summaryrefslogtreecommitdiff
path: root/absl/strings/internal/str_format/float_conversion.cc
blob: 20012b5876cc829eeaca4bbba097a9335b7bee1c (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
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
#include "absl/strings/internal/str_format/float_conversion.h"

#include <string.h>
#include <algorithm>
#include <array>
#include <cassert>
#include <cmath>
#include <limits>
#include <string>

#include "absl/base/attributes.h"
#include "absl/base/internal/bits.h"
#include "absl/base/optimization.h"
#include "absl/meta/type_traits.h"
#include "absl/numeric/int128.h"
#include "absl/types/span.h"

namespace absl {
namespace str_format_internal {

namespace {

// Calculates `10 * (*v) + carry` and stores the result in `*v` and returns
// the carry.
template <typename Int>
inline Int MultiplyBy10WithCarry(Int *v, Int carry) {
  using NextInt = absl::conditional_t<sizeof(Int) == 4, uint64_t, uint128>;
  static_assert(sizeof(void *) >= sizeof(Int),
                "Don't want to use uint128 in 32-bit mode. It is too slow.");
  NextInt tmp = 10 * static_cast<NextInt>(*v) + carry;
  *v = static_cast<Int>(tmp);
  return static_cast<Int>(tmp >> (sizeof(Int) * 8));
}

// Calculates `(2^64 * carry + *v) / 10`.
// Stores the quotient in `*v` and returns the remainder.
// Requires: `0 <= carry <= 9`
inline uint64_t DivideBy10WithCarry(uint64_t *v, uint64_t carry) {
  constexpr uint64_t divisor = 10;
  // 2^64 / divisor = word_quotient + word_remainder / divisor
  constexpr uint64_t word_quotient = (uint64_t{1} << 63) / (divisor / 2);
  constexpr uint64_t word_remainder = uint64_t{} - word_quotient * divisor;

  const uint64_t mod = *v % divisor;
  const uint64_t next_carry = word_remainder * carry + mod;
  *v = *v / divisor + carry * word_quotient + next_carry / divisor;
  return next_carry % divisor;
}

int LeadingZeros(uint64_t v) { return base_internal::CountLeadingZeros64(v); }
int LeadingZeros(uint128 v) {
  auto high = static_cast<uint64_t>(v >> 64);
  auto low = static_cast<uint64_t>(v);
  return high != 0 ? base_internal::CountLeadingZeros64(high)
                   : 64 + base_internal::CountLeadingZeros64(low);
}

int TrailingZeros(uint64_t v) {
  return base_internal::CountTrailingZerosNonZero64(v);
}
int TrailingZeros(uint128 v) {
  auto high = static_cast<uint64_t>(v >> 64);
  auto low = static_cast<uint64_t>(v);
  return low == 0 ? 64 + base_internal::CountTrailingZerosNonZero64(high)
                  : base_internal::CountTrailingZerosNonZero64(low);
}

// The buffer must have an extra digit that is known to not need rounding.
// This is done below by having an extra '0' digit on the left.
void RoundUp(char *last_digit) {
  char *p = last_digit;
  while (*p == '9' || *p == '.') {
    if (*p == '9') *p = '0';
    --p;
  }
  ++*p;
}

void RoundToEven(char *last_digit) {
  char *p = last_digit;
  if (*p == '.') --p;
  if (*p % 2 == 1) RoundUp(p);
}

char *PrintIntegralDigitsFromRightDynamic(uint128 v, Span<uint32_t> array,
                                          int exp, char *p) {
  if (v == 0) {
    *--p = '0';
    return p;
  }

  int w = exp / 32;
  const int offset = exp % 32;
  // Left shift v by exp bits.
  array[w] = static_cast<uint32_t>(v << offset);
  for (v >>= (32 - offset); v; v >>= 32) array[++w] = static_cast<uint32_t>(v);

  // While we have more than one word available, go in chunks of 1e9.
  // We are guaranteed to have at least those many digits.
  // `w` holds the largest populated word, so keep it updated.
  while (w > 0) {
    uint32_t carry = 0;
    for (int i = w; i >= 0; --i) {
      uint64_t tmp = uint64_t{array[i]} + (uint64_t{carry} << 32);
      array[i] = tmp / uint64_t{1000000000};
      carry = tmp % uint64_t{1000000000};
    }
    // If the highest word is now empty, remove it from view.
    if (array[w] == 0) --w;

    for (int i = 0; i < 9; ++i, carry /= 10) {
      *--p = carry % 10 + '0';
    }
  }

