidiPairedBracketType);
  64     TESTCASE_AUTO(TestEmojiProperties);
  65     TESTCASE_AUTO_END;
  66 }
  67
  68 //====================================================
  69 // private data used by the tests
  70 //====================================================
  71
  72 // test DerivedCoreProperties.txt -------------------------------------------
  73
  74 // copied from genprops.c
  75 static int32_t
  76 getTokenIndex(const char *const tokens[], int32_t countTokens, const char *s) {
  77     const char *t, *z;
  78     int32_t i, j;
  79
  80     s=u_skipWhitespace(s);
  81     for(i=0; i<countTokens; ++i) {
  82         t=tokens[i];
  83         if(t!=NULL) {
  84             for(j=0;; ++j) {
  85                 if(t[j]!=0) {
  86                     if(s[j]!=t[j]) {
  87                         break;
  88                     }
  89                 } else {
  90                     z=u_skipWhitespace(s+j);
  91                     if(*z==';' || *z==0) {
  92                         return i;
  93                     } else {
  94                         break;
  95                     }
  96                 }
  97             }
  98         }
  99     }
 100     return -1;
 101 }
 102
 103 static const char *const
 104 derivedPropsNames[]={
 105     "Math",
 106     "Alphabetic",
 107     "Lowercase",
 108     "Uppercase",
 109     "ID_Start",
 110     "ID_Continue",
 111     "XID_Start",
 112     "XID_Continue",
 113     "Default_Ignorable_Code_Point",
 114     "Full_Composition_Exclusion",
 115     "Grapheme_Extend",
 116     "Grapheme_Link", /* Unicode 5 moves this property here from PropList.txt */
 117     "Grapheme_Base",
 118     "Cased",
 119     "Case_Ignorable",
 120     "Changes_When_Lowercased",
 121     "Changes_When_Uppercased",
 122     "Changes_When_Titlecased",
 123     "Changes_When_Casefolded",
 124     "Changes_When_Casemapped",
 125     "Changes_When_NFKC_Casefolded"
 126 };
 127
 128 static const UProperty
 129 derivedPropsIndex[]={
 130     UCHAR_MATH,
 131     UCHAR_ALPHABETIC,
 132     UCHAR_LOWERCASE,
 133     UCHAR_UPPERCASE,
 134     UCHAR_ID_START,
 135     UCHAR_ID_CONTINUE,
 136     UCHAR_XID_START,
 137     UCHAR_XID_CONTINUE,
 138     UCHAR_DEFAULT_IGNORABLE_CODE_POINT,
 139     UCHAR_FULL_COMPOSITION_EXCLUSION,
 140     UCHAR_GRAPHEME_EXTEND,
 141     UCHAR_GRAPHEME_LINK,
 142     UCHAR_GRAPHEME_BASE,
 143     UCHAR_CASED,
 144     UCHAR_CASE_IGNORABLE,
 145     UCHAR_CHANGES_WHEN_LOWERCASED,
 146     UCHAR_CHANGES_WHEN_UPPERCASED,
 147     UCHAR_CHANGES_WHEN_TITLECASED,
 148     UCHAR_CHANGES_WHEN_CASEFOLDED,
 149     UCHAR_CHANGES_WHEN_CASEMAPPED,
 150     UCHAR_CHANGES_WHEN_NFKC_CASEFOLDED
 151 };
 152
 153 static int32_t numErrors[UPRV_LENGTHOF(derivedPropsIndex)]={ 0 };
 154
 155 enum { MAX_ERRORS=50 };
 156
 157 U_CFUNC void U_CALLCONV
 158 derivedPropsLineFn(void *context,
 159                    char *fields[][2], int32_t /* fieldCount */,
 160                    UErrorCode *pErrorCode)
 161 {
 162     UnicodeTest *me=(UnicodeTest *)context;
 163     uint32_t start, end;
 164     int32_t i;
 165
 166     u_parseCodePointRange(fields[0][0], &start, &end, pErrorCode);
 167     if(U_FAILURE(*pErrorCode)) {
 168         me->errln("UnicodeTest: syntax error in DerivedCoreProperties.txt or DerivedNormalizationProps.txt field 0 at %s\n", fields[0][0]);
 169         return;
 170     }
 171
 172     /* parse derived binary property name, ignore unknown names */
