coffee/fast_float
1
0
Fork 0
mirror of https://github.com/fastfloat/fast_float.git synced 2026-02-06 17:59:52 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge from upstream/main, fix conflicts

Browse Source
This commit is contained in:
Maya Warrier 2023-05-01 20:27:29 -04:00
commit 680ccc73ed
8 changed files with 236 additions and 141 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
CONTRIBUTORS
View File

@ -5,3 +5,4 @@ Neal Richardson
Tim Paine Tim Paine
Fabio Pellacini Fabio Pellacini
Lénárd Szolnoki Lénárd Szolnoki
Jan Pharago

18
README.md
View File

@ -97,6 +97,24 @@ constexpr double constexptest() {
} }
``` ```
## Non-ASCII Inputs
We also support UTF-16 and UTF-32 inputs, as well as ASCII/UTF-8, as in the following example:
``` C++
#include "fast_float/fast_float.h"
#include <iostream>
int main() {
const std::u16string input = u"3.1416 xyz ";
double result;
auto answer = fast_float::from_chars(input.data(), input.data()+input.size(), result);
if(answer.ec != std::errc()) { std::cerr << "parsing failure\n"; return EXIT_FAILURE; }
std::cout << "parsed the number " << result << std::endl;
return EXIT_SUCCESS;
}
```
## Using commas as decimal separator ## Using commas as decimal separator

94
include/fast_float/ascii_number.h
View File

@ -18,9 +18,9 @@ namespace fast_float {
// Next function can be micro-optimized, but compilers are entirely // Next function can be micro-optimized, but compilers are entirely
// able to optimize it well. // able to optimize it well.
template <typename CharT> template <typename UC>
fastfloat_really_inline constexpr bool is_integer(CharT c) noexcept { fastfloat_really_inline constexpr bool is_integer(UC c) noexcept {
return c >= CharT('0') && c <= CharT('9'); return !(c > UC('9') || c < UC('0'));
} }
fastfloat_really_inline constexpr uint64_t byteswap(uint64_t val) { fastfloat_really_inline constexpr uint64_t byteswap(uint64_t val) {
@ -70,11 +70,11 @@ uint64_t simd_read8_to_u64(const char16_t* chars) {
} }
#endif #endif
// Read 8 CharT into a u64. Truncates CharT if != char. // Read 8 UC into a u64. Truncates UC if not char.
template <typename CharT> template <typename UC>
fastfloat_really_inline FASTFLOAT_CONSTEXPR20 fastfloat_really_inline FASTFLOAT_CONSTEXPR20
uint64_t read8_to_u64(const CharT *chars) { uint64_t read8_to_u64(const UC *chars) {
if (cpp20_and_in_constexpr() || !std::is_same<CharT, char>::value) { if (cpp20_and_in_constexpr() || !std::is_same<UC, char>::value) {
uint64_t val = 0; uint64_t val = 0;
for(int i = 0; i < 8; ++i) { for(int i = 0; i < 8; ++i) {
val |= uint64_t(char(*chars)) << (i*8); val |= uint64_t(char(*chars)) << (i*8);
@ -135,7 +135,7 @@ uint32_t parse_eight_digits_unrolled(const char16_t* chars) noexcept {
#ifdef FASTFLOAT_HAS_SIMD #ifdef FASTFLOAT_HAS_SIMD
return parse_eight_digits_unrolled(simd_read8_to_u64(chars)); return parse_eight_digits_unrolled(simd_read8_to_u64(chars));
#else #else
// never reaches here, remove warning // never reaches here, removes warning
return 0; return 0;
#endif #endif
} }
@ -150,7 +150,7 @@ uint32_t parse_eight_digits_unrolled(const char32_t* chars) noexcept {
// credit @aqrit // credit @aqrit
fastfloat_really_inline constexpr bool is_made_of_eight_digits_fast(uint64_t val) noexcept { fastfloat_really_inline constexpr bool is_made_of_eight_digits_fast(uint64_t val) noexcept {
return !((((val + 0x4646464646464646) | (val - 0x3030303030303030)) & return !((((val + 0x4646464646464646) | (val - 0x3030303030303030)) &
0x8080808080808080)); 0x8080808080808080));
} }
fastfloat_really_inline FASTFLOAT_CONSTEXPR20 fastfloat_really_inline FASTFLOAT_CONSTEXPR20
@ -199,50 +199,47 @@ bool parse_if_eight_digits_unrolled(const char32_t*, uint64_t&) noexcept {
return false; return false;
} }
template <typename UC>
struct parsed_number_string_t {
typedef span<const char> byte_span;
template <typename CharT>
struct parsed_number_string {
int64_t exponent{0}; int64_t exponent{0};
uint64_t mantissa{0}; uint64_t mantissa{0};
int64_t exp_number{0}; int64_t exp_number{0};
const CharT *lastmatch{nullptr}; UC const * lastmatch{nullptr};
bool negative{false}; bool negative{false};
bool valid{false}; bool valid{false};
bool too_many_digits{false}; bool too_many_digits{false};
// contains the range of the significant digits // contains the range of the significant digits
span<const CharT> integer{}; // non-nullable span<const UC> integer{}; // non-nullable
span<const CharT> fraction{}; // nullable span<const UC> fraction{}; // nullable
}; };
using byte_span = span<const char>;
using parsed_number_string = parsed_number_string_t<char>;
// Assuming that you use no more than 19 digits, this will // Assuming that you use no more than 19 digits, this will
// parse an ASCII string. // parse an ASCII string.
