C++ U_FAILURE函数代码示例

U_CAPI int32_t U_EXPORT2
uspoof_areConfusableUnicodeString(const USpoofChecker *sc,
                                  const icu::UnicodeString &id1,
                                  const icu::UnicodeString &id2,
                                  UErrorCode *status) {
    const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
    if (U_FAILURE(*status)) {
        return 0;
    }
    //
    // See section 4 of UAX 39 for the algorithm for checking whether two strings are confusable,
    //   and for definitions of the types (single, whole, mixed-script) of confusables.

    // We only care about a few of the check flags.  Ignore the others.
    // If no tests relavant to this function have been specified, return an error.
    // TODO:  is this really the right thing to do?  It's probably an error on the caller's part,
    //        but logically we would just return 0 (no error).
    if ((This->fChecks & USPOOF_CONFUSABLE) == 0) {
        *status = U_INVALID_STATE_ERROR;
        return 0;
    }

    // Compute the skeletons and check for confusability.
    UnicodeString id1Skeleton;
    uspoof_getSkeletonUnicodeString(sc, 0 /* deprecated */, id1, id1Skeleton, status);
    UnicodeString id2Skeleton;
    uspoof_getSkeletonUnicodeString(sc, 0 /* deprecated */, id2, id2Skeleton, status);
    if (U_FAILURE(*status)) { return 0; }
    if (id1Skeleton != id2Skeleton) {
        return 0;
    }

    // If we get here, the strings are confusable.  Now we just need to set the flags for the appropriate classes
    // of confusables according to UTS 39 section 4.
    // Start by computing the resolved script sets of id1 and id2.
    ScriptSet id1RSS;
    This->getResolvedScriptSet(id1, id1RSS, *status);
    ScriptSet id2RSS;
    This->getResolvedScriptSet(id2, id2RSS, *status);

    // Turn on all applicable flags
    int32_t result = 0;
    if (id1RSS.intersects(id2RSS)) {
        result |= USPOOF_SINGLE_SCRIPT_CONFUSABLE;
    } else {
        result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
        if (!id1RSS.isEmpty() && !id2RSS.isEmpty()) {
            result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
        }
    }

    // Turn off flags that the user doesn't want
    if ((This->fChecks & USPOOF_SINGLE_SCRIPT_CONFUSABLE) == 0) {
        result &= ~USPOOF_SINGLE_SCRIPT_CONFUSABLE;
    }
    if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) == 0) {
        result &= ~USPOOF_MIXED_SCRIPT_CONFUSABLE;
    }
    if ((This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE) == 0) {
        result &= ~USPOOF_WHOLE_SCRIPT_CONFUSABLE;
    }

    return result;
}

//.........这里部分代码省略.........
			// Flag that we need to send the cookies, because we have set at least one.
			requireSend = true;
		}
	}
	
	if( requireSend )
	{
		cookies.sendCookies( req );
	}
	
	// Lazy init code.
	if( spepConfigData->isLazyInit() )
	{
		
		std::string globalESOECookieName( spepConfigData->getGlobalESOECookieName() );
		Cookies cookies( req );
		std::vector<std::string> cookieValues;
		cookies.getCookieValuesByName( cookieValues, globalESOECookieName );
		if( cookieValues.empty() ) {
			bool matchedLazyInitResource = false;
			UnicodeString properURIUnicode( spep::UnicodeStringConversion::toUnicodeString( properURI ) );
			
			std::vector<UnicodeString>::const_iterator lazyInitResourceIterator;
			for( lazyInitResourceIterator = spepConfigData->getLazyInitResources().begin();
				lazyInitResourceIterator != spepConfigData->getLazyInitResources().end();
				++lazyInitResourceIterator )
			{
				// TODO Opportunity for caching of compiled regex patterns is here.
				UParseError parseError;
				UErrorCode errorCode = U_ZERO_ERROR;
				// Perform the regular expression matching here.
				UBool result = RegexPattern::matches( *lazyInitResourceIterator, properURIUnicode, parseError, errorCode );
				
				if ( U_FAILURE( errorCode ) )
				{
					// TODO throw u_errorName( errorCode )
					return HTTP_INTERNAL_SERVER_ERROR;
				}
				
				// FALSE is defined by ICU. This line for portability.
				if (result != FALSE)
				{
					matchedLazyInitResource = true;
					break;
				}
			}
			
			if( matchedLazyInitResource )
			{
				if( !spepConfigData->isLazyInitDefaultPermit() )
				{
					return DECLINED;
				}
			}
			else
			{
				if( spepConfigData->isLazyInitDefaultPermit() )
				{
					return DECLINED;
				}
			}
		}
	}
	
	boost::posix_time::ptime epoch( boost::gregorian::date( 1970, 1, 1 ) );
	boost::posix_time::time_duration timestamp = boost::posix_time::microsec_clock::local_time() - epoch;

void
Package::writePackage(const char *filename, char outType, const char *comment) {
    char prefix[MAX_PKG_NAME_LENGTH+4];
    UDataOffsetTOCEntry entry;
    UDataSwapper *dsLocalToOut, *ds[TYPE_COUNT];
    FILE *file;
    Item *pItem;
    char *name;
    UErrorCode errorCode;
    int32_t i, length, prefixLength, maxItemLength, basenameOffset, offset, outInt32;
    uint8_t outCharset;
    UBool outIsBigEndian;

    extractPackageName(filename, prefix, MAX_PKG_NAME_LENGTH);

