michael@0: /*
michael@0: ***************************************************************************
michael@0: * Copyright (C) 2008-2013, International Business Machines Corporation
michael@0: * and others. All Rights Reserved.
michael@0: ***************************************************************************
michael@0: *   file name:  uspoof.cpp
michael@0: *   encoding:   US-ASCII
michael@0: *   tab size:   8 (not used)
michael@0: *   indentation:4
michael@0: *
michael@0: *   created on: 2008Feb13
michael@0: *   created by: Andy Heninger
michael@0: *
michael@0: *   Unicode Spoof Detection
michael@0: */
michael@0: #include "unicode/utypes.h"
michael@0: #include "unicode/normalizer2.h"
michael@0: #include "unicode/uspoof.h"
michael@0: #include "unicode/ustring.h"
michael@0: #include "unicode/utf16.h"
michael@0: #include "cmemory.h"
michael@0: #include "cstring.h"
michael@0: #include "identifier_info.h"
michael@0: #include "mutex.h"
michael@0: #include "scriptset.h"
michael@0: #include "uassert.h"
michael@0: #include "ucln_in.h"
michael@0: #include "uspoof_impl.h"
michael@0: #include "umutex.h"
michael@0: 
michael@0: 
michael@0: #if !UCONFIG_NO_NORMALIZATION
michael@0: 
michael@0: U_NAMESPACE_USE
michael@0: 
michael@0: 
michael@0: //
michael@0: // Static Objects used by the spoof impl, their thread safe initialization and their cleanup.
michael@0: //
michael@0: static UnicodeSet *gInclusionSet = NULL;
michael@0: static UnicodeSet *gRecommendedSet = NULL;
michael@0: static const Normalizer2 *gNfdNormalizer = NULL;
michael@0: static UMutex gInitMutex = U_MUTEX_INITIALIZER;
michael@0: 
michael@0: static UBool U_CALLCONV
michael@0: uspoof_cleanup(void) {
michael@0:     delete gInclusionSet;
michael@0:     gInclusionSet = NULL;
michael@0:     delete gRecommendedSet;
michael@0:     gRecommendedSet = NULL;
michael@0:     gNfdNormalizer = NULL;
michael@0:     return TRUE;
michael@0: }
michael@0: 
michael@0: static void initializeStatics() {
michael@0:     Mutex m(&gInitMutex);
michael@0:     UErrorCode status = U_ZERO_ERROR;
michael@0:     if (gInclusionSet == NULL) {
michael@0:         gInclusionSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\
michael@0:             \\-.\\u00B7\\u05F3\\u05F4\\u0F0B\\u200C\\u200D\\u2019]"), status);
michael@0:         gRecommendedSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\
michael@0:             [0-z\\u00C0-\\u017E\\u01A0\\u01A1\\u01AF\\u01B0\\u01CD-\
michael@0:             \\u01DC\\u01DE-\\u01E3\\u01E6-\\u01F5\\u01F8-\\u021B\\u021E\
michael@0:             \\u021F\\u0226-\\u0233\\u02BB\\u02BC\\u02EC\\u0300-\\u0304\
michael@0:             \\u0306-\\u030C\\u030F-\\u0311\\u0313\\u0314\\u031B\\u0323-\
michael@0:             \\u0328\\u032D\\u032E\\u0330\\u0331\\u0335\\u0338\\u0339\
michael@0:             \\u0342-\\u0345\\u037B-\\u03CE\\u03FC-\\u045F\\u048A-\\u0525\
michael@0:             \\u0531-\\u0586\\u05D0-\\u05F2\\u0621-\\u063F\\u0641-\\u0655\
michael@0:             \\u0660-\\u0669\\u0670-\\u068D\\u068F-\\u06D5\\u06E5\\u06E6\
michael@0:             \\u06EE-\\u06FF\\u0750-\\u07B1\\u0901-\\u0939\\u093C-\\u094D\
michael@0:             \\u0950\\u0960-\\u0972\\u0979-\\u0A4D\\u0A5C-\\u0A74\\u0A81-\
michael@0:             \\u0B43\\u0B47-\\u0B61\\u0B66-\\u0C56\\u0C60\\u0C61\\u0C66-\
michael@0:             \\u0CD6\\u0CE0-\\u0CEF\\u0D02-\\u0D28\\u0D2A-\\u0D39\\u0D3D-\
michael@0:             \\u0D43\\u0D46-\\u0D4D\\u0D57-\\u0D61\\u0D66-\\u0D8E\\u0D91-\
michael@0:             \\u0DA5\\u0DA7-\\u0DDE\\u0DF2\\u0E01-\\u0ED9\\u0F00\\u0F20-\
michael@0:             \\u0F8B\\u0F90-\\u109D\\u10D0-\\u10F0\\u10F7-\\u10FA\\u1200-\
