1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/intl/icu/source/i18n/uspoof.cpp Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,736 @@
1.4 +/*
1.5 +***************************************************************************
1.6 +* Copyright (C) 2008-2013, International Business Machines Corporation
1.7 +* and others. All Rights Reserved.
1.8 +***************************************************************************
1.9 +* file name: uspoof.cpp
1.10 +* encoding: US-ASCII
1.11 +* tab size: 8 (not used)
1.12 +* indentation:4
1.13 +*
1.14 +* created on: 2008Feb13
1.15 +* created by: Andy Heninger
1.16 +*
1.17 +* Unicode Spoof Detection
1.18 +*/
1.19 +#include "unicode/utypes.h"
1.20 +#include "unicode/normalizer2.h"
1.21 +#include "unicode/uspoof.h"
1.22 +#include "unicode/ustring.h"
1.23 +#include "unicode/utf16.h"
1.24 +#include "cmemory.h"
1.25 +#include "cstring.h"
1.26 +#include "identifier_info.h"
1.27 +#include "mutex.h"
1.28 +#include "scriptset.h"
1.29 +#include "uassert.h"
1.30 +#include "ucln_in.h"
1.31 +#include "uspoof_impl.h"
1.32 +#include "umutex.h"
1.33 +
1.34 +
1.35 +#if !UCONFIG_NO_NORMALIZATION
1.36 +
1.37 +U_NAMESPACE_USE
1.38 +
1.39 +
1.40 +//
1.41 +// Static Objects used by the spoof impl, their thread safe initialization and their cleanup.
1.42 +//
1.43 +static UnicodeSet *gInclusionSet = NULL;
1.44 +static UnicodeSet *gRecommendedSet = NULL;
1.45 +static const Normalizer2 *gNfdNormalizer = NULL;
1.46 +static UMutex gInitMutex = U_MUTEX_INITIALIZER;
1.47 +
1.48 +static UBool U_CALLCONV
1.49 +uspoof_cleanup(void) {
1.50 + delete gInclusionSet;
1.51 + gInclusionSet = NULL;
1.52 + delete gRecommendedSet;
1.53 + gRecommendedSet = NULL;
1.54 + gNfdNormalizer = NULL;
1.55 + return TRUE;
1.56 +}
1.57 +
1.58 +static void initializeStatics() {
1.59 + Mutex m(&gInitMutex);
1.60 + UErrorCode status = U_ZERO_ERROR;
1.61 + if (gInclusionSet == NULL) {
1.62 + gInclusionSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\
1.63 + \\-.\\u00B7\\u05F3\\u05F4\\u0F0B\\u200C\\u200D\\u2019]"), status);
1.64 + gRecommendedSet = new UnicodeSet(UNICODE_STRING_SIMPLE("[\
1.65 + [0-z\\u00C0-\\u017E\\u01A0\\u01A1\\u01AF\\u01B0\\u01CD-\
1.66 + \\u01DC\\u01DE-\\u01E3\\u01E6-\\u01F5\\u01F8-\\u021B\\u021E\
1.67 + \\u021F\\u0226-\\u0233\\u02BB\\u02BC\\u02EC\\u0300-\\u0304\
1.68 + \\u0306-\\u030C\\u030F-\\u0311\\u0313\\u0314\\u031B\\u0323-\
1.69 + \\u0328\\u032D\\u032E\\u0330\\u0331\\u0335\\u0338\\u0339\
1.70 + \\u0342-\\u0345\\u037B-\\u03CE\\u03FC-\\u045F\\u048A-\\u0525\
1.71 + \\u0531-\\u0586\\u05D0-\\u05F2\\u0621-\\u063F\\u0641-\\u0655\
1.72 + \\u0660-\\u0669\\u0670-\\u068D\\u068F-\\u06D5\\u06E5\\u06E6\
1.73 + \\u06EE-\\u06FF\\u0750-\\u07B1\\u0901-\\u0939\\u093C-\\u094D\
1.74 + \\u0950\\u0960-\\u0972\\u0979-\\u0A4D\\u0A5C-\\u0A74\\u0A81-\
1.75 + \\u0B43\\u0B47-\\u0B61\\u0B66-\\u0C56\\u0C60\\u0C61\\u0C66-\
1.76 + \\u0CD6\\u0CE0-\\u0CEF\\u0D02-\\u0D28\\u0D2A-\\u0D39\\u0D3D-\
1.77 + \\u0D43\\u0D46-\\u0D4D\\u0D57-\\u0D61\\u0D66-\\u0D8E\\u0D91-\
1.78 + \\u0DA5\\u0DA7-\\u0DDE\\u0DF2\\u0E01-\\u0ED9\\u0F00\\u0F20-\
