aboutsummaryrefslogtreecommitdiffstats
path: root/contrib/libs/icu/i18n/numparse_decimal.cpp
blob: aedca60586ad7dfe7d020f91d6eb6ff82c8c59f5 (plain) (blame)
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
// © 2018 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html

#include "unicode/utypes.h"

#if !UCONFIG_NO_FORMATTING

// Allow implicit conversion from char16_t* to UnicodeString for this file:
// Helpful in toString methods and elsewhere.
#define UNISTR_FROM_STRING_EXPLICIT

#include "numparse_types.h"
#include "numparse_decimal.h"
#include "static_unicode_sets.h"
#include "numparse_utils.h"
#include "unicode/uchar.h"
#include "putilimp.h"
#include "number_decimalquantity.h"
#include "string_segment.h"

using namespace icu;
using namespace icu::numparse;
using namespace icu::numparse::impl;


DecimalMatcher::DecimalMatcher(const DecimalFormatSymbols& symbols, const Grouper& grouper,
                               parse_flags_t parseFlags) {
    if (0 != (parseFlags & PARSE_FLAG_MONETARY_SEPARATORS)) {
        groupingSeparator = symbols.getConstSymbol(DecimalFormatSymbols::kMonetaryGroupingSeparatorSymbol);
        decimalSeparator = symbols.getConstSymbol(DecimalFormatSymbols::kMonetarySeparatorSymbol);
    } else {
        groupingSeparator = symbols.getConstSymbol(DecimalFormatSymbols::kGroupingSeparatorSymbol);
        decimalSeparator = symbols.getConstSymbol(DecimalFormatSymbols::kDecimalSeparatorSymbol);
    }
    bool strictSeparators = 0 != (parseFlags & PARSE_FLAG_STRICT_SEPARATORS);
    unisets::Key groupingKey = strictSeparators ? unisets::STRICT_ALL_SEPARATORS
                                                : unisets::ALL_SEPARATORS;

    // Attempt to find separators in the static cache

    groupingUniSet = unisets::get(groupingKey);
    unisets::Key decimalKey = unisets::chooseFrom(
            decimalSeparator,
            strictSeparators ? unisets::STRICT_COMMA : unisets::COMMA,
            strictSeparators ? unisets::STRICT_PERIOD : unisets::PERIOD);
    if (decimalKey >= 0) {
        decimalUniSet = unisets::get(decimalKey);
    } else if (!decimalSeparator.isEmpty()) {
        auto* set = new UnicodeSet();
        set->add(decimalSeparator.char32At(0));
        set->freeze();
        decimalUniSet = set;
        fLocalDecimalUniSet.adoptInstead(set);
    } else {
        decimalUniSet = unisets::get(unisets::EMPTY);
    }

    if (groupingKey >= 0 && decimalKey >= 0) {
        // Everything is available in the static cache
        separatorSet = groupingUniSet;
        leadSet = unisets::get(
                strictSeparators ? unisets::DIGITS_OR_ALL_SEPARATORS
                                 : unisets::DIGITS_OR_STRICT_ALL_SEPARATORS);
    } else {
        auto* set = new UnicodeSet();
        set->addAll(*groupingUniSet);
        set->addAll(*decimalUniSet);
        set->freeze();
        separatorSet = set;
        fLocalSeparatorSet.adoptInstead(set);
        leadSet = nullptr;
    }

    UChar32 cpZero = symbols.getCodePointZero();
    if (cpZero == -1 || !u_isdigit(cpZero) || u_digit(cpZero, 10) != 0) {
        // Uncommon case: okay to allocate.
        auto* digitStrings = new UnicodeString[10];
        fLocalDigitStrings.adoptInstead(digitStrings);
        for (int32_t i = 0; i <= 9; i++) {
            digitStrings[i] = symbols.getConstDigitSymbol(i);
        }
    }

    requireGroupingMatch = 0 != (parseFlags & PARSE_FLAG_STRICT_GROUPING_SIZE);
    groupingDisabled = 0 != (parseFlags & PARSE_FLAG_GROUPING_DISABLED);
    integerOnly = 0 != (parseFlags & PARSE_FLAG_INTEGER_ONLY);
    grouping1 = grouper.getPrimary();
    grouping2 = grouper.getSecondary();