michael@0:             \\u135A\\u135F\\u1380-\\u138F\\u1401-\\u167F\\u1780-\\u17A2\
michael@0:             \\u17A5-\\u17A7\\u17A9-\\u17B3\\u17B6-\\u17CA\\u17D2\\u17D7-\
michael@0:             \\u17DC\\u17E0-\\u17E9\\u1810-\\u18A8\\u18AA-\\u18F5\\u1E00-\
michael@0:             \\u1E99\\u1F00-\\u1FFC\\u2D30-\\u2D65\\u2D80-\\u2DDE\\u3005-\
michael@0:             \\u3007\\u3041-\\u31B7\\u3400-\\u9FCB\\uA000-\\uA48C\\uA67F\
michael@0:             \\uA717-\\uA71F\\uA788\\uAA60-\\uAA7B\\uAC00-\\uD7A3\\uFA0E-\
michael@0:             \\uFA29\\U00020000-\
michael@0:             \\U0002B734]-[[:Cn:][:nfkcqc=n:][:XIDC=n:]]]"), status);
michael@0:         gNfdNormalizer = Normalizer2::getNFDInstance(status);
michael@0:     }
michael@0:     ucln_i18n_registerCleanup(UCLN_I18N_SPOOF, uspoof_cleanup);
michael@0: 
michael@0:     return;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI USpoofChecker * U_EXPORT2
michael@0: uspoof_open(UErrorCode *status) {
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return NULL;
michael@0:     }
michael@0:     initializeStatics();
michael@0:     SpoofImpl *si = new SpoofImpl(SpoofData::getDefault(*status), *status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         delete si;
michael@0:         si = NULL;
michael@0:     }
michael@0:     return reinterpret_cast<USpoofChecker *>(si);
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI USpoofChecker * U_EXPORT2
michael@0: uspoof_openFromSerialized(const void *data, int32_t length, int32_t *pActualLength,
michael@0:                           UErrorCode *status) {
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return NULL;
michael@0:     }
michael@0:     initializeStatics();
michael@0:     SpoofData *sd = new SpoofData(data, length, *status);
michael@0:     SpoofImpl *si = new SpoofImpl(sd, *status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         delete sd;
michael@0:         delete si;
michael@0:         return NULL;
michael@0:     }
michael@0:     if (sd == NULL || si == NULL) {
michael@0:         *status = U_MEMORY_ALLOCATION_ERROR;
michael@0:         delete sd;
michael@0:         delete si;
michael@0:         return NULL;
michael@0:     }
michael@0:         
michael@0:     if (pActualLength != NULL) {
michael@0:         *pActualLength = sd->fRawData->fLength;
michael@0:     }
michael@0:     return reinterpret_cast<USpoofChecker *>(si);
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI USpoofChecker * U_EXPORT2
michael@0: uspoof_clone(const USpoofChecker *sc, UErrorCode *status) {
michael@0:     const SpoofImpl *src = SpoofImpl::validateThis(sc, *status);
michael@0:     if (src == NULL) {
michael@0:         return NULL;
michael@0:     }
michael@0:     SpoofImpl *result = new SpoofImpl(*src, *status);   // copy constructor
michael@0:     if (U_FAILURE(*status)) {
michael@0:         delete result;
michael@0:         result = NULL;
michael@0:     }
michael@0:     return reinterpret_cast<USpoofChecker *>(result);
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI void U_EXPORT2
michael@0: uspoof_close(USpoofChecker *sc) {
michael@0:     UErrorCode status = U_ZERO_ERROR;
michael@0:     SpoofImpl *This = SpoofImpl::validateThis(sc, status);
michael@0:     delete This;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI void U_EXPORT2
michael@0: uspoof_setChecks(USpoofChecker *sc, int32_t checks, UErrorCode *status) {
michael@0:     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         return;
michael@0:     }
michael@0: 
michael@0:     // Verify that the requested checks are all ones (bits) that 
michael@0:     //   are acceptable, known values.
