mirror of
https://github.com/fastfloat/fast_float.git
synced 2025-12-07 17:26:51 +08:00
Merge pull request #197 from lemire/other_chars
Support UTF-16 and UTF-32 inputs
This commit is contained in:
commit
fe571b1da7
@ -5,3 +5,4 @@ Neal Richardson
|
||||
Tim Paine
|
||||
Fabio Pellacini
|
||||
Lénárd Szolnoki
|
||||
Jan Pharago
|
||||
18
README.md
18
README.md
@ -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
|
||||
|
||||
|
||||
|
||||
@ -12,8 +12,9 @@ namespace fast_float {
|
||||
|
||||
// Next function can be micro-optimized, but compilers are entirely
|
||||
// able to optimize it well.
|
||||
fastfloat_really_inline constexpr bool is_integer(char c) noexcept {
|
||||
return c >= '0' && c <= '9';
|
||||
template <typename UC>
|
||||
fastfloat_really_inline constexpr bool is_integer(UC c) noexcept {
|
||||
return !(c > UC('9') || c < UC('0'));
|
||||
}
|
||||
|
||||
fastfloat_really_inline constexpr uint64_t byteswap(uint64_t val) {
|
||||
@ -75,6 +76,16 @@ uint32_t parse_eight_digits_unrolled(uint64_t val) {
|
||||
return uint32_t(val);
|
||||
}
|
||||
|
||||
fastfloat_really_inline constexpr
|
||||
uint32_t parse_eight_digits_unrolled(const char16_t *) noexcept {
|
||||
return 0;
|
||||
}
|
||||
|
||||
fastfloat_really_inline constexpr
|
||||
uint32_t parse_eight_digits_unrolled(const char32_t *) noexcept {
|
||||
return 0;
|
||||
}
|
||||
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
uint32_t parse_eight_digits_unrolled(const char *chars) noexcept {
|
||||
return parse_eight_digits_unrolled(read_u64(chars));
|
||||
@ -86,40 +97,51 @@ fastfloat_really_inline constexpr bool is_made_of_eight_digits_fast(uint64_t val
|
||||
0x8080808080808080));
|
||||
}
|
||||
|
||||
fastfloat_really_inline constexpr
|
||||
bool is_made_of_eight_digits_fast(const char16_t *) noexcept {
|
||||
return false;
|
||||
}
|
||||
|
||||
fastfloat_really_inline constexpr
|
||||
bool is_made_of_eight_digits_fast(const char32_t *) noexcept {
|
||||
return false;
|
||||
}
|
||||
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
bool is_made_of_eight_digits_fast(const char *chars) noexcept {
|
||||
return is_made_of_eight_digits_fast(read_u64(chars));
|
||||
}
|
||||
|
||||
typedef span<const char> byte_span;
|
||||
|
||||
struct parsed_number_string {
|
||||
template <typename UC>
|
||||
struct parsed_number_string_t {
|
||||
int64_t exponent{0};
|
||||
uint64_t mantissa{0};
|
||||
const char *lastmatch{nullptr};
|
||||
UC const * lastmatch{nullptr};
|
||||
bool negative{false};
|
||||
bool valid{false};
|
||||
bool too_many_digits{false};
|
||||
// contains the range of the significant digits
|
||||
byte_span integer{}; // non-nullable
|
||||
byte_span fraction{}; // nullable
|
||||
span<const UC> integer{}; // non-nullable
|
||||
span<const UC> fraction{}; // nullable
|
||||
};
|
||||
|
||||
using byte_span = span<char>;
|
||||
using parsed_number_string = parsed_number_string_t<char>;
|
||||
// Assuming that you use no more than 19 digits, this will
|
||||
// parse an ASCII string.
