Coverage Report

Created: 2025-06-24 06:43

/src/icu/source/i18n/collationbuilder.cpp
Line
Count
Source (jump to first uncovered line)
1
// © 2016 and later: Unicode, Inc. and others.
2
// License & terms of use: http://www.unicode.org/copyright.html
3
/*
4
*******************************************************************************
5
* Copyright (C) 2013-2014, International Business Machines
6
* Corporation and others.  All Rights Reserved.
7
*******************************************************************************
8
* collationbuilder.cpp
9
*
10
* (replaced the former ucol_bld.cpp)
11
*
12
* created on: 2013may06
13
* created by: Markus W. Scherer
14
*/
15
16
#ifdef DEBUG_COLLATION_BUILDER
17
#include <stdio.h>
18
#endif
19
20
#include "unicode/utypes.h"
21
22
#if !UCONFIG_NO_COLLATION
23
24
#include "unicode/caniter.h"
25
#include "unicode/normalizer2.h"
26
#include "unicode/tblcoll.h"
27
#include "unicode/parseerr.h"
28
#include "unicode/uchar.h"
29
#include "unicode/ucol.h"
30
#include "unicode/unistr.h"
31
#include "unicode/usetiter.h"
32
#include "unicode/utf16.h"
33
#include "unicode/uversion.h"
34
#include "cmemory.h"
35
#include "collation.h"
36
#include "collationbuilder.h"
37
#include "collationdata.h"
38
#include "collationdatabuilder.h"
39
#include "collationfastlatin.h"
40
#include "collationroot.h"
41
#include "collationrootelements.h"
42
#include "collationruleparser.h"
43
#include "collationsettings.h"
44
#include "collationtailoring.h"
45
#include "collationweights.h"
46
#include "normalizer2impl.h"
47
#include "uassert.h"
48
#include "ucol_imp.h"
49
#include "utf16collationiterator.h"
50
51
U_NAMESPACE_BEGIN
52
53
namespace {
54
55
class BundleImporter : public CollationRuleParser::Importer {
56
public:
57
0
    BundleImporter() {}
58
    virtual ~BundleImporter();
59
    virtual void getRules(
60
            const char *localeID, const char *collationType,
61
            UnicodeString &rules,
62
            const char *&errorReason, UErrorCode &errorCode);
63
};
64
65
BundleImporter::~BundleImporter() {}
66
67
void
68
BundleImporter::getRules(
69
        const char *localeID, const char *collationType,
70
        UnicodeString &rules,
71
0
        const char *& /*errorReason*/, UErrorCode &errorCode) {
72
0
    CollationLoader::loadRules(localeID, collationType, rules, errorCode);
73
0
}
74
75
}  // namespace
76
77
// RuleBasedCollator implementation ---------------------------------------- ***
78
79
// These methods are here, rather than in rulebasedcollator.cpp,
80
// for modularization:
81
// Most code using Collator does not need to build a Collator from rules.
82
// By moving these constructors and helper methods to a separate file,
83
// most code will not have a static dependency on the builder code.
84
85
RuleBasedCollator::RuleBasedCollator()
86
        : data(NULL),
87
          settings(NULL),
88
          tailoring(NULL),
89
          cacheEntry(NULL),
90
0
          validLocale(""),
91
0
          explicitlySetAttributes(0),
92
0
          actualLocaleIsSameAsValid(FALSE) {
93
0
}
94
95
RuleBasedCollator::RuleBasedCollator(const UnicodeString &rules, UErrorCode &errorCode)
96
        : data(NULL),
97
          settings(NULL),
98
          tailoring(NULL),
99
          cacheEntry(NULL),
100
0
          validLocale(""),
101
0
          explicitlySetAttributes(0),
102
0
          actualLocaleIsSameAsValid(FALSE) {
103
0
    internalBuildTailoring(rules, UCOL_DEFAULT, UCOL_DEFAULT, NULL, NULL, errorCode);
104
0
}
105
106
RuleBasedCollator::RuleBasedCollator(const UnicodeString &rules, ECollationStrength strength,
107
                                     UErrorCode &errorCode)
108
        : data(NULL),
109
          settings(NULL),
110
          tailoring(NULL),
111
          cacheEntry(NULL),
112
0
          validLocale(""),
113
0
          explicitlySetAttributes(0),
114
0
          actualLocaleIsSameAsValid(FALSE) {
115
0
    internalBuildTailoring(rules, strength, UCOL_DEFAULT, NULL, NULL, errorCode);
116
0
}
117
118
RuleBasedCollator::RuleBasedCollator(const UnicodeString &rules,
119
                                     UColAttributeValue decompositionMode,
120
                                     UErrorCode &errorCode)
121
        : data(NULL),
122
          settings(NULL),
123
          tailoring(NULL),
124
          cacheEntry(NULL),
125
0
          validLocale(""),
126
0
          explicitlySetAttributes(0),
127
0
          actualLocaleIsSameAsValid(FALSE) {
128
0
    internalBuildTailoring(rules, UCOL_DEFAULT, decompositionMode, NULL, NULL, errorCode);
129
0
}
130
131
RuleBasedCollator::RuleBasedCollator(const UnicodeString &rules,
132
                                     ECollationStrength strength,
133
                                     UColAttributeValue decompositionMode,
134
                                     UErrorCode &errorCode)
135
        : data(NULL),
136
          settings(NULL),
137
          tailoring(NULL),
138
          cacheEntry(NULL),
139
0
          validLocale(""),
140
0
          explicitlySetAttributes(0),
141
0
          actualLocaleIsSameAsValid(FALSE) {
142
0
    internalBuildTailoring(rules, strength, decompositionMode, NULL, NULL, errorCode);
143
0
}
144
145
RuleBasedCollator::RuleBasedCollator(const UnicodeString &rules,
146
                                     UParseError &parseError, UnicodeString &reason,
147
                                     UErrorCode &errorCode)
148
        : data(NULL),
149
          settings(NULL),
150
          tailoring(NULL),
151
          cacheEntry(NULL),
152
0
          validLocale(""),
153
0
          explicitlySetAttributes(0),
154
0
          actualLocaleIsSameAsValid(FALSE) {
155
0
    internalBuildTailoring(rules, UCOL_DEFAULT, UCOL_DEFAULT, &parseError, &reason, errorCode);
156
0
}
157
158
void
159
RuleBasedCollator::internalBuildTailoring(const UnicodeString &rules,
160
                                          int32_t strength,
161
                                          UColAttributeValue decompositionMode,
162
                                          UParseError *outParseError, UnicodeString *outReason,
163
0
                                          UErrorCode &errorCode) {
164
0
    const CollationTailoring *base = CollationRoot::getRoot(errorCode);
165
0
    if(U_FAILURE(errorCode)) { return; }
166
0
    if(outReason != NULL) { outReason->remove(); }
167
0
    CollationBuilder builder(base, errorCode);
168
0
    UVersionInfo noVersion = { 0, 0, 0, 0 };
169
0
    BundleImporter importer;
170
0
    LocalPointer<CollationTailoring> t(builder.parseAndBuild(rules, noVersion,
171
0
                                                             &importer,
172
0
                                                             outParseError, errorCode));
173
0
    if(U_FAILURE(errorCode)) {
174
0
        const char *reason = builder.getErrorReason();
175
0
        if(reason != NULL && outReason != NULL) {
176
0
            *outReason = UnicodeString(reason, -1, US_INV);
177
0
        }
178
0
        return;
179
0
    }
180
0
    t->actualLocale.setToBogus();
181
0
    adoptTailoring(t.orphan(), errorCode);
182
    // Set attributes after building the collator,
183
    // to keep the default settings consistent with the rule string.
184
0
    if(strength != UCOL_DEFAULT) {
185
0
        setAttribute(UCOL_STRENGTH, (UColAttributeValue)strength, errorCode);
186
0
    }
187
0
    if(decompositionMode != UCOL_DEFAULT) {
188
0
        setAttribute(UCOL_NORMALIZATION_MODE, decompositionMode, errorCode);
189
0
    }
190
0
}
191
192
// CollationBuilder implementation ----------------------------------------- ***
193
194
// Some compilers don't care if constants are defined in the .cpp file.
195
// MS Visual C++ does not like it, but gcc requires it. clang does not care.
196
#ifndef _MSC_VER
197
const int32_t CollationBuilder::HAS_BEFORE2;
198
const int32_t CollationBuilder::HAS_BEFORE3;
199
#endif
200
201
CollationBuilder::CollationBuilder(const CollationTailoring *b, UErrorCode &errorCode)
202
0
        : nfd(*Normalizer2::getNFDInstance(errorCode)),
203
0
          fcd(*Normalizer2Factory::getFCDInstance(errorCode)),
204
0
          nfcImpl(*Normalizer2Factory::getNFCImpl(errorCode)),
205
0
          base(b),
206
0
          baseData(b->data),
207
0
          rootElements(b->data->rootElements, b->data->rootElementsLength),
208
0
          variableTop(0),
209
0
          dataBuilder(new CollationDataBuilder(errorCode)), fastLatinEnabled(TRUE),
210
          errorReason(NULL),
211
0
          cesLength(0),
212
0
          rootPrimaryIndexes(errorCode), nodes(errorCode) {
213
0
    nfcImpl.ensureCanonIterData(errorCode);
214
0
    if(U_FAILURE(errorCode)) {
215
0
        errorReason = "CollationBuilder fields initialization failed";
216
0
        return;
217
0
    }
218
0
    if(dataBuilder == NULL) {
219
0
        errorCode = U_MEMORY_ALLOCATION_ERROR;
220
0
        return;
221
0
    }
222
0
    dataBuilder->initForTailoring(baseData, errorCode);
223
0
    if(U_FAILURE(errorCode)) {
224
0
        errorReason = "CollationBuilder initialization failed";
225
0
    }
226
0
}
227
228
0
CollationBuilder::~CollationBuilder() {
229
0
    delete dataBuilder;
230
0
}
231
232
CollationTailoring *
233
CollationBuilder::parseAndBuild(const UnicodeString &ruleString,
234
                                const UVersionInfo rulesVersion,
235
                                CollationRuleParser::Importer *importer,
236
                                UParseError *outParseError,
237
0
                                UErrorCode &errorCode) {
238
0
    if(U_FAILURE(errorCode)) { return NULL; }
239
0
    if(baseData->rootElements == NULL) {
240
0
        errorCode = U_MISSING_RESOURCE_ERROR;
241
0
        errorReason = "missing root elements data, tailoring not supported";
242
0
        return NULL;
243
0
    }
244
0
    LocalPointer<CollationTailoring> tailoring(new CollationTailoring(base->settings));
245
0
    if(tailoring.isNull() || tailoring->isBogus()) {
246
0
        errorCode = U_MEMORY_ALLOCATION_ERROR;
247
0
        return NULL;
248
0
    }
249
0
    CollationRuleParser parser(baseData, errorCode);
250
0
    if(U_FAILURE(errorCode)) { return NULL; }
251
    // Note: This always bases &[last variable] and &[first regular]
252
    // on the root collator's maxVariable/variableTop.
