icuSources/i18n/uspoof.cpp

   1 /*
   2 ***************************************************************************
   3 * Copyright (C) 2008-2015, International Business Machines Corporation
   4 * and others. All Rights Reserved.
   5 ***************************************************************************
   6 *   file name:  uspoof.cpp
   7 *   encoding:   US-ASCII
   8 *   tab size:   8 (not used)
   9 *   indentation:4
  10 *
  11 *   created on: 2008Feb13
  12 *   created by: Andy Heninger
  13 *
  14 *   Unicode Spoof Detection
  15 */
  16 #include "unicode/utypes.h"
  17 #include "unicode/normalizer2.h"
  18 #include "unicode/uspoof.h"
  19 #include "unicode/ustring.h"
  20 #include "unicode/utf16.h"
  21 #include "cmemory.h"
  22 #include "cstring.h"
  23 #include "identifier_info.h"
  24 #include "mutex.h"
  25 #include "scriptset.h"
  26 #include "uassert.h"
  27 #include "ucln_in.h"
  28 #include "uspoof_impl.h"
  29 #include "umutex.h"
  30
  31
  32 #if !UCONFIG_NO_NORMALIZATION
  33
  34 U_NAMESPACE_USE
  35
  36
  37 //
  38 // Static Objects used by the spoof impl, their thread safe initialization and their cleanup.
  39 //
  40 static UnicodeSet *gInclusionSet = NULL;
  41 static UnicodeSet *gRecommendedSet = NULL;
  42 static const Normalizer2 *gNfdNormalizer = NULL;
  43 static SpoofData *gDefaultSpoofData = NULL;
  44 static UInitOnce gSpoofInitStaticsOnce = U_INITONCE_INITIALIZER;
  45 static UInitOnce gSpoofInitDefaultOnce = U_INITONCE_INITIALIZER;
  46
  47 static UBool U_CALLCONV
  48 uspoof_cleanup(void) {
  49     delete gInclusionSet;
  50     gInclusionSet = NULL;
  51     delete gRecommendedSet;
  52     gRecommendedSet = NULL;
  53     gNfdNormalizer = NULL;
  54     if (gDefaultSpoofData) {
  55         gDefaultSpoofData->removeReference();   // Will delete, assuming all user-level spoof checkers were closed.
  56     }
  57     gDefaultSpoofData = NULL;
  58     gSpoofInitStaticsOnce.reset();
  59     gSpoofInitDefaultOnce.reset();
  60     return TRUE;
  61 }
  62
  63 static void U_CALLCONV initializeStatics(UErrorCode &status) {
  64     static const char *inclusionPat =
  65            "[\\u0027\\u002D-\\u002E\\u003A\\u00B7\\u0375\\u058A\\u05F3-\\u05F4"
  66            "\\u06FD-\\u06FE\\u0F0B\\u200C-\\u200D\\u2010\\u2019\\u2027\\u30A0\\u30FB]";
  67     gInclusionSet = new UnicodeSet(UnicodeString(inclusionPat, -1, US_INV), status);
  68     gInclusionSet->freeze();
  69
  70     // Note: data from http://unicode.org/Public/security/latest/xidmodifications.txt version 8.0.0
  71     //       There is no tooling to generate this from the .txt file, hand extracted with editor macros.
  72     //       Ultimately, data will be available as character properties, eliminating this.
  73     // Note: concatenated string constants do not work with UNICODE_STRING_SIMPLE on all platforms.
  74     static const char *recommendedPat =
  75             "[\\u0030-\\u0039\\u0041-\\u005A\\u005F\\u0061-\\u007A\\u00C0-\\u00D6\\u00D8-\\u00F6"
  76             "\\u00F8-\\u0131\\u0134-\\u013E\\u0141-\\u0148\\u014A-\\u017E\\u018F\\u01A0-\\u01A1"
  77             "\\u01AF-\\u01B0\\u01CD-\\u01DC\\u01DE-\\u01E3\\u01E6-\\u01F0\\u01F4-\\u01F5\\u01F8-\\u021B"
  78             "\\u021E-\\u021F\\u0226-\\u0233\\u0259\\u02BB-\\u02BC\\u02EC\\u0300-\\u0304\\u0306-\\u030C"
  79             "\\u030F-\\u0311\\u0313-\\u0314\\u031B\\u0323-\\u0328\\u032D-\\u032E\\u0330-\\u0331"
  80             "\\u0335\\u0338-\\u0339\\u0342\\u0345\\u037B-\\u037D\\u0386\\u0388-\\u038A\\u038C"
  81             "\\u038E-\\u03A1\\u03A3-\\u03CE\\u03FC-\\u045F\\u048A-\\u0529\\u052E-\\u052F\\u0531-\\u0556"
  82             "\\u0559\\u0561-\\u0586\\u05B4\\u05D0-\\u05EA\\u05F0-\\u05F2\\u0620-\\u063F\\u0641-\\u0655"
