summaryrefslogtreecommitdiffstats
path: root/intl/icu/source/i18n/collationruleparser.cpp
blob: ea3ea66bc08cf5ea32bd38c91ba5dafb4a82405f (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
// Copyright (C) 2016 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html
/*
*******************************************************************************
* Copyright (C) 2013-2015, International Business Machines
* Corporation and others.  All Rights Reserved.
*******************************************************************************
* collationruleparser.cpp
*
* (replaced the former ucol_tok.cpp)
*
* created on: 2013apr10
* created by: Markus W. Scherer
*/

#include "unicode/utypes.h"

#if !UCONFIG_NO_COLLATION

#include "unicode/normalizer2.h"
#include "unicode/parseerr.h"
#include "unicode/uchar.h"
#include "unicode/ucol.h"
#include "unicode/uloc.h"
#include "unicode/unistr.h"
#include "unicode/utf16.h"
#include "charstr.h"
#include "cmemory.h"
#include "collation.h"
#include "collationdata.h"
#include "collationruleparser.h"
#include "collationsettings.h"
#include "collationtailoring.h"
#include "cstring.h"
#include "patternprops.h"
#include "uassert.h"
#include "uvectr32.h"

U_NAMESPACE_BEGIN

namespace {

static const UChar BEFORE[] = { 0x5b, 0x62, 0x65, 0x66, 0x6f, 0x72, 0x65, 0 };  // "[before"
const int32_t BEFORE_LENGTH = 7;

}  // namespace

CollationRuleParser::Sink::~Sink() {}

void
CollationRuleParser::Sink::suppressContractions(const UnicodeSet &, const char *&, UErrorCode &) {}

void
CollationRuleParser::Sink::optimize(const UnicodeSet &, const char *&, UErrorCode &) {}

CollationRuleParser::Importer::~Importer() {}

CollationRuleParser::CollationRuleParser(const CollationData *base, UErrorCode &errorCode)
        : nfd(*Normalizer2::getNFDInstance(errorCode)),
          nfc(*Normalizer2::getNFCInstance(errorCode)),
          rules(NULL), baseData(base), settings(NULL),
          parseError(NULL), errorReason(NULL),
          sink(NULL), importer(NULL),
          ruleIndex(0) {
}

CollationRuleParser::~CollationRuleParser() {
}

void
CollationRuleParser::parse(const UnicodeString &ruleString,
                           CollationSettings &outSettings,
                           UParseError *outParseError,
                           UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return; }
    settings = &outSettings;
    parseError = outParseError;
    if(parseError != NULL) {
        parseError->line = 0;
        parseError->offset = -1;
        parseError->preContext[0] = 0;
        parseError->postContext[0] = 0;
    }
    errorReason = NULL;
    parse(ruleString, errorCode);
}

void
CollationRuleParser::parse(const UnicodeString &ruleString, UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return; }
    rules = &ruleString;
    ruleIndex = 0;

    while(ruleIndex < rules->length()) {
        UChar c = rules->charAt(ruleIndex);
        if(PatternProps::isWhiteSpace(c)) {
            ++ruleIndex;
            continue;
        }
        switch(c) {
        case 0x26:  // '&'
            parseRuleChain(errorCode);
            break;
        case 0x5b:  // '['
            parseSetting(errorCode);
            break;
        case 0x23:  // '#' starts a comment, until the end of the line
            ruleIndex = skipComment(ruleIndex + 1);
            break;
        case 0x40:  // '@' is equivalent to [backwards 2]
            settings->setFlag(CollationSettings::BACKWARD_SECONDARY,
                              UCOL_ON, 0, errorCode);
            ++ruleIndex;
            break;
        case 0x21:  // '!' used to turn on Thai/Lao character reversal
            // Accept but ignore. The root collator has contractions
            // that are equivalent to the character reversal, where appropriate.
            ++ruleIndex;
            break;
        default:
            setParseError("expected a reset or setting or comment", errorCode);
            break;
        }
        if(U_FAILURE(errorCode)) { return; }
    }
}