253
    // If we wanted this to change after [maxVariable x], then we would keep
254
    // the tailoring.settings pointer here and read its variableTop when we need it.
255
    // See http://unicode.org/cldr/trac/ticket/6070
256
0
    variableTop = base->settings->variableTop;
257
0
    parser.setSink(this);
258
0
    parser.setImporter(importer);
259
0
    CollationSettings &ownedSettings = *SharedObject::copyOnWrite(tailoring->settings);
260
0
    parser.parse(ruleString, ownedSettings, outParseError, errorCode);
261
0
    errorReason = parser.getErrorReason();
262
0
    if(U_FAILURE(errorCode)) { return NULL; }
263
0
    if(dataBuilder->hasMappings()) {
264
0
        makeTailoredCEs(errorCode);
265
0
        closeOverComposites(errorCode);
266
0
        finalizeCEs(errorCode);
267
        // Copy all of ASCII, and Latin-1 letters, into each tailoring.
268
0
        optimizeSet.add(0, 0x7f);
269
0
        optimizeSet.add(0xc0, 0xff);
270
        // Hangul is decomposed on the fly during collation,
271
        // and the tailoring data is always built with HANGUL_TAG specials.
272
0
        optimizeSet.remove(Hangul::HANGUL_BASE, Hangul::HANGUL_END);
273
0
        dataBuilder->optimize(optimizeSet, errorCode);
274
0
        tailoring->ensureOwnedData(errorCode);
275
0
        if(U_FAILURE(errorCode)) { return NULL; }
276
0
        if(fastLatinEnabled) { dataBuilder->enableFastLatin(); }
277
0
        dataBuilder->build(*tailoring->ownedData, errorCode);
278
0
        tailoring->builder = dataBuilder;
279
0
        dataBuilder = NULL;
280
0
    } else {
281
0
        tailoring->data = baseData;
282
0
    }
283
0
    if(U_FAILURE(errorCode)) { return NULL; }
284
0
    ownedSettings.fastLatinOptions = CollationFastLatin::getOptions(
285
0
        tailoring->data, ownedSettings,
286
0
        ownedSettings.fastLatinPrimaries, UPRV_LENGTHOF(ownedSettings.fastLatinPrimaries));
287
0
    tailoring->rules = ruleString;
288
0
    tailoring->rules.getTerminatedBuffer();  // ensure NUL-termination
289
0
    tailoring->setVersion(base->version, rulesVersion);
290
0
    return tailoring.orphan();
291
0
}
292
293
void
294
CollationBuilder::addReset(int32_t strength, const UnicodeString &str,
295
0
                           const char *&parserErrorReason, UErrorCode &errorCode) {
296
0
    if(U_FAILURE(errorCode)) { return; }
297
0
    U_ASSERT(!str.isEmpty());
298
0
    if(str.charAt(0) == CollationRuleParser::POS_LEAD) {
299
0
        ces[0] = getSpecialResetPosition(str, parserErrorReason, errorCode);
300
0
        cesLength = 1;
301
0
        if(U_FAILURE(errorCode)) { return; }
302
0
        U_ASSERT((ces[0] & Collation::CASE_AND_QUATERNARY_MASK) == 0);
303
0
    } else {
304
        // normal reset to a character or string
305
0
        UnicodeString nfdString = nfd.normalize(str, errorCode);
306
0
        if(U_FAILURE(errorCode)) {
307
0
            parserErrorReason = "normalizing the reset position";
308
0
            return;
309
0
        }
310
0
        cesLength = dataBuilder->getCEs(nfdString, ces, 0);
311
0
        if(cesLength > Collation::MAX_EXPANSION_LENGTH) {
312
0
            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
313
0
            parserErrorReason = "reset position maps to too many collation elements (more than 31)";
314
0
            return;
315
0
        }
316
0
    }
317
0
    if(strength == UCOL_IDENTICAL) { return; }  // simple reset-at-position
318
319
    // &[before strength]position
320
0
    U_ASSERT(UCOL_PRIMARY <= strength && strength <= UCOL_TERTIARY);
321
0
    int32_t index = findOrInsertNodeForCEs(strength, parserErrorReason, errorCode);
322
0
    if(U_FAILURE(errorCode)) { return; }
323
324
0
    int64_t node = nodes.elementAti(index);
325
    // If the index is for a "weaker" node,
326
    // then skip backwards over this and further "weaker" nodes.
327
0
    while(strengthFromNode(node) > strength) {
328
0
        index = previousIndexFromNode(node);
329
0
        node = nodes.elementAti(index);
330
0
    }
331
332
    // Find or insert a node whose index we will put into a temporary CE.
333
0
    if(strengthFromNode(node) == strength && isTailoredNode(node)) {
334
        // Reset to just before this same-strength tailored node.
335
0
        index = previousIndexFromNode(node);
336
0
    } else if(strength == UCOL_PRIMARY) {
337
        // root primary node (has no previous index)
338
0
        uint32_t p = weight32FromNode(node);
339
0
        if(p == 0) {
340
0
            errorCode = U_UNSUPPORTED_ERROR;
341
0
            parserErrorReason = "reset primary-before ignorable not possible";
342
0
            return;
343
0
        }
344
0
        if(p <= rootElements.getFirstPrimary()) {
345
            // There is no primary gap between ignorables and the space-first-primary.
346
0
            errorCode = U_UNSUPPORTED_ERROR;
347
0
            parserErrorReason = "reset primary-before first non-ignorable not supported";
348
0
            return;
349
0
        }
350
0
        if(p == Collation::FIRST_TRAILING_PRIMARY) {
351
            // We do not support tailoring to an unassigned-implicit CE.
352
0
            errorCode = U_UNSUPPORTED_ERROR;
353
0
            parserErrorReason = "reset primary-before [first trailing] not supported";
354
0
            return;
355
0
        }
356
0
        p = rootElements.getPrimaryBefore(p, baseData->isCompressiblePrimary(p));
357
0
        index = findOrInsertNodeForPrimary(p, errorCode);
358
        // Go to the last node in this list:
359
        // Tailor after the last node between adjacent root nodes.
360
0
        for(;;) {
361
0
            node = nodes.elementAti(index);
362
0
            int32_t nextIndex = nextIndexFromNode(node);
363
0
            if(nextIndex == 0) { break; }
364
0
            index = nextIndex;
365
0
        }
366
0
    } else {
367
        // &[before 2] or &[before 3]
368
0
        index = findCommonNode(index, UCOL_SECONDARY);
369
0
        if(strength >= UCOL_TERTIARY) {
370
0
            index = findCommonNode(index, UCOL_TERTIARY);
371
0
        }
372
        // findCommonNode() stayed on the stronger node or moved to
373
        // an explicit common-weight node of the reset-before strength.
374
0
        node = nodes.elementAti(index);
375
0
        if(strengthFromNode(node) == strength) {
376
            // Found a same-strength node with an explicit weight.
377
0
            uint32_t weight16 = weight16FromNode(node);
378
0
            if(weight16 == 0) {
379
0
                errorCode = U_UNSUPPORTED_ERROR;
380
0
                if(strength == UCOL_SECONDARY) {
381
0
                    parserErrorReason = "reset secondary-before secondary ignorable not possible";
382
0
                } else {
383
0
                    parserErrorReason = "reset tertiary-before completely ignorable not possible";
384
0
                }
385
0
                return;
386
0
            }
387
0
            U_ASSERT(weight16 > Collation::BEFORE_WEIGHT16);
388
            // Reset to just before this node.
389
            // Insert the preceding same-level explicit weight if it is not there already.
390
            // Which explicit weight immediately precedes this one?
391
0
            weight16 = getWeight16Before(index, node, strength);
392
            // Does this preceding weight have a node?
393
0
            uint32_t previousWeight16;
394
0
            int32_t previousIndex = previousIndexFromNode(node);
395
0
            for(int32_t i = previousIndex;; i = previousIndexFromNode(node)) {
396
0
                node = nodes.elementAti(i);
397
0
                int32_t previousStrength = strengthFromNode(node);
398
0
                if(previousStrength < strength) {
399
0
                    U_ASSERT(weight16 >= Collation::COMMON_WEIGHT16 || i == previousIndex);
400
                    // Either the reset element has an above-common weight and
401
                    // the parent node provides the implied common weight,
402
                    // or the reset element has a weight<=common in the node
403
                    // right after the parent, and we need to insert the preceding weight.
404
0
                    previousWeight16 = Collation::COMMON_WEIGHT16;
405
0
                    break;
406
0
                } else if(previousStrength == strength && !isTailoredNode(node)) {
407
0
                    previousWeight16 = weight16FromNode(node);
408
0
                    break;
409
0
                }
410
                // Skip weaker nodes and same-level tailored nodes.
411
0
            }
412
0
            if(previousWeight16 == weight16) {
413
                // The preceding weight has a node,
414
                // maybe with following weaker or tailored nodes.
415
                // Reset to the last of them.
416
0
                index = previousIndex;
417
0
            } else {
418
                // Insert a node with the preceding weight, reset to that.
419
0
                node = nodeFromWeight16(weight16) | nodeFromStrength(strength);
420
0
                index = insertNodeBetween(previousIndex, index, node, errorCode);
421
0
            }
422
0
        } else {
423
            // Found a stronger node with implied strength-common weight.
424
0
            uint32_t weight16 = getWeight16Before(index, node, strength);
425
0
            index = findOrInsertWeakNode(index, weight16, strength, errorCode);
426
0
        }
427
        // Strength of the temporary CE = strength of its reset position.
428
        // Code above raises an error if the before-strength is stronger.
429
0
        strength = ceStrength(ces[cesLength - 1]);
430
0
    }
431
0
    if(U_FAILURE(errorCode)) {
432
0
        parserErrorReason = "inserting reset position for &[before n]";
433
0
        return;
434
0
    }
435
0
    ces[cesLength - 1] = tempCEFromIndexAndStrength(index, strength);
436
0
}
437
438
uint32_t
439
0
CollationBuilder::getWeight16Before(int32_t index, int64_t node, int32_t level) {
440
0
    U_ASSERT(strengthFromNode(node) < level || !isTailoredNode(node));
441
    // Collect the root CE weights if this node is for a root CE.
442
    // If it is not, then return the low non-primary boundary for a tailored CE.
443
0
    uint32_t t;
444
0
    if(strengthFromNode(node) == UCOL_TERTIARY) {
445
0
        t = weight16FromNode(node);
446
0
    } else {
447
0
        t = Collation::COMMON_WEIGHT16;  // Stronger node with implied common weight.
448
0
    }
449
0
    while(strengthFromNode(node) > UCOL_SECONDARY) {
450
0
        index = previousIndexFromNode(node);
451
0
        node = nodes.elementAti(index);
452
0
    }
453
0
    if(isTailoredNode(node)) {
454
0
        return Collation::BEFORE_WEIGHT16;
455
0
    }
456
0
    uint32_t s;
457
0
    if(strengthFromNode(node) == UCOL_SECONDARY) {
458
0
        s = weight16FromNode(node);
459
0
    } else {
460
0
        s = Collation::COMMON_WEIGHT16;  // Stronger node with implied common weight.