    // if there is an explicit comment, then use it, else use what's in the current header
    if(comment!=NULL) {
        /* get the header size minus the current comment */
        DataHeader *pHeader;
        int32_t length;

        pHeader=(DataHeader *)header;
        headerLength=4+pHeader->info.size;
        length=(int32_t)strlen(comment);
        if((int32_t)(headerLength+length)>=(int32_t)sizeof(header)) {
            fprintf(stderr, "icupkg: comment too long\n");
            exit(U_BUFFER_OVERFLOW_ERROR);
        }
        memcpy(header+headerLength, comment, length+1);
        headerLength+=length;
        if(headerLength&0xf) {
            /* NUL-pad the header to a multiple of 16 */
            length=(headerLength+0xf)&~0xf;
            memset(header+headerLength, 0, length-headerLength);
            headerLength=length;
        }
        pHeader->dataHeader.headerSize=(uint16_t)headerLength;
    }

    makeTypeProps(outType, outCharset, outIsBigEndian);

    // open (TYPE_COUNT-2) swappers
    // one is a no-op for local type==outType
    // one type (TYPE_LE) is bogus
    errorCode=U_ZERO_ERROR;
    i=makeTypeEnum(outType);
    ds[TYPE_B]= i==TYPE_B ? NULL : udata_openSwapper(TRUE, U_ASCII_FAMILY, outIsBigEndian, outCharset, &errorCode);
    ds[TYPE_L]= i==TYPE_L ? NULL : udata_openSwapper(FALSE, U_ASCII_FAMILY, outIsBigEndian, outCharset, &errorCode);
    ds[TYPE_LE]=NULL;
    ds[TYPE_E]= i==TYPE_E ? NULL : udata_openSwapper(TRUE, U_EBCDIC_FAMILY, outIsBigEndian, outCharset, &errorCode);
    if(U_FAILURE(errorCode)) {
        fprintf(stderr, "icupkg: udata_openSwapper() failed - %s\n", u_errorName(errorCode));
        exit(errorCode);
    }
    for(i=0; i<TYPE_COUNT; ++i) {
        if(ds[i]!=NULL) {
            ds[i]->printError=printPackageError;
            ds[i]->printErrorContext=stderr;
        }
    }

    dsLocalToOut=ds[makeTypeEnum(U_CHARSET_FAMILY, U_IS_BIG_ENDIAN)];

    // create the file and write its contents
    file=fopen(filename, "wb");
    if(file==NULL) {
        fprintf(stderr, "icupkg: unable to create file \"%s\"\n", filename);
        exit(U_FILE_ACCESS_ERROR);
    }

    // swap and write the header
    if(dsLocalToOut!=NULL) {
        udata_swapDataHeader(dsLocalToOut, header, headerLength, header, &errorCode);
        if(U_FAILURE(errorCode)) {
            fprintf(stderr, "icupkg: udata_swapDataHeader(local to out) failed - %s\n", u_errorName(errorCode));
            exit(errorCode);
        }
    }
    length=(int32_t)fwrite(header, 1, headerLength, file);
    if(length!=headerLength) {
        fprintf(stderr, "icupkg: unable to write complete header to file \"%s\"\n", filename);
        exit(U_FILE_ACCESS_ERROR);
    }

    // prepare and swap the package name with a tree separator
    // for prepending to item names
    strcat(prefix, U_TREE_ENTRY_SEP_STRING);
    prefixLength=(int32_t)strlen(prefix);
    if(dsLocalToOut!=NULL) {
        dsLocalToOut->swapInvChars(dsLocalToOut, prefix, prefixLength, prefix, &errorCode);
        if(U_FAILURE(errorCode)) {
            fprintf(stderr, "icupkg: swapInvChars(output package name) failed - %s\n", u_errorName(errorCode));
            exit(errorCode);
        }

        // swap and sort the item names (sorting needs to be done in the output charset)
        dsLocalToOut->swapInvChars(dsLocalToOut, inStrings, inStringTop, inStrings, &errorCode);
        if(U_FAILURE(errorCode)) {
            fprintf(stderr, "icupkg: swapInvChars(item names) failed - %s\n", u_errorName(errorCode));
            exit(errorCode);
//.........这里部分代码省略.........

static UBool U_CALLCONV
loadData(UStringPrepProfile* profile, 
         const char* path, 
         const char* name, 
         const char* type, 
         UErrorCode* errorCode) {
    /* load Unicode SPREP data from file */    
    UTrie _sprepTrie={ 0,0,0,0,0,0,0 };
    UDataMemory *dataMemory;
    const int32_t *p=NULL;
    const uint8_t *pb;
    UVersionInfo normUnicodeVersion;
    int32_t normUniVer, sprepUniVer, normCorrVer;

    if(errorCode==NULL || U_FAILURE(*errorCode)) {
        return 0;
    }

    /* open the data outside the mutex block */
    //TODO: change the path
    dataMemory=udata_openChoice(path, type, name, isSPrepAcceptable, NULL, errorCode);
    if(U_FAILURE(*errorCode)) {
        return FALSE;
    }

    p=(const int32_t *)udata_getMemory(dataMemory);
    pb=(const uint8_t *)(p+_SPREP_INDEX_TOP);
    utrie_unserialize(&_sprepTrie, pb, p[_SPREP_INDEX_TRIE_SIZE], errorCode);
    _sprepTrie.getFoldingOffset=getSPrepFoldingOffset;


    if(U_FAILURE(*errorCode)) {
        udata_close(dataMemory);
        return FALSE;
    }

