summaryrefslogtreecommitdiffstats
path: root/intl/icu/source/i18n/collationdata.cpp
blob: 1b8b6a76de39cc1394c66e295a72c018285b25e2 (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
// © 2016 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html
/*
*******************************************************************************
* Copyright (C) 2012-2015, International Business Machines
* Corporation and others.  All Rights Reserved.
*******************************************************************************
* collationdata.cpp
*
* created on: 2012jul28
* created by: Markus W. Scherer
*/

#include "unicode/utypes.h"

#if !UCONFIG_NO_COLLATION

#include "unicode/ucol.h"
#include "unicode/udata.h"
#include "unicode/uscript.h"
#include "cmemory.h"
#include "collation.h"
#include "collationdata.h"
#include "uassert.h"
#include "utrie2.h"
#include "uvectr32.h"

U_NAMESPACE_BEGIN

uint32_t
CollationData::getIndirectCE32(uint32_t ce32) const {
    U_ASSERT(Collation::isSpecialCE32(ce32));
    int32_t tag = Collation::tagFromCE32(ce32);
    if(tag == Collation::DIGIT_TAG) {
        // Fetch the non-numeric-collation CE32.
        ce32 = ce32s[Collation::indexFromCE32(ce32)];
    } else if(tag == Collation::LEAD_SURROGATE_TAG) {
        ce32 = Collation::UNASSIGNED_CE32;
    } else if(tag == Collation::U0000_TAG) {
        // Fetch the normal ce32 for U+0000.
        ce32 = ce32s[0];
    }
    return ce32;
}

uint32_t
CollationData::getFinalCE32(uint32_t ce32) const {
    if(Collation::isSpecialCE32(ce32)) {
        ce32 = getIndirectCE32(ce32);
    }
    return ce32;
}

int64_t
CollationData::getSingleCE(UChar32 c, UErrorCode &errorCode) const {
    if(U_FAILURE(errorCode)) { return 0; }
    // Keep parallel with CollationDataBuilder::getSingleCE().
    const CollationData *d;
    uint32_t ce32 = getCE32(c);
    if(ce32 == Collation::FALLBACK_CE32) {
        d = base;
        ce32 = base->getCE32(c);
    } else {
        d = this;
    }
    while(Collation::isSpecialCE32(ce32)) {
        switch(Collation::tagFromCE32(ce32)) {
        case Collation::LATIN_EXPANSION_TAG:
        case Collation::BUILDER_DATA_TAG:
        case Collation::PREFIX_TAG:
        case Collation::CONTRACTION_TAG:
        case Collation::HANGUL_TAG:
        case Collation::LEAD_SURROGATE_TAG:
            errorCode = U_UNSUPPORTED_ERROR;
            return 0;
        case Collation::FALLBACK_TAG:
        case Collation::RESERVED_TAG_3:
            errorCode = U_INTERNAL_PROGRAM_ERROR;
            return 0;
        case Collation::LONG_PRIMARY_TAG:
            return Collation::ceFromLongPrimaryCE32(ce32);
        case Collation::LONG_SECONDARY_TAG:
            return Collation::ceFromLongSecondaryCE32(ce32);
        case Collation::EXPANSION32_TAG:
            if(Collation::lengthFromCE32(ce32) == 1) {
                ce32 = d->ce32s[Collation::indexFromCE32(ce32)];
                break;
            } else {
                errorCode = U_UNSUPPORTED_ERROR;
                return 0;
            }
        case Collation::EXPANSION_TAG: {
            if(Collation::lengthFromCE32(ce32) == 1) {
                return d->ces[Collation::indexFromCE32(ce32)];
            } else {
                errorCode = U_UNSUPPORTED_ERROR;
                return 0;
            }
        }
        case Collation::DIGIT_TAG:
            // Fetch the non-numeric-collation CE32 and continue.
            ce32 = d->ce32s[Collation::indexFromCE32(ce32)];
            break;
        case Collation::U0000_TAG:
            U_ASSERT(c == 0);
            // Fetch the normal ce32 for U+0000 and continue.
            ce32 = d->ce32s[0];
            break;
        case Collation::OFFSET_TAG:
            return d->getCEFromOffsetCE32(c, ce32);
        case Collation::IMPLICIT_TAG:
            return Collation::unassignedCEFromCodePoint(c);
        }
    }
    return Collation::ceFromSimpleCE32(ce32);
}

uint32_t
CollationData::getFirstPrimaryForGroup(int32_t script) const {
    int32_t index = getScriptIndex(script);
    return index == 0 ? 0 : (uint32_t)scriptStarts[index] << 16;
}

uint32_t
CollationData::getLastPrimaryForGroup(int32_t script) const {
    int32_t index = getScriptIndex(script);
    if(index == 0) {
        return 0;
    }
    uint32_t limit = scriptStarts[index + 1];
    return (limit << 16) - 1;
}