template <typename CharT> template <typename UC>
fastfloat_really_inline FASTFLOAT_CONSTEXPR20 fastfloat_really_inline FASTFLOAT_CONSTEXPR20
parsed_number_string<CharT> parse_number_string(const CharT *p, const CharT *pend, parse_options options) noexcept { parsed_number_string_t<UC> parse_number_string(UC const *p, UC const * pend, parse_options_t<UC> options) noexcept {
const chars_format fmt = options.format; chars_format const fmt = options.format;
const CharT decimal_point = options.decimal_point; UC const decimal_point = options.decimal_point;
parsed_number_string<CharT> answer; parsed_number_string_t<UC> answer;
answer.valid = false; answer.valid = false;
answer.too_many_digits = false; answer.too_many_digits = false;
answer.negative = (*p == CharT('-')); answer.negative = (*p == UC('-'));
#ifdef FASTFLOAT_ALLOWS_LEADING_PLUS // disabled by default #ifdef FASTFLOAT_ALLOWS_LEADING_PLUS // disabled by default
if ((*p == CharT('-')) || (*p == CharT('+'))) { if ((*p == UC('-')) || (*p == UC('+'))) {
#else #else
if (*p == CharT('-')) { // C++17 20.19.3.(7.1) explicitly forbids '+' sign here if (*p == UC('-')) { // C++17 20.19.3.(7.1) explicitly forbids '+' sign here
#endif #endif
++p; ++p;
if (p == pend) { if (p == pend) {
return answer; return answer;
} }
// a sign must be followed by an integer or the dot if (!is_integer(*p) && (*p != decimal_point)) { // a sign must be followed by an integer or the dot
if (!is_integer(*p) && *p != decimal_point) return answer;
return answer; }
} }
const CharT *const start_digits = p; UC const * const start_digits = p;
uint64_t i = 0; // an unsigned int avoids signed overflows (which are bad) uint64_t i = 0; // an unsigned int avoids signed overflows (which are bad)
@ -250,43 +247,42 @@ parsed_number_string<CharT> parse_number_string(const CharT *p, const CharT *pen
// a multiplication by 10 is cheaper than an arbitrary integer // a multiplication by 10 is cheaper than an arbitrary integer
// multiplication // multiplication
i = 10 * i + i = 10 * i +
uint64_t(*p - CharT('0')); // might overflow, we will handle the overflow later uint64_t(*p - UC('0')); // might overflow, we will handle the overflow later
++p; ++p;
} }
const CharT *const end_of_integer_part = p; UC const * const end_of_integer_part = p;
int64_t digit_count = int64_t(end_of_integer_part - start_digits); int64_t digit_count = int64_t(end_of_integer_part - start_digits);
answer.integer = span<const CharT>(start_digits, size_t(digit_count)); answer.integer = span<const UC>(start_digits, size_t(digit_count));
int64_t exponent = 0; int64_t exponent = 0;
const bool has_decimal_point = (p != pend) && (*p == decimal_point); if ((p != pend) && (*p == decimal_point)) {
if (has_decimal_point) {
++p; ++p;
const CharT* before = p; UC const * before = p;
// can occur at most twice without overflowing, but let it occur more, since // can occur at most twice without overflowing, but let it occur more, since
// for integers with many digits, digit parsing is the primary bottleneck. // for integers with many digits, digit parsing is the primary bottleneck.