    /* in the mutex block, set the data for this process */
    umtx_lock(&usprepMutex);
    if(profile->sprepData==NULL) {
        profile->sprepData=dataMemory;
        dataMemory=NULL;
        uprv_memcpy(&profile->indexes, p, sizeof(profile->indexes));
        uprv_memcpy(&profile->sprepTrie, &_sprepTrie, sizeof(UTrie));
    } else {
        p=(const int32_t *)udata_getMemory(profile->sprepData);
    }
    umtx_unlock(&usprepMutex);
    /* initialize some variables */
    profile->mappingData=(uint16_t *)((uint8_t *)(p+_SPREP_INDEX_TOP)+profile->indexes[_SPREP_INDEX_TRIE_SIZE]);
    
    u_getUnicodeVersion(normUnicodeVersion);
    normUniVer = (normUnicodeVersion[0] << 24) + (normUnicodeVersion[1] << 16) + 
                 (normUnicodeVersion[2] << 8 ) + (normUnicodeVersion[3]);
    sprepUniVer = (dataVersion[0] << 24) + (dataVersion[1] << 16) + 
                  (dataVersion[2] << 8 ) + (dataVersion[3]);
    normCorrVer = profile->indexes[_SPREP_NORM_CORRECTNS_LAST_UNI_VERSION];
    
    if(U_FAILURE(*errorCode)){
        udata_close(dataMemory);
        return FALSE;
    }
    if( normUniVer < sprepUniVer && /* the Unicode version of SPREP file must be less than the Unicode Vesion of the normalization data */
        normUniVer < normCorrVer && /* the Unicode version of the NormalizationCorrections.txt file should be less than the Unicode Vesion of the normalization data */
        ((profile->indexes[_SPREP_OPTIONS] & _SPREP_NORMALIZATION_ON) > 0) /* normalization turned on*/
      ){
        *errorCode = U_INVALID_FORMAT_ERROR;
        udata_close(dataMemory);
        return FALSE;
    }
    profile->isDataLoaded = TRUE;

    /* if a different thread set it first, then close the extra data */
    if(dataMemory!=NULL) {
        udata_close(dataMemory); /* NULL if it was set correctly */
    }


    return profile->isDataLoaded;
}

int main(int /* argc*/ , const char * /*argv*/ []) {
    UErrorCode status = U_ZERO_ERROR;
    int diffs = 0;
    int gbaddiffs =0;
    setup(status);
    if(U_FAILURE(status)) return 1;

    int expected = PROVIDER_COUNT;

    for(int l=0;l<LOCALE_COUNT;l++) {
        printf("\n");
        uint8_t oldBytes[200];
        int32_t oldLen = -1;
        for(int v=0;v<=expected;v++) {

            // Construct the locale ID
            char locID[200];
            strcpy(locID, locale[l]);
            if((v!=expected)) { // -1 = no version
                strcat(locID, "@sp=icu");
                strcat(locID, provider_version[v]);
            }
            
            printf("%-28s =  ", locID);
            
            UErrorCode subStatus = U_ZERO_ERROR;
            uint8_t bytes[200];
            uint8_t bytesb[200];
#define USE_CXX 0

#if USE_CXX
            Collator *col = Collator::createInstance(Locale(locID),subStatus);
            if(U_FAILURE(subStatus)) {
                printf("ERR: %s\n", u_errorName(subStatus));
                continue;
            }
            int32_t len = col->getSortKey(stuff, -1, bytes, 200);
#else
#if 1
            char xbuf2[200];
            strcpy(xbuf2,"X/");
            strcat(xbuf2,locID);
            strcat(xbuf2,"/");
            //printf(" -> %s\n", xbuf2);
            UCollator *col = ucol_openFromShortString(xbuf2, FALSE,NULL, &subStatus);
#else
            UCollator *col = ucol_open(locID, &subStatus);
#endif
            if(U_FAILURE(subStatus)) {
                printf("ERR: %s\n", u_errorName(subStatus));
                continue;
            }
            

            char xbuf3[200];
            {
              int32_t def = ucol_getShortDefinitionString(col,locID/*NULL*/,xbuf3,200,&subStatus);
              if(U_FAILURE(subStatus)) {
                printf("Err getting short string name: %s\n", u_errorName(subStatus));
              } else {
                printf(" --> %s\n", xbuf3);
              }              
            }

            int32_t len = ucol_getSortKey(col, stuff, -1, bytes, 200);
#endif

            printf("     ");

            int tdiffs=0;

            for(int i=0;i<len;i++) {
	      if(i<oldLen&&bytes[i]!=oldBytes[i]) {
                diffs++;
                printf("*");
              } else {
                printf(" ");
              }
              printf("%02X", (0xFF&bytes[i]));
            }
            printf("\n");

            char xbuf4[200];
            UCollator *col2 = ucol_openFromShortString(xbuf3, FALSE, NULL, &subStatus);
            if(U_FAILURE(subStatus)) {
              printf("Err opening from new short string : %s\n", u_errorName(subStatus));
              continue;
            } else {
              int32_t def4 = ucol_getShortDefinitionString(col,locID/*NULL*/,xbuf4,200,&subStatus);
              if(strcmp(xbuf4,xbuf3)) {
                printf(" --> reopened = %s (%s)\n", xbuf4, u_errorName(subStatus));
              }
            }
            int32_t len2 = ucol_getSortKey(col2, stuff, -1, bytesb, 200);

            int baddiffs=0;
            for(int i=0;i<len;i++) {
	      if(i<len&&bytes[i]!=bytesb[i]) {
                  baddiffs++;
                  printf("!");
//.........这里部分代码省略.........

bool f_intl_is_failure(int64 error_code) {
  if (U_FAILURE((UErrorCode)error_code)) return true;
  return false;
}