  83             "\\u0660-\\u0669\\u0670-\\u0672\\u0674\\u0679-\\u068D\\u068F-\\u06D3\\u06D5\\u06E5-\\u06E6"
  84             "\\u06EE-\\u06FC\\u06FF\\u0750-\\u07B1\\u08A0-\\u08AC\\u08B2\\u0901-\\u094D\\u094F-\\u0950"
  85             "\\u0956-\\u0957\\u0960-\\u0963\\u0966-\\u096F\\u0971-\\u0977\\u0979-\\u097F\\u0981-\\u0983"
  86             "\\u0985-\\u098C\\u098F-\\u0990\\u0993-\\u09A8\\u09AA-\\u09B0\\u09B2\\u09B6-\\u09B9"
  87             "\\u09BC-\\u09C4\\u09C7-\\u09C8\\u09CB-\\u09CE\\u09D7\\u09E0-\\u09E3\\u09E6-\\u09F1"
  88             "\\u0A01-\\u0A03\\u0A05-\\u0A0A\\u0A0F-\\u0A10\\u0A13-\\u0A28\\u0A2A-\\u0A30\\u0A32"
  89             "\\u0A35\\u0A38-\\u0A39\\u0A3C\\u0A3E-\\u0A42\\u0A47-\\u0A48\\u0A4B-\\u0A4D\\u0A5C"
  90             "\\u0A66-\\u0A74\\u0A81-\\u0A83\\u0A85-\\u0A8D\\u0A8F-\\u0A91\\u0A93-\\u0AA8\\u0AAA-\\u0AB0"
  91             "\\u0AB2-\\u0AB3\\u0AB5-\\u0AB9\\u0ABC-\\u0AC5\\u0AC7-\\u0AC9\\u0ACB-\\u0ACD\\u0AD0"
  92             "\\u0AE0-\\u0AE3\\u0AE6-\\u0AEF\\u0B01-\\u0B03\\u0B05-\\u0B0C\\u0B0F-\\u0B10\\u0B13-\\u0B28"
  93             "\\u0B2A-\\u0B30\\u0B32-\\u0B33\\u0B35-\\u0B39\\u0B3C-\\u0B43\\u0B47-\\u0B48\\u0B4B-\\u0B4D"
  94             "\\u0B56-\\u0B57\\u0B5F-\\u0B61\\u0B66-\\u0B6F\\u0B71\\u0B82-\\u0B83\\u0B85-\\u0B8A"
  95             "\\u0B8E-\\u0B90\\u0B92-\\u0B95\\u0B99-\\u0B9A\\u0B9C\\u0B9E-\\u0B9F\\u0BA3-\\u0BA4"
  96             "\\u0BA8-\\u0BAA\\u0BAE-\\u0BB9\\u0BBE-\\u0BC2\\u0BC6-\\u0BC8\\u0BCA-\\u0BCD\\u0BD0"
  97             "\\u0BD7\\u0BE6-\\u0BEF\\u0C01-\\u0C03\\u0C05-\\u0C0C\\u0C0E-\\u0C10\\u0C12-\\u0C28"
  98             "\\u0C2A-\\u0C33\\u0C35-\\u0C39\\u0C3D-\\u0C44\\u0C46-\\u0C48\\u0C4A-\\u0C4D\\u0C55-\\u0C56"
  99             "\\u0C60-\\u0C61\\u0C66-\\u0C6F\\u0C82-\\u0C83\\u0C85-\\u0C8C\\u0C8E-\\u0C90\\u0C92-\\u0CA8"
 100             "\\u0CAA-\\u0CB3\\u0CB5-\\u0CB9\\u0CBC-\\u0CC4\\u0CC6-\\u0CC8\\u0CCA-\\u0CCD\\u0CD5-\\u0CD6"
 101             "\\u0CE0-\\u0CE3\\u0CE6-\\u0CEF\\u0CF1-\\u0CF2\\u0D02-\\u0D03\\u0D05-\\u0D0C\\u0D0E-\\u0D10"
 102             "\\u0D12-\\u0D3A\\u0D3D-\\u0D43\\u0D46-\\u0D48\\u0D4A-\\u0D4E\\u0D57\\u0D60-\\u0D61"
 103             "\\u0D66-\\u0D6F\\u0D7A-\\u0D7F\\u0D82-\\u0D83\\u0D85-\\u0D8E\\u0D91-\\u0D96\\u0D9A-\\u0DA5"
 104             "\\u0DA7-\\u0DB1\\u0DB3-\\u0DBB\\u0DBD\\u0DC0-\\u0DC6\\u0DCA\\u0DCF-\\u0DD4\\u0DD6"
 105             "\\u0DD8-\\u0DDE\\u0DF2\\u0E01-\\u0E32\\u0E34-\\u0E3A\\u0E40-\\u0E4E\\u0E50-\\u0E59"
 106             "\\u0E81-\\u0E82\\u0E84\\u0E87-\\u0E88\\u0E8A\\u0E8D\\u0E94-\\u0E97\\u0E99-\\u0E9F"
 107             "\\u0EA1-\\u0EA3\\u0EA5\\u0EA7\\u0EAA-\\u0EAB\\u0EAD-\\u0EB2\\u0EB4-\\u0EB9\\u0EBB-\\u0EBD"
 108             "\\u0EC0-\\u0EC4\\u0EC6\\u0EC8-\\u0ECD\\u0ED0-\\u0ED9\\u0EDE-\\u0EDF\\u0F00\\u0F20-\\u0F29"
 109             "\\u0F35\\u0F37\\u0F3E-\\u0F42\\u0F44-\\u0F47\\u0F49-\\u0F4C\\u0F4E-\\u0F51\\u0F53-\\u0F56"
 110             "\\u0F58-\\u0F5B\\u0F5D-\\u0F68\\u0F6A-\\u0F6C\\u0F71-\\u0F72\\u0F74\\u0F7A-\\u0F80"
 111             "\\u0F82-\\u0F84\\u0F86-\\u0F92\\u0F94-\\u0F97\\u0F99-\\u0F9C\\u0F9E-\\u0FA1\\u0FA3-\\u0FA6"
 112             "\\u0FA8-\\u0FAB\\u0FAD-\\u0FB8\\u0FBA-\\u0FBC\\u0FC6\\u1000-\\u1049\\u1050-\\u109D"
 113             "\\u10C7\\u10CD\\u10D0-\\u10F0\\u10F7-\\u10FA\\u10FD-\\u10FF\\u1200-\\u1248\\u124A-\\u124D"
 114             "\\u1250-\\u1256\\u1258\\u125A-\\u125D\\u1260-\\u1288\\u128A-\\u128D\\u1290-\\u12B0"
 115             "\\u12B2-\\u12B5\\u12B8-\\u12BE\\u12C0\\u12C2-\\u12C5\\u12C8-\\u12D6\\u12D8-\\u1310"
 116             "\\u1312-\\u1315\\u1318-\\u135A\\u135D-\\u135F\\u1380-\\u138F\\u1780-\\u17A2\\u17A5-\\u17A7"
 117             "\\u17A9-\\u17B3\\u17B6-\\u17CA\\u17D2\\u17D7\\u17DC\\u17E0-\\u17E9\\u1E00-\\u1E99"
 118             "\\u1E9E\\u1EA0-\\u1EF9\\u1F00-\\u1F15\\u1F18-\\u1F1D\\u1F20-\\u1F45\\u1F48-\\u1F4D"
