1// Copyright 2011 the V8 project authors. All rights reserved.
2// Redistribution and use in source and binary forms, with or without
3// modification, are permitted provided that the following conditions are
4// met:
5//
6//     * Redistributions of source code must retain the above copyright
7//       notice, this list of conditions and the following disclaimer.
8//     * Redistributions in binary form must reproduce the above
9//       copyright notice, this list of conditions and the following
10//       disclaimer in the documentation and/or other materials provided
11//       with the distribution.
12//     * Neither the name of Google Inc. nor the names of its
13//       contributors may be used to endorse or promote products derived
14//       from this software without specific prior written permission.
15//
16// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
17// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
18// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
19// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
20// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
21// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
22// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
26// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
27
28#include <math.h>
29
30#include "../include/v8stdint.h"
31#include "checks.h"
32#include "utils.h"
33
34#include "double.h"
35#include "fixed-dtoa.h"
36
37namespace v8 {
38namespace internal {
39
40// Represents a 128bit type. This class should be replaced by a native type on
41// platforms that support 128bit integers.
42class UInt128 {
43 public:
44  UInt128() : high_bits_(0), low_bits_(0) { }
45  UInt128(uint64_t high, uint64_t low) : high_bits_(high), low_bits_(low) { }
46
47  void Multiply(uint32_t multiplicand) {
48    uint64_t accumulator;
49
50    accumulator = (low_bits_ & kMask32) * multiplicand;
51    uint32_t part = static_cast<uint32_t>(accumulator & kMask32);
52    accumulator >>= 32;
53    accumulator = accumulator + (low_bits_ >> 32) * multiplicand;
54    low_bits_ = (accumulator << 32) + part;
55    accumulator >>= 32;
56    accumulator = accumulator + (high_bits_ & kMask32) * multiplicand;
57    part = static_cast<uint32_t>(accumulator & kMask32);
58    accumulator >>= 32;
59    accumulator = accumulator + (high_bits_ >> 32) * multiplicand;
60    high_bits_ = (accumulator << 32) + part;
61    ASSERT((accumulator >> 32) == 0);
62  }
63
64  void Shift(int shift_amount) {
65    ASSERT(-64 <= shift_amount && shift_amount <= 64);
66    if (shift_amount == 0) {
67      return;
68    } else if (shift_amount == -64) {
69      high_bits_ = low_bits_;
70      low_bits_ = 0;
71    } else if (shift_amount == 64) {
72      low_bits_ = high_bits_;
73      high_bits_ = 0;
74    } else if (shift_amount <= 0) {
75      high_bits_ <<= -shift_amount;
76      high_bits_ += low_bits_ >> (64 + shift_amount);
77      low_bits_ <<= -shift_amount;
78    } else {
79      low_bits_ >>= shift_amount;
80      low_bits_ += high_bits_ << (64 - shift_amount);
81      high_bits_ >>= shift_amount;
82    }
83  }
84
85  // Modifies *this to *this MOD (2^power).
86  // Returns *this DIV (2^power).
87  int DivModPowerOf2(int power) {
88    if (power >= 64) {
89      int result = static_cast<int>(high_bits_ >> (power - 64));
90      high_bits_ -= static_cast<uint64_t>(result) << (power - 64);
91      return result;
92    } else {
93      uint64_t part_low = low_bits_ >> power;
94      uint64_t part_high = high_bits_ << (64 - power);
95      int result = static_cast<int>(part_low + part_high);
96      high_bits_ = 0;
97      low_bits_ -= part_low << power;
98      return result;
99    }
100  }
101
102  bool IsZero() const {
103    return high_bits_ == 0 && low_bits_ == 0;
104  }
105
106  int BitAt(int position) {
107    if (position >= 64) {
108      return static_cast<int>(high_bits_ >> (position - 64)) & 1;
109    } else {
110      return static_cast<int>(low_bits_ >> position) & 1;
111    }
112  }
113
114 private:
115  static const uint64_t kMask32 = 0xFFFFFFFF;
116  // Value == (high_bits_ << 64) + low_bits_
117  uint64_t high_bits_;
118  uint64_t low_bits_;
119};
120
121
122static const int kDoubleSignificandSize = 53;  // Includes the hidden bit.