void
CollationRuleParser::parseRuleChain(UErrorCode &errorCode) {
    int32_t resetStrength = parseResetAndPosition(errorCode);
    UBool isFirstRelation = TRUE;
    for(;;) {
        int32_t result = parseRelationOperator(errorCode);
        if(U_FAILURE(errorCode)) { return; }
        if(result < 0) {
            if(ruleIndex < rules->length() && rules->charAt(ruleIndex) == 0x23) {
                // '#' starts a comment, until the end of the line
                ruleIndex = skipComment(ruleIndex + 1);
                continue;
            }
            if(isFirstRelation) {
                setParseError("reset not followed by a relation", errorCode);
            }
            return;
        }
        int32_t strength = result & STRENGTH_MASK;
        if(resetStrength < UCOL_IDENTICAL) {
            // reset-before rule chain
            if(isFirstRelation) {
                if(strength != resetStrength) {
                    setParseError("reset-before strength differs from its first relation", errorCode);
                    return;
                }
            } else {
                if(strength < resetStrength) {
                    setParseError("reset-before strength followed by a stronger relation", errorCode);
                    return;
                }
            }
        }
        int32_t i = ruleIndex + (result >> OFFSET_SHIFT);  // skip over the relation operator
        if((result & STARRED_FLAG) == 0) {
            parseRelationStrings(strength, i, errorCode);
        } else {
            parseStarredCharacters(strength, i, errorCode);
        }
        if(U_FAILURE(errorCode)) { return; }
        isFirstRelation = FALSE;
    }
}

int32_t
CollationRuleParser::parseResetAndPosition(UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return UCOL_DEFAULT; }
    int32_t i = skipWhiteSpace(ruleIndex + 1);
    int32_t j;
    UChar c;
    int32_t resetStrength;
    if(rules->compare(i, BEFORE_LENGTH, BEFORE, 0, BEFORE_LENGTH) == 0 &&
            (j = i + BEFORE_LENGTH) < rules->length() &&
            PatternProps::isWhiteSpace(rules->charAt(j)) &&
            ((j = skipWhiteSpace(j + 1)) + 1) < rules->length() &&
            0x31 <= (c = rules->charAt(j)) && c <= 0x33 &&
            rules->charAt(j + 1) == 0x5d) {
        // &[before n] with n=1 or 2 or 3
        resetStrength = UCOL_PRIMARY + (c - 0x31);
        i = skipWhiteSpace(j + 2);
    } else {
        resetStrength = UCOL_IDENTICAL;
    }
    if(i >= rules->length()) {
        setParseError("reset without position", errorCode);
        return UCOL_DEFAULT;
    }
    UnicodeString str;
    if(rules->charAt(i) == 0x5b) {  // '['
        i = parseSpecialPosition(i, str, errorCode);
    } else {
        i = parseTailoringString(i, str, errorCode);
    }
    sink->addReset(resetStrength, str, errorReason, errorCode);
    if(U_FAILURE(errorCode)) { setErrorContext(); }
    ruleIndex = i;
    return resetStrength;
}

int32_t
CollationRuleParser::parseRelationOperator(UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return UCOL_DEFAULT; }
    ruleIndex = skipWhiteSpace(ruleIndex);
    if(ruleIndex >= rules->length()) { return UCOL_DEFAULT; }
    int32_t strength;
    int32_t i = ruleIndex;
    UChar c = rules->charAt(i++);
    switch(c) {
    case 0x3c:  // '<'
        if(i < rules->length() && rules->charAt(i) == 0x3c) {  // <<
            ++i;
            if(i < rules->length() && rules->charAt(i) == 0x3c) {  // <<<
                ++i;
                if(i < rules->length() && rules->charAt(i) == 0x3c) {  // <<<<
                    ++i;
                    strength = UCOL_QUATERNARY;
                } else {
                    strength = UCOL_TERTIARY;
                }
            } else {
                strength = UCOL_SECONDARY;
            }
        } else {
            strength = UCOL_PRIMARY;
        }
        if(i < rules->length() && rules->charAt(i) == 0x2a) {  // '*'
            ++i;
            strength |= STARRED_FLAG;
        }
        break;
    case 0x3b:  // ';' same as <<
        strength = UCOL_SECONDARY;
        break;
    case 0x2c:  // ',' same as <<<
        strength = UCOL_TERTIARY;
        break;
    case 0x3d:  // '='
        strength = UCOL_IDENTICAL;
        if(i < rules->length() && rules->charAt(i) == 0x2a) {  // '*'
            ++i;
            strength |= STARRED_FLAG;
        }
        break;
    default:
        return UCOL_DEFAULT;
    }
    return ((i - ruleIndex) << OFFSET_SHIFT) | strength;
}