    // Fraction grouping parsing is disabled for now but could be enabled later.
    // See https://unicode-org.atlassian.net/browse/ICU-10794
    // fractionGrouping = 0 != (parseFlags & PARSE_FLAG_FRACTION_GROUPING_ENABLED);
}

bool DecimalMatcher::match(StringSegment& segment, ParsedNumber& result, UErrorCode& status) const {
    return match(segment, result, 0, status);
}

bool DecimalMatcher::match(StringSegment& segment, ParsedNumber& result, int8_t exponentSign,
                           UErrorCode&) const {
    if (result.seenNumber() && exponentSign == 0) {
        // A number has already been consumed.
        return false;
    } else if (exponentSign != 0) {
        // scientific notation always comes after the number
        U_ASSERT(!result.quantity.bogus);
    }

    // Initial offset before any character consumption.
    int32_t initialOffset = segment.getOffset();

    // Return value: whether to ask for more characters.
    bool maybeMore = false;

    // All digits consumed so far.
    number::impl::DecimalQuantity digitsConsumed;
    digitsConsumed.bogus = true;

    // The total number of digits after the decimal place, used for scaling the result.
    int32_t digitsAfterDecimalPlace = 0;

    // The actual grouping and decimal separators used in the string.
    // If non-null, we have seen that token.
    UnicodeString actualGroupingString;
    UnicodeString actualDecimalString;
    actualGroupingString.setToBogus();
    actualDecimalString.setToBogus();

    // Information for two groups: the previous group and the current group.
    //
    // Each group has three pieces of information:
    //
    // Offset: the string position of the beginning of the group, including a leading separator
    // if there was a leading separator. This is needed in case we need to rewind the parse to
    // that position.
    //
    // Separator type:
    // 0 => beginning of string
    // 1 => lead separator is a grouping separator
    // 2 => lead separator is a decimal separator
    //
    // Count: the number of digits in the group. If -1, the group has been validated.
    int32_t currGroupOffset = 0;
    int32_t currGroupSepType = 0;
    int32_t currGroupCount = 0;
    int32_t prevGroupOffset = -1;
    int32_t prevGroupSepType = -1;
    int32_t prevGroupCount = -1;

    while (segment.length() > 0) {
        maybeMore = false;

        // Attempt to match a digit.
        int8_t digit = -1;

        // Try by code point digit value.
        UChar32 cp = segment.getCodePoint();
        if (u_isdigit(cp)) {
            segment.adjustOffset(U16_LENGTH(cp));
            digit = static_cast<int8_t>(u_digit(cp, 10));
        }

        // Try by digit string.
        if (digit == -1 && !fLocalDigitStrings.isNull()) {
            for (int32_t i = 0; i < 10; i++) {
                const UnicodeString& str = fLocalDigitStrings[i];
                if (str.isEmpty()) {
                    continue;
                }
                int32_t overlap = segment.getCommonPrefixLength(str);
                if (overlap == str.length()) {
                    segment.adjustOffset(overlap);
                    digit = static_cast<int8_t>(i);
                    break;
                }
                maybeMore = maybeMore || (overlap == segment.length());
            }
        }

        if (digit >= 0) {
            // Digit was found.
            if (digitsConsumed.bogus) {
                digitsConsumed.bogus = false;
                digitsConsumed.clear();
            }
            digitsConsumed.appendDigit(digit, 0, true);
            currGroupCount++;
            if (!actualDecimalString.isBogus()) {
                digitsAfterDecimalPlace++;
            }
            continue;
        }

        // Attempt to match a literal grouping or decimal separator.
        bool isDecimal = false;
        bool isGrouping = false;

        // 1) Attempt the decimal separator string literal.
        // if (we have not seen a decimal separator yet) { ... }
        if (actualDecimalString.isBogus() && !decimalSeparator.isEmpty()) {
            int32_t overlap = segment.getCommonPrefixLength(decimalSeparator);
            maybeMore = maybeMore || (overlap == segment.length());
            if (overlap == decimalSeparator.length()) {
                isDecimal = true;
                actualDecimalString = decimalSeparator;
            }
        }