461
0
    }
462
0
    while(strengthFromNode(node) > UCOL_PRIMARY) {
463
0
        index = previousIndexFromNode(node);
464
0
        node = nodes.elementAti(index);
465
0
    }
466
0
    if(isTailoredNode(node)) {
467
0
        return Collation::BEFORE_WEIGHT16;
468
0
    }
469
    // [p, s, t] is a root CE. Return the preceding weight for the requested level.
470
0
    uint32_t p = weight32FromNode(node);
471
0
    uint32_t weight16;
472
0
    if(level == UCOL_SECONDARY) {
473
0
        weight16 = rootElements.getSecondaryBefore(p, s);
474
0
    } else {
475
0
        weight16 = rootElements.getTertiaryBefore(p, s, t);
476
0
        U_ASSERT((weight16 & ~Collation::ONLY_TERTIARY_MASK) == 0);
477
0
    }
478
0
    return weight16;
479
0
}
480
481
int64_t
482
CollationBuilder::getSpecialResetPosition(const UnicodeString &str,
483
0
                                          const char *&parserErrorReason, UErrorCode &errorCode) {
484
0
    U_ASSERT(str.length() == 2);
485
0
    int64_t ce;
486
0
    int32_t strength = UCOL_PRIMARY;
487
0
    UBool isBoundary = FALSE;
488
0
    UChar32 pos = str.charAt(1) - CollationRuleParser::POS_BASE;
489
0
    U_ASSERT(0 <= pos && pos <= CollationRuleParser::LAST_TRAILING);
490
0
    switch(pos) {
491
0
    case CollationRuleParser::FIRST_TERTIARY_IGNORABLE:
492
        // Quaternary CEs are not supported.
493
        // Non-zero quaternary weights are possible only on tertiary or stronger CEs.
494
0
        return 0;
495
0
    case CollationRuleParser::LAST_TERTIARY_IGNORABLE:
496
0
        return 0;
497
0
    case CollationRuleParser::FIRST_SECONDARY_IGNORABLE: {
498
        // Look for a tailored tertiary node after [0, 0, 0].
499
0
        int32_t index = findOrInsertNodeForRootCE(0, UCOL_TERTIARY, errorCode);
500
0
        if(U_FAILURE(errorCode)) { return 0; }
501
0
        int64_t node = nodes.elementAti(index);
502
0
        if((index = nextIndexFromNode(node)) != 0) {
503
0
            node = nodes.elementAti(index);
504
0
            U_ASSERT(strengthFromNode(node) <= UCOL_TERTIARY);
505
0
            if(isTailoredNode(node) && strengthFromNode(node) == UCOL_TERTIARY) {
506
0
                return tempCEFromIndexAndStrength(index, UCOL_TERTIARY);
507
0
            }
508
0
        }
509
0
        return rootElements.getFirstTertiaryCE();
510
        // No need to look for nodeHasAnyBefore() on a tertiary node.
511
0
    }
512
0
    case CollationRuleParser::LAST_SECONDARY_IGNORABLE:
513
0
        ce = rootElements.getLastTertiaryCE();
514
0
        strength = UCOL_TERTIARY;
515
0
        break;
516
0
    case CollationRuleParser::FIRST_PRIMARY_IGNORABLE: {
517
        // Look for a tailored secondary node after [0, 0, *].
518
0
        int32_t index = findOrInsertNodeForRootCE(0, UCOL_SECONDARY, errorCode);
519
0
        if(U_FAILURE(errorCode)) { return 0; }
520
0
        int64_t node = nodes.elementAti(index);
521
0
        while((index = nextIndexFromNode(node)) != 0) {
522
0
            node = nodes.elementAti(index);
523
0
            strength = strengthFromNode(node);
524
0
            if(strength < UCOL_SECONDARY) { break; }
525
0
            if(strength == UCOL_SECONDARY) {
526
0
                if(isTailoredNode(node)) {
527
0
                    if(nodeHasBefore3(node)) {
528
0
                        index = nextIndexFromNode(nodes.elementAti(nextIndexFromNode(node)));
529
0
                        U_ASSERT(isTailoredNode(nodes.elementAti(index)));
530
0
                    }
531
0
                    return tempCEFromIndexAndStrength(index, UCOL_SECONDARY);
532
0
                } else {
533
0
                    break;
534
0
                }
535
0
            }
536
0
        }
537
0
        ce = rootElements.getFirstSecondaryCE();
538
0
        strength = UCOL_SECONDARY;
539
0
        break;
540
0
    }
541
0
    case CollationRuleParser::LAST_PRIMARY_IGNORABLE:
542
0
        ce = rootElements.getLastSecondaryCE();
543
0
        strength = UCOL_SECONDARY;
544
0
        break;
545
0
    case CollationRuleParser::FIRST_VARIABLE:
546
0
        ce = rootElements.getFirstPrimaryCE();
547
0
        isBoundary = TRUE;  // FractionalUCA.txt: FDD1 00A0, SPACE first primary
548
0
        break;
549
0
    case CollationRuleParser::LAST_VARIABLE:
550
0
        ce = rootElements.lastCEWithPrimaryBefore(variableTop + 1);
551
0
        break;
552
0
    case CollationRuleParser::FIRST_REGULAR:
553
0
        ce = rootElements.firstCEWithPrimaryAtLeast(variableTop + 1);
554
0
        isBoundary = TRUE;  // FractionalUCA.txt: FDD1 263A, SYMBOL first primary
555
0
        break;
556
0
    case CollationRuleParser::LAST_REGULAR:
557
        // Use the Hani-first-primary rather than the actual last "regular" CE before it,
558
        // for backward compatibility with behavior before the introduction of
559
        // script-first-primary CEs in the root collator.
560
0
        ce = rootElements.firstCEWithPrimaryAtLeast(
561
0
            baseData->getFirstPrimaryForGroup(USCRIPT_HAN));
562
0
        break;
563
0
    case CollationRuleParser::FIRST_IMPLICIT:
564
0
        ce = baseData->getSingleCE(0x4e00, errorCode);
565
0
        break;
566
0
    case CollationRuleParser::LAST_IMPLICIT:
567
        // We do not support tailoring to an unassigned-implicit CE.
568
0
        errorCode = U_UNSUPPORTED_ERROR;
569
0
        parserErrorReason = "reset to [last implicit] not supported";
570
0
        return 0;
571
0
    case CollationRuleParser::FIRST_TRAILING:
572
0
        ce = Collation::makeCE(Collation::FIRST_TRAILING_PRIMARY);
573
0
        isBoundary = TRUE;  // trailing first primary (there is no mapping for it)
574
0
        break;
575
0
    case CollationRuleParser::LAST_TRAILING:
576
0
        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
577
0
        parserErrorReason = "LDML forbids tailoring to U+FFFF";
578
0
        return 0;
579
0
    default:
580
0
        UPRV_UNREACHABLE;
581
0
    }
582
583
0
    int32_t index = findOrInsertNodeForRootCE(ce, strength, errorCode);
584
0
    if(U_FAILURE(errorCode)) { return 0; }
585
0
    int64_t node = nodes.elementAti(index);
586
0
    if((pos & 1) == 0) {
587
        // even pos = [first xyz]
588
0
        if(!nodeHasAnyBefore(node) && isBoundary) {
589
            // A <group> first primary boundary is artificially added to FractionalUCA.txt.
590
            // It is reachable via its special contraction, but is not normally used.
591
            // Find the first character tailored after the boundary CE,
592
            // or the first real root CE after it.
593
0
            if((index = nextIndexFromNode(node)) != 0) {
594
                // If there is a following node, then it must be tailored
595
                // because there are no root CEs with a boundary primary
596
                // and non-common secondary/tertiary weights.
597
0
                node = nodes.elementAti(index);
598
0
                U_ASSERT(isTailoredNode(node));
599
0
                ce = tempCEFromIndexAndStrength(index, strength);
600
0
            } else {
601
0
                U_ASSERT(strength == UCOL_PRIMARY);
602
0
                uint32_t p = (uint32_t)(ce >> 32);
603
0
                int32_t pIndex = rootElements.findPrimary(p);
604
0
                UBool isCompressible = baseData->isCompressiblePrimary(p);
605
0
                p = rootElements.getPrimaryAfter(p, pIndex, isCompressible);
606
0
                ce = Collation::makeCE(p);
607
0
                index = findOrInsertNodeForRootCE(ce, UCOL_PRIMARY, errorCode);
608
0
                if(U_FAILURE(errorCode)) { return 0; }
609
0
                node = nodes.elementAti(index);
610
0
            }
611
0
        }
612
0
        if(nodeHasAnyBefore(node)) {
613
            // Get the first node that was tailored before this one at a weaker strength.
614
0
            if(nodeHasBefore2(node)) {
615
0
                index = nextIndexFromNode(nodes.elementAti(nextIndexFromNode(node)));
616
0
                node = nodes.elementAti(index);
617
0
            }
618
0
            if(nodeHasBefore3(node)) {
619
0
                index = nextIndexFromNode(nodes.elementAti(nextIndexFromNode(node)));
620
0
            }
621
0
            U_ASSERT(isTailoredNode(nodes.elementAti(index)));
622
0
            ce = tempCEFromIndexAndStrength(index, strength);
623
0
        }
624
0
    } else {
625
        // odd pos = [last xyz]
626
        // Find the last node that was tailored after the [last xyz]
627
        // at a strength no greater than the position's strength.
628
0
        for(;;) {
629
0
            int32_t nextIndex = nextIndexFromNode(node);
630
0
            if(nextIndex == 0) { break; }
631
0
            int64_t nextNode = nodes.elementAti(nextIndex);
632
0
            if(strengthFromNode(nextNode) < strength) { break; }
633
0
            index = nextIndex;
634
0
            node = nextNode;
635
0
        }
636
        // Do not make a temporary CE for a root node.
637
        // This last node might be the node for the root CE itself,
638
        // or a node with a common secondary or tertiary weight.
639
0
        if(isTailoredNode(node)) {
640
0
            ce = tempCEFromIndexAndStrength(index, strength);
641
0
        }
642
0
    }
643
0
    return ce;
644
0
}
645
646
void
647
CollationBuilder::addRelation(int32_t strength, const UnicodeString &prefix,
648
                              const UnicodeString &str, const UnicodeString &extension,
649
0
                              const char *&parserErrorReason, UErrorCode &errorCode) {
650
0
    if(U_FAILURE(errorCode)) { return; }
651
0
    UnicodeString nfdPrefix;
652
0
    if(!prefix.isEmpty()) {
653
0
        nfd.normalize(prefix, nfdPrefix, errorCode);
654
0
        if(U_FAILURE(errorCode)) {
655
0
            parserErrorReason = "normalizing the relation prefix";
656
0
            return;
657
0
        }
658
0
    }
659
0
    UnicodeString nfdString = nfd.normalize(str, errorCode);
660
0
    if(U_FAILURE(errorCode)) {
661
0
        parserErrorReason = "normalizing the relation string";
662
0
        return;
663
0
    }
664
665
    // The runtime code decomposes Hangul syllables on the fly,
666
    // with recursive processing but without making the Jamo pieces visible for matching.
667
    // It does not work with certain types of contextual mappings.