int32_t
CollationData::getGroupForPrimary(uint32_t p) const {
    p >>= 16;
    if(p < scriptStarts[1] || scriptStarts[scriptStartsLength - 1] <= p) {
        return -1;
    }
    int32_t index = 1;
    while(p >= scriptStarts[index + 1]) { ++index; }
    for(int32_t i = 0; i < numScripts; ++i) {
        if(scriptsIndex[i] == index) {
            return i;
        }
    }
    for(int32_t i = 0; i < MAX_NUM_SPECIAL_REORDER_CODES; ++i) {
        if(scriptsIndex[numScripts + i] == index) {
            return UCOL_REORDER_CODE_FIRST + i;
        }
    }
    return -1;
}

int32_t
CollationData::getScriptIndex(int32_t script) const {
    if(script < 0) {
        return 0;
    } else if(script < numScripts) {
        return scriptsIndex[script];
    } else if(script < UCOL_REORDER_CODE_FIRST) {
        return 0;
    } else {
        script -= UCOL_REORDER_CODE_FIRST;
        if(script < MAX_NUM_SPECIAL_REORDER_CODES) {
            return scriptsIndex[numScripts + script];
        } else {
            return 0;
        }
    }
}

int32_t
CollationData::getEquivalentScripts(int32_t script,
                                    int32_t dest[], int32_t capacity,
                                    UErrorCode &errorCode) const {
    if(U_FAILURE(errorCode)) { return 0; }
    int32_t index = getScriptIndex(script);
    if(index == 0) { return 0; }
    if(script >= UCOL_REORDER_CODE_FIRST) {
        // Special groups have no aliases.
        if(capacity > 0) {
            dest[0] = script;
        } else {
            errorCode = U_BUFFER_OVERFLOW_ERROR;
        }
        return 1;
    }

    int32_t length = 0;
    for(int32_t i = 0; i < numScripts; ++i) {
        if(scriptsIndex[i] == index) {
            if(length < capacity) {
                dest[length] = i;
            }
            ++length;
        }
    }
    if(length > capacity) {
        errorCode = U_BUFFER_OVERFLOW_ERROR;
    }
    return length;
}

void
CollationData::makeReorderRanges(const int32_t *reorder, int32_t length,
                                 UVector32 &ranges, UErrorCode &errorCode) const {
    makeReorderRanges(reorder, length, false, ranges, errorCode);
}

void
CollationData::makeReorderRanges(const int32_t *reorder, int32_t length,
                                 UBool latinMustMove,
                                 UVector32 &ranges, UErrorCode &errorCode) const {
    if(U_FAILURE(errorCode)) { return; }
    ranges.removeAllElements();
    if(length == 0 || (length == 1 && reorder[0] == USCRIPT_UNKNOWN)) {
        return;
    }

    // Maps each script-or-group range to a new lead byte.
    uint8_t table[MAX_NUM_SCRIPT_RANGES];
    uprv_memset(table, 0, sizeof(table));

    {
        // Set "don't care" values for reserved ranges.
        int32_t index = scriptsIndex[
                numScripts + REORDER_RESERVED_BEFORE_LATIN - UCOL_REORDER_CODE_FIRST];
        if(index != 0) {
            table[index] = 0xff;
        }
        index = scriptsIndex[
                numScripts + REORDER_RESERVED_AFTER_LATIN - UCOL_REORDER_CODE_FIRST];
        if(index != 0) {
            table[index] = 0xff;
        }
    }

    // Never reorder special low and high primary lead bytes.
    U_ASSERT(scriptStartsLength >= 2);
    U_ASSERT(scriptStarts[0] == 0);
    int32_t lowStart = scriptStarts[1];
    U_ASSERT(lowStart == ((Collation::MERGE_SEPARATOR_BYTE + 1) << 8));
    int32_t highLimit = scriptStarts[scriptStartsLength - 1];
    U_ASSERT(highLimit == (Collation::TRAIL_WEIGHT_BYTE << 8));

    // Get the set of special reorder codes in the input list.
    // This supports a fixed number of special reorder codes;
    // it works for data with codes beyond UCOL_REORDER_CODE_LIMIT.
    uint32_t specials = 0;
    for(int32_t i = 0; i < length; ++i) {
        int32_t reorderCode = reorder[i] - UCOL_REORDER_CODE_FIRST;
        if(0 <= reorderCode && reorderCode < MAX_NUM_SPECIAL_REORDER_CODES) {
            specials |= (uint32_t)1 << reorderCode;
        }
    }

    // Start the reordering with the special low reorder codes that do not occur in the input.
    for(int32_t i = 0; i < MAX_NUM_SPECIAL_REORDER_CODES; ++i) {
        int32_t index = scriptsIndex[numScripts + i];
        if(index != 0 && (specials & ((uint32_t)1 << i)) == 0) {
            lowStart = addLowScriptRange(table, index, lowStart);
        }
    }