Variant php_intl_idn_to(CStrRef domain, Variant errorcode, int mode) {
  long option = 0;
  UChar* ustring = NULL;
  int ustring_len = 0;
  UErrorCode status;
  char     *converted_utf8 = NULL;
  int32_t   converted_utf8_len;
  UChar*    converted = NULL;
  int32_t   converted_ret_len;

  // Convert the string to UTF-16
  status = U_ZERO_ERROR;
  intl_convert_utf8_to_utf16(&ustring, &ustring_len,
      (char*)domain.data(), domain.size(), &status);
  if (U_FAILURE(status)) {
    free(ustring);
    errorcode = status;
    return false;
  }

  // Call the appropriate IDN function
  int converted_len = (ustring_len > 1) ? ustring_len : 1;
  for (;;) {
    UParseError parse_error;
    status = U_ZERO_ERROR;
    converted = (UChar*)malloc(sizeof(UChar)*converted_len);
    // If the malloc failed, bail out
    if (!converted) {
      free(ustring);
      errorcode = U_MEMORY_ALLOCATION_ERROR;
      return false;
    }
    if (mode == INTL_IDN_TO_ASCII) {
      converted_ret_len = uidna_IDNToASCII(ustring,
          ustring_len, converted, converted_len,
          (int32_t)option, &parse_error, &status);
    } else {
      converted_ret_len = uidna_IDNToUnicode(ustring,
          ustring_len, converted, converted_len,
          (int32_t)option, &parse_error, &status);
    }
    if (status != U_BUFFER_OVERFLOW_ERROR)
      break;
    // If we have a buffer overflow error, try again with a larger buffer
    free(converted);
    converted = NULL;
    converted_len = converted_len * 2;
  }
  free(ustring);
  if (U_FAILURE(status)) {
    free(converted);
    errorcode = status;
    return false;
  }

  // Convert the string back to UTF-8
  status = U_ZERO_ERROR;
  intl_convert_utf16_to_utf8(&converted_utf8, &converted_utf8_len,
      converted, converted_ret_len, &status);
  free(converted);
  if (U_FAILURE(status)) {
    free(converted_utf8);
    errorcode = status;
    return false;
  }

  // Return the string
  return String(converted_utf8, converted_utf8_len, AttachString);
}

void IntlPluralRules::initializePluralRules(ExecState& exec, JSValue locales, JSValue optionsValue)
{
    VM& vm = exec.vm();
    auto scope = DECLARE_THROW_SCOPE(vm);

    // 13.1.1 InitializePluralRules (pluralRules, locales, options)
    // https://tc39.github.io/ecma402/#sec-initializepluralrules
    Vector<String> requestedLocales = canonicalizeLocaleList(exec, locales);
    RETURN_IF_EXCEPTION(scope, void());

    JSObject* options;
    if (optionsValue.isUndefined())
        options = constructEmptyObject(&exec, exec.lexicalGlobalObject()->nullPrototypeObjectStructure());
    else {
        options = optionsValue.toObject(&exec);
        RETURN_IF_EXCEPTION(scope, void());
    }

    HashMap<String, String> localeOpt;
    String localeMatcher = intlStringOption(exec, options, vm.propertyNames->localeMatcher, { "lookup", "best fit" }, "localeMatcher must be either \"lookup\" or \"best fit\"", "best fit");
    RETURN_IF_EXCEPTION(scope, void());
    localeOpt.add(vm.propertyNames->localeMatcher.string(), localeMatcher);

    const HashSet<String> availableLocales = exec.jsCallee()->globalObject(vm)->intlNumberFormatAvailableLocales();
    HashMap<String, String> resolved = resolveLocale(exec, availableLocales, requestedLocales, localeOpt, nullptr, 0, IntlPRInternal::localeData);
    m_locale = resolved.get(vm.propertyNames->locale.string());
    if (m_locale.isEmpty()) {
        throwTypeError(&exec, scope, "failed to initialize PluralRules due to invalid locale"_s);
        return;
    }

    String typeString = intlStringOption(exec, options, Identifier::fromString(&vm, "type"), { "cardinal", "ordinal" }, "type must be \"cardinal\" or \"ordinal\"", "cardinal");
    RETURN_IF_EXCEPTION(scope, void());
    m_type = typeString == "ordinal" ? UPLURAL_TYPE_ORDINAL : UPLURAL_TYPE_CARDINAL;

    unsigned minimumIntegerDigits = intlNumberOption(exec, options, Identifier::fromString(&vm, "minimumIntegerDigits"), 1, 21, 1);
    RETURN_IF_EXCEPTION(scope, void());
    m_minimumIntegerDigits = minimumIntegerDigits;

    unsigned minimumFractionDigitsDefault = 0;
    unsigned minimumFractionDigits = intlNumberOption(exec, options, Identifier::fromString(&vm, "minimumFractionDigits"), 0, 20, minimumFractionDigitsDefault);
    RETURN_IF_EXCEPTION(scope, void());
    m_minimumFractionDigits = minimumFractionDigits;

    unsigned maximumFractionDigitsDefault = std::max(minimumFractionDigits, 3u);
    unsigned maximumFractionDigits = intlNumberOption(exec, options, Identifier::fromString(&vm, "maximumFractionDigits"), minimumFractionDigits, 20, maximumFractionDigitsDefault);
    RETURN_IF_EXCEPTION(scope, void());
    m_maximumFractionDigits = maximumFractionDigits;

    JSValue minimumSignificantDigitsValue = options->get(&exec, Identifier::fromString(&vm, "minimumSignificantDigits"));
    RETURN_IF_EXCEPTION(scope, void());