 119             "\\u1F50-\\u1F57\\u1F59\\u1F5B\\u1F5D\\u1F5F-\\u1F70\\u1F72\\u1F74\\u1F76\\u1F78"
 120             "\\u1F7A\\u1F7C\\u1F80-\\u1FB4\\u1FB6-\\u1FBA\\u1FBC\\u1FC2-\\u1FC4\\u1FC6-\\u1FC8"
 121             "\\u1FCA\\u1FCC\\u1FD0-\\u1FD2\\u1FD6-\\u1FDA\\u1FE0-\\u1FE2\\u1FE4-\\u1FEA\\u1FEC"
 122             "\\u1FF2-\\u1FF4\\u1FF6-\\u1FF8\\u1FFA\\u1FFC\\u2D27\\u2D2D\\u2D80-\\u2D96\\u2DA0-\\u2DA6"
 123             "\\u2DA8-\\u2DAE\\u2DB0-\\u2DB6\\u2DB8-\\u2DBE\\u2DC0-\\u2DC6\\u2DC8-\\u2DCE\\u2DD0-\\u2DD6"
 124             "\\u2DD8-\\u2DDE\\u3005-\\u3007\\u3041-\\u3096\\u3099-\\u309A\\u309D-\\u309E\\u30A1-\\u30FA"
 125             "\\u30FC-\\u30FE\\u3105-\\u312D\\u31A0-\\u31BA\\u3400-\\u4DB5\\u4E00-\\u9FD5\\uA660-\\uA661"
 126             "\\uA674-\\uA67B\\uA67F\\uA69F\\uA717-\\uA71F\\uA788\\uA78D-\\uA78E\\uA790-\\uA793"
 127             "\\uA7A0-\\uA7AA\\uA7FA\\uA9E7-\\uA9FE\\uAA60-\\uAA76\\uAA7A-\\uAA7F\\uAB01-\\uAB06"
 128             "\\uAB09-\\uAB0E\\uAB11-\\uAB16\\uAB20-\\uAB26\\uAB28-\\uAB2E\\uAC00-\\uD7A3\\uFA0E-\\uFA0F"
 129             "\\uFA11\\uFA13-\\uFA14\\uFA1F\\uFA21\\uFA23-\\uFA24\\uFA27-\\uFA29\\U00020000-\\U0002A6D6"
 130             "\\U0002A700-\\U0002B734\\U0002B740-\\U0002B81D\\U0002B820-\\U0002CEA1]";
 131
 132     gRecommendedSet = new UnicodeSet(UnicodeString(recommendedPat, -1, US_INV), status);
 133     gRecommendedSet->freeze();
 134     gNfdNormalizer = Normalizer2::getNFDInstance(status);
 135     ucln_i18n_registerCleanup(UCLN_I18N_SPOOF, uspoof_cleanup);
 136 }
 137
 138 static void U_CALLCONV initializeDefaultData(UErrorCode &status) {
 139     gDefaultSpoofData = SpoofData::getDefault(status);
 140     ucln_i18n_registerCleanup(UCLN_I18N_SPOOF, uspoof_cleanup);
 141 }
 142
 143 U_CFUNC void uspoof_internalInitStatics(UErrorCode *status) {
 144     umtx_initOnce(gSpoofInitStaticsOnce, &initializeStatics, *status);
 145 }
 146
 147 U_CAPI USpoofChecker * U_EXPORT2
 148 uspoof_open(UErrorCode *status) {
 149     umtx_initOnce(gSpoofInitStaticsOnce, &initializeStatics, *status);
 150     umtx_initOnce(gSpoofInitDefaultOnce, &initializeDefaultData, *status);
 151     if (U_FAILURE(*status)) {
 152         return NULL;
 153     }
 154     SpoofImpl *si = new SpoofImpl(gDefaultSpoofData, *status);
 155     if (si) {
 156         gDefaultSpoofData->addReference();
 157     }
 158     if (U_SUCCESS(*status) && si == NULL) {
 159         *status = U_MEMORY_ALLOCATION_ERROR;
 160     }
 161     if (U_FAILURE(*status)) {
 162         delete si;
 163         si = NULL;
 164     }
 165     return reinterpret_cast<USpoofChecker *>(si);
 166 }
 167
 168
 169 U_CAPI USpoofChecker * U_EXPORT2
 170 uspoof_openFromSerialized(const void *data, int32_t length, int32_t *pActualLength,
 171                           UErrorCode *status) {
 172     if (U_FAILURE(*status)) {
 173         return NULL;
 174     }
 175     umtx_initOnce(gSpoofInitStaticsOnce, &initializeStatics, *status);
 176     SpoofData *sd = new SpoofData(data, length, *status);
 177     SpoofImpl *si = new SpoofImpl(sd, *status);
 178     if (U_FAILURE(*status)) {
 179         delete sd;
 180         delete si;
 181         return NULL;
 182     }
 183     if (sd == NULL || si == NULL) {
 184         *status = U_MEMORY_ALLOCATION_ERROR;
 185         delete sd;
 186         delete si;
 187         return NULL;
 188     }
 189
 190     if (pActualLength != NULL) {
 191         *pActualLength = sd->fRawData->fLength;
 192     }
 193     return reinterpret_cast<USpoofChecker *>(si);
 194 }
 195
 196
 197 U_CAPI USpoofChecker * U_EXPORT2
 198 uspoof_clone(const USpoofChecker *sc, UErrorCode *status) {
 199     const SpoofImpl *src = SpoofImpl::validateThis(sc, *status);
 200     if (src == NULL) {
 201         return NULL;
 202     }
 203     SpoofImpl *result = new SpoofImpl(*src, *status);   // copy constructor