|
||||
template <typename UC>
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
parsed_number_string parse_number_string(const char *p, const char *pend, parse_options options) noexcept {
|
||||
const chars_format fmt = options.format;
|
||||
const char decimal_point = options.decimal_point;
|
||||
parsed_number_string_t<UC> parse_number_string(UC const *p, UC const * pend, parse_options_t<UC> options) noexcept {
|
||||
chars_format const fmt = options.format;
|
||||
UC const decimal_point = options.decimal_point;
|
||||
|
||||
parsed_number_string answer;
|
||||
parsed_number_string_t<UC> answer;
|
||||
answer.valid = false;
|
||||
answer.too_many_digits = false;
|
||||
answer.negative = (*p == '-');
|
||||
answer.negative = (*p == UC('-'));
|
||||
#ifdef FASTFLOAT_ALLOWS_LEADING_PLUS // disabled by default
|
||||
if ((*p == '-') || (*p == '+')) {
|
||||
if ((*p == UC('-')) || (*p == UC('+'))) {
|
||||
#else
|
||||
if (*p == '-') { // 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
|
||||
++p;
|
||||
if (p == pend) {
|
||||
@ -129,7 +151,7 @@ parsed_number_string parse_number_string(const char *p, const char *pend, parse_
|
||||
return answer;
|
||||
}
|
||||
}
|
||||
const char *const start_digits = p;
|
||||
UC const * const start_digits = p;
|
||||
|
||||
uint64_t i = 0; // an unsigned int avoids signed overflows (which are bad)
|
||||
|
||||
@ -137,29 +159,31 @@ parsed_number_string parse_number_string(const char *p, const char *pend, parse_
|
||||
// a multiplication by 10 is cheaper than an arbitrary integer
|
||||
// multiplication
|
||||
i = 10 * i +
|
||||
uint64_t(*p - '0'); // might overflow, we will handle the overflow later
|
||||
uint64_t(*p - UC('0')); // might overflow, we will handle the overflow later
|
||||
++p;
|
||||
}
|
||||
const char *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);
|
||||
answer.integer = byte_span(start_digits, size_t(digit_count));
|
||||
answer.integer = span<const UC>(start_digits, size_t(digit_count));
|
||||
int64_t exponent = 0;
|
||||
if ((p != pend) && (*p == decimal_point)) {
|
||||
++p;
|
||||
const char* before = p;
|
||||
UC const * before = p;
|
||||
// can occur at most twice without overflowing, but let it occur more, since
|
||||
// for integers with many digits, digit parsing is the primary bottleneck.
|
||||
while ((std::distance(p, pend) >= 8) && is_made_of_eight_digits_fast(p)) {
|
||||
i = i * 100000000 + parse_eight_digits_unrolled(p); // in rare cases, this will overflow, but that's ok
|
||||
p += 8;
|
||||
if (std::is_same<UC,char>::value) {
|
||||
while ((std::distance(p, pend) >= 8) && is_made_of_eight_digits_fast(p)) {
|
||||
i = i * 100000000 + parse_eight_digits_unrolled(p); // in rare cases, this will overflow, but that's ok
|
||||
p += 8;
|
||||
}
|
||||
}
|
||||
while ((p != pend) && is_integer(*p)) {
|
||||
uint8_t digit = uint8_t(*p - '0');
|
||||
uint8_t digit = uint8_t(*p - UC('0'));
|
||||
++p;
|
||||
i = i * 10 + digit; // in rare cases, this will overflow, but that's ok
|
||||
}
|
||||
exponent = before - p;
|
||||
answer.fraction = byte_span(before, size_t(p - before));
|
||||
answer.fraction = span<const UC>(before, size_t(p - before));
|
||||
digit_count -= exponent;
|
||||
}
|
||||
// we must have encountered at least one integer!
