1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/intl/icu/source/common/caniter.cpp Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,577 @@
1.4 +/*
1.5 + *****************************************************************************
1.6 + * Copyright (C) 1996-2011, International Business Machines Corporation and *
1.7 + * others. All Rights Reserved. *
1.8 + *****************************************************************************
1.9 + */
1.10 +
1.11 +#include "unicode/utypes.h"
1.12 +
1.13 +#if !UCONFIG_NO_NORMALIZATION
1.14 +
1.15 +#include "unicode/caniter.h"
1.16 +#include "unicode/normalizer2.h"
1.17 +#include "unicode/uchar.h"
1.18 +#include "unicode/uniset.h"
1.19 +#include "unicode/usetiter.h"
1.20 +#include "unicode/ustring.h"
1.21 +#include "unicode/utf16.h"
1.22 +#include "cmemory.h"
1.23 +#include "hash.h"
1.24 +#include "normalizer2impl.h"
1.25 +
1.26 +/**
1.27 + * This class allows one to iterate through all the strings that are canonically equivalent to a given
1.28 + * string. For example, here are some sample results:
1.29 +Results for: {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
1.30 +1: \u0041\u030A\u0064\u0307\u0327
1.31 + = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
1.32 +2: \u0041\u030A\u0064\u0327\u0307
1.33 + = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
1.34 +3: \u0041\u030A\u1E0B\u0327
1.35 + = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
1.36 +4: \u0041\u030A\u1E11\u0307
1.37 + = {LATIN CAPITAL LETTER A}{COMBINING RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
1.38 +5: \u00C5\u0064\u0307\u0327
1.39 + = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
1.40 +6: \u00C5\u0064\u0327\u0307
1.41 + = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
1.42 +7: \u00C5\u1E0B\u0327
1.43 + = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
1.44 +8: \u00C5\u1E11\u0307
1.45 + = {LATIN CAPITAL LETTER A WITH RING ABOVE}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
1.46 +9: \u212B\u0064\u0307\u0327
1.47 + = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING DOT ABOVE}{COMBINING CEDILLA}
1.48 +10: \u212B\u0064\u0327\u0307
1.49 + = {ANGSTROM SIGN}{LATIN SMALL LETTER D}{COMBINING CEDILLA}{COMBINING DOT ABOVE}
1.50 +11: \u212B\u1E0B\u0327
1.51 + = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH DOT ABOVE}{COMBINING CEDILLA}
1.52 +12: \u212B\u1E11\u0307
1.53 + = {ANGSTROM SIGN}{LATIN SMALL LETTER D WITH CEDILLA}{COMBINING DOT ABOVE}
1.54 + *<br>Note: the code is intended for use with small strings, and is not suitable for larger ones,
1.55 + * since it has not been optimized for that situation.
1.56 + *@author M. Davis
1.57 + *@draft
1.58 + */
1.59 +
1.60 +// public
1.61 +
1.62 +U_NAMESPACE_BEGIN
1.63 +
1.64 +// TODO: add boilerplate methods.
1.65 +
1.66 +UOBJECT_DEFINE_RTTI_IMPLEMENTATION(CanonicalIterator)
1.67 +
1.68 +/**
1.69 + *@param source string to get results for
1.70 + */
1.71 +CanonicalIterator::CanonicalIterator(const UnicodeString &sourceStr, UErrorCode &status) :
1.72 + pieces(NULL),
1.73 + pieces_length(0),
1.74 + pieces_lengths(NULL),
1.75 + current(NULL),
1.76 + current_length(0),
1.77 + nfd(*Normalizer2Factory::getNFDInstance(status)),
1.78 + nfcImpl(*Normalizer2Factory::getNFCImpl(status))
1.79 +{
1.80 + if(U_SUCCESS(status) && nfcImpl.ensureCanonIterData(status)) {
1.81 + setSource(sourceStr, status);
1.82 + }
1.83 +}
1.84 +
1.85 +CanonicalIterator::~CanonicalIterator() {
1.86 + cleanPieces();
1.87 +}
1.88 +
1.89 +void CanonicalIterator::cleanPieces() {
1.90 + int32_t i = 0;
1.91 + if(pieces != NULL) {
1.92 + for(i = 0; i < pieces_length; i++) {
1.93 + if(pieces[i] != NULL) {
1.94 + delete[] pieces[i];
1.95 + }
1.96 + }
1.97 + uprv_free(pieces);
1.98 + pieces = NULL;
1.99 + pieces_length = 0;
1.100 + }
1.101 + if(pieces_lengths != NULL) {
1.102 + uprv_free(pieces_lengths);
1.103 + pieces_lengths = NULL;
1.104 + }
1.105 + if(current != NULL) {
1.106 + uprv_free(current);
1.107 + current = NULL;
1.108 + current_length = 0;
1.109 + }
1.110 +}
1.111 +
1.112 +/**
1.113 + *@return gets the source: NOTE: it is the NFD form of source
1.114 + */
1.115 +UnicodeString CanonicalIterator::getSource() {
1.116 + return source;
1.117 +}
1.118 +
1.119 +/**
1.120 + * Resets the iterator so that one can start again from the beginning.