        // 2) Attempt to match the actual grouping string literal.
        if (!actualGroupingString.isBogus()) {
            int32_t overlap = segment.getCommonPrefixLength(actualGroupingString);
            maybeMore = maybeMore || (overlap == segment.length());
            if (overlap == actualGroupingString.length()) {
                isGrouping = true;
            }
        }

        // 2.5) Attempt to match a new the grouping separator string literal.
        // if (we have not seen a grouping or decimal separator yet) { ... }
        if (!groupingDisabled && actualGroupingString.isBogus() && actualDecimalString.isBogus() &&
            !groupingSeparator.isEmpty()) {
            int32_t overlap = segment.getCommonPrefixLength(groupingSeparator);
            maybeMore = maybeMore || (overlap == segment.length());
            if (overlap == groupingSeparator.length()) {
                isGrouping = true;
                actualGroupingString = groupingSeparator;
            }
        }

        // 3) Attempt to match a decimal separator from the equivalence set.
        // if (we have not seen a decimal separator yet) { ... }
        // The !isGrouping is to confirm that we haven't yet matched the current character.
        if (!isGrouping && actualDecimalString.isBogus()) {
            if (decimalUniSet->contains(cp)) {
                isDecimal = true;
                actualDecimalString = UnicodeString(cp);
            }
        }

        // 4) Attempt to match a grouping separator from the equivalence set.
        // if (we have not seen a grouping or decimal separator yet) { ... }
        if (!groupingDisabled && actualGroupingString.isBogus() && actualDecimalString.isBogus()) {
            if (groupingUniSet->contains(cp)) {
                isGrouping = true;
                actualGroupingString = UnicodeString(cp);
            }
        }

        // Leave if we failed to match this as a separator.
        if (!isDecimal && !isGrouping) {
            break;
        }

        // Check for conditions when we don't want to accept the separator.
        if (isDecimal && integerOnly) {
            break;
        } else if (currGroupSepType == 2 && isGrouping) {
            // Fraction grouping
            break;
        }

        // Validate intermediate grouping sizes.
        bool prevValidSecondary = validateGroup(prevGroupSepType, prevGroupCount, false);
        bool currValidPrimary = validateGroup(currGroupSepType, currGroupCount, true);
        if (!prevValidSecondary || (isDecimal && !currValidPrimary)) {
            // Invalid grouping sizes.
            if (isGrouping && currGroupCount == 0) {
                // Trailing grouping separators: these are taken care of below
                U_ASSERT(currGroupSepType == 1);
            } else if (requireGroupingMatch) {
                // Strict mode: reject the parse
                digitsConsumed.clear();
                digitsConsumed.bogus = true;
            }
            break;
        } else if (requireGroupingMatch && currGroupCount == 0 && currGroupSepType == 1) {
            break;
        } else {
            // Grouping sizes OK so far.
            prevGroupOffset = currGroupOffset;
            prevGroupCount = currGroupCount;
            if (isDecimal) {
                // Do not validate this group any more.
                prevGroupSepType = -1;
            } else {
                prevGroupSepType = currGroupSepType;
            }
        }

        // OK to accept the separator.
        // Special case: don't update currGroup if it is empty; this allows two grouping
        // separators in a row in lenient mode.
        if (currGroupCount != 0) {
            currGroupOffset = segment.getOffset();
        }
        currGroupSepType = isGrouping ? 1 : 2;
        currGroupCount = 0;
        if (isGrouping) {
            segment.adjustOffset(actualGroupingString.length());
        } else {
            segment.adjustOffset(actualDecimalString.length());
        }
    }

    // End of main loop.
    // Back up if there was a trailing grouping separator.
    // Shift prev -> curr so we can check it as a final group.
    if (currGroupSepType != 2 && currGroupCount == 0) {
        maybeMore = true;
        segment.setOffset(currGroupOffset);
        currGroupOffset = prevGroupOffset;
        currGroupSepType = prevGroupSepType;
        currGroupCount = prevGroupCount;
        prevGroupOffset = -1;
        prevGroupSepType = 0;
        prevGroupCount = 1;
    }

    // Validate final grouping sizes.
    bool prevValidSecondary = validateGroup(prevGroupSepType, prevGroupCount, false);
    bool currValidPrimary = validateGroup(currGroupSepType, currGroupCount, true);
    if (!requireGroupingMatch) {
        // The cases we need to handle here are lone digits.
        // Examples: "1,1"  "1,1,"  "1,1,1"  "1,1,1,"  ",1" (all parse as 1)
        // See more examples in numberformattestspecification.txt
        int32_t digitsToRemove = 0;
        if (!prevValidSecondary) {
            segment.setOffset(prevGroupOffset);
            digitsToRemove += prevGroupCount;
            digitsToRemove += currGroupCount;
        } else if (!currValidPrimary && (prevGroupSepType != 0 || prevGroupCount != 0)) {
            maybeMore = true;
            segment.setOffset(currGroupOffset);
            digitsToRemove += currGroupCount;
        }
        if (digitsToRemove != 0) {
            digitsConsumed.adjustMagnitude(-digitsToRemove);
            digitsConsumed.truncate();
        }
        prevValidSecondary = true;
        currValidPrimary = true;
    }
    if (currGroupSepType != 2 && (!prevValidSecondary || !currValidPrimary)) {
        // Grouping failure.
        digitsConsumed.bogus = true;
    }

    // Strings that start with a separator but have no digits,
    // or strings that failed a grouping size check.
    if (digitsConsumed.bogus) {
        maybeMore = maybeMore || (segment.length() == 0);
        segment.setOffset(initialOffset);
        return maybeMore;
    }

    // We passed all inspections. Start post-processing.

    // Adjust for fraction part.
    digitsConsumed.adjustMagnitude(-digitsAfterDecimalPlace);

    // Set the digits, either normal or exponent.
    if (exponentSign != 0 && segment.getOffset() != initialOffset) {
        bool overflow = false;
        if (digitsConsumed.fitsInLong()) {
            int64_t exponentLong = digitsConsumed.toLong(false);
            U_ASSERT(exponentLong >= 0);
            if (exponentLong <= INT32_MAX) {
                auto exponentInt = static_cast<int32_t>(exponentLong);
                if (result.quantity.adjustMagnitude(exponentSign * exponentInt)) {
                    overflow = true;
                }
            } else {
                overflow = true;
            }
        } else {
            overflow = true;
        }
        if (overflow) {
            if (exponentSign == -1) {
                // Set to zero
                result.quantity.clear();
            } else {
                // Set to infinity
                result.quantity.bogus = true;
                result.flags |= FLAG_INFINITY;
            }
        }
    } else {
        result.quantity = digitsConsumed;
    }

    // Set other information into the result and return.
    if (!actualDecimalString.isBogus()) {
        result.flags |= FLAG_HAS_DECIMAL_SEPARATOR;
    }
    result.setCharsConsumed(segment);
    return segment.length() == 0 || maybeMore;
}

bool DecimalMatcher::validateGroup(int32_t sepType, int32_t count, bool isPrimary) const {
    if (requireGroupingMatch) {
        if (sepType == -1) {
            // No such group (prevGroup before first shift).
            return true;
        } else if (sepType == 0) {
            // First group.
            if (isPrimary) {
                // No grouping separators is OK.
                return true;
            } else {
                return count != 0 && count <= grouping2;
            }
        } else if (sepType == 1) {
            // Middle group.
            if (isPrimary) {
                return count == grouping1;
            } else {
                return count == grouping2;
            }
        } else {
            U_ASSERT(sepType == 2);
            // After the decimal separator.
            return true;
        }
    } else {
        if (sepType == 1) {
            // #11230: don't accept middle groups with only 1 digit.
            return count != 1;
        } else {
            return true;
        }
    }
}

bool DecimalMatcher::smokeTest(const StringSegment& segment) const {
    // The common case uses a static leadSet for efficiency.
    if (fLocalDigitStrings.isNull() && leadSet != nullptr) {
        return segment.startsWith(*leadSet);
    }
    if (segment.startsWith(*separatorSet) || u_isdigit(segment.getCodePoint())) {
        return true;
    }
    if (fLocalDigitStrings.isNull()) {
        return false;
    }
    for (int32_t i = 0; i < 10; i++) {
        if (segment.startsWith(fLocalDigitStrings[i])) {
            return true;
        }
    }
    return false;
}

UnicodeString DecimalMatcher::toString() const {
    return u"<Decimal>";
}


#endif /* #if !UCONFIG_NO_FORMATTING */