while ((std::distance(p, pend) >= 8) && parse_if_eight_digits_unrolled(p, i)) { // in rare cases, this will overflow, but that's ok while ((std::distance(p, pend) >= 8) && parse_if_eight_digits_unrolled(p, i)) { // in rare cases, this will overflow, but that's ok
p += 8; p += 8;
} }
while ((p != pend) && is_integer(*p)) { while ((p != pend) && is_integer(*p)) {
i = i * 10 + uint64_t(*p - CharT('0')); // in rare cases, this will overflow, but that's ok uint8_t digit = uint8_t(*p - UC('0'));
++p; ++p;
} }
exponent = before - p; exponent = before - p;
answer.fraction = span<const CharT>(before, size_t(p - before)); answer.fraction = span<const UC>(before, size_t(p - before));
digit_count -= exponent; digit_count -= exponent;
} }
// we must have encountered at least one integer // we must have encountered at least one integer!
if (digit_count == 0) { if (digit_count == 0) {
return answer; return answer;
} }
int64_t exp_number = 0; // explicit exponential part int64_t exp_number = 0; // explicit exponential part
if ((fmt & chars_format::scientific) && (p != pend) && ((CharT('e') == *p) || (CharT('E') == *p))) { if ((fmt & chars_format::scientific) && (p != pend) && ((UC('e') == *p) || (UC('E') == *p))) {
const CharT * location_of_e = p; UC const * location_of_e = p;
++p; ++p;
bool neg_exp = false; bool neg_exp = false;
if ((p != pend) && (CharT('-') == *p)) { if ((p != pend) && (UC('-') == *p)) {
neg_exp = true; neg_exp = true;
++p; ++p;
} else if ((p != pend) && (CharT('+') == *p)) { // '+' on exponent is allowed by C++17 20.19.3.(7.1) } else if ((p != pend) && (UC('+') == *p)) { // '+' on exponent is allowed by C++17 20.19.3.(7.1)
++p; ++p;
} }
if ((p == pend) || !is_integer(*p)) { if ((p == pend) || !is_integer(*p)) {
@ -298,7 +294,7 @@ parsed_number_string<CharT> parse_number_string(const CharT *p, const CharT *pen
p = location_of_e; p = location_of_e;
} else { } else {
while ((p != pend) && is_integer(*p)) { while ((p != pend) && is_integer(*p)) {
uint8_t digit = uint8_t(*p - CharT('0')); uint8_t digit = uint8_t(*p - UC('0'));
if (exp_number < 0x10000000) { if (exp_number < 0x10000000) {
exp_number = 10 * exp_number + digit; exp_number = 10 * exp_number + digit;
} }
@ -325,13 +321,13 @@ parsed_number_string<CharT> parse_number_string(const CharT *p, const CharT *pen
// We have to handle the case where we have 0.0000somenumber. // We have to handle the case where we have 0.0000somenumber.
// We need to be mindful of the case where we only have zeroes... // We need to be mindful of the case where we only have zeroes...
// E.g., 0.000000000...000. // E.g., 0.000000000...000.
const CharT *start = start_digits; UC const * start = start_digits;
while ((start != pend) && (*start == CharT('0') || *start == decimal_point)) { while ((start != pend) && (*start == UC('0') || *start == decimal_point)) {
if(*start == CharT('0')) { digit_count --; } if(*start == UC('0')) { digit_count --; }
start++; start++;
} }
// exponent/mantissa must be truncated later // exponent/mantissa must be truncated later!
// this is unlikely, so don't inline truncation code with the rest of parse_number_string() // this is unlikely, so don't inline truncation code with the rest of parse_number_string()
answer.too_many_digits = digit_count > 19; answer.too_many_digits = digit_count > 19;
} }

83
include/fast_float/digit_comparison.h
View File

@ -23,9 +23,9 @@ constexpr static uint64_t powers_of_ten_uint64[] = {
// this algorithm is not even close to optimized, but it has no practical // this algorithm is not even close to optimized, but it has no practical
// effect on performance: in order to have a faster algorithm, we'd need // effect on performance: in order to have a faster algorithm, we'd need
// to slow down performance for faster algorithms, and this is still fast. // to slow down performance for faster algorithms, and this is still fast.