1.79 + \\u0F8B\\u0F90-\\u109D\\u10D0-\\u10F0\\u10F7-\\u10FA\\u1200-\
1.80 + \\u135A\\u135F\\u1380-\\u138F\\u1401-\\u167F\\u1780-\\u17A2\
1.81 + \\u17A5-\\u17A7\\u17A9-\\u17B3\\u17B6-\\u17CA\\u17D2\\u17D7-\
1.82 + \\u17DC\\u17E0-\\u17E9\\u1810-\\u18A8\\u18AA-\\u18F5\\u1E00-\
1.83 + \\u1E99\\u1F00-\\u1FFC\\u2D30-\\u2D65\\u2D80-\\u2DDE\\u3005-\
1.84 + \\u3007\\u3041-\\u31B7\\u3400-\\u9FCB\\uA000-\\uA48C\\uA67F\
1.85 + \\uA717-\\uA71F\\uA788\\uAA60-\\uAA7B\\uAC00-\\uD7A3\\uFA0E-\
1.86 + \\uFA29\\U00020000-\
1.87 + \\U0002B734]-[[:Cn:][:nfkcqc=n:][:XIDC=n:]]]"), status);
1.88 + gNfdNormalizer = Normalizer2::getNFDInstance(status);
1.89 + }
1.90 + ucln_i18n_registerCleanup(UCLN_I18N_SPOOF, uspoof_cleanup);
1.91 +
1.92 + return;
1.93 +}
1.94 +
1.95 +
1.96 +U_CAPI USpoofChecker * U_EXPORT2
1.97 +uspoof_open(UErrorCode *status) {
1.98 + if (U_FAILURE(*status)) {
1.99 + return NULL;
1.100 + }
1.101 + initializeStatics();
1.102 + SpoofImpl *si = new SpoofImpl(SpoofData::getDefault(*status), *status);
1.103 + if (U_FAILURE(*status)) {
1.104 + delete si;
1.105 + si = NULL;
1.106 + }
1.107 + return reinterpret_cast<USpoofChecker *>(si);
1.108 +}
1.109 +
1.110 +
1.111 +U_CAPI USpoofChecker * U_EXPORT2
1.112 +uspoof_openFromSerialized(const void *data, int32_t length, int32_t *pActualLength,
1.113 + UErrorCode *status) {
1.114 + if (U_FAILURE(*status)) {
1.115 + return NULL;
1.116 + }
1.117 + initializeStatics();
1.118 + SpoofData *sd = new SpoofData(data, length, *status);
1.119 + SpoofImpl *si = new SpoofImpl(sd, *status);
1.120 + if (U_FAILURE(*status)) {
1.121 + delete sd;
1.122 + delete si;
1.123 + return NULL;
1.124 + }
1.125 + if (sd == NULL || si == NULL) {
1.126 + *status = U_MEMORY_ALLOCATION_ERROR;
1.127 + delete sd;
1.128 + delete si;
1.129 + return NULL;
1.130 + }
1.131 +
1.132 + if (pActualLength != NULL) {
1.133 + *pActualLength = sd->fRawData->fLength;
1.134 + }
1.135 + return reinterpret_cast<USpoofChecker *>(si);
1.136 +}
1.137 +
1.138 +
1.139 +U_CAPI USpoofChecker * U_EXPORT2
1.140 +uspoof_clone(const USpoofChecker *sc, UErrorCode *status) {
1.141 + const SpoofImpl *src = SpoofImpl::validateThis(sc, *status);
1.142 + if (src == NULL) {
1.143 + return NULL;
1.144 + }
1.145 + SpoofImpl *result = new SpoofImpl(*src, *status); // copy constructor
1.146 + if (U_FAILURE(*status)) {
1.147 + delete result;
1.148 + result = NULL;
1.149 + }
1.150 + return reinterpret_cast<USpoofChecker *>(result);
1.151 +}
1.152 +
1.153 +
1.154 +U_CAPI void U_EXPORT2
1.155 +uspoof_close(USpoofChecker *sc) {
1.156 + UErrorCode status = U_ZERO_ERROR;
1.157 + SpoofImpl *This = SpoofImpl::validateThis(sc, status);
1.158 + delete This;
1.159 +}
1.160 +
1.161 +
1.162 +U_CAPI void U_EXPORT2
1.163 +uspoof_setChecks(USpoofChecker *sc, int32_t checks, UErrorCode *status) {
1.164 + SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.165 + if (This == NULL) {
1.166 + return;
1.167 + }
1.168 +
1.169 + // Verify that the requested checks are all ones (bits) that
1.170 + // are acceptable, known values.