  // Print the leftover of the last word.
  for (auto last = array[0]; last != 0; last /= 10) {
    *--p = last % 10 + '0';
  }

  return p;
}

struct FractionalResult {
  const char *end;
  int precision;
};

FractionalResult PrintFractionalDigitsDynamic(uint128 v, Span<uint32_t> array,
                                              char *p, int exp, int precision) {
  int w = exp / 32;
  const int offset = exp % 32;

  // Right shift `v` by `exp` bits.
  array[w] = static_cast<uint32_t>(v << (32 - offset));
  v >>= offset;
  // Make sure we don't overflow the array. We already calculated that non-zero
  // bits fit, so we might not have space for leading zero bits.
  for (int pos = w; v; v >>= 32) array[--pos] = static_cast<uint32_t>(v);

  // Multiply the whole sequence by 10.
  // On each iteration, the leftover carry word is the next digit.
  // `w` holds the largest populated word, so keep it updated.
  for (; w >= 0 && precision > 0; --precision) {
    uint32_t carry = 0;
    for (int i = w; i >= 0; --i) {
      carry = MultiplyBy10WithCarry(&array[i], carry);
    }
    // If the lowest word is now empty, remove it from view.
    if (array[w] == 0) --w;
    *p++ = carry + '0';
  }

  constexpr uint32_t threshold = 0x80000000;
  if (array[0] < threshold) {
    // We round down, so nothing to do.
  } else if (array[0] > threshold ||
             std::any_of(&array[1], &array[w + 1],
                         [](uint32_t word) { return word != 0; })) {
    RoundUp(p - 1);
  } else {
    RoundToEven(p - 1);
  }
  return {p, precision};
}

// Generic digit printer.
// `bits` determines how many bits of termporary space it needs for the
// calcualtions.
template <int bits, typename = void>
class DigitPrinter {
  static constexpr int kInts = (bits + 31) / 32;

 public:
  // Quick upper bound for the number of decimal digits we need.
  // This would be std::ceil(std::log10(std::pow(2, bits))), but that is not
  // constexpr.
  static constexpr int kDigits10 = 1 + (bits + 9) / 10 * 3 + bits / 900;
  using InputType = uint128;

  static char *PrintIntegralDigitsFromRight(InputType v, int exp, char *end) {
    std::array<uint32_t, kInts> array{};
    return PrintIntegralDigitsFromRightDynamic(v, absl::MakeSpan(array), exp,
                                               end);
  }

  static FractionalResult PrintFractionalDigits(InputType v, char *p, int exp,
                                                int precision) {
    std::array<uint32_t, kInts> array{};
    return PrintFractionalDigitsDynamic(v, absl::MakeSpan(array), p, exp,
                                        precision);
  }
};

// Specialiation for 64-bit working space.
// This is a performance optimization over the generic primary template.
// Only enabled in 64-bit platforms. The generic one is faster in 32-bit
// platforms.
template <int bits>
class DigitPrinter<bits, absl::enable_if_t<bits == 64 && (sizeof(void *) >=
                                                          sizeof(uint64_t))>> {
 public:
  static constexpr size_t kDigits10 = 20;
  using InputType = uint64_t;

  static char *PrintIntegralDigitsFromRight(uint64_t v, int exp, char *p) {
    v <<= exp;
    do {
      *--p = DivideBy10WithCarry(&v, 0) + '0';
    } while (v != 0);
    return p;
  }

  static FractionalResult PrintFractionalDigits(uint64_t v, char *p, int exp,
                                                int precision) {
    v <<= (64 - exp);
    while (precision > 0) {
      if (!v) return {p, precision};
      *p++ = MultiplyBy10WithCarry(&v, uint64_t{}) + '0';
      --precision;
    }

    // We need to round.
    if (v < 0x8000000000000000) {
      // We round down, so nothing to do.
    } else if (v > 0x8000000000000000) {
      // We round up.
      RoundUp(p - 1);
    } else {
      RoundToEven(p - 1);
    }

    assert(precision == 0);
    // Precision can only be zero here. Return a constant instead.
    return {p, 0};
  }
};