668
0
    int32_t nfdLength = nfdString.length();
669
0
    if(nfdLength >= 2) {
670
0
        UChar c = nfdString.charAt(0);
671
0
        if(Hangul::isJamoL(c) || Hangul::isJamoV(c)) {
672
            // While handling a Hangul syllable, contractions starting with Jamo L or V
673
            // would not see the following Jamo of that syllable.
674
0
            errorCode = U_UNSUPPORTED_ERROR;
675
0
            parserErrorReason = "contractions starting with conjoining Jamo L or V not supported";
676
0
            return;
677
0
        }
678
0
        c = nfdString.charAt(nfdLength - 1);
679
0
        if(Hangul::isJamoL(c) ||
680
0
                (Hangul::isJamoV(c) && Hangul::isJamoL(nfdString.charAt(nfdLength - 2)))) {
681
            // A contraction ending with Jamo L or L+V would require
682
            // generating Hangul syllables in addTailComposites() (588 for a Jamo L),
683
            // or decomposing a following Hangul syllable on the fly, during contraction matching.
684
0
            errorCode = U_UNSUPPORTED_ERROR;
685
0
            parserErrorReason = "contractions ending with conjoining Jamo L or L+V not supported";
686
0
            return;
687
0
        }
688
        // A Hangul syllable completely inside a contraction is ok.
689
0
    }
690
    // Note: If there is a prefix, then the parser checked that
691
    // both the prefix and the string begin with NFC boundaries (not Jamo V or T).
692
    // Therefore: prefix.isEmpty() || !isJamoVOrT(nfdString.charAt(0))
693
    // (While handling a Hangul syllable, prefixes on Jamo V or T
694
    // would not see the previous Jamo of that syllable.)
695
696
0
    if(strength != UCOL_IDENTICAL) {
697
        // Find the node index after which we insert the new tailored node.
698
0
        int32_t index = findOrInsertNodeForCEs(strength, parserErrorReason, errorCode);
699
0
        U_ASSERT(cesLength > 0);
700
0
        int64_t ce = ces[cesLength - 1];
701
0
        if(strength == UCOL_PRIMARY && !isTempCE(ce) && (uint32_t)(ce >> 32) == 0) {
702
            // There is no primary gap between ignorables and the space-first-primary.
703
0
            errorCode = U_UNSUPPORTED_ERROR;
704
0
            parserErrorReason = "tailoring primary after ignorables not supported";
705
0
            return;
706
0
        }
707
0
        if(strength == UCOL_QUATERNARY && ce == 0) {
708
            // The CE data structure does not support non-zero quaternary weights
709
            // on tertiary ignorables.
710
0
            errorCode = U_UNSUPPORTED_ERROR;
711
0
            parserErrorReason = "tailoring quaternary after tertiary ignorables not supported";
712
0
            return;
713
0
        }
714
        // Insert the new tailored node.
715
0
        index = insertTailoredNodeAfter(index, strength, errorCode);
716
0
        if(U_FAILURE(errorCode)) {
717
0
            parserErrorReason = "modifying collation elements";
718
0
            return;
719
0
        }
720
        // Strength of the temporary CE:
721
        // The new relation may yield a stronger CE but not a weaker one.
722
0
        int32_t tempStrength = ceStrength(ce);
723
0
        if(strength < tempStrength) { tempStrength = strength; }
724
0
        ces[cesLength - 1] = tempCEFromIndexAndStrength(index, tempStrength);
725
0
    }
726
727
0
    setCaseBits(nfdString, parserErrorReason, errorCode);
728
0
    if(U_FAILURE(errorCode)) { return; }
729
730
0
    int32_t cesLengthBeforeExtension = cesLength;
731
0
    if(!extension.isEmpty()) {
732
0
        UnicodeString nfdExtension = nfd.normalize(extension, errorCode);
733
0
        if(U_FAILURE(errorCode)) {
734
0
            parserErrorReason = "normalizing the relation extension";
735
0
            return;
736
0
        }
737
0
        cesLength = dataBuilder->getCEs(nfdExtension, ces, cesLength);
738
0
        if(cesLength > Collation::MAX_EXPANSION_LENGTH) {
739
0
            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
740
0
            parserErrorReason =
741
0
                "extension string adds too many collation elements (more than 31 total)";
742
0
            return;
743
0
        }
744
0
    }
745
0
    uint32_t ce32 = Collation::UNASSIGNED_CE32;
746
0
    if((prefix != nfdPrefix || str != nfdString) &&
747
0
            !ignorePrefix(prefix, errorCode) && !ignoreString(str, errorCode)) {
748
        // Map from the original input to the CEs.
749
        // We do this in case the canonical closure is incomplete,
750
        // so that it is possible to explicitly provide the missing mappings.
751
0
        ce32 = addIfDifferent(prefix, str, ces, cesLength, ce32, errorCode);
752
0
    }
753
0
    addWithClosure(nfdPrefix, nfdString, ces, cesLength, ce32, errorCode);
754
0
    if(U_FAILURE(errorCode)) {
755
0
        parserErrorReason = "writing collation elements";
756
0
        return;
757
0
    }
758
0
    cesLength = cesLengthBeforeExtension;
759
0
}
760
761
int32_t
762
CollationBuilder::findOrInsertNodeForCEs(int32_t strength, const char *&parserErrorReason,
763
0
                                         UErrorCode &errorCode) {
764
0
    if(U_FAILURE(errorCode)) { return 0; }
765
0
    U_ASSERT(UCOL_PRIMARY <= strength && strength <= UCOL_QUATERNARY);
766
767
    // Find the last CE that is at least as "strong" as the requested difference.
768
    // Note: Stronger is smaller (UCOL_PRIMARY=0).
769
0
    int64_t ce;
770
0
    for(;; --cesLength) {
771
0
        if(cesLength == 0) {
772
0
            ce = ces[0] = 0;
773
0
            cesLength = 1;
774
0
            break;
775
0
        } else {
776
0
            ce = ces[cesLength - 1];
777
0
        }
778
0
        if(ceStrength(ce) <= strength) { break; }
779
0
    }
780
781
0
    if(isTempCE(ce)) {
782
        // No need to findCommonNode() here for lower levels
783
        // because insertTailoredNodeAfter() will do that anyway.
784
0
        return indexFromTempCE(ce);
785
0
    }
786
787
    // root CE
788
0
    if((uint8_t)(ce >> 56) == Collation::UNASSIGNED_IMPLICIT_BYTE) {
789
0
        errorCode = U_UNSUPPORTED_ERROR;
790
0
        parserErrorReason = "tailoring relative to an unassigned code point not supported";
791
0
        return 0;
792
0
    }
793
0
    return findOrInsertNodeForRootCE(ce, strength, errorCode);
794
0
}
795
796
int32_t
797
0
CollationBuilder::findOrInsertNodeForRootCE(int64_t ce, int32_t strength, UErrorCode &errorCode) {
798
0
    if(U_FAILURE(errorCode)) { return 0; }
799
0
    U_ASSERT((uint8_t)(ce >> 56) != Collation::UNASSIGNED_IMPLICIT_BYTE);
800
801
    // Find or insert the node for each of the root CE's weights,
802
    // down to the requested level/strength.
803
    // Root CEs must have common=zero quaternary weights (for which we never insert any nodes).
804
0
    U_ASSERT((ce & 0xc0) == 0);
805
0
    int32_t index = findOrInsertNodeForPrimary((uint32_t)(ce >> 32), errorCode);
806
0
    if(strength >= UCOL_SECONDARY) {
807
0
        uint32_t lower32 = (uint32_t)ce;
808
0
        index = findOrInsertWeakNode(index, lower32 >> 16, UCOL_SECONDARY, errorCode);
809
0
        if(strength >= UCOL_TERTIARY) {
810
0
            index = findOrInsertWeakNode(index, lower32 & Collation::ONLY_TERTIARY_MASK,
811
0
                                         UCOL_TERTIARY, errorCode);
812
0
        }
813
0
    }
814
0
    return index;
815
0
}
816
817
namespace {
818
819
/**
820
 * Like Java Collections.binarySearch(List, key, Comparator).
821
 *
822
 * @return the index>=0 where the item was found,
823
 *         or the index<0 for inserting the string at ~index in sorted order
824
 *         (index into rootPrimaryIndexes)
825
 */
826
int32_t
827
binarySearchForRootPrimaryNode(const int32_t *rootPrimaryIndexes, int32_t length,
828
0
                               const int64_t *nodes, uint32_t p) {
829
0
    if(length == 0) { return ~0; }
830
0
    int32_t start = 0;
831
0
    int32_t limit = length;
832
0
    for (;;) {
833
0
        int32_t i = (start + limit) / 2;
834
0
        int64_t node = nodes[rootPrimaryIndexes[i]];
835
0
        uint32_t nodePrimary = (uint32_t)(node >> 32);  // weight32FromNode(node)
836
0
        if (p == nodePrimary) {
837
0
            return i;
838
0
        } else if (p < nodePrimary) {
839
0
            if (i == start) {
840
0
                return ~start;  // insert s before i
841
0
            }
842
0
            limit = i;
843
0
        } else {
844
0
            if (i == start) {
845
0
                return ~(start + 1);  // insert s after i
846
0
            }
847
0
            start = i;
848
0
        }
849
0
    }
850
0
}
851
852
}  // namespace
853
854
int32_t
855
0
CollationBuilder::findOrInsertNodeForPrimary(uint32_t p, UErrorCode &errorCode) {
856
0
    if(U_FAILURE(errorCode)) { return 0; }
857
858
0
    int32_t rootIndex = binarySearchForRootPrimaryNode(
859
0
        rootPrimaryIndexes.getBuffer(), rootPrimaryIndexes.size(), nodes.getBuffer(), p);
860
0
    if(rootIndex >= 0) {
861
0
        return rootPrimaryIndexes.elementAti(rootIndex);
862
0
    } else {
863
        // Start a new list of nodes with this primary.
864
0
        int32_t index = nodes.size();
865
0
        nodes.addElement(nodeFromWeight32(p), errorCode);
866
0
        rootPrimaryIndexes.insertElementAt(index, ~rootIndex, errorCode);
867
0
        return index;
868
0
    }
869
0
}
870
871
int32_t
872
0
CollationBuilder::findOrInsertWeakNode(int32_t index, uint32_t weight16, int32_t level, UErrorCode &errorCode) {
873
0
    if(U_FAILURE(errorCode)) { return 0; }
874
0
    U_ASSERT(0 <= index && index < nodes.size());
875
0
    U_ASSERT(UCOL_SECONDARY <= level && level <= UCOL_TERTIARY);
876
877
0
    if(weight16 == Collation::COMMON_WEIGHT16) {
878
0
        return findCommonNode(index, level);
879
0
    }
880
881
    // If this will be the first below-common weight for the parent node,
882
    // then we will also need to insert a common weight after it.
883
0
    int64_t node = nodes.elementAti(index);
884
0
    U_ASSERT(strengthFromNode(node) < level);  // parent node is stronger
885
0
    if(weight16 != 0 && weight16 < Collation::COMMON_WEIGHT16) {
886
0
        int32_t hasThisLevelBefore = level == UCOL_SECONDARY ? HAS_BEFORE2 : HAS_BEFORE3;
887
0
        if((node & hasThisLevelBefore) == 0) {
888
            // The parent node has an implied level-common weight.