1.171 + if (checks & ~USPOOF_ALL_CHECKS) {
1.172 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.173 + return;
1.174 + }
1.175 +
1.176 + This->fChecks = checks;
1.177 +}
1.178 +
1.179 +
1.180 +U_CAPI int32_t U_EXPORT2
1.181 +uspoof_getChecks(const USpoofChecker *sc, UErrorCode *status) {
1.182 + const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.183 + if (This == NULL) {
1.184 + return 0;
1.185 + }
1.186 + return This->fChecks;
1.187 +}
1.188 +
1.189 +U_CAPI void U_EXPORT2
1.190 +uspoof_setRestrictionLevel(USpoofChecker *sc, URestrictionLevel restrictionLevel) {
1.191 + UErrorCode status = U_ZERO_ERROR;
1.192 + SpoofImpl *This = SpoofImpl::validateThis(sc, status);
1.193 + if (This != NULL) {
1.194 + This->fRestrictionLevel = restrictionLevel;
1.195 + }
1.196 +}
1.197 +
1.198 +U_CAPI URestrictionLevel U_EXPORT2
1.199 +uspoof_getRestrictionLevel(const USpoofChecker *sc) {
1.200 + UErrorCode status = U_ZERO_ERROR;
1.201 + const SpoofImpl *This = SpoofImpl::validateThis(sc, status);
1.202 + if (This == NULL) {
1.203 + return USPOOF_UNRESTRICTIVE;
1.204 + }
1.205 + return This->fRestrictionLevel;
1.206 +}
1.207 +
1.208 +U_CAPI void U_EXPORT2
1.209 +uspoof_setAllowedLocales(USpoofChecker *sc, const char *localesList, UErrorCode *status) {
1.210 + SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.211 + if (This == NULL) {
1.212 + return;
1.213 + }
1.214 + This->setAllowedLocales(localesList, *status);
1.215 +}
1.216 +
1.217 +U_CAPI const char * U_EXPORT2
1.218 +uspoof_getAllowedLocales(USpoofChecker *sc, UErrorCode *status) {
1.219 + SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.220 + if (This == NULL) {
1.221 + return NULL;
1.222 + }
1.223 + return This->getAllowedLocales(*status);
1.224 +}
1.225 +
1.226 +
1.227 +U_CAPI const USet * U_EXPORT2
1.228 +uspoof_getAllowedChars(const USpoofChecker *sc, UErrorCode *status) {
1.229 + const UnicodeSet *result = uspoof_getAllowedUnicodeSet(sc, status);
1.230 + return result->toUSet();
1.231 +}
1.232 +
1.233 +U_CAPI const UnicodeSet * U_EXPORT2
1.234 +uspoof_getAllowedUnicodeSet(const USpoofChecker *sc, UErrorCode *status) {
1.235 + const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.236 + if (This == NULL) {
1.237 + return NULL;
1.238 + }
1.239 + return This->fAllowedCharsSet;
1.240 +}
1.241 +
1.242 +
1.243 +U_CAPI void U_EXPORT2
1.244 +uspoof_setAllowedChars(USpoofChecker *sc, const USet *chars, UErrorCode *status) {
1.245 + const UnicodeSet *set = UnicodeSet::fromUSet(chars);
1.246 + uspoof_setAllowedUnicodeSet(sc, set, status);
1.247 +}
1.248 +
1.249 +
1.250 +U_CAPI void U_EXPORT2
1.251 +uspoof_setAllowedUnicodeSet(USpoofChecker *sc, const UnicodeSet *chars, UErrorCode *status) {
1.252 + SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.253 + if (This == NULL) {
1.254 + return;
1.255 + }
1.256 + if (chars->isBogus()) {
1.257 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.258 + return;
1.259 + }
1.260 + UnicodeSet *clonedSet = static_cast<UnicodeSet *>(chars->clone());
1.261 + if (clonedSet == NULL || clonedSet->isBogus()) {
1.262 + *status = U_MEMORY_ALLOCATION_ERROR;
1.263 + return;
1.264 + }
1.265 + clonedSet->freeze();
1.266 + delete This->fAllowedCharsSet;
1.267 + This->fAllowedCharsSet = clonedSet;
1.268 + This->fChecks |= USPOOF_CHAR_LIMIT;
1.269 +}
1.270 +
1.271 +
1.272 +U_CAPI int32_t U_EXPORT2
1.273 +uspoof_check(const USpoofChecker *sc,
1.274 + const UChar *id, int32_t length,
1.275 + int32_t *position,
1.276 + UErrorCode *status) {
1.277 +
1.278 + const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.279 + if (This == NULL) {
1.280 + return 0;
1.281 + }
1.282 + if (length < -1) {
1.283 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.284 + return 0;
1.285 + }
1.286 + UnicodeString idStr((length == -1), id, length); // Aliasing constructor.