|
||||
@ -167,14 +191,14 @@ parsed_number_string parse_number_string(const char *p, const char *pend, parse_
|
||||
return answer;
|
||||
}
|
||||
int64_t exp_number = 0; // explicit exponential part
|
||||
if ((fmt & chars_format::scientific) && (p != pend) && (('e' == *p) || ('E' == *p))) {
|
||||
const char * location_of_e = p;
|
||||
if ((fmt & chars_format::scientific) && (p != pend) && ((UC('e') == *p) || (UC('E') == *p))) {
|
||||
UC const * location_of_e = p;
|
||||
++p;
|
||||
bool neg_exp = false;
|
||||
if ((p != pend) && ('-' == *p)) {
|
||||
if ((p != pend) && (UC('-') == *p)) {
|
||||
neg_exp = true;
|
||||
++p;
|
||||
} else if ((p != pend) && ('+' == *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;
|
||||
}
|
||||
if ((p == pend) || !is_integer(*p)) {
|
||||
@ -186,7 +210,7 @@ parsed_number_string parse_number_string(const char *p, const char *pend, parse_
|
||||
p = location_of_e;
|
||||
} else {
|
||||
while ((p != pend) && is_integer(*p)) {
|
||||
uint8_t digit = uint8_t(*p - '0');
|
||||
uint8_t digit = uint8_t(*p - UC('0'));
|
||||
if (exp_number < 0x10000000) {
|
||||
exp_number = 10 * exp_number + digit;
|
||||
}
|
||||
@ -212,9 +236,9 @@ parsed_number_string parse_number_string(const char *p, const char *pend, parse_
|
||||
// We have to handle the case where we have 0.0000somenumber.
|
||||
// We need to be mindful of the case where we only have zeroes...
|
||||
// E.g., 0.000000000...000.
|
||||
const char *start = start_digits;
|
||||
while ((start != pend) && (*start == '0' || *start == decimal_point)) {
|
||||
if(*start == '0') { digit_count --; }
|
||||
UC const * start = start_digits;
|
||||
while ((start != pend) && (*start == UC('0') || *start == decimal_point)) {
|
||||
if(*start == UC('0')) { digit_count --; }
|
||||
start++;
|
||||
}
|
||||
if (digit_count > 19) {
|
||||
@ -224,19 +248,19 @@ parsed_number_string parse_number_string(const char *p, const char *pend, parse_
|
||||
// pre-tokenized spans from above.
|
||||
i = 0;
|
||||
p = answer.integer.ptr;
|
||||
const char* int_end = p + answer.integer.len();
|
||||
UC const * int_end = p + answer.integer.len();
|
||||
const uint64_t minimal_nineteen_digit_integer{1000000000000000000};
|
||||
while((i < minimal_nineteen_digit_integer) && (p != int_end)) {
|
||||
i = i * 10 + uint64_t(*p - '0');
|
||||
i = i * 10 + uint64_t(*p - UC('0'));
|
||||
++p;
|
||||
}
|
||||
if (i >= minimal_nineteen_digit_integer) { // We have a big integers
|
||||
exponent = end_of_integer_part - p + exp_number;
|
||||
} else { // We have a value with a fractional component.
|
||||
p = answer.fraction.ptr;
|
||||
const char* frac_end = p + answer.fraction.len();
|
||||
UC const * frac_end = p + answer.fraction.len();
|
||||
while((i < minimal_nineteen_digit_integer) && (p != frac_end)) {
|
||||
i = i * 10 + uint64_t(*p - '0');
|
||||
i = i * 10 + uint64_t(*p - UC('0'));
|
||||
++p;
|
||||
}
|
||||
exponent = answer.fraction.ptr - p + exp_number;
|
||||
|
||||
@ -23,8 +23,9 @@ constexpr static uint64_t powers_of_ten_uint64[] = {
|
||||
// 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
|
||||
// to slow down performance for faster algorithms, and this is still fast.