889
0
            int64_t commonNode =
890
0
                nodeFromWeight16(Collation::COMMON_WEIGHT16) | nodeFromStrength(level);
891
0
            if(level == UCOL_SECONDARY) {
892
                // Move the HAS_BEFORE3 flag from the parent node
893
                // to the new secondary common node.
894
0
                commonNode |= node & HAS_BEFORE3;
895
0
                node &= ~(int64_t)HAS_BEFORE3;
896
0
            }
897
0
            nodes.setElementAt(node | hasThisLevelBefore, index);
898
            // Insert below-common-weight node.
899
0
            int32_t nextIndex = nextIndexFromNode(node);
900
0
            node = nodeFromWeight16(weight16) | nodeFromStrength(level);
901
0
            index = insertNodeBetween(index, nextIndex, node, errorCode);
902
            // Insert common-weight node.
903
0
            insertNodeBetween(index, nextIndex, commonNode, errorCode);
904
            // Return index of below-common-weight node.
905
0
            return index;
906
0
        }
907
0
    }
908
909
    // Find the root CE's weight for this level.
910
    // Postpone insertion if not found:
911
    // Insert the new root node before the next stronger node,
912
    // or before the next root node with the same strength and a larger weight.
913
0
    int32_t nextIndex;
914
0
    while((nextIndex = nextIndexFromNode(node)) != 0) {
915
0
        node = nodes.elementAti(nextIndex);
916
0
        int32_t nextStrength = strengthFromNode(node);
917
0
        if(nextStrength <= level) {
918
            // Insert before a stronger node.
919
0
            if(nextStrength < level) { break; }
920
            // nextStrength == level
921
0
            if(!isTailoredNode(node)) {
922
0
                uint32_t nextWeight16 = weight16FromNode(node);
923
0
                if(nextWeight16 == weight16) {
924
                    // Found the node for the root CE up to this level.
925
0
                    return nextIndex;
926
0
                }
927
                // Insert before a node with a larger same-strength weight.
928
0
                if(nextWeight16 > weight16) { break; }
929
0
            }
930
0
        }
931
        // Skip the next node.
932
0
        index = nextIndex;
933
0
    }
934
0
    node = nodeFromWeight16(weight16) | nodeFromStrength(level);
935
0
    return insertNodeBetween(index, nextIndex, node, errorCode);
936
0
}
937
938
int32_t
939
0
CollationBuilder::insertTailoredNodeAfter(int32_t index, int32_t strength, UErrorCode &errorCode) {
940
0
    if(U_FAILURE(errorCode)) { return 0; }
941
0
    U_ASSERT(0 <= index && index < nodes.size());
942
0
    if(strength >= UCOL_SECONDARY) {
943
0
        index = findCommonNode(index, UCOL_SECONDARY);
944
0
        if(strength >= UCOL_TERTIARY) {
945
0
            index = findCommonNode(index, UCOL_TERTIARY);
946
0
        }
947
0
    }
948
    // Postpone insertion:
949
    // Insert the new node before the next one with a strength at least as strong.
950
0
    int64_t node = nodes.elementAti(index);
951
0
    int32_t nextIndex;
952
0
    while((nextIndex = nextIndexFromNode(node)) != 0) {
953
0
        node = nodes.elementAti(nextIndex);
954
0
        if(strengthFromNode(node) <= strength) { break; }
955
        // Skip the next node which has a weaker (larger) strength than the new one.
956
0
        index = nextIndex;
957
0
    }
958
0
    node = IS_TAILORED | nodeFromStrength(strength);
959
0
    return insertNodeBetween(index, nextIndex, node, errorCode);
960
0
}
961
962
int32_t
963
CollationBuilder::insertNodeBetween(int32_t index, int32_t nextIndex, int64_t node,
964
0
                                    UErrorCode &errorCode) {
965
0
    if(U_FAILURE(errorCode)) { return 0; }
966
0
    U_ASSERT(previousIndexFromNode(node) == 0);
967
0
    U_ASSERT(nextIndexFromNode(node) == 0);
968
0
    U_ASSERT(nextIndexFromNode(nodes.elementAti(index)) == nextIndex);
969
    // Append the new node and link it to the existing nodes.
970
0
    int32_t newIndex = nodes.size();
971
0
    node |= nodeFromPreviousIndex(index) | nodeFromNextIndex(nextIndex);
972
0
    nodes.addElement(node, errorCode);
973
0
    if(U_FAILURE(errorCode)) { return 0; }
974
    // nodes[index].nextIndex = newIndex
975
0
    node = nodes.elementAti(index);
976
0
    nodes.setElementAt(changeNodeNextIndex(node, newIndex), index);
977
    // nodes[nextIndex].previousIndex = newIndex
978
0
    if(nextIndex != 0) {
979
0
        node = nodes.elementAti(nextIndex);
980
0
        nodes.setElementAt(changeNodePreviousIndex(node, newIndex), nextIndex);
981
0
    }
982
0
    return newIndex;
983
0
}
984
985
int32_t
986
0
CollationBuilder::findCommonNode(int32_t index, int32_t strength) const {
987
0
    U_ASSERT(UCOL_SECONDARY <= strength && strength <= UCOL_TERTIARY);
988
0
    int64_t node = nodes.elementAti(index);
989
0
    if(strengthFromNode(node) >= strength) {
990
        // The current node is no stronger.
991
0
        return index;
992
0
    }
993
0
    if(strength == UCOL_SECONDARY ? !nodeHasBefore2(node) : !nodeHasBefore3(node)) {
994
        // The current node implies the strength-common weight.
995
0
        return index;
996
0
    }
997
0
    index = nextIndexFromNode(node);
998
0
    node = nodes.elementAti(index);
999
0
    U_ASSERT(!isTailoredNode(node) && strengthFromNode(node) == strength &&
1000
0
            weight16FromNode(node) < Collation::COMMON_WEIGHT16);
1001
    // Skip to the explicit common node.
1002
0
    do {
1003
0
        index = nextIndexFromNode(node);
1004
0
        node = nodes.elementAti(index);
1005
0
        U_ASSERT(strengthFromNode(node) >= strength);
1006
0
    } while(isTailoredNode(node) || strengthFromNode(node) > strength ||
1007
0
            weight16FromNode(node) < Collation::COMMON_WEIGHT16);
1008
0
    U_ASSERT(weight16FromNode(node) == Collation::COMMON_WEIGHT16);
1009
0
    return index;
1010
0
}
1011
1012
void
1013
CollationBuilder::setCaseBits(const UnicodeString &nfdString,
1014
0
                              const char *&parserErrorReason, UErrorCode &errorCode) {
1015
0
    if(U_FAILURE(errorCode)) { return; }
1016
0
    int32_t numTailoredPrimaries = 0;
1017
0
    for(int32_t i = 0; i < cesLength; ++i) {
1018
0
        if(ceStrength(ces[i]) == UCOL_PRIMARY) { ++numTailoredPrimaries; }
1019
0
    }
1020
    // We should not be able to get too many case bits because
1021
    // cesLength<=31==MAX_EXPANSION_LENGTH.
1022
    // 31 pairs of case bits fit into an int64_t without setting its sign bit.
1023
0
    U_ASSERT(numTailoredPrimaries <= 31);
1024
1025
0
    int64_t cases = 0;
1026
0
    if(numTailoredPrimaries > 0) {
1027
0
        const UChar *s = nfdString.getBuffer();
1028
0
        UTF16CollationIterator baseCEs(baseData, FALSE, s, s, s + nfdString.length());
1029
0
        int32_t baseCEsLength = baseCEs.fetchCEs(errorCode) - 1;
1030
0
        if(U_FAILURE(errorCode)) {
1031
0
            parserErrorReason = "fetching root CEs for tailored string";
1032
0
            return;
1033
0
        }
1034
0
        U_ASSERT(baseCEsLength >= 0 && baseCEs.getCE(baseCEsLength) == Collation::NO_CE);
1035
1036
0
        uint32_t lastCase = 0;
1037
0
        int32_t numBasePrimaries = 0;
1038
0
        for(int32_t i = 0; i < baseCEsLength; ++i) {
1039
0
            int64_t ce = baseCEs.getCE(i);
1040
0
            if((ce >> 32) != 0) {
1041
0
                ++numBasePrimaries;
1042
0
                uint32_t c = ((uint32_t)ce >> 14) & 3;
1043
0
                U_ASSERT(c == 0 || c == 2);  // lowercase or uppercase, no mixed case in any base CE
1044
0
                if(numBasePrimaries < numTailoredPrimaries) {
1045
0
                    cases |= (int64_t)c << ((numBasePrimaries - 1) * 2);
1046
0
                } else if(numBasePrimaries == numTailoredPrimaries) {
1047
0
                    lastCase = c;
1048
0
                } else if(c != lastCase) {
1049
                    // There are more base primary CEs than tailored primaries.
1050
                    // Set mixed case if the case bits of the remainder differ.
1051
0
                    lastCase = 1;
1052
                    // Nothing more can change.
1053
0
                    break;
1054
0
                }
1055
0
            }
1056
0
        }
1057
0
        if(numBasePrimaries >= numTailoredPrimaries) {
1058
0
            cases |= (int64_t)lastCase << ((numTailoredPrimaries - 1) * 2);
1059
0
        }
1060
0
    }
1061
1062
0
    for(int32_t i = 0; i < cesLength; ++i) {
1063
0
        int64_t ce = ces[i] & INT64_C(0xffffffffffff3fff);  // clear old case bits
1064
0
        int32_t strength = ceStrength(ce);
1065
0
        if(strength == UCOL_PRIMARY) {
1066
0
            ce |= (cases & 3) << 14;
1067
0
            cases >>= 2;
1068
0
        } else if(strength == UCOL_TERTIARY) {
1069
            // Tertiary CEs must have uppercase bits.
1070
            // See the LDML spec, and comments in class CollationCompare.
1071
0
            ce |= 0x8000;
1072
0
        }
1073
        // Tertiary ignorable CEs must have 0 case bits.
1074
        // We set 0 case bits for secondary CEs too
1075
        // since currently only U+0345 is cased and maps to a secondary CE,
1076
        // and it is lowercase. Other secondaries are uncased.
1077
        // See [[:Cased:]&[:uca1=:]] where uca1 queries the root primary weight.
1078
0
        ces[i] = ce;
1079
0
    }
1080
0
}
1081
1082
void
1083
CollationBuilder::suppressContractions(const UnicodeSet &set, const char *&parserErrorReason,
1084
0
                                       UErrorCode &errorCode) {
1085
0
    if(U_FAILURE(errorCode)) { return; }
1086
0
    dataBuilder->suppressContractions(set, errorCode);
1087
0
    if(U_FAILURE(errorCode)) {
1088
0
        parserErrorReason = "application of [suppressContractions [set]] failed";
1089
0
    }
1090
0
}
1091
1092
void
1093
CollationBuilder::optimize(const UnicodeSet &set, const char *& /* parserErrorReason */,
1094
0
                           UErrorCode &errorCode) {
1095
0
    if(U_FAILURE(errorCode)) { return; }
1096
0
    optimizeSet.addAll(set);
1097
0
}
1098
1099
uint32_t
1100
CollationBuilder::addWithClosure(const UnicodeString &nfdPrefix, const UnicodeString &nfdString,
1101
                                 const int64_t newCEs[], int32_t newCEsLength, uint32_t ce32,
1102
0
                                 UErrorCode &errorCode) {
1103
    // Map from the NFD input to the CEs.
