summaryrefslogtreecommitdiff
path: root/absl/strings/internal/str_format/arg.cc
blob: e28a29b1716924f3edfa4d2b9902c8b8ad3ac3a1 (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
// Copyright 2020 The Abseil Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//      https://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

//
// POSIX spec:
//   http://pubs.opengroup.org/onlinepubs/009695399/functions/fprintf.html
//
#include "absl/strings/internal/str_format/arg.h"

#include <cassert>
#include <cerrno>
#include <cstdlib>
#include <string>
#include <type_traits>

#include "absl/base/port.h"
#include "absl/strings/internal/str_format/float_conversion.h"
#include "absl/strings/numbers.h"

namespace absl {
ABSL_NAMESPACE_BEGIN
namespace str_format_internal {
namespace {

// Reduce *capacity by s.size(), clipped to a 0 minimum.
void ReducePadding(string_view s, size_t *capacity) {
  *capacity = Excess(s.size(), *capacity);
}

// Reduce *capacity by n, clipped to a 0 minimum.
void ReducePadding(size_t n, size_t *capacity) {
  *capacity = Excess(n, *capacity);
}

template <typename T>
struct MakeUnsigned : std::make_unsigned<T> {};
template <>
struct MakeUnsigned<absl::int128> {
  using type = absl::uint128;
};
template <>
struct MakeUnsigned<absl::uint128> {
  using type = absl::uint128;
};

template <typename T>
struct IsSigned : std::is_signed<T> {};
template <>
struct IsSigned<absl::int128> : std::true_type {};
template <>
struct IsSigned<absl::uint128> : std::false_type {};

// Integral digit printer.
// Call one of the PrintAs* routines after construction once.
// Use with_neg_and_zero/without_neg_or_zero/is_negative to access the results.
class IntDigits {
 public:
  // Print the unsigned integer as octal.
  // Supports unsigned integral types and uint128.
  template <typename T>
  void PrintAsOct(T v) {
    static_assert(!IsSigned<T>::value, "");
    char *p = storage_ + sizeof(storage_);
    do {
      *--p = static_cast<char>('0' + (static_cast<size_t>(v) & 7));
      v >>= 3;
    } while (v);
    start_ = p;
    size_ = storage_ + sizeof(storage_) - p;
  }

  // Print the signed or unsigned integer as decimal.
  // Supports all integral types.
  template <typename T>
  void PrintAsDec(T v) {
    static_assert(std::is_integral<T>::value, "");
    start_ = storage_;
    size_ = numbers_internal::FastIntToBuffer(v, storage_) - storage_;
  }

  void PrintAsDec(int128 v) {
    auto u = static_cast<uint128>(v);
    bool add_neg = false;
    if (v < 0) {
      add_neg = true;
      u = uint128{} - u;
    }
    PrintAsDec(u, add_neg);
  }

  void PrintAsDec(uint128 v, bool add_neg = false) {
    // This function can be sped up if needed. We can call FastIntToBuffer
    // twice, or fix FastIntToBuffer to support uint128.
    char *p = storage_ + sizeof(storage_);
    do {
      p -= 2;
      numbers_internal::PutTwoDigits(static_cast<size_t>(v % 100), p);
      v /= 100;
    } while (v);
    if (p[0] == '0') {
      // We printed one too many hexits.
      ++p;
    }
    if (add_neg) {
      *--p = '-';
    }
    size_ = storage_ + sizeof(storage_) - p;
    start_ = p;
  }

  // Print the unsigned integer as hex using lowercase.
  // Supports unsigned integral types and uint128.
  template <typename T>
  void PrintAsHexLower(T v) {
    static_assert(!IsSigned<T>::value, "");
    char *p = storage_ + sizeof(storage_);

    do {
      p -= 2;
      constexpr const char* table = numbers_internal::kHexTable;
      std::memcpy(p, table + 2 * (static_cast<size_t>(v) & 0xFF), 2);
      if (sizeof(T) == 1) break;
      v >>= 8;
    } while (v);
    if (p[0] == '0') {
      // We printed one too many digits.
      ++p;
    }
    start_ = p;
    size_ = storage_ + sizeof(storage_) - p;
  }

  // Print the unsigned integer as hex using uppercase.
  // Supports unsigned integral types and uint128.
  template <typename T>
  void PrintAsHexUpper(T v) {
    static_assert(!IsSigned<T>::value, "");
    char *p = storage_ + sizeof(storage_);

    // kHexTable is only lowercase, so do it manually for uppercase.
    do {
      *--p = "0123456789ABCDEF"[static_cast<size_t>(v) & 15];
      v >>= 4;
    } while (v);
    start_ = p;
    size_ = storage_ + sizeof(storage_) - p;
  }

  // The printed value including the '-' sign if available.
  // For inputs of value `0`, this will return "0"
  string_view with_neg_and_zero() const { return {start_, size_}; }