1.287 + int32_t result = uspoof_checkUnicodeString(sc, idStr, position, status);
1.288 + return result;
1.289 +}
1.290 +
1.291 +
1.292 +U_CAPI int32_t U_EXPORT2
1.293 +uspoof_checkUTF8(const USpoofChecker *sc,
1.294 + const char *id, int32_t length,
1.295 + int32_t *position,
1.296 + UErrorCode *status) {
1.297 +
1.298 + if (U_FAILURE(*status)) {
1.299 + return 0;
1.300 + }
1.301 + UnicodeString idStr = UnicodeString::fromUTF8(StringPiece(id, length>=0 ? length : uprv_strlen(id)));
1.302 + int32_t result = uspoof_checkUnicodeString(sc, idStr, position, status);
1.303 + return result;
1.304 +}
1.305 +
1.306 +
1.307 +U_CAPI int32_t U_EXPORT2
1.308 +uspoof_areConfusable(const USpoofChecker *sc,
1.309 + const UChar *id1, int32_t length1,
1.310 + const UChar *id2, int32_t length2,
1.311 + UErrorCode *status) {
1.312 + SpoofImpl::validateThis(sc, *status);
1.313 + if (U_FAILURE(*status)) {
1.314 + return 0;
1.315 + }
1.316 + if (length1 < -1 || length2 < -1) {
1.317 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.318 + return 0;
1.319 + }
1.320 +
1.321 + UnicodeString id1Str((length1==-1), id1, length1); // Aliasing constructor
1.322 + UnicodeString id2Str((length2==-1), id2, length2); // Aliasing constructor
1.323 + return uspoof_areConfusableUnicodeString(sc, id1Str, id2Str, status);
1.324 +}
1.325 +
1.326 +
1.327 +U_CAPI int32_t U_EXPORT2
1.328 +uspoof_areConfusableUTF8(const USpoofChecker *sc,
1.329 + const char *id1, int32_t length1,
1.330 + const char *id2, int32_t length2,
1.331 + UErrorCode *status) {
1.332 + SpoofImpl::validateThis(sc, *status);
1.333 + if (U_FAILURE(*status)) {
1.334 + return 0;
1.335 + }
1.336 + if (length1 < -1 || length2 < -1) {
1.337 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.338 + return 0;
1.339 + }
1.340 + UnicodeString id1Str = UnicodeString::fromUTF8(StringPiece(id1, length1>=0? length1 : uprv_strlen(id1)));
1.341 + UnicodeString id2Str = UnicodeString::fromUTF8(StringPiece(id2, length2>=0? length2 : uprv_strlen(id2)));
1.342 + int32_t results = uspoof_areConfusableUnicodeString(sc, id1Str, id2Str, status);
1.343 + return results;
1.344 +}
1.345 +
1.346 +
1.347 +U_CAPI int32_t U_EXPORT2
1.348 +uspoof_areConfusableUnicodeString(const USpoofChecker *sc,
1.349 + const icu::UnicodeString &id1,
1.350 + const icu::UnicodeString &id2,
1.351 + UErrorCode *status) {
1.352 + const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.353 + if (U_FAILURE(*status)) {
1.354 + return 0;
1.355 + }
1.356 + //
1.357 + // See section 4 of UAX 39 for the algorithm for checking whether two strings are confusable,
1.358 + // and for definitions of the types (single, whole, mixed-script) of confusables.