    JSValue maximumSignificantDigitsValue = options->get(&exec, Identifier::fromString(&vm, "maximumSignificantDigits"));
    RETURN_IF_EXCEPTION(scope, void());

    if (!minimumSignificantDigitsValue.isUndefined() || !maximumSignificantDigitsValue.isUndefined()) {
        unsigned minimumSignificantDigits = intlNumberOption(exec, options, Identifier::fromString(&vm, "minimumSignificantDigits"), 1, 21, 1);
        RETURN_IF_EXCEPTION(scope, void());
        unsigned maximumSignificantDigits = intlNumberOption(exec, options, Identifier::fromString(&vm, "maximumSignificantDigits"), minimumSignificantDigits, 21, 21);
        RETURN_IF_EXCEPTION(scope, void());
        m_minimumSignificantDigits = minimumSignificantDigits;
        m_maximumSignificantDigits = maximumSignificantDigits;
    }

    UErrorCode status = U_ZERO_ERROR;
    m_numberFormat = std::unique_ptr<UNumberFormat, UNumberFormatDeleter>(unum_open(UNUM_DECIMAL, nullptr, 0, m_locale.utf8().data(), nullptr, &status));
    if (U_FAILURE(status)) {
        throwTypeError(&exec, scope, "failed to initialize PluralRules"_s);
        return;
    }

    if (m_minimumSignificantDigits) {
        unum_setAttribute(m_numberFormat.get(), UNUM_SIGNIFICANT_DIGITS_USED, true);
        unum_setAttribute(m_numberFormat.get(), UNUM_MIN_SIGNIFICANT_DIGITS, m_minimumSignificantDigits.value());
        unum_setAttribute(m_numberFormat.get(), UNUM_MAX_SIGNIFICANT_DIGITS, m_maximumSignificantDigits.value());
    } else {
        unum_setAttribute(m_numberFormat.get(), UNUM_MIN_INTEGER_DIGITS, m_minimumIntegerDigits);
        unum_setAttribute(m_numberFormat.get(), UNUM_MIN_FRACTION_DIGITS, m_minimumFractionDigits);
        unum_setAttribute(m_numberFormat.get(), UNUM_MAX_FRACTION_DIGITS, m_maximumFractionDigits);
    }

    status = U_ZERO_ERROR;
    m_pluralRules = std::unique_ptr<UPluralRules, UPluralRulesDeleter>(uplrules_openForType(m_locale.utf8().data(), m_type, &status));
    if (U_FAILURE(status)) {
        throwTypeError(&exec, scope, "failed to initialize PluralRules"_s);
        return;
    }

    m_initializedPluralRules = true;
}

int icu_breakpoints(lua_State *L) {
  const char* input = luaL_checkstring(L, 1);
  int input_l = strlen(input);
  const char* locale = luaL_checkstring(L, 2);
  UChar *buffer;
  int32_t l, breakcount = 0;
  UErrorCode err = U_ZERO_ERROR;
  u_strFromUTF8(NULL, 0, &l, input, input_l, &err);
  /* Above call returns an error every time. */
  err = U_ZERO_ERROR;
  buffer = malloc(l * sizeof(UChar));
  u_strFromUTF8(buffer, l, &l, input, input_l, &err);

  UBreakIterator* wordbreaks, *linebreaks;
  int32_t i, previous;
  wordbreaks = ubrk_open(UBRK_WORD, locale, buffer, l, &err);
  if(U_FAILURE(err)) {
    luaL_error(L, "Word break parser failure: %s", u_errorName(err));
  }

  linebreaks = ubrk_open(UBRK_LINE, locale, buffer, l, &err);
  if(U_FAILURE(err)) {
    luaL_error(L, "Line break parser failure: %s", u_errorName(err));
  }

  previous = 0;
  i = 0;
  while (i <= l) {
    int32_t out_l;
    int32_t type;
    if (!ubrk_isBoundary(linebreaks, i) && !ubrk_isBoundary(wordbreaks,i)) {
      i++; continue;
    }
    lua_checkstack(L, 3);
    /* At some kind of boundary */
    lua_newtable(L);
    lua_pushstring(L, "type");
    lua_pushstring(L, ubrk_isBoundary(linebreaks,i) ? "line" : "word");
    lua_settable(L, -3);

    int32_t utf8_index = 0;
    err = U_ZERO_ERROR;
    u_strToUTF8(NULL, 0, &utf8_index, buffer, i, &err);
    assert(U_SUCCESS(err) || err == U_BUFFER_OVERFLOW_ERROR);

    lua_pushstring(L, "index");
    lua_pushinteger(L, utf8_index);
    lua_settable(L, -3);

    if (ubrk_isBoundary(linebreaks, i)) {
      lua_pushstring(L, "subtype");
      type = ubrk_getRuleStatus(linebreaks);
      if (type >= UBRK_LINE_SOFT && type < UBRK_LINE_SOFT_LIMIT) {
        lua_pushstring(L, "soft");
      } else {
        lua_pushstring(L, "hard");
      }
      lua_settable(L, -3);
    }
    lua_pushstring(L, "token");
    lua_pushlstring(L, input+previous, utf8_index-previous);

    lua_settable(L, -3);

    previous = utf8_index;
    breakcount++;
    i++;
  }
  ubrk_close(wordbreaks);
  ubrk_close(linebreaks);
  return breakcount;
}

