Repository navigation
Expand file tree
/
Copy pathtext_normalization.cpp
More file actions
621 lines (591 loc) · 22.6 KB
/
Copy pathtext_normalization.cpp
File metadata and controls
621 lines (591 loc) · 22.6 KB
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
#include "engine/framework/text/text_normalization.h"
#include "engine/framework/io/text.h"
#include <algorithm>
#include <cctype>
#include <cstdint>
#include <functional>
#include <limits>
#include <regex>
#include <stdexcept>
#include <string>
#include <utility>
#include <vector>
namespace engine::text {
namespace {
bool parse_non_negative_i64(std::string_view text, int64_t & out) {
if (text.empty()) {
return false;
}
int64_t value = 0;
for (const char ch : text) {
if (ch < '0' || ch > '9') {
return false;
}
const int digit = ch - '0';
if (value > (std::numeric_limits<int64_t>::max() - digit) / 10) {
return false;
}
value = value * 10 + digit;
}
out = value;
return true;
}
std::string english_cardinal_under_1000(int value) {
static const std::string units[] = {
"zero", "one", "two", "three", "four", "five", "six", "seven", "eight", "nine",
"ten", "eleven", "twelve", "thirteen", "fourteen", "fifteen", "sixteen",
"seventeen", "eighteen", "nineteen",
};
static const std::string tens[] = {
"", "", "twenty", "thirty", "forty", "fifty", "sixty", "seventy", "eighty", "ninety",
};
if (value < 20) {
return units[static_cast<size_t>(value)];
}
if (value < 100) {
const int ten = value / 10;
const int one = value % 10;
return one == 0 ? tens[static_cast<size_t>(ten)]
: tens[static_cast<size_t>(ten)] + " " + units[static_cast<size_t>(one)];
}
const int hundred = value / 100;
const int rest = value % 100;
return rest == 0 ? units[static_cast<size_t>(hundred)] + " hundred"
: units[static_cast<size_t>(hundred)] + " hundred " + english_cardinal_under_1000(rest);
}
std::string english_digits_individually(std::string_view digits) {
std::string out;
for (size_t i = 0; i < digits.size(); ++i) {
if (i != 0) {
out.push_back(' ');
}
out += english_cardinal_under_1000(digits[i] - '0');
}
return out;
}
std::string english_cardinal(int64_t value) {
if (value < 0) {
return "minus " + english_cardinal(-value);
}
if (value < 1000) {
return english_cardinal_under_1000(static_cast<int>(value));
}
if (value < 1000000) {
const int64_t thousands = value / 1000;
const int64_t rest = value % 1000;
return rest == 0 ? english_cardinal(thousands) + " thousand"
: english_cardinal(thousands) + " thousand " + english_cardinal(rest);
}
return std::to_string(value);
}
std::string english_cardinal_from_digits(std::string_view digits) {
int64_t value = 0;
if (!parse_non_negative_i64(digits, value)) {
return english_digits_individually(digits);
}
return english_cardinal(value);
}
std::string english_year(int value) {
if (value >= 1900 && value <= 1999) {
const int rest = value - 1900;
return rest == 0 ? "nineteen hundred" : "nineteen " + english_cardinal_under_1000(rest);
}
if (value >= 2000 && value <= 2099) {
const int rest = value - 2000;
if (rest == 0) {
return "two thousand";
}
if (rest < 10) {
return "two thousand " + english_cardinal_under_1000(rest);
}
return "twenty " + english_cardinal_under_1000(rest);
}
return english_cardinal(value);
}
std::string english_ordinal(int64_t value) {
static const std::string ordinals[] = {
"zeroth", "first", "second", "third", "fourth", "fifth", "sixth", "seventh", "eighth",
"ninth", "tenth", "eleventh", "twelfth", "thirteenth", "fourteenth", "fifteenth",
"sixteenth", "seventeenth", "eighteenth", "nineteenth",
};
static const std::string tens_ordinals[] = {
"", "", "twentieth", "thirtieth", "fortieth", "fiftieth",
"sixtieth", "seventieth", "eightieth", "ninetieth",
};
if (value < 20) {
return ordinals[static_cast<size_t>(value)];
}
if (value < 100) {
const int64_t ten = value / 10;
const int64_t one = value % 10;
return one == 0 ? tens_ordinals[static_cast<size_t>(ten)]
: english_cardinal_under_1000(static_cast<int>(ten * 10)) + " " + english_ordinal(one);
}
if (value % 100 == 0) {
return english_cardinal(value / 100) + " hundredth";
}
return english_cardinal(value - (value % 100)) + " " + english_ordinal(value % 100);
}
std::string english_ordinal_from_digits(std::string_view digits) {
int64_t value = 0;
if (!parse_non_negative_i64(digits, value)) {
return english_digits_individually(digits);
}
return english_ordinal(value);
}
std::string english_fraction(std::string_view numerator_digits, std::string_view denominator_digits) {
int64_t numerator = 0;
int64_t denominator = 0;
if (!parse_non_negative_i64(numerator_digits, numerator) ||
!parse_non_negative_i64(denominator_digits, denominator)) {
return english_digits_individually(numerator_digits) + " over " + english_digits_individually(denominator_digits);
}
if (denominator == 2) {
return numerator == 1 ? "one half" : english_cardinal(numerator) + " halves";
}
std::string out = english_cardinal(numerator) + " " + english_ordinal(denominator);
if (numerator != 1) {
out.push_back('s');
}
return out;
}
std::string normalize_english_regex(
std::string text,
const std::regex & pattern,
const std::function<std::string(const std::smatch &)> & replacement) {
std::string out;
auto begin = text.cbegin();
auto end = text.cend();
std::smatch match;
while (std::regex_search(begin, end, match, pattern)) {
out.append(begin, match[0].first);
out += replacement(match);
begin = match[0].second;
}
out.append(begin, end);
return out;
}
std::string normalize_english_dates(std::string text) {
const std::regex pattern(
R"(\b(January|February|March|April|May|June|July|August|September|October|November|December)\s+(\d{1,2})(?:st|nd|rd|th)?(?:,\s*(\d{4}))?)",
std::regex_constants::icase);
return normalize_english_regex(std::move(text), pattern, [](const std::smatch & match) {
int64_t day = 0;
int64_t year = 0;
std::string out = match[1].str() + " " +
(parse_non_negative_i64(match[2].str(), day) ? english_ordinal(day) : english_digits_individually(match[2].str()));
if (match[3].matched) {
out += " ";
out += parse_non_negative_i64(match[3].str(), year) ? english_year(static_cast<int>(year))
: english_digits_individually(match[3].str());
}
return out;
});
}
std::string apply_index_tts_punctuation_map(std::string text) {
const std::pair<const char *, const char *> replacements[] = {
{"：", ","}, {"；", ","}, {";", ","}, {"，", ","}, {"。", "."}, {"！", "!"},
{"？", "?"}, {"\n", " "}, {"·", "-"}, {"、", ","}, {"...", "…"}, {",,,", "…"},
{"，，，", "…"}, {"……", "…"}, {"“", "'"}, {"”", "'"}, {"\"", "'"}, {"‘", "'"},
{"’", "'"}, {"（", "'"}, {"）", "'"}, {"(", "'"}, {")", "'"}, {"《", "'"},
{"》", "'"}, {"【", "'"}, {"】", "'"}, {"[", "'"}, {"]", "'"}, {"—", "-"},
{"～", "-"}, {"~", "-"}, {"「", "'"}, {"」", "'"}, {":", ","},
};
for (const auto & [from, to] : replacements) {
text = replace_all(std::move(text), from, to);
}
return text;
}
std::string uppercase_ascii(std::string text) {
for (char & ch : text) {
ch = static_cast<char>(std::toupper(static_cast<unsigned char>(ch)));
}
return text;
}
uint32_t decode_utf8_codepoint(std::string_view text, size_t offset, size_t & width) {
const auto byte = [&](size_t i) { return static_cast<unsigned char>(text[offset + i]); };
const unsigned char first = byte(0);
if ((first & 0x80U) == 0U) {
width = 1;
return first;
}
if ((first & 0xE0U) == 0xC0U && offset + 1 < text.size()) {
width = 2;
return ((first & 0x1FU) << 6U) | (byte(1) & 0x3FU);
}
if ((first & 0xF0U) == 0xE0U && offset + 2 < text.size()) {
width = 3;
return ((first & 0x0FU) << 12U) | ((byte(1) & 0x3FU) << 6U) | (byte(2) & 0x3FU);
}
if ((first & 0xF8U) == 0xF0U && offset + 3 < text.size()) {
width = 4;
return ((first & 0x07U) << 18U) | ((byte(1) & 0x3FU) << 12U) |
((byte(2) & 0x3FU) << 6U) | (byte(3) & 0x3FU);
}
throw std::runtime_error("text normalization received invalid UTF-8");
}
void append_utf8(std::string & out, uint32_t cp) {
if (cp <= 0x7FU) {
out.push_back(static_cast<char>(cp));
} else if (cp <= 0x7FFU) {
out.push_back(static_cast<char>(0xC0U | (cp >> 6U)));
out.push_back(static_cast<char>(0x80U | (cp & 0x3FU)));
} else if (cp <= 0xFFFFU) {
out.push_back(static_cast<char>(0xE0U | (cp >> 12U)));
out.push_back(static_cast<char>(0x80U | ((cp >> 6U) & 0x3FU)));
out.push_back(static_cast<char>(0x80U | (cp & 0x3FU)));
} else {
out.push_back(static_cast<char>(0xF0U | (cp >> 18U)));
out.push_back(static_cast<char>(0x80U | ((cp >> 12U) & 0x3FU)));
out.push_back(static_cast<char>(0x80U | ((cp >> 6U) & 0x3FU)));
out.push_back(static_cast<char>(0x80U | (cp & 0x3FU)));
}
}
uint32_t halfwidth_katakana_to_fullwidth(uint32_t cp) {
switch (cp) {
case 0xFF66: return 0x30F2;
case 0xFF67: return 0x30A1;
case 0xFF68: return 0x30A3;
case 0xFF69: return 0x30A5;
case 0xFF6A: return 0x30A7;
case 0xFF6B: return 0x30A9;
case 0xFF6C: return 0x30E3;
case 0xFF6D: return 0x30E5;
case 0xFF6E: return 0x30E7;
case 0xFF6F: return 0x30C3;
case 0xFF70: return 0x30FC;
case 0xFF71: return 0x30A2;
case 0xFF72: return 0x30A4;
case 0xFF73: return 0x30A6;
case 0xFF74: return 0x30A8;
case 0xFF75: return 0x30AA;
case 0xFF76: return 0x30AB;
case 0xFF77: return 0x30AD;
case 0xFF78: return 0x30AF;
case 0xFF79: return 0x30B1;
case 0xFF7A: return 0x30B3;
case 0xFF7B: return 0x30B5;
case 0xFF7C: return 0x30B7;
case 0xFF7D: return 0x30B9;
case 0xFF7E: return 0x30BB;
case 0xFF7F: return 0x30BD;
case 0xFF80: return 0x30BF;
case 0xFF81: return 0x30C1;
case 0xFF82: return 0x30C4;
case 0xFF83: return 0x30C6;
case 0xFF84: return 0x30C8;
case 0xFF85: return 0x30CA;
case 0xFF86: return 0x30CB;
case 0xFF87: return 0x30CC;
case 0xFF88: return 0x30CD;
case 0xFF89: return 0x30CE;
case 0xFF8A: return 0x30CF;
case 0xFF8B: return 0x30D2;
case 0xFF8C: return 0x30D5;
case 0xFF8D: return 0x30D8;
case 0xFF8E: return 0x30DB;
case 0xFF8F: return 0x30DE;
case 0xFF90: return 0x30DF;
case 0xFF91: return 0x30E0;
case 0xFF92: return 0x30E1;
case 0xFF93: return 0x30E2;
case 0xFF94: return 0x30E4;
case 0xFF95: return 0x30E6;
case 0xFF96: return 0x30E8;
case 0xFF97: return 0x30E9;
case 0xFF98: return 0x30EA;
case 0xFF99: return 0x30EB;
case 0xFF9A: return 0x30EC;
case 0xFF9B: return 0x30ED;
case 0xFF9C: return 0x30EF;
case 0xFF9D: return 0x30F3;
default: return cp;
}
}
uint32_t apply_katakana_voicing(uint32_t cp, uint32_t mark) {
if (mark == 0x3099 || mark == 0xFF9E) {
switch (cp) {
case 0x30A6: return 0x30F4;
case 0x30AB: return 0x30AC;
case 0x30AD: return 0x30AE;
case 0x30AF: return 0x30B0;
case 0x30B1: return 0x30B2;
case 0x30B3: return 0x30B4;
case 0x30B5: return 0x30B6;
case 0x30B7: return 0x30B8;
case 0x30B9: return 0x30BA;
case 0x30BB: return 0x30BC;
case 0x30BD: return 0x30BE;
case 0x30BF: return 0x30C0;
case 0x30C1: return 0x30C2;
case 0x30C4: return 0x30C5;
case 0x30C6: return 0x30C7;
case 0x30C8: return 0x30C9;
case 0x30CF: return 0x30D0;
case 0x30D2: return 0x30D3;
case 0x30D5: return 0x30D6;
case 0x30D8: return 0x30D9;
case 0x30DB: return 0x30DC;
case 0x30EF: return 0x30F7;
case 0x30F0: return 0x30F8;
case 0x30F1: return 0x30F9;
case 0x30F2: return 0x30FA;
default: return cp;
}
}
if (mark == 0x309A || mark == 0xFF9F) {
switch (cp) {
case 0x30CF: return 0x30D1;
case 0x30D2: return 0x30D4;
case 0x30D5: return 0x30D7;
case 0x30D8: return 0x30DA;
case 0x30DB: return 0x30DD;
default: return cp;
}
}
return cp;
}
std::vector<uint32_t> japanese_codepoints(std::string_view text) {
std::vector<uint32_t> out;
for (size_t pos = 0; pos < text.size();) {
size_t width = 0;
uint32_t cp = decode_utf8_codepoint(text, pos, width);
pos += width;
if (cp >= 0x3041U && cp <= 0x3096U) {
cp += 0x60U;
} else if (cp == 0x309DU) {
cp = 0x30FDU;
} else if (cp == 0x309EU) {
cp = 0x30FEU;
} else {
cp = halfwidth_katakana_to_fullwidth(cp);
}
out.push_back(cp);
}
for (size_t i = 1; i < out.size();) {
if (out[i] == 0x3099U || out[i] == 0x309AU || out[i] == 0xFF9EU || out[i] == 0xFF9FU) {
const uint32_t voiced = apply_katakana_voicing(out[i - 1], out[i]);
if (voiced != out[i - 1]) {
out[i - 1] = voiced;
out.erase(out.begin() + static_cast<std::ptrdiff_t>(i));
continue;
}
}
++i;
}
return out;
}
} // namespace
std::string replace_all(std::string text, std::string_view from, std::string_view to) {
if (from.empty()) {
return text;
}
for (size_t pos = text.find(from); pos != std::string::npos; pos = text.find(from, pos + to.size())) {
text.replace(pos, from.size(), to);
}
return text;
}
std::string collapse_ascii_whitespace(std::string_view text) {
std::string out;
out.reserve(text.size());
bool previous_space = false;
for (const char ch : text) {
const bool space = ch == ' ' || ch == '\t' || ch == '\n' || ch == '\r';
if (space) {
if (!previous_space) {
out.push_back(' ');
}
} else {
out.push_back(ch);
}
previous_space = space;
}
return engine::io::trim_ascii_whitespace(std::move(out));
}
std::string normalize_index_tts_punctuation(std::string text) {
return apply_index_tts_punctuation_map(std::move(text));
}
std::string normalize_english_numbers(std::string text) {
// URLs before anything else: "https://example.com" ->
// "https comma slash slash example dot com" (wetext en verbalizes the
// punctuation). The domain dots must not become decimal points later.
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(https?)://([A-Za-z0-9-]+(?:\.[A-Za-z0-9-]+)*))", std::regex_constants::icase),
[](const std::smatch & match) {
std::string domain = match[2].str();
return match[1].str() + " comma slash slash " +
std::regex_replace(domain, std::regex(R"(\.)"), " dot ");
});
// Telephone runs: "135-4567-8900" -> digit-by-digit groups (wetext en).
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(\d{3})-(\d{4})-(\d{4})\b)"),
[](const std::smatch & match) {
return english_digits_individually(match[1].str()) + ", " +
english_digits_individually(match[2].str()) + ", " +
english_digits_individually(match[3].str());
});
// o'clock times: "3:00" -> "three,oh zero" (wetext en output, comma included).
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(\d{1,2}):00\b)"),
[](const std::smatch & match) {
return english_cardinal_from_digits(match[1].str()) + ",oh zero";
});
// Numeric dates: "2024/10/01" -> "the first of october twenty twenty four".
static const char * const kMonths[] = {
"january", "february", "march", "april", "may", "june",
"july", "august", "september", "october", "november", "december",
};
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(\d{4})/(\d{1,2})/(\d{1,2})\b)"),
[](const std::smatch & match) {
int64_t year = 0;
int64_t month = 0;
int64_t day = 0;
if (!parse_non_negative_i64(match[1].str(), year) ||
!parse_non_negative_i64(match[2].str(), month) || month < 1 || month > 12 ||
!parse_non_negative_i64(match[3].str(), day) || day < 1 || day > 31) {
return match[0].str();
}
return "the " + english_ordinal(day) + " of " + kMonths[month - 1] + " " +
english_year(static_cast<int>(year));
});
// Currency: "$50" -> "fifty dollars".
text = normalize_english_regex(
std::move(text),
std::regex(R"(\$(\d+(?:\.\d+)?))"),
[](const std::smatch & match) {
std::string amount;
const std::string value = match[1].str();
const size_t dot = value.find('.');
if (dot == std::string::npos) {
amount = english_cardinal_from_digits(value);
} else {
amount = english_cardinal_from_digits(value.substr(0, dot)) + " point " +
english_digits_individually(value.substr(dot + 1));
}
return amount + (value == "1" ? " dollar" : " dollars");
});
// Negative numbers: "-5 degrees" -> "negative five degrees". The minus is
// only a sign when not attached to a word/number ("GPT-5" stays intact).
text = normalize_english_regex(
std::move(text),
std::regex(R"((^|[^A-Za-z0-9])-(\d+))"),
[](const std::smatch & match) {
return match[1].str() + "negative " + english_cardinal_from_digits(match[2].str());
});
text = normalize_english_dates(std::move(text));
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(\d+(?:\.\d+)?)%)"),
[](const std::smatch & match) {
return normalize_english_regex(
match[1].str(),
std::regex(R"((\d+)\.(\d+))"),
[](const std::smatch & decimal) {
return english_cardinal_from_digits(decimal[1].str()) + " point " +
english_digits_individually(decimal[2].str());
}) +
" percent";
});
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(\d+)/(\d+)\b)"),
[](const std::smatch & match) {
return english_fraction(match[1].str(), match[2].str());
});
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(\d+)\.(\d+)\b)"),
[](const std::smatch & match) {
return english_cardinal_from_digits(match[1].str()) + " point " +
english_digits_individually(match[2].str());
});
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(\d+)(st|nd|rd|th)\b)", std::regex_constants::icase),
[](const std::smatch & match) {
return english_ordinal_from_digits(match[1].str());
});
// Split letter<->digit boundaries so digits attached to letters verbalize
// like the official wetext English normalizer ("DS4" -> "DS four",
// "R2D2" -> "R two D two", "4K" -> "four K"). Runs after dates/decimals/
// ordinals so "1st", "2.5" and friends have already been expanded; the
// standalone cardinal rule below then spells out each digit group.
text = std::regex_replace(std::move(text), std::regex(R"(([A-Za-z])(\d))"), "$1 $2");
text = std::regex_replace(std::move(text), std::regex(R"((\d)([A-Za-z]))"), "$1 $2");
text = normalize_english_regex(
std::move(text),
std::regex(R"(\b(\d+)\b)"),
[](const std::smatch & match) {
return english_cardinal_from_digits(match[1].str());
});
return text;
}
std::string normalize_english_text(std::string_view text, const EnglishTextNormalizationOptions & options) {
std::string out = collapse_ascii_whitespace(text);
if (options.expand_common_contractions) {
out = std::regex_replace(
out,
std::regex(R"((what|where|who|which|how|t?here|it|s?he|that|this)'s)", std::regex_constants::icase),
"$1 is");
}
if (options.spell_numbers) {
out = normalize_english_numbers(std::move(out));
}
if (options.verbalize_symbols) {
// Match the official wetext English normalizer: standalone ASCII
// symbols are verbalized ("a_b" -> "a underscore b",
// "C++" -> "C plus plus", "a=b" -> "a equal sign b"). Runs after
// number spelling so "50%" has already become "fifty percent".
const std::pair<const char *, const char *> symbol_words[] = {
{"_", " underscore "},
{"+", " plus "},
{"=", " equal sign "},
{"*", " asterisk "},
{"&", " and "},
{"#", " hash "},
{"%", " percent "},
{"|", " vertical bar "},
{"~", " tilde "},
};
for (const auto & [from, to] : symbol_words) {
out = replace_all(std::move(out), from, to);
}
out = collapse_ascii_whitespace(out);
}
if (options.index_tts_punctuation) {
out = apply_index_tts_punctuation_map(std::move(out));
}
if (options.uppercase_ascii) {
out = uppercase_ascii(std::move(out));
}
return out;
}
std::string normalize_japanese_text(std::string_view text) {
std::string mapped;
mapped.reserve(text.size());
for (uint32_t cp : japanese_codepoints(collapse_ascii_whitespace(text))) {
switch (cp) {
case 0xFF01: cp = 0x0021; break;
case 0xFF1F: cp = 0x003F; break;
case 0xFF0C: cp = 0x3001; break;
case 0xFF0E: cp = 0x3002; break;
case 0xFF61: cp = 0x3002; break;
case 0xFF64: cp = 0x3001; break;
case 0xFF62: cp = 0x300C; break;
case 0xFF63: cp = 0x300D; break;
case 0x3000: cp = 0x0020; break;
default: break;
}
append_utf8(mapped, cp);
}
mapped = replace_all(std::move(mapped), "……", "…");
mapped = replace_all(std::move(mapped), "...", "…");
mapped = replace_all(std::move(mapped), "..", "…");
return collapse_ascii_whitespace(mapped);
}
} // namespace engine::text