michael@0:     if (checks & ~USPOOF_ALL_CHECKS) {
michael@0:         *status = U_ILLEGAL_ARGUMENT_ERROR; 
michael@0:         return;
michael@0:     }
michael@0: 
michael@0:     This->fChecks = checks;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_getChecks(const USpoofChecker *sc, UErrorCode *status) {
michael@0:     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         return 0;
michael@0:     }
michael@0:     return This->fChecks;
michael@0: }
michael@0: 
michael@0: U_CAPI void U_EXPORT2
michael@0: uspoof_setRestrictionLevel(USpoofChecker *sc, URestrictionLevel restrictionLevel) {
michael@0:     UErrorCode status = U_ZERO_ERROR;
michael@0:     SpoofImpl *This = SpoofImpl::validateThis(sc, status);
michael@0:     if (This != NULL) {
michael@0:         This->fRestrictionLevel = restrictionLevel;
michael@0:     }
michael@0: }
michael@0: 
michael@0: U_CAPI URestrictionLevel U_EXPORT2
michael@0: uspoof_getRestrictionLevel(const USpoofChecker *sc) {
michael@0:     UErrorCode status = U_ZERO_ERROR;
michael@0:     const SpoofImpl *This = SpoofImpl::validateThis(sc, status);
michael@0:     if (This == NULL) {
michael@0:         return USPOOF_UNRESTRICTIVE;
michael@0:     }
michael@0:     return This->fRestrictionLevel;
michael@0: }
michael@0: 
michael@0: U_CAPI void U_EXPORT2
michael@0: uspoof_setAllowedLocales(USpoofChecker *sc, const char *localesList, UErrorCode *status) {
michael@0:     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         return;
michael@0:     }
michael@0:     This->setAllowedLocales(localesList, *status);
michael@0: }
michael@0: 
michael@0: U_CAPI const char * U_EXPORT2
michael@0: uspoof_getAllowedLocales(USpoofChecker *sc, UErrorCode *status) {
michael@0:     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         return NULL;
michael@0:     }
michael@0:     return This->getAllowedLocales(*status);
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI const USet * U_EXPORT2
michael@0: uspoof_getAllowedChars(const USpoofChecker *sc, UErrorCode *status) {
michael@0:     const UnicodeSet *result = uspoof_getAllowedUnicodeSet(sc, status);
michael@0:     return result->toUSet();
michael@0: }
michael@0: 
michael@0: U_CAPI const UnicodeSet * U_EXPORT2
michael@0: uspoof_getAllowedUnicodeSet(const USpoofChecker *sc, UErrorCode *status) {
michael@0:     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         return NULL;
michael@0:     }
michael@0:     return This->fAllowedCharsSet;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI void U_EXPORT2
michael@0: uspoof_setAllowedChars(USpoofChecker *sc, const USet *chars, UErrorCode *status) {
michael@0:     const UnicodeSet *set = UnicodeSet::fromUSet(chars);
michael@0:     uspoof_setAllowedUnicodeSet(sc, set, status);
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI void U_EXPORT2
michael@0: uspoof_setAllowedUnicodeSet(USpoofChecker *sc, const UnicodeSet *chars, UErrorCode *status) {
michael@0:     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         return;
michael@0:     }
michael@0:     if (chars->isBogus()) {
michael@0:         *status = U_ILLEGAL_ARGUMENT_ERROR;
michael@0:         return;
michael@0:     }
michael@0:     UnicodeSet *clonedSet = static_cast<UnicodeSet *>(chars->clone());
michael@0:     if (clonedSet == NULL || clonedSet->isBogus()) {
michael@0:         *status = U_MEMORY_ALLOCATION_ERROR;
michael@0:         return;
michael@0:     }
michael@0:     clonedSet->freeze();
michael@0:     delete This->fAllowedCharsSet;
michael@0:     This->fAllowedCharsSet = clonedSet;
michael@0:     This->fChecks |= USPOOF_CHAR_LIMIT;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_check(const USpoofChecker *sc,
michael@0:              const UChar *id, int32_t length,
michael@0:              int32_t *position,
michael@0:              UErrorCode *status) {
michael@0:              
michael@0:     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         return 0;
michael@0:     }
michael@0:     if (length < -1) {
michael@0:         *status = U_ILLEGAL_ARGUMENT_ERROR;
michael@0:         return 0;
michael@0:     }
michael@0:     UnicodeString idStr((length == -1), id, length);  // Aliasing constructor.
