Skip to content

Faster integer formatting #767

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Mar 25, 2025
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
59 changes: 45 additions & 14 deletions ext/json/ext/fbuffer/fbuffer.h
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@

#include "ruby.h"
#include "ruby/encoding.h"
#include "../vendor/jeaiii-ltoa.h"

/* shims */
/* This is the fallback definition from Ruby 3.4 */
Expand Down Expand Up @@ -150,6 +151,13 @@ static void fbuffer_append(FBuffer *fb, const char *newstr, unsigned long len)
}
}

/* Appends a character into a buffer. The buffer needs to have sufficient capacity, via fbuffer_inc_capa(...). */
static inline void fbuffer_append_reserved_char(FBuffer *fb, char chr)
{
fb->ptr[fb->len] = chr;
fb->len += 1;
}

static void fbuffer_append_str(FBuffer *fb, VALUE str)
{
const char *newstr = StringValuePtr(str);
Expand All @@ -167,25 +175,48 @@ static inline void fbuffer_append_char(FBuffer *fb, char newchr)
fb->len++;
}

static long fltoa(long number, char *buf)
static inline char *fbuffer_cursor(FBuffer *fb)
{
return fb->ptr + fb->len;
}

static inline void fbuffer_advance_to(FBuffer *fb, char *end)
{
static const char digits[] = "0123456789";
long sign = number;
char* tmp = buf;

if (sign < 0) number = -number;
do *tmp-- = digits[number % 10]; while (number /= 10);
if (sign < 0) *tmp-- = '-';
return buf - tmp;
fb->len = end - fb->ptr;
}

#define LONG_BUFFER_SIZE 20
/*
* Appends the decimal string representation of \a number into the buffer.
*/
static void fbuffer_append_long(FBuffer *fb, long number)
{
char buf[LONG_BUFFER_SIZE];
char *buffer_end = buf + LONG_BUFFER_SIZE;
long len = fltoa(number, buffer_end - 1);
fbuffer_append(fb, buffer_end - len, len);
/*
* The jeaiii_ultoa() function produces digits left-to-right,
* allowing us to write directly into the buffer, but we don't know
* the number of resulting characters.
*
* We do know, however, that the `number` argument is always in the
* range 0xc000000000000000 to 0x3fffffffffffffff, or, in decimal,
* -4611686018427387904 to 4611686018427387903. The max number of chars
* generated is therefore 20 (including a potential sign character).
*/

static const int MAX_CHARS_FOR_LONG = 20;

fbuffer_inc_capa(fb, MAX_CHARS_FOR_LONG);

if (number < 0) {
fbuffer_append_reserved_char(fb, '-');

/*
* Since number is always > LONG_MIN, `-number` will not overflow
* and is always the positive abs() value.
*/
number = -number;
}

char *end = jeaiii_ultoa(fbuffer_cursor(fb), number);
fbuffer_advance_to(fb, end);
}

static VALUE fbuffer_finalize(FBuffer *fb)
Expand Down
1 change: 1 addition & 0 deletions ext/json/ext/generator/depend
Original file line number Diff line number Diff line change
@@ -1,2 +1,3 @@
generator.o: generator.c $(srcdir)/../fbuffer/fbuffer.h
generator.o: generator.c $(srcdir)/../vendor/fpconv.c
generator.o: generator.c $(srcdir)/../vendor/jeaiii-ltoa.h
277 changes: 277 additions & 0 deletions ext/json/ext/vendor/jeaiii-ltoa.h
Original file line number Diff line number Diff line change
@@ -0,0 +1,277 @@
/*

This file is released under the terms of the MIT License. It is based on the
work of James Edward Anhalt III, with the original license listed below.