 204     if (U_FAILURE(*status)) {
 205         delete result;
 206         result = NULL;
 207     }
 208     return reinterpret_cast<USpoofChecker *>(result);
 209 }
 210
 211
 212 U_CAPI void U_EXPORT2
 213 uspoof_close(USpoofChecker *sc) {
 214     UErrorCode status = U_ZERO_ERROR;
 215     SpoofImpl *This = SpoofImpl::validateThis(sc, status);
 216     delete This;
 217 }
 218
 219
 220 U_CAPI void U_EXPORT2
 221 uspoof_setChecks(USpoofChecker *sc, int32_t checks, UErrorCode *status) {
 222     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 223     if (This == NULL) {
 224         return;
 225     }
 226
 227     // Verify that the requested checks are all ones (bits) that
 228     //   are acceptable, known values.
 229     if (checks & ~(USPOOF_ALL_CHECKS | USPOOF_AUX_INFO)) {
 230         *status = U_ILLEGAL_ARGUMENT_ERROR;
 231         return;
 232     }
 233
 234     This->fChecks = checks;
 235 }
 236
 237
 238 U_CAPI int32_t U_EXPORT2
 239 uspoof_getChecks(const USpoofChecker *sc, UErrorCode *status) {
 240     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 241     if (This == NULL) {
 242         return 0;
 243     }
 244     return This->fChecks;
 245 }
 246
 247 U_CAPI void U_EXPORT2
 248 uspoof_setRestrictionLevel(USpoofChecker *sc, URestrictionLevel restrictionLevel) {
 249     UErrorCode status = U_ZERO_ERROR;
 250     SpoofImpl *This = SpoofImpl::validateThis(sc, status);
 251     if (This != NULL) {
 252         This->fRestrictionLevel = restrictionLevel;
 253     }
 254 }
 255
 256 U_CAPI URestrictionLevel U_EXPORT2
 257 uspoof_getRestrictionLevel(const USpoofChecker *sc) {
 258     UErrorCode status = U_ZERO_ERROR;
 259     const SpoofImpl *This = SpoofImpl::validateThis(sc, status);
 260     if (This == NULL) {
 261         return USPOOF_UNRESTRICTIVE;
 262     }
 263     return This->fRestrictionLevel;
 264 }
 265
 266 U_CAPI void U_EXPORT2
 267 uspoof_setAllowedLocales(USpoofChecker *sc, const char *localesList, UErrorCode *status) {
 268     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 269     if (This == NULL) {
 270         return;
 271     }
 272     This->setAllowedLocales(localesList, *status);
 273 }
 274
 275 U_CAPI const char * U_EXPORT2
 276 uspoof_getAllowedLocales(USpoofChecker *sc, UErrorCode *status) {
 277     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 278     if (This == NULL) {
 279         return NULL;
 280     }
 281     return This->getAllowedLocales(*status);
 282 }
 283
 284
 285 U_CAPI const USet * U_EXPORT2
 286 uspoof_getAllowedChars(const USpoofChecker *sc, UErrorCode *status) {
 287     const UnicodeSet *result = uspoof_getAllowedUnicodeSet(sc, status);
 288     return result->toUSet();
 289 }
 290
 291 U_CAPI const UnicodeSet * U_EXPORT2
 292 uspoof_getAllowedUnicodeSet(const USpoofChecker *sc, UErrorCode *status) {
 293     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 294     if (This == NULL) {
 295         return NULL;
 296     }
 297     return This->fAllowedCharsSet;
 298 }
 299
 300
 301 U_CAPI void U_EXPORT2
 302 uspoof_setAllowedChars(USpoofChecker *sc, const USet *chars, UErrorCode *status) {
 303     const UnicodeSet *set = UnicodeSet::fromUSet(chars);
 304     uspoof_setAllowedUnicodeSet(sc, set, status);
 305 }
 306
 307
 308 U_CAPI void U_EXPORT2
 309 uspoof_setAllowedUnicodeSet(USpoofChecker *sc, const UnicodeSet *chars, UErrorCode *status) {
 310     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 311     if (This == NULL) {
 312         return;
 313     }
 314     if (chars->isBogus()) {
 315         *status = U_ILLEGAL_ARGUMENT_ERROR;
 316         return;
 317     }
 318     UnicodeSet *clonedSet = static_cast<UnicodeSet *>(chars->clone());
 319     if (clonedSet == NULL || clonedSet->isBogus()) {