void
CollationRuleParser::parseRelationStrings(int32_t strength, int32_t i, UErrorCode &errorCode) {
    // Parse
    //     prefix | str / extension
    // where prefix and extension are optional.
    UnicodeString prefix, str, extension;
    i = parseTailoringString(i, str, errorCode);
    if(U_FAILURE(errorCode)) { return; }
    UChar next = (i < rules->length()) ? rules->charAt(i) : 0;
    if(next == 0x7c) {  // '|' separates the context prefix from the string.
        prefix = str;
        i = parseTailoringString(i + 1, str, errorCode);
        if(U_FAILURE(errorCode)) { return; }
        next = (i < rules->length()) ? rules->charAt(i) : 0;
    }
    if(next == 0x2f) {  // '/' separates the string from the extension.
        i = parseTailoringString(i + 1, extension, errorCode);
    }
    if(!prefix.isEmpty()) {
        UChar32 prefix0 = prefix.char32At(0);
        UChar32 c = str.char32At(0);
        if(!nfc.hasBoundaryBefore(prefix0) || !nfc.hasBoundaryBefore(c)) {
            setParseError("in 'prefix|str', prefix and str must each start with an NFC boundary",
                          errorCode);
            return;
        }
    }
    sink->addRelation(strength, prefix, str, extension, errorReason, errorCode);
    if(U_FAILURE(errorCode)) { setErrorContext(); }
    ruleIndex = i;
}

void
CollationRuleParser::parseStarredCharacters(int32_t strength, int32_t i, UErrorCode &errorCode) {
    UnicodeString empty, raw;
    i = parseString(skipWhiteSpace(i), raw, errorCode);
    if(U_FAILURE(errorCode)) { return; }
    if(raw.isEmpty()) {
        setParseError("missing starred-relation string", errorCode);
        return;
    }
    UChar32 prev = -1;
    int32_t j = 0;
    for(;;) {
        while(j < raw.length()) {
            UChar32 c = raw.char32At(j);
            if(!nfd.isInert(c)) {
                setParseError("starred-relation string is not all NFD-inert", errorCode);
                return;
            }
            sink->addRelation(strength, empty, UnicodeString(c), empty, errorReason, errorCode);
            if(U_FAILURE(errorCode)) {
                setErrorContext();
                return;
            }
            j += U16_LENGTH(c);
            prev = c;
        }
        if(i >= rules->length() || rules->charAt(i) != 0x2d) {  // '-'
            break;
        }
        if(prev < 0) {
            setParseError("range without start in starred-relation string", errorCode);
            return;
        }
        i = parseString(i + 1, raw, errorCode);
        if(U_FAILURE(errorCode)) { return; }
        if(raw.isEmpty()) {
            setParseError("range without end in starred-relation string", errorCode);
            return;
        }
        UChar32 c = raw.char32At(0);
        if(c < prev) {
            setParseError("range start greater than end in starred-relation string", errorCode);
            return;
        }
        // range prev-c
        UnicodeString s;
        while(++prev <= c) {
            if(!nfd.isInert(prev)) {
                setParseError("starred-relation string range is not all NFD-inert", errorCode);
                return;
            }
            if(U_IS_SURROGATE(prev)) {
                setParseError("starred-relation string range contains a surrogate", errorCode);
                return;
            }
            if(0xfffd <= prev && prev <= 0xffff) {
                setParseError("starred-relation string range contains U+FFFD, U+FFFE or U+FFFF", errorCode);
                return;
            }
            s.setTo(prev);
            sink->addRelation(strength, empty, s, empty, errorReason, errorCode);
            if(U_FAILURE(errorCode)) {
                setErrorContext();
                return;
            }
        }
        prev = -1;
        j = U16_LENGTH(c);
    }
    ruleIndex = skipWhiteSpace(i);
}