1.121 + */
1.122 +void CanonicalIterator::reset() {
1.123 + done = FALSE;
1.124 + for (int i = 0; i < current_length; ++i) {
1.125 + current[i] = 0;
1.126 + }
1.127 +}
1.128 +
1.129 +/**
1.130 + *@return the next string that is canonically equivalent. The value null is returned when
1.131 + * the iteration is done.
1.132 + */
1.133 +UnicodeString CanonicalIterator::next() {
1.134 + int32_t i = 0;
1.135 +
1.136 + if (done) {
1.137 + buffer.setToBogus();
1.138 + return buffer;
1.139 + }
1.140 +
1.141 + // delete old contents
1.142 + buffer.remove();
1.143 +
1.144 + // construct return value
1.145 +
1.146 + for (i = 0; i < pieces_length; ++i) {
1.147 + buffer.append(pieces[i][current[i]]);
1.148 + }
1.149 + //String result = buffer.toString(); // not needed
1.150 +
1.151 + // find next value for next time
1.152 +
1.153 + for (i = current_length - 1; ; --i) {
1.154 + if (i < 0) {
1.155 + done = TRUE;
1.156 + break;
1.157 + }
1.158 + current[i]++;
1.159 + if (current[i] < pieces_lengths[i]) break; // got sequence
1.160 + current[i] = 0;
1.161 + }
1.162 + return buffer;
1.163 +}
1.164 +
1.165 +/**
1.166 + *@param set the source string to iterate against. This allows the same iterator to be used
1.167 + * while changing the source string, saving object creation.
1.168 + */
1.169 +void CanonicalIterator::setSource(const UnicodeString &newSource, UErrorCode &status) {
1.170 + int32_t list_length = 0;
1.171 + UChar32 cp = 0;
1.172 + int32_t start = 0;
1.173 + int32_t i = 0;
1.174 + UnicodeString *list = NULL;
1.175 +
1.176 + nfd.normalize(newSource, source, status);
1.177 + if(U_FAILURE(status)) {
1.178 + return;
1.179 + }
1.180 + done = FALSE;
1.181 +
1.182 + cleanPieces();
1.183 +
1.184 + // catch degenerate case
1.185 + if (newSource.length() == 0) {
1.186 + pieces = (UnicodeString **)uprv_malloc(sizeof(UnicodeString *));
1.187 + pieces_lengths = (int32_t*)uprv_malloc(1 * sizeof(int32_t));
1.188 + pieces_length = 1;
1.189 + current = (int32_t*)uprv_malloc(1 * sizeof(int32_t));
1.190 + current_length = 1;
1.191 + if (pieces == NULL || pieces_lengths == NULL || current == NULL) {
1.192 + status = U_MEMORY_ALLOCATION_ERROR;
1.193 + goto CleanPartialInitialization;
1.194 + }
1.195 + current[0] = 0;
1.196 + pieces[0] = new UnicodeString[1];
1.197 + pieces_lengths[0] = 1;
1.198 + if (pieces[0] == 0) {
1.199 + status = U_MEMORY_ALLOCATION_ERROR;
1.200 + goto CleanPartialInitialization;
1.201 + }
1.202 + return;
1.203 + }
1.204 +
1.205 +
1.206 + list = new UnicodeString[source.length()];
1.207 + if (list == 0) {
1.208 + status = U_MEMORY_ALLOCATION_ERROR;
1.209 + goto CleanPartialInitialization;
1.210 + }
1.211 +
1.212 + // i should initialy be the number of code units at the
1.213 + // start of the string
1.214 + i = U16_LENGTH(source.char32At(0));
1.215 + //int32_t i = 1;
1.216 + // find the segments
1.217 + // This code iterates through the source string and
1.218 + // extracts segments that end up on a codepoint that
1.219 + // doesn't start any decompositions. (Analysis is done
1.220 + // on the NFD form - see above).