1104
0
    ce32 = addIfDifferent(nfdPrefix, nfdString, newCEs, newCEsLength, ce32, errorCode);
1105
0
    ce32 = addOnlyClosure(nfdPrefix, nfdString, newCEs, newCEsLength, ce32, errorCode);
1106
0
    addTailComposites(nfdPrefix, nfdString, errorCode);
1107
0
    return ce32;
1108
0
}
1109
1110
uint32_t
1111
CollationBuilder::addOnlyClosure(const UnicodeString &nfdPrefix, const UnicodeString &nfdString,
1112
                                 const int64_t newCEs[], int32_t newCEsLength, uint32_t ce32,
1113
0
                                 UErrorCode &errorCode) {
1114
0
    if(U_FAILURE(errorCode)) { return ce32; }
1115
1116
    // Map from canonically equivalent input to the CEs. (But not from the all-NFD input.)
1117
0
    if(nfdPrefix.isEmpty()) {
1118
0
        CanonicalIterator stringIter(nfdString, errorCode);
1119
0
        if(U_FAILURE(errorCode)) { return ce32; }
1120
0
        UnicodeString prefix;
1121
0
        for(;;) {
1122
0
            UnicodeString str = stringIter.next();
1123
0
            if(str.isBogus()) { break; }
1124
0
            if(ignoreString(str, errorCode) || str == nfdString) { continue; }
1125
0
            ce32 = addIfDifferent(prefix, str, newCEs, newCEsLength, ce32, errorCode);
1126
0
            if(U_FAILURE(errorCode)) { return ce32; }
1127
0
        }
1128
0
    } else {
1129
0
        CanonicalIterator prefixIter(nfdPrefix, errorCode);
1130
0
        CanonicalIterator stringIter(nfdString, errorCode);
1131
0
        if(U_FAILURE(errorCode)) { return ce32; }
1132
0
        for(;;) {
1133
0
            UnicodeString prefix = prefixIter.next();
1134
0
            if(prefix.isBogus()) { break; }
1135
0
            if(ignorePrefix(prefix, errorCode)) { continue; }
1136
0
            UBool samePrefix = prefix == nfdPrefix;
1137
0
            for(;;) {
1138
0
                UnicodeString str = stringIter.next();
1139
0
                if(str.isBogus()) { break; }
1140
0
                if(ignoreString(str, errorCode) || (samePrefix && str == nfdString)) { continue; }
1141
0
                ce32 = addIfDifferent(prefix, str, newCEs, newCEsLength, ce32, errorCode);
1142
0
                if(U_FAILURE(errorCode)) { return ce32; }
1143
0
            }
1144
0
            stringIter.reset();
1145
0
        }
1146
0
    }
1147
0
    return ce32;
1148
0
}
1149
1150
void
1151
CollationBuilder::addTailComposites(const UnicodeString &nfdPrefix, const UnicodeString &nfdString,
1152
0
                                    UErrorCode &errorCode) {
1153
0
    if(U_FAILURE(errorCode)) { return; }
1154
1155
    // Look for the last starter in the NFD string.
1156
0
    UChar32 lastStarter;
1157
0
    int32_t indexAfterLastStarter = nfdString.length();
1158
0
    for(;;) {
1159
0
        if(indexAfterLastStarter == 0) { return; }  // no starter at all
1160
0
        lastStarter = nfdString.char32At(indexAfterLastStarter - 1);
1161
0
        if(nfd.getCombiningClass(lastStarter) == 0) { break; }
1162
0
        indexAfterLastStarter -= U16_LENGTH(lastStarter);
1163
0
    }
1164
    // No closure to Hangul syllables since we decompose them on the fly.
1165
0
    if(Hangul::isJamoL(lastStarter)) { return; }
1166
1167
    // Are there any composites whose decomposition starts with the lastStarter?
1168
    // Note: Normalizer2Impl does not currently return start sets for NFC_QC=Maybe characters.
1169
    // We might find some more equivalent mappings here if it did.
1170
0
    UnicodeSet composites;
1171
0
    if(!nfcImpl.getCanonStartSet(lastStarter, composites)) { return; }
1172
1173
0
    UnicodeString decomp;
1174
0
    UnicodeString newNFDString, newString;
1175
0
    int64_t newCEs[Collation::MAX_EXPANSION_LENGTH];
1176
0
    UnicodeSetIterator iter(composites);
1177
0
    while(iter.next()) {
1178
0
        U_ASSERT(!iter.isString());
1179
0
        UChar32 composite = iter.getCodepoint();
1180
0
        nfd.getDecomposition(composite, decomp);
1181
0
        if(!mergeCompositeIntoString(nfdString, indexAfterLastStarter, composite, decomp,
1182
0
                                     newNFDString, newString, errorCode)) {
1183
0
            continue;
1184
0
        }
1185
0
        int32_t newCEsLength = dataBuilder->getCEs(nfdPrefix, newNFDString, newCEs, 0);
1186
0
        if(newCEsLength > Collation::MAX_EXPANSION_LENGTH) {
1187
            // Ignore mappings that we cannot store.
1188
0
            continue;
1189
0
        }
1190
        // Note: It is possible that the newCEs do not make use of the mapping
1191
        // for which we are adding the tail composites, in which case we might be adding
1192
        // unnecessary mappings.
1193
        // For example, when we add tail composites for ae^ (^=combining circumflex),
1194
        // UCA discontiguous-contraction matching does not find any matches
1195
        // for ae_^ (_=any combining diacritic below) *unless* there is also
1196
        // a contraction mapping for ae.
1197
        // Thus, if there is no ae contraction, then the ae^ mapping is ignored
1198
        // while fetching the newCEs for ae_^.
1199
        // TODO: Try to detect this effectively.
1200
        // (Alternatively, print a warning when prefix contractions are missing.)
1201
1202
        // We do not need an explicit mapping for the NFD strings.
1203
        // It is fine if the NFD input collates like this via a sequence of mappings.
1204
        // It also saves a little bit of space, and may reduce the set of characters with contractions.
1205
0
        uint32_t ce32 = addIfDifferent(nfdPrefix, newString,
1206
0
                                       newCEs, newCEsLength, Collation::UNASSIGNED_CE32, errorCode);
1207
0
        if(ce32 != Collation::UNASSIGNED_CE32) {
1208
            // was different, was added
1209
0
            addOnlyClosure(nfdPrefix, newNFDString, newCEs, newCEsLength, ce32, errorCode);
1210
0
        }
1211
0
    }
1212
0
}
1213
1214
UBool
1215
CollationBuilder::mergeCompositeIntoString(const UnicodeString &nfdString,
1216
                                           int32_t indexAfterLastStarter,
1217
                                           UChar32 composite, const UnicodeString &decomp,
1218
                                           UnicodeString &newNFDString, UnicodeString &newString,
1219
0
                                           UErrorCode &errorCode) const {
1220
0
    if(U_FAILURE(errorCode)) { return FALSE; }
1221
0
    U_ASSERT(nfdString.char32At(indexAfterLastStarter - 1) == decomp.char32At(0));
1222
0
    int32_t lastStarterLength = decomp.moveIndex32(0, 1);
1223
0
    if(lastStarterLength == decomp.length()) {
1224
        // Singleton decompositions should be found by addWithClosure()
1225
        // and the CanonicalIterator, so we can ignore them here.
1226
0
        return FALSE;
1227
0
    }
1228
0
    if(nfdString.compare(indexAfterLastStarter, 0x7fffffff,
1229
0
                         decomp, lastStarterLength, 0x7fffffff) == 0) {
1230
        // same strings, nothing new to be found here
1231
0
        return FALSE;
1232
0
    }
1233
1234
    // Make new FCD strings that combine a composite, or its decomposition,
1235
    // into the nfdString's last starter and the combining marks following it.
1236
    // Make an NFD version, and a version with the composite.
1237
0
    newNFDString.setTo(nfdString, 0, indexAfterLastStarter);
1238
0
    newString.setTo(nfdString, 0, indexAfterLastStarter - lastStarterLength).append(composite);
1239
1240
    // The following is related to discontiguous contraction matching,
1241
    // but builds only FCD strings (or else returns FALSE).
1242
0
    int32_t sourceIndex = indexAfterLastStarter;
1243
0
    int32_t decompIndex = lastStarterLength;
1244
    // Small optimization: We keep the source character across loop iterations
1245
    // because we do not always consume it,
1246
    // and then need not fetch it again nor look up its combining class again.
1247
0
    UChar32 sourceChar = U_SENTINEL;
1248
    // The cc variables need to be declared before the loop so that at the end
1249
    // they are set to the last combining classes seen.
1250
0
    uint8_t sourceCC = 0;
1251
0
    uint8_t decompCC = 0;
1252
0
    for(;;) {
1253
0
        if(sourceChar < 0) {
1254
0
            if(sourceIndex >= nfdString.length()) { break; }
1255
0
            sourceChar = nfdString.char32At(sourceIndex);
1256
0
            sourceCC = nfd.getCombiningClass(sourceChar);
1257
0
            U_ASSERT(sourceCC != 0);
1258
0
        }
1259
        // We consume a decomposition character in each iteration.
1260
0
        if(decompIndex >= decomp.length()) { break; }
1261
0
        UChar32 decompChar = decomp.char32At(decompIndex);
1262
0
        decompCC = nfd.getCombiningClass(decompChar);
1263
        // Compare the two characters and their combining classes.
1264
0
        if(decompCC == 0) {
1265
            // Unable to merge because the source contains a non-zero combining mark
1266
            // but the composite's decomposition contains another starter.
1267
            // The strings would not be equivalent.
1268
0
            return FALSE;
1269
0
        } else if(sourceCC < decompCC) {
1270
            // Composite + sourceChar would not be FCD.
1271
0
            return FALSE;
1272
0
        } else if(decompCC < sourceCC) {
1273
0
            newNFDString.append(decompChar);
1274
0
            decompIndex += U16_LENGTH(decompChar);
1275
0
        } else if(decompChar != sourceChar) {
1276
            // Blocked because same combining class.
1277
0
            return FALSE;
1278
0
        } else {  // match: decompChar == sourceChar
1279
0
            newNFDString.append(decompChar);
1280
0
            decompIndex += U16_LENGTH(decompChar);
1281
0
            sourceIndex += U16_LENGTH(decompChar);
1282
0
            sourceChar = U_SENTINEL;
1283
0
        }
1284
0
    }
1285
    // We are at the end of at least one of the two inputs.
1286
0
    if(sourceChar >= 0) {  // more characters from nfdString but not from decomp
1287
0
        if(sourceCC < decompCC) {
1288
            // Appending the next source character to the composite would not be FCD.