// Specialiation for 128-bit working space.
// This is a performance optimization over the generic primary template.
template <int bits>
class DigitPrinter<bits, absl::enable_if_t<bits == 128 && (sizeof(void *) >=
                                                           sizeof(uint64_t))>> {
 public:
  static constexpr size_t kDigits10 = 40;
  using InputType = uint128;

  static char *PrintIntegralDigitsFromRight(uint128 v, int exp, char *p) {
    v <<= exp;
    auto high = static_cast<uint64_t>(v >> 64);
    auto low = static_cast<uint64_t>(v);

    do {
      uint64_t carry = DivideBy10WithCarry(&high, 0);
      carry = DivideBy10WithCarry(&low, carry);
      *--p = carry + '0';
    } while (high != 0u);

    while (low != 0u) {
      *--p = DivideBy10WithCarry(&low, 0) + '0';
    }
    return p;
  }

  static FractionalResult PrintFractionalDigits(uint128 v, char *p, int exp,
                                                int precision) {
    v <<= (128 - exp);
    auto high = static_cast<uint64_t>(v >> 64);
    auto low = static_cast<uint64_t>(v);

    // While we have digits to print and `low` is not empty, do the long
    // multiplication.
    while (precision > 0 && low != 0) {
      uint64_t carry = MultiplyBy10WithCarry(&low, uint64_t{});
      carry = MultiplyBy10WithCarry(&high, carry);

      *p++ = carry + '0';
      --precision;
    }

    // Now `low` is empty, so use a faster approach for the rest of the digits.
    // This block is pretty much the same as the main loop for the 64-bit case
    // above.
    while (precision > 0) {
      if (!high) return {p, precision};
      *p++ = MultiplyBy10WithCarry(&high, uint64_t{}) + '0';
      --precision;
    }

    // We need to round.
    if (high < 0x8000000000000000) {
      // We round down, so nothing to do.
    } else if (high > 0x8000000000000000 || low != 0) {
      // We round up.
      RoundUp(p - 1);
    } else {
      RoundToEven(p - 1);
    }

    assert(precision == 0);
    // Precision can only be zero here. Return a constant instead.
    return {p, 0};
  }
};

struct FormatState {
  char sign_char;
  int precision;
  const ConversionSpec &conv;
  FormatSinkImpl *sink;
};

void FinalPrint(string_view data, int trailing_zeros,
                const FormatState &state) {
  if (state.conv.width() < 0) {
    // No width specified. Fast-path.
    if (state.sign_char != '\0') state.sink->Append(1, state.sign_char);
    state.sink->Append(data);
    state.sink->Append(trailing_zeros, '0');
    return;
  }

  int left_spaces = 0, zeros = 0, right_spaces = 0;
  int total_size = (state.sign_char != 0 ? 1 : 0) +
                   static_cast<int>(data.size()) + trailing_zeros;
  int missing_chars = std::max(state.conv.width() - total_size, 0);
  if (state.conv.flags().left) {
    right_spaces = missing_chars;
  } else if (state.conv.flags().zero) {
    zeros = missing_chars;
  } else {
    left_spaces = missing_chars;
  }

  state.sink->Append(left_spaces, ' ');
  if (state.sign_char != '\0') state.sink->Append(1, state.sign_char);
  state.sink->Append(zeros, '0');
  state.sink->Append(data);
  state.sink->Append(trailing_zeros, '0');
  state.sink->Append(right_spaces, ' ');
}

template <int num_bits, typename Int>
void FormatFPositiveExp(Int v, int exp, const FormatState &state) {
  using IntegralPrinter = DigitPrinter<num_bits>;
  char buffer[IntegralPrinter::kDigits10 + /* . */ 1];
  buffer[IntegralPrinter::kDigits10] = '.';

  const char *digits = IntegralPrinter::PrintIntegralDigitsFromRight(
      static_cast<typename IntegralPrinter::InputType>(v), exp,
      buffer + sizeof(buffer) - 1);
  size_t size = buffer + sizeof(buffer) - digits;