  // The printed value not including the '-' sign.
  // For inputs of value `0`, this will return "".
  string_view without_neg_or_zero() const {
    static_assert('-' < '0', "The check below verifies both.");
    size_t advance = start_[0] <= '0' ? 1 : 0;
    return {start_ + advance, size_ - advance};
  }

  bool is_negative() const { return start_[0] == '-'; }

 private:
  const char *start_;
  size_t size_;
  // Max size: 128 bit value as octal -> 43 digits, plus sign char
  char storage_[128 / 3 + 1 + 1];
};

// Note: 'o' conversions do not have a base indicator, it's just that
// the '#' flag is specified to modify the precision for 'o' conversions.
string_view BaseIndicator(const IntDigits &as_digits,
                          const FormatConversionSpecImpl conv) {
  // always show 0x for %p.
  bool alt = conv.has_alt_flag() ||
             conv.conversion_char() == FormatConversionCharInternal::p;
  bool hex = (conv.conversion_char() == FormatConversionCharInternal::x ||
              conv.conversion_char() == FormatConversionCharInternal::X ||
              conv.conversion_char() == FormatConversionCharInternal::p);
  // From the POSIX description of '#' flag:
  //   "For x or X conversion specifiers, a non-zero result shall have
  //   0x (or 0X) prefixed to it."
  if (alt && hex && !as_digits.without_neg_or_zero().empty()) {
    return conv.conversion_char() == FormatConversionCharInternal::X ? "0X"
                                                                     : "0x";
  }
  return {};
}

string_view SignColumn(bool neg, const FormatConversionSpecImpl conv) {
  if (conv.conversion_char() == FormatConversionCharInternal::d ||
      conv.conversion_char() == FormatConversionCharInternal::i) {
    if (neg) return "-";
    if (conv.has_show_pos_flag()) return "+";
    if (conv.has_sign_col_flag()) return " ";
  }
  return {};
}

bool ConvertCharImpl(unsigned char v, const FormatConversionSpecImpl conv,
                     FormatSinkImpl *sink) {
  size_t fill = 0;
  if (conv.width() >= 0) fill = conv.width();
  ReducePadding(1, &fill);
  if (!conv.has_left_flag()) sink->Append(fill, ' ');
  sink->Append(1, v);
  if (conv.has_left_flag()) sink->Append(fill, ' ');
  return true;
}

bool ConvertIntImplInnerSlow(const IntDigits &as_digits,
                             const FormatConversionSpecImpl conv,
                             FormatSinkImpl *sink) {
  // Print as a sequence of Substrings:
  //   [left_spaces][sign][base_indicator][zeroes][formatted][right_spaces]
  size_t fill = 0;
  if (conv.width() >= 0) fill = conv.width();

  string_view formatted = as_digits.without_neg_or_zero();
  ReducePadding(formatted, &fill);

  string_view sign = SignColumn(as_digits.is_negative(), conv);
  ReducePadding(sign, &fill);

  string_view base_indicator = BaseIndicator(as_digits, conv);
  ReducePadding(base_indicator, &fill);

  int precision = conv.precision();
  bool precision_specified = precision >= 0;
  if (!precision_specified)
    precision = 1;

  if (conv.has_alt_flag() &&
      conv.conversion_char() == FormatConversionCharInternal::o) {
    // From POSIX description of the '#' (alt) flag:
    //   "For o conversion, it increases the precision (if necessary) to
    //   force the first digit of the result to be zero."
    if (formatted.empty() || *formatted.begin() != '0') {
      int needed = static_cast<int>(formatted.size()) + 1;
      precision = std::max(precision, needed);
    }
  }

  size_t num_zeroes = Excess(formatted.size(), precision);
  ReducePadding(num_zeroes, &fill);

  size_t num_left_spaces = !conv.has_left_flag() ? fill : 0;
  size_t num_right_spaces = conv.has_left_flag() ? fill : 0;

  // From POSIX description of the '0' (zero) flag:
  //   "For d, i, o, u, x, and X conversion specifiers, if a precision
  //   is specified, the '0' flag is ignored."
  if (!precision_specified && conv.has_zero_flag()) {
    num_zeroes += num_left_spaces;
    num_left_spaces = 0;
  }

  sink->Append(num_left_spaces, ' ');
  sink->Append(sign);
  sink->Append(base_indicator);
  sink->Append(num_zeroes, '0');
  sink->Append(formatted);
  sink->Append(num_right_spaces, ' ');
  return true;
}

template <typename T>
bool ConvertIntArg(T v, const FormatConversionSpecImpl conv,
                   FormatSinkImpl *sink) {
  using U = typename MakeUnsigned<T>::type;
  IntDigits as_digits;