int32_t
CollationRuleParser::parseTailoringString(int32_t i, UnicodeString &raw, UErrorCode &errorCode) {
    i = parseString(skipWhiteSpace(i), raw, errorCode);
    if(U_SUCCESS(errorCode) && raw.isEmpty()) {
        setParseError("missing relation string", errorCode);
    }
    return skipWhiteSpace(i);
}

int32_t
CollationRuleParser::parseString(int32_t i, UnicodeString &raw, UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return i; }
    raw.remove();
    while(i < rules->length()) {
        UChar32 c = rules->charAt(i++);
        if(isSyntaxChar(c)) {
            if(c == 0x27) {  // apostrophe
                if(i < rules->length() && rules->charAt(i) == 0x27) {
                    // Double apostrophe, encodes a single one.
                    raw.append((UChar)0x27);
                    ++i;
                    continue;
                }
                // Quote literal text until the next single apostrophe.
                for(;;) {
                    if(i == rules->length()) {
                        setParseError("quoted literal text missing terminating apostrophe", errorCode);
                        return i;
                    }
                    c = rules->charAt(i++);
                    if(c == 0x27) {
                        if(i < rules->length() && rules->charAt(i) == 0x27) {
                            // Double apostrophe inside quoted literal text,
                            // still encodes a single apostrophe.
                            ++i;
                        } else {
                            break;
                        }
                    }
                    raw.append((UChar)c);
                }
            } else if(c == 0x5c) {  // backslash
                if(i == rules->length()) {
                    setParseError("backslash escape at the end of the rule string", errorCode);
                    return i;
                }
                c = rules->char32At(i);
                raw.append(c);
                i += U16_LENGTH(c);
            } else {
                // Any other syntax character terminates a string.
                --i;
                break;
            }
        } else if(PatternProps::isWhiteSpace(c)) {
            // Unquoted white space terminates a string.
            --i;
            break;
        } else {
            raw.append((UChar)c);
        }
    }
    for(int32_t j = 0; j < raw.length();) {
        UChar32 c = raw.char32At(j);
        if(U_IS_SURROGATE(c)) {
            setParseError("string contains an unpaired surrogate", errorCode);
            return i;
        }
        if(0xfffd <= c && c <= 0xffff) {
            setParseError("string contains U+FFFD, U+FFFE or U+FFFF", errorCode);
            return i;
        }
        j += U16_LENGTH(c);
    }
    return i;
}

namespace {

static const char *const positions[] = {
    "first tertiary ignorable",
    "last tertiary ignorable",
    "first secondary ignorable",
    "last secondary ignorable",
    "first primary ignorable",
    "last primary ignorable",
    "first variable",
    "last variable",
    "first regular",
    "last regular",
    "first implicit",
    "last implicit",
    "first trailing",
    "last trailing"
};

}  // namespace

int32_t
CollationRuleParser::parseSpecialPosition(int32_t i, UnicodeString &str, UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return 0; }
    UnicodeString raw;
    int32_t j = readWords(i + 1, raw);
    if(j > i && rules->charAt(j) == 0x5d && !raw.isEmpty()) {  // words end with ]
        ++j;
        for(int32_t pos = 0; pos < UPRV_LENGTHOF(positions); ++pos) {
            if(raw == UnicodeString(positions[pos], -1, US_INV)) {
                str.setTo((UChar)POS_LEAD).append((UChar)(POS_BASE + pos));
                return j;
            }
        }
        if(raw == UNICODE_STRING_SIMPLE("top")) {
            str.setTo((UChar)POS_LEAD).append((UChar)(POS_BASE + LAST_REGULAR));
            return j;
        }
        if(raw == UNICODE_STRING_SIMPLE("variable top")) {
            str.setTo((UChar)POS_LEAD).append((UChar)(POS_BASE + LAST_VARIABLE));
            return j;
        }
    }
    setParseError("not a valid special reset position", errorCode);
    return i;
}