  // In `alt` mode (flag #) we keep the `.` even if there are no fractional
  // digits. In non-alt mode, we strip it.
  if (ABSL_PREDICT_FALSE(state.precision == 0 && !state.conv.flags().alt)) {
    --size;
  }

  FinalPrint(string_view(digits, size), state.precision, state);
}

template <int num_bits, typename Int>
void FormatFNegativeExp(Int v, int exp, const FormatState &state) {
  constexpr int input_bits = sizeof(Int) * 8;

  using IntegralPrinter = DigitPrinter<input_bits>;
  using FractionalPrinter = DigitPrinter<num_bits>;

  static constexpr size_t integral_size =
      1 + /* in case we need to round up an extra digit */
      IntegralPrinter::kDigits10 + 1;
  char buffer[integral_size + /* . */ 1 + num_bits];
  buffer[integral_size] = '.';
  char *const integral_digits_end = buffer + integral_size;
  char *integral_digits_start;
  char *const fractional_digits_start = buffer + integral_size + 1;

  if (exp < input_bits) {
    integral_digits_start = IntegralPrinter::PrintIntegralDigitsFromRight(
        v >> exp, 0, integral_digits_end);
  } else {
    integral_digits_start = integral_digits_end - 1;
    *integral_digits_start = '0';
  }

  // PrintFractionalDigits may pull a carried 1 all the way up through the
  // integral portion.
  integral_digits_start[-1] = '0';
  auto fractional_result = FractionalPrinter::PrintFractionalDigits(
      static_cast<typename FractionalPrinter::InputType>(v),
      fractional_digits_start, exp, state.precision);
  if (integral_digits_start[-1] != '0') --integral_digits_start;

  size_t size = fractional_result.end - integral_digits_start;

  // In `alt` mode (flag #) we keep the `.` even if there are no fractional
  // digits. In non-alt mode, we strip it.
  if (ABSL_PREDICT_FALSE(state.precision == 0 && !state.conv.flags().alt)) {
    --size;
  }
  FinalPrint(string_view(integral_digits_start, size),
             fractional_result.precision, state);
}

template <typename Int>
void FormatF(Int mantissa, int exp, const FormatState &state) {
  // Remove trailing zeros as they are not useful.
  // This helps use faster implementations/less stack space in some cases.
  if (mantissa != 0) {
    int trailing = TrailingZeros(mantissa);
    mantissa >>= trailing;
    exp += trailing;
  }

  // The table driven dispatch gives us two benefits: fast distpatch and
  // prevent inlining.
  // We must not inline any of the functions below (other than the ones for
  // 64-bit) to avoid blowing up this stack frame.

  if (exp >= 0) {
    // We will left shift the mantissa. Calculate how many bits we need.
    // Special case 64-bit as we will use a uint64_t for it. Use a table for the
    // rest and unconditionally use uint128.
    const int total_bits = sizeof(Int) * 8 - LeadingZeros(mantissa) + exp;

    if (total_bits <= 64) {
      return FormatFPositiveExp<64>(mantissa, exp, state);
    } else {
      using Formatter = void (*)(uint128, int, const FormatState &);
      static constexpr Formatter kFormatters[] = {
          FormatFPositiveExp<1 << 7>,  FormatFPositiveExp<1 << 8>,
          FormatFPositiveExp<1 << 9>,  FormatFPositiveExp<1 << 10>,
          FormatFPositiveExp<1 << 11>, FormatFPositiveExp<1 << 12>,
          FormatFPositiveExp<1 << 13>, FormatFPositiveExp<1 << 14>,
          FormatFPositiveExp<1 << 15>,
      };
      static constexpr int max_total_bits =
          sizeof(Int) * 8 + std::numeric_limits<long double>::max_exponent;
      assert(total_bits <= max_total_bits);
      static_assert(max_total_bits <= (1 << 15), "");
      const int log2 =
          64 - LeadingZeros((static_cast<uint64_t>(total_bits) - 1) / 128);
      assert(log2 < std::end(kFormatters) - std::begin(kFormatters));
      kFormatters[log2](mantissa, exp, state);
    }
  } else {
    exp = -exp;

    // We know we don't need more than Int itself for the integral part.
    // We need `precision` fractional digits, but there are at most `exp`
    // non-zero digits after the decimal point. The rest will be zeros.
    // Special case 64-bit as we will use a uint64_t for it. Use a table for the
    // rest and unconditionally use uint128.