 320         *status = U_MEMORY_ALLOCATION_ERROR;
 321         return;
 322     }
 323     clonedSet->freeze();
 324     delete This->fAllowedCharsSet;
 325     This->fAllowedCharsSet = clonedSet;
 326     This->fChecks |= USPOOF_CHAR_LIMIT;
 327 }
 328
 329
 330 U_CAPI int32_t U_EXPORT2
 331 uspoof_check(const USpoofChecker *sc,
 332              const UChar *id, int32_t length,
 333              int32_t *position,
 334              UErrorCode *status) {
 335
 336     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 337     if (This == NULL) {
 338         return 0;
 339     }
 340     if (length < -1) {
 341         *status = U_ILLEGAL_ARGUMENT_ERROR;
 342         return 0;
 343     }
 344     UnicodeString idStr((length == -1), id, length);  // Aliasing constructor.
 345     int32_t result = uspoof_checkUnicodeString(sc, idStr, position, status);
 346     return result;
 347 }
 348
 349
 350 U_CAPI int32_t U_EXPORT2
 351 uspoof_checkUTF8(const USpoofChecker *sc,
 352                  const char *id, int32_t length,
 353                  int32_t *position,
 354                  UErrorCode *status) {
 355
 356     if (U_FAILURE(*status)) {
 357         return 0;
 358     }
 359     UnicodeString idStr = UnicodeString::fromUTF8(StringPiece(id, length>=0 ? length : uprv_strlen(id)));
 360     int32_t result = uspoof_checkUnicodeString(sc, idStr, position, status);
 361     return result;
 362 }
 363
 364
 365 U_CAPI int32_t U_EXPORT2
 366 uspoof_areConfusable(const USpoofChecker *sc,
 367                      const UChar *id1, int32_t length1,
 368                      const UChar *id2, int32_t length2,
 369                      UErrorCode *status) {
 370     SpoofImpl::validateThis(sc, *status);
 371     if (U_FAILURE(*status)) {
 372         return 0;
 373     }
 374     if (length1 < -1 || length2 < -1) {
 375         *status = U_ILLEGAL_ARGUMENT_ERROR;
 376         return 0;
 377     }
 378
 379     UnicodeString id1Str((length1==-1), id1, length1);  // Aliasing constructor
 380     UnicodeString id2Str((length2==-1), id2, length2);  // Aliasing constructor
 381     return uspoof_areConfusableUnicodeString(sc, id1Str, id2Str, status);
 382 }
 383
 384
 385 U_CAPI int32_t U_EXPORT2
 386 uspoof_areConfusableUTF8(const USpoofChecker *sc,
 387                          const char *id1, int32_t length1,
 388                          const char *id2, int32_t length2,
 389                          UErrorCode *status) {
 390     SpoofImpl::validateThis(sc, *status);
 391     if (U_FAILURE(*status)) {
 392         return 0;
 393     }
 394     if (length1 < -1 || length2 < -1) {
 395         *status = U_ILLEGAL_ARGUMENT_ERROR;
 396         return 0;
 397     }
 398     UnicodeString id1Str = UnicodeString::fromUTF8(StringPiece(id1, length1>=0? length1 : uprv_strlen(id1)));
 399     UnicodeString id2Str = UnicodeString::fromUTF8(StringPiece(id2, length2>=0? length2 : uprv_strlen(id2)));
 400     int32_t results = uspoof_areConfusableUnicodeString(sc, id1Str, id2Str, status);
 401     return results;
 402 }
 403
 404
 405 U_CAPI int32_t U_EXPORT2
 406 uspoof_areConfusableUnicodeString(const USpoofChecker *sc,
 407                                   const icu::UnicodeString &id1,
 408                                   const icu::UnicodeString &id2,
 409                                   UErrorCode *status) {
 410     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 411     if (U_FAILURE(*status)) {
 412         return 0;
 413     }
 414     //
 415     // See section 4 of UAX 39 for the algorithm for checking whether two strings are confusable,
 416     //   and for definitions of the types (single, whole, mixed-script) of confusables.
 417
 418     // We only care about a few of the check flags.  Ignore the others.