MIT License

Copyright (c) 2024,2025 Enrico Thierbach - https://github.com/radiospiel
Copyright (c) 2022 James Edward Anhalt III - https://github.com/jeaiii/itoa

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#ifndef JEAIII_TO_TEXT_H_
#define JEAIII_TO_TEXT_H_

#include <stdint.h>

typedef uint_fast32_t u32_t;
typedef uint_fast64_t u64_t;

#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wshorten-64-to-32"

#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wmissing-braces"

#define u32(x) ((u32_t)(x))
#define u64(x) ((u64_t)(x))

struct digit_pair
{
char dd[2];
};

static const struct digit_pair *digits_dd = (struct digit_pair *)(
"00" "01" "02" "03" "04" "05" "06" "07" "08" "09"
"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"
);

static const struct digit_pair *digits_fd = (struct digit_pair *)(
"0_" "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"
);

static const u64_t mask24 = (u64(1) << 24) - 1;
static const u64_t mask32 = (u64(1) << 32) - 1;
static const u64_t mask57 = (u64(1) << 57) - 1;

#define COPY(buffer, digits) memcpy(buffer, &(digits), sizeof(struct digit_pair))

static char *
jeaiii_ultoa(char *b, u64_t n)
{
if (n < u32(1e2)) {
COPY(b, digits_fd[n]);
return n < 10 ? b + 1 : b + 2;
}

if (n < u32(1e6)) {
if (n < u32(1e4)) {
u32_t f0 = u32(10 * (1 << 24) / 1e3 + 1) * n;
COPY(b, digits_fd[f0 >> 24]);

b -= n < u32(1e3);
u32_t f2 = (f0 & mask24) * 100;
COPY(b + 2, digits_dd[f2 >> 24]);

return b + 4;
}

u64_t f0 = u64(10 * (1ull << 32ull)/ 1e5 + 1) * n;
COPY(b, digits_fd[f0 >> 32]);

b -= n < u32(1e5);
u64_t f2 = (f0 & mask32) * 100;
COPY(b + 2, digits_dd[f2 >> 32]);

u64_t f4 = (f2 & mask32) * 100;
COPY(b + 4, digits_dd[f4 >> 32]);
return b + 6;
}

if (n < u64(1ull << 32ull)) {
if (n < u32(1e8)) {
u64_t f0 = u64(10 * (1ull << 48ull) / 1e7 + 1) * n >> 16;
COPY(b, digits_fd[f0 >> 32]);

b -= n < u32(1e7);
u64_t f2 = (f0 & mask32) * 100;
COPY(b + 2, digits_dd[f2 >> 32]);

u64_t f4 = (f2 & mask32) * 100;
COPY(b + 4, digits_dd[f4 >> 32]);

u64_t f6 = (f4 & mask32) * 100;
COPY(b + 6, digits_dd[f6 >> 32]);

return b + 8;
}

u64_t f0 = u64(10 * (1ull << 57ull) / 1e9 + 1) * n;
COPY(b, digits_fd[f0 >> 57]);

b -= n < u32(1e9);
u64_t f2 = (f0 & mask57) * 100;
COPY(b + 2, digits_dd[f2 >> 57]);

u64_t f4 = (f2 & mask57) * 100;
COPY(b + 4, digits_dd[f4 >> 57]);

u64_t f6 = (f4 & mask57) * 100;
COPY(b + 6, digits_dd[f6 >> 57]);

u64_t f8 = (f6 & mask57) * 100;
COPY(b + 8, digits_dd[f8 >> 57]);

return b + 10;
}

// if we get here U must be u64 but some compilers don't know that, so reassign n to a u64 to avoid warnings
u32_t z = n % u32(1e8);
u64_t u = n / u32(1e8);

if (u < u32(1e2)) {
// u can't be 1 digit (if u < 10 it would have been handled above as a 9 digit 32bit number)
COPY(b, digits_dd[u]);
b += 2;
}
else if (u < u32(1e6)) {
if (u < u32(1e4)) {
u32_t f0 = u32(10 * (1 << 24) / 1e3 + 1) * u;
COPY(b, digits_fd[f0 >> 24]);

b -= u < u32(1e3);
u32_t f2 = (f0 & mask24) * 100;
COPY(b + 2, digits_dd[f2 >> 24]);
b += 4;
}
else {
u64_t f0 = u64(10 * (1ull << 32ull) / 1e5 + 1) * u;
COPY(b, digits_fd[f0 >> 32]);