michael@0:     int32_t result = uspoof_checkUnicodeString(sc, idStr, position, status);
michael@0:     return result;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_checkUTF8(const USpoofChecker *sc,
michael@0:                  const char *id, int32_t length,
michael@0:                  int32_t *position,
michael@0:                  UErrorCode *status) {
michael@0: 
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return 0;
michael@0:     }
michael@0:     UnicodeString idStr = UnicodeString::fromUTF8(StringPiece(id, length>=0 ? length : uprv_strlen(id)));
michael@0:     int32_t result = uspoof_checkUnicodeString(sc, idStr, position, status);
michael@0:     return result;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_areConfusable(const USpoofChecker *sc,
michael@0:                      const UChar *id1, int32_t length1,
michael@0:                      const UChar *id2, int32_t length2,
michael@0:                      UErrorCode *status) {
michael@0:     SpoofImpl::validateThis(sc, *status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return 0;
michael@0:     }
michael@0:     if (length1 < -1 || length2 < -1) {
michael@0:         *status = U_ILLEGAL_ARGUMENT_ERROR;
michael@0:         return 0;
michael@0:     }
michael@0:         
michael@0:     UnicodeString id1Str((length1==-1), id1, length1);  // Aliasing constructor
michael@0:     UnicodeString id2Str((length2==-1), id2, length2);  // Aliasing constructor
michael@0:     return uspoof_areConfusableUnicodeString(sc, id1Str, id2Str, status);
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_areConfusableUTF8(const USpoofChecker *sc,
michael@0:                          const char *id1, int32_t length1,
michael@0:                          const char *id2, int32_t length2,
michael@0:                          UErrorCode *status) {
michael@0:     SpoofImpl::validateThis(sc, *status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return 0;
michael@0:     }
michael@0:     if (length1 < -1 || length2 < -1) {
michael@0:         *status = U_ILLEGAL_ARGUMENT_ERROR;
michael@0:         return 0;
michael@0:     }
michael@0:     UnicodeString id1Str = UnicodeString::fromUTF8(StringPiece(id1, length1>=0? length1 : uprv_strlen(id1)));
michael@0:     UnicodeString id2Str = UnicodeString::fromUTF8(StringPiece(id2, length2>=0? length2 : uprv_strlen(id2)));
michael@0:     int32_t results = uspoof_areConfusableUnicodeString(sc, id1Str, id2Str, status);
michael@0:     return results;
michael@0: }
michael@0:  
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_areConfusableUnicodeString(const USpoofChecker *sc,
michael@0:                                   const icu::UnicodeString &id1,
michael@0:                                   const icu::UnicodeString &id2,
michael@0:                                   UErrorCode *status) {
michael@0:     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return 0;
michael@0:     }
michael@0:     // 
michael@0:     // See section 4 of UAX 39 for the algorithm for checking whether two strings are confusable,
michael@0:     //   and for definitions of the types (single, whole, mixed-script) of confusables.
michael@0:     
michael@0:     // We only care about a few of the check flags.  Ignore the others.
michael@0:     // If no tests relavant to this function have been specified, return an error.
michael@0:     // TODO:  is this really the right thing to do?  It's probably an error on the caller's part,
michael@0:     //        but logically we would just return 0 (no error).