 173     i=getTokenIndex(derivedPropsNames, UPRV_LENGTHOF(derivedPropsNames), fields[1][0]);
 174     if(i<0) {
 175         UnicodeString propName(fields[1][0], (int32_t)(fields[1][1]-fields[1][0]));
 176         propName.trim();
 177         if(me->unknownPropertyNames->find(propName)==NULL) {
 178             UErrorCode errorCode=U_ZERO_ERROR;
 179             me->unknownPropertyNames->puti(propName, 1, errorCode);
 180             me->errln("UnicodeTest warning: unknown property name '%s' in DerivedCoreProperties.txt or DerivedNormalizationProps.txt\n", fields[1][0]);
 181         }
 182         return;
 183     }
 184
 185     me->derivedProps[i].add(start, end);
 186 }
 187
 188 void UnicodeTest::TestAdditionalProperties() {
 189 #if !UCONFIG_NO_NORMALIZATION
 190     // test DerivedCoreProperties.txt and DerivedNormalizationProps.txt
 191     if(UPRV_LENGTHOF(derivedProps)<UPRV_LENGTHOF(derivedPropsNames)) {
 192         errln("error: UnicodeTest::derivedProps[] too short, need at least %d UnicodeSets\n",
 193               UPRV_LENGTHOF(derivedPropsNames));
 194         return;
 195     }
 196     if(UPRV_LENGTHOF(derivedPropsIndex)!=UPRV_LENGTHOF(derivedPropsNames)) {
 197         errln("error in ucdtest.cpp: UPRV_LENGTHOF(derivedPropsIndex)!=UPRV_LENGTHOF(derivedPropsNames)\n");
 198         return;
 199     }
 200
 201     char path[500];
 202     if(getUnidataPath(path) == NULL) {
 203         errln("unable to find path to source/data/unidata/");
 204         return;
 205     }
 206     char *basename=strchr(path, 0);
 207     strcpy(basename, "DerivedCoreProperties.txt");
 208
 209     char *fields[2][2];
 210     UErrorCode errorCode=U_ZERO_ERROR;
 211     u_parseDelimitedFile(path, ';', fields, 2, derivedPropsLineFn, this, &errorCode);
 212     if(U_FAILURE(errorCode)) {
 213         errln("error parsing DerivedCoreProperties.txt: %s\n", u_errorName(errorCode));
 214         return;
 215     }
 216
 217     strcpy(basename, "DerivedNormalizationProps.txt");
 218     u_parseDelimitedFile(path, ';', fields, 2, derivedPropsLineFn, this, &errorCode);
 219     if(U_FAILURE(errorCode)) {
 220         errln("error parsing DerivedNormalizationProps.txt: %s\n", u_errorName(errorCode));
 221         return;
 222     }
 223
 224     // now we have all derived core properties in the UnicodeSets
 225     // run them all through the API
 226     int32_t rangeCount, range;
 227     uint32_t i;
 228     UChar32 start, end;
 229
 230     // test all TRUE properties
 231     for(i=0; i<UPRV_LENGTHOF(derivedPropsNames); ++i) {
 232         rangeCount=derivedProps[i].getRangeCount();
 233         for(range=0; range<rangeCount && numErrors[i]<MAX_ERRORS; ++range) {
 234             start=derivedProps[i].getRangeStart(range);
 235             end=derivedProps[i].getRangeEnd(range);
 236             for(; start<=end; ++start) {
 237                 if(!u_hasBinaryProperty(start, derivedPropsIndex[i])) {
 238                     dataerrln("UnicodeTest error: u_hasBinaryProperty(U+%04lx, %s)==FALSE is wrong", start, derivedPropsNames[i]);
 239                     if(++numErrors[i]>=MAX_ERRORS) {
 240                       dataerrln("Too many errors, moving to the next test");
 241                       break;
 242                     }
 243                 }
 244             }
 245         }
 246     }
 247
 248     // invert all properties
 249     for(i=0; i<UPRV_LENGTHOF(derivedPropsNames); ++i) {
 250         derivedProps[i].complement();