1.359 +
1.360 + // We only care about a few of the check flags. Ignore the others.
1.361 + // If no tests relavant to this function have been specified, return an error.
1.362 + // TODO: is this really the right thing to do? It's probably an error on the caller's part,
1.363 + // but logically we would just return 0 (no error).
1.364 + if ((This->fChecks & (USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE |
1.365 + USPOOF_WHOLE_SCRIPT_CONFUSABLE)) == 0) {
1.366 + *status = U_INVALID_STATE_ERROR;
1.367 + return 0;
1.368 + }
1.369 + int32_t flagsForSkeleton = This->fChecks & USPOOF_ANY_CASE;
1.370 +
1.371 + int32_t result = 0;
1.372 + IdentifierInfo *identifierInfo = This->getIdentifierInfo(*status);
1.373 + if (U_FAILURE(*status)) {
1.374 + return 0;
1.375 + }
1.376 + identifierInfo->setIdentifier(id1, *status);
1.377 + int32_t id1ScriptCount = identifierInfo->getScriptCount();
1.378 + identifierInfo->setIdentifier(id2, *status);
1.379 + int32_t id2ScriptCount = identifierInfo->getScriptCount();
1.380 + This->releaseIdentifierInfo(identifierInfo);
1.381 + identifierInfo = NULL;
1.382 +
1.383 + if (This->fChecks & USPOOF_SINGLE_SCRIPT_CONFUSABLE) {
1.384 + UnicodeString id1Skeleton;
1.385 + UnicodeString id2Skeleton;
1.386 + if (id1ScriptCount <= 1 && id2ScriptCount <= 1) {
1.387 + flagsForSkeleton |= USPOOF_SINGLE_SCRIPT_CONFUSABLE;
1.388 + uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id1, id1Skeleton, status);
1.389 + uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id2, id2Skeleton, status);
1.390 + if (id1Skeleton == id2Skeleton) {
1.391 + result |= USPOOF_SINGLE_SCRIPT_CONFUSABLE;
1.392 + }
1.393 + }
1.394 + }
1.395 +
1.396 + if (result & USPOOF_SINGLE_SCRIPT_CONFUSABLE) {
1.397 + // If the two inputs are single script confusable they cannot also be
1.398 + // mixed or whole script confusable, according to the UAX39 definitions.
1.399 + // So we can skip those tests.
1.400 + return result;
1.401 + }
1.402 +
1.403 + // Two identifiers are whole script confusable if each is of a single script
1.404 + // and they are mixed script confusable.
1.405 + UBool possiblyWholeScriptConfusables =
1.406 + id1ScriptCount <= 1 && id2ScriptCount <= 1 && (This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE);
1.407 +
1.408 + //
1.409 + // Mixed Script Check
1.410 + //
1.411 + if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) || possiblyWholeScriptConfusables ) {
1.412 + // For getSkeleton(), resetting the USPOOF_SINGLE_SCRIPT_CONFUSABLE flag will get us
1.413 + // the mixed script table skeleton, which is what we want.
1.414 + // The Any Case / Lower Case bit in the skelton flags was set at the top of the function.