void
CollationRuleParser::parseSetting(UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return; }
    UnicodeString raw;
    int32_t i = ruleIndex + 1;
    int32_t j = readWords(i, raw);
    if(j <= i || raw.isEmpty()) {
        setParseError("expected a setting/option at '['", errorCode);
    }
    if(rules->charAt(j) == 0x5d) {  // words end with ]
        ++j;
        if(raw.startsWith(UNICODE_STRING_SIMPLE("reorder")) &&
                (raw.length() == 7 || raw.charAt(7) == 0x20)) {
            parseReordering(raw, errorCode);
            ruleIndex = j;
            return;
        }
        if(raw == UNICODE_STRING_SIMPLE("backwards 2")) {
            settings->setFlag(CollationSettings::BACKWARD_SECONDARY,
                              UCOL_ON, 0, errorCode);
            ruleIndex = j;
            return;
        }
        UnicodeString v;
        int32_t valueIndex = raw.lastIndexOf((UChar)0x20);
        if(valueIndex >= 0) {
            v.setTo(raw, valueIndex + 1);
            raw.truncate(valueIndex);
        }
        if(raw == UNICODE_STRING_SIMPLE("strength") && v.length() == 1) {
            int32_t value = UCOL_DEFAULT;
            UChar c = v.charAt(0);
            if(0x31 <= c && c <= 0x34) {  // 1..4
                value = UCOL_PRIMARY + (c - 0x31);
            } else if(c == 0x49) {  // 'I'
                value = UCOL_IDENTICAL;
            }
            if(value != UCOL_DEFAULT) {
                settings->setStrength(value, 0, errorCode);
                ruleIndex = j;
                return;
            }
        } else if(raw == UNICODE_STRING_SIMPLE("alternate")) {
            UColAttributeValue value = UCOL_DEFAULT;
            if(v == UNICODE_STRING_SIMPLE("non-ignorable")) {
                value = UCOL_NON_IGNORABLE;
            } else if(v == UNICODE_STRING_SIMPLE("shifted")) {
                value = UCOL_SHIFTED;
            }
            if(value != UCOL_DEFAULT) {
                settings->setAlternateHandling(value, 0, errorCode);
                ruleIndex = j;
                return;
            }
        } else if(raw == UNICODE_STRING_SIMPLE("maxVariable")) {
            int32_t value = UCOL_DEFAULT;
            if(v == UNICODE_STRING_SIMPLE("space")) {
                value = CollationSettings::MAX_VAR_SPACE;
            } else if(v == UNICODE_STRING_SIMPLE("punct")) {
                value = CollationSettings::MAX_VAR_PUNCT;
            } else if(v == UNICODE_STRING_SIMPLE("symbol")) {
                value = CollationSettings::MAX_VAR_SYMBOL;
            } else if(v == UNICODE_STRING_SIMPLE("currency")) {
                value = CollationSettings::MAX_VAR_CURRENCY;
            }
            if(value != UCOL_DEFAULT) {
                settings->setMaxVariable(value, 0, errorCode);
                settings->variableTop = baseData->getLastPrimaryForGroup(
                    UCOL_REORDER_CODE_FIRST + value);
                U_ASSERT(settings->variableTop != 0);
                ruleIndex = j;
                return;
            }
        } else if(raw == UNICODE_STRING_SIMPLE("caseFirst")) {
            UColAttributeValue value = UCOL_DEFAULT;
            if(v == UNICODE_STRING_SIMPLE("off")) {
                value = UCOL_OFF;
            } else if(v == UNICODE_STRING_SIMPLE("lower")) {
                value = UCOL_LOWER_FIRST;
            } else if(v == UNICODE_STRING_SIMPLE("upper")) {
                value = UCOL_UPPER_FIRST;
            }
            if(value != UCOL_DEFAULT) {
                settings->setCaseFirst(value, 0, errorCode);
                ruleIndex = j;
                return;
            }
        } else if(raw == UNICODE_STRING_SIMPLE("caseLevel")) {
            UColAttributeValue value = getOnOffValue(v);
            if(value != UCOL_DEFAULT) {
                settings->setFlag(CollationSettings::CASE_LEVEL, value, 0, errorCode);
                ruleIndex = j;
                return;
            }
        } else if(raw == UNICODE_STRING_SIMPLE("normalization")) {
            UColAttributeValue value = getOnOffValue(v);
            if(value != UCOL_DEFAULT) {
                settings->setFlag(CollationSettings::CHECK_FCD, value, 0, errorCode);
                ruleIndex = j;
                return;
            }
        } else if(raw == UNICODE_STRING_SIMPLE("numericOrdering")) {