 419     // If no tests relavant to this function have been specified, return an error.
 420     // TODO:  is this really the right thing to do?  It's probably an error on the caller's part,
 421     //        but logically we would just return 0 (no error).
 422     if ((This->fChecks & (USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE |
 423                           USPOOF_WHOLE_SCRIPT_CONFUSABLE)) == 0) {
 424         *status = U_INVALID_STATE_ERROR;
 425         return 0;
 426     }
 427     int32_t  flagsForSkeleton = This->fChecks & USPOOF_ANY_CASE;
 428
 429     int32_t  result = 0;
 430     IdentifierInfo *identifierInfo = This->getIdentifierInfo(*status);
 431     if (U_FAILURE(*status)) {
 432         return 0;
 433     }
 434     identifierInfo->setIdentifier(id1, *status);
 435     int32_t id1ScriptCount = identifierInfo->getScriptCount();
 436     int32_t id1FirstScript = identifierInfo->getScripts()->nextSetBit(0);
 437     identifierInfo->setIdentifier(id2, *status);
 438     int32_t id2ScriptCount = identifierInfo->getScriptCount();
 439     int32_t id2FirstScript = identifierInfo->getScripts()->nextSetBit(0);
 440     This->releaseIdentifierInfo(identifierInfo);
 441     identifierInfo = NULL;
 442
 443     if (This->fChecks & USPOOF_SINGLE_SCRIPT_CONFUSABLE) {
 444         UnicodeString   id1Skeleton;
 445         UnicodeString   id2Skeleton;
 446         if (id1ScriptCount <= 1 && id2ScriptCount <= 1 && id1FirstScript == id2FirstScript) {
 447             flagsForSkeleton |= USPOOF_SINGLE_SCRIPT_CONFUSABLE;
 448             uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id1, id1Skeleton, status);
 449             uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id2, id2Skeleton, status);
 450             if (id1Skeleton == id2Skeleton) {
 451                 result |= USPOOF_SINGLE_SCRIPT_CONFUSABLE;
 452             }
 453         }
 454     }
 455
 456     if (result & USPOOF_SINGLE_SCRIPT_CONFUSABLE) {
 457          // If the two inputs are single script confusable they cannot also be
 458          // mixed or whole script confusable, according to the UAX39 definitions.
 459          // So we can skip those tests.
 460          return result;
 461     }
 462
 463     // Two identifiers are whole script confusable if each is of a single script
 464     // and they are mixed script confusable.
 465     UBool possiblyWholeScriptConfusables =
 466         id1ScriptCount <= 1 && id2ScriptCount <= 1 && (This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE);
 467
 468     //
 469     // Mixed Script Check
 470     //
 471     if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) || possiblyWholeScriptConfusables ) {
 472         // For getSkeleton(), resetting the USPOOF_SINGLE_SCRIPT_CONFUSABLE flag will get us
 473         // the mixed script table skeleton, which is what we want.
 474         // The Any Case / Lower Case bit in the skelton flags was set at the top of the function.
 475         UnicodeString id1Skeleton;
 476         UnicodeString id2Skeleton;
 477         flagsForSkeleton &= ~USPOOF_SINGLE_SCRIPT_CONFUSABLE;
 478         uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id1, id1Skeleton, status);
 479         uspoof_getSkeletonUnicodeString(sc, flagsForSkeleton, id2, id2Skeleton, status);
 480         if (id1Skeleton == id2Skeleton) {
 481             result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
 482             if (possiblyWholeScriptConfusables) {
 483                 result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
 484             }
 485         }
 486     }
 487
 488     return result;
 489 }
 490
 491
 492
 493
 494 U_CAPI int32_t U_EXPORT2
 495 uspoof_checkUnicodeString(const USpoofChecker *sc,
 496                           const icu::UnicodeString &id,
 497                           int32_t *position,
 498                           UErrorCode *status) {
 499     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 500     if (This == NULL) {
 501         return 0;
 502     }
 503     int32_t result = 0;
 504
 505     IdentifierInfo *identifierInfo = NULL;
 506     if ((This->fChecks) & (USPOOF_RESTRICTION_LEVEL | USPOOF_MIXED_NUMBERS)) {
 507         identifierInfo = This->getIdentifierInfo(*status);
 508         if (U_FAILURE(*status)) {
 509             goto cleanupAndReturn;
 510         }
 511         identifierInfo->setIdentifier(id, *status);
 512         identifierInfo->setIdentifierProfile(*This->fAllowedCharsSet);
 513     }
 514
 515
 516     if ((This->fChecks) & USPOOF_RESTRICTION_LEVEL) {
 517         URestrictionLevel idRestrictionLevel = identifierInfo->getRestrictionLevel(*status);
 518         if (idRestrictionLevel > This->fRestrictionLevel) {
 519             result |= USPOOF_RESTRICTION_LEVEL;
 520         }
 521         if (This->fChecks & USPOOF_AUX_INFO) {
 522             result |= idRestrictionLevel;
 523         }
 524     }
 525
 526     if ((This->fChecks) & USPOOF_MIXED_NUMBERS) {
 527         const UnicodeSet *numerics = identifierInfo->getNumerics();
 528         if (numerics->size() > 1) {
 529             result |= USPOOF_MIXED_NUMBERS;
 530         }
 531
 532         // TODO: ICU4J returns the UnicodeSet of the numerics found in the identifier.