static void TestGetSortKey() {
    /* This is meant to test a buffer reallocation crash while using
    French secondary sorting with a large buffer.
    The fact that Japanese characters are used is irrelevant. */
    static const UChar pucUTF16[] = {
        0x3049,0x30b9,0x3088,0xfffd,0xfffd,0x308f,0xfffd,0x3042,
        0xfffd,0xfffd,0x305e,0xfffd,0x30b6,0x30bb,0x305b,0x30b1,
        0x3050,0x30af,0x304e,0x30bd,0xfffd,0x30c6,0xfffd,0xfffd,
        0x30e1,0xfffd,0xfffd,0x30d9,0xfffd,0x3092,0x3075,0x304a,
        0x3074,0x3070,0x30f5,0x30c4,0x306e,0x30df,0x3053,0xfffd,
        0x30a6,0x30b6,0x30e0,0xfffd,0x30bc,0x30ef,0x3087,0x30cc,
        0x305f,0x30de,0xfffd,0x3090,0x3063,0x30dc,0x30b6,0x30b9,
        0x30d2,0x3072,0x3061,0xfffd,0xfffd,0xfffd,0x307b,0x3092,
        0x30a5,0x30a9,0x30b1,0x30e7,0xfffd,0xfffd,0xfffd,0xfffd,
        0xfffd,0x305e,0xfffd,0x30c7,0x30ae,0x305b,0x308b,0x30c0,
        0x30f5,0xfffd,0xfffd,0xfffd,0x307d,0x304e,0xfffd,0xfffd,
        0x30c0,0x30c8,0x306f,0x307a,0x30dd,0x30e4,0x3084,0xfffd,
        0x308c,0x30f1,0xfffd,0x30c6,0xfffd,0x307a,0xfffd,0x3052,
        0x3056,0x305d,0x30b7,0xfffd,0x305b,0x30b0,0x30b9,0xfffd,
        0x30b2,0x306d,0x3044,0xfffd,0x3073,0xfffd,0x30be,0x30cf,
        0x3080,0xfffd,0x30a8,0x30f5,0x30a5,0x30c7,0x307c,0xfffd,
        0x30d1,0x305f,0x30b2,0xfffd,0x3053,0x30ca,0xfffd,0x30dd,
        0x3058,0x30c0,0x305d,0x30e1,0xfffd,0x30bb,0x305f,0x30d1,
        0x30f2,0x3058,0x3086,0x30ce,0x30db,0x30cb,0x30e9,0xfffd,
        0x308c,0xfffd,0xfffd,0x30af,0x30c4,0x3076,0x304c,0x30f5,
        0x30e8,0x308c,0xfffd,0x30e2,0x3073,0x30a3,0x304e,0x30ea,
        0xfffd,0x304f,0xfffd,0x306c,0x3044,0xfffd,0xfffd,0x30c9,
        0xfffd,0x30f5,0xfffd,0xfffd,0xfffd,0x30eb,0x30a8,0xfffd,
        0x306d,0x307d,0x30d8,0x3069,0xfffd,0xfffd,0x3086,0x30a9,
        0xfffd,0x3076,0x30e9,0x30cc,0x3074,0x30e0,0xfffd,0xfffd,
        0xfffd,0x30f0,0x3086,0x30ac,0x3076,0x3068,0x30c7,0xfffd,
        0x30b7,0x30d2,0x3048,0x308e,0x30e8,0x30d9,0x30ce,0x30d0,
        0x308b,0x30ee,0x30e6,0x3079,0x30f3,0x30af,0xfffd,0x3079,
        0xfffd,0xfffd,0x30ca,0x30bf,0xfffd,0x30b5,0xfffd,0xfffd,
        0x3093,0xfffd,0x30ba,0xfffd,0x3076,0x3047,0x304a,0xfffd,
        0xfffd,0x3086,0xfffd,0x3081,0xfffd,0x30f6,0x3066,0xfffd,
        0xfffd,0x30b6,0x30ef,0x30e2,0x30bf,0xfffd,0x3053,0x304a,
        0xfffd,0xfffd,0x304a,0x30e8,0xfffd,0x30e2,0xfffd,0xfffd,
        0x305c,0x3081,0x30c6,0xfffd,0x3091,0x3046,0x306a,0x3059,
        0xfffd,0xfffd,0x30dd,0x30d1,0x308a,0x30ee,0xfffd,0xfffd,
        0x308a,0x3042,0x30da,0xfffd,0x3064,0x30ef,0x305c,0x306b,
        0xfffd,0x30ca,0x3085,0x3067,0x30ea,0x30c2,0x30c8,0xfffd,
        0x30f5,0xfffd,0xfffd,0xfffd,0x30ca,0xfffd,0x3050,0x30f1,
        0x3050,0x3053,0x3072,0xfffd,0xfffd,0xfffd,0x3074,0xfffd,
        0x304b,0x30dd,0x306d,0xfffd,0x3049,0x30a1,0x30cc,0x30de,
        0x30ae,0x307b,0x308a,0xfffd,0x3065,0xfffd,0xfffd,0x30c0,
        0xfffd,0x3048,0x30dc,0x304f,0x3085,0x3059,0x304b,0x30d3,
        0x30eb,0x30a4,0x3073,0xfffd,0x30ba,0x308f,0x30a7,0x30c3,
        0x3074,0x30cf,0x306c,0x3053,0x30c0,0xfffd,0x3066,0xfffd,
        0x308f,0xfffd,0x30b5,0xfffd,0x3092,0x30c4,0xfffd,0x30d6,
        0x3056,0x30ad,0x30d2,0x30ba,0xfffd,0x30e6,0x304c,0x3088,
        0x30b6,0x3048,0x3077,0x30d1,0xfffd,0x3050,0xfffd,0x3042,
        0xfffd,0xfffd,0x308f,0xfffd,0x30c1,0xfffd,0x3074,0x3061,
        0x3056,0x30e5,0xfffd,0xfffd,0x3057,0xfffd,0xfffd,0xfffd,
        0xfffd,0x30bd,0x30b3,0x30ee,0xfffd,0x30f2,0x3084,0x3050,
        0xfffd,0x30e7,0xfffd,0xfffd,0x3060,0x3049,0x30f2,0x30ad,
        0x30bf,0x30f1,0x30a2,0xfffd,0x30af,0xfffd,0x3060,0x30a1,
        0x30e9,0x30c3,0xfffd,0x3072,0x3093,0x3070,0xfffd,0x308f,
        0x3060,0xfffd,0x3067,0x306f,0x3082,0x308b,0x3051,0xfffd,
        0x3058,0xfffd,0xfffd,0x30a8,0x3051,0x3054,0x30ad,0x30f0,
        0x3053,0xfffd,0x30e1,0x30d7,0x308d,0x307f,0x30be,0x30b0,
        0xfffd,0x30db,0xfffd,0x30d1,0xfffd,0x3054,0x30a5,0xfffd,
        0x306a,0xfffd,0x305c,0xfffd,0x3052,0x3088,0xfffd,0x306e,
        0xfffd,0x30a9,0x30a1,0x30b4,0x3083,0x30bd,0xfffd,0xfffd,
        0x306a,0x3070,0x30cd,0xfffd,0x3072,0x30ed,0x30c6,0x30be,
        0x30c4,0x305e,0x30b3,0x30e1,0x308a,0xfffd,0x305b,0xfffd,
        0x3042,0x3088,0xfffd,0x304c,0xfffd,0x3089,0x3071,0xfffd,
        0xfffd,0x30c6,0x3062,0x3079,0xfffd,0x304b,0x304a,0xfffd,
        0x30ad,0x3045,0x3045,0x3087,0xfffd,0x306a,0x308b,0x0000,
        0x30bd,0x3065,0x30b8,0x3086,0x30d3,0x3076,0xfffd,0xfffd,
        0x308f,0x3053,0x307c,0x3053,0x3084,0x30ae,0x30c4,0x3045,
        0x30a8,0x30d0,0x30e1,0x308c,0x30e6,0x30b7,0xfffd,0xfffd,
        0xfffd,0x3046,0x305f,0xfffd,0x3086,0x30ab,0xfffd,0xfffd,
        0x30c8,0xfffd,0x30a1,0x3052,0x3059,0xfffd,0x30a4,0xfffd,
        0xfffd,0x308c,0x3085,0x30ab,0x30b5,0x3091,0x30bf,0x30e3,
        0xfffd,0xfffd,0x3087,0xfffd,0x30f6,0x3051,0x30bd,0x3092,
        0x3063,0xfffd,0x30a9,0x3063,0x306e,0xfffd,0xfffd,0xfffd,
        0x306c,0xfffd,0x307e,0x30ad,0x3077,0x30c2,0x30e9,0x30d5,
        0xfffd,0xfffd,0x30c6,0x305c,0xfffd,0xfffd,0x3089,0xfffd,
        0x3048,0x30cb,0x308c,0xfffd,0xfffd,0x3044,0xfffd,0x3080,
        0x3063,0x3079,0xfffd,0x308a,0x30cb,0x3042,0x3057,0xfffd,
        0x307c,0x30c1,0x30a8,0x30cf,0xfffd,0x3083,0xfffd,0xfffd,
        0x306c,0xfffd,0x305e,0x3092,0xfffd,0x30dc,0x30b0,0x3081,
        0x30e3,0x30f0,0x304e,0x30cc,0x308e,0x30c4,0x30ad
    };