|
||||
template <typename UC>
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR14
|
||||
int32_t scientific_exponent(parsed_number_string& num) noexcept {
|
||||
int32_t scientific_exponent(parsed_number_string_t<UC> & num) noexcept {
|
||||
uint64_t mantissa = num.mantissa;
|
||||
int32_t exponent = int32_t(num.exponent);
|
||||
while (mantissa >= 10000) {
|
||||
@ -153,19 +154,19 @@ void round_down(adjusted_mantissa& am, int32_t shift) noexcept {
|
||||
}
|
||||
am.power2 += shift;
|
||||
}
|
||||
|
||||
template <typename UC>
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
void skip_zeros(const char*& first, const char* last) noexcept {
|
||||
void skip_zeros(UC const * & first, UC const * last) noexcept {
|
||||
uint64_t val;
|
||||
while (!cpp20_and_in_constexpr() && std::distance(first, last) >= 8) {
|
||||
while (!cpp20_and_in_constexpr() && std::distance(first, last) >= int_cmp_len<UC>()) {
|
||||
::memcpy(&val, first, sizeof(uint64_t));
|
||||
if (val != 0x3030303030303030) {
|
||||
if (val != int_cmp_zeros<UC>()) {
|
||||
break;
|
||||
}
|
||||
first += 8;
|
||||
first += int_cmp_len<UC>();
|
||||
}
|
||||
while (first != last) {
|
||||
if (*first != '0') {
|
||||
if (*first != UC('0')) {
|
||||
break;
|
||||
}
|
||||
first++;
|
||||
@ -174,29 +175,40 @@ void skip_zeros(const char*& first, const char* last) noexcept {
|
||||
|
||||
// determine if any non-zero digits were truncated.
|
||||
// all characters must be valid digits.
|
||||
template <typename UC>
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
bool is_truncated(const char* first, const char* last) noexcept {
|
||||
bool is_truncated(UC const * first, UC const * last) noexcept {
|
||||
// do 8-bit optimizations, can just compare to 8 literal 0s.
|
||||
uint64_t val;
|
||||
while (!cpp20_and_in_constexpr() && std::distance(first, last) >= 8) {
|
||||
while (!cpp20_and_in_constexpr() && std::distance(first, last) >= int_cmp_len<UC>()) {
|
||||
::memcpy(&val, first, sizeof(uint64_t));
|
||||
if (val != 0x3030303030303030) {
|
||||
if (val != int_cmp_zeros<UC>()) {
|
||||
return true;
|
||||
}
|
||||
first += 8;
|
||||
first += int_cmp_len<UC>();
|
||||
}
|
||||
while (first != last) {
|
||||
if (*first != '0') {
|
||||
if (*first != UC('0')) {
|
||||
return true;
|
||||
}
|
||||
first++;
|
||||
++first;
|
||||
}
|
||||
return false;
|
||||
}
|
||||
template <typename UC>
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
bool is_truncated(span<const UC> s) noexcept {
|
||||
return is_truncated(s.ptr, s.ptr + s.len());
|
||||
}
|
||||
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
bool is_truncated(byte_span s) noexcept {
|
||||
return is_truncated(s.ptr, s.ptr + s.len());
|
||||
void parse_eight_digits(const char16_t*& , limb& , size_t& , size_t& ) noexcept {
|
||||
// currently unused
|
||||
}
|
||||
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
void parse_eight_digits(const char32_t*& , limb& , size_t& , size_t& ) noexcept {
|
||||
// currently unused
|
||||
}
|
||||
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR20
|
||||
@ -207,9 +219,10 @@ void parse_eight_digits(const char*& p, limb& value, size_t& counter, size_t& co
|
||||
count += 8;
|
||||
}
|
||||
|
||||
template <typename UC>
|
||||
fastfloat_really_inline FASTFLOAT_CONSTEXPR14
|
||||
void parse_one_digit(const char*& p, limb& value, size_t& counter, size_t& count) noexcept {
|
||||
value = value * 10 + limb(*p - '0');
|
||||
void parse_one_digit(UC const *& p, limb& value, size_t& counter, size_t& count) noexcept {
|
||||
value = value * 10 + limb(*p - UC('0'));
|
||||
p++;
|
||||
counter++;
|
||||
count++;
|
||||
@ -230,8 +243,9 @@ void round_up_bigint(bigint& big, size_t& count) noexcept {
|
||||
}
|
||||
|
||||
// parse the significant digits into a big integer
|
||||
template <typename UC>
|
||||
inline FASTFLOAT_CONSTEXPR20
|
||||
void parse_mantissa(bigint& result, parsed_number_string& 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.