 533         //       We have no easy way to do the same in C.
 534         // if (checkResult != null) {
 535         //     checkResult.numerics = numerics;
 536         // }
 537     }
 538
 539
 540     if (This->fChecks & (USPOOF_CHAR_LIMIT)) {
 541         int32_t i;
 542         UChar32 c;
 543         int32_t length = id.length();
 544         for (i=0; i<length ;) {
 545             c = id.char32At(i);
 546             i += U16_LENGTH(c);
 547             if (!This->fAllowedCharsSet->contains(c)) {
 548                 result |= USPOOF_CHAR_LIMIT;
 549                 break;
 550             }
 551         }
 552     }
 553
 554     if (This->fChecks &
 555         (USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE | USPOOF_INVISIBLE)) {
 556         // These are the checks that need to be done on NFD input
 557         UnicodeString nfdText;
 558         gNfdNormalizer->normalize(id, nfdText, *status);
 559         int32_t nfdLength = nfdText.length();
 560
 561         if (This->fChecks & USPOOF_INVISIBLE) {
 562
 563             // scan for more than one occurence of the same non-spacing mark
 564             // in a sequence of non-spacing marks.
 565             int32_t     i;
 566             UChar32     c;
 567             UChar32     firstNonspacingMark = 0;
 568             UBool       haveMultipleMarks = FALSE;
 569             UnicodeSet  marksSeenSoFar;   // Set of combining marks in a single combining sequence.
 570
 571             for (i=0; i<nfdLength ;) {
 572                 c = nfdText.char32At(i);
 573                 i += U16_LENGTH(c);
 574                 if (u_charType(c) != U_NON_SPACING_MARK) {
 575                     firstNonspacingMark = 0;
 576                     if (haveMultipleMarks) {
 577                         marksSeenSoFar.clear();
 578                         haveMultipleMarks = FALSE;
 579                     }
 580                     continue;
 581                 }
 582                 if (firstNonspacingMark == 0) {
 583                     firstNonspacingMark = c;
 584                     continue;
 585                 }
 586                 if (!haveMultipleMarks) {
 587                     marksSeenSoFar.add(firstNonspacingMark);
 588                     haveMultipleMarks = TRUE;
 589                 }
 590                 if (marksSeenSoFar.contains(c)) {
 591                     // report the error, and stop scanning.
 592                     // No need to find more than the first failure.
 593                     result |= USPOOF_INVISIBLE;
 594                     break;
 595                 }
 596                 marksSeenSoFar.add(c);
 597             }
 598         }
 599
 600
 601         if (This->fChecks & (USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE)) {
 602             // The basic test is the same for both whole and mixed script confusables.
 603             // Compute the set of scripts that every input character has a confusable in.
 604             // For this computation an input character is always considered to be
 605             // confusable with itself in its own script.
 606             //
 607             // If the number of such scripts is two or more, and the input consisted of
 608             // characters all from a single script, we have a whole script confusable.
 609             // (The two scripts will be the original script and the one that is confusable)
 610             //
 611             // If the number of such scripts >= one, and the original input contained characters from
 612             // more than one script, we have a mixed script confusable.  (We can transform
 613             // some of the characters, and end up with a visually similar string all in
 614             // one script.)
 615
 616             if (identifierInfo == NULL) {
 617                 identifierInfo = This->getIdentifierInfo(*status);
 618                 if (U_FAILURE(*status)) {
 619                     goto cleanupAndReturn;
 620                 }
 621                 identifierInfo->setIdentifier(id, *status);
 622             }
 623
 624             int32_t scriptCount = identifierInfo->getScriptCount();
 625
 626             ScriptSet scripts;
 627             This->wholeScriptCheck(nfdText, &scripts, *status);
 628             int32_t confusableScriptCount = scripts.countMembers();
 629             //printf("confusableScriptCount = %d\n", confusableScriptCount);
 630
 631             if ((This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE) &&
 632                 confusableScriptCount >= 2 &&
 633                 scriptCount == 1) {
 634                 result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
 635             }
 636
 637             if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) &&
 638                 confusableScriptCount >= 1 &&
 639                 scriptCount > 1) {
 640                 result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
 641             }
 642         }
 643     }
 644
 645 cleanupAndReturn:
 646     This->releaseIdentifierInfo(identifierInfo);
 647     if (position != NULL) {
 648         *position = 0;
 649     }
 650     return result;
 651 }
 652
 653
 654 U_CAPI int32_t U_EXPORT2
 655 uspoof_getSkeleton(const USpoofChecker *sc,
 656                    uint32_t type,
 657                    const UChar *id,  int32_t length,
 658                    UChar *dest, int32_t destCapacity,
 659                    UErrorCode *status) {
 660
 661     SpoofImpl::validateThis(sc, *status);
 662     if (U_FAILURE(*status)) {
 663         return 0;
 664     }
 665     if (length<-1 || destCapacity<0 || (destCapacity==0 && dest!=NULL)) {
 666         *status = U_ILLEGAL_ARGUMENT_ERROR;
 667         return 0;
 668     }
 669
 670     UnicodeString idStr((length==-1), id, length);  // Aliasing constructor
 671     UnicodeString destStr;
 672     uspoof_getSkeletonUnicodeString(sc, type, idStr, destStr, status);
 673     destStr.extract(dest, destCapacity, *status);
 674     return destStr.length();
 675 }
 676
 677
 678
 679 U_I18N_API UnicodeString &  U_EXPORT2
 680 uspoof_getSkeletonUnicodeString(const USpoofChecker *sc,
 681                                 uint32_t type,
 682                                 const UnicodeString &id,
 683                                 UnicodeString &dest,
 684                                 UErrorCode *status) {
 685     const SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 686     if (U_FAILURE(*status)) {
 687         return dest;
 688     }
 689
 690    int32_t tableMask = 0;
 691    switch (type) {
 692       case 0:
 693         tableMask = USPOOF_ML_TABLE_FLAG;
 694         break;
 695       case USPOOF_SINGLE_SCRIPT_CONFUSABLE:
 696         tableMask = USPOOF_SL_TABLE_FLAG;
 697         break;
 698       case USPOOF_ANY_CASE:
 699         tableMask = USPOOF_MA_TABLE_FLAG;
 700         break;
 701       case USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_ANY_CASE:
 702         tableMask = USPOOF_SA_TABLE_FLAG;
 703         break;
 704       default:
 705         *status = U_ILLEGAL_ARGUMENT_ERROR;
 706         return dest;
 707     }
 708
 709     UnicodeString nfdId;
 710     gNfdNormalizer->normalize(id, nfdId, *status);
 711
 712     // Apply the skeleton mapping to the NFD normalized input string
 713     // Accumulate the skeleton, possibly unnormalized, in a UnicodeString.