1289
0
            return FALSE;
1290
0
        }
1291
0
        newNFDString.append(nfdString, sourceIndex, 0x7fffffff);
1292
0
        newString.append(nfdString, sourceIndex, 0x7fffffff);
1293
0
    } else if(decompIndex < decomp.length()) {  // more characters from decomp, not from nfdString
1294
0
        newNFDString.append(decomp, decompIndex, 0x7fffffff);
1295
0
    }
1296
0
    U_ASSERT(nfd.isNormalized(newNFDString, errorCode));
1297
0
    U_ASSERT(fcd.isNormalized(newString, errorCode));
1298
0
    U_ASSERT(nfd.normalize(newString, errorCode) == newNFDString);  // canonically equivalent
1299
0
    return TRUE;
1300
0
}
1301
1302
UBool
1303
0
CollationBuilder::ignorePrefix(const UnicodeString &s, UErrorCode &errorCode) const {
1304
    // Do not map non-FCD prefixes.
1305
0
    return !isFCD(s, errorCode);
1306
0
}
1307
1308
UBool
1309
0
CollationBuilder::ignoreString(const UnicodeString &s, UErrorCode &errorCode) const {
1310
    // Do not map non-FCD strings.
1311
    // Do not map strings that start with Hangul syllables: We decompose those on the fly.
1312
0
    return !isFCD(s, errorCode) || Hangul::isHangul(s.charAt(0));
1313
0
}
1314
1315
UBool
1316
0
CollationBuilder::isFCD(const UnicodeString &s, UErrorCode &errorCode) const {
1317
0
    return U_SUCCESS(errorCode) && fcd.isNormalized(s, errorCode);
1318
0
}
1319
1320
void
1321
0
CollationBuilder::closeOverComposites(UErrorCode &errorCode) {
1322
0
    UnicodeSet composites(UNICODE_STRING_SIMPLE("[:NFD_QC=N:]"), errorCode);  // Java: static final
1323
0
    if(U_FAILURE(errorCode)) { return; }
1324
    // Hangul is decomposed on the fly during collation.
1325
0
    composites.remove(Hangul::HANGUL_BASE, Hangul::HANGUL_END);
1326
0
    UnicodeString prefix;  // empty
1327
0
    UnicodeString nfdString;
1328
0
    UnicodeSetIterator iter(composites);
1329
0
    while(iter.next()) {
1330
0
        U_ASSERT(!iter.isString());
1331
0
        nfd.getDecomposition(iter.getCodepoint(), nfdString);
1332
0
        cesLength = dataBuilder->getCEs(nfdString, ces, 0);
1333
0
        if(cesLength > Collation::MAX_EXPANSION_LENGTH) {
1334
            // Too many CEs from the decomposition (unusual), ignore this composite.
1335
            // We could add a capacity parameter to getCEs() and reallocate if necessary.
1336
            // However, this can only really happen in contrived cases.
1337
0
            continue;
1338
0
        }
1339
0
        const UnicodeString &composite(iter.getString());
1340
0
        addIfDifferent(prefix, composite, ces, cesLength, Collation::UNASSIGNED_CE32, errorCode);
1341
0
    }
1342
0
}
1343
1344
uint32_t
1345
CollationBuilder::addIfDifferent(const UnicodeString &prefix, const UnicodeString &str,
1346
                                 const int64_t newCEs[], int32_t newCEsLength, uint32_t ce32,
1347
0
                                 UErrorCode &errorCode) {
1348
0
    if(U_FAILURE(errorCode)) { return ce32; }
1349
0
    int64_t oldCEs[Collation::MAX_EXPANSION_LENGTH];
1350
0
    int32_t oldCEsLength = dataBuilder->getCEs(prefix, str, oldCEs, 0);
1351
0
    if(!sameCEs(newCEs, newCEsLength, oldCEs, oldCEsLength)) {
1352
0
        if(ce32 == Collation::UNASSIGNED_CE32) {
1353
0
            ce32 = dataBuilder->encodeCEs(newCEs, newCEsLength, errorCode);
1354
0
        }
1355
0
        dataBuilder->addCE32(prefix, str, ce32, errorCode);
1356
0
    }
1357
0
    return ce32;
1358
0
}
1359
1360
UBool
1361
CollationBuilder::sameCEs(const int64_t ces1[], int32_t ces1Length,
1362
0
                          const int64_t ces2[], int32_t ces2Length) {
1363
0
    if(ces1Length != ces2Length) {
1364
0
        return FALSE;
1365
0
    }
1366
0
    U_ASSERT(ces1Length <= Collation::MAX_EXPANSION_LENGTH);
1367
0
    for(int32_t i = 0; i < ces1Length; ++i) {
1368
0
        if(ces1[i] != ces2[i]) { return FALSE; }
1369
0
    }
1370
0
    return TRUE;
1371
0
}
1372
1373
#ifdef DEBUG_COLLATION_BUILDER
1374
1375
uint32_t
1376
alignWeightRight(uint32_t w) {
1377
    if(w != 0) {
1378
        while((w & 0xff) == 0) { w >>= 8; }
1379
    }
1380
    return w;
1381
}
1382
1383
#endif
1384
1385
void
1386
0
CollationBuilder::makeTailoredCEs(UErrorCode &errorCode) {
1387
0
    if(U_FAILURE(errorCode)) { return; }
1388
1389
0
    CollationWeights primaries, secondaries, tertiaries;
1390
0
    int64_t *nodesArray = nodes.getBuffer();
1391
#ifdef DEBUG_COLLATION_BUILDER
1392
        puts("\nCollationBuilder::makeTailoredCEs()");
1393
#endif
1394
1395
0
    for(int32_t rpi = 0; rpi < rootPrimaryIndexes.size(); ++rpi) {
1396
0
        int32_t i = rootPrimaryIndexes.elementAti(rpi);
1397
0
        int64_t node = nodesArray[i];
1398
0
        uint32_t p = weight32FromNode(node);
1399
0
        uint32_t s = p == 0 ? 0 : Collation::COMMON_WEIGHT16;
1400
0
        uint32_t t = s;
1401
0
        uint32_t q = 0;
1402
0
        UBool pIsTailored = FALSE;
1403
0
        UBool sIsTailored = FALSE;
1404
0
        UBool tIsTailored = FALSE;
1405
#ifdef DEBUG_COLLATION_BUILDER
1406
        printf("\nprimary     %lx\n", (long)alignWeightRight(p));
1407
#endif
1408
0
        int32_t pIndex = p == 0 ? 0 : rootElements.findPrimary(p);
1409
0
        int32_t nextIndex = nextIndexFromNode(node);
1410
0
        while(nextIndex != 0) {
1411
0
            i = nextIndex;
1412
0
            node = nodesArray[i];
1413
0
            nextIndex = nextIndexFromNode(node);
1414
0
            int32_t strength = strengthFromNode(node);
1415
0
            if(strength == UCOL_QUATERNARY) {
1416
0
                U_ASSERT(isTailoredNode(node));
1417
#ifdef DEBUG_COLLATION_BUILDER
1418
                printf("      quat+     ");
1419
#endif
1420
0
                if(q == 3) {
1421
0
                    errorCode = U_BUFFER_OVERFLOW_ERROR;
1422
0
                    errorReason = "quaternary tailoring gap too small";
1423
0
                    return;
1424
0
                }
1425
0
                ++q;
1426
0
            } else {
1427
0
                if(strength == UCOL_TERTIARY) {
1428
0
                    if(isTailoredNode(node)) {
1429
#ifdef DEBUG_COLLATION_BUILDER
1430
                        printf("    ter+        ");
1431
#endif
1432
0
                        if(!tIsTailored) {
1433
                            // First tailored tertiary node for [p, s].
1434
0
                            int32_t tCount = countTailoredNodes(nodesArray, nextIndex,
1435
0
                                                                UCOL_TERTIARY) + 1;
1436
0
                            uint32_t tLimit;
1437
0
                            if(t == 0) {
1438
                                // Gap at the beginning of the tertiary CE range.
1439
0
                                t = rootElements.getTertiaryBoundary() - 0x100;
1440
0
                                tLimit = rootElements.getFirstTertiaryCE() & Collation::ONLY_TERTIARY_MASK;
1441
0
                            } else if(!pIsTailored && !sIsTailored) {
1442
                                // p and s are root weights.
1443
0
                                tLimit = rootElements.getTertiaryAfter(pIndex, s, t);
1444
0
                            } else if(t == Collation::BEFORE_WEIGHT16) {
1445
0
                                tLimit = Collation::COMMON_WEIGHT16;
1446
0
                            } else {
1447
                                // [p, s] is tailored.
1448
0
                                U_ASSERT(t == Collation::COMMON_WEIGHT16);
1449
0
                                tLimit = rootElements.getTertiaryBoundary();
1450
0
                            }
1451
0
                            U_ASSERT(tLimit == 0x4000 || (tLimit & ~Collation::ONLY_TERTIARY_MASK) == 0);
1452
0
                            tertiaries.initForTertiary();
1453
0
                            if(!tertiaries.allocWeights(t, tLimit, tCount)) {
1454
0
                                errorCode = U_BUFFER_OVERFLOW_ERROR;
1455
0
                                errorReason = "tertiary tailoring gap too small";
1456
0
                                return;
1457
0
                            }
1458
0
                            tIsTailored = TRUE;
1459
0
                        }
1460
0
                        t = tertiaries.nextWeight();
1461
0
                        U_ASSERT(t != 0xffffffff);
1462
0
                    } else {
1463
0
                        t = weight16FromNode(node);
1464
0
                        tIsTailored = FALSE;
1465
#ifdef DEBUG_COLLATION_BUILDER
1466
                        printf("    ter     %lx\n", (long)alignWeightRight(t));
1467
#endif
1468
0
                    }
1469
0
                } else {
1470
0
                    if(strength == UCOL_SECONDARY) {
1471
0
                        if(isTailoredNode(node)) {
1472
#ifdef DEBUG_COLLATION_BUILDER
1473
                            printf("  sec+          ");
1474
#endif
1475
0
                            if(!sIsTailored) {
1476
                                // First tailored secondary node for p.
1477
0
                                int32_t sCount = countTailoredNodes(nodesArray, nextIndex,
1478
0
                                                                    UCOL_SECONDARY) + 1;
1479
0
                                uint32_t sLimit;
1480
0
                                if(s == 0) {
1481
                                    // Gap at the beginning of the secondary CE range.
1482
0
                                    s = rootElements.getSecondaryBoundary() - 0x100;
1483
0
                                    sLimit = rootElements.getFirstSecondaryCE() >> 16;
1484
0
                                } else if(!pIsTailored) {
1485
                                    // p is a root primary.
1486
0
                                    sLimit = rootElements.getSecondaryAfter(pIndex, s);
1487
0
                                } else if(s == Collation::BEFORE_WEIGHT16) {
1488
0
                                    sLimit = Collation::COMMON_WEIGHT16;
1489
0
                                } else {
1490
                                    // p is a tailored primary.