template <typename CharT> template <typename UC>
fastfloat_really_inline FASTFLOAT_CONSTEXPR14 fastfloat_really_inline FASTFLOAT_CONSTEXPR14
int32_t scientific_exponent(parsed_number_string<CharT>& num) noexcept { int32_t scientific_exponent(parsed_number_string_t<UC> & num) noexcept {
uint64_t mantissa = num.mantissa; uint64_t mantissa = num.mantissa;
int32_t exponent = int32_t(num.exponent); int32_t exponent = int32_t(num.exponent);
while (mantissa >= 10000) { while (mantissa >= 10000) {
@ -154,22 +154,19 @@ void round_down(adjusted_mantissa& am, int32_t shift) noexcept {
} }
am.power2 += shift; am.power2 += shift;
} }
template <typename UC>
template <typename CharT>
fastfloat_really_inline FASTFLOAT_CONSTEXPR20 fastfloat_really_inline FASTFLOAT_CONSTEXPR20
void skip_zeros(const CharT*& first, const CharT* last) noexcept { void skip_zeros(UC const * & first, UC const * last) noexcept {
if (std::is_same<CharT, char>::value) { uint64_t val;
uint64_t val; while (!cpp20_and_in_constexpr() && std::distance(first, last) >= int_cmp_len<UC>()) {
while (!cpp20_and_in_constexpr() && std::distance(first, last) >= 8) { ::memcpy(&val, first, sizeof(uint64_t));
::memcpy(&val, first, sizeof(uint64_t)); if (val != int_cmp_zeros<UC>()) {
if (val != 0x3030303030303030) { break;
break;
}
first += 8;
} }
first += int_cmp_len<UC>();
} }
while (first != last) { while (first != last) {
if (*first != CharT('0')) { if (*first != UC('0')) {
break; break;
} }
first++; first++;
@ -178,48 +175,46 @@ void skip_zeros(const CharT*& first, const CharT* last) noexcept {
// determine if any non-zero digits were truncated. // determine if any non-zero digits were truncated.
// all characters must be valid digits. // all characters must be valid digits.
template <typename CharT> template <typename UC>
fastfloat_really_inline FASTFLOAT_CONSTEXPR20 fastfloat_really_inline FASTFLOAT_CONSTEXPR20
bool is_truncated(const CharT* first, const CharT* last) noexcept { bool is_truncated(UC const * first, UC const * last) noexcept {
if (std::is_same<CharT, char>::value) { // do 8-bit optimizations, can just compare to 8 literal 0s.
// do 8-bit optimizations, can just compare to 8 literal 0s. uint64_t val;
uint64_t val; while (!cpp20_and_in_constexpr() && std::distance(first, last) >= int_cmp_len<UC>()) {
while (!cpp20_and_in_constexpr() && std::distance(first, last) >= 8) { ::memcpy(&val, first, sizeof(uint64_t));
::memcpy(&val, first, sizeof(uint64_t)); if (val != int_cmp_zeros<UC>()) {
if (val != 0x3030303030303030) {
return true;
}
first += 8;
}
}
while (first != last) {
if (*first != CharT('0')) {
return true; return true;
} }
first++; first += int_cmp_len<UC>();
}
while (first != last) {
if (*first != UC('0')) {
return true;
}
++first;
} }
return false; return false;
} }
template <typename UC>
template <typename CharT>
fastfloat_really_inline FASTFLOAT_CONSTEXPR20 fastfloat_really_inline FASTFLOAT_CONSTEXPR20
bool is_truncated(span<const CharT> s) noexcept { bool is_truncated(span<const UC> s) noexcept {
return is_truncated(s.ptr, s.ptr + s.len()); return is_truncated(s.ptr, s.ptr + s.len());
} }
template <typename CharT>
template <typename UC>
fastfloat_really_inline FASTFLOAT_CONSTEXPR20 fastfloat_really_inline FASTFLOAT_CONSTEXPR20
void parse_eight_digits(const CharT*& p, limb& value, size_t& counter, size_t& count) noexcept { void parse_eight_digits(const UC*& p, limb& value, size_t& counter, size_t& count) noexcept {
value = value * 100000000 + parse_eight_digits_unrolled(p); value = value * 100000000 + parse_eight_digits_unrolled(p);
p += 8; p += 8;
counter += 8; counter += 8;
count += 8; count += 8;
} }
template <typename CharT> template <typename UC>
fastfloat_really_inline FASTFLOAT_CONSTEXPR14 fastfloat_really_inline FASTFLOAT_CONSTEXPR14
void parse_one_digit(const CharT*& p, limb& value, size_t& counter, size_t& count) noexcept { void parse_one_digit(UC const *& p, limb& value, size_t& counter, size_t& count) noexcept {
value = value * 10 + limb(*p - CharT('0')); value = value * 10 + limb(*p - UC('0'));
p++; p++;
counter++; counter++;
count++; count++;
@ -240,9 +235,9 @@ void round_up_bigint(bigint& big, size_t& count) noexcept {
} }
// parse the significant digits into a big integer // parse the significant digits into a big integer
template <typename CharT> template <typename UC>
inline FASTFLOAT_CONSTEXPR20 inline FASTFLOAT_CONSTEXPR20
void parse_mantissa(bigint& result, parsed_number_string<CharT>& num, size_t max_digits, size_t& digits) noexcept { void parse_mantissa(bigint& result, parsed_number_string_t<UC>& num, size_t max_digits, size_t& digits) noexcept {
// try to minimize the number of big integer and scalar multiplication. // try to minimize the number of big integer and scalar multiplication.