|
||||
// therefore, try to parse 8 digits at a time, and multiply by the largest
|
||||
// scalar value (9 or 19 digits) for each step.
|
||||
@ -245,13 +259,15 @@ void parse_mantissa(bigint& result, parsed_number_string& num, size_t max_digits
|
||||
#endif
|
||||
|
||||
// process all integer digits.
|
||||
const char* p = num.integer.ptr;
|
||||
const char* pend = p + num.integer.len();
|
||||
UC const * p = num.integer.ptr;
|
||||
UC const * pend = p + num.integer.len();
|
||||
skip_zeros(p, pend);
|
||||
// process all digits, in increments of step per loop
|
||||
while (p != pend) {
|
||||
while ((std::distance(p, pend) >= 8) && (step - counter >= 8) && (max_digits - digits >= 8)) {
|
||||
parse_eight_digits(p, value, counter, digits);
|
||||
if (std::is_same<UC,char>::value) {
|
||||
while ((std::distance(p, pend) >= 8) && (step - counter >= 8) && (max_digits - digits >= 8)) {
|
||||
parse_eight_digits(p, value, counter, digits);
|
||||
}
|
||||
}
|
||||
while (counter < step && p != pend && digits < max_digits) {
|
||||
parse_one_digit(p, value, counter, digits);
|
||||
@ -283,8 +299,10 @@ void parse_mantissa(bigint& result, parsed_number_string& num, size_t max_digits
|
||||
}
|
||||
// process all digits, in increments of step per loop
|
||||
while (p != pend) {
|
||||
while ((std::distance(p, pend) >= 8) && (step - counter >= 8) && (max_digits - digits >= 8)) {
|
||||
parse_eight_digits(p, value, counter, digits);
|
||||
if (std::is_same<UC,char>::value) {
|
||||
while ((std::distance(p, pend) >= 8) && (step - counter >= 8) && (max_digits - digits >= 8)) {
|
||||
parse_eight_digits(p, value, counter, digits);
|
||||
}
|
||||
}
|
||||
while (counter < step && p != pend && digits < max_digits) {
|
||||
parse_one_digit(p, value, counter, digits);
|
||||
@ -395,9 +413,9 @@ adjusted_mantissa negative_digit_comp(bigint& bigmant, adjusted_mantissa am, int
|
||||
// `b` as a big-integer type, scaled to the same binary exponent as
|
||||
// the actual digits. we then compare the big integer representations
|
||||
// of both, and use that to direct rounding.
|
||||
template <typename T>
|
||||
template <typename T, typename UC>
|
||||
inline FASTFLOAT_CONSTEXPR20
|
||||
adjusted_mantissa digit_comp(parsed_number_string& num, adjusted_mantissa am) noexcept {
|
||||
adjusted_mantissa digit_comp(parsed_number_string_t<UC>& num, adjusted_mantissa am) noexcept {
|
||||
// remove the invalid exponent bias
|
||||
am.power2 -= invalid_am_bias;
|
||||
|
||||
|
||||
@ -13,22 +13,25 @@ enum chars_format {
|
||||
general = fixed | scientific
|
||||
};
|
||||
|
||||
|
||||
struct from_chars_result {
|
||||
const char *ptr;
|
||||
template <typename UC>
|
||||
struct from_chars_result_t {
|
||||
UC const * ptr;
|
||||
std::errc ec;
|
||||
};
|
||||
using from_chars_result = from_chars_result_t<char>;
|
||||
|
||||
struct parse_options {
|
||||
constexpr explicit parse_options(chars_format fmt = chars_format::general,
|
||||
char dot = '.')
|
||||
template <typename UC>
|
||||
struct parse_options_t {
|
||||
constexpr explicit parse_options_t(chars_format fmt = chars_format::general,
|
||||
UC dot = UC('.'))
|
||||
: format(fmt), decimal_point(dot) {}
|
||||
|
||||
/** Which number formats are accepted */
|
||||
chars_format format;
|
||||
/** 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
|
||||
@ -49,18 +52,18 @@ struct parse_options {
|
||||
* 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`.
