v8
V8 is Google’s open source high-performance JavaScript and WebAssembly engine, written in C++.
Loading...
Searching...
No Matches
fast-dtoa.h
Go to the documentation of this file.
1// Copyright 2011 the V8 project authors. All rights reserved.
2// Use of this source code is governed by a BSD-style license that can be
3// found in the LICENSE file.
4
5#ifndef V8_BASE_NUMBERS_FAST_DTOA_H_
6#define V8_BASE_NUMBERS_FAST_DTOA_H_
7
8#include "src/base/vector.h"
9
10namespace v8 {
11namespace base {
12
14 // Computes the shortest representation of the given input. The returned
15 // result will be the most accurate number of this length. Longer
16 // representations might be more accurate.
18 // Computes a representation where the precision (number of digits) is
19 // given as input. The precision is independent of the decimal point.
21};
22
23// FastDtoa will produce at most kFastDtoaMaximalLength digits. This does not
24// include the terminating '\0' character.
26
27// Provides a decimal representation of v.
28// The result should be interpreted as buffer * 10^(point - length).
29//
30// Precondition:
31// * v must be a strictly positive finite double.
32//
33// Returns true if it succeeds, otherwise the result can not be trusted.
34// There will be *length digits inside the buffer followed by a null terminator.
35// If the function returns true and mode equals
36// - FAST_DTOA_SHORTEST, then
37// the parameter requested_digits is ignored.
38// The result satisfies
39// v == (double) (buffer * 10^(point - length)).
40// The digits in the buffer are the shortest representation possible. E.g.
41// if 0.099999999999 and 0.1 represent the same double then "1" is returned
42// with point = 0.
43// The last digit will be closest to the actual v. That is, even if several
44// digits might correctly yield 'v' when read again, the buffer will contain
45// the one closest to v.
46// - FAST_DTOA_PRECISION, then
47// the buffer contains requested_digits digits.
48// the difference v - (buffer * 10^(point-length)) is closest to zero for
49// all possible representations of requested_digits digits.
50// If there are two values that are equally close, then FastDtoa returns
51// false.
52// For both modes the buffer must be large enough to hold the result.
53V8_BASE_EXPORT bool FastDtoa(double d, FastDtoaMode mode, int requested_digits,
54 Vector<char> buffer, int* length,
55 int* decimal_point);
56
57} // namespace base
58} // namespace v8
59
60#endif // V8_BASE_NUMBERS_FAST_DTOA_H_
#define V8_BASE_EXPORT
Definition base-export.h:26
const int kFastDtoaMaximalLength
Definition fast-dtoa.h:25
@ FAST_DTOA_SHORTEST
Definition fast-dtoa.h:17
@ FAST_DTOA_PRECISION
Definition fast-dtoa.h:20
bool FastDtoa(double v, FastDtoaMode mode, int requested_digits, Vector< char > buffer, int *length, int *decimal_point)
Definition fast-dtoa.cc:713