    UErrorCode status = U_ZERO_ERROR;
    UCollator *pCollator;
    int32_t lenActualSortKey;
    uint8_t pucSortKey[4096];
    static const int32_t LENSORTKEY = (int32_t)sizeof(pucSortKey);

    ucol_prepareShortStringOpen("LFR_AN_CX_EX_FO_HX_NX_S3", 0, NULL, &status);

    pCollator = ucol_openFromShortString("LFR_AN_CX_EX_FO_HX_NX_S3", 0, NULL, &status);

    if (U_FAILURE(status)) {
        log_data_err("error opening collator -> %s. (Are you missing data?)\n", u_errorName(status));
        return;
    }
//.........这里部分代码省略.........

//--------------------------------------------------------------------------
//
//    Assignment Operator
//
//--------------------------------------------------------------------------
RegexPattern &RegexPattern::operator = (const RegexPattern &other) {
    if (this == &other) {
        // Source and destination are the same.  Don't do anything.
        return *this;
    }

    // Clean out any previous contents of object being assigned to.
    zap();

    // Give target object a default initialization
    init();

    // Copy simple fields
    fDeferredStatus   = other.fDeferredStatus;

    if (U_FAILURE(fDeferredStatus)) {
        return *this;
    }

    if (other.fPatternString == NULL) {
        fPatternString = NULL;
        fPattern = utext_clone(fPattern, other.fPattern, FALSE, TRUE, &fDeferredStatus);
    } else {
        fPatternString = new UnicodeString(*(other.fPatternString));
        if (fPatternString == NULL) {
            fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
        } else {
            fPattern = utext_openConstUnicodeString(NULL, fPatternString, &fDeferredStatus);
        }
    }
    if (U_FAILURE(fDeferredStatus)) {
        return *this;
    }

    fFlags            = other.fFlags;
    fLiteralText      = other.fLiteralText;
    fMinMatchLen      = other.fMinMatchLen;
    fFrameSize        = other.fFrameSize;
    fDataSize         = other.fDataSize;
    fStaticSets       = other.fStaticSets;
    fStaticSets8      = other.fStaticSets8;

    fStartType        = other.fStartType;
    fInitialStringIdx = other.fInitialStringIdx;
    fInitialStringLen = other.fInitialStringLen;
    *fInitialChars    = *other.fInitialChars;
    fInitialChar      = other.fInitialChar;
    *fInitialChars8   = *other.fInitialChars8;
    fNeedsAltInput    = other.fNeedsAltInput;