|
||||
*/
|
||||
template<typename T>
|
||||
template<typename T, typename UC = char>
|
||||
FASTFLOAT_CONSTEXPR20
|
||||
from_chars_result from_chars(const char *first, const char *last,
|
||||
from_chars_result_t<UC> from_chars(UC const * first, UC const * last,
|
||||
T &value, chars_format fmt = chars_format::general) noexcept;
|
||||
|
||||
/**
|
||||
* Like from_chars, but accepts an `options` argument to govern number parsing.
|
||||
*/
|
||||
template<typename T>
|
||||
template<typename T, typename UC = char>
|
||||
FASTFLOAT_CONSTEXPR20
|
||||
from_chars_result from_chars_advanced(const char *first, const char *last,
|
||||
T &value, parse_options options) noexcept;
|
||||
from_chars_result_t<UC> from_chars_advanced(UC const * first, UC const * last,
|
||||
T &value, parse_options_t<UC> options) noexcept;
|
||||
|
||||
} // namespace fast_float
|
||||
#include "parse_number.h"
|
||||
|
||||
@ -106,11 +106,12 @@ fastfloat_really_inline constexpr bool cpp20_and_in_constexpr() {
|
||||
}
|
||||
|
||||
// Compares two ASCII strings in a case insensitive manner.
|
||||
template <typename UC>
|
||||
inline FASTFLOAT_CONSTEXPR14 bool
|
||||
fastfloat_strncasecmp(const char *input1, const char *input2, size_t length) {
|
||||
fastfloat_strncasecmp(UC const * input1, UC const * input2, size_t length) {
|
||||
char running_diff{0};
|
||||
for (size_t i = 0; i < length; i++) {
|
||||
running_diff |= (input1[i] ^ input2[i]);
|
||||
for (size_t i = 0; i < length; ++i) {
|
||||
running_diff |= (char(input1[i]) ^ char(input2[i]));
|
||||
}
|
||||
return (running_diff == 0) || (running_diff == 32);
|
||||
}
|
||||
@ -503,6 +504,68 @@ constexpr bool space_lut<T>::value[];
|
||||
|
||||
inline constexpr bool is_space(uint8_t c) { return space_lut<>::value[c]; }
|
||||
#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
|
||||
|
||||
#endif
|
||||
|
||||
@ -19,41 +19,41 @@ namespace detail {
|
||||
* The case comparisons could be made much faster given that we know that the
|
||||
* strings a null-free and fixed.
|
||||
**/
|
||||
template <typename T>
|
||||
from_chars_result FASTFLOAT_CONSTEXPR14
|
||||
parse_infnan(const char *first, const char *last, T &value) noexcept {
|
||||
from_chars_result answer{};
|
||||
template <typename T, typename UC>
|
||||
from_chars_result_t<UC> FASTFLOAT_CONSTEXPR14
|
||||
parse_infnan(UC const * first, UC const * last, T &value) noexcept {
|
||||
from_chars_result_t<UC> answer{};
|
||||
answer.ptr = first;
|
||||
answer.ec = std::errc(); // be optimistic
|
||||
bool minusSign = false;
|
||||
if (*first == '-') { // 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;
|
||||
++first;
|
||||
}
|
||||
#ifdef FASTFLOAT_ALLOWS_LEADING_PLUS // disabled by default
|
||||
if (*first == '+') {
|
||||
if (*first == UC('+')) {
|
||||
++first;
|
||||
}
|
||||
#endif
|
||||
if (last - first >= 3) {
|
||||
if (fastfloat_strncasecmp(first, "nan", 3)) {
|
||||
if (fastfloat_strncasecmp(first, str_const_nan<UC>(), 3)) {
|
||||
answer.ptr = (first += 3);
|
||||
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).