// therefore, try to parse 8 digits at a time, and multiply by the largest // therefore, try to parse 8 digits at a time, and multiply by the largest
// scalar value (9 or 19 digits) for each step. // scalar value (9 or 19 digits) for each step.
@ -256,8 +251,8 @@ void parse_mantissa(bigint& result, parsed_number_string<CharT>& num, size_t max
#endif #endif
// process all integer digits. // process all integer digits.
const CharT* p = num.integer.ptr; UC const * p = num.integer.ptr;
const CharT* pend = p + num.integer.len(); UC const * pend = p + num.integer.len();
skip_zeros(p, pend); skip_zeros(p, pend);
// process all digits, in increments of step per loop // process all digits, in increments of step per loop
while (p != pend) { while (p != pend) {
@ -406,9 +401,9 @@ adjusted_mantissa negative_digit_comp(bigint& bigmant, adjusted_mantissa am, int
// `b` as a big-integer type, scaled to the same binary exponent as // `b` as a big-integer type, scaled to the same binary exponent as
// the actual digits. we then compare the big integer representations // the actual digits. we then compare the big integer representations
// of both, and use that to direct rounding. // of both, and use that to direct rounding.
template <typename T, typename CharT> template <typename T, typename UC>
inline FASTFLOAT_CONSTEXPR20 inline FASTFLOAT_CONSTEXPR20
adjusted_mantissa digit_comp(parsed_number_string<CharT>& num, adjusted_mantissa am) noexcept { adjusted_mantissa digit_comp(parsed_number_string_t<UC>& num, adjusted_mantissa am) noexcept {
// remove the invalid exponent bias // remove the invalid exponent bias
am.power2 -= invalid_am_bias; am.power2 -= invalid_am_bias;

31
include/fast_float/fast_float.h
View File

@ -13,22 +13,25 @@ enum chars_format {
general = fixed | scientific general = fixed | scientific
}; };
template <typename CharT> template <typename UC>
struct from_chars_result { struct from_chars_result_t {
const CharT *ptr; UC const * ptr;
std::errc ec; std::errc ec;
}; };
using from_chars_result = from_chars_result_t<char>;
struct parse_options { template <typename UC>
constexpr explicit parse_options( struct parse_options_t {
chars_format fmt = chars_format::general, char dot = '.') constexpr explicit parse_options_t(chars_format fmt = chars_format::general,
UC dot = UC('.'))
: format(fmt), decimal_point(dot) {} : format(fmt), decimal_point(dot) {}
/** Which number formats are accepted */ /** Which number formats are accepted */
chars_format format; chars_format format;
/** The character used as decimal point */ /** The character used as decimal point */
char decimal_point; UC decimal_point;
}; };
using parse_options = parse_options_t<char>;
/** /**
* This function parses the character sequence [first,last) for a number. It parses floating-point numbers expecting * This function parses the character sequence [first,last) for a number. It parses floating-point numbers expecting
@ -49,21 +52,19 @@ struct parse_options {
* to determine whether we allow the fixed point and scientific notation respectively. * to determine whether we allow the fixed point and scientific notation respectively.
* The default is `fast_float::chars_format::general` which allows both `fixed` and `scientific`. * The default is `fast_float::chars_format::general` which allows both `fixed` and `scientific`.