1.415 + UnicodeString id1Skeleton;
1.416 + UnicodeString id2Skeleton;
1.417 + flagsForSkeleton &= ~USPOOF_SINGLE_SCRIPT_CONFUSABLE;
1.418 + uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id1, id1Skeleton, status);
1.419 + uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id2, id2Skeleton, status);
1.420 + if (id1Skeleton == id2Skeleton) {
1.421 + result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
1.422 + if (possiblyWholeScriptConfusables) {
1.423 + result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
1.424 + }
1.425 + }
1.426 + }
1.427 +
1.428 + return result;
1.429 +}
1.430 +
1.431 +
1.432 +
1.433 +
1.434 +U_CAPI int32_t U_EXPORT2
1.435 +uspoof_checkUnicodeString(const USpoofChecker *sc,
1.436 + const icu::UnicodeString &id,
1.437 + int32_t *position,
1.438 + UErrorCode *status) {
1.439 + const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.440 + if (This == NULL) {
1.441 + return 0;
1.442 + }
1.443 + int32_t result = 0;
1.444 +
1.445 + IdentifierInfo *identifierInfo = NULL;
1.446 + if ((This->fChecks) & (USPOOF_RESTRICTION_LEVEL | USPOOF_MIXED_NUMBERS)) {
1.447 + identifierInfo = This->getIdentifierInfo(*status);
1.448 + if (U_FAILURE(*status)) {
1.449 + goto cleanupAndReturn;
1.450 + }
1.451 + identifierInfo->setIdentifier(id, *status);
1.452 + identifierInfo->setIdentifierProfile(*This->fAllowedCharsSet);
1.453 + }
1.454 +
1.455 +
1.456 + if ((This->fChecks) & USPOOF_RESTRICTION_LEVEL) {
1.457 + URestrictionLevel idRestrictionLevel = identifierInfo->getRestrictionLevel(*status);
1.458 + if (idRestrictionLevel > This->fRestrictionLevel) {
1.459 + result |= USPOOF_RESTRICTION_LEVEL;
1.460 + }
1.461 + if (This->fChecks & USPOOF_AUX_INFO) {
1.462 + result |= idRestrictionLevel;
1.463 + }
1.464 + }
1.465 +
1.466 + if ((This->fChecks) & USPOOF_MIXED_NUMBERS) {
1.467 + const UnicodeSet *numerics = identifierInfo->getNumerics();
1.468 + if (numerics->size() > 1) {
1.469 + result |= USPOOF_MIXED_NUMBERS;
1.470 + }
1.471 +
1.472 + // TODO: ICU4J returns the UnicodeSet of the numerics found in the identifier.
1.473 + // We have no easy way to do the same in C.
1.474 + // if (checkResult != null) {
1.475 + // checkResult.numerics = numerics;
1.476 + // }
1.477 + }
1.478 +
1.479 +
1.480 + if (This->fChecks & (USPOOF_CHAR_LIMIT)) {
1.481 + int32_t i;
1.482 + UChar32 c;
1.483 + int32_t length = id.length();
1.484 + for (i=0; i<length ;) {
1.485 + c = id.char32At(i);
1.486 + i += U16_LENGTH(c);
1.487 + if (!This->fAllowedCharsSet->contains(c)) {
1.488 + result |= USPOOF_CHAR_LIMIT;
1.489 + break;
1.490 + }
1.491 + }
1.492 + }
1.493 +
1.494 + if (This->fChecks &
1.495 + (USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE | USPOOF_INVISIBLE)) {
1.496 + // These are the checks that need to be done on NFD input
1.497 + UnicodeString nfdText;
1.498 + gNfdNormalizer->normalize(id, nfdText, *status);
1.499 + int32_t nfdLength = nfdText.length();
1.500 +
1.501 + if (This->fChecks & USPOOF_INVISIBLE) {
1.502 +
1.503 + // scan for more than one occurence of the same non-spacing mark
1.504 + // in a sequence of non-spacing marks.
1.505 + int32_t i;
1.506 + UChar32 c;
1.507 + UChar32 firstNonspacingMark = 0;
1.508 + UBool haveMultipleMarks = FALSE;
1.509 + UnicodeSet marksSeenSoFar; // Set of combining marks in a single combining sequence.
1.510 +
1.511 + for (i=0; i<nfdLength ;) {
1.512 + c = nfdText.char32At(i);
1.513 + i += U16_LENGTH(c);
1.514 + if (u_charType(c) != U_NON_SPACING_MARK) {
1.515 + firstNonspacingMark = 0;
1.516 + if (haveMultipleMarks) {
1.517 + marksSeenSoFar.clear();
1.518 + haveMultipleMarks = FALSE;
1.519 + }
1.520 + continue;
1.521 + }
1.522 + if (firstNonspacingMark == 0) {
1.523 + firstNonspacingMark = c;
1.524 + continue;
1.525 + }
1.526 + if (!haveMultipleMarks) {
1.527 + marksSeenSoFar.add(firstNonspacingMark);
1.528 + haveMultipleMarks = TRUE;
1.529 + }
1.530 + if (marksSeenSoFar.contains(c)) {
1.531 + // report the error, and stop scanning.
1.532 + // No need to find more than the first failure.