123
124
125static void FillDigits32FixedLength(uint32_t number, int requested_length,
126                                    Vector<char> buffer, int* length) {
127  for (int i = requested_length - 1; i >= 0; --i) {
128    buffer[(*length) + i] = '0' + number % 10;
129    number /= 10;
130  }
131  *length += requested_length;
132}
133
134
135static void FillDigits32(uint32_t number, Vector<char> buffer, int* length) {
136  int number_length = 0;
137  // We fill the digits in reverse order and exchange them afterwards.
138  while (number != 0) {
139    int digit = number % 10;
140    number /= 10;
141    buffer[(*length) + number_length] = '0' + digit;
142    number_length++;
143  }
144  // Exchange the digits.
145  int i = *length;
146  int j = *length + number_length - 1;
147  while (i < j) {
148    char tmp = buffer[i];
149    buffer[i] = buffer[j];
150    buffer[j] = tmp;
151    i++;
152    j--;
153  }
154  *length += number_length;
155}
156
157
158static void FillDigits64FixedLength(uint64_t number, int requested_length,
159                                    Vector<char> buffer, int* length) {
160  const uint32_t kTen7 = 10000000;
161  // For efficiency cut the number into 3 uint32_t parts, and print those.
162  uint32_t part2 = static_cast<uint32_t>(number % kTen7);
163  number /= kTen7;
164  uint32_t part1 = static_cast<uint32_t>(number % kTen7);
165  uint32_t part0 = static_cast<uint32_t>(number / kTen7);
166
167  FillDigits32FixedLength(part0, 3, buffer, length);
168  FillDigits32FixedLength(part1, 7, buffer, length);
169  FillDigits32FixedLength(part2, 7, buffer, length);
170}
171
172
173static void FillDigits64(uint64_t number, Vector<char> buffer, int* length) {
174  const uint32_t kTen7 = 10000000;
175  // For efficiency cut the number into 3 uint32_t parts, and print those.
176  uint32_t part2 = static_cast<uint32_t>(number % kTen7);
177  number /= kTen7;
178  uint32_t part1 = static_cast<uint32_t>(number % kTen7);
179  uint32_t part0 = static_cast<uint32_t>(number / kTen7);
180
181  if (part0 != 0) {
182    FillDigits32(part0, buffer, length);
183    FillDigits32FixedLength(part1, 7, buffer, length);
184    FillDigits32FixedLength(part2, 7, buffer, length);
185  } else if (part1 != 0) {
186    FillDigits32(part1, buffer, length);
187    FillDigits32FixedLength(part2, 7, buffer, length);
188  } else {
189    FillDigits32(part2, buffer, length);
190  }
191}
192
193
194static void RoundUp(Vector<char> buffer, int* length, int* decimal_point) {
195  // An empty buffer represents 0.
196  if (*length == 0) {
197    buffer[0] = '1';
198    *decimal_point = 1;
199    *length = 1;
200    return;
201  }
202  // Round the last digit until we either have a digit that was not '9' or until
203  // we reached the first digit.
204  buffer[(*length) - 1]++;
205  for (int i = (*length) - 1; i > 0; --i) {
206    if (buffer[i] != '0' + 10) {
207      return;
208    }
209    buffer[i] = '0';
210    buffer[i - 1]++;
211  }
212  // If the first digit is now '0' + 10, we would need to set it to '0' and add
213  // a '1' in front. However we reach the first digit only if all following
214  // digits had been '9' before rounding up. Now all trailing digits are '0' and
215  // we simply switch the first digit to '1' and update the decimal-point
216  // (indicating that the point is now one digit to the right).
217  if (buffer[0] == '0' + 10) {
218    buffer[0] = '1';
219    (*decimal_point)++;
220  }
221}
222
223
224// The given fractionals number represents a fixed-point number with binary
225// point at bit (-exponent).
226// Preconditions:
227//   -128 <= exponent <= 0.