*/ */
template<typename T, typename CharT> template<typename T, typename UC = char>
FASTFLOAT_CONSTEXPR20 FASTFLOAT_CONSTEXPR20
from_chars_result<CharT> from_chars(const CharT *first, const CharT *last, from_chars_result_t<UC> from_chars(UC const * first, UC const * last,
T &value, chars_format fmt = chars_format::general) noexcept; T &value, chars_format fmt = chars_format::general) noexcept;
/** /**
* Like from_chars, but accepts an `options` argument to govern number parsing. * Like from_chars, but accepts an `options` argument to govern number parsing.
*/ */
template<typename T, typename CharT> template<typename T, typename UC = char>
FASTFLOAT_CONSTEXPR20 FASTFLOAT_CONSTEXPR20
from_chars_result<CharT> from_chars_advanced(const CharT *first, const CharT *last, from_chars_result_t<UC> from_chars_advanced(UC const * first, UC const * last,
T &value, parse_options options) noexcept; T &value, parse_options_t<UC> options) noexcept;
} } // namespace fast_float
// namespace fast_float
#include "parse_number.h" #include "parse_number.h"
#endif // FASTFLOAT_FAST_FLOAT_H #endif // FASTFLOAT_FAST_FLOAT_H

71
include/fast_float/float_common.h
View File

@ -136,13 +136,12 @@ fastfloat_really_inline constexpr bool has_simd() {
} }
// Compares two ASCII strings in a case insensitive manner. // Compares two ASCII strings in a case insensitive manner.
// maya: for now, keep input2 ASCII only template <typename UC>
template <typename CharT>
inline FASTFLOAT_CONSTEXPR14 bool inline FASTFLOAT_CONSTEXPR14 bool
fastfloat_strncasecmp(const CharT *input1, const char *input2, size_t length) { fastfloat_strncasecmp(UC const * input1, UC const * input2, size_t length) {
char running_diff{0}; char running_diff{0};
for (size_t i = 0; i < length; i++) { for (size_t i = 0; i < length; ++i) {
running_diff |= (char(input1[i]) ^ input2[i]); running_diff |= (char(input1[i]) ^ char(input2[i]));
} }
return (running_diff == 0) || (running_diff == 32); return (running_diff == 0) || (running_diff == 32);
} }
@ -535,6 +534,68 @@ constexpr bool space_lut<T>::value[];
inline constexpr bool is_space(uint8_t c) { return space_lut<>::value[c]; } inline constexpr bool is_space(uint8_t c) { return space_lut<>::value[c]; }
#endif #endif
template<typename UC>
static constexpr uint64_t int_cmp_zeros()
{
static_assert((sizeof(UC) == 1) || (sizeof(UC) == 2) || (sizeof(UC) == 4), "Unsupported character size");
return (sizeof(UC) == 1) ? 0x3030303030303030 : (sizeof(UC) == 2) ? (uint64_t(UC('0')) << 48 | uint64_t(UC('0')) << 32 | uint64_t(UC('0')) << 16 | UC('0')) : (uint64_t(UC('0')) << 32 | UC('0'));
}
template<typename UC>
static constexpr int int_cmp_len()
{
return sizeof(uint64_t) / sizeof(UC);
}
template<typename UC>
static constexpr UC const * str_const_nan()
{
return nullptr;
}
template<>
constexpr char const * str_const_nan<char>()
{
return "nan";
}
template<>
constexpr wchar_t const * str_const_nan<wchar_t>()
{
return L"nan";
}
template<>
constexpr char16_t const * str_const_nan<char16_t>()
{
return u"nan";
}
template<>
constexpr char32_t const * str_const_nan<char32_t>()
{
return U"nan";
}
template<typename UC>
static constexpr UC const * str_const_inf()
{
return nullptr;
}
template<>
constexpr char const * str_const_inf<char>()
{
return "infinity";
}
template<>
constexpr wchar_t const * str_const_inf<wchar_t>()
{
return L"infinity";
}
template<>
constexpr char16_t const * str_const_inf<char16_t>()
{
return u"infinity";
}
template<>
constexpr char32_t const * str_const_inf<char32_t>()
{
return U"infinity";
}
} // namespace fast_float } // namespace fast_float
#endif #endif

51
include/fast_float/parse_number.h
View File

@ -19,43 +19,41 @@ namespace detail {
* The case comparisons could be made much faster given that we know that the * The case comparisons could be made much faster given that we know that the
* strings a null-free and fixed. * strings a null-free and fixed.