  // This odd casting is due to a bug in -Wswitch behavior in gcc49 which causes
  // it to complain about a switch/case type mismatch, even though both are
  // FormatConverionChar.  Likely this is because at this point
  // FormatConversionChar is declared, but not defined.
  switch (static_cast<uint8_t>(conv.conversion_char())) {
    case static_cast<uint8_t>(FormatConversionCharInternal::c):
      return ConvertCharImpl(static_cast<unsigned char>(v), conv, sink);

    case static_cast<uint8_t>(FormatConversionCharInternal::o):
      as_digits.PrintAsOct(static_cast<U>(v));
      break;

    case static_cast<uint8_t>(FormatConversionCharInternal::x):
      as_digits.PrintAsHexLower(static_cast<U>(v));
      break;
    case static_cast<uint8_t>(FormatConversionCharInternal::X):
      as_digits.PrintAsHexUpper(static_cast<U>(v));
      break;

    case static_cast<uint8_t>(FormatConversionCharInternal::u):
      as_digits.PrintAsDec(static_cast<U>(v));
      break;

    case static_cast<uint8_t>(FormatConversionCharInternal::d):
    case static_cast<uint8_t>(FormatConversionCharInternal::i):
      as_digits.PrintAsDec(v);
      break;

    case static_cast<uint8_t>(FormatConversionCharInternal::a):
    case static_cast<uint8_t>(FormatConversionCharInternal::e):
    case static_cast<uint8_t>(FormatConversionCharInternal::f):
    case static_cast<uint8_t>(FormatConversionCharInternal::g):
    case static_cast<uint8_t>(FormatConversionCharInternal::A):
    case static_cast<uint8_t>(FormatConversionCharInternal::E):
    case static_cast<uint8_t>(FormatConversionCharInternal::F):
    case static_cast<uint8_t>(FormatConversionCharInternal::G):
      return ConvertFloatImpl(static_cast<double>(v), conv, sink);

    default:
       ABSL_INTERNAL_ASSUME(false);
  }

  if (conv.is_basic()) {
    sink->Append(as_digits.with_neg_and_zero());
    return true;
  }
  return ConvertIntImplInnerSlow(as_digits, conv, sink);
}

template <typename T>
bool ConvertFloatArg(T v, const FormatConversionSpecImpl conv,
                     FormatSinkImpl *sink) {
  return FormatConversionCharIsFloat(conv.conversion_char()) &&
         ConvertFloatImpl(v, conv, sink);
}

inline bool ConvertStringArg(string_view v, const FormatConversionSpecImpl conv,
                             FormatSinkImpl *sink) {
  if (conv.is_basic()) {
    sink->Append(v);
    return true;
  }
  return sink->PutPaddedString(v, conv.width(), conv.precision(),
                               conv.has_left_flag());
}

}  // namespace

// ==================== Strings ====================
StringConvertResult FormatConvertImpl(const std::string &v,
                                      const FormatConversionSpecImpl conv,
                                      FormatSinkImpl *sink) {
  return {ConvertStringArg(v, conv, sink)};
}

StringConvertResult FormatConvertImpl(string_view v,
                                      const FormatConversionSpecImpl conv,
                                      FormatSinkImpl *sink) {
  return {ConvertStringArg(v, conv, sink)};
}

ArgConvertResult<FormatConversionCharSetUnion(
    FormatConversionCharSetInternal::s, FormatConversionCharSetInternal::p)>
FormatConvertImpl(const char *v, const FormatConversionSpecImpl conv,
                  FormatSinkImpl *sink) {
  if (conv.conversion_char() == FormatConversionCharInternal::p)
    return {FormatConvertImpl(VoidPtr(v), conv, sink).value};
  size_t len;
  if (v == nullptr) {
    len = 0;
  } else if (conv.precision() < 0) {
    len = std::strlen(v);
  } else {
    // If precision is set, we look for the NUL-terminator on the valid range.
    len = std::find(v, v + conv.precision(), '\0') - v;
  }
  return {ConvertStringArg(string_view(v, len), conv, sink)};
}

// ==================== Raw pointers ====================
ArgConvertResult<FormatConversionCharSetInternal::p> FormatConvertImpl(
    VoidPtr v, const FormatConversionSpecImpl conv, FormatSinkImpl *sink) {
  if (!v.value) {
    sink->Append("(nil)");
    return {true};
  }
  IntDigits as_digits;
  as_digits.PrintAsHexLower(v.value);
  return {ConvertIntImplInnerSlow(as_digits, conv, sink)};
}

// ==================== Floats ====================
FloatingConvertResult FormatConvertImpl(float v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertFloatArg(v, conv, sink)};
}
FloatingConvertResult FormatConvertImpl(double v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertFloatArg(v, conv, sink)};
}
FloatingConvertResult FormatConvertImpl(long double v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertFloatArg(v, conv, sink)};
}

// ==================== Chars ====================
IntegralConvertResult FormatConvertImpl(char v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(signed char v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(unsigned char v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}

// ==================== Ints ====================
IntegralConvertResult FormatConvertImpl(short v,  // NOLINT
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(unsigned short v,  // NOLINT
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(int v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(unsigned v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(long v,  // NOLINT
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(unsigned long v,  // NOLINT
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(long long v,  // NOLINT
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(unsigned long long v,  // NOLINT
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(absl::int128 v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}
IntegralConvertResult FormatConvertImpl(absl::uint128 v,
                                        const FormatConversionSpecImpl conv,
                                        FormatSinkImpl *sink) {
  return {ConvertIntArg(v, conv, sink)};
}

ABSL_INTERNAL_FORMAT_DISPATCH_OVERLOADS_EXPAND_();



}  // namespace str_format_internal

ABSL_NAMESPACE_END
}  // namespace absl