    // Skip the reserved range before Latin if Latin is the first script,
    // so that we do not move it unnecessarily.
    int32_t skippedReserved = 0;
    if(specials == 0 && reorder[0] == USCRIPT_LATIN && !latinMustMove) {
        int32_t index = scriptsIndex[USCRIPT_LATIN];
        U_ASSERT(index != 0);
        int32_t start = scriptStarts[index];
        U_ASSERT(lowStart <= start);
        skippedReserved = start - lowStart;
        lowStart = start;
    }

    // Reorder according to the input scripts, continuing from the bottom of the primary range.
    int32_t originalLength = length;  // length will be decremented if "others" is in the list.
    UBool hasReorderToEnd = false;
    for(int32_t i = 0; i < length;) {
        int32_t script = reorder[i++];
        if(script == USCRIPT_UNKNOWN) {
            // Put the remaining scripts at the top.
            hasReorderToEnd = true;
            while(i < length) {
                script = reorder[--length];
                if(script == USCRIPT_UNKNOWN ||  // Must occur at most once.
                        script == UCOL_REORDER_CODE_DEFAULT) {
                    errorCode = U_ILLEGAL_ARGUMENT_ERROR;
                    return;
                }
                int32_t index = getScriptIndex(script);
                if(index == 0) { continue; }
                if(table[index] != 0) {  // Duplicate or equivalent script.
                    errorCode = U_ILLEGAL_ARGUMENT_ERROR;
                    return;
                }
                highLimit = addHighScriptRange(table, index, highLimit);
            }
            break;
        }
        if(script == UCOL_REORDER_CODE_DEFAULT) {
            // The default code must be the only one in the list, and that is handled by the caller.
            // Otherwise it must not be used.
            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
            return;
        }
        int32_t index = getScriptIndex(script);
        if(index == 0) { continue; }
        if(table[index] != 0) {  // Duplicate or equivalent script.
            errorCode = U_ILLEGAL_ARGUMENT_ERROR;
            return;
        }
        lowStart = addLowScriptRange(table, index, lowStart);
    }

    // Put all remaining scripts into the middle.
    for(int32_t i = 1; i < scriptStartsLength - 1; ++i) {
        int32_t leadByte = table[i];
        if(leadByte != 0) { continue; }
        int32_t start = scriptStarts[i];
        if(!hasReorderToEnd && start > lowStart) {
            // No need to move this script.
            lowStart = start;
        }
        lowStart = addLowScriptRange(table, i, lowStart);
    }
    if(lowStart > highLimit) {
        if((lowStart - (skippedReserved & 0xff00)) <= highLimit) {
            // Try not skipping the before-Latin reserved range.
            makeReorderRanges(reorder, originalLength, true, ranges, errorCode);
            return;
        }
        // We need more primary lead bytes than available, despite the reserved ranges.
        errorCode = U_BUFFER_OVERFLOW_ERROR;
        return;
    }

    // Turn lead bytes into a list of (limit, offset) pairs.
    // Encode each pair in one list element:
    // Upper 16 bits = limit, lower 16 = signed lead byte offset.
    int32_t offset = 0;
    for(int32_t i = 1;; ++i) {
        int32_t nextOffset = offset;
        while(i < scriptStartsLength - 1) {
            int32_t newLeadByte = table[i];
            if(newLeadByte == 0xff) {
                // "Don't care" lead byte for reserved range, continue with current offset.
            } else {
                nextOffset = newLeadByte - (scriptStarts[i] >> 8);
                if(nextOffset != offset) { break; }
            }
            ++i;
        }
        if(offset != 0 || i < scriptStartsLength - 1) {
            ranges.addElement(((int32_t)scriptStarts[i] << 16) | (offset & 0xffff), errorCode);
        }
        if(i == scriptStartsLength - 1) { break; }
        offset = nextOffset;
    }
}

int32_t
CollationData::addLowScriptRange(uint8_t table[], int32_t index, int32_t lowStart) const {
    int32_t start = scriptStarts[index];
    if((start & 0xff) < (lowStart & 0xff)) {
        lowStart += 0x100;
    }
    table[index] = (uint8_t)(lowStart >> 8);
    int32_t limit = scriptStarts[index + 1];
    lowStart = ((lowStart & 0xff00) + ((limit & 0xff00) - (start & 0xff00))) | (limit & 0xff);
    return lowStart;
}

int32_t
CollationData::addHighScriptRange(uint8_t table[], int32_t index, int32_t highLimit) const {
    int32_t limit = scriptStarts[index + 1];
    if((limit & 0xff) > (highLimit & 0xff)) {
        highLimit -= 0x100;
    }
    int32_t start = scriptStarts[index];
    highLimit = ((highLimit & 0xff00) - ((limit & 0xff00) - (start & 0xff00))) | (start & 0xff);
    table[index] = (uint8_t)(highLimit >> 8);
    return highLimit;
}

U_NAMESPACE_END

#endif  // !UCONFIG_NO_COLLATION