**/ **/
template <typename T, typename CharT> template <typename T, typename UC>
from_chars_result<CharT> FASTFLOAT_CONSTEXPR14 from_chars_result_t<UC> FASTFLOAT_CONSTEXPR14
parse_infnan(const CharT *first, const CharT *last, T &value) noexcept { parse_infnan(UC const * first, UC const * last, T &value) noexcept {
from_chars_result<CharT> answer{}; from_chars_result_t<UC> answer{};
answer.ptr = first; answer.ptr = first;
answer.ec = std::errc(); // be optimistic answer.ec = std::errc(); // be optimistic
bool minusSign = false; bool minusSign = false;
if (*first == CharT('-')) { // assume first < last, so dereference without checks; C++17 20.19.3.(7.1) explicitly forbids '+' here if (*first == UC('-')) { // assume first < last, so dereference without checks; C++17 20.19.3.(7.1) explicitly forbids '+' here
minusSign = true; minusSign = true;
++first; ++first;
} }
#ifdef FASTFLOAT_ALLOWS_LEADING_PLUS // disabled by default #ifdef FASTFLOAT_ALLOWS_LEADING_PLUS // disabled by default
if (*first == CharT('+')) { if (*first == UC('+')) {
++first; ++first;
} }
#endif #endif
if (last - first >= 3) { if (last - first >= 3) {
if (fastfloat_strncasecmp(first, "nan", 3)) { if (fastfloat_strncasecmp(first, str_const_nan<UC>(), 3)) {
answer.ptr = (first += 3); answer.ptr = (first += 3);
value = minusSign ? -std::numeric_limits<T>::quiet_NaN() : std::numeric_limits<T>::quiet_NaN(); value = minusSign ? -std::numeric_limits<T>::quiet_NaN() : std::numeric_limits<T>::quiet_NaN();
// Check for possible nan(n-char-seq-opt), C++17 20.19.3.7, C11 7.20.1.3.3. At least MSVC produces nan(ind) and nan(snan). // Check for possible nan(n-char-seq-opt), C++17 20.19.3.7, C11 7.20.1.3.3. At least MSVC produces nan(ind) and nan(snan).
if(first != last && *first == CharT('(')) { if(first != last && *first == UC('(')) {
for(const CharT* ptr = first + 1; ptr != last; ++ptr) { for(UC const * ptr = first + 1; ptr != last; ++ptr) {
if (*ptr == CharT(')')) { if (*ptr == UC(')')) {
answer.ptr = ptr + 1; // valid nan(n-char-seq-opt) answer.ptr = ptr + 1; // valid nan(n-char-seq-opt)
break; break;
} }
else if(!((CharT('a') <= *ptr && *ptr <= CharT('z')) || else if(!((UC('a') <= *ptr && *ptr <= UC('z')) || (UC('A') <= *ptr && *ptr <= UC('Z')) || (UC('0') <= *ptr && *ptr <= UC('9')) || *ptr == UC('_')))
(CharT('A') <= *ptr && *ptr <= CharT('Z')) ||
(CharT('0') <= *ptr && *ptr <= CharT('9')) || *ptr == CharT('_')))
break; // forbidden char, not nan(n-char-seq-opt) break; // forbidden char, not nan(n-char-seq-opt)
} }
} }
return answer; return answer;
} }
if (fastfloat_strncasecmp(first, "inf", 3)) { if (fastfloat_strncasecmp(first, str_const_inf<UC>(), 3)) {
if ((last - first >= 8) && fastfloat_strncasecmp(first + 3, "inity", 5)) { if ((last - first >= 8) && fastfloat_strncasecmp(first + 3, str_const_inf<UC>() + 3, 5)) {
answer.ptr = first + 8; answer.ptr = first + 8;
} else { } else {
answer.ptr = first + 3; answer.ptr = first + 3;
@ -134,22 +132,25 @@ fastfloat_really_inline bool rounds_to_nearest() noexcept {
} // namespace detail } // namespace detail
template<typename T, typename CharT> template<typename T, typename UC>
FASTFLOAT_CONSTEXPR20 FASTFLOAT_CONSTEXPR20
from_chars_result<CharT> from_chars(const CharT *first, const CharT *last, from_chars_result_t<UC> from_chars(UC const * first, UC const * last,
T &value, chars_format fmt /*= chars_format::general*/) noexcept { T &value, chars_format fmt /*= chars_format::general*/) noexcept {
return from_chars_advanced(first, last, value, parse_options{fmt}); return from_chars_advanced(first, last, value, parse_options_t<UC>{fmt});
} }
template<typename T, typename CharT> template<typename T, typename UC>