b -= u < u32(1e5);
u64_t f2 = (f0 & mask32) * 100;
COPY(b + 2, digits_dd[f2 >> 32]);

u64_t f4 = (f2 & mask32) * 100;
COPY(b + 4, digits_dd[f4 >> 32]);
b += 6;
}
}
else if (u < u32(1e8)) {
u64_t f0 = u64(10 * (1ull << 48ull) / 1e7 + 1) * u >> 16;
COPY(b, digits_fd[f0 >> 32]);

b -= u < u32(1e7);
u64_t f2 = (f0 & mask32) * 100;
COPY(b + 2, digits_dd[f2 >> 32]);

u64_t f4 = (f2 & mask32) * 100;
COPY(b + 4, digits_dd[f4 >> 32]);

u64_t f6 = (f4 & mask32) * 100;
COPY(b + 6, digits_dd[f6 >> 32]);

b += 8;
}
else if (u < u64(1ull << 32ull)) {
u64_t f0 = u64(10 * (1ull << 57ull) / 1e9 + 1) * u;
COPY(b, digits_fd[f0 >> 57]);

b -= u < u32(1e9);
u64_t f2 = (f0 & mask57) * 100;
COPY(b + 2, digits_dd[f2 >> 57]);

u64_t f4 = (f2 & mask57) * 100;
COPY(b + 4, digits_dd[f4 >> 57]);

u64_t f6 = (f4 & mask57) * 100;
COPY(b + 6, digits_dd[f6 >> 57]);

u64_t f8 = (f6 & mask57) * 100;
COPY(b + 8, digits_dd[f8 >> 57]);
b += 10;
}
else {
u32_t y = u % u32(1e8);
u /= u32(1e8);

// u is 2, 3, or 4 digits (if u < 10 it would have been handled above)
if (u < u32(1e2)) {
COPY(b, digits_dd[u]);
b += 2;
}
else {
u32_t f0 = u32(10 * (1 << 24) / 1e3 + 1) * u;
COPY(b, digits_fd[f0 >> 24]);

b -= u < u32(1e3);
u32_t f2 = (f0 & mask24) * 100;
COPY(b + 2, digits_dd[f2 >> 24]);

b += 4;
}
// do 8 digits
u64_t f0 = (u64((1ull << 48ull) / 1e6 + 1) * y >> 16) + 1;
COPY(b, digits_dd[f0 >> 32]);

u64_t f2 = (f0 & mask32) * 100;
COPY(b + 2, digits_dd[f2 >> 32]);

u64_t f4 = (f2 & mask32) * 100;
COPY(b + 4, digits_dd[f4 >> 32]);

u64_t f6 = (f4 & mask32) * 100;
COPY(b + 6, digits_dd[f6 >> 32]);
b += 8;
}

// do 8 digits
u64_t f0 = (u64((1ull << 48ull) / 1e6 + 1) * z >> 16) + 1;
COPY(b, digits_dd[f0 >> 32]);

u64_t f2 = (f0 & mask32) * 100;
COPY(b + 2, digits_dd[f2 >> 32]);

u64_t f4 = (f2 & mask32) * 100;
COPY(b + 4, digits_dd[f4 >> 32]);

u64_t f6 = (f4 & mask32) * 100;
COPY(b + 6, digits_dd[f6 >> 32]);

return b + 8;
}

#undef u32
#undef u64
#undef COPY

#pragma clang diagnostic pop
#pragma GCC diagnostic pop

#endif // JEAIII_TO_TEXT_H_

12 changes: 12 additions & 0 deletions test/json/json_generator_test.rb
Original file line number Diff line number Diff line change
Expand Up @@ -707,4 +707,16 @@ def test_json_generate_float
assert_equal expected, value.to_json
end
end

def test_numbers_of_various_sizes
numbers = [
0, 1, -1, 9, -9, 13, -13, 91, -91, 513, -513, 7513, -7513,
17591, -17591, -4611686018427387904, 4611686018427387903,
2**62, 2**63, 2**64, -(2**62), -(2**63), -(2**64)
]

numbers.each do |number|
assert_equal "[#{number}]", JSON.generate([number])
end
end
end