1.221 + for (; i < source.length(); i += U16_LENGTH(cp)) {
1.222 + cp = source.char32At(i);
1.223 + if (nfcImpl.isCanonSegmentStarter(cp)) {
1.224 + source.extract(start, i-start, list[list_length++]); // add up to i
1.225 + start = i;
1.226 + }
1.227 + }
1.228 + source.extract(start, i-start, list[list_length++]); // add last one
1.229 +
1.230 +
1.231 + // allocate the arrays, and find the strings that are CE to each segment
1.232 + pieces = (UnicodeString **)uprv_malloc(list_length * sizeof(UnicodeString *));
1.233 + pieces_length = list_length;
1.234 + pieces_lengths = (int32_t*)uprv_malloc(list_length * sizeof(int32_t));
1.235 + current = (int32_t*)uprv_malloc(list_length * sizeof(int32_t));
1.236 + current_length = list_length;
1.237 + if (pieces == NULL || pieces_lengths == NULL || current == NULL) {
1.238 + status = U_MEMORY_ALLOCATION_ERROR;
1.239 + goto CleanPartialInitialization;
1.240 + }
1.241 +
1.242 + for (i = 0; i < current_length; i++) {
1.243 + current[i] = 0;
1.244 + }
1.245 + // for each segment, get all the combinations that can produce
1.246 + // it after NFD normalization
1.247 + for (i = 0; i < pieces_length; ++i) {
1.248 + //if (PROGRESS) printf("SEGMENT\n");
1.249 + pieces[i] = getEquivalents(list[i], pieces_lengths[i], status);
1.250 + }
1.251 +
1.252 + delete[] list;
1.253 + return;
1.254 +// Common section to cleanup all local variables and reset object variables.
1.255 +CleanPartialInitialization:
1.256 + if (list != NULL) {
1.257 + delete[] list;
1.258 + }
1.259 + cleanPieces();
1.260 +}
1.261 +
1.262 +/**
1.263 + * Dumb recursive implementation of permutation.
1.264 + * TODO: optimize
1.265 + * @param source the string to find permutations for
1.266 + * @return the results in a set.
1.267 + */
1.268 +void U_EXPORT2 CanonicalIterator::permute(UnicodeString &source, UBool skipZeros, Hashtable *result, UErrorCode &status) {
1.269 + if(U_FAILURE(status)) {
1.270 + return;
1.271 + }
1.272 + //if (PROGRESS) printf("Permute: %s\n", UToS(Tr(source)));
1.273 + int32_t i = 0;
1.274 +
1.275 + // optimization:
1.276 + // if zero or one character, just return a set with it
1.277 + // we check for length < 2 to keep from counting code points all the time
1.278 + if (source.length() <= 2 && source.countChar32() <= 1) {
1.279 + UnicodeString *toPut = new UnicodeString(source);
1.280 + /* test for NULL */
1.281 + if (toPut == 0) {
1.282 + status = U_MEMORY_ALLOCATION_ERROR;
1.283 + return;
1.284 + }
1.285 + result->put(source, toPut, status);
1.286 + return;
1.287 + }
1.288 +
1.289 + // otherwise iterate through the string, and recursively permute all the other characters
1.290 + UChar32 cp;
1.291 + Hashtable subpermute(status);
1.292 + if(U_FAILURE(status)) {
1.293 + return;
1.294 + }
1.295 + subpermute.setValueDeleter(uprv_deleteUObject);
1.296 +
1.297 + for (i = 0; i < source.length(); i += U16_LENGTH(cp)) {
1.298 + cp = source.char32At(i);
1.299 + const UHashElement *ne = NULL;
1.300 + int32_t el = -1;
1.301 + UnicodeString subPermuteString = source;
1.302 +
1.303 + // optimization:
1.304 + // if the character is canonical combining class zero,
1.305 + // don't permute it
1.306 + if (skipZeros && i != 0 && u_getCombiningClass(cp) == 0) {
1.307 + //System.out.println("Skipping " + Utility.hex(UTF16.valueOf(source, i)));
1.308 + continue;
1.309 + }
1.310 +
1.311 + subpermute.removeAll();
1.312 +
1.313 + // see what the permutations of the characters before and after this one are
1.314 + //Hashtable *subpermute = permute(source.substring(0,i) + source.substring(i + UTF16.getCharCount(cp)));
1.315 + permute(subPermuteString.replace(i, U16_LENGTH(cp), NULL, 0), skipZeros, &subpermute, status);
1.316 + /* Test for buffer overflows */
1.317 + if(U_FAILURE(status)) {
1.318 + return;
1.319 + }
1.320 + // The upper replace is destructive. The question is do we have to make a copy, or we don't care about the contents
1.321 + // of source at this point.