            UColAttributeValue value = getOnOffValue(v);
            if(value != UCOL_DEFAULT) {
                settings->setFlag(CollationSettings::NUMERIC, value, 0, errorCode);
                ruleIndex = j;
                return;
            }
        } else if(raw == UNICODE_STRING_SIMPLE("hiraganaQ")) {
            UColAttributeValue value = getOnOffValue(v);
            if(value != UCOL_DEFAULT) {
                if(value == UCOL_ON) {
                    setParseError("[hiraganaQ on] is not supported", errorCode);
                }
                ruleIndex = j;
                return;
            }
        } else if(raw == UNICODE_STRING_SIMPLE("import")) {
            CharString lang;
            lang.appendInvariantChars(v, errorCode);
            if(errorCode == U_MEMORY_ALLOCATION_ERROR) { return; }
            // BCP 47 language tag -> ICU locale ID
            char localeID[ULOC_FULLNAME_CAPACITY];
            int32_t parsedLength;
            int32_t length = uloc_forLanguageTag(lang.data(), localeID, ULOC_FULLNAME_CAPACITY,
                                                 &parsedLength, &errorCode);
            if(U_FAILURE(errorCode) ||
                    parsedLength != lang.length() || length >= ULOC_FULLNAME_CAPACITY) {
                errorCode = U_ZERO_ERROR;
                setParseError("expected language tag in [import langTag]", errorCode);
                return;
            }
            // localeID minus all keywords
            char baseID[ULOC_FULLNAME_CAPACITY];
            length = uloc_getBaseName(localeID, baseID, ULOC_FULLNAME_CAPACITY, &errorCode);
            if(U_FAILURE(errorCode) || length >= ULOC_KEYWORDS_CAPACITY) {
                errorCode = U_ZERO_ERROR;
                setParseError("expected language tag in [import langTag]", errorCode);
                return;
            }
            if(length == 3 && uprv_memcmp(baseID, "und", 3) == 0) {
                uprv_strcpy(baseID, "root");
            }
            // @collation=type, or length=0 if not specified
            char collationType[ULOC_KEYWORDS_CAPACITY];
            length = uloc_getKeywordValue(localeID, "collation",
                                          collationType, ULOC_KEYWORDS_CAPACITY,
                                          &errorCode);
            if(U_FAILURE(errorCode) || length >= ULOC_KEYWORDS_CAPACITY) {
                errorCode = U_ZERO_ERROR;
                setParseError("expected language tag in [import langTag]", errorCode);
                return;
            }
            if(importer == NULL) {
                setParseError("[import langTag] is not supported", errorCode);
            } else {
                UnicodeString importedRules;
                importer->getRules(baseID, length > 0 ? collationType : "standard",
                                   importedRules, errorReason, errorCode);
                if(U_FAILURE(errorCode)) {
                    if(errorReason == NULL) {
                        errorReason = "[import langTag] failed";
                    }
                    setErrorContext();
                    return;
                }
                const UnicodeString *outerRules = rules;
                int32_t outerRuleIndex = ruleIndex;
                parse(importedRules, errorCode);
                if(U_FAILURE(errorCode)) {
                    if(parseError != NULL) {
                        parseError->offset = outerRuleIndex;
                    }
                }
                rules = outerRules;
                ruleIndex = j;
            }
            return;
        }
    } else if(rules->charAt(j) == 0x5b) {  // words end with [
        UnicodeSet set;
        j = parseUnicodeSet(j, set, errorCode);
        if(U_FAILURE(errorCode)) { return; }
        if(raw == UNICODE_STRING_SIMPLE("optimize")) {
            sink->optimize(set, errorReason, errorCode);
            if(U_FAILURE(errorCode)) { setErrorContext(); }
            ruleIndex = j;
            return;
        } else if(raw == UNICODE_STRING_SIMPLE("suppressContractions")) {
            sink->suppressContractions(set, errorReason, errorCode);
            if(U_FAILURE(errorCode)) { setErrorContext(); }
            ruleIndex = j;
            return;
        }
    }
    setParseError("not a valid setting/option", errorCode);
}