 251     }
 252
 253     // test all FALSE properties
 254     for(i=0; i<UPRV_LENGTHOF(derivedPropsNames); ++i) {
 255         rangeCount=derivedProps[i].getRangeCount();
 256         for(range=0; range<rangeCount && numErrors[i]<MAX_ERRORS; ++range) {
 257             start=derivedProps[i].getRangeStart(range);
 258             end=derivedProps[i].getRangeEnd(range);
 259             for(; start<=end; ++start) {
 260                 if(u_hasBinaryProperty(start, derivedPropsIndex[i])) {
 261                     errln("UnicodeTest error: u_hasBinaryProperty(U+%04lx, %s)==TRUE is wrong\n", start, derivedPropsNames[i]);
 262                     if(++numErrors[i]>=MAX_ERRORS) {
 263                       errln("Too many errors, moving to the next test");
 264                       break;
 265                     }
 266                 }
 267             }
 268         }
 269     }
 270 #endif /* !UCONFIG_NO_NORMALIZATION */
 271 }
 272
 273 void UnicodeTest::TestBinaryValues() {
 274     /*
 275      * Unicode 5.1 explicitly defines binary property value aliases.
 276      * Verify that they are all recognized.
 277      */
 278     UErrorCode errorCode=U_ZERO_ERROR;
 279     UnicodeSet alpha(UNICODE_STRING_SIMPLE("[:Alphabetic:]"), errorCode);
 280     if(U_FAILURE(errorCode)) {
 281         dataerrln("UnicodeSet([:Alphabetic:]) failed - %s", u_errorName(errorCode));
 282         return;
 283     }
 284
 285     static const char *const falseValues[]={ "N", "No", "F", "False" };
 286     static const char *const trueValues[]={ "Y", "Yes", "T", "True" };
 287     int32_t i;
 288     for(i=0; i<UPRV_LENGTHOF(falseValues); ++i) {
 289         UnicodeString pattern=UNICODE_STRING_SIMPLE("[:Alphabetic=:]");
 290         pattern.insert(pattern.length()-2, UnicodeString(falseValues[i], -1, US_INV));
 291         errorCode=U_ZERO_ERROR;
 292         UnicodeSet set(pattern, errorCode);
 293         if(U_FAILURE(errorCode)) {
 294             errln("UnicodeSet([:Alphabetic=%s:]) failed - %s\n", falseValues[i], u_errorName(errorCode));
 295             continue;
 296         }
 297         set.complement();
 298         if(set!=alpha) {
 299             errln("UnicodeSet([:Alphabetic=%s:]).complement()!=UnicodeSet([:Alphabetic:])\n", falseValues[i]);
 300         }
 301     }
 302     for(i=0; i<UPRV_LENGTHOF(trueValues); ++i) {
 303         UnicodeString pattern=UNICODE_STRING_SIMPLE("[:Alphabetic=:]");
 304         pattern.insert(pattern.length()-2, UnicodeString(trueValues[i], -1, US_INV));
 305         errorCode=U_ZERO_ERROR;
 306         UnicodeSet set(pattern, errorCode);
 307         if(U_FAILURE(errorCode)) {
 308             errln("UnicodeSet([:Alphabetic=%s:]) failed - %s\n", trueValues[i], u_errorName(errorCode));
 309             continue;
 310         }
 311         if(set!=alpha) {
 312             errln("UnicodeSet([:Alphabetic=%s:])!=UnicodeSet([:Alphabetic:])\n", trueValues[i]);
 313         }
 314     }
 315 }
 316
 317 void UnicodeTest::TestConsistency() {
 318 #if !UCONFIG_NO_NORMALIZATION
 319     /*
 320      * Test for an example that getCanonStartSet() delivers
 321      * all characters that compose from the input one,
 322      * even in multiple steps.
 323      * For example, the set for "I" (0049) should contain both
 324      * I-diaeresis (00CF) and I-diaeresis-acute (1E2E).
 325      * In general, the set for the middle such character should be a subset
 326      * of the set for the first.