michael@0:     if ((This->fChecks & (USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE | 
michael@0:                           USPOOF_WHOLE_SCRIPT_CONFUSABLE)) == 0) {
michael@0:         *status = U_INVALID_STATE_ERROR;
michael@0:         return 0;
michael@0:     }
michael@0:     int32_t  flagsForSkeleton = This->fChecks & USPOOF_ANY_CASE;
michael@0: 
michael@0:     int32_t  result = 0;
michael@0:     IdentifierInfo *identifierInfo = This->getIdentifierInfo(*status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return 0;
michael@0:     }
michael@0:     identifierInfo->setIdentifier(id1, *status);
michael@0:     int32_t id1ScriptCount = identifierInfo->getScriptCount();
michael@0:     identifierInfo->setIdentifier(id2, *status);
michael@0:     int32_t id2ScriptCount = identifierInfo->getScriptCount();
michael@0:     This->releaseIdentifierInfo(identifierInfo);
michael@0:     identifierInfo = NULL;
michael@0: 
michael@0:     if (This->fChecks & USPOOF_SINGLE_SCRIPT_CONFUSABLE) {
michael@0:         UnicodeString   id1Skeleton;
michael@0:         UnicodeString   id2Skeleton;
michael@0:         if (id1ScriptCount <= 1 && id2ScriptCount <= 1) {
michael@0:             flagsForSkeleton |= USPOOF_SINGLE_SCRIPT_CONFUSABLE;
michael@0:             uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id1, id1Skeleton, status);
michael@0:             uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id2, id2Skeleton, status);
michael@0:             if (id1Skeleton == id2Skeleton) {
michael@0:                 result |= USPOOF_SINGLE_SCRIPT_CONFUSABLE;
michael@0:             }
michael@0:         }
michael@0:     }
michael@0: 
michael@0:     if (result & USPOOF_SINGLE_SCRIPT_CONFUSABLE) {
michael@0:          // If the two inputs are single script confusable they cannot also be
michael@0:          // mixed or whole script confusable, according to the UAX39 definitions.
michael@0:          // So we can skip those tests.
michael@0:          return result;
michael@0:     }
michael@0: 
michael@0:     // Two identifiers are whole script confusable if each is of a single script 
michael@0:     // and they are mixed script confusable.
michael@0:     UBool possiblyWholeScriptConfusables = 
michael@0:         id1ScriptCount <= 1 && id2ScriptCount <= 1 && (This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE);
michael@0: 
michael@0:     //
michael@0:     // Mixed Script Check
michael@0:     //
michael@0:     if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) || possiblyWholeScriptConfusables ) {
michael@0:         // For getSkeleton(), resetting the USPOOF_SINGLE_SCRIPT_CONFUSABLE flag will get us
michael@0:         // the mixed script table skeleton, which is what we want.
michael@0:         // The Any Case / Lower Case bit in the skelton flags was set at the top of the function.
michael@0:         UnicodeString id1Skeleton;
michael@0:         UnicodeString id2Skeleton;
michael@0:         flagsForSkeleton &= ~USPOOF_SINGLE_SCRIPT_CONFUSABLE;
michael@0:         uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id1, id1Skeleton, status);
michael@0:         uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id2, id2Skeleton, status);
michael@0:         if (id1Skeleton == id2Skeleton) {
michael@0:             result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
michael@0:             if (possiblyWholeScriptConfusables) {
michael@0:                 result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
michael@0:             }
michael@0:         }
michael@0:     }
michael@0: 
michael@0:     return result;
michael@0: }
michael@0: 
michael@0: 
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_checkUnicodeString(const USpoofChecker *sc,
michael@0:                           const icu::UnicodeString &id, 
michael@0:                           int32_t *position,
michael@0:                           UErrorCode *status) {
michael@0:     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         return 0;
michael@0:     }
michael@0:     int32_t result = 0;
michael@0: 
michael@0:     IdentifierInfo *identifierInfo = NULL;
michael@0:     if ((This->fChecks) & (USPOOF_RESTRICTION_LEVEL | USPOOF_MIXED_NUMBERS)) {
michael@0:         identifierInfo = This->getIdentifierInfo(*status);
michael@0:         if (U_FAILURE(*status)) {
michael@0:             goto cleanupAndReturn;
michael@0:         }
michael@0:         identifierInfo->setIdentifier(id, *status);
michael@0:         identifierInfo->setIdentifierProfile(*This->fAllowedCharsSet);
michael@0:     }
michael@0: 
michael@0: 
michael@0:     if ((This->fChecks) & USPOOF_RESTRICTION_LEVEL) {
michael@0:         URestrictionLevel idRestrictionLevel = identifierInfo->getRestrictionLevel(*status);
michael@0:         if (idRestrictionLevel > This->fRestrictionLevel) {
michael@0:             result |= USPOOF_RESTRICTION_LEVEL;
michael@0:         }
michael@0:         if (This->fChecks & USPOOF_AUX_INFO) {
michael@0:             result |= idRestrictionLevel;
michael@0:         }
michael@0:     }
michael@0: 
michael@0:     if ((This->fChecks) & USPOOF_MIXED_NUMBERS) {
michael@0:         const UnicodeSet *numerics = identifierInfo->getNumerics();
michael@0:         if (numerics->size() > 1) {
michael@0:             result |= USPOOF_MIXED_NUMBERS;
michael@0:         }
michael@0: 
michael@0:         // TODO: ICU4J returns the UnicodeSet of the numerics found in the identifier.