void
CollationRuleParser::parseReordering(const UnicodeString &raw, UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return; }
    int32_t i = 7;  // after "reorder"
    if(i == raw.length()) {
        // empty [reorder] with no codes
        settings->resetReordering();
        return;
    }
    // Parse the codes in [reorder aa bb cc].
    UVector32 reorderCodes(errorCode);
    if(U_FAILURE(errorCode)) { return; }
    CharString word;
    while(i < raw.length()) {
        ++i;  // skip the word-separating space
        int32_t limit = raw.indexOf((UChar)0x20, i);
        if(limit < 0) { limit = raw.length(); }
        word.clear().appendInvariantChars(raw.tempSubStringBetween(i, limit), errorCode);
        if(U_FAILURE(errorCode)) { return; }
        int32_t code = getReorderCode(word.data());
        if(code < 0) {
            setParseError("unknown script or reorder code", errorCode);
            return;
        }
        reorderCodes.addElement(code, errorCode);
        if(U_FAILURE(errorCode)) { return; }
        i = limit;
    }
    settings->setReordering(*baseData, reorderCodes.getBuffer(), reorderCodes.size(), errorCode);
}

static const char *const gSpecialReorderCodes[] = {
    "space", "punct", "symbol", "currency", "digit"
};

int32_t
CollationRuleParser::getReorderCode(const char *word) {
    for(int32_t i = 0; i < UPRV_LENGTHOF(gSpecialReorderCodes); ++i) {
        if(uprv_stricmp(word, gSpecialReorderCodes[i]) == 0) {
            return UCOL_REORDER_CODE_FIRST + i;
        }
    }
    int32_t script = u_getPropertyValueEnum(UCHAR_SCRIPT, word);
    if(script >= 0) {
        return script;
    }
    if(uprv_stricmp(word, "others") == 0) {
        return UCOL_REORDER_CODE_OTHERS;  // same as Zzzz = USCRIPT_UNKNOWN
    }
    return -1;
}

UColAttributeValue
CollationRuleParser::getOnOffValue(const UnicodeString &s) {
    if(s == UNICODE_STRING_SIMPLE("on")) {
        return UCOL_ON;
    } else if(s == UNICODE_STRING_SIMPLE("off")) {
        return UCOL_OFF;
    } else {
        return UCOL_DEFAULT;
    }
}

int32_t
CollationRuleParser::parseUnicodeSet(int32_t i, UnicodeSet &set, UErrorCode &errorCode) {
    // Collect a UnicodeSet pattern between a balanced pair of [brackets].
    int32_t level = 0;
    int32_t j = i;
    for(;;) {
        if(j == rules->length()) {
            setParseError("unbalanced UnicodeSet pattern brackets", errorCode);
            return j;
        }
        UChar c = rules->charAt(j++);
        if(c == 0x5b) {  // '['
            ++level;
        } else if(c == 0x5d) {  // ']'
            if(--level == 0) { break; }
        }
    }
    set.applyPattern(rules->tempSubStringBetween(i, j), errorCode);
    if(U_FAILURE(errorCode)) {
        errorCode = U_ZERO_ERROR;
        setParseError("not a valid UnicodeSet pattern", errorCode);
        return j;
    }
    j = skipWhiteSpace(j);
    if(j == rules->length() || rules->charAt(j) != 0x5d) {
        setParseError("missing option-terminating ']' after UnicodeSet pattern", errorCode);
        return j;
    }
    return ++j;
}