    if (exp <= 64) {
      return FormatFNegativeExp<64>(mantissa, exp, state);
    } else {
      using Formatter = void (*)(uint128, int, const FormatState &);
      static constexpr Formatter kFormatters[] = {
          FormatFNegativeExp<1 << 7>,  FormatFNegativeExp<1 << 8>,
          FormatFNegativeExp<1 << 9>,  FormatFNegativeExp<1 << 10>,
          FormatFNegativeExp<1 << 11>, FormatFNegativeExp<1 << 12>,
          FormatFNegativeExp<1 << 13>, FormatFNegativeExp<1 << 14>};
      static_assert(
          -std::numeric_limits<long double>::min_exponent <= (1 << 14), "");
      const int log2 =
          64 - LeadingZeros((static_cast<uint64_t>(exp) - 1) / 128);
      assert(log2 < std::end(kFormatters) - std::begin(kFormatters));
      kFormatters[log2](mantissa, exp, state);
    }
  }
}

char *CopyStringTo(string_view v, char *out) {
  std::memcpy(out, v.data(), v.size());
  return out + v.size();
}

template <typename Float>
bool FallbackToSnprintf(const Float v, const ConversionSpec &conv,
                        FormatSinkImpl *sink) {
  int w = conv.width() >= 0 ? conv.width() : 0;
  int p = conv.precision() >= 0 ? conv.precision() : -1;
  char fmt[32];
  {
    char *fp = fmt;
    *fp++ = '%';
    fp = CopyStringTo(conv.flags().ToString(), fp);
    fp = CopyStringTo("*.*", fp);
    if (std::is_same<long double, Float>()) {
      *fp++ = 'L';
    }
    *fp++ = conv.conv().Char();
    *fp = 0;
    assert(fp < fmt + sizeof(fmt));
  }
  std::string space(512, '\0');
  string_view result;
  while (true) {
    int n = snprintf(&space[0], space.size(), fmt, w, p, v);
    if (n < 0) return false;
    if (static_cast<size_t>(n) < space.size()) {
      result = string_view(space.data(), n);
      break;
    }
    space.resize(n + 1);
  }
  sink->Append(result);
  return true;
}

// 128-bits in decimal: ceil(128*log(2)/log(10))
//   or std::numeric_limits<__uint128_t>::digits10
constexpr int kMaxFixedPrecision = 39;

constexpr int kBufferLength = /*sign*/ 1 +
                              /*integer*/ kMaxFixedPrecision +
                              /*point*/ 1 +
                              /*fraction*/ kMaxFixedPrecision +
                              /*exponent e+123*/ 5;

struct Buffer {
  void push_front(char c) {
    assert(begin > data);
    *--begin = c;
  }
  void push_back(char c) {
    assert(end < data + sizeof(data));
    *end++ = c;
  }
  void pop_back() {
    assert(begin < end);
    --end;
  }

  char &back() {
    assert(begin < end);
    return end[-1];
  }

  char last_digit() const { return end[-1] == '.' ? end[-2] : end[-1]; }

  int size() const { return static_cast<int>(end - begin); }

  char data[kBufferLength];
  char *begin;
  char *end;
};

enum class FormatStyle { Fixed, Precision };

// If the value is Inf or Nan, print it and return true.
// Otherwise, return false.
template <typename Float>
bool ConvertNonNumericFloats(char sign_char, Float v,
                             const ConversionSpec &conv, FormatSinkImpl *sink) {
  char text[4], *ptr = text;
  if (sign_char != '\0') *ptr++ = sign_char;
  if (std::isnan(v)) {
    ptr = std::copy_n(conv.conv().upper() ? "NAN" : "nan", 3, ptr);
  } else if (std::isinf(v)) {
    ptr = std::copy_n(conv.conv().upper() ? "INF" : "inf", 3, ptr);
  } else {
    return false;
  }

  return sink->PutPaddedString(string_view(text, ptr - text), conv.width(), -1,
                               conv.flags().left);
}