    //  Copy the pattern.  It's just values, nothing deep to copy.
    fCompiledPat->assign(*other.fCompiledPat, fDeferredStatus);
    fGroupMap->assign(*other.fGroupMap, fDeferredStatus);

    //  Copy the Unicode Sets.
    //    Could be made more efficient if the sets were reference counted and shared,
    //    but I doubt that pattern copying will be particularly common.
    //    Note:  init() already added an empty element zero to fSets
    int32_t i;
    int32_t  numSets = other.fSets->size();
    fSets8 = new Regex8BitSet[numSets];
    if (fSets8 == NULL) {
    	fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
    	return *this;
    }
    for (i=1; i<numSets; i++) {
        if (U_FAILURE(fDeferredStatus)) {
            return *this;
        }
        UnicodeSet *sourceSet = (UnicodeSet *)other.fSets->elementAt(i);
        UnicodeSet *newSet    = new UnicodeSet(*sourceSet);
        if (newSet == NULL) {
            fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
            break;
        }
        fSets->addElement(newSet, fDeferredStatus);
        fSets8[i] = other.fSets8[i];
    }

    // Copy the named capture group hash map.
    int32_t hashPos = UHASH_FIRST;
    while (const UHashElement *hashEl = uhash_nextElement(other.fNamedCaptureMap, &hashPos)) {
        if (U_FAILURE(fDeferredStatus)) {
            break;
        }
        const UnicodeString *name = (const UnicodeString *)hashEl->key.pointer;
        UnicodeString *key = new UnicodeString(*name);
        int32_t val = hashEl->value.integer;
        if (key == NULL) {
            fDeferredStatus = U_MEMORY_ALLOCATION_ERROR;
        } else {
            uhash_puti(fNamedCaptureMap, key, val, &fDeferredStatus);
        }
    }
    return *this;
//.........这里部分代码省略.........

U_CAPI const USet* U_EXPORT2
uspoof_getCheckResultNumerics(const USpoofCheckResult *checkResult, UErrorCode *status) {
    const CheckResult* This = CheckResult::validateThis(checkResult, *status);
    if (U_FAILURE(*status)) { return NULL; }
    return This->fNumerics.toUSet();
}

U_CAPI URestrictionLevel U_EXPORT2
uspoof_getCheckResultRestrictionLevel(const USpoofCheckResult *checkResult, UErrorCode *status) {
    const CheckResult* This = CheckResult::validateThis(checkResult, *status);
    if (U_FAILURE(*status)) { return USPOOF_UNRESTRICTIVE; }
    return This->fRestrictionLevel;
}

U_CAPI int32_t U_EXPORT2
uspoof_getCheckResultChecks(const USpoofCheckResult *checkResult, UErrorCode *status) {
    const CheckResult* This = CheckResult::validateThis(checkResult, *status);
    if (U_FAILURE(*status)) { return 0; }
    return This->fChecks;
}

const LanguageBreakEngine *
ICULanguageBreakFactory::loadEngineFor(UChar32 c, int32_t breakType) {
    UErrorCode status = U_ZERO_ERROR;
    UScriptCode code = uscript_getScript(c, &status);
    if (U_SUCCESS(status)) {
        DictionaryMatcher *m = loadDictionaryMatcherFor(code, breakType);
        if (m != NULL) {
            const LanguageBreakEngine *engine = NULL;
            switch(code) {
            case USCRIPT_THAI:
                engine = new ThaiBreakEngine(m, status);
                break;
            case USCRIPT_LAO:
                engine = new LaoBreakEngine(m, status);
                break;
            case USCRIPT_MYANMAR:
                engine = new BurmeseBreakEngine(m, status);
                break;
            case USCRIPT_KHMER:
                engine = new KhmerBreakEngine(m, status);
                break;

#if !UCONFIG_NO_NORMALIZATION
                // CJK not available w/o normalization
            case USCRIPT_HANGUL:
                engine = new CjkBreakEngine(m, kKorean, status);
                break;

            // use same BreakEngine and dictionary for both Chinese and Japanese
            case USCRIPT_HIRAGANA:
            case USCRIPT_KATAKANA:
            case USCRIPT_HAN:
                engine = new CjkBreakEngine(m, kChineseJapanese, status);
                break;
#if 0
            // TODO: Have to get some characters with script=common handled
            // by CjkBreakEngine (e.g. U+309B). Simply subjecting
            // them to CjkBreakEngine does not work. The engine has to
            // special-case them.
            case USCRIPT_COMMON:
            {
                UBlockCode block = ublock_getCode(code);
                if (block == UBLOCK_HIRAGANA || block == UBLOCK_KATAKANA)
                   engine = new CjkBreakEngine(dict, kChineseJapanese, status);
                break;
            }
#endif
#endif

            default:
                break;
            }
            if (engine == NULL) {
                delete m;
            }
            else if (U_FAILURE(status)) {
                delete engine;
                engine = NULL;
            }
            return engine;
        }
    }
    return NULL;
}