int32_t
CollationRuleParser::readWords(int32_t i, UnicodeString &raw) const {
    static const UChar sp = 0x20;
    raw.remove();
    i = skipWhiteSpace(i);
    for(;;) {
        if(i >= rules->length()) { return 0; }
        UChar c = rules->charAt(i);
        if(isSyntaxChar(c) && c != 0x2d && c != 0x5f) {  // syntax except -_
            if(raw.isEmpty()) { return i; }
            if(raw.endsWith(&sp, 1)) {  // remove trailing space
                raw.truncate(raw.length() - 1);
            }
            return i;
        }
        if(PatternProps::isWhiteSpace(c)) {
            raw.append(sp);
            i = skipWhiteSpace(i + 1);
        } else {
            raw.append(c);
            ++i;
        }
    }
}

int32_t
CollationRuleParser::skipComment(int32_t i) const {
    // skip to past the newline
    while(i < rules->length()) {
        UChar c = rules->charAt(i++);
        // LF or FF or CR or NEL or LS or PS
        if(c == 0xa || c == 0xc || c == 0xd || c == 0x85 || c == 0x2028 || c == 0x2029) {
            // Unicode Newline Guidelines: "A readline function should stop at NLF, LS, FF, or PS."
            // NLF (new line function) = CR or LF or CR+LF or NEL.
            // No need to collect all of CR+LF because a following LF will be ignored anyway.
            break;
        }
    }
    return i;
}

void
CollationRuleParser::setParseError(const char *reason, UErrorCode &errorCode) {
    if(U_FAILURE(errorCode)) { return; }
    // Error code consistent with the old parser (from ca. 2001),
    // rather than U_PARSE_ERROR;
    errorCode = U_INVALID_FORMAT_ERROR;
    errorReason = reason;
    if(parseError != NULL) { setErrorContext(); }
}

void
CollationRuleParser::setErrorContext() {
    if(parseError == NULL) { return; }

    // Note: This relies on the calling code maintaining the ruleIndex
    // at a position that is useful for debugging.
    // For example, at the beginning of a reset or relation etc.
    parseError->offset = ruleIndex;
    parseError->line = 0;  // We are not counting line numbers.

    // before ruleIndex
    int32_t start = ruleIndex - (U_PARSE_CONTEXT_LEN - 1);
    if(start < 0) {
        start = 0;
    } else if(start > 0 && U16_IS_TRAIL(rules->charAt(start))) {
        ++start;
    }
    int32_t length = ruleIndex - start;
    rules->extract(start, length, parseError->preContext);
    parseError->preContext[length] = 0;

    // starting from ruleIndex
    length = rules->length() - ruleIndex;
    if(length >= U_PARSE_CONTEXT_LEN) {
        length = U_PARSE_CONTEXT_LEN - 1;
        if(U16_IS_LEAD(rules->charAt(ruleIndex + length - 1))) {
            --length;
        }
    }
    rules->extract(ruleIndex, length, parseError->postContext);
    parseError->postContext[length] = 0;
}

UBool
CollationRuleParser::isSyntaxChar(UChar32 c) {
    return 0x21 <= c && c <= 0x7e &&
            (c <= 0x2f || (0x3a <= c && c <= 0x40) ||
            (0x5b <= c && c <= 0x60) || (0x7b <= c));
}

int32_t
CollationRuleParser::skipWhiteSpace(int32_t i) const {
    while(i < rules->length() && PatternProps::isWhiteSpace(rules->charAt(i))) {
        ++i;
    }
    return i;
}

U_NAMESPACE_END

#endif  // !UCONFIG_NO_COLLATION