1.533 + result |= USPOOF_INVISIBLE;
1.534 + break;
1.535 + }
1.536 + marksSeenSoFar.add(c);
1.537 + }
1.538 + }
1.539 +
1.540 +
1.541 + if (This->fChecks & (USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE)) {
1.542 + // The basic test is the same for both whole and mixed script confusables.
1.543 + // Compute the set of scripts that every input character has a confusable in.
1.544 + // For this computation an input character is always considered to be
1.545 + // confusable with itself in its own script.
1.546 + //
1.547 + // If the number of such scripts is two or more, and the input consisted of
1.548 + // characters all from a single script, we have a whole script confusable.
1.549 + // (The two scripts will be the original script and the one that is confusable)
1.550 + //
1.551 + // If the number of such scripts >= one, and the original input contained characters from
1.552 + // more than one script, we have a mixed script confusable. (We can transform
1.553 + // some of the characters, and end up with a visually similar string all in
1.554 + // one script.)
1.555 +
1.556 + if (identifierInfo == NULL) {
1.557 + identifierInfo = This->getIdentifierInfo(*status);
1.558 + if (U_FAILURE(*status)) {
1.559 + goto cleanupAndReturn;
1.560 + }
1.561 + identifierInfo->setIdentifier(id, *status);
1.562 + }
1.563 +
1.564 + int32_t scriptCount = identifierInfo->getScriptCount();
1.565 +
1.566 + ScriptSet scripts;
1.567 + This->wholeScriptCheck(nfdText, &scripts, *status);
1.568 + int32_t confusableScriptCount = scripts.countMembers();
1.569 + //printf("confusableScriptCount = %d\n", confusableScriptCount);
1.570 +
1.571 + if ((This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE) &&
1.572 + confusableScriptCount >= 2 &&
1.573 + scriptCount == 1) {
1.574 + result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
1.575 + }
1.576 +
1.577 + if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) &&
1.578 + confusableScriptCount >= 1 &&
1.579 + scriptCount > 1) {
1.580 + result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
1.581 + }
1.582 + }
1.583 + }
1.584 +
1.585 +cleanupAndReturn:
1.586 + This->releaseIdentifierInfo(identifierInfo);
1.587 + if (position != NULL) {
1.588 + *position = 0;
1.589 + }
1.590 + return result;
1.591 +}
1.592 +
1.593 +
1.594 +U_CAPI int32_t U_EXPORT2
1.595 +uspoof_getSkeleton(const USpoofChecker *sc,
1.596 + uint32_t type,
1.597 + const UChar *id, int32_t length,
1.598 + UChar *dest, int32_t destCapacity,
1.599 + UErrorCode *status) {
1.600 +
1.601 + SpoofImpl::validateThis(sc, *status);
1.602 + if (U_FAILURE(*status)) {
1.603 + return 0;
1.604 + }
1.605 + if (length<-1 || destCapacity<0 || (destCapacity==0 && dest!=NULL)) {
1.606 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.607 + return 0;
1.608 + }
1.609 +
1.610 + UnicodeString idStr((length==-1), id, length); // Aliasing constructor
1.611 + UnicodeString destStr;
1.612 + uspoof_getSkeletonUnicodeString(sc, type, idStr, destStr, status);
1.613 + destStr.extract(dest, destCapacity, *status);
1.614 + return destStr.length();
1.615 +}
1.616 +
1.617 +
1.618 +
1.619 +U_I18N_API UnicodeString & U_EXPORT2
1.620 +uspoof_getSkeletonUnicodeString(const USpoofChecker *sc,
1.621 + uint32_t type,
1.622 + const UnicodeString &id,
1.623 + UnicodeString &dest,
1.624 + UErrorCode *status) {
1.625 + const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.626 + if (U_FAILURE(*status)) {
1.627 + return dest;
1.628 + }
1.629 +
1.630 + int32_t tableMask = 0;
1.631 + switch (type) {
1.632 + case 0:
1.633 + tableMask = USPOOF_ML_TABLE_FLAG;
1.634 + break;
1.635 + case USPOOF_SINGLE_SCRIPT_CONFUSABLE:
1.636 + tableMask = USPOOF_SL_TABLE_FLAG;
1.637 + break;
1.638 + case USPOOF_ANY_CASE:
1.639 + tableMask = USPOOF_MA_TABLE_FLAG;
1.640 + break;
1.641 + case USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_ANY_CASE:
1.642 + tableMask = USPOOF_SA_TABLE_FLAG;
1.643 + break;
1.644 + default:
1.645 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.646 + return dest;
1.647 + }
1.648 +
1.649 + UnicodeString nfdId;
1.650 + gNfdNormalizer->normalize(id, nfdId, *status);
1.651 +
1.652 + // Apply the skeleton mapping to the NFD normalized input string
1.653 + // Accumulate the skeleton, possibly unnormalized, in a UnicodeString.