// Round up the last digit of the value.
// It will carry over and potentially overflow. 'exp' will be adjusted in that
// case.
template <FormatStyle mode>
void RoundUp(Buffer *buffer, int *exp) {
  char *p = &buffer->back();
  while (p >= buffer->begin && (*p == '9' || *p == '.')) {
    if (*p == '9') *p = '0';
    --p;
  }

  if (p < buffer->begin) {
    *p = '1';
    buffer->begin = p;
    if (mode == FormatStyle::Precision) {
      std::swap(p[1], p[2]);  // move the .
      ++*exp;
      buffer->pop_back();
    }
  } else {
    ++*p;
  }
}

void PrintExponent(int exp, char e, Buffer *out) {
  out->push_back(e);
  if (exp < 0) {
    out->push_back('-');
    exp = -exp;
  } else {
    out->push_back('+');
  }
  // Exponent digits.
  if (exp > 99) {
    out->push_back(exp / 100 + '0');
    out->push_back(exp / 10 % 10 + '0');
    out->push_back(exp % 10 + '0');
  } else {
    out->push_back(exp / 10 + '0');
    out->push_back(exp % 10 + '0');
  }
}

template <typename Float, typename Int>
constexpr bool CanFitMantissa() {
  return
#if defined(__clang__) && !defined(__SSE3__)
      // Workaround for clang bug: https://bugs.llvm.org/show_bug.cgi?id=38289
      // Casting from long double to uint64_t is miscompiled and drops bits.
      (!std::is_same<Float, long double>::value ||
       !std::is_same<Int, uint64_t>::value) &&
#endif
      std::numeric_limits<Float>::digits <= std::numeric_limits<Int>::digits;
}

template <typename Float>
struct Decomposed {
  using MantissaType =
      absl::conditional_t<std::is_same<long double, Float>::value, uint128,
                          uint64_t>;
  static_assert(std::numeric_limits<Float>::digits <= sizeof(MantissaType) * 8,
                "");
  MantissaType mantissa;
  int exponent;
};

// Decompose the double into an integer mantissa and an exponent.
template <typename Float>
Decomposed<Float> Decompose(Float v) {
  int exp;
  Float m = std::frexp(v, &exp);
  m = std::ldexp(m, std::numeric_limits<Float>::digits);
  exp -= std::numeric_limits<Float>::digits;

  return {static_cast<typename Decomposed<Float>::MantissaType>(m), exp};
}

// Print 'digits' as decimal.
// In Fixed mode, we add a '.' at the end.
// In Precision mode, we add a '.' after the first digit.
template <FormatStyle mode, typename Int>
int PrintIntegralDigits(Int digits, Buffer *out) {
  int printed = 0;
  if (digits) {
    for (; digits; digits /= 10) out->push_front(digits % 10 + '0');
    printed = out->size();
    if (mode == FormatStyle::Precision) {
      out->push_front(*out->begin);
      out->begin[1] = '.';
    } else {
      out->push_back('.');
    }
  } else if (mode == FormatStyle::Fixed) {
    out->push_front('0');
    out->push_back('.');
    printed = 1;
  }
  return printed;
}

// Back out 'extra_digits' digits and round up if necessary.
bool RemoveExtraPrecision(int extra_digits, bool has_leftover_value,
                          Buffer *out, int *exp_out) {
  if (extra_digits <= 0) return false;

  // Back out the extra digits
  out->end -= extra_digits;

  bool needs_to_round_up = [&] {
    // We look at the digit just past the end.
    // There must be 'extra_digits' extra valid digits after end.
    if (*out->end > '5') return true;
    if (*out->end < '5') return false;
    if (has_leftover_value || std::any_of(out->end + 1, out->end + extra_digits,
                                          [](char c) { return c != '0'; }))
      return true;

    // Ends in ...50*, round to even.
    return out->last_digit() % 2 == 1;
  }();

  if (needs_to_round_up) {
    RoundUp<FormatStyle::Precision>(out, exp_out);
  }
  return true;
}

// Print the value into the buffer.
// This will not include the exponent, which will be returned in 'exp_out' for
// Precision mode.
template <typename Int, typename Float, FormatStyle mode>
bool FloatToBufferImpl(Int int_mantissa, int exp, int precision, Buffer *out,
                       int *exp_out) {
  assert((CanFitMantissa<Float, Int>()));

  const int int_bits = std::numeric_limits<Int>::digits;