1491
0
                                    U_ASSERT(s == Collation::COMMON_WEIGHT16);
1492
0
                                    sLimit = rootElements.getSecondaryBoundary();
1493
0
                                }
1494
0
                                if(s == Collation::COMMON_WEIGHT16) {
1495
                                    // Do not tailor into the getSortKey() range of
1496
                                    // compressed common secondaries.
1497
0
                                    s = rootElements.getLastCommonSecondary();
1498
0
                                }
1499
0
                                secondaries.initForSecondary();
1500
0
                                if(!secondaries.allocWeights(s, sLimit, sCount)) {
1501
0
                                    errorCode = U_BUFFER_OVERFLOW_ERROR;
1502
0
                                    errorReason = "secondary tailoring gap too small";
1503
#ifdef DEBUG_COLLATION_BUILDER
1504
                                    printf("!secondaries.allocWeights(%lx, %lx, sCount=%ld)\n",
1505
                                           (long)alignWeightRight(s), (long)alignWeightRight(sLimit),
1506
                                           (long)alignWeightRight(sCount));
1507
#endif
1508
0
                                    return;
1509
0
                                }
1510
0
                                sIsTailored = TRUE;
1511
0
                            }
1512
0
                            s = secondaries.nextWeight();
1513
0
                            U_ASSERT(s != 0xffffffff);
1514
0
                        } else {
1515
0
                            s = weight16FromNode(node);
1516
0
                            sIsTailored = FALSE;
1517
#ifdef DEBUG_COLLATION_BUILDER
1518
                            printf("  sec       %lx\n", (long)alignWeightRight(s));
1519
#endif
1520
0
                        }
1521
0
                    } else /* UCOL_PRIMARY */ {
1522
0
                        U_ASSERT(isTailoredNode(node));
1523
#ifdef DEBUG_COLLATION_BUILDER
1524
                        printf("pri+            ");
1525
#endif
1526
0
                        if(!pIsTailored) {
1527
                            // First tailored primary node in this list.
1528
0
                            int32_t pCount = countTailoredNodes(nodesArray, nextIndex,
1529
0
                                                                UCOL_PRIMARY) + 1;
1530
0
                            UBool isCompressible = baseData->isCompressiblePrimary(p);
1531
0
                            uint32_t pLimit =
1532
0
                                rootElements.getPrimaryAfter(p, pIndex, isCompressible);
1533
0
                            primaries.initForPrimary(isCompressible);
1534
0
                            if(!primaries.allocWeights(p, pLimit, pCount)) {
1535
0
                                errorCode = U_BUFFER_OVERFLOW_ERROR;  // TODO: introduce a more specific UErrorCode?
1536
0
                                errorReason = "primary tailoring gap too small";
1537
0
                                return;
1538
0
                            }
1539
0
                            pIsTailored = TRUE;
1540
0
                        }
1541
0
                        p = primaries.nextWeight();
1542
0
                        U_ASSERT(p != 0xffffffff);
1543
0
                        s = Collation::COMMON_WEIGHT16;
1544
0
                        sIsTailored = FALSE;
1545
0
                    }
1546
0
                    t = s == 0 ? 0 : Collation::COMMON_WEIGHT16;
1547
0
                    tIsTailored = FALSE;
1548
0
                }
1549
0
                q = 0;
1550
0
            }
1551
0
            if(isTailoredNode(node)) {
1552
0
                nodesArray[i] = Collation::makeCE(p, s, t, q);
1553
#ifdef DEBUG_COLLATION_BUILDER
1554
                printf("%016llx\n", (long long)nodesArray[i]);
1555
#endif
1556
0
            }
1557
0
        }
1558
0
    }
1559
0
}
1560
1561
int32_t
1562
0
CollationBuilder::countTailoredNodes(const int64_t *nodesArray, int32_t i, int32_t strength) {
1563
0
    int32_t count = 0;
1564
0
    for(;;) {
1565
0
        if(i == 0) { break; }
1566
0
        int64_t node = nodesArray[i];
1567
0
        if(strengthFromNode(node) < strength) { break; }
1568
0
        if(strengthFromNode(node) == strength) {
1569
0
            if(isTailoredNode(node)) {
1570
0
                ++count;
1571
0
            } else {
1572
0
                break;
1573
0
            }
1574
0
        }
1575
0
        i = nextIndexFromNode(node);
1576
0
    }
1577
0
    return count;
1578
0
}
1579
1580
class CEFinalizer : public CollationDataBuilder::CEModifier {
1581
public:
1582
0
    CEFinalizer(const int64_t *ces) : finalCEs(ces) {}
1583
    virtual ~CEFinalizer();
1584
0
    virtual int64_t modifyCE32(uint32_t ce32) const {
1585
0
        U_ASSERT(!Collation::isSpecialCE32(ce32));
1586
0
        if(CollationBuilder::isTempCE32(ce32)) {
1587
            // retain case bits
1588
0
            return finalCEs[CollationBuilder::indexFromTempCE32(ce32)] | ((ce32 & 0xc0) << 8);
1589
0
        } else {
1590
0
            return Collation::NO_CE;
1591
0
        }
1592
0
    }
1593
0
    virtual int64_t modifyCE(int64_t ce) const {
1594
0
        if(CollationBuilder::isTempCE(ce)) {
1595
            // retain case bits
1596
0
            return finalCEs[CollationBuilder::indexFromTempCE(ce)] | (ce & 0xc000);
1597
0
        } else {
1598
0
            return Collation::NO_CE;
1599
0
        }
1600
0
    }
1601
1602
private:
1603
    const int64_t *finalCEs;
1604
};
1605
1606
0
CEFinalizer::~CEFinalizer() {}
1607
1608
void
1609
0
CollationBuilder::finalizeCEs(UErrorCode &errorCode) {
1610
0
    if(U_FAILURE(errorCode)) { return; }
1611
0
    LocalPointer<CollationDataBuilder> newBuilder(new CollationDataBuilder(errorCode), errorCode);
1612
0
    if(U_FAILURE(errorCode)) {
1613
0
        return;
1614
0
    }
1615
0
    newBuilder->initForTailoring(baseData, errorCode);
1616
0
    CEFinalizer finalizer(nodes.getBuffer());
1617
0
    newBuilder->copyFrom(*dataBuilder, finalizer, errorCode);
1618
0
    if(U_FAILURE(errorCode)) { return; }
1619
0
    delete dataBuilder;
1620
0
    dataBuilder = newBuilder.orphan();
1621
0
}
1622
1623
int32_t
1624
0
CollationBuilder::ceStrength(int64_t ce) {
1625
0
    return
1626
0
        isTempCE(ce) ? strengthFromTempCE(ce) :
1627
0
        (ce & INT64_C(0xff00000000000000)) != 0 ? UCOL_PRIMARY :
1628
0
        ((uint32_t)ce & 0xff000000) != 0 ? UCOL_SECONDARY :
1629
0
        ce != 0 ? UCOL_TERTIARY :
1630
0
        UCOL_IDENTICAL;
1631
0
}
1632
1633
U_NAMESPACE_END
1634
1635
U_NAMESPACE_USE
1636
1637
U_CAPI UCollator * U_EXPORT2
1638
ucol_openRules(const UChar *rules, int32_t rulesLength,
1639
               UColAttributeValue normalizationMode, UCollationStrength strength,
1640
0
               UParseError *parseError, UErrorCode *pErrorCode) {
1641
0
    if(U_FAILURE(*pErrorCode)) { return NULL; }
1642
0
    if(rules == NULL && rulesLength != 0) {
1643
0
        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
1644
0
        return NULL;
1645
0
    }
1646
0
    RuleBasedCollator *coll = new RuleBasedCollator();
1647
0
    if(coll == NULL) {
1648
0
        *pErrorCode = U_MEMORY_ALLOCATION_ERROR;
1649
0
        return NULL;
1650
0
    }
1651
0
    UnicodeString r((UBool)(rulesLength < 0), rules, rulesLength);
1652
0
    coll->internalBuildTailoring(r, strength, normalizationMode, parseError, NULL, *pErrorCode);
1653
0
    if(U_FAILURE(*pErrorCode)) {
1654
0
        delete coll;
1655
0
        return NULL;
1656
0
    }
1657
0
    return coll->toUCollator();
1658
0
}
1659
1660
static const int32_t internalBufferSize = 512;
1661
1662
// The @internal ucol_getUnsafeSet() was moved here from ucol_sit.cpp
1663
// because it calls UnicodeSet "builder" code that depends on all Unicode properties,
1664
// and the rest of the collation "runtime" code only depends on normalization.
1665
// This function is not related to the collation builder,
1666
// but it did not seem worth moving it into its own .cpp file,
1667
// nor rewriting it to use lower-level UnicodeSet and Normalizer2Impl methods.
1668
U_CAPI int32_t U_EXPORT2
1669
ucol_getUnsafeSet( const UCollator *coll,
1670
                  USet *unsafe,
1671
                  UErrorCode *status)
1672
0
{
1673
0
    UChar buffer[internalBufferSize];
1674
0
    int32_t len = 0;
1675
1676
0
    uset_clear(unsafe);
1677
1678
    // cccpattern = "[[:^tccc=0:][:^lccc=0:]]", unfortunately variant
1679
0
    static const UChar cccpattern[25] = { 0x5b, 0x5b, 0x3a, 0x5e, 0x74, 0x63, 0x63, 0x63, 0x3d, 0x30, 0x3a, 0x5d,
1680
0
                                    0x5b, 0x3a, 0x5e, 0x6c, 0x63, 0x63, 0x63, 0x3d, 0x30, 0x3a, 0x5d, 0x5d, 0x00 };
1681
1682
    // add chars that fail the fcd check
1683
0
    uset_applyPattern(unsafe, cccpattern, 24, USET_IGNORE_SPACE, status);
1684
1685
    // add lead/trail surrogates
1686
    // (trail surrogates should need to be unsafe only if the caller tests for UTF-16 code *units*,
1687
    // not when testing code *points*)
1688
0
    uset_addRange(unsafe, 0xd800, 0xdfff);
1689
1690
0
    USet *contractions = uset_open(0,0);
1691
1692
0
    int32_t i = 0, j = 0;
1693
0
    ucol_getContractionsAndExpansions(coll, contractions, NULL, FALSE, status);
1694
0
    int32_t contsSize = uset_size(contractions);
1695
0
    UChar32 c = 0;
1696
    // Contraction set consists only of strings
1697
    // to get unsafe code points, we need to
1698
    // break the strings apart and add them to the unsafe set
1699
0
    for(i = 0; i < contsSize; i++) {
1700
0
        len = uset_getItem(contractions, i, NULL, NULL, buffer, internalBufferSize, status);
1701
0
        if(len > 0) {
1702
0
            j = 0;
1703
0
            while(j < len) {
1704
0
                U16_NEXT(buffer, j, len, c);
1705
0
                if(j < len) {
1706
0
                    uset_add(unsafe, c);
1707
0
                }
1708
0
            }
1709
0
        }
1710
0
    }
1711
1712
0
    uset_close(contractions);
1713
1714
0
    return uset_size(unsafe);
1715
0
}
1716
1717
#endif  // !UCONFIG_NO_COLLATION