1.322 +
1.323 + // prefix this character to all of them
1.324 + ne = subpermute.nextElement(el);
1.325 + while (ne != NULL) {
1.326 + UnicodeString *permRes = (UnicodeString *)(ne->value.pointer);
1.327 + UnicodeString *chStr = new UnicodeString(cp);
1.328 + //test for NULL
1.329 + if (chStr == NULL) {
1.330 + status = U_MEMORY_ALLOCATION_ERROR;
1.331 + return;
1.332 + }
1.333 + chStr->append(*permRes); //*((UnicodeString *)(ne->value.pointer));
1.334 + //if (PROGRESS) printf(" Piece: %s\n", UToS(*chStr));
1.335 + result->put(*chStr, chStr, status);
1.336 + ne = subpermute.nextElement(el);
1.337 + }
1.338 + }
1.339 + //return result;
1.340 +}
1.341 +
1.342 +// privates
1.343 +
1.344 +// we have a segment, in NFD. Find all the strings that are canonically equivalent to it.
1.345 +UnicodeString* CanonicalIterator::getEquivalents(const UnicodeString &segment, int32_t &result_len, UErrorCode &status) {
1.346 + Hashtable result(status);
1.347 + Hashtable permutations(status);
1.348 + Hashtable basic(status);
1.349 + if (U_FAILURE(status)) {
1.350 + return 0;
1.351 + }
1.352 + result.setValueDeleter(uprv_deleteUObject);
1.353 + permutations.setValueDeleter(uprv_deleteUObject);
1.354 + basic.setValueDeleter(uprv_deleteUObject);
1.355 +
1.356 + UChar USeg[256];
1.357 + int32_t segLen = segment.extract(USeg, 256, status);
1.358 + getEquivalents2(&basic, USeg, segLen, status);
1.359 +
1.360 + // now get all the permutations
1.361 + // add only the ones that are canonically equivalent
1.362 + // TODO: optimize by not permuting any class zero.
1.363 +
1.364 + const UHashElement *ne = NULL;
1.365 + int32_t el = -1;
1.366 + //Iterator it = basic.iterator();
1.367 + ne = basic.nextElement(el);
1.368 + //while (it.hasNext())
1.369 + while (ne != NULL) {
1.370 + //String item = (String) it.next();
1.371 + UnicodeString item = *((UnicodeString *)(ne->value.pointer));
1.372 +
1.373 + permutations.removeAll();
1.374 + permute(item, CANITER_SKIP_ZEROES, &permutations, status);
1.375 + const UHashElement *ne2 = NULL;
1.376 + int32_t el2 = -1;
1.377 + //Iterator it2 = permutations.iterator();
1.378 + ne2 = permutations.nextElement(el2);
1.379 + //while (it2.hasNext())
1.380 + while (ne2 != NULL) {
1.381 + //String possible = (String) it2.next();
1.382 + //UnicodeString *possible = new UnicodeString(*((UnicodeString *)(ne2->value.pointer)));
1.383 + UnicodeString possible(*((UnicodeString *)(ne2->value.pointer)));
1.384 + UnicodeString attempt;
1.385 + nfd.normalize(possible, attempt, status);
1.386 +
1.387 + // TODO: check if operator == is semanticaly the same as attempt.equals(segment)
1.388 + if (attempt==segment) {
1.389 + //if (PROGRESS) printf("Adding Permutation: %s\n", UToS(Tr(*possible)));
1.390 + // TODO: use the hashtable just to catch duplicates - store strings directly (somehow).