michael@0:         //       We have no easy way to do the same in C.
michael@0:         // if (checkResult != null) {
michael@0:         //     checkResult.numerics = numerics;
michael@0:         // }
michael@0:     }
michael@0: 
michael@0: 
michael@0:     if (This->fChecks & (USPOOF_CHAR_LIMIT)) {
michael@0:         int32_t i;
michael@0:         UChar32 c;
michael@0:         int32_t length = id.length();
michael@0:         for (i=0; i<length ;) {
michael@0:             c = id.char32At(i);
michael@0:             i += U16_LENGTH(c);
michael@0:             if (!This->fAllowedCharsSet->contains(c)) {
michael@0:                 result |= USPOOF_CHAR_LIMIT;
michael@0:                 break;
michael@0:             }
michael@0:         }
michael@0:     }
michael@0: 
michael@0:     if (This->fChecks & 
michael@0:         (USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE | USPOOF_INVISIBLE)) {
michael@0:         // These are the checks that need to be done on NFD input
michael@0:         UnicodeString nfdText;
michael@0:         gNfdNormalizer->normalize(id, nfdText, *status);
michael@0:         int32_t nfdLength = nfdText.length();
michael@0: 
michael@0:         if (This->fChecks & USPOOF_INVISIBLE) {
michael@0:            
michael@0:             // scan for more than one occurence of the same non-spacing mark
michael@0:             // in a sequence of non-spacing marks.
michael@0:             int32_t     i;
michael@0:             UChar32     c;
michael@0:             UChar32     firstNonspacingMark = 0;
michael@0:             UBool       haveMultipleMarks = FALSE;  
michael@0:             UnicodeSet  marksSeenSoFar;   // Set of combining marks in a single combining sequence.
michael@0:             
michael@0:             for (i=0; i<nfdLength ;) {
michael@0:                 c = nfdText.char32At(i);
michael@0:                 i += U16_LENGTH(c);
michael@0:                 if (u_charType(c) != U_NON_SPACING_MARK) {
michael@0:                     firstNonspacingMark = 0;
michael@0:                     if (haveMultipleMarks) {
michael@0:                         marksSeenSoFar.clear();
michael@0:                         haveMultipleMarks = FALSE;
michael@0:                     }
michael@0:                     continue;
michael@0:                 }
michael@0:                 if (firstNonspacingMark == 0) {
michael@0:                     firstNonspacingMark = c;
michael@0:                     continue;
michael@0:                 }
michael@0:                 if (!haveMultipleMarks) {
michael@0:                     marksSeenSoFar.add(firstNonspacingMark);
michael@0:                     haveMultipleMarks = TRUE;
michael@0:                 }
michael@0:                 if (marksSeenSoFar.contains(c)) {
michael@0:                     // report the error, and stop scanning.
michael@0:                     // No need to find more than the first failure.
michael@0:                     result |= USPOOF_INVISIBLE;
michael@0:                     break;
michael@0:                 }
michael@0:                 marksSeenSoFar.add(c);
michael@0:             }
michael@0:         }
michael@0:        
michael@0:         
michael@0:         if (This->fChecks & (USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE)) {
michael@0:             // The basic test is the same for both whole and mixed script confusables.
michael@0:             // Compute the set of scripts that every input character has a confusable in.
michael@0:             // For this computation an input character is always considered to be
michael@0:             // confusable with itself in its own script.