FASTFLOAT_CONSTEXPR20 FASTFLOAT_CONSTEXPR20
from_chars_result<CharT> from_chars_advanced(const CharT *first, const CharT *last, from_chars_result_t<UC> from_chars_advanced(UC const * first, UC const * last,
T &value, parse_options options) noexcept { T &value, parse_options_t<UC> options) noexcept {
{
static_assert (std::is_same<T, double>::value || std::is_same<T, float>::value, "only float and double are supported"); static_assert (std::is_same<T, double>::value || std::is_same<T, float>::value, "only float and double are supported");
static_assert (std::is_same<UC, char>::value ||
std::is_same<UC, wchar_t>::value ||
std::is_same<UC, char16_t>::value ||
std::is_same<UC, char32_t>::value , "only char, wchar_t, char16_t and char32_t are supported");
from_chars_result_t<UC> answer;
from_chars_result<CharT> answer;
#ifdef FASTFLOAT_SKIP_WHITE_SPACE // disabled by default #ifdef FASTFLOAT_SKIP_WHITE_SPACE // disabled by default
while ((first != last) && fast_float::is_space(uint8_t(*first))) { while ((first != last) && fast_float::is_space(uint8_t(*first))) {
first++; first++;
@ -160,7 +161,7 @@ from_chars_result<CharT> from_chars_advanced(const CharT *first, const CharT *la
answer.ptr = first; answer.ptr = first;
return answer; return answer;
} }
parsed_number_string<CharT> pns = parse_number_string(first, last, options); parsed_number_string_t<UC> pns = parse_number_string<UC>(first, last, options);
if (!pns.valid) { if (!pns.valid) {
return detail::parse_infnan(first, last, value); return detail::parse_infnan(first, last, value);
} }

28
tests/basictest.cpp
View File

@ -644,9 +644,9 @@ enum class Diag { runtime, comptime };
} // anonymous namespace } // anonymous namespace
template <Diag diag, class T> template <Diag diag, class T, typename result_type, typename stringtype>
constexpr void check_basic_test_result(std::string_view str, constexpr void check_basic_test_result(stringtype str,
fast_float::from_chars_result result, result_type result,
T actual, T expected, std::errc expected_ec) { T actual, T expected, std::errc expected_ec) {
if constexpr (diag == Diag::runtime) { if constexpr (diag == Diag::runtime) {
INFO( INFO(
@ -702,11 +702,33 @@ constexpr void check_basic_test_result(std::string_view str,
#undef FASTFLOAT_CHECK_EQ #undef FASTFLOAT_CHECK_EQ
} }
template<Diag diag, class T> template<Diag diag, class T>
constexpr void basic_test(std::string_view str, T expected, std::errc expected_ec = std::errc()) { constexpr void basic_test(std::string_view str, T expected, std::errc expected_ec = std::errc()) {
T actual; T actual;
auto result = fast_float::from_chars(str.data(), str.data() + str.size(), actual); auto result = fast_float::from_chars(str.data(), str.data() + str.size(), actual);
check_basic_test_result<diag>(str, result, actual, expected, expected_ec); check_basic_test_result<diag>(str, result, actual, expected, expected_ec);
constexpr size_t global_string_capacity = 2048;
if(str.size() > global_string_capacity) {
return;
}
// We give plenty of memory: 2048 characters.
char16_t u16[global_string_capacity]{};
for (size_t i = 0; i < str.size(); i++) {
u16[i] = char16_t(str[i]);
}
auto result16 = fast_float::from_chars(u16, u16 + str.size(), actual);
check_basic_test_result<diag>(std::u16string_view(u16, str.size()), result16, actual, expected, expected_ec);
char32_t u32[global_string_capacity]{};
for (size_t i = 0; i < str.size(); i++) {
u32[i] = char32_t(str[i]);
}
auto result32 = fast_float::from_chars(u32, u32 + str.size(), actual);
check_basic_test_result<diag>(std::u32string_view(u32, str.size()), result32, actual, expected, expected_ec);
} }
template<Diag diag, class T> template<Diag diag, class T>