1.391 + result.put(possible, new UnicodeString(possible), status); //add(possible);
1.392 + } else {
1.393 + //if (PROGRESS) printf("-Skipping Permutation: %s\n", UToS(Tr(*possible)));
1.394 + }
1.395 +
1.396 + ne2 = permutations.nextElement(el2);
1.397 + }
1.398 + ne = basic.nextElement(el);
1.399 + }
1.400 +
1.401 + /* Test for buffer overflows */
1.402 + if(U_FAILURE(status)) {
1.403 + return 0;
1.404 + }
1.405 + // convert into a String[] to clean up storage
1.406 + //String[] finalResult = new String[result.size()];
1.407 + UnicodeString *finalResult = NULL;
1.408 + int32_t resultCount;
1.409 + if((resultCount = result.count())) {
1.410 + finalResult = new UnicodeString[resultCount];
1.411 + if (finalResult == 0) {
1.412 + status = U_MEMORY_ALLOCATION_ERROR;
1.413 + return NULL;
1.414 + }
1.415 + }
1.416 + else {
1.417 + status = U_ILLEGAL_ARGUMENT_ERROR;
1.418 + return NULL;
1.419 + }
1.420 + //result.toArray(finalResult);
1.421 + result_len = 0;
1.422 + el = -1;
1.423 + ne = result.nextElement(el);
1.424 + while(ne != NULL) {
1.425 + finalResult[result_len++] = *((UnicodeString *)(ne->value.pointer));
1.426 + ne = result.nextElement(el);
1.427 + }
1.428 +
1.429 +
1.430 + return finalResult;
1.431 +}
1.432 +
1.433 +Hashtable *CanonicalIterator::getEquivalents2(Hashtable *fillinResult, const UChar *segment, int32_t segLen, UErrorCode &status) {
1.434 +
1.435 + if (U_FAILURE(status)) {
1.436 + return NULL;
1.437 + }
1.438 +
1.439 + //if (PROGRESS) printf("Adding: %s\n", UToS(Tr(segment)));
1.440 +
1.441 + UnicodeString toPut(segment, segLen);
1.442 +
1.443 + fillinResult->put(toPut, new UnicodeString(toPut), status);
1.444 +
1.445 + UnicodeSet starts;
1.446 +
1.447 + // cycle through all the characters
1.448 + UChar32 cp;
1.449 + for (int32_t i = 0; i < segLen; i += U16_LENGTH(cp)) {
1.450 + // see if any character is at the start of some decomposition
1.451 + U16_GET(segment, 0, i, segLen, cp);
1.452 + if (!nfcImpl.getCanonStartSet(cp, starts)) {
1.453 + continue;
1.454 + }
1.455 + // if so, see which decompositions match
1.456 + UnicodeSetIterator iter(starts);
1.457 + while (iter.next()) {
1.458 + UChar32 cp2 = iter.getCodepoint();
1.459 + Hashtable remainder(status);
1.460 + remainder.setValueDeleter(uprv_deleteUObject);
1.461 + if (extract(&remainder, cp2, segment, segLen, i, status) == NULL) {
1.462 + continue;
1.463 + }
1.464 +
1.465 + // there were some matches, so add all the possibilities to the set.
1.466 + UnicodeString prefix(segment, i);
1.467 + prefix += cp2;
1.468 +
1.469 + int32_t el = -1;
1.470 + const UHashElement *ne = remainder.nextElement(el);
1.471 + while (ne != NULL) {
1.472 + UnicodeString item = *((UnicodeString *)(ne->value.pointer));
1.473 + UnicodeString *toAdd = new UnicodeString(prefix);
1.474 + /* test for NULL */
1.475 + if (toAdd == 0) {
1.476 + status = U_MEMORY_ALLOCATION_ERROR;
1.477 + return NULL;
1.478 + }
1.479 + *toAdd += item;
1.480 + fillinResult->put(*toAdd, toAdd, status);
1.481 +
1.482 + //if (PROGRESS) printf("Adding: %s\n", UToS(Tr(*toAdd)));
1.483 +
1.484 + ne = remainder.nextElement(el);
1.485 + }
1.486 + }
1.487 + }
1.488 +
1.489 + /* Test for buffer overflows */
1.490 + if(U_FAILURE(status)) {
1.491 + return NULL;
1.492 + }
1.493 + return fillinResult;
1.494 +}
1.495 +
1.496 +/**
1.497 + * See if the decomposition of cp2 is at segment starting at segmentPos
1.498 + * (with canonical rearrangment!)