michael@0:             //
michael@0:             // If the number of such scripts is two or more, and the input consisted of
michael@0:             // characters all from a single script, we have a whole script confusable.
michael@0:             // (The two scripts will be the original script and the one that is confusable)
michael@0:             //
michael@0:             // If the number of such scripts >= one, and the original input contained characters from
michael@0:             // more than one script, we have a mixed script confusable.  (We can transform
michael@0:             // some of the characters, and end up with a visually similar string all in
michael@0:             // one script.)
michael@0: 
michael@0:             if (identifierInfo == NULL) {
michael@0:                 identifierInfo = This->getIdentifierInfo(*status);
michael@0:                 if (U_FAILURE(*status)) {
michael@0:                     goto cleanupAndReturn;
michael@0:                 }
michael@0:                 identifierInfo->setIdentifier(id, *status);
michael@0:             }
michael@0: 
michael@0:             int32_t scriptCount = identifierInfo->getScriptCount();
michael@0:             
michael@0:             ScriptSet scripts;
michael@0:             This->wholeScriptCheck(nfdText, &scripts, *status);
michael@0:             int32_t confusableScriptCount = scripts.countMembers();
michael@0:             //printf("confusableScriptCount = %d\n", confusableScriptCount);
michael@0:             
michael@0:             if ((This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE) &&
michael@0:                 confusableScriptCount >= 2 &&
michael@0:                 scriptCount == 1) {
michael@0:                 result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
michael@0:             }
michael@0:         
michael@0:             if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) &&
michael@0:                 confusableScriptCount >= 1 &&
michael@0:                 scriptCount > 1) {
michael@0:                 result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
michael@0:             }
michael@0:         }
michael@0:     }
michael@0: 
michael@0: cleanupAndReturn:
michael@0:     This->releaseIdentifierInfo(identifierInfo);
michael@0:     if (position != NULL) {
michael@0:         *position = 0;
michael@0:     }
michael@0:     return result;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_getSkeleton(const USpoofChecker *sc,
michael@0:                    uint32_t type,
michael@0:                    const UChar *id,  int32_t length,
michael@0:                    UChar *dest, int32_t destCapacity,
michael@0:                    UErrorCode *status) {
michael@0: 
michael@0:     SpoofImpl::validateThis(sc, *status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return 0;
michael@0:     }
michael@0:     if (length<-1 || destCapacity<0 || (destCapacity==0 && dest!=NULL)) {
michael@0:         *status = U_ILLEGAL_ARGUMENT_ERROR;
michael@0:         return 0;
michael@0:     }
michael@0: 
michael@0:     UnicodeString idStr((length==-1), id, length);  // Aliasing constructor
michael@0:     UnicodeString destStr;
michael@0:     uspoof_getSkeletonUnicodeString(sc, type, idStr, destStr, status);
michael@0:     destStr.extract(dest, destCapacity, *status);
michael@0:     return destStr.length();
michael@0: }
michael@0: 
michael@0: 
michael@0: 
michael@0: U_I18N_API UnicodeString &  U_EXPORT2
michael@0: uspoof_getSkeletonUnicodeString(const USpoofChecker *sc,
michael@0:                                 uint32_t type,
michael@0:                                 const UnicodeString &id,
michael@0:                                 UnicodeString &dest,
michael@0:                                 UErrorCode *status) {
michael@0:     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return dest;
michael@0:     }
michael@0: 
michael@0:    int32_t tableMask = 0;
michael@0:    switch (type) {
michael@0:       case 0:
michael@0:         tableMask = USPOOF_ML_TABLE_FLAG;
michael@0:         break;
michael@0:       case USPOOF_SINGLE_SCRIPT_CONFUSABLE:
michael@0:         tableMask = USPOOF_SL_TABLE_FLAG;
michael@0:         break;
michael@0:       case USPOOF_ANY_CASE:
michael@0:         tableMask = USPOOF_MA_TABLE_FLAG;
michael@0:         break;
michael@0:       case USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_ANY_CASE:
michael@0:         tableMask = USPOOF_SA_TABLE_FLAG;
michael@0:         break;
michael@0:       default:
michael@0:         *status = U_ILLEGAL_ARGUMENT_ERROR;
michael@0:         return dest;
michael@0:     }
michael@0: 
michael@0:     UnicodeString nfdId;
michael@0:     gNfdNormalizer->normalize(id, nfdId, *status);
michael@0: 
michael@0:     // Apply the skeleton mapping to the NFD normalized input string
michael@0:     // Accumulate the skeleton, possibly unnormalized, in a UnicodeString.