|
||||
if(first != last && *first == '(') {
|
||||
for(const char* ptr = first + 1; ptr != last; ++ptr) {
|
||||
if (*ptr == ')') {
|
||||
if(first != last && *first == UC('(')) {
|
||||
for(UC const * ptr = first + 1; ptr != last; ++ptr) {
|
||||
if (*ptr == UC(')')) {
|
||||
answer.ptr = ptr + 1; // valid nan(n-char-seq-opt)
|
||||
break;
|
||||
}
|
||||
else if(!(('a' <= *ptr && *ptr <= 'z') || ('A' <= *ptr && *ptr <= 'Z') || ('0' <= *ptr && *ptr <= '9') || *ptr == '_'))
|
||||
else if(!((UC('a') <= *ptr && *ptr <= UC('z')) || (UC('A') <= *ptr && *ptr <= UC('Z')) || (UC('0') <= *ptr && *ptr <= UC('9')) || *ptr == UC('_')))
|
||||
break; // forbidden char, not nan(n-char-seq-opt)
|
||||
}
|
||||
}
|
||||
return answer;
|
||||
}
|
||||
if (fastfloat_strncasecmp(first, "inf", 3)) {
|
||||
if ((last - first >= 8) && fastfloat_strncasecmp(first + 3, "inity", 5)) {
|
||||
if (fastfloat_strncasecmp(first, str_const_inf<UC>(), 3)) {
|
||||
if ((last - first >= 8) && fastfloat_strncasecmp(first + 3, str_const_inf<UC>() + 3, 5)) {
|
||||
answer.ptr = first + 8;
|
||||
} else {
|
||||
answer.ptr = first + 3;
|
||||
@ -132,22 +132,25 @@ fastfloat_really_inline bool rounds_to_nearest() noexcept {
|
||||
|
||||
} // namespace detail
|
||||
|
||||
template<typename T>
|
||||
template<typename T, typename UC>
|
||||
FASTFLOAT_CONSTEXPR20
|
||||
from_chars_result from_chars(const char *first, const char *last,
|
||||
from_chars_result_t<UC> from_chars(UC const * first, UC const * last,
|
||||
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>
|
||||
template<typename T, typename UC>
|
||||
FASTFLOAT_CONSTEXPR20
|
||||
from_chars_result from_chars_advanced(const char *first, const char *last,
|
||||
T &value, parse_options options) noexcept {
|
||||
from_chars_result_t<UC> from_chars_advanced(UC const * first, UC const * last,
|
||||
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<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 answer;
|
||||
from_chars_result_t<UC> answer;
|
||||
#ifdef FASTFLOAT_SKIP_WHITE_SPACE // disabled by default
|
||||
while ((first != last) && fast_float::is_space(uint8_t(*first))) {
|
||||
first++;
|
||||
@ -158,7 +161,7 @@ from_chars_result from_chars_advanced(const char *first, const char *last,
|
||||
answer.ptr = first;
|
||||
return answer;
|
||||
}
|
||||
parsed_number_string pns = parse_number_string(first, last, options);
|
||||
parsed_number_string_t<UC> pns = parse_number_string<UC>(first, last, options);
|
||||
if (!pns.valid) {
|
||||
return detail::parse_infnan(first, last, value);
|
||||
}
|
||||
|
||||
@ -644,9 +644,9 @@ enum class Diag { runtime, comptime };
|
||||
|
||||
} // anonymous namespace
|
||||
|
||||
template <Diag diag, class T>
|
||||
constexpr void check_basic_test_result(std::string_view str,
|
||||
fast_float::from_chars_result result,
|
||||
template <Diag diag, class T, typename result_type, typename stringtype>
|
||||
constexpr void check_basic_test_result(stringtype str,
|
||||
result_type result,
|
||||
T actual, T expected, std::errc expected_ec) {
|
||||
if constexpr (diag == Diag::runtime) {
|
||||
INFO(
|
||||
@ -702,11 +702,33 @@ constexpr void check_basic_test_result(std::string_view str,
|
||||
#undef FASTFLOAT_CHECK_EQ
|
||||
}
|
||||
|
||||
|
||||
template<Diag diag, class T>
|
||||
constexpr void basic_test(std::string_view str, T expected, std::errc expected_ec = std::errc()) {
|
||||
T 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);
|
||||
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>
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user