1.499 + * If so, take the remainder, and return the equivalents
1.500 + */
1.501 +Hashtable *CanonicalIterator::extract(Hashtable *fillinResult, UChar32 comp, const UChar *segment, int32_t segLen, int32_t segmentPos, UErrorCode &status) {
1.502 +//Hashtable *CanonicalIterator::extract(UChar32 comp, const UnicodeString &segment, int32_t segLen, int32_t segmentPos, UErrorCode &status) {
1.503 + //if (PROGRESS) printf(" extract: %s, ", UToS(Tr(UnicodeString(comp))));
1.504 + //if (PROGRESS) printf("%s, %i\n", UToS(Tr(segment)), segmentPos);
1.505 +
1.506 + if (U_FAILURE(status)) {
1.507 + return NULL;
1.508 + }
1.509 +
1.510 + UnicodeString temp(comp);
1.511 + int32_t inputLen=temp.length();
1.512 + UnicodeString decompString;
1.513 + nfd.normalize(temp, decompString, status);
1.514 + const UChar *decomp=decompString.getBuffer();
1.515 + int32_t decompLen=decompString.length();
1.516 +
1.517 + // See if it matches the start of segment (at segmentPos)
1.518 + UBool ok = FALSE;
1.519 + UChar32 cp;
1.520 + int32_t decompPos = 0;
1.521 + UChar32 decompCp;
1.522 + U16_NEXT(decomp, decompPos, decompLen, decompCp);
1.523 +
1.524 + int32_t i = segmentPos;
1.525 + while(i < segLen) {
1.526 + U16_NEXT(segment, i, segLen, cp);
1.527 +
1.528 + if (cp == decompCp) { // if equal, eat another cp from decomp
1.529 +
1.530 + //if (PROGRESS) printf(" matches: %s\n", UToS(Tr(UnicodeString(cp))));
1.531 +
1.532 + if (decompPos == decompLen) { // done, have all decomp characters!
1.533 + temp.append(segment+i, segLen-i);
1.534 + ok = TRUE;
1.535 + break;
1.536 + }
1.537 + U16_NEXT(decomp, decompPos, decompLen, decompCp);
1.538 + } else {
1.539 + //if (PROGRESS) printf(" buffer: %s\n", UToS(Tr(UnicodeString(cp))));
1.540 +
1.541 + // brute force approach
1.542 + temp.append(cp);
1.543 +
1.544 + /* TODO: optimize
1.545 + // since we know that the classes are monotonically increasing, after zero
1.546 + // e.g. 0 5 7 9 0 3
1.547 + // we can do an optimization
1.548 + // there are only a few cases that work: zero, less, same, greater
1.549 + // if both classes are the same, we fail
1.550 + // if the decomp class < the segment class, we fail
1.551 +
1.552 + segClass = getClass(cp);
1.553 + if (decompClass <= segClass) return null;
1.554 + */
1.555 + }
1.556 + }
1.557 + if (!ok)
1.558 + return NULL; // we failed, characters left over
1.559 +
1.560 + //if (PROGRESS) printf("Matches\n");
1.561 +
1.562 + if (inputLen == temp.length()) {
1.563 + fillinResult->put(UnicodeString(), new UnicodeString(), status);
1.564 + return fillinResult; // succeed, but no remainder
1.565 + }
1.566 +
1.567 + // brute force approach
1.568 + // check to make sure result is canonically equivalent
1.569 + UnicodeString trial;
1.570 + nfd.normalize(temp, trial, status);
1.571 + if(U_FAILURE(status) || trial.compare(segment+segmentPos, segLen - segmentPos) != 0) {
1.572 + return NULL;
1.573 + }
1.574 +
1.575 + return getEquivalents2(fillinResult, temp.getBuffer()+inputLen, temp.length()-inputLen, status);
1.576 +}
1.577 +
1.578 +U_NAMESPACE_END
1.579 +
1.580 +#endif /* #if !UCONFIG_NO_NORMALIZATION */