228//   0 <= fractionals * 2^exponent < 1
229//   The buffer holds the result.
230// The function will round its result. During the rounding-process digits not
231// generated by this function might be updated, and the decimal-point variable
232// might be updated. If this function generates the digits 99 and the buffer
233// already contained "199" (thus yielding a buffer of "19999") then a
234// rounding-up will change the contents of the buffer to "20000".
235static void FillFractionals(uint64_t fractionals, int exponent,
236                            int fractional_count, Vector<char> buffer,
237                            int* length, int* decimal_point) {
238  ASSERT(-128 <= exponent && exponent <= 0);
239  // 'fractionals' is a fixed-point number, with binary point at bit
240  // (-exponent). Inside the function the non-converted remainder of fractionals
241  // is a fixed-point number, with binary point at bit 'point'.
242  if (-exponent <= 64) {
243    // One 64 bit number is sufficient.
244    ASSERT(fractionals >> 56 == 0);
245    int point = -exponent;
246    for (int i = 0; i < fractional_count; ++i) {
247      if (fractionals == 0) break;
248      // Instead of multiplying by 10 we multiply by 5 and adjust the point
249      // location. This way the fractionals variable will not overflow.
250      // Invariant at the beginning of the loop: fractionals < 2^point.
251      // Initially we have: point <= 64 and fractionals < 2^56
252      // After each iteration the point is decremented by one.
253      // Note that 5^3 = 125 < 128 = 2^7.
254      // Therefore three iterations of this loop will not overflow fractionals
255      // (even without the subtraction at the end of the loop body). At this
256      // time point will satisfy point <= 61 and therefore fractionals < 2^point
257      // and any further multiplication of fractionals by 5 will not overflow.
258      fractionals *= 5;
259      point--;
260      int digit = static_cast<int>(fractionals >> point);
261      buffer[*length] = '0' + digit;
262      (*length)++;
263      fractionals -= static_cast<uint64_t>(digit) << point;
264    }
265    // If the first bit after the point is set we have to round up.
266    if (((fractionals >> (point - 1)) & 1) == 1) {
267      RoundUp(buffer, length, decimal_point);
268    }
269  } else {  // We need 128 bits.
270    ASSERT(64 < -exponent && -exponent <= 128);
271    UInt128 fractionals128 = UInt128(fractionals, 0);
272    fractionals128.Shift(-exponent - 64);
273    int point = 128;
274    for (int i = 0; i < fractional_count; ++i) {
275      if (fractionals128.IsZero()) break;
276      // As before: instead of multiplying by 10 we multiply by 5 and adjust the
277      // point location.
278      // This multiplication will not overflow for the same reasons as before.
279      fractionals128.Multiply(5);
280      point--;
281      int digit = fractionals128.DivModPowerOf2(point);
282      buffer[*length] = '0' + digit;
283      (*length)++;
284    }
285    if (fractionals128.BitAt(point - 1) == 1) {
286      RoundUp(buffer, length, decimal_point);
287    }
288  }
289}
290
291
292// Removes leading and trailing zeros.
293// If leading zeros are removed then the decimal point position is adjusted.
294static void TrimZeros(Vector<char> buffer, int* length, int* decimal_point) {
295  while (*length > 0 && buffer[(*length) - 1] == '0') {
296    (*length)--;
297  }
298  int first_non_zero = 0;
299  while (first_non_zero < *length && buffer[first_non_zero] == '0') {
300    first_non_zero++;
301  }
302  if (first_non_zero != 0) {
303    for (int i = first_non_zero; i < *length; ++i) {
304      buffer[i - first_non_zero] = buffer[i];
305    }
306    *length -= first_non_zero;
307    *decimal_point -= first_non_zero;
308  }
309}
310
311
312bool FastFixedDtoa(double v,
313                   int fractional_count,
314                   Vector<char> buffer,
315                   int* length,
316                   int* decimal_point) {
317  const uint32_t kMaxUInt32 = 0xFFFFFFFF;
318  uint64_t significand = Double(v).Significand();
319  int exponent = Double(v).Exponent();
320  // v = significand * 2^exponent (with significand a 53bit integer).