michael@0:     int32_t inputIndex = 0;
michael@0:     UnicodeString skelStr;
michael@0:     int32_t normalizedLen = nfdId.length();
michael@0:     for (inputIndex=0; inputIndex < normalizedLen; ) {
michael@0:         UChar32 c = nfdId.char32At(inputIndex);
michael@0:         inputIndex += U16_LENGTH(c);
michael@0:         This->confusableLookup(c, tableMask, skelStr);
michael@0:     }
michael@0: 
michael@0:     gNfdNormalizer->normalize(skelStr, dest, *status);
michael@0:     return dest;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_getSkeletonUTF8(const USpoofChecker *sc,
michael@0:                        uint32_t type,
michael@0:                        const char *id,  int32_t length,
michael@0:                        char *dest, int32_t destCapacity,
michael@0:                        UErrorCode *status) {
michael@0:     SpoofImpl::validateThis(sc, *status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return 0;
michael@0:     }
michael@0:     if (length<-1 || destCapacity<0 || (destCapacity==0 && dest!=NULL)) {
michael@0:         *status = U_ILLEGAL_ARGUMENT_ERROR;
michael@0:         return 0;
michael@0:     }
michael@0: 
michael@0:     UnicodeString srcStr = UnicodeString::fromUTF8(StringPiece(id, length>=0 ? length : uprv_strlen(id)));
michael@0:     UnicodeString destStr;
michael@0:     uspoof_getSkeletonUnicodeString(sc, type, srcStr, destStr, status);
michael@0:     if (U_FAILURE(*status)) {
michael@0:         return 0;
michael@0:     }
michael@0: 
michael@0:     int32_t lengthInUTF8 = 0;
michael@0:     u_strToUTF8(dest, destCapacity, &lengthInUTF8,
michael@0:                 destStr.getBuffer(), destStr.length(), status);
michael@0:     return lengthInUTF8;
michael@0: }
michael@0: 
michael@0: 
michael@0: U_CAPI int32_t U_EXPORT2
michael@0: uspoof_serialize(USpoofChecker *sc,void *buf, int32_t capacity, UErrorCode *status) {
michael@0:     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
michael@0:     if (This == NULL) {
michael@0:         U_ASSERT(U_FAILURE(*status));
michael@0:         return 0;
michael@0:     }
michael@0:     int32_t dataSize = This->fSpoofData->fRawData->fLength;
michael@0:     if (capacity < dataSize) {
michael@0:         *status = U_BUFFER_OVERFLOW_ERROR;
michael@0:         return dataSize;
michael@0:     }
michael@0:     uprv_memcpy(buf, This->fSpoofData->fRawData, dataSize);
michael@0:     return dataSize;
michael@0: }
michael@0: 
michael@0: U_CAPI const USet * U_EXPORT2
michael@0: uspoof_getInclusionSet(UErrorCode *) {
michael@0:     initializeStatics();
michael@0:     return gInclusionSet->toUSet();
michael@0: }
michael@0: 
michael@0: U_CAPI const USet * U_EXPORT2
michael@0: uspoof_getRecommendedSet(UErrorCode *) {
michael@0:     initializeStatics();
michael@0:     return gRecommendedSet->toUSet();
michael@0: }
michael@0: 
michael@0: U_I18N_API const UnicodeSet * U_EXPORT2
michael@0: uspoof_getInclusionUnicodeSet(UErrorCode *) {
michael@0:     initializeStatics();
michael@0:     return gInclusionSet;
michael@0: }
michael@0: 
michael@0: U_I18N_API const UnicodeSet * U_EXPORT2
michael@0: uspoof_getRecommendedUnicodeSet(UErrorCode *) {
michael@0:     initializeStatics();
michael@0:     return gRecommendedSet;
michael@0: }
michael@0: 
michael@0: 
michael@0: 
michael@0: #endif // !UCONFIG_NO_NORMALIZATION