 714     int32_t inputIndex = 0;
 715     UnicodeString skelStr;
 716     int32_t normalizedLen = nfdId.length();
 717     for (inputIndex=0; inputIndex < normalizedLen; ) {
 718         UChar32 c = nfdId.char32At(inputIndex);
 719         inputIndex += U16_LENGTH(c);
 720         This->confusableLookup(c, tableMask, skelStr);
 721     }
 722
 723     gNfdNormalizer->normalize(skelStr, dest, *status);
 724     return dest;
 725 }
 726
 727
 728 U_CAPI int32_t U_EXPORT2
 729 uspoof_getSkeletonUTF8(const USpoofChecker *sc,
 730                        uint32_t type,
 731                        const char *id,  int32_t length,
 732                        char *dest, int32_t destCapacity,
 733                        UErrorCode *status) {
 734     SpoofImpl::validateThis(sc, *status);
 735     if (U_FAILURE(*status)) {
 736         return 0;
 737     }
 738     if (length<-1 || destCapacity<0 || (destCapacity==0 && dest!=NULL)) {
 739         *status = U_ILLEGAL_ARGUMENT_ERROR;
 740         return 0;
 741     }
 742
 743     UnicodeString srcStr = UnicodeString::fromUTF8(StringPiece(id, length>=0 ? length : uprv_strlen(id)));
 744     UnicodeString destStr;
 745     uspoof_getSkeletonUnicodeString(sc, type, srcStr, destStr, status);
 746     if (U_FAILURE(*status)) {
 747         return 0;
 748     }
 749
 750     int32_t lengthInUTF8 = 0;
 751     u_strToUTF8(dest, destCapacity, &lengthInUTF8,
 752                 destStr.getBuffer(), destStr.length(), status);
 753     return lengthInUTF8;
 754 }
 755
 756
 757 U_CAPI int32_t U_EXPORT2
 758 uspoof_serialize(USpoofChecker *sc,void *buf, int32_t capacity, UErrorCode *status) {
 759     SpoofImpl *This = SpoofImpl::validateThis(sc, *status);
 760     if (This == NULL) {
 761         U_ASSERT(U_FAILURE(*status));
 762         return 0;
 763     }
 764     int32_t dataSize = This->fSpoofData->fRawData->fLength;
 765     if (capacity < dataSize) {
 766         *status = U_BUFFER_OVERFLOW_ERROR;
 767         return dataSize;
 768     }
 769     uprv_memcpy(buf, This->fSpoofData->fRawData, dataSize);
 770     return dataSize;
 771 }
 772
 773 U_CAPI const USet * U_EXPORT2
 774 uspoof_getInclusionSet(UErrorCode *status) {
 775     umtx_initOnce(gSpoofInitStaticsOnce, &initializeStatics, *status);
 776     return gInclusionSet->toUSet();
 777 }
 778
 779 U_CAPI const USet * U_EXPORT2
 780 uspoof_getRecommendedSet(UErrorCode *status) {
 781     umtx_initOnce(gSpoofInitStaticsOnce, &initializeStatics, *status);
 782     return gRecommendedSet->toUSet();
 783 }
 784
 785 U_I18N_API const UnicodeSet * U_EXPORT2
 786 uspoof_getInclusionUnicodeSet(UErrorCode *status) {
 787     umtx_initOnce(gSpoofInitStaticsOnce, &initializeStatics, *status);
 788     return gInclusionSet;
 789 }
 790
 791 U_I18N_API const UnicodeSet * U_EXPORT2
 792 uspoof_getRecommendedUnicodeSet(UErrorCode *status) {
 793     umtx_initOnce(gSpoofInitStaticsOnce, &initializeStatics, *status);
 794     return gRecommendedSet;
 795 }
 796
 797
 798
 799 #endif // !UCONFIG_NO_NORMALIZATION