1.654 + int32_t inputIndex = 0;
1.655 + UnicodeString skelStr;
1.656 + int32_t normalizedLen = nfdId.length();
1.657 + for (inputIndex=0; inputIndex < normalizedLen; ) {
1.658 + UChar32 c = nfdId.char32At(inputIndex);
1.659 + inputIndex += U16_LENGTH(c);
1.660 + This->confusableLookup(c, tableMask, skelStr);
1.661 + }
1.662 +
1.663 + gNfdNormalizer->normalize(skelStr, dest, *status);
1.664 + return dest;
1.665 +}
1.666 +
1.667 +
1.668 +U_CAPI int32_t U_EXPORT2
1.669 +uspoof_getSkeletonUTF8(const USpoofChecker *sc,
1.670 + uint32_t type,
1.671 + const char *id, int32_t length,
1.672 + char *dest, int32_t destCapacity,
1.673 + UErrorCode *status) {
1.674 + SpoofImpl::validateThis(sc, *status);
1.675 + if (U_FAILURE(*status)) {
1.676 + return 0;
1.677 + }
1.678 + if (length<-1 || destCapacity<0 || (destCapacity==0 && dest!=NULL)) {
1.679 + *status = U_ILLEGAL_ARGUMENT_ERROR;
1.680 + return 0;
1.681 + }
1.682 +
1.683 + UnicodeString srcStr = UnicodeString::fromUTF8(StringPiece(id, length>=0 ? length : uprv_strlen(id)));
1.684 + UnicodeString destStr;
1.685 + uspoof_getSkeletonUnicodeString(sc, type, srcStr, destStr, status);
1.686 + if (U_FAILURE(*status)) {
1.687 + return 0;
1.688 + }
1.689 +
1.690 + int32_t lengthInUTF8 = 0;
1.691 + u_strToUTF8(dest, destCapacity, &lengthInUTF8,
1.692 + destStr.getBuffer(), destStr.length(), status);
1.693 + return lengthInUTF8;
1.694 +}
1.695 +
1.696 +
1.697 +U_CAPI int32_t U_EXPORT2
1.698 +uspoof_serialize(USpoofChecker *sc,void *buf, int32_t capacity, UErrorCode *status) {
1.699 + SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
1.700 + if (This == NULL) {
1.701 + U_ASSERT(U_FAILURE(*status));
1.702 + return 0;
1.703 + }
1.704 + int32_t dataSize = This->fSpoofData->fRawData->fLength;
1.705 + if (capacity < dataSize) {
1.706 + *status = U_BUFFER_OVERFLOW_ERROR;
1.707 + return dataSize;
1.708 + }
1.709 + uprv_memcpy(buf, This->fSpoofData->fRawData, dataSize);
1.710 + return dataSize;
1.711 +}
1.712 +
1.713 +U_CAPI const USet * U_EXPORT2
1.714 +uspoof_getInclusionSet(UErrorCode *) {
1.715 + initializeStatics();
1.716 + return gInclusionSet->toUSet();
1.717 +}
1.718 +
1.719 +U_CAPI const USet * U_EXPORT2
1.720 +uspoof_getRecommendedSet(UErrorCode *) {
1.721 + initializeStatics();
1.722 + return gRecommendedSet->toUSet();
1.723 +}
1.724 +
1.725 +U_I18N_API const UnicodeSet * U_EXPORT2
1.726 +uspoof_getInclusionUnicodeSet(UErrorCode *) {
1.727 + initializeStatics();
1.728 + return gInclusionSet;
1.729 +}
1.730 +
1.731 +U_I18N_API const UnicodeSet * U_EXPORT2
1.732 +uspoof_getRecommendedUnicodeSet(UErrorCode *) {
1.733 + initializeStatics();
1.734 + return gRecommendedSet;
1.735 +}
1.736 +
1.737 +
1.738 +
1.739 +#endif // !UCONFIG_NO_NORMALIZATION