  // In precision mode, we start printing one char to the right because it will
  // also include the '.'
  // In fixed mode we put the dot afterwards on the right.
  out->begin = out->end =
      out->data + 1 + kMaxFixedPrecision + (mode == FormatStyle::Precision);

  if (exp >= 0) {
    if (std::numeric_limits<Float>::digits + exp > int_bits) {
      // The value will overflow the Int
      return false;
    }
    int digits_printed = PrintIntegralDigits<mode>(int_mantissa << exp, out);
    int digits_to_zero_pad = precision;
    if (mode == FormatStyle::Precision) {
      *exp_out = digits_printed - 1;
      digits_to_zero_pad -= digits_printed - 1;
      if (RemoveExtraPrecision(-digits_to_zero_pad, false, out, exp_out)) {
        return true;
      }
    }
    for (; digits_to_zero_pad-- > 0;) out->push_back('0');
    return true;
  }

  exp = -exp;
  // We need at least 4 empty bits for the next decimal digit.
  // We will multiply by 10.
  if (exp > int_bits - 4) return false;

  const Int mask = (Int{1} << exp) - 1;

  // Print the integral part first.
  int digits_printed = PrintIntegralDigits<mode>(int_mantissa >> exp, out);
  int_mantissa &= mask;

  int fractional_count = precision;
  if (mode == FormatStyle::Precision) {
    if (digits_printed == 0) {
      // Find the first non-zero digit, when in Precision mode.
      *exp_out = 0;
      if (int_mantissa) {
        while (int_mantissa <= mask) {
          int_mantissa *= 10;
          --*exp_out;
        }
      }
      out->push_front(static_cast<char>(int_mantissa >> exp) + '0');
      out->push_back('.');
      int_mantissa &= mask;
    } else {
      // We already have a digit, and a '.'
      *exp_out = digits_printed - 1;
      fractional_count -= *exp_out;
      if (RemoveExtraPrecision(-fractional_count, int_mantissa != 0, out,
                               exp_out)) {
        // If we had enough digits, return right away.
        // The code below will try to round again otherwise.
        return true;
      }
    }
  }

  auto get_next_digit = [&] {
    int_mantissa *= 10;
    int digit = static_cast<int>(int_mantissa >> exp);
    int_mantissa &= mask;
    return digit;
  };

  // Print fractional_count more digits, if available.
  for (; fractional_count > 0; --fractional_count) {
    out->push_back(get_next_digit() + '0');
  }

  int next_digit = get_next_digit();
  if (next_digit > 5 ||
      (next_digit == 5 && (int_mantissa || out->last_digit() % 2 == 1))) {
    RoundUp<mode>(out, exp_out);
  }

  return true;
}

template <FormatStyle mode, typename Float>
bool FloatToBuffer(Decomposed<Float> decomposed, int precision, Buffer *out,
                   int *exp) {
  if (precision > kMaxFixedPrecision) return false;

  // Try with uint64_t.
  if (CanFitMantissa<Float, std::uint64_t>() &&
      FloatToBufferImpl<std::uint64_t, Float, mode>(
          static_cast<std::uint64_t>(decomposed.mantissa),
          static_cast<std::uint64_t>(decomposed.exponent), precision, out, exp))
    return true;

#if defined(ABSL_HAVE_INTRINSIC_INT128)
  // If that is not enough, try with __uint128_t.
  return CanFitMantissa<Float, __uint128_t>() &&
         FloatToBufferImpl<__uint128_t, Float, mode>(
             static_cast<__uint128_t>(decomposed.mantissa),
             static_cast<__uint128_t>(decomposed.exponent), precision, out,
             exp);
#endif
  return false;
}

void WriteBufferToSink(char sign_char, string_view str,
                       const ConversionSpec &conv, FormatSinkImpl *sink) {
  int left_spaces = 0, zeros = 0, right_spaces = 0;
  int missing_chars =
      conv.width() >= 0 ? std::max(conv.width() - static_cast<int>(str.size()) -
                                       static_cast<int>(sign_char != 0),
                                   0)
                        : 0;
  if (conv.flags().left) {
    right_spaces = missing_chars;
  } else if (conv.flags().zero) {
    zeros = missing_chars;
  } else {
    left_spaces = missing_chars;
  }