321  // If the exponent is larger than 20 (i.e. we may have a 73bit number) then we
322  // don't know how to compute the representation. 2^73 ~= 9.5*10^21.
323  // If necessary this limit could probably be increased, but we don't need
324  // more.
325  if (exponent > 20) return false;
326  if (fractional_count > 20) return false;
327  *length = 0;
328  // At most kDoubleSignificandSize bits of the significand are non-zero.
329  // Given a 64 bit integer we have 11 0s followed by 53 potentially non-zero
330  // bits:  0..11*..0xxx..53*..xx
331  if (exponent + kDoubleSignificandSize > 64) {
332    // The exponent must be > 11.
333    //
334    // We know that v = significand * 2^exponent.
335    // And the exponent > 11.
336    // We simplify the task by dividing v by 10^17.
337    // The quotient delivers the first digits, and the remainder fits into a 64
338    // bit number.
339    // Dividing by 10^17 is equivalent to dividing by 5^17*2^17.
340    const uint64_t kFive17 = V8_2PART_UINT64_C(0xB1, A2BC2EC5);  // 5^17
341    uint64_t divisor = kFive17;
342    int divisor_power = 17;
343    uint64_t dividend = significand;
344    uint32_t quotient;
345    uint64_t remainder;
346    // Let v = f * 2^e with f == significand and e == exponent.
347    // Then need q (quotient) and r (remainder) as follows:
348    //   v            = q * 10^17       + r
349    //   f * 2^e      = q * 10^17       + r
350    //   f * 2^e      = q * 5^17 * 2^17 + r
351    // If e > 17 then
352    //   f * 2^(e-17) = q * 5^17        + r/2^17
353    // else
354    //   f  = q * 5^17 * 2^(17-e) + r/2^e
355    if (exponent > divisor_power) {
356      // We only allow exponents of up to 20 and therefore (17 - e) <= 3
357      dividend <<= exponent - divisor_power;
358      quotient = static_cast<uint32_t>(dividend / divisor);
359      remainder = (dividend % divisor) << divisor_power;
360    } else {
361      divisor <<= divisor_power - exponent;
362      quotient = static_cast<uint32_t>(dividend / divisor);
363      remainder = (dividend % divisor) << exponent;
364    }
365    FillDigits32(quotient, buffer, length);
366    FillDigits64FixedLength(remainder, divisor_power, buffer, length);
367    *decimal_point = *length;
368  } else if (exponent >= 0) {
369    // 0 <= exponent <= 11
370    significand <<= exponent;
371    FillDigits64(significand, buffer, length);
372    *decimal_point = *length;
373  } else if (exponent > -kDoubleSignificandSize) {
374    // We have to cut the number.
375    uint64_t integrals = significand >> -exponent;
376    uint64_t fractionals = significand - (integrals << -exponent);
377    if (integrals > kMaxUInt32) {
378      FillDigits64(integrals, buffer, length);
379    } else {
380      FillDigits32(static_cast<uint32_t>(integrals), buffer, length);
381    }
382    *decimal_point = *length;
383    FillFractionals(fractionals, exponent, fractional_count,
384                    buffer, length, decimal_point);
385  } else if (exponent < -128) {
386    // This configuration (with at most 20 digits) means that all digits must be
387    // 0.
388    ASSERT(fractional_count <= 20);
389    buffer[0] = '\0';
390    *length = 0;
391    *decimal_point = -fractional_count;
392  } else {
393    *decimal_point = 0;
394    FillFractionals(significand, exponent, fractional_count,
395                    buffer, length, decimal_point);
396  }
397  TrimZeros(buffer, length, decimal_point);
398  buffer[*length] = '\0';
399  if ((*length) == 0) {
400    // The string is empty and the decimal_point thus has no importance. Mimick
401    // Gay's dtoa and and set it to -fractional_count.
402    *decimal_point = -fractional_count;
403  }
404  return true;
405}
406
407} }  // namespace v8::internal
408