 327      */
 328     IcuTestErrorCode errorCode(*this, "TestConsistency");
 329     const Normalizer2 *nfd=Normalizer2::getNFDInstance(errorCode);
 330     const Normalizer2Impl *nfcImpl=Normalizer2Factory::getNFCImpl(errorCode);
 331     if(!nfcImpl->ensureCanonIterData(errorCode) || errorCode.isFailure()) {
 332         dataerrln("Normalizer2::getInstance(NFD) or Normalizer2Factory::getNFCImpl() failed - %s\n",
 333                   errorCode.errorName());
 334         errorCode.reset();
 335         return;
 336     }
 337
 338     UnicodeSet set1, set2;
 339     if (nfcImpl->getCanonStartSet(0x49, set1)) {
 340         /* enumerate all characters that are plausible to be latin letters */
 341         for(UChar start=0xa0; start<0x2000; ++start) {
 342             UnicodeString decomp=nfd->normalize(UnicodeString(start), errorCode);
 343             if(decomp.length()>1 && decomp[0]==0x49) {
 344                 set2.add(start);
 345             }
 346         }
 347
 348         if (set1!=set2) {
 349             errln("[canon start set of 0049] != [all c with canon decomp with 0049]");
 350         }
 351         // This was available in cucdtst.c but the test had to move to intltest
 352         // because the new internal normalization functions are in C++.
 353         //compareUSets(set1, set2,
 354         //             "[canon start set of 0049]", "[all c with canon decomp with 0049]",
 355         //             TRUE);
 356     } else {
 357         errln("NFC.getCanonStartSet() returned FALSE");
 358     }
 359 #endif
 360 }
 361
 362 /**
 363  * Test various implementations of Pattern_Syntax & Pattern_White_Space.
 364  */
 365 void UnicodeTest::TestPatternProperties() {
 366     IcuTestErrorCode errorCode(*this, "TestPatternProperties()");
 367     UnicodeSet syn_pp;
 368     UnicodeSet syn_prop(UNICODE_STRING_SIMPLE("[:Pattern_Syntax:]"), errorCode);
 369     UnicodeSet syn_list(
 370         "[!-/\\:-@\\[-\\^`\\{-~"
 371         "\\u00A1-\\u00A7\\u00A9\\u00AB\\u00AC\\u00AE\\u00B0\\u00B1\\u00B6\\u00BB\\u00BF\\u00D7\\u00F7"
 372         "\\u2010-\\u2027\\u2030-\\u203E\\u2041-\\u2053\\u2055-\\u205E\\u2190-\\u245F\\u2500-\\u2775"
 373         "\\u2794-\\u2BFF\\u2E00-\\u2E7F\\u3001-\\u3003\\u3008-\\u3020\\u3030\\uFD3E\\uFD3F\\uFE45\\uFE46]", errorCode);
 374     UnicodeSet ws_pp;
 375     UnicodeSet ws_prop(UNICODE_STRING_SIMPLE("[:Pattern_White_Space:]"), errorCode);
 376     UnicodeSet ws_list(UNICODE_STRING_SIMPLE("[\\u0009-\\u000D\\ \\u0085\\u200E\\u200F\\u2028\\u2029]"), errorCode);
 377     UnicodeSet syn_ws_pp;
 378     UnicodeSet syn_ws_prop(syn_prop);
 379     syn_ws_prop.addAll(ws_prop);
 380     for(UChar32 c=0; c<=0xffff; ++c) {
 381         if(PatternProps::isSyntax(c)) {
 382             syn_pp.add(c);
 383         }
 384         if(PatternProps::isWhiteSpace(c)) {
 385             ws_pp.add(c);
 386         }
 387         if(PatternProps::isSyntaxOrWhiteSpace(c)) {
 388             syn_ws_pp.add(c);
 389         }
 390     }
 391     compareUSets(syn_pp, syn_prop,
 392                  "PatternProps.isSyntax()", "[:Pattern_Syntax:]", TRUE);
 393     compareUSets(syn_pp, syn_list,
 394                  "PatternProps.isSyntax()", "[Pattern_Syntax ranges]", TRUE);
 395     compareUSets(ws_pp, ws_prop,
 396                  "PatternProps.isWhiteSpace()", "[:Pattern_White_Space:]", TRUE);
 397     compareUSets(ws_pp, ws_list,
 398                  "PatternProps.isWhiteSpace()", "[Pattern_White_Space ranges]", TRUE);
 399     compareUSets(syn_ws_pp, syn_ws_prop,
 400                  "PatternProps.isSyntaxOrWhiteSpace()",
 401                  "[[:Pattern_Syntax:][:Pattern_White_Space:]]", TRUE);
 402 }
 403
 404 // So far only minimal port of Java & cucdtst.c compareUSets().
 405 UBool
 406 UnicodeTest::compareUSets(const UnicodeSet &a, const UnicodeSet &b,
 407                           const char *a_name, const char *b_name,
 408                           UBool diffIsError) {
 409     UBool same= a==b;
 410     if(!same && diffIsError) {
 411         errln("Sets are different: %s vs. %s\n", a_name, b_name);
 412     }
 413     return same;
 414 }
 415
 416 namespace {
 417
 418 /**
 419  * Maps a special script code to the most common script of its encoded characters.
 420  */
 421 UScriptCode getCharScript(UScriptCode script) {
 422     switch(script) {
 423     case USCRIPT_HAN_WITH_BOPOMOFO:
 424     case USCRIPT_SIMPLIFIED_HAN:
 425     case USCRIPT_TRADITIONAL_HAN:
 426         return USCRIPT_HAN;
 427     case USCRIPT_JAPANESE:
 428         return USCRIPT_HIRAGANA;
 429     case USCRIPT_JAMO:
 430     case USCRIPT_KOREAN:
 431         return USCRIPT_HANGUL;
 432     case USCRIPT_SYMBOLS_EMOJI:
 433         return USCRIPT_SYMBOLS;
 434     default:
 435         return script;
 436     }
 437 }
 438
 439 }  // namespace
 440
 441 void UnicodeTest::TestScriptMetadata() {
 442     IcuTestErrorCode errorCode(*this, "TestScriptMetadata()");
 443     UnicodeSet rtl("[[:bc=R:][:bc=AL:]-[:Cn:]-[:sc=Common:]]", errorCode);
 444     // So far, sample characters are uppercase.
 445     // Georgian is special.
 446     UnicodeSet cased("[[:Lu:]-[:sc=Common:]-[:sc=Geor:]]", errorCode);
 447     for(int32_t sci = 0; sci < USCRIPT_CODE_LIMIT; ++sci) {
 448         UScriptCode sc = (UScriptCode)sci;
 449         // Run the test with -v to see which script has failures:
 450         // .../intltest$ make && ./intltest utility/UnicodeTest/TestScriptMetadata -v | grep -C 6 FAIL
 451         logln(uscript_getShortName(sc));
 452         UScriptUsage usage = uscript_getUsage(sc);
 453         UnicodeString sample = uscript_getSampleUnicodeString(sc);
 454         UnicodeSet scriptSet;
 455         scriptSet.applyIntPropertyValue(UCHAR_SCRIPT, sc, errorCode);
 456         if(usage == USCRIPT_USAGE_NOT_ENCODED) {
 457             assertTrue("not encoded, no sample", sample.isEmpty());
 458             assertFalse("not encoded, not RTL", uscript_isRightToLeft(sc));
 459             assertFalse("not encoded, not LB letters", uscript_breaksBetweenLetters(sc));
 460             assertFalse("not encoded, not cased", uscript_isCased(sc));
 461             assertTrue("not encoded, no characters", scriptSet.isEmpty());
 462         } else {
 463             assertFalse("encoded, has a sample character", sample.isEmpty());
 464             UChar32 firstChar = sample.char32At(0);
 465             UScriptCode charScript = getCharScript(sc);
 466             assertEquals("script(sample(script))",
 467                          (int32_t)charScript, (int32_t)uscript_getScript(firstChar, errorCode));
 468             assertEquals("RTL vs. set", (UBool)rtl.contains(firstChar), (UBool)uscript_isRightToLeft(sc));
 469             assertEquals("cased vs. set", (UBool)cased.contains(firstChar), (UBool)uscript_isCased(sc));
 470             assertEquals("encoded, has characters", (UBool)(sc == charScript), (UBool)(!scriptSet.isEmpty()));
 471             if(uscript_isRightToLeft(sc)) {
 472                 rtl.removeAll(scriptSet);
 473             }
 474             if(uscript_isCased(sc)) {
 475                 cased.removeAll(scriptSet);
 476             }
 477         }
 478     }
 479     UnicodeString pattern;
 480     assertEquals("no remaining RTL characters",
 481                  UnicodeString("[]"), rtl.toPattern(pattern));
 482     assertEquals("no remaining cased characters",
 483                  UnicodeString("[]"), cased.toPattern(pattern));
 484
 485     assertTrue("Hani breaks between letters", uscript_breaksBetweenLetters(USCRIPT_HAN));
 486     assertTrue("Thai breaks between letters", uscript_breaksBetweenLetters(USCRIPT_THAI));
 487     assertFalse("Latn does not break between letters", uscript_breaksBetweenLetters(USCRIPT_LATIN));
 488 }
 489
 490 void UnicodeTest::TestBidiPairedBracketType() {
 491     // BidiBrackets-6.3.0.txt says:
 492     //
 493     // The set of code points listed in this file was originally derived
 494     // using the character properties General_Category (gc), Bidi_Class (bc),
 495     // Bidi_Mirrored (Bidi_M), and Bidi_Mirroring_Glyph (bmg), as follows:
 496     // two characters, A and B, form a pair if A has gc=Ps and B has gc=Pe,
 497     // both have bc=ON and Bidi_M=Y, and bmg of A is B. Bidi_Paired_Bracket
 498     // maps A to B and vice versa, and their Bidi_Paired_Bracket_Type
 499     // property values are Open and Close, respectively.
 500     IcuTestErrorCode errorCode(*this, "TestBidiPairedBracketType()");
 501     UnicodeSet bpt("[:^bpt=n:]", errorCode);
 502     assertTrue("bpt!=None is not empty", !bpt.isEmpty());
 503     // The following should always be true.
 504     UnicodeSet mirrored("[:Bidi_M:]", errorCode);
 505     UnicodeSet other_neutral("[:bc=ON:]", errorCode);
 506     assertTrue("bpt!=None is a subset of Bidi_M", mirrored.containsAll(bpt));
 507     assertTrue("bpt!=None is a subset of bc=ON", other_neutral.containsAll(bpt));
 508     // The following are true at least initially in Unicode 6.3.
 509     UnicodeSet bpt_open("[:bpt=o:]", errorCode);
 510     UnicodeSet bpt_close("[:bpt=c:]", errorCode);
 511     UnicodeSet ps("[:Ps:]", errorCode);
 512     UnicodeSet pe("[:Pe:]", errorCode);
 513     assertTrue("bpt=Open is a subset of Ps", ps.containsAll(bpt_open));
 514     assertTrue("bpt=Close is a subset of Pe", pe.containsAll(bpt_close));
 515 }
 516
 517 void UnicodeTest::TestEmojiProperties() {
 518     assertFalse("space is not Emoji", u_hasBinaryProperty(0x20, UCHAR_EMOJI));
 519     assertTrue("shooting star is Emoji", u_hasBinaryProperty(0x1F320, UCHAR_EMOJI));
 520     IcuTestErrorCode errorCode(*this, "TestEmojiProperties()");
 521     UnicodeSet emoji("[:Emoji:]", errorCode);
 522     assertTrue("lots of Emoji", emoji.size() > 700);
 523
 524     assertTrue("shooting star is Emoji_Presentation",
 525                u_hasBinaryProperty(0x1F320, UCHAR_EMOJI_PRESENTATION));
 526     assertTrue("Fitzpatrick 6 is Emoji_Modifier",
 527                u_hasBinaryProperty(0x1F3FF, UCHAR_EMOJI_MODIFIER));
 528     assertTrue("happy person is Emoji_Modifier_Base",
 529                u_hasBinaryProperty(0x1F64B, UCHAR_EMOJI_MODIFIER_BASE));
 530 }