  sink->Append(left_spaces, ' ');
  if (sign_char != '\0') sink->Append(1, sign_char);
  sink->Append(zeros, '0');
  sink->Append(str);
  sink->Append(right_spaces, ' ');
}

template <typename Float>
bool FloatToSink(const Float v, const ConversionSpec &conv,
                 FormatSinkImpl *sink) {
  // Print the sign or the sign column.
  Float abs_v = v;
  char sign_char = 0;
  if (std::signbit(abs_v)) {
    sign_char = '-';
    abs_v = -abs_v;
  } else if (conv.flags().show_pos) {
    sign_char = '+';
  } else if (conv.flags().sign_col) {
    sign_char = ' ';
  }

  // Print nan/inf.
  if (ConvertNonNumericFloats(sign_char, abs_v, conv, sink)) {
    return true;
  }

  int precision = conv.precision() < 0 ? 6 : conv.precision();

  int exp = 0;

  auto decomposed = Decompose(abs_v);

  Buffer buffer;

  switch (conv.conv().id()) {
    case ConversionChar::f:
    case ConversionChar::F:
      FormatF(decomposed.mantissa, decomposed.exponent,
              {sign_char, precision, conv, sink});
      return true;

    case ConversionChar::e:
    case ConversionChar::E:
      if (!FloatToBuffer<FormatStyle::Precision>(decomposed, precision, &buffer,
                                                 &exp)) {
        return FallbackToSnprintf(v, conv, sink);
      }
      if (!conv.flags().alt && buffer.back() == '.') buffer.pop_back();
      PrintExponent(exp, conv.conv().upper() ? 'E' : 'e', &buffer);
      break;

    case ConversionChar::g:
    case ConversionChar::G:
      precision = std::max(0, precision - 1);
      if (!FloatToBuffer<FormatStyle::Precision>(decomposed, precision, &buffer,
                                                 &exp)) {
        return FallbackToSnprintf(v, conv, sink);
      }
      if (precision + 1 > exp && exp >= -4) {
        if (exp < 0) {
          // Have 1.23456, needs 0.00123456
          // Move the first digit
          buffer.begin[1] = *buffer.begin;
          // Add some zeros
          for (; exp < -1; ++exp) *buffer.begin-- = '0';
          *buffer.begin-- = '.';
          *buffer.begin = '0';
        } else if (exp > 0) {
          // Have 1.23456, needs 1234.56
          // Move the '.' exp positions to the right.
          std::rotate(buffer.begin + 1, buffer.begin + 2,
                      buffer.begin + exp + 2);
        }
        exp = 0;
      }
      if (!conv.flags().alt) {
        while (buffer.back() == '0') buffer.pop_back();
        if (buffer.back() == '.') buffer.pop_back();
      }
      if (exp) PrintExponent(exp, conv.conv().upper() ? 'E' : 'e', &buffer);
      break;

    case ConversionChar::a:
    case ConversionChar::A:
      return FallbackToSnprintf(v, conv, sink);

    default:
      return false;
  }

  WriteBufferToSink(sign_char,
                    string_view(buffer.begin, buffer.end - buffer.begin), conv,
                    sink);

  return true;
}

}  // namespace

bool ConvertFloatImpl(long double v, const ConversionSpec &conv,
                      FormatSinkImpl *sink) {
  if (std::numeric_limits<long double>::digits ==
      2 * std::numeric_limits<double>::digits) {
    // This is the `double-double` representation of `long double`.
    // We do not handle it natively. Fallback to snprintf.
    return FallbackToSnprintf(v, conv, sink);
  }

  return FloatToSink(v, conv, sink);
}

bool ConvertFloatImpl(float v, const ConversionSpec &conv,
                      FormatSinkImpl *sink) {
  // DivideBy10WithCarry is not actually used in some builds. This here silences
  // the "unused" warning. We just need to put it in any function that is really
  // used.
  (void)&DivideBy10WithCarry;
  return FloatToSink(v, conv, sink);
}

bool ConvertFloatImpl(double v, const ConversionSpec &conv,
                      FormatSinkImpl *sink) {
  return FloatToSink(v, conv, sink);
}

}  // namespace